Transcript
Page 1: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

Radioembolization with Yttrium-90Microspheres: A State-of-the-ArtBrachytherapy Treatment for Primaryand Secondary Liver MalignanciesPart 3: Comprehensive Literature Review and Future Direction

Riad Salem, MD, MBA, and Kenneth G. Thurston, MA

Treatment options for primary and secondary liver tumors that cannot be resected or ablated are based on transarterialtechniques. Although the majority of these are based on bland and chemoembolization techniques, yttrium-90microspheres represent an alternate transarterial option. Although the amount of literature on 90Y does not rival thatof bland or chemoembolization, there nevertheless are ample data that support its use for primary and metastatic livertumors. A comprehensive review of the entire available literature dating from the early 1960s is presented, as is adiscussion of the possibilities for future research with use of radioembolization as a platform.

J Vasc Interv Radiol 2006; 17:1571–1594

Abbreviations: AFP � �-fetoprotein, CEA � carcinoembryonic antigen, ECOG � Eastern Cooperative Oncology Group, FDG � [18F]fluorodeoxyglucose, 5-FU� 5-fluorouracil, GEP � gastroenteropancreatic, HAC � hepatic artery chemotherapy, HCC � hepatocellular carcinoma, PET � positron emission tomography,RECIST � Response Evaluation Criteria in Solid Tumors, RF � radiofrequency, SIRT � selective internal radiation therapy, TACE � transarterial chemoemboli-zation

ALTHOUGH yttrium-90 microspheretherapy has only recently (within thepast 5 years) gained increasing aware-ness and clinical use, investigationsinto 90Y and other radioisotopes forthe treatment of cancer date back tothe 1960s (1,2). Initial studies of resin90Y in humans were reported in thelate 1970s. The seminal work in a ca-nine liver model demonstrating thesafety and feasibility of 90Y therapy forhepatic malignancies was reported in

the late 1980s (3,4). Human studies of90Y microsphere therapy in liver appli-cations followed from the late 1980sthrough the 1990s (5–16). These inves-tigations established the safety of 90Yfor intrahepatic applications and theoptimal dosimetry for tumor radiationkill while minimizing exposure to nor-mal liver tissue. The assessment of po-tential pulmonary shunt, particularlyin patients with hepatocellular carci-noma (HCC), was reinforced in thesestudies. The importance of emboliza-tion of collateral vessels such as thegastroduodenal and right gastric arter-ies to prevent reflux to the gastricstructures was also realized. Gastriculceration requiring surgical interven-tion was routinely reported in many ofthese studies.

With improvements in technologypermitting smaller vessels to be cath-eterized, as well as refinements in im-aging techniques, the safety and effi-cacy of 90Y microsphere delivery hasimproved significantly. During the

past 5 years, numerous studies involv-ing larger cohorts, randomized trials,and 90Y microspheres in combinationwith other systemic and liver-directedtherapies have provided confirmatoryevidence of the safety and efficacy of90Y therapy for the treatment of pri-mary and metastatic (predominantlycolorectal) liver disease (8,13,14,17–27). New applications for 90Y ther-apy in selective lobar/segmental infu-sion with the intent of preservingfunctional liver reserve and downstag-ing disease to permit resection, radio-frequency (RF) ablation, and livertransplantation are also being ex-plored (5–10).

On the basis of encouraging pre-liminary results with 90Y therapy inmetastases other than those from colo-rectal cancer, such as breast and neu-roendocrine metastases, several direc-tions for future clinical applicationsare also warranted (11–20). 90Y ther-apy in combination with radiation-sensitizing agents and growth factor

From the Department of Radiology (R.S.), Robert H.Lurie Comprehensive Cancer Center, 676 North StClair, Suite 800, Chicago, Illinois 60611; and MDSNordion (K.G.T.), Kanata, Canada. Received June25, 2006; accepted June 26. Address correspondenceto R.S.; E-mail: [email protected]

R.S. is a consultant for MDS Nordion and has lec-tured for Sirtex Medical. K.G.T. is Director of Clin-ical Affairs for MDS Nordion. No research supportwas provided for this manuscript. Neither manufac-turer had any input in this document.

© SIR, 2006

DOI: 10.1097/01.RVI.0000236744.34720.73

Review Article

1571

Page 2: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

inhibitors present opportunities toevaluate its application in combinato-rial treatment paradigms. Other possi-ble studies include randomized trialsof 90Y versus transarterial chemoem-bolization (TACE), bland emboliza-tion, drug-eluting beads, other radio-active spheres, and best supportivecare. Finally, the potential applicationof 90Y therapy to organs other than theliver via an intraarterially placed cath-eter presents several areas for futureresearch.

This review concludes the three-part series on radioembolization andprovides a comprehensive review ofthe historical development of 90Y ther-apy, contemporary clinical results,and the direction for future research inclinical applications.

CONTEMPORARY CLINICALRESULTS

Early Clinical Work

As early as 1963, researchers wereinvestigating the utility of 90Y micro-spheres in canine prostates (2). In1967, Flynn (1) assessed the role of 90Ymicrospheres for the treatment of lungmalignancies. Ariel (21,22) and Ariel etal (23) described initial results of 90Ymicrosphere treatment with indica-tions, as well as the experience withintraarterial injection of microspheresfor the treatment of unresectable pan-creatic cancer. These agents have alsobeen used for splenic injection in apatient with lymphoma, cerebral infu-sion for brain tumors, extremity infu-sions for osteogenic sarcomas, and sy-novial injection for pain (24–31).Seminal work performed by investiga-tors established the proof of principleof radioactive intraarterial injection forthe treatment of liver tumors with useof various radioconjugates including90Y, 32P, rhenium, and holmium(29,32–43). Direct intratumoral injec-tion and portal venous injection of 90Yhave been studied, with both tech-niques demonstrating antitumoral ac-tivity (44–46). Magnetically guidedmicrospheres have also been investi-gated (47–49). Antitumor effects of 90Yare well-established in patients withlymphoma, as well as in combinationwith bone tracers (50–52). Nonradio-active glass microspheres have beenstudied in rabbit kidneys (53). Finally,splenic radioembolization with 90Y

has successfully been used for thetreatment of hypersplenism andthrombocytopenia (54).

In 1987, Wollner and colleagues (4)studied the effects of radiosensitizerscombined with 90Y resin micro-spheres. Five dogs were treated with50 Gy of 90Y alone, whereas anotherfive were treated with 50 Gy 90Y withbromodeoxyuridine. The approximateradiation dose to the liver was 50 Gy.Dogs receiving 90Y alone experiencedno changes in aminotransferase levels,whereas those that received bromode-oxyuridine and 90Y experienced tran-sient changes. At necropsy, the typeand degree of hepatic toxicity amongthe animals receiving radioactive mi-crospheres was comparable with thatpreviously described in patients re-ceiving external-beam hepatic irradia-tion at conventional doses (20–30 rad).Resin microspheres were found in thelungs of some canines, causing radia-tion-induced granulomas and leadingto the initiation of lung shunting cal-culation before the use of 90Y micro-spheres. The authors concluded thatbromodeoxyuridine could produce ac-ceptable, nonlethal, and tolerable tox-icities in this dog model, suggestingthat clinical studies of this combina-tion are not likely to be contraindi-cated by synergistic toxicity (4). Theunexpected fragmentation of the resinspheres without myelosuppression ledthe authors to initiate work with glassmicrospheres that cannot leach.

In 1988, Wollner et al (3) studied theeffects of 90Y glass microspheres in acanine model. Hepatic arterial injec-tion of radioactive glass microsphereswas found to produce portal changessimilar to those observed in humansafter external-beam therapy. Althoughthe extent of damage was proportionalto absorbed dose, radiation exposuresin excess of 300 Gy did not cause totalhepatic necrosis and were compatiblewith survival. No microspheres dis-tributed to the bone marrow, and nomyelosuppression was encountered.The authors concluded that hepatic ex-posures to humans of 50–100 Gy by90Y microsphere injection appear to befeasible and tolerable (3).

Comprehensive Literature Review90Y glass microspheres or Therasphere

in HCC.—In 1989, Houle et al (55)presented data on a pilot study of

seven patients with HCC. No toxici-ties were observed for absorbeddoses of 50–100 Gy to the liver andas high as 320 Gy to the tumor itself.Tumor response was seen only at thehigher absorbed doses. The authorsconcluded that 90Y glass micro-spheres can safely deliver large dosesof internal radiation to hepatic tu-mors as long as extrahepatic shunt-ing can be excluded, and that extra-hepatic shunting will be the mainlimitation to this form of radiationtherapy.

In 1992, Shepherd et al (56) con-ducted a phase I dose-escalation studyof 90Y microspheres in 10 patients withprimary HCC. The inclusion criteriaincluded cytologically or histologi-cally confirmed HCC and measurablehepatic lesions, a Karnofsky perfor-mance status of 60% or greater, normalbone marrow function, and adequatepulmonary status. Exclusionary crite-ria included compromised liver func-tion, history of significant peripheralvascular disease, previous thrombo-embolism, bleeding diathesis, or al-lergy to contrast agents. Treatmentwas administered in a nuclear medi-cine laboratory through a previouslyplaced hepatic artery catheter. Brems-strahlung scans were obtained afterdosing to assess distribution. Beforeinjection of 90Y, the presence of extra-hepatic shunting was assessed withuse of 99mTc macroaggregated albu-min scanning. Scintigraphy was thenperformed, and 90Y was not adminis-tered if there was significant shuntingto the lungs, stomach, or bowel. Fourpatients were treated with 50 Gy, twopatients received a 75-Gy dose, andthree patients received a 100-Gy dose.Survival was not an endpoint but wasreported in this series to range from 16days to 1,050 days (median survival,126 d). The patients who survived thelongest had the greatest tumor-to-liverperfusion ratio and therefore thegreatest estimated dose delivered tothe tumor. None of the patients expe-rienced myelosuppression. One pa-tient experienced a duodenal ulcer 2weeks after treatment, which ulti-mately required surgery. This seminalstudy provided the initial safety datato enable outcome studies with 90Y inHCC and provided critical patient se-lection and technique refinement. Forexample, the study defined areas forexcluding patients at risk for extrahe-

1572 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 3: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

patic shunting and suggested the im-portance of the assessment of peritu-moral hepatic vasculature (56).

Direct intratumoral injection wasperformed by Dong et al (57) in a 1992study in which 28 patients with HCCwere treated with a percutaneous in-jection of 90Y glass microspheres un-der ultrasound (US) guidance. In somepatients, peritumoral ethanol and in-traportal chemotherapy was alsogiven. There was a 91% reduction intumor size for the cohort. On fol-low-up Doppler imaging, there was areduction and/or elimination in bloodflow within the tumors. Eleven of 13patients had a response as measuredby �-fetoprotein (AFP) levels; six pa-tients’ condition completely normal-ized. Histopathologically, eight pa-tients exhibited complete necrosis. Theauthors concluded that percutaneous90Y is an appropriate first-line treat-ment for patients with HCC (57).

Tian et al (46) performed anotherdirect intratumoral injection of 90Yglass microspheres in which 27 pa-tients with HCC and six patients withliver metastases were treated with US-guided glass microsphere injection.The procedure was repeated at 3- to4-week intervals as necessary. Fol-low-up consisted of US and clinicaland laboratory assessments. Ninety-one percent of tumors decreased insize, with concomitant blood flow andsonographic changes observed. AFP in10 of 13 patients returned to normallevels. The authors concluded that in-tratumoral administration of 90Y glassmicrospheres under US guidanceyielded a promising cure rate for livertumors with no significant side effects.Radiation doses of 28,215–75,720 cGywere thought to explain the tumor re-sponse. The authors further postulatedthat these results demonstrated thatdirect, percutaneous, intratumoral ra-dionuclide injection is feasible fortreatment of malignant lesions (46).

Yan et al (58) published similar en-couraging data with glass micro-spheres in 1993. Six rabbits were in-jected with 185–1,480 MBq of 90Y glassmicrospheres. Three rabbits were in-jected with 35–300 mg of 90Y glass mi-crospheres for toxicity analysis. Patho-logic examinations were performed onall rabbits. No toxicities were noted inthe rabbits after injection of114.1–845.2 Gy 90Y glass micro-spheres. In the rabbit study, some ar-

eas of hepatocyte degeneration andportal fibrosis were seen pathologi-cally. The authors concluded that rab-bits could tolerate as much as eighttimes the upper limit of a clinical doseof 100 Gy. In a parallel clinical study,18 patients received 2,442–5,550 MBqof 35-�m 90Y glass microspheres forthe treatment of HCC. Whole bloodcounts, liver function tests, and imag-ing examination were performed. Themean tumor-to-liver tissue ratio was3:1. In some tumors, a 14:1 tumor-to-liver ratio was noted. Mean absorbeddoses in the normal liver parenchymaand tumor liver tissue were 30 Gy and88 Gy, respectively. Clinical follow-updemonstrated 50% reduction in tumormass in 13 of 18 patients (72%), as wellas a significant decrease in posttreat-ment AFP level. The authors con-cluded that 90Y glass microspheres aresafe and effective in large dosesthrough the hepatic artery for internalradiation treatment of HCC, and thatresponses were a result of localizedand hypervascular tumors (58).

Investigators reported on 22 patientsto determine response parameters, sur-vival, and toxicity after intraarterialinjection of 90Y microspheres (59). Theconditions of 20 patients were evalu-ated for efficacy, including nine pa-tients with Okuda stage I/II diseaseand 11 patients with Okuda stage IIIdisease. The median dose deliveredwas 104 Gy (range, 45–145 Gy). Therewere 31 serious adverse events; themost common were liver enzyme in-creases and gastrointestinal ulceration.Treatment efficacy was measured bytumor response, duration of response,time to progression, and overall sur-vival. One complete tumor responseand three partial responses were re-ported. The median time to progres-sion was 44 weeks, and the mediansurvival was 54 weeks (range, 7–180weeks). Total dose greater than 104Gy, Okuda stage I disease, and tumor-to-liver uptake ratio greater than 2were three factors associated with pro-longed survival.

Carr (60) performed a critical studyin 2004 supporting the safety and effi-cacy of TheraSphere for inoperableHCC. Sixty-five patients with biopsy-proven HCC received a median radi-ation dose of 134 Gy. Forty-two pa-tients had Okuda stage I disease, and23 had Okuda stage II disease. Toxic-ities included nine episodes of abdom-

inal pain (which did not fit criteria forpostembolization syndrome), two epi-sodes of cholecystitis, and transient in-creases in liver function in 25 patients.A finding previously unreported waslymphopenia in 75% of patients, noneof which was associated with adverseclinical events or opportunistic infec-tions. Median survival times were 649days and 302 days for patients withOkuda I and Okuda II disease, respec-tively, compared with 244 days and 64days, respectively, for historical con-trol individuals.

A study presenting data in patientstreated with TheraSphere in the pres-ence of portal vein thrombosis waspublished in 2004 (61). Fifteen patientswith unresectable HCC and portalvein thrombosis of one or both first-order and related segmental portal ve-nous branches received a total of 29infusions. Liver toxicity was assessedby serum total bilirubin level gradedfor severity according to the NationalInstitutes of Health/National CancerInstitute Clinical Toxicity Criteria (ver-sion 2.0). Other adverse events werereported according to the standardsestablished by the Society of Interven-tional Radiology. There were no pro-cedural complications with delivery ofmicrospheres, and treatment was welltolerated by all patients. Increasedposttreatment bilirubin levels wereobserved in all treatments in five pa-tients, four of whom had computedtomographic (CT) or AFP evidence ofintrahepatic disease progression. Afterinitial treatment, two patients experi-enced bilirubin toxicity (grades 1/2),one patient exhibited an increment inbilirubin toxicity grade (grades 1–3),and one patient had an improvementin grade after initial treatment. Therewere no new treatment-related toxici-ties in nine patients after a secondtreatment. The authors concluded thatTheraSphere treatment was well toler-ated and appears to be safe to use inpatients with compromised portal ve-nous flow in one or both first-orderand related segmental portal venousbranches. In patients who did not ex-hibit disease progression, there ap-peared to be no clinically significantchange in bilirubin level (61). The effi-cacy of TheraSphere in this contextwas subsequently presented in 2006(62), in which investigators introducedthe concept of portal vein retraction asa secondary sign of tumor response.

Salem and Thurston • 1573Volume 17 Number 10

Page 4: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

In 2004, Liu et al (63) presented aretrospective review of 14 patientstreated for unresectable HCC. Tumorresponse by CT/magnetic resonanceimaging and AFP level was monitoredafter treatment. One patient had com-plete response by CT; however, thispatient received TACE before Thera-Sphere treatment. Eight patients had apartial response, and two patients hadtumor progression. AFP-producingHCC was seen in eight of the 11 pa-tients treated. Five patients had de-creased AFP after treatment. None ofthe patients treated experienced anyserious adverse events, including livertoxicity.

Geschwind et al (64) reported on 80patients from a relatively large data-base of 121 patients who were treatedwith TheraSphere. Patients’ diseasewas staged with the Child-Pugh,Okuda, or Cancer of the Liver ItalianProgram scoring systems. Survivaltimes were found to be 628 days and324 days for Okuda I (68%) and II(32%) disease, respectively. These datawere instrumental for delineating theessential components to conduct alarge randomized controlled trial com-paring TheraSphere versus a standardTACE regimen with potential end-points of survival, tumor response,AFP reduction, and quality-of-life es-timates.

To evaluate the cumulative HCCclinical experience with TheraSphere,data from patients in the original hu-manitarian device exemption applica-tion were combined with data frompatients from the humanitarian deviceexemption clinical experience at threesites through February 2003. This pro-vided a cohort of 121 patients forwhom toxicity and survival data wereanalyzed to identify predictive factorsassociated with 3-month mortality rate(65). Patients were stratified into highand low risk categories on the basis offive liver reserve factors, including (i)tumor presentation (infiltrative vs fo-cal), (ii) bulk disease (tumor volume�70% vs �70%), (ii) increase in liverenzymes (aspartate or alanine amino-transferase level �5� upper limit ofnormal vs �5� upper limit of nor-mal), (iv) a combination of tumor vol-ume and albumin (tumor volume�50% and albumin level �3 g/dL ver-sus other), and (v) bilirubin increase(�2 mg/dL vs �2 mg/dL). Two non–liver reserve factors were also iden-

tified, including a diagnosis of cholan-giocarcinoma (one patient hadreceived a misdiagnosis of HCC when,in fact, the correct diagnosis was chol-angiocarcinoma) and a lung dosegreater than 30 Gy. When these sevenfactors were taken into account, 33 of121 patients were classified as highrisk (at least one high risk factor) ver-sus 88 patients classified as low risk(no high risk factor). As expected, thehigh-risk group had lower survivaland a higher proportion of patientswho experienced treatment-relatedadverse events ending in death. Six-teen of the 33 patients in the high-riskgroup (49%) did not survive the first 3months after treatment, comparedwith six of the 88 patients in the low-risk group (7%; Fisher exact test; P �.0001). Median survival times for thelow- and high-risk groups were 466days and 108 days, respectively (haz-ard ratio, 6.0; P � .0001). There were atotal of 25 serious adverse events end-ing in the death of the patient. Elevenof 12 patients experiencing a treat-ment-related serious adverse eventending in death were in the high-riskgroup. Moreover, if the high mortalityrisk criteria based on the seven liverreserve and non–liver reserve factorsare taken into account, 16 of the pa-tients (64%) would have been classi-fied as having high mortality risk be-fore treatment. Causes of death inthese 16 patients included liver failure(n � 8), hepatitis (n � 1), hepatorenalfailure (n � 1), gastric ulcer (n � 1),radiation pneumonitis (n � 1), not oth-erwise specified (n � 2), infection (n �1), and gastrointestinal bleeding (n �1). The results of this study suggestedthat the conditions of patients withHCC should be evaluated for the pres-ence of the aforementioned risk fac-tors, which include infiltrative tumor,at least 70% liver replacement by tu-mor, increase in liver enzyme levels(aspartate or alanine aminotransferaselevel �5� upper limit of normal), acombination of tumor volume at least50% and albumin level less than 3g/dL, bilirubin level increase of atleast 2 mg/dL, diagnosis of cholangio-carcinoma, or predicted lung dose ofgreater than 30 Gy. Patients with anyof these characteristics would be ex-pected to experience early death (�3months) and have an increased likeli-hood of treatment-related adverseevents resulting in death. The absence

of these risk factors was predictive ofimproved survival (median, 466 days)compared with patients in the high-risk group (median, 108 days).

The patients in the low-risk grouprepresented the group for whom therisk/benefit ratio of TheraSpheretreatment was most favorable; that is,optimal survival (�3 months) withminimal likelihood of treatment-re-lated adverse events resulting indeath. A subsequent analysis exam-ined the relationship between the inci-dence of treatment-emergent liver tox-icities and the factors influencingthose events in this group (66). Thepurpose of this analysis was to char-acterize clinically significant (grade�3) liver toxicities observed afterTheraSphere treatment and to identifyfactors associated with increased riskof these events. Because contemporarytreatment approaches involved se-quential lobar infusion of TheraSpherein more than one treatment cycle,treatment-emergent toxicities afterfirst and second treatments were ex-amined. The factors analyzed for asso-ciation with the first treatment toxici-ties were tumor presentation (ie,unilobar, bilobar), presence of portalvein occlusion (yes vs no), presence ofascites (yes vs no), pretreatment totalbilirubin value, percent of tumor re-placing liver volume, treatment ap-proach (lobar vs whole), and liverdose. The factors evaluated for secondtreatment toxicities were duration be-tween treatments (in days), same livervolume treated as the first treatment(yes vs no), pretreatment (day of treat-ment) total bilirubin value, second-treatment liver dose and patient dose(accumulated for same volume orweighted average for different vol-umes). The median dose administeredon first treatment (whole liver or lo-bar) for all 88 patients was 127 Gy(range, 34–153 Gy). Twenty-three pa-tients (26%) received a second treat-ment at a median dose of 129 Gy(range, 27–146 Gy). Two patients re-ceived a third treatment (median, 138Gy; range, 130–145 Gy). Thirty pa-tients (34%) received whole-liver treat-ment in a single administration (me-dian, 105 Gy; range, 46–151 Gy), twoadministrations (one patient; median,100 Gy), or two sequential lobar ad-ministrations (median, 130 Gy; range,89–139 Gy). Fifty-six patients receivedlobar treatment in a single treatment

1574 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 5: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

to one lobe only (median, 133 Gy;range, 34–153 Gy) or repeated treat-ments to the same lobe (median, 246Gy; range, 144–268 Gy). The medianaccumulated dose per patient for all 88patients who received whole-liver orlobar treatment was 130 Gy (range,34–268 Gy). Thirty-seven of 88 pa-tients (42%) experienced 68 liver tox-icities after first or second treatment,including increased bilirubin level (n� 24), ascites (n � 11), increased as-partate or alanine aminotransferase (n� 7), increased alkaline phosphatase(n � 3), hepatic encephalopathy (n �3), liver failure (n � 3), and hemobilia(n � 1). Two patients experiencingliver failure underwent liver biopsiesfor possible radiation-induced liverdisease, without confirmation of thiscondition. When the independent ef-fect of the factors was examined, pre-treatment bilirubin level was associ-ated with the occurrence of livertoxicities for first (P � .0001) and sec-ond (P � .04) treatments, and the me-dian bilirubin level in those patientsexperiencing at least one event wasapproximately two times higher thanin patients who did not experience anevent. There was also an associationbetween liver dose and liver toxicities(first treatment medians, 123 Gy vs 133Gy; P � .08; second treatment medi-ans, 126 Gy vs 132 Gy; P � .26). Therealso appeared to be increased risk ofliver toxicity with decreasing time in-terval between treatments (medians,154 days vs 77 days; P � .05). Furtheranalyses of the joint effects of the var-ious factors with respect to first treat-ment indicated that bilirubin level andliver dose were linearly related (P �.05) with an increased risk (odds andhazard ratios) of patients experiencingliver toxicity. This trend remained forbilirubin on second treatment (three of13 patients [23%] with bilirubin level�1.1 mg/dL experienced liver toxicityvs six of 10 patients [60%] with biliru-bin level �1.1 mg/dL; P � .10). Therewas no association between increasedliver dose and liver toxicities on sec-ond treatment (P � .25). Of note, de-spite the apparent relationship be-tween bilirubin level, liver dose, andliver-related toxicities, the majority (53or 78%) of the 68 observed liver toxic-ities resolved on follow-up. Of the re-maining 15 events, four patients expe-rienced toxicities resulting in death;however, none of the events were

judged to be treatment-related. Onepatient had an increased bilirubinlevel at 565 days after treatment anddied 2 years later. Ten events in ninepatients represented sustained grade 3events (eight bilirubin, one alkalinephosphatase, and one aspartate or ala-nine aminotransferase) with onsetfrom treatment ranging from 28 daysto 119 days. Five of these events werejudged to be unlikely or unrelated totreatment, and the five remaining pa-tients had tumor progression in theuntreated lobe.

In summary, this analysis indicatedthat, for patients who satisfy the eligi-bility criteria defined as low risk, ap-proximately one third would experi-ence at least one severe or life-threatening (grade 3/4) liver-relatedtoxicity after treatment. The risk ofthese toxicities appears to be related topatients’ pretreatment total bilirubinlevel, increasing treatment dose ashigh as 150 Gy per treatment, and thefive liver-related risk factors (65,66).Most of these toxicities (78%) will re-solve; however, some may persist as aresult of tumor progression and/oradvancing cirrhosis. These data con-firmed that treatment-related liver tox-icities increase with increasing pre-treatment total bilirubin levels. Inaddition, in this patient population,radiation-induced liver disease wasnot seen, including in two patientswhose liver biopsy results were nega-tive for this condition. This promisingtreatment option was put in perspec-tive by an editorial published by Daw-son (67).

Rhee et al (9) reported on 14 pa-tients with HCC treated with Thera-Sphere with use of a segmental ap-proach. Dose vials intended for lobarinfusion were administered at the seg-mental level, thereby increasing theeffective dose to the tumor and mini-mizing radiation to the normal paren-chyma. The range of absorbed dose totumor was 105–857 Gy. There was nochange in bilirubin level after treat-ment with this alternative approach.The authors concluded that subseg-mental infusions are safe and feasible,with no alterations in liver functions,setting the stage for radiation segmen-tectomy.

Investigators reported on 43 con-secutive patients treated with Thera-Sphere for unresectable HCC (10).Patients were stratified into three risk

groups according to the method oftreatment and risk stratification(group 0, segmental infusion; group 1,lobar infusion/low risk; group 2, lobarinfusion/high risk), as well as on thebasis of the Okuda and Child-Pughscoring systems. Patients were treatedby liver segment or lobe on one ormore occasions on the basis of tumordistribution, liver function, and vascu-lar flow dynamics, with a volume-weighted average of 138 Gy. Patientswere monitored for adverse events,objective tumor response, and sur-vival. On the basis of follow-up data,20 patients (47%) had an objective tu-mor response based on percent reduc-tion in tumor size (World Health Or-ganization), and 34 patients (79%) hada tumor response when percent reduc-tion and/or tumor necrosis were usedas a composite measure of tumor re-sponse. When patient risk stratifica-tion was used, median survival timeswere 46.5 months (group 0), 16.9months (group 1), and 11.1 months(group 3). Median survival in allpatients in non–high-risk groups(groups 0/1) was 20.8 months. Therewere no treatment-related life-threat-ening adverse events. The authorsconcluded that TheraSphere appearsto be safe and effective in selected pa-tients and represents a promising newtherapy for HCC (10,67).

Investigators reported on 140 pa-tients with HCC who underwent 238administrations to the right and/orleft hepatic artery with TheraSphereon an outpatient basis (68). All pa-tients were stratified by Okuda, Child-Pugh, and low/high risk stratificationas previously described. Patients hadbaseline liver function tests on the dayof treatment, as well as at clinical fol-low-up at 30- to 90-day intervals. Fol-low-up CT/MR images of the liverwere obtained at 30 days and subse-quently at 90-day intervals. Patientswere followed up until death. The me-dian lobar volume, activity delivered,and absorbed dose were 838 mL, 1.61GBq, and 110 Gy, respectively. Thecause of liver disease was hepatitis C(30%), alcohol (24.3%), combined hep-atitis C and alcohol (8.6%), crypto-genic etiology (23.5%), hepatitis B(6.4%), hemochromatosis (3.6%), andother (3.6). There were 54, 85, and onepatient, respectively, with Okuda classI, II, and III disease. There were 69, 66,and five patients, respectively, with

Salem and Thurston • 1575Volume 17 Number 10

Page 6: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

Child-Pugh class A, B, and C disease.There were 97 and 43 patients in low-and high-risk groups, respectively.There was an 11% incidence of grade 3bilirubin toxicities at 3 months afterthe last treatment, most of which wereattributed to cirrhosis and tumor pro-gression. Tumor response rates in pa-tients with complete 3- to 6-monthdata were 32% according to RECISTand 68% according to European Asso-ciation for the Study of the Liver cri-teria. Median survival for patientswith Okuda I and II/III disease were800 days versus 368 days (P � .0001).Median survival times for Child classA and B/C disease were 800 days ver-sus 368 days (P � .0001). Median sur-vival times for low- and high-riskgroups were 800 days versus 258 days(P � .0001). The authors concluded inthis large cohort that TheraSphere uni-lobar infusion for HCC appears to rep-resent an efficacious therapy with ac-ceptable toxicity and promisingsurvival data (68).

Therasphere in HCC in downstagingto permit liver transplantation, resection,or RF ablation.—The use of Thera-Sphere as a bridge to transplantationwas recently described (6,69). The pa-tient was initially excluded fromtransplantation by size criteria. Thepatient was treated with TheraSphereand underwent a successful trans-plantation 42 days after treatment.Pathologic explants demonstratedcomplete necrosis of the tumor withno viable cells.

Kulik et al (5) recently described alarge cohort of 35 patients who werenot candidates for transplantation, re-section, or RF ablation and weretreated with TheraSphere. Sixty-sixpercent of these patients had their dis-ease successfully downstaged to per-mit transplantation, resection, or RFablation. Eight patients received atransplant, and one underwent resec-tion. Five of seven explants in thetransplanted patients demonstratedcomplete necrosis by pathologic exam-ination. Tumor partial response ratewas 50% by World Health Organisa-tion criteria. Median time to partialresponse was 75 days, and mediantime to maximum response was 120days. One-, 2-, and 3-year survivalrates were 84%, 54%, and 27%, respec-tively. Median survival for the entirecohort was 800 days.

Tumor viability on explants after

TheraSphere and other liver-directedtherapies such as TACE and RF abla-tion have been previously described(5,70,71).

Therasphere in colorectal cancer.—In1992, Anderson et al (72) performeda small study with TheraSphere inpatients with colorectal cancer. Amaximum of 150 Gy was adminis-tered with glass 90Y microspheres ad-ministered via the hepatic artery andtargeted to tumor with use of angio-tensin II in seven patients. The au-thors observed no toxicities, and he-patic metastatic progression was de-layed in six patients. Median survivalwas 11 months (range, 5 to �25months).

Goin et al (73) performed a dose-escalation study with TheraSphere in43 patients with colorectal metastases.The study assessed dose-related ef-fects on survival, tumor response, andtoxicity. There were no life-threaten-ing or fatal toxicities. The median sur-vival was 408 days (95% CI, 316–565d). Tumor response was evaluated bydecrease in tumor size assessed bycross-sectional imaging. By these crite-ria, two patients had a complete re-sponse, eight had a partial response,and 35 (81%) had at least stable dis-ease. Higher doses were associatedwith greater tumor response and in-creased survival (P � .05). In addition,tumor hypervascularity (P � .01),greater baseline performance status (P� .002), and less liver involvement (P� .004) were associated with enhancedresponse or survival. Clinical toxicitiesincluded duodenal/gastric ulcers insix patients (14%), which resolvedwith medical management. Thesewere most likely caused by inadver-tent deposition of microspheres intothe gastrointestinal tract via unappre-ciated collateral vessels. Other relatedcomplications included single occur-rences of mild fever and fatigue. Nodose relationship to toxicities was ob-served in the study.

Wong et al (74) presented data onTheraSphere treatment of eight pa-tients with unresectable colorectalliver metastases. Tumor response wasevaluated with imaging (CT/MR im-aging) and metabolic evaluation via[18F]fluorodeoxyglucose (FDG) posi-tron emission tomography (PET) andserum carcinoembryonic antigen(CEA). Five of the eight patients hadan improvement in their tumor activ-

ity, as assessed by a decrease in18FDG-PET metabolic activity andconfirmed by parallel changes in se-rum CEA level. However, as observedin other studies, the use of imaging byCT/MR illustrated that only some ofthe tumors that responded by meta-bolic criteria showed a correspondingdecrease in size. This study suggestedthat the use of tumor size as an indi-cation of treatment response wouldlead to underestimation of the effect ofTheraSphere. The authors concludedthat there was a significant metabolicresponse to TheraSphere treatment inpatients with unresectable colorectalliver metastases. This treatment ap-peared to provide significant pallia-tion for patients with otherwise incur-able disease. In a subsequent study,Wong et al (75) presented data onTheraSphere treatment of 27 patientswith metastatic colorectal cancer to theliver. Tumor response was evaluatedvia 18FDG-PET and serum CEA mea-surement. The study evaluated the useof 18FDG-PET to quantify the meta-bolic response to treatment, compar-ing visual estimates with standardizedhepatic uptake values. Visual esti-mates were graded as indicative ofprogression; no change; or mild, mod-erate, or dramatic improvement. Vi-sual estimates indicated that 20 pa-tients showed a response to treatment,whereas seven patients experiencedprogression or no change in their dis-ease. In the response group, there wasa significant correlation (r � 0.75; P �.0001) between the responses identi-fied through visual estimation andthose determined by hepatic standard-ized uptake values. No significant cor-relation was observed with CEA val-ues (P � .13), which was attributed tothe effect of extrahepatic lesions. Theauthors concluded that treatment sig-nificantly reduced hepatic tumor me-tabolism and appeared to be palliativein patients with unresectable liver me-tastases.

In a recent article, Lewandowski etal (76) reported on 27 patients withunresectable colorectal cancer treatedwith TheraSphere. The targeted ab-sorbed radiation dose was 135–150 Gy.They found that TheraSphere pro-vided stabilization of liver disease inpatients in whom chemotherapy hadfailed. Tumor response measured byPET imaging exceeded that of CT im-aging for first (88% vs 35%) and sec-

1576 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 7: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

ond (73% vs 36%) treated lobes, re-spectively. Tumor replacement lessthan 25% (vs �25%) was associatedwith a significant increase in mediansurvival (339 days vs 162 days). Treat-ment-related toxicities included mildfatigue (48%), nausea (15%), andvague abdominal pain (19%). Therewas one case of radiation-induced gas-tritis from inadvertent deposition ofmicrospheres to the gastrointestinaltract (4%). Interestingly, the responserate in this study duplicates the 35%rate obtained in 1978 in a similar co-hort treated with intraarterial chemo-therapy and microspheres (77).

Atassi et al (78) described a 71-pa-tient cohort with metastatic colorectalcancer treated with TheraSphere at atargeted dose of 120 Gy on a lobarbasis. The median lobar volume, activ-ity administered, and net dose were825 mL, 127 Gy, and 2.16 GBq, respec-tively. Treatment-related toxicities in-cluded mild fatigue (n � 28; 39%) andtransient vague abdominal discomfort(n � 14; 20%). Grade 3/4 bilirubin tox-icity occurred in three patients (4%),all of whom had documented tumorprogression. There were no cases ofradiation-induced gastritis/ulceration,hepatitis, or pneumonitis. The partialtumor response rate by RECIST was35%. Survival rates from first treat-ment at 1 and 2 years were 39.1% and22.1%, respectively. Factors associatedwith prolonged survival includedbaseline Eastern Cooperative Oncol-ogy Group (ECOG) performance sta-tus, number of tumors, and presenceor absence of extrahepatic disease.ECOG scores of 0/1 were associatedwith median survival times of 566days and 219 days, respectively (P �.0001). The number of tumor nodules(�4 vs �4) was associated with me-dian survival times of 566 days and216 days, respectively (P � .0001). Thepresence or absence of extrahepaticdisease was related to median survivalof 187 and 407 days, respectively (P �.003).

The authors concluded that, in se-lected patients with liver-dominantcolorectal metastatic disease, Thera-Sphere appears to provide a viabletreatment alternative (78).

Therasphere in mixed neoplasia.—In1988, Herba et al (15) treated 15 pa-tients, 12 with metastatic colorectalcancers, one with a carcinoid tumor,one with an islet-cell tumor, and one

with a hepatoma, with three doselevels: 5,000 cGy (n � 10), 7,500 cGy(n � 3), and 10,000 cGy (n � 2).Mean follow-up time was 7 months(range, 2–12 months). Stable diseasein the liver was seen in 10 patients,four of whom had concurrent pro-gression of extrahepatic disease,which resulted in two deaths. Twoadditional deaths were not directlyrelated to the malignant process. Pro-gression of liver disease was foundin five patients, with three deaths oc-curring at 7–8 months. No proce-dural, hematologic, or pulmonarycomplications occurred. Late gas-troduodenal ulceration occurred at6–8 weeks in three patients who hadhistories of chronic alcohol abuse.This method of therapy seems to befeasible and efficient. Caution is nec-essary with high doses or with pa-tients with a history of, or predispo-sition to, gastroduodenal ulcers.

In 1989, Blanchard et al (79) de-scribed the use of 90Y microspheres asmonotherapy in 14 liver metastasesand one case of HCC. Twenty addi-tional patients were screened but werefound to be unsuitable for this ther-apy. Five patients exhibited a responseby World Health Organisation criteria(33%). Treatment was complicated bygastric ulceration in six patients (40%).Survival in the treated group was 62weeks, whereas it was 30 weeks forthe untreated cohort. The authors con-cluded that 90Y monotherapy might beeffective in reducing tumor size inproperly selected patients.

Andrews et al (11) presented dataon 24 patients including 17 with colo-rectal metastases to the liver, six withmetastatic neuroendocrine tumors,and one with HCC. Imaging at week16 indicated a partial response in fivepatients, minimal response in four,stable disease in seven, and progres-sive disease in the remaining eight.Other than mild gastrointestinalsymptoms in four patients (unrelatedto TheraSphere treatment), no hema-tologic, hepatic, or pulmonary toxici-ties were observed. The authors con-sidered the hepatic tolerance toradiation delivered by 90Y to be excel-lent at doses as high as 150 Gy used inthe study.

Herba and Thirlwell (80) performeda prospective dose-escalation studywith TheraSphere starting at 50 Gyand escalating in 25-Gy increments to

150 Gy. There were 37 patients withliver metastases, 33 of whom had colo-rectal metastases to the liver. The au-thors observed no major hematologicor pulmonary complications but didobserve some gastroduodenal ulcer-ation, which occurred early in theirclinical experience with TheraSphere,as a result of inadvertent deposition ofspheres in the gastrointestinal tract. Abeneficial response was observed byCT in cases in which tumors could beresolved. Stabilization or decrease intumor size was observed in 22 of 30patients (73%). Because of the smallsample size of the study, no statisti-cally significant relationship betweendose and clinical or radiologic benefi-cial effects was observed. The authorsconcluded that TheraSphere treatmentwas a feasible and safe technique withbeneficial effects.

Lewandowski et al (81) reported ona large cohort of patients with livermetastases treated with TheraSphere.One hundred thirteen patients under-went 171 administrations. Primary tu-mors included colon (n � 43), breast (n� 18), neuroendocrine (n � 11), pan-creatic (n � 7), lung (n � 5), cholan-giocarcinoma (n � 5), melanoma (n �4), renal (n � 4), esophageal (n � 3),ovarian (n � 2), adenocarcinoma withunknown primary tumor (n � 2), andlymphoma, gastric, duodenal, blad-der, angiosarcoma, squamous-cell car-cinoma, thyroid, adrenal, and parotid(n � 1 each). Standard-of-care multi-ple-agent chemotherapy had failed inall patients. Patients underwent base-line and follow-up liver function tests,tumor markers, CT/MR imaging, andPET imaging. They were followed upfor survival from time of first treat-ment. The median age of patients was62 years. The mean activity and doseinfused were 2.4 GBq and 117 Gy, re-spectively. All patients were dis-charged within 6 hours after the pro-cedure. Clinical toxicities includedfatigue (54%), abdominal pain (11%),and nausea/vomiting (11%). Therewere no ulcers. Follow-up CEA mea-surements in patients with colorectalcancer demonstrated a mean 61% de-crease after therapy. On imaging fol-low-up, RECIST, European Associa-tion for the Study of the Liver, andPET response rates of 29%, 67%, and79% were obtained, respectively. Me-dian survival from first treatment forall patients was 10 months. The au-

Salem and Thurston • 1577Volume 17 Number 10

Page 8: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

thors concluded that TheraSphere he-patic treatments can be performedsafely on an outpatient basis in pa-tients with unresectable liver neopla-sia. Treatments are well tolerated,with acceptable toxicities. Response to90Y in these patients is supported bythe decrease in metabolic activity onPET and tumor markers (81).

SIR-Spheres in HCC.—The earliestreports of the use of resin micro-spheres were in the late 1970s(82–86). Since then, there has notbeen a significant body of literaturediscussing the use of resin micro-spheres in this clinical setting.

In 1998, Lau et al (87) reported on71 patients with HCC treated withSIR-Spheres. Patients were treatedwith percutaneous techniques or intra-hepatic ports. Response was measuredby AFP measurements and/or cross-sectional imaging. Patients were ad-ministered activities ranging from 0.8to 5.0 GBq (21.6–135.1 mCi; median,3.0 GBq or 81.1 mCi). Partial responserate as measured by 50% reduction intumor volume occurred in 19 patients(26.7%) after the first treatment. Over-all objective response in terms ofchanges in AFP level was 89% (partialresponse, 67%; complete response,22%) among the 46 patients with in-creased pretreatment AFP levels.Treatment was repeated in 15 patients.The maximum number of treatmentswas five, and the maximum total ac-tivity was 13.0 GBq (351.4 mCi) givenin three treatments. Four patients un-derwent resection after treatment. Intwo cases, complete histologic remis-sion was noted, and in the other two,occasional viable tumor cells werefound in the necrotic centers. The me-dian survival of the 71 patients was 9.4months (range, 1.8–46.4 months).Treatment was well tolerated, andthere was no bone marrow toxicity orclinical evidence of radiation hepatitisor pneumonitis. The authors con-cluded that (i) selective internal radia-tion therapy (SIRT) with 90Y micro-spheres is effective in selected cases ofunresectable HCC and is well toler-ated, (ii) objective response rate interms of decrease in AFP is higherthan that based on reduction in tumorvolume shown by CT, (ii) nontumor-ous liver appears more tolerant to in-ternal radiation than to external-beamradiation, and (iv) selective internal ra-

diation treatment may make some un-resectable tumors resectable (87).

In 2001, Lau et al (88) reported re-sults from data during a treatment pe-riod from 1992 through 1996 in which82 patients with unresectable HCCwere treated with 90Y. Analyses of out-comes were used to classify patients as“short survivors” (died within 1 year;n � 51; 62%), or “long survivors”(lived for �1 year from treatment; n �31; 34%). Comparisons betweengroups suggested that low pretreat-ment AFP levels and high tumor-to-liver uptake ratios favored longer sur-vival.

Szeto et al (89) reported on a hepa-titis B surface antigen–positive patientwho experienced HCC 7 years aftercadaveric kidney transplantation. Thepatient was treated with SIR-Spheres.Serum AFP levels normalized within 2weeks. Follow-up imaging demon-strated significant necrosis of the tu-mor and compensatory hypertrophyof the normal parenchyma. The treat-ment was well tolerated except fortransient liver function deterioration.The patient died of liver failure after15 disease-free months.

SIR-Spheres in downstaging to per-mit liver transplantation, resection, orRF ablation.—In 2004, Lau et al (7) re-ported on salvage hepatic surgery in49 patients with HCC treated withdownstaging procedures. Thirty-twopatients had combination chemother-apy alone (65.3%), eight patients hadsingle-agent chemotherapy alone(16.3%), four patients received in-traarterial SIR-Spheres alone (8.2%),and five patients received sequentialtherapy (10.2%). Twenty-eight pa-tients (57.1%) underwent major he-patic resection. Thirteen patients(26.5%) had complete necrosis of thetumor after treatment. Twenty-onepatients (42.9%) had recurrence aftersurgery, and 14 of these were intra-hepatic recurrence. The median sur-vival time was 85.9 months. The1-year, 3-year, and 5-year survivalrates were 98%, 64%, and 57%, re-spectively.

90Y resin microspheres or SIR-Spheres in colorectal cancer.—In 1989,investigators reported on dosimetrywork in 10 patients with metastaticcolorectal cancer (90). The micro-spheres were concentrated in the mi-crovasculature of the tumor nodulesby the concurrent administration of

angiotensin II. The radiation dose de-livered to liver parenchyma wasmeasured at the time of operation byan intraoperative radiation detectionprobe. The investigators concludedthat intraoperative dosimetry con-firmed the poor correlation betweentotal radioactivity used and radiationdose received by normal liver paren-chyma.

In 1992, investigators publishedtheir experience in the treatment of 29patients with metastatic colorectalcancer with a surgically implantablehepatic artery pump (91). Vasoactiveagents were used to enhance tumorflow of microspheres. The overallmean decrease in CEA level was 70%of pretreatment levels, with 88% of pa-tients (23 of 26) experiencing morethan a 50% decrease in pretreatmentCEA levels. In 48% of patients, therewas a decrease in tumor volume bymore than 50%. The authors con-cluded that SIR-Spheres treatment waseffective in causing tumor regressionin patients with liver metastases sec-ondary to large bowel cancer.

Gray et al (92) published a phase IIIrandomized clinical trial of 74 patientsconducted to assess whether a singleinjection of SIR-Spheres in combina-tion with intrahepatic floxuridinecould increase the tumor responserate, time to disease progression in theliver, and survival compared withfloxuridine alone. Treatment-relatedtoxicities and changes in quality of lifewere also examined. All patients hadundergone complete surgical resectionof a primary adenocarcinoma of thelarge bowel, and only those with un-resectable metastases limited to theliver and lymph nodes in the portahepatis were included in the study. Inaddition, patients were required tohave World Health Organisation per-formance status of 0–2, adequate he-matologic and hepatic function, andno evidence of cirrhosis or ascites.Both treatment arms received 12-daycycles of continuous-infusion floxuri-dine at 0.3 mg/kg at four weekly in-tervals, which was continued for 18cycles or until there was evidence oftumor progression, extrahepatic me-tastases requiring a systemic chemo-therapy change, unacceptable toxicity,or port failure, or the patient requestedcessation of treatment. The SIR-Spheres treatment arm also received apredetermined quantity of 90Y that

1578 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 9: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

varied (2 GBq, 2.5 GBq, or 3 GBq) de-pending on the size of the tumor. 90Ymicrospheres were administered onetime only, within 4 weeks of insertionof the hepatic artery access port. Themean SIR-Spheres dose administeredwas 2.156 � 0.32 GBq. There was nodifference between the 90Y arm andcontrol arm in the mean floxuridinedose (1,863 � 1,735 mg vs 1,822 �1,323 mg) or the mean number of cy-cles of chemotherapy (8.7 � 5.6 vs 8.0� 5.0). Six of 34 patients in the hepaticartery chemotherapy (HAC) arm(18%) had at least a partial response,whereas 16 of 36 patients in the HAC/90Y arm (44%) had at least a partialresponse. The partial plus complete re-sponse rate was significantly greaterfor patients receiving SIR-Sphereswhen measured by tumor area (44%vs 17.6%; P � .01) tumor volume (50%vs 24%; P � .03), and CEA level (72%vs 47%; P � .004). The median time todisease progression in the liver wassignificantly longer for patients whoreceived SIR-Spheres than in thosewho received HAC alone when mea-sured by tumor area (9.7 months vs15.9 months; P � .001), tumor volume(7.6 months vs 12.0 months; P � .04),or CEA level (5.7 months vs 6.7months; P � .06). The 1-, 2-, 3-, and5-year survival rates for patientstreated with SIR-Spheres were 72%,39%, 17%, and 3.5%, respectively,compared with 68%, 29%, 6.5%, and 0for HAC alone. Cox regression analy-sis suggested an improvement in sur-vival for patients treated with SIR-Spheres who survived more than 15months (P � .06). There was no in-crease in grade 3/4 treatment-relatedtoxicity and no loss of quality of lifefor patients treated with SIR-Spherescompared with patients treated withHAC alone.

The authors concluded that thecombination of a single injection ofSIR-Spheres plus HAC was substan-tially more effective in increasing tu-mor response and progression-freesurvival than the same regimen ofHAC alone (92).

Investigators reported on a 50-pa-tient cohort with extensive colorectalliver metastases not suitable for resec-tion or cryotherapy (93). The studycompared experience with 90Y alone (n� 7) and in combination with 5-flu-orouracil (5-FU; n � 43). For all pa-tients, 90Y microspheres were admin-

istered as a single treatment within 10days of hepatic artery port placement.The dose was titrated to the estimatedextent of disease (�25% liver replace-ment, 2 GBq; 25%–50% liver replace-ment, 2.5 GBq; �50% liver replace-ment, 3 GBq) and given over a periodof 10 minutes a few minutes after ad-ministration of 50 �g angiotensin II.Forty-three of the 50 enrolled patientsalso received 5-FU given at the time of90Y continuously over a period of 4days (1 g/d) every 4 weeks. Acutepain and/or nausea were experiencedby 14 patients (28%) at the time ofadministration of SIR-Spheres andwere managed with narcotics and an-tiemetic agents. Six patients (12%) ex-perienced an acute duodenal ulcerwithin 2 months after SIR-Spherestherapy and the initial cycle of 5-FU asa result of misperfusion of the duode-num by 90Y, 5-FU, or both. MedianCEA levels were reduced to 25% ofbaseline values at 1 month after treat-ment with 90Y and remained at lessthan 30% of baseline levels when mon-itored for 6 months. Among all pa-tients with liver metastases, mediansurvival from the time of diagnosiswas 14.5 months (range, 1.9–91.4months), and median survival fromthe time of treatment was 9.8 months(range 1.0–30.3 months).

The same group (94) described 38patients with extensive colorectal livermetastases who received SIR-Spheres.Liver involvement was less than 25%in 19 patients, 25%–50% in nine pa-tients, and greater than 50% in 10 pa-tients. Patients received 90Y in the he-patic artery via an arterial port andsubsequent every-4-week cycles ofHAC with 5-FU. The treatments werewell tolerated, and no treatment-re-lated mortality was observed. Re-sponse to SIR-Spheres therapy, as in-dicated by decreasing tumor markersand serial CT every 3 months, wasseen in more than 90% of patients. Es-timated survival rates at 6, 12, and 18months were 70%, 46%, and 46%, re-spectively, and were principallydriven by the development of extrahe-patic metastases. The authors con-cluded that SIR-Spheres treatment waswell tolerated in patients with exten-sive colorectal liver metastases andachieved encouraging liver tumor re-sponses that are well maintained byHAC.

Van Hazel at al (18) reported a ran-

domized study of 21 patients in which11 patients received SIR-Spheres plus5-FU/leucovorin and 10 received5-FU/leucovorin alone. The mean ad-ministered radiation dose in those re-ceiving SIR-Spheres was 2.25 GBq. Theauthors concluded that the adminis-tration of SIR-Spheres along with astandard chemotherapeutic regimensignificantly increased treatment-re-lated response (10 vs zero patientsshowed a partial response on fol-low-up CT), time to disease progres-sion (18.6 months vs 3.6 months), andsurvival (29.4 months vs 12.8 months)in comparison with chemotherapyalone. Even though more toxicitieswere associated with the combinationtherapy, there was no difference inquality of life over a 3-month period(18).

Geller et al (95) described a groupof 19 patients with chemotherapy-re-fractory colorectal cancer treated withSIR-Spheres. One hepatic lobe wastreated through the right or left he-patic artery at any treatment session.Patients received 40 Gy � 20%. Therewere no toxicities that were greaterthan grade 2. Eighty percent of pa-tients had grade 1 granulocyte orplatelet decrease, 50% had grade 1/2lymphocyte decrease, and only case ofgrade 1 elevation in bilirubin level wasseen. The conditions of 13 patientscould be evaluated for response at 2months at the lobar level. There weresix minor responses (46%), three casesof stable disease (23%), two mixed re-sponses, and two patients with pro-gressive tumors. The authors concludedthat these represented encouraginginitial results for heavily pretreatedpatients with advanced-stage colonmetastases in the liver.

Coldwell et al (14) presented an ab-stract describing 84 patients receiving127 infusions of SIR-Spheres for colo-rectal metastases to the liver. The tar-get dose was 90 Gy to the tumor and30 Gy to the normal parenchyma. Ob-jective response rates were 35% by CT,70% by CEA, and 90% by FDG-PET.Mean follow-up time was 12 months,with median survival not having beenreached at the time of presentation. Nolife-threatening toxicities were noted.All patients who showed fluoroscopiccessation of blood flow (ie, stasis) dur-ing 90Y infusion experienced post-embolization syndrome. The authorsconcluded that radioembolization pro-

Salem and Thurston • 1579Volume 17 Number 10

Page 10: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

vided encouraging response rates withan improvement in overall survivaland acceptable toxicity in the group ofpatients treated.

Van Hazel et al (20) presented re-sults of a phase I/II dose-escalationstudy combining SIR-Spheres withsystemic chemotherapy. Seventeen pa-tients with liver metastases from colo-rectal cancer, who had not receivedprevious chemotherapy for metastaticdisease, were entered into the study.Chemotherapy consisted of the FOL-FOX4 regimen (oxaliplatin/5-FU/leu-covorin) modified with an oxaliplatindosage escalated from 30 mg/m2 to 85mg/m2 to assess safety and tolerabil-ity. SIR-Spheres were implanted by in-jection into the hepatic arterial systemon day 3 or 4 of the first chemotherapycycle. Most patients experienced nau-sea or abdominal pain within 48 hoursof 90Y administration. Mild peripheralneuropathy was evident in six of thefirst 11 patients, appearing betweencycles 10 and 12. Grade 3/4 transientneutropenia was seen in five of 11 pa-tients. Three grade 3 events were doc-umented in two of three patientstreated at the 30-mg/m2 tier (diarrhea,nausea/vomiting, fever). Six grade 3events were seen in four of eight pa-tients treated at the 60-mg/m2 tier (di-arrhea, nausea/vomiting, abdominalpain, fever, leukopenia, anemia). Onlyone grade 3 event has been shown todate in one of six patients at the 85-mg/m2 tier (nausea/vomiting). Partialresponses (by RECIST criteria) wereseen in all of the first 11 patients. Themedian time to liver progression at thetime of presentation was 11 months.The authors concluded that SIR-Spheres therapy in combination withoxaliplatin appears to possess an ac-ceptable toxicity profile. Further trialsare under way (20).

Goldstein et al (96) presented aphase I dose-escalation study with iri-notecan in which 25 patients whoseprevious chemotherapy had failed butwho were irinotecan naive had beenentered into the study. Irinotecan wasgiven weekly twice every 3 weeksstarting the day before 90Y administra-tion for a maximum of nine cycles.Irinotecan dosage was escalated from50 mg/m2 to 100 mg/m2 to assesssafety and tolerability and to identify amaximum tolerated dosage. If nograde 3/4 toxicity was experienced af-ter two cycles, patients could have

their irinotecan dosages escalated to75 mg/m2 and then 100 mg/m2. Early-stage acute and self-limiting nausea,vomiting, and liver pain were experi-enced by most patients. Mild lethargyand anorexia was also common. Grade3/4 toxic events were seen in four ofsix patients treated at a dosage of 50mg/m2 (jaundice, n � 1; ascites, n � 1,thrombocytopenia, n � 1; increasedALP level, n � 1), four of 13 treated at75 mg/m2 (ascites, n � 1; thrombocy-topenia, n � 1; abdominal pain, n � 1;fatigue, n � 1), and two of six (at thetime of publication) treated at 100mg/m2 (deep vein thrombosis, n � 1;abdominal pain, n � 1). Partial re-sponses were seen in nine of 17 pa-tients, median time to liver progres-sion was 7.5 months, and mediansurvival time was 12 months. The au-thors concluded that the use of SIR-Spheres in combination with irinote-can was associated with acceptabletoxicity. Trials of first-line therapywith irinotecan and 5-FU/leucovorincombinations were then initiated (96).

In the same group, Sharma et al (97)presented another phase I dose-escala-tion study of oxaliplatin. The objectivewas to determine the MTD of oxalipla-tin with infusional 5-FU in combina-tion with 90Y in the first-line treatmentof patients with inoperable liver me-tastases from colorectal cancer. Sec-ondary objectives included efficacy asdemonstrated by response rate, timeto progression, and overall survival.Oxaliplatin dosages were escalated at30 mg/m2, 60 mg/m2, and 85 mg/m2.If no grade 3/4 toxicity was experi-enced after two cycles, individual pa-tients could have their dosages esca-lated to 85 mg/m2. Chemotherapytreatment commenced 2 days before90Y therapy. Fourteen patients wereenrolled in the study. Five patientshad extrahepatic metastases. Acute,mild, self-limiting nausea, vomiting,and liver pain were experienced bymost patients. Mild lethargy and an-orexia was also common. Grade 3/4toxic events were reported in one ofthree patients treated at a dosage of 30mg/m2 (neutropenia and diarrhea),four of eight patients treated at 60mg/m2 (neutropenia, n � 2; diarrheaand neutropenia, n � 1; neutropenia, n� 1). Recruitment of three of six pa-tients occurred at a dosage of 85 mg/m2. Partial responses were seen in 10of 11 patients whose conditions were

evaluable. The authors concluded thatSIR-Spheres in combination with FOL-FOX4 (oxaliplatin/5-FU/leucovorin)is associated with acceptable overalltoxicity. Although the MTD was notreached, it is expected to be 85 mg/m2

(19). Since the publication of these ini-tial data, the authors have updatedtheir findings (97). A total of 20 pa-tients were enrolled in the study. Themean dose of SIRT administered was1.7 GBq. National Cancer Institutegrade 1–3 abdominal pain within 48hours of SIRT was noted in eight pa-tients, one of whom had microspheresdetected in the gastric mucosa by en-doscopy. Grade 3/4 neutropenia wasseen in 12 patients, seven of whomwere treated at the highest dose level,thereby defining the dose-limiting tox-icity. The nadir in mean leukocyte lev-els was noted 2 months after SIRT.Although one episode of transientgrade 4 hepatotoxicity occurred, therewas no radiologic evidence of radia-tion hepatitis in any of the patientstreated. Partial responses by RECISTcriteria were seen in 18 patients, andstability was noted in the remainingtwo patients. Two patients underwenthepatic surgery after protocol therapy,with no significant postoperative com-plications. Pathologic examination ofhepatic samples revealed smallamounts of viable adenocarcinomawith small foci of necrosis and exten-sive surrounding fibrosis. In the sub-set of patients with no extrahepaticmetastases, mean time to progressionwas 14.2 months. The authors closedtheir study and concluded that themaximum tolerated dose for SIRTwith concomitant chemotherapy withFOLFOX4 (oxaliplatin/5-FU/leucov-orin) is 60 mg/m2 of oxaliplatin for thefirst three cycles. They also noted thatdownstaging of disease should be con-sidered in selected patients after treat-ment (97).

Murthy et al (98) reported on 12patients with advanced unresectablecolorectal hepatic metastases treatedwith SIR-Spheres. A total of 17 infu-sions were administered. The averagemedian prescribed dose was 39.6 mCi.The delivered dose in six infusions(35%) was less than the prescribeddose as a result of embolic arterial oc-clusion. Radiologic response was sta-ble disease in five of nine patients(56%), and CEA levels decreased infour of seven patients (57%). Median

1580 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 11: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

survival times from diagnosis andtreatment were 24.6 and 4.5 months,respectively. In seven of the 17 infu-sions (41%), the patient experiencedtransient abdominal pain and nausea.One patient experienced a gastric ulcerthat was managed nonoperatively(98).

Lim et al (99) reported on a 30-patient cohort composed of patientswith liver metastases from colorectalorigin in which 5-FU–based therapieshad failed. There was a 33% responserate, with median response durationof 8.3 months and a time to progres-sion of 5.3 months. Response rate andprogression-free survival rates werelower in patients in whom earlier che-motherapy had failed (21% and 3.9months, respectively). Four cases oflate gastrointestinal ulcers were noted.

Kennedy et al (100) recently re-ported on a seven-center study inwhich 208 patients in whom irinotecanand/or oxaliplatin-based chemother-apy had failed were treated with SIR-Spheres on a lobar and whole-liver ba-sis. All were monitored for clinical andlaboratory toxicities, anatomic andfunctional response, and overall sur-vival. The most common clinical tox-icities were constitutional (eg, fatigue,fever, weight loss), gastrointestinal(eg, nausea, vomiting, gastric ulcers),and bilirubin level increases, whichoccurred in 45%, 30%, and 4.5% of pa-tients, respectively. Five percent of pa-tients experienced gastrointestinal ul-ceration. There were no cases ofradiation-induced liver disease. Theresponse rate on imaging was 35%,whereas the PET response rate was91%. Survival rates were 10.5 monthsand 4.5 months for responders andnonresponders, respectively. The au-thors concluded that, for patients withmetastatic colorectal cancers, resin mi-crospheres provide acceptable clinicaltoxicities, significant objective imagingresponses, and promising survivalrates (100).

One group presented the prelimi-nary results in a 50-patient ongoingphase II study (101). Seventeen pa-tients had been enrolled at the time ofthe initial report. Nine patients had atumor burden less than 25%, whereasthe remaining eight had a 25%–50%tumor burden. The majority had bilo-bar disease; the mean dose adminis-tered was 1.6 GBq. Three months aftertreatment, there was one complete re-

sponse by necrosis criteria, three pa-tients had stable disease, two had dis-ease progression, and two othersunderwent resection. The authors con-cluded that SIRT may be used as adownstaging procedure for surgicalresection in selected patients and thatRF ablation may play a complimen-tary role (101).

SIR-Spheres in mixed neoplasia.—Ru-bin et al (17) presented a case report ofa patient with metastatic breast cancerto the liver treated with SIR-Spheres.The authors concluded that the use ofan integrative approach to cancer treat-ment including SIR-Spheres was suc-cessful in the performance of palliativetherapy in a patient with metastaticbreast cancer.

Coldwell et al (12) reported on 34women with unresectable breast can-cer metastatic to the liver treated withSIR-Spheres. Inclusion criteria in-cluded only those patients with anECOG performance score of 0/1 withan expected survival of at least 3months. The average dose of radiationadministered was 1.75 GBq. Althoughall patients showed a response totreatment with a reduction of thenumber and size of the hepatic lesionson PET, all patients also experiencedmild to moderate postembolizationsyndrome.

Lim et al (102) reported on 46 pa-tients with unresectable hepatic malig-nancies (colorectal cancer, n � 32; hep-atocellular carcinoma, n � 5; otherdisease, n � 9) treated with SIR-Spheres. These selected patients hadECOG performance scores of 2 or less,life expectancy of more than 3 months,and no brain metastases at the time oftreatment. Follow-up data were avail-able for 43 of the patients. Of these, 12patients (27%) showed partial re-sponse to therapy (10 with metastaticcolorectal cancer and one with hepa-toma), whereas another 12 patients(27%) had stable disease. The medianduration of response for all patientswas 8.6 months (range, 2–21 months).Early toxicities were minimal, andfour patients experienced delayed gas-tric ulceration (102).

The use of SIR-Spheres for meta-static neuroendocrine cancers is ofparticular interest. The high embolicload coupled with low specific activitymakes this an ideal therapeutic optionfor this condition. An 84-patient co-hort was described in 2005 (13). Tumor

dose was 1,000 Gy to tumor volume.There were 14 cases (17%) of grade 3gastrointestinal toxicity. Sixty-sevenpercent of patients experienced re-sponse on follow-up PET, with diseasestabilization in the remaining patients.Eighty percent of patients who experi-enced symptoms reported relief of thesymptoms. The authors concludedthat the administration of 90Y SIR-Spheres is a viable treatment optionfor metastatic neuroendocrine disease(13). Another cohort was presented in2006 in which investigators used 24fractions to treat 18 patients (103). Theinvestigators observed an 89% objec-tive tumor response rate by imagingand chromogranin A measurement.There were no treatment-relateddeaths and no significant toxicities.Median survival had not been reachedby the time of publication (103).

Popperl et al (104) recently reportedon 23 patients with unresectable he-patic malignancies (21 with metastaticdisease and two with HCC) treatedwith SIR-Spheres. The mean activity oftreatment was 2,270 MBq. Follow-updata in 13 of 23 patients showed amarked decrease of FDG uptake, a de-crease in tumor marker levels, and un-changed or slightly decreasing lesionsize (on CT) in 10 of 13 patients (one ofwhom had HCC). Disease stabilitywas found in two patients, whereasanother patient experienced progres-sive disease. Long-term follow-up in-vestigations were available in two of23 patients, in whom hepatic and ex-trahepatic progression was seen 6 and9 months after 90Y therapy. Minor sideeffects included abdominal pain andfever. Pancreatitis and gastric ulcer-ation were also observed (104).

Wong et al (105) described 19 pa-tients with unresectable chemothera-py-refractive hepatic metastatic dis-ease of various origins treated withSIR-Spheres. The median absorbeddose for the tumor was 76 Gy. PETwas used to monitor patients at3-month intervals. By PET criteria, 15of patients (79%) showed response totherapy, whereas four (21%) showedno response. The authors concludedthat there is a significant reduction ofhepatic metastatic load as evaluatedby PET after radioembolization (105).

Murthy et al (101) described a co-hort of patients with metastatic lungcancer treated with SIRT. Six patientswith unresectable hepatic metastases

Salem and Thurston • 1581Volume 17 Number 10

Page 12: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

were treated with eight infusions ofSIR-Spheres after failed systemic che-motherapy, RF ablation, or arterialembolization. SIRT was administeredas second- to sixth-line therapy. Themedian interval from diagnosis to SIR-Spheres treatment was 20.5 months(range, 6–51 months). A median doseof 36.1 mCi (range, 12.9–54 mCi) wasdelivered. Single-photon emissionCT/CT fusion Bremsstrahlung scansafter therapy confirmed preferentialdeposition of SIR-Spheres within me-tastases. Responses to therapy in-cluded a decrease in the size of thehepatic metastases in one patient andstable disease in two patients. One pa-tient had a mixed response, and twopatients had progression of disease.One grade 3 liver toxicity and onegrade 4 liver toxicity occurred. All pa-tients experienced grade 1/2 fatigue.Time to progression of liver diseaseranged from 3 months to 9 months.The authors concluded that adminis-tration of SIR-Spheres is a feasible al-ternative to systemic therapy for pa-tients with liver-dominant metastasesfrom lung cancers (101). Although se-rious hepatic toxicity was noted in pa-tients with advanced liver metastases,the treatment was tolerated with onlyreversible fatigue in the majority ofpatients. Finally, the authors con-cluded that, when the treatment wasdeemed effective, the duration of localdisease control after one treatmentequaled or exceeded what would beexpected with chemotherapy (101).

Investigators reported on 49 pa-tients who underwent 78 administra-tions of SIR-Spheres via intraarterialinfusion in the left or right hepatic ar-tery (106). Of these patients, 26 had adiagnosis of colorectal cancer meta-static to the liver. Pancreatic, breast,carcinoid, lung, thyroid, squamous-cell, renal-cell, gastrointestinal stromaltumor, and endometrial cancer wereamong the other primary malignan-cies with hepatic metastases. All pa-tients had undergone previous multi-ple-agent chemotherapy, which hadfailed. The median target volume was993 mL. The mean dose of resin micro-spheres administered was 0.83 GBq,which translates to 42 Gy absorbeddose. Forty-seven of 49 patients re-ceived treatment on an outpatient ba-sis. Clinical toxicities included fatigue(n � 18; 37%), vague abdominal pain(n � 10; 20%), and nausea/vomiting (n

� 10; 20%). Three patients (6%) expe-rienced ascites and/or leg edema aftertreatment as a consequence of liverfailure in advanced-stage metastaticdisease. Follow-up CEA measurementin patients with colorectal cancer dem-onstrated a mean 59% decrease aftertherapy over a mean period of 62 days.On CT/MR imaging follow-up, aRECIST response rate of 29% and aPET response rate of 79% were noted.Mean and median survival times were305 days and 175 days, respectively(106).

Two reports by the same group(107,108) were presented in which lab-oratory toxicities and midterm resultsof the use of SIR-Spheres were dis-cussed. Thirty-four patients with ex-tensive hepatic metastases weretreated, and their toxicities were as-sessed by Southwest Oncology Groupcriteria. Twenty-two patients (65%) ex-perienced 33 liver toxicities. At3-month follow-up, 17 grade 0 and 16grade 1 toxicities were observed. Notoxicities of grade 3 or greater wererecorded. At 6-month follow-up, sixgrade 0, 15 grade 1, four grade 2, andone grade 3 toxicities were noted. Themost frequent toxicity was increasedaminotransferase levels, followed byincreased alkaline phosphatase levelsand increased bilirubin levels. Eight ofthe 33 toxicities resolved on follow-up,whereas five patients had disease pro-gression. The case of grade 3 toxicitywas attributed to tumor progression. Itwas concluded that whole-liver treat-ment with intraarterial injection of 90Ymicrospheres in a single session is asafe palliative treatment option in pa-tients with extensive hepatic metasta-ses. No life-threatening liver-relatedtoxicities were observed. The samegroup discussed the clinical follow-upof the 34-patient cohort (108). Two ma-jor adverse events were observed (gas-tric ulcers). Mean follow-up time was242 days (range, 62–586 days). Twopatients died within 90 days aftertreatment. One patient underwent RFablation of residual metastases. Themean hepatic tumor volume of all pa-tients increased by 2% after 3 monthsof 5-FU therapy, decreased by 5% after6 months, and increased again by 26%after 9 months. According to RECISTcriteria, partial response was detectedin 11, four, and two patients at 3, 6,and 9 months, respectively. Stable dis-ease was seen in six, four, and two

patients at 3, 6, and 9 months, respec-tively. Progressive disease was notedin five patients at 3- and 6-month fol-low-up. No complete remission wasobserved. Initially, in almost all pa-tients, a significant decrease in tumormarkers was recorded, supporting theclinical benefit of the therapy(107,108).

Jakobs et al (109) presented mid-term results on 88 patients who re-ceived 118 administrations of SIR-Spheres. Forty-five of these patientshad colorectal cancer, whereas the re-maining cases comprised pancreatic,breast, neuroendocrine, lung thyroid,squamous-cell, gastrointestinal stro-mal, thymus, melanoma, and endome-trial cancers, and HCC. All patientshad undergone chemotherapy thatfailed. The mean dose infused was0.83 GBq for lobar and whole-livertreatments. The mean absorbed dosewas 42 Gy. Sixty-seven of 88 patientswere treated on an outpatient basis.There were two cases of gastric ulcer-ation and one case of pancreatitis,which were likely a result of nontargetembolization. The mean CEA decreasewas 32% for the entire cohort of pa-tients with colorectal cancer. The im-aging response rate according toRECIST criteria was 18%, whereas thefunctional response rate according toPET was 62%. Mean and median sur-vival times were 285 days and 180days, respectively. The authors con-cluded that SIR-Spheres could be per-formed safely on an outpatient basisand that the device was safe and effec-tive for a variety of metastatic tumorsas evidenced by functional imaging(ie, PET) and tumor markers (ie, CEA)(109).

Cianni et al (110) reported on 29patients with liver metastases that hadshown progression despite multiple-agent chemotherapy treated with SIR-Spheres. Primaries included colon,pancreas, breast, esophagus, andstomach cancers. Patients with biliru-bin levels greater than 1.8 mg/dL andgreater than 20% lung shunting wereexcluded. One patient died of hepaticfailure 30 days after treatment. Twopatients with more than 60% tumorburden died within 4–6 months of mi-crosphere implantation. The investiga-tors obtained a response in all cases byimaging or tumor markers. The au-thors concluded that radioemboliza-

1582 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 13: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

tion was safe and effective for this pre-treated population (110).

Wang et al (111) reported on 34 pa-tients with liver metastases treatedwith SIR-Spheres. The mean activityinfused was 2.2 GBq (range, 0.7–6.1GBq). There were two cases of gastri-tis, two cases of peptic ulcer disease,and one case of radiation cholecystitis.Surprisingly, no patient had coil em-bolization (eg, gastroduodenal artery,right gastric artery) before administra-tion of SIR-Spheres. Twenty-two pa-tients showed complete initial re-sponse, four showed partial initialresponse, and eight showed diseaseprogression. Average times to pro-gression were 8 months and 4 monthsin responders and nonresponders, re-spectively. Survival times were 16months and 7 months in respondersand nonresponders, respectively. Theauthors concluded that this therapywas a safe and effective therapy, withmost patients showing no complica-tions (111).

King et al (112) described their ex-perience in 34 patients with metastaticneuroendocrine tumors treated withSIRT and systemic 5-FU. Patients re-ceived whole-liver infusion, and clini-cal toxicities and imaging follow-upwere obtained. The median follow-uptime was 9.8 months. Complicationsincluded three ulcers, one case of pan-creatitis, and two cases of self-limitingjaundice. All patients had abdominalpain, fatigue, and lethargy for 4 weeks.Of 30 patients whose conditions wereevaluable at 3-month follow-up, therewere six partial responders (20%), 19with stable disease (63%), and fivewith progressive disease (17%). Chro-mogranin A levels decreased in 60% ofpatients. The authors concluded thatSIRT appears to be safe and effectivefor neuroendocrine tumors (112).

Coldwell et al (113) described 23patients with nodular cholangiocarci-noma treated with SIR-Spheres. Allpatients had received two chemother-apy regimens that had failed. The me-dian age was 47 years; 18 patients re-ceived one treatment, and five patientshad two treatments each. The meanactivity infused was 1.5 GBq, with amean tumor dose of 150 Gy. Meanfollow-up time was 14 months. Me-dian survival was not reached, with 19of 23 patients alive at the time of anal-ysis. Response rates by PET and cross-sectional imaging were 90% and 45%,

respectively. The authors concludedthat the lack of response to systemicchemotherapy, the slow growth of thistumor type, and their hypervascularnature make these tumors ideal for lo-coregional therapy (113).

Studies combining Therasphere andSIR-Spheres.—Lewandowski et al (8)presented findings of a novel tech-nique (“radiation segmentectomy”)with use of intraarterially infused 90Yto deliver tumoricidal doses rangingfrom 163 Gy to 4,993 Gy with mini-mal radiation exposure to normal tis-sue. Eighteen patients with a medianage of 65 years (range, 41–80 y) un-derwent TheraSphere or SIR-Spherestreatment for unresectable hepatoma(n � 12) or metastatic liver disease (n� 6). Patients presented with goodperformance status (ECOG stage 0/1)and normal liver function test resultsat baseline. All patients received se-lective infusion of 90Y microspheresto segments within the right (n �13), left (n � 4), or both (n � 1)lobes. The median segmental masstreated was 231.8 g (range,108–1,236), median activity of 90Y in-fused was 1.4 GBq (range, 0.8–3),median absorbed tumor dose was925.9 Gy (range, 163–4,993), and me-dian dose to normal segmental pa-renchyma was 14.1 Gy (range,0.2–124). Other than reported symp-toms of fatigue at 7–10 days aftertreatment, there were no clinical ad-verse effects. Liver function test re-sults remained within normal limitsfor all patients for a median fol-low-up time of 60 days (range,20–200). No patient had signs orsymptoms of radiation hepatitis dur-ing the follow-up period. The au-thors concluded that tumoricidal ra-diation doses approaching 5,000 Gycould be achieved to portions of theliver with virtually no adverse clini-cal events. Radiation segmentectomyusing selective and nonembolic infu-sion of 90Y microspheres to targetedhepatic segments appears safe withminimal toxicity in a patient popula-tion whose disease is refractory toother forms of treatment. This ap-proach permits the safe delivery ofextremely high doses of radiationthat cannot be achieved with use oftraditional external-beam therapy (8).

Kennedy et al (114) described a co-hort of 243 patients with unresectablecolorectal cancer who were treated

with TheraSphere or SIR-Spheres. Allpatients had received first- and sec-ond-line chemotherapy that hadfailed, including irinotecan and/or ox-aliplatin. Patients were treated with-out concurrent chemotherapy in a sin-gle outpatient treatment to the wholeliver or on a lobar basis. The targetdose was 60–150 Gy to the tumor andless than 30 Gy to the normal paren-chyma. Follow-up regarding toxicities,response, progression, and survivalwas available for all patients: 243 pa-tients were reviewed (127 treated withTheraSphere, 116 treated with SIR-Spheres), with follow-up times rang-ing from 3 to 34 months. Patients inthe TheraSphere group received dosesof 36–150 Gy (median, 147 Gy; 171infusions). TheraSphere objective re-sponse rates at 3 months were 35% byCT, 70% by CEA, and 90% by PET,with median survival times of 11.0months versus 5.8 months historically.Patients in the SIR-Spheres group re-ceived 40–85 Gy over 140 infusions.Response rates were similar to thosewith TheraSphere, but the median sur-vival had not yet been reached by thetime of publication. The most commongrade 3 toxicities were nausea, pain,and fatigue in 30% of cases. Therewere no reports of radiation hepatitis.Repeat treatment of 18 patients wassuccessful 8–33 months after initialmicrosphere therapy. The authors con-cluded that 90Y microsphere treatmentfor salvage therapy of colorectal livermetastases is encouraging regardingresponse rate, safety profile, and pos-sible improvement in survival (114).

Kennedy et al (16) also presented acohort of 40 patients (median age, 56years) with neuroendocrine tumorswho underwent 64 treatments withTheraSphere or SIR-Spheres. The me-dian number of treatments per personwas 2 (range, 1–4), and treatmentswere administered on a lobar (n � 31)and whole-liver (n � 9) basis. Four-teen patients received TheraSphere,with a median dose of 143 Gy (range,131–281 Gy), whereas 26 patients re-ceived SIR-Spheres, with a median ac-tivity of 36.69 mCi (range, 9–64.2mCi). Follow-up times ranged from 2months to 48 months. At the time ofpresentation, 33 of 40 patients werestill alive. There were three completeresponses, three cases of stable dis-eases, and 34 partial responses on se-rial CT and OctreoScan. There were no

Salem and Thurston • 1583Volume 17 Number 10

Page 14: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

grade 3/4 toxicities, no cases of radia-tion hepatitis, and no life-threateningevents. Five patients (13%) were ableto discontinue palliative octreotidetherapy because of regression of para-neoplastic carcinoid symptoms aftermicrosphere therapy. Only three pa-tients had not had previous chemo-therapy or liver-directed procedures.The authors concluded that intraarte-rial brachytherapy is an effective treat-ment for liver-dominant neuroendo-crine disease, with an excellent safetyprofile, wide patient applicability,consistent palliative results, and sig-nificant debulking (16).

Wong et al (115) recently describeda study to assess the safety and tumorresponse of intraarterial TheraSphereand SIR-Spheres for the treatment ofsurgically unresectable and chemo-therapy-refractory liver metastases.Forty-six patients with metastatic can-cer to the liver from various solid tu-mors with tumor progression despitemultiple-agent chemotherapy were in-cluded. All patients had baseline CT,PET, hepatic angiography, and in-traarterial 99mTc macroaggregated al-bumin scan for assessment of extrahe-patic aberrant perfusion and lungshunting fraction. Treatment on a lo-bar basis was undertaken in 27 pa-tients with glass microspheres and in19 patients with resin microspheres.Patients were monitored over a periodof 3 months after the last treatmentwith use of dedicated attenuation-cor-rected PET. The results demonstrateda significant decrease in total liverstandardized uptake value after treat-ment by glass or resin-based micro-spheres (P � .027 and P � .014, respec-tively). There was no significantdifference in the amplitudes of themean percentage reduction of tumormetabolism between these two agents(20% � 25% vs 10% � 30% for glass vsresin-based microspheres; P � .38).None of the patients in the glass-basedmicrosphere group experienced com-plications, whereas three patients hadcomplications related to hyperbiliru-binemia (one transient and two per-manent) in the resin-based micro-sphere group. The authors concludedthat there is significant mean reduc-tion of hepatic metastatic tumor load(ie, metabolism) as evaluated objec-tively by PET after radioembolizationfor the treatment of unresectable met-astatic disease to the liver. Radioem-

bolization represents a safe and effec-tive therapy for metastatic cancer tothe liver by arresting disease progres-sion and decreasing standardized up-take value (115).

Two review articles and two bookchapters have recently been publishedon the topic of 90Y therapy for thetreatment of hepatic malignancies(116–119). The authors discussed therole of 90Y in patients with HCC andmetastatic disease to the liver and re-viewed the available literature. In ad-dition, a discussion of the future direc-tion of 90Y was presented, includingquestions that remain on timing andoptimal dose for 90Y, the ideal mode ofadministration (percutaneous vs intra-hepatic ports), factors that might pre-dict successful outcomes, and the com-parison of 90Y therapy with otherregional therapies (116–119). Thesewill continue to form the basis of fur-ther investigations.

FUTURE DIRECTION

Potential Role of 90Y in Treatmentof Liver-dominant Breast andColorectal Metastases

There have been significant ad-vances in the treatment of primarybreast and colorectal carcinoma withthe thymidylate synthase inhibitors(eg, 5-FU, floxuridine, capecitabine)that are known to be radiation sensi-tizers (120). The typical treatmentparadigm involves resection of the pri-mary tumor followed by chemother-apy (in the case of positive lymphnodes) or radiation (in the case of pos-itive margins) for breast carcinoma orradiation of the surgical bed in thecase of local recurrence of colorectalmalignancy. The most typical sites ofdistant metastases after breast cancerresection include the liver, bone, andbrain, and those for colorectal cancerinclude the liver and lungs. The natu-ral course of disease appears to beimproved and altered by improvedchemotherapy agents, with liver me-tastases representing the life-threaten-ing condition.

Historical approaches to treatmentof liver metastases have included ad-ministration of thymidylate synthaseinhibitors in an attempt to delay fur-ther disease progression. Stabilizationof disease has shown durations of 6–8months; however, inevitably, all pa-

tients experience progression whenthese regimens are used (121). Prelim-inary data in the application of 90Ymicrospheres for liver-dominant colo-rectal and breast metastases are veryencouraging according to RECIST andFDG-PET imaging criteria (12,17–20).Given the low toxicity profile for 90Yrelative to the standard of care chemo-therapy regimens, patients tolerate thetherapy exceedingly well. Other thanslight fatigue and flulike symptomsfor approximately 2 weeks after treat-ment, there are virtually no clinicaltoxicities.

As mentioned previously, the sys-temic chemotherapy regimens in pa-tients eligible for 90Y treatment whopresent with liver-dominant meta-static disease are currently suspendedat least 30 days before 90Y therapy.This is because of the known radia-tion-sensitizing properties of the thy-midylate synthetase inhibitors and thepotential for radiation-induced livertoxicity caused by the synergistic ef-fect of the agent combined with 90Ytherapy. However, it is this synergismthat may present some interesting op-portunities to further extend the ther-apeutic benefit of 90Y therapy in com-bination with radiation-sensitizingagents. There is precedent for this ap-proach in the treatment of breast, pan-creatic, colon, liver, and other malig-nancies. The principle of enhancedtumor response and time to diseaseprogression with use of a combinationof radiation-sensitizing agents and ex-ternal-beam ionizing radiation is wellestablished (122). Clearly, additionalassessment of toxicity, maximum tol-erated dose (ie, dose escalation), andoptimal therapeutic window will needto be explored by carefully controlledstudies before widespread clinical ap-plication is considered.

Other studies to be consideredinclude the combination of liver-directed therapy, such as 90Y, withcurrent standards of care, includingbevacizumab and cetuximab. Al-though growth factor inhibitors mayindeed play a role in controlling tumorgrowth, the vascular effects may affectthe applicability of transarterial thera-pies. Growth factor inhibitors (eg, be-vacizumab, cetuximab) may inhibitthe ability to treat liver tumors by del-eteriously affecting the vasculature.Studies combining chemotherapyagents (ie, FOLFOX [5-FU/leucovor-

1584 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 15: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

in/oxaliplatin], FOLFIRI [5-FU/leuco-vorin/irinotecan]) with and without90Y (first- or second-line treatment)may show promise in further extend-ing survival, which is currently re-ported as longer than 20 months (120).

Potential Role of 90Y in Treatmentof Neuroendocrine Liver Metastases

Neuroendocrine tumors typicallyarise in the endocrine glands and cellsthat are located throughout the body.The most common types include car-cinoid, gastroenteropancreatic (GEP),and adrenal. The most common pri-mary sites for carcinoid tumors in-clude the gastrointestinal tract or thelungs; they may also arise in the pan-creas. Other neuroendocrine tumorsmay originate from anywhere in thebody. GEP tumors that most com-monly originate in the cells of thestomach, intestines, or pancreas maybe further categorized into gastrino-mas, glucagonomas, insulinomas, va-soactive intestinal polypeptide tu-mors, and somatostatinomas. There isalso a class of multiple endocrine neo-plasia syndromes that are usually ge-netically based, and their presence in-creases the likelihood of developmentof GEP. Carcinoid tumors, particularlyif they metastasize to the liver, mayresult in the excessive production ofserotonin and produce carcinoid syn-drome. This condition is characterizedby diarrhea, flushing of the skin, andasthma-like breathing difficulties.Loss of appetite and weight loss mayalso result.

The intent of neuroendocrine treat-ments is to reduce tumor burden,eliminate or reduce the symptoms re-sulting from excessive production ofhormones when necessary, and pro-long life. Classic treatments for neu-roendocrine tumors that have metas-tasized to the liver include surgery (ie,resection) or transplantation and theadministration of somatostatin ana-logues such as octreotide or lan-reotide. Other approaches include ab-lative techniques (ie, cryoablation andRF ablation), bland embolization,TACE, and, more recently, 90Y micro-sphere therapy.

The primary goals of surgical inter-vention in neuroendocrine disease arecomplete removal of the tumor, reduc-tion of bulk disease, and elimination ofcarcinoid syndrome when present.

Liver transplantation has been at-tempted with mixed results. Five-yearsurvival rates were reported at 36%and 47% in two transplant studies(123,124). Reported recurrence-freesurvival rates at 5 years were less than24% for 103 patients studied (124).Surgical intervention, including localresection and more radical hepatec-tomy, has resulted in the eliminationof carcinoid symptoms, with 5-yearsurvival rates of 60%–80% (125–127).The samples reported in the majorityof surgical trials are low (n � 50), withno prospective randomization versusuntreated control individuals; there-fore, any survival data are to be inter-preted with caution. Moreover, pa-tients amenable to surgical resectionare most likely self-selected for lesstumor burden than that experiencedby those with more advanced stages ofthe disease. Despite initial apparentcurative effect, recurrence rates of50%–60% within 4–5 years have beenreported (128). Two studies involvinglarger cohorts (120 and 144 patients)reported palliation or control of carci-noid symptoms, with survival rates of61% at 5 years in one study and 91%and 68% at 3 years and 6 years, respec-tively, in the other (129,130).

Less invasive alternatives to sur-gery, such as RF ablation and cryoab-lation, have also been used in cases ofunresectable disease, with limited suc-cess. These treatments involve the ad-ministration of extreme heat (ie, RFablation) or cold (ie, cryoablation) toachieve tumor kill. In one study of 34patients, immediate symptom reliefwas achieved in 95% of patients, withcontrol in 80% for a mean of 10 months(131). New liver lesions appeared in28% of cases, with local liver recur-rence in 13%. Mean survival from di-agnosis was 5.5 � 0.8 years; survivalfrom treatment was 1.6 � 0.2 years.Cryoablation studies have reportedmedian survival times of 13–24months, with mostly asymptomatic re-sults after treatment (132,133). Al-though RF ablation and cryoablationresult in some reduction of tumor bur-den and symptom relief, disease recur-rence and return of symptoms is inev-itable.

Transarterial embolization andTACE involve the intrahepatic admin-istration of particles (300–500 �m) or acombination of particles and cytotoxicagents to compromise blood supply to

the tumor and/or deliver a tumori-cidal dose. Modest tumor responserates (ie, reduction or stabilization)ranging from 40% to 60% have beenreported, with 5-year survival rates of50%–65% (134,135). Tumor type af-fects tumor response and survival.Transarterial embolization and TACEwere used to treat metastatic carcinoidtumors (n � 69) and pancreatic islet-cell cancers (n � 54). Clinical outcomewas superior for carcinoid tumors ver-sus pancreatic tumors with respect totumor response (66.7% vs 35.3%), pro-gression-free survival (22.7 months vs16.1 months), and survival (overallsurvival, 33.8 months vs 23.2 months).TACE was also found to be superior totransarterial embolization in terms oftumor response (50% vs 25%) and sur-vival (31.5 months vs 18.2 months)(134).

The classic treatment of GEP tu-mors has involved the administrationof somatostatin analogues such asoctreotide and lanreotide to controltumor growth and the symptoms ofexcessive hormonal production. Dela-noit et al (136) conducted a compre-hensive review of somatostatin ana-logues in the treatment of GEP tumorsfrom 1970 to 2003. Median objectivetumor response, stable disease, bio-chemical response, and symptomaticresponse rates among patients treatedwith octreotide were 4.5% (range,0–28.6%), 40.0% (range, 21.4%–87.5%),72.5% (range, 28.6%–81%) and 71.0%(range, 58%–100%), respectively. Cor-responding values for the longer-act-ing lanreotide were 5.0% (range,0–8%), 70.0% (range, 37%–87.5%),42.0% (range, 18%–62.5%) and 50.0%(range, 30%–100%), respectively.Therefore, although biochemical andsymptomatic responses were favor-able, there was no evidence of objec-tive tumor response for either agent(4.5% and 5.0%, respectively). The au-thors noted a scarcity of data regard-ing survival benefit of the somatosta-tin analogues in GEP and the fact thatthere were no prospective randomizedtrials of patients with GEP treatedwith these agents (136). The lack oftumor response to somatostatin ana-logues has been shown in other stud-ies, with tumor responses of 3%–10%,in contrast to high biochemical (range,53%–78%; mean, 77%) and symptomresponse rates (73%) (137,138). Otheragents such as interferon with or with-

Salem and Thurston • 1585Volume 17 Number 10

Page 16: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

out octreotide have failed to improvetumor response (range, 0–11%), evenin a large cohort of patients receivingthe therapy (n �350) (139). Althoughthe somatostatin analogues providepalliative relief from the symptoms ofhormone overproduction, their benefitin tumor reduction is negligible. Thisphenomenon is further supported by athorough review of the literature from1966 to 2000 regarding somatostatinanalogues in the treatment of numer-ous cancer types. The authors attrib-uted the lack of consistent results re-garding objective tumor regression toseveral factors, including nonrandom-ized trials in patients with highly dis-seminated disease who had been pre-treated, bias in reporting positiveresults, and unreliability of anecdotaldata. However, they did suggest thatthe concept of receptor radiation ther-apy, particularly indium In 111 oct-reotide or 90Y DOTA lanreotide,showed promising results in prelimi-nary studies with somatostatin recep-tor–positive malignancies.

Although published data on thetreatment of neuroendocrine tumorswith 90Y microspheres are limited,preliminary results based on its use inclinical practice are encouraging. Thetherapy showed an excellent safetyprofile, consistent palliative results,and significant debulking. The major-ity of patients showed a partial re-sponse on imaging follow-up (13,16).Therefore, on the basis of this earlywork, 90Y microspheres appear to pro-vide a low-toxicity alternative withsome clinical benefit as measured bytumor response and symptom reduc-tion.

Potential Clinical Investigations andExtrahepatic Applications

On the basis of the extensive clini-cal data for 90Y microspheres accumu-lated to date, it is clear that the modal-ity offers an attractive alternative inthe available therapies for liver dis-ease. Given careful patient selectionand proper angiographic technique,the therapy offers patients a minimallyinvasive, low-toxicity treatment withvery favorable tumor response andpotential survival benefit, even in thecontext of extrahepatic disease (10,18,140).

Given the encouraging safety andtherapeutic benefit of 90Y in primary

and metastatic liver disease, there is anopportunity to explore its applicationin combination with other availabletherapies. Studies to assess the poten-tial synergistic therapeutic benefit of90Y and known radiation sensitizers inmetastatic breast and colorectal cancerare warranted. Combinatorial capecit-abine and 90Y presents a low-toxicityoption for breast cancer. The potentialto improve hepatic tumor response viathe synergistic action of selective up-take of 5-FU in the presence of radia-tion warrants further investigation.90Y in combination with 5-FU, floxuri-dine, and capecitabine in colorectalmetastases to the liver requires furtherstudy. Given the potential for “superirradiation” of liver parenchyma in thepresence of these agents, carefullycontrolled phase I dose-escalationstudies, particularly in breast cancermetastases, are required.

Combinatorial applications of 90Yand ablative techniques such as RFand cryoablation may provide an op-tion for patients who would otherwiserequire surgical resection but who areat high surgical risk as a result of co-morbidities or prefer less invasivemeans of treatment for their disease.90Y has been shown to reduce tumorburden in downstaging to permittransplantation or resection for HCC(5–7,69). In patients with tumors thatare not amenable to ablative therapybecause of excessive size (6–8 cm), 90Ymicrospheres could be used to reducethese lesions (�3 cm), followed by theuse of ablative therapy to effect furtherresponse. This presumes that viabletissue is still present after 90Y therapy.There is strong evidence that 90Y treat-ment typically results in complete ne-crosis of the lesion(s) treated (5,141). Aprospective randomized study com-paring time to progression or progres-sion-free survival of patients treatedwith 90Y versus RF ablation in lesions3–5 cm in size would provide impor-tant information concerning this ques-tion.

Combinatorial 90Y and TACE pre-sents an interesting opportunity to as-sess the cumulative effect of these mo-dalities in effecting tumor kill.Intraarterial infusion of 90Y in an aer-obic environment (ie, nonstasis) couldbe followed by TACE after the radio-active effect diminishes to subthera-peutic levels (approximately 2 weeksafter 90Y treatment). TACE administra-

tion with cytotoxic tumor exposure ina hypoxic environment would thenaddress any viable (ie, radioresistant)cells that remained.

The recently published AmericanAssociation for the Study of Liver Dis-ease practice guideline for the man-agement of HCC has identified a sub-group of patients within the BarcelonaClinic Liver Cancer staging systemwho present with advanced-stage dis-ease characterized by clinical symp-toms and vascular invasion, includingportal vein thrombosis (142). The chal-lenge in treating patients with portalvein thrombosis with use of embolic-type therapy is the risk of liver failureresulting from compromise of portaland hepatic flow. Preliminary studieshave shown that 90Y can be safely ad-ministered in patients with portal veinthrombosis if microsphere infusion isnot pursued to the endpoint of angio-graphic stasis of the tumor bed (61).Moreover, in the case of single portalbranch occlusion in the absence ofmain occlusion, the use of embolicagents may be considered if patientsare selected carefully and hepatopetalflow is present. This presents an op-portunity to assess the relative safetyand efficacy of 90Y therapy versus anumber of agents in a prospective ran-domized paradigm. This could in-clude, but not be limited to, 90Y versusTACE, 90Y versus bland embolization,90Y versus drug-eluting beads, 90Y ver-sus other radioactive spheres, and 90Yversus best supportive care. Althoughinvestigators have suggested that 90Yeffects on tumor are the result of radi-ation rather than macroembolic effect,a study of 90Y therapy versus bland 20-to 40-�m microsphere treatmentwould be of interest (143).

Another possible area of investiga-tion includes the prophylactic radio-embolization of remnant liver tissue inpatients undergoing hepatic resectionfor HCC or colorectal metastases. Al-though initially attractive, this approachmay hinder and limit the ability for fu-ture 90Y to the prophylactically radio-embolized lobe. In addition, the bloodsupply to small metastases is derivedfrom the portal vein, not the hepaticartery, bringing into question whetherprophylactic treatment would yieldany radiation effect to microscopic me-tastases (144). Therefore, if such astudy is undertaken, improved sur-vival or decreased time to disease re-

1586 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 17: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

currence would represent the end-points, because imaging of lesions isnot feasible. It is clear that further re-search is needed to address possibletreatment options for advanced-stageHCC. Any studies in this patient pop-ulation require careful considerationof the risk of therapy-induced liverfailure versus the benefit of lesion sta-bilization.

Given the tremendous hypervascu-larity of neuroendocrine hepatic tu-mors relative to normal parenchyma,90Y represents an ideal therapy for pal-liative reduction of bulk disease andcarcinoid syndrome when this condi-tion is present. Limited studies with90Y in neuroendocrine disease sug-gest that these benefits may result(11,13,16,80). Randomized controlledstudies of 90Y versus current standardof care (ie, surgical resection or soma-tostatin analogues) are warranted.

One of the most fruitful areas forfurther research is the application of90Y in the downstaging of disease forliver transplantation. One study hasshown that patients with stage T3HCC lesions can have their diseasedownstaged to T2 (on Milan criteria)in more than 50% of cases (5). Largerprospective randomized controlledstudies comparing 90Y with otheravailable therapies (eg, RF ablationand TACE) are also warranted. Giventhe increasing incidence of hepatitis Cin the United States, the incidence ofHCC in this country is expected toincrease sharply in the next severalyears (145). Given earlier detection ofHCC, with the intent of curative livertransplantation, there will be a signif-icant need for alternative downstagingoptions. The ultimate question that re-mains unanswered is whether sur-vival would be significantly extendedfor patients with downstaged diseasetreated with transplantation comparedwith patients who initially presentwith stage T2 disease. Only large pro-spective randomized controlled stud-ies of well-established therapies willbegin to provide the answers to thisquestion.

Improved dosimetry planning(translating into enhanced efficacy) for90Y should be the focus of research inthe next few years (9,146). This couldcome in the form of enhanced 99mTcmacroaggregated albumin particles orin the development of planning resinor glass microspheres. This would al-

low the direct calculation of tumorand normal parenchymal exposure af-ter 90Y treatment. Direct knowledge ofthis would allow administration ofsufficient 90Y radioactivity to result intumoricidal doses with minimal expo-sure to normal parenchyma. Thismight then evolve into the ability toperform repeated and scheduled 90Ytreatments in a manner analogous toTACE, but without the associatedrisks of radiation hepatitis. The abilityto image the microspheres (cold orhot) with imaging techniques such asMR imaging would also prove clini-cally useful in the dose-planning andclinical follow-up stages.

One final area worthy of future in-vestigation is in the treatment of non-colorectal, nonneuroendocrine cancersmetastatic to the liver. Often referredto as mixed neoplasia, these terms re-fer to liver-dominant metastatic dis-ease to the liver from various primarytumors (eg, breast, melanoma, pan-creas, and lung). Although several re-ports have been described, controlledphase II studies that have studied timeto progression, tumor response, orprogression-free survival outcomeswould be clinically relevant, given thedearth of options for some of thesepatients (11,15,16,80).

Survival is the gold standard bywhich the efficacy of any liver-di-rected therapy would ideally be mea-sured. However, the resource costsand lengthy studies required toachieve this endpoint in a constantlyevolving state-of-the-art paradigmmake this a daunting task. There isclearly a need to develop surrogatemeasures of efficacy pending the re-finement of technology, treatmenttechniques, and operator expertise sothe treatment effect in randomizedstudies is validated to the satisfactionof scientists, statisticians, and clini-cians. Tumor response provides aquantitative measure that has beenhistorically touted as a viable alterna-tive; however, correlative studies link-ing tumor response to extended sur-vival have not been forthcoming.Quality of life has gained increasingawareness as a potentially useful mea-sure of treatment efficacy. For patientswho undergo systemic chemotherapyand external radiation as part of thetreatment regimen, quality of life withrespect to time commitment and sideeffects certainly affect their quality of

life during the finite time they haveavailable. The incorporation of care-fully designed and validated quality-of-life measures in prospective ran-domized controlled trials of 90Y isanother fruitful area of investigation.Time to progression and progression-free survival represent alternativeendpoints that may balance the needfor clinically relevant data with theoptions of patients who have diseaseprogression despite therapy to tryother experimental agents. Finally, al-though new and evolving treatmentsmight ideally be compared with a no-treatment control arm, the ethical andmoral issues in an increasingly awarepatient base make the likelihood ofthis quite low.

The potential applications of 90Ytherapy to areas outside the liver arenumerous. Any site in the body that isangiographically accessible may po-tentially be considered for 90Y therapy,potentially permitting treatment ofmeningiomas, glioblastoma multi-forme, renal-cell carcinomas, head/neck tumors, and lung tumors. Giventhat skin necrosis after 90Y has beendescribed, the cutaneous contributionof the extrahepatic vasculature beingexamined would have to be taken intoconsideration (28). Clearly, very care-fully controlled in vitro, animal, andphase I safety studies with small pa-tient cohorts are required before wide-spread clinical applications in these ar-eas are explored.

SUMMARY

As with any oncologic therapy, thedisease presentation, patient perfor-mance status, and functional liver re-serve must be taken into account whena patient is considered for 90Y micro-sphere therapy. In patients with pri-mary HCC, treatment planning is of-ten complicated by the trade-offbetween exposing functional (but cir-rhotic) liver parenchyma to radiationwhile maximizing radiation exposureto malignant tissue to achieve com-plete kill. Although patients withliver-dominant metastatic diseasepresent with good liver function, theyhave typically undergone multiplesystemic chemotherapy regimens thatmay have compromised hepatic re-serve and vascular supply. After thedecision to treat with 90Y therapy ismade, the clinical outcome relative to

Salem and Thurston • 1587Volume 17 Number 10

Page 18: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

patient risk and benefit is the para-mount consideration.

On the basis of the currently avail-able data for 90Y therapy in the treat-ment of liver disease, it is clear that themodality provides a favorable risk/benefit profile for the patient. Numer-ous studies have demonstrated that90Y therapy can effectively reduce orstabilize primary and metastatic liverdisease with a minimal toxicity profile.The early clinical results indicated that90Y could be safely delivered to thetumor bed while minimizing insult tonormal liver parenchyma. Subsequentstudies have shown that, with appro-priate patient selection, lobar adminis-tration, and careful attention to theembolization of collateral gastrointes-tinal vessels, complications of radia-tion to gastrointestinal structures areminimized. Despite these findings andthe consensus among investigatorssupporting lobar infusions and pro-phylactic embolization of the gas-troduodenal artery and right gastricarteries, practitioners continue to re-sist this approach and incorrectly as-sume that superselective infusion ofmicrospheres completely eliminatesthe risk of nontarget administration.Hence, avoidable complications of ul-ceration and pancreatitis continue tobe reported (98,111,147–149). Thesespecialized interventional techniqueshave also minimized the risk of treat-ment-emergent liver failure caused byexcessive exposure of functional liverreserve. Screening patients for poten-tial pulmonary shunt via 99mTc macro-aggregated albumin scanning has vir-tually eliminated the risk of radiationpneumonitis. Refinements in angio-graphic and imaging technology havefurther improved the safety and effi-cacy of 90Y therapy in selectively di-recting microspheres to the targetarea. With improvements in catheterdelivery technology, more selectivedelivery to the tumor bed (ie, radiationsegmentectomy) has enabled the treat-ment of patients with increased biliru-bin levels who otherwise would havebeen at risk of liver failure as a resultof excessive radiation exposure offunctional reserve.

Contemporary research in 90Y ther-apy is extending its application inHCC to new areas including down-staging to permit orthotopic livertransplantation, resection, and RF ab-lation. It is common that, at some

point in the cancer disease process, pa-tients present with life-threatening liv-er-dominant disease. Preliminarystudies of 90Y administration as mono-therapy in patients with liver-domi-nant metastatic disease in whom stan-dard-of-care systemic chemotherapyhad failed are encouraging. Reduc-tion and stabilization of advancinghepatic tumors provide the patientswith a “chemotherapy break,” stabi-lize performance status, and in manycases permit patients to proceed toundergo other systemic or regionallydirected therapies. This introducesthe concept of the use of 90Y as a“bridge” to chemotherapy.

However, there are continuingchallenges and opportunities in the90Y treatment of liver disease. Severalphase I studies of 90Y microspheres incombination with infusional chemo-therapy regimens in metastatic colo-rectal cancer to the liver have demon-strated very effective tumor responseand stabilization of disease. However,the use of these regimens must be tem-pered against the neurologic, hemato-logic, and gastrointestinal toxicitiesthat result. Whether the synergisticeffect of 90Y and radiation sensitizersoutweighs the risk of increased tox-icity remains to be determined.Clearly, the final determination ofsafety and efficacy will await thecompletion of randomized con-trolled trials comparing combinato-rial 90Y and chemotherapy versuschemotherapy alone.

Expanding 90Y therapy to the treat-ment of other liver metastases such asbreast and neuroendocrine tumors isalso being explored. Preliminary stud-ies indicate that 90Y therapy can effec-tively reduce or stabilize disease andvirtually eliminate the symptoms as-sociated with carcinoid syndrome inthe case of functional neuroendocrinetumors. Finally, the potential applica-tions of 90Y therapy to extrahepatic ar-eas are numerous. It is clear that themodality provides a versatile tool inthe array of therapeutic possibilitiesavailable to the interventional oncolo-gist.

CONCLUSION

Advancements in liver-directedtherapies continue to be of particularinterest. With the advent of sophisti-cated improvements in techniques of

TACE such as drug-eluting micro-spheres, other technologies have be-gun to emerge. Administration of 90Ymicrospheres represents a technologythat continues to expand into the treat-ment of patients with liver malignan-cies. More than 100 publications, 50posters, and innumerable presenta-tions on 90Y have been published orpresented to scientific and clinical au-diences. Worldwide, this therapy isused in more than 80 centers, withseveral interested in adding 90Y treat-ment to their offerings. This com-prehensive literature review clearlysupports the conclusion that the ad-ministration of 90Y microspheres issafe and effective for the treatment ofliver tumors.

Acknowledgments: The authors thankthe following individuals for their contri-bution to this comprehensive review arti-cle: Beth Oman, Agnieszka Stanisz, KarenBarrett, Krystina Sajdak, Margaret Gilbert-sen, and Vanessa Gates. In particular, theauthors thank Angi Courtney for her con-tribution to this article.

References1. Flynn WJ. The treatment of pulmo-

nary metastases with microspheres ofyttrium 90. Minerva Med 1967;58:4498–4500.

2. Greene WM. Embolization of dogprostates with yttrium-90 micro-spheres. J Urol 1963; 90:451–457.

3. Wollner I, Knutsen C, Smith P, et al.Effects of hepatic arterial yttrium 90glass microspheres in dogs. Cancer1988; 61:1336–1344.

4. Wollner IS, Knutsen CA, Ullrich KA,et al. Effects of hepatic arterial yttri-um-90 microsphere administrationalone and combined with regionalbromodeoxyuridine infusion in dogs.Cancer Res 1987; 47:3285–3290.

5. Kulik L, Atassi B, van Holsbeeck L, etal. Yttrium-90 microspheres (Thera-Sphere) treatment of unresectablehepatocellular carcinoma: downstag-ing to resection, RF ablation andbridge to transplantation. J Surg On-col 2006 (in press).

6. Kulik LM, Mulcahy MF, Hunter RD,et al. Use of yttrium-90 micro-spheres (TheraSphere) in a patientwith unresectable hepatocellular car-cinoma leading to liver transplanta-tion: a case report. Liver Transpl 2005;11:1127–1131.

7. Lau WY, Ho SK, Yu SC, et al. Sal-vage surgery following downstagingof unresectable hepatocellular carci-noma. Ann Surg 2004; 240:299–305.

8. Lewandowski R, Salem R, Thurston

1588 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 19: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

K, et al. Radiation segmentectomyof unresectable of liver carcinoma us-ing selective infusion of intraarterialyttrium-90 TheraSphere. Presented atthe Radiological Society of NorthAmerica; November 28–December 3,2004; Chicago, IL.

9. Rhee TK, Omary RA, Gates V, et al.The effect of catheter-directed CT an-giography on yttrium-90 radioembo-lization treatment of hepatocellularcarcinoma. J Vasc Interv Radiol 2005;16:1085–1091.

10. Salem R, Lewandowski RJ, Atassi B,et al. Treatment of unresectable hep-atocellular carcinoma with use of 90Ymicrospheres (TheraSphere): safety,tumor response, and survival. J VascInterv Radiol 2005; 16:1627–1639.

11. Andrews JC, Walker SC, AckermannRJ, et al. Hepatic radioembolizationwith yttrium-90 containing glass mi-crospheres: preliminary results andclinical follow-up. J Nucl Med 1994;35:1637–1644.

12. Coldwell D, Nutting C, Kennedy AK.Treatment of hepatic metastases frombreast cancer with yttrium-90 SIR-Spheres radioembolization. Presentedat the Society of Interventional Radiol-ogy Annual Meeting; March 31 throughApril 5, 2005; New Orleans, LA.

13. Coldwell D, Nutting C, Kennedy AK.Use of yttrium-90 SirSpheres to treatmetastatic unresectable neuroen-dorcine tumors to the liver. Presentedat the World Congress of Gastrointes-tinal Cancer; June 27–30, 2005; Barce-lona, Spain.

14. Coldwell D, Nutting C, Kennedy AS.Initial clinical results in the treatmentof unresectable hepatic tumors withresin-based yttrium-90 radioemboli-zation. Presented at the Cardiovascu-lar and Interventional Radiology So-ciety of Europe (CIRSE) AnnualMeeting; September 25–29, 2004; Bar-celona, Spain.

15. Herba MJ, Illescas FF, Thirlwell MP,et al. Hepatic malignancies: im-proved treatment with intraarterialY-90. Radiology 1988; 169:311–314.

16. Kennedy A, Coldwell D, Nutting C, etal. Hepatic brachytherapy for GIneuroendocrine tumors with Y90 mi-crospheres: long-term USA results.Presented at International Congresson Anti-cancer Treatment; February1–4, 2005; Paris.

17. Rubin D, Nutting C, Jones B. Meta-static breast cancer in a 54-year-oldwoman: integrative treatment with yt-trium-90 radioembolization. IntegrCancer Ther 2004; 3:262–267.

18. Van Hazel G, Blackwell A, AndersonJ, et al. Randomised phase 2 trial ofSIR-Spheres plus fluorouracil/leucov-orin chemotherapy versus fluoroura-

cil/leucovorin chemotherapy alone inadvanced colorectal cancer. J SurgOncol 2004; 88:78–85.

19. Van Hazel G, Price D, Bower G, et al.Selective internal radiation therapy(SIRT) plus systemic chemotherapywith FOLFOX4: a phase I dose escala-tion study. Presented at the AmericanSociety of Clinical Oncology Gastro-intestinal Malignancies Symposium;February 27–29, 2005; Hollywood, FL.

20. Van Hazel G, Price D, Bower G, et al.Selective internal radiation therapy(SIRT) for liver metastases with con-comitant systemic oxaliplatin, 5-flu-orouracil and folinic acid: a phase I/IIdose escalation study. Presented atthe American Society of Clinical On-cology Gastrointestinal MalignanciesSymposium; February 27–29, 2005;Hollywood, FL.

21. Ariel I. Intra-arterial injection of ra-dioactive microspheres of ceramic inthe treatment of malignant tumors. in-dications and clinical results. MinervaMed 1965; 56:2030–2037.

22. Ariel IM. Treatment of inoperableprimary pancreatic and liver cancerby the intra-arterial administration ofradioactive isotopes (Y90 radiatingmicrospheres). Ann Surg 1965;162:267–278.

23. Ariel IM, Flynn W, Pack GT. Currentpossibilities of radioactive ceramic mi-crospheres labeled with yttrium 90.Minerva Med 1967; 58:4496–4498.

24. Ariel IM, Padula G. Irradiation ofthe spleen by the intra-arterial admin-istration of 90 yttrium microspheresin patients with malignant lympho-ma: a preliminary report. Cancer1973; 31:90–96.

25. Simon N, Siffert R, Baron MG, et al.Preoperative irradiation of osteogenicsarcoma with intra-arterially injectedyttrium-90 microspheres: case report.Cancer 1968; 21:453–455.

26. Jaffe R, Simon N. Treatment of braintumors with internally deposited yt-trium 90. Neurology 1968; 18:309.

27. Pandey U, Bapat KN, Samuel G, et al.Evaluation of 90Y phosphate particlesas a possible radiation synoviorthosisagent. Nucl Med Commun 2005; 26:459–463.

28. Sojan S, Bartholomeusz D. Cutane-ous radiation necrosis as a complica-tion of yttrium-90 synovectomy. HellJ Nucl Med 2005; 8:58–59.

29. Dunn AL, Manco-Johnson M, BuschMT, et al. Leukemia and P32 radio-nuclide synovectomy for hemophilicarthropathy. J Thromb Haemost 2005;3:1541–1542.

30. Chadzopoulos D, Badiavas K. Com-plications in 90Y radiation synovec-tomy. Hell J Nucl Med 2005; 8:134.

31. Sheppeard H, Aldin A, Ward DJ.

Osmic acid versus yttrium-90 in rheu-matoid synovitis of the knee. ScandJ Rheumatol 1981; 10:234–236.

32. Chamberlain MN, Gray BN, HeggieJC, et al. Hepatic metastases—aphysiological approach to treatment.Br J Surg 1983; 70:596–598.

33. Mantravadi RV, Spigos DG, Tan WS,et al. Intraarterial yttrium 90 in thetreatment of hepatic malignancy. Ra-diology 1982; 142:783–786.

34. Mantravadi RV, Spigos DG, KareshSM, et al. Work in progress: intra-arterial P-32 chromic phosphate forprevention of postoperative liver me-tastases in high-risk colorectal cancerpatients. Radiology 1983; 148:555–559.

35. Hafeli UO, Casillas S, Dietz DW, et al.Hepatic tumor radioembolization in arat model using radioactive rhenium(186Re/188Re) glass microspheres.Int J Radiat Oncol Biol Phys 1999; 44:189–199.

36. Nijsen F, Rook D, Brandt C, et al.Targeting of liver tumour in rats byselective delivery of holmium-166loaded microspheres: a biodistribu-tion study. Eur J Nucl Med 2001; 28:743–749.

37. Vuorela J, Kauppinen T, Sokka T.Distribution of radiation in synovec-tomy of the knee with 166Ho-FHMAusing image fusion. Cancer BiotherRadiopharm 2005; 20:333–337.

38. Yan L, Li L, Chen S. Radioemboliza-tion with 32P-labelled glass micro-spheres for advanced hepatocellularcarcinoma. Zhonghua Wai Ke Za Zhi1996; 34:526–529.

39. Yan L, Li Z, Li L, et al. The relation-ship between effects and radiationdoses of intra-arterial phosphorus-32glass microspheres embolization ther-apy for patients with advanced livercancer. Chung-Hua Wai Ko Tsa ChihChin J Surg 2000; 38:837–840.

40. Keng GH, Sundram FX, Yu SW, et al.Preliminary experience in radionu-clide therapy of hepatocellular carci-noma using hepatic intra-arterial ra-dio-conjugates. Ann Acad MedSingapore 2002; 31:382–386.

41. Kumar A, Bal C, Srivastava DN, et al.Management of multiple intrahepaticrecurrences after radiofrequency abla-tion of hepatocellular carcinoma withrhenium-188-HDD-lipiodol. Eur J Gas-troenterol Hepatol 2006; 18:219–223.

42. Kumar A, Srivastava DN, Bal C.Management of postsurgical recur-rence of hepatocellular carcinomawith rhenium 188-HDD labeled io-dized oil. J Vasc Interv Radiol 2006;17:157–161.

43. Yoon CJ, Chung JW, Park JH, et al.Transcatheter arterial embolizationwith 188Rhenium-HDD-labeled io-

Salem and Thurston • 1589Volume 17 Number 10

Page 20: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

dized oil in rabbit VX2 liver tumor. JVasc Interv Radiol 2004; 15:1121–1128.

44. Lin WY, Tsai SC, Hsieh JF, et al. Ef-fects of 90Y-microspheres on liver tu-mors: comparison of intratumoral in-jection method and intra-arterialinjection method. J Nucl Med 2000;41:1892–1897.

45. Yan ZP, Lin G, Zhao HY, et al. Yt-trium-90 glass microspheres injectedvia the portal vein: an experimentalstudy. Acta Radiol 1993; 34:395–398.

46. Tian JH, Xu BX, Zhang JM, et al. Ul-trasound-guided internal radiother-apy using yttrium-90-glass micro-spheres for liver malignancies. J NuclMed 1996; 37:958–963.

47. Hafeli UO, Sweeney SM, BeresfordBA, et al. Effective targeting of mag-netic radioactive 90Y-microspheres totumor cells by an externally appliedmagnetic field: preliminary in vitroand in vivo results. Nucl Med Biol1995; 22:147–155.

48. Hafeli UO, Sweeney SM, BeresfordBA, et al. Magnetically directedpoly(lactic acid) 90Y-microspheres:novel agents for targeted intracavi-tary radiotherapy. J Biomed MaterRes 1994; 28:901–908.

49. Rudge S, Peterson C, Vessely C, et al.Adsorption and desorption of chemo-therapeutic drugs from a magneti-cally targeted carrier (MTC). J ControlRelease 2001 74: 335–340.

50. Hernandez MC, Knox SJ. Radiobiol-ogy of radioimmunotherapy with 90Yibritumomab tiuxetan (Zevalin). Se-min Oncol 2003 30(6 suppl 17):6–10.

51. Hernandez MC, Knox SJ. Radiobiol-ogy of radioimmunotherapy: target-ing CD20 B-cell antigen in non-Hodgkin’s lymphoma. Int J RadiatOncol Biol Phys 2004; 59:1274–1287.

52. Toegel S, Mien LK, Wadsak W, et al.In vitro evaluation of no carrieradded, carrier added and cross-com-plexed [90Y]-EDTMP provides evi-dence for a novel “foreign carrier the-ory.” Nucl Med Biol 2006; 33:95–99.

53. Araki N, Nagata Y, Fujiwara K, et al.Evaluation of glass microspheres forintra-arterial radiotherapy in animalkidneys. Int J Radiat Oncol Biol Phys2001; 49:459–463.

54. Becker CD, Rosler H, Biasiutti FD, etal. Congestive hypersplenism: treat-ment by means of radioembolizationof the spleen with Y-90. Radiology1995; 195:183–186.

55. Houle S, Yip TK, Shepherd FA, et al.Hepatocellular carcinoma: pilot trialof treatment with Y-90 microspheres.Radiology 1989; 172:857–860.

56. Shepherd FA, Rotstein LE, Houle S, etal. A phase I dose escalation trial ofyttrium-90 microspheres in the treat-

ment of primary hepatocellular carci-noma. Cancer 1992; 70:2250–2254.

57. Dong BW, Liang P, Jing XH. Com-bined treatment of hepatic carcinoma:percutaneous intratumoral injectionof Y-90 glass treatment microsphereswith sonographic guidance. Zhong-hua Yi Xue Za Zhi 1994 74:471–473,517.

58. Yan ZP, Lin G, Zhao HY, et al. Anexperimental study and clinical pilottrials on yttrium-90 glass micro-spheres through the hepatic artery fortreatment of primary liver cancer.Cancer 1993; 72:3210–3215.

59. Dancey JE, Shepherd FA, Paul K, et al.Treatment of nonresectable hepatocel-lular carcinoma with intrahepatic90Y-microspheres. J Nucl Med 2000;41:1673–1681.

60. Carr BI. Hepatic arterial 90Yttriumglass microspheres (Therasphere) forunresectable hepatocellular carcino-ma: interim safety and survival dataon 65 patients. Liver Transpl 2004 10(2suppl 1):S107–110.

61. Salem R, Lewandowski R, Roberts C,et al. Use of Yttrium-90 glass micro-spheres (TheraSphere) for the treat-ment of unresectable hepatocellularcarcinoma in patients with portal veinthrombosis. J Vasc Interv Radiol 2004;15:335–345.

62. Lewandowski R, Alli A, Atassi B, et al.Retraction of portal vein thrombus fol-lowing treatment with Y90 TheraS-phere: a secondary sign of tumor re-sponse. Presented at the Society ofInterventional Radiology Annual Meet-ing 2006; Toronto, ON, Canada.

63. Liu MD, Uaje MB, Al-Ghazi MS, et al.Use of yttrium-90 TheraSphere for thetreatment of unresectable hepatocel-lular carcinoma. Am Surg 2004; 70:947–953.

64. Geschwind JF, Salem R, Carr BI, et al.Yttrium-90 microspheres for the treat-ment of hepatocellular carcinoma.Gastroenterology 2004; 127(5 suppl 1):S194–S205.

65. Goin JE, Salem R, Carr BI, et al.Treatment of unresectable hepatocel-lular carcinoma with intrahepatic yt-trium 90 microspheres: a risk-stratifi-cation analysis. J Vasc Interv Radiol2005; 16:195–203.

66. Goin JE, Salem R, Carr BI, et al.Treatment of unresectable hepatocel-lular carcinoma with intrahepatic yt-trium 90 microspheres: factors associ-ated with liver toxicities. J Vasc IntervRadiol 2005; 16:205–213.

67. Dawson LA. Hepatic arterial yt-trium 90 microspheres: another treat-ment option for hepatocellular carci-noma. J Vasc Interv Radiol 2005; 16:161–164.

68. Atassi B, Lewandowski RJ, Kulik L, et

al. Treatment of unresectable hepa-tocellular carcinoma using intra-arte-rial Y90 (TheraSphere®): long-termfollow-up. Presented at the Society ofInterventional Radiology AnnualMeeting 2006; Toronto, ON, Canada.

69. Kim DY, Kwon DS, Salem R, et al.Successful embolization of hepatocel-luar carcinoma with yttrium-90 glassmicrospheres prior to liver transplan-tation. J Gastrointest Surg 2006; 10:413–416.

70. Pompili M, Mirante VG, Rondinara G,et al. Percutaneous ablation proce-dures in cirrhotic patients with hepa-tocellular carcinoma submitted toliver transplantation: assessment ofefficacy at explant analysis and ofsafety for tumor recurrence. LiverTranspl 2005; 11:1117–1126.

71. Wong LL, Tanaka K, Lau L, et al.Pre-transplant treatment of hepatocel-lular carcinoma: assessment of tumornecrosis in explanted livers. ClinTransplant 2004; 18:227–234.

72. Anderson JH, Goldberg JA, BessentRG, et al. Glass yttrium-90 micro-spheres for patients with colorectalliver metastases. Radiother Oncol1992; 25:137–139.

73. Goin JE, Dancey JE, Hermann GA, etal. Treatment of unresectable meta-static colorectal carcinoma to the liverwith intrahepatic Y-90 microspheres:a dose-ranging study. World J NuclMed 2003; 2:216–225.

74. Wong CY, Salem R, Raman S, et al.Evaluating 90Y-glass microspheretreatment response of unresectablecolorectal liver metastases by[18F]FDG PET: a comparison with CTor MRI. Eur J Nucl Med Mol Imaging2002; 29:815–820.

75. Wong CY, Salem R, Qing F, et al.Metabolic response after intraarterial90Y-glass microsphere treatment forcolorectal liver metastases: compari-son of quantitative and visual analy-ses by 18F-FDG PET. J Nucl Med2004; 45:1892–1897.

76. Lewandowski RJ, Thurston KG, GoinJE, et al. 90Y microsphere (TheraS-phere) treatment for unresectablecolorectal cancer metastases of the liv-er: response to treatment at targeteddoses of 135-150 Gy as measured by[18F]fluorodeoxyglucose positronemission tomography and computedtomographic imaging. J Vasc IntervRadiol 2005; 16:1641–1651.

77. Ariel IM, Padula G. Treatment ofsymptomatic metastatic cancer to theliver from primary colon and rectal can-cer by the intraarterial administrationof chemotherapy and radioactive iso-topes. J Surg Oncol 1978; 10:327–336.

78. Atassi B, Lewandowski R, MulachyM, et al. Treatment with yttrium-90

1590 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 21: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

microspheres of unresectable meta-static colorectal cancer to the liver:safety, treatment response and sur-vival. Presented at the American So-ciety of Clinical Oncology Gastroin-testinal Malignancies Symposium;2006; San Francisco.

79. Blanchard RJ, Morrow IM, SutherlandJB. Treatment of liver tumors withyttrium-90 microspheres alone. CanAssoc Radiol J 1989; 40:206–210.

80. Herba MJ, Thirlwell MP. Radioem-bolization for hepatic metastases. Se-min Oncol 2002; 29:152–159.

81. Lewandowski RJ, Atassi BA, WongCO, et al. Use of yttrium-90 micro-spheres for the treatment of liver neo-plasia: long-term follow-up. Pre-sented at the Annual Meeting of theCardiovascular and Interventional So-ciety of Europe 2006; Rome.

82. Grady ED. Intrahepatic arterial 90-yttrium resin spheres to treat livercancer. Int J Nucl Med Biol 1978;5:253–254.

83. Grady ED. Internal radiation ther-apy of hepatic cancer. Dis Colon Rec-tum 1979; 22:371–375.

84. Grady ED, Auda SP, Cheek WV. Va-soconstrictors to improve localizationof radioactive microspheres in thetreatment of liver cancer. J Med AssocGa 1981; 70:791–795.

85. Grady ED, Nolan TR, Crumbley AJ, etal. Internal radiation therapy of livercancer (90-Yttrium resin spheres in-tra-arterially) to supplement externalradiation therapy, local and systemicchemotherapy. J Med Assoc Ga 1977;66:625–629.

86. Grady ED, McLaren J, Auda SP, et al.Combination of internal radiationtherapy and hyperthermia to treatliver cancer. South Med J 1983; 76:1101–1105.

87. Lau WY, Ho S, Leung TW, et al. Se-lective internal radiation therapy fornonresectable hepatocellular carci-noma with intraarterial infusion of90yttrium microspheres. Int J RadiatOncol Biol Phys 1998; 40:583–592.

88. Lau WY, Ho S, Leung WT, et al. Whatdetermines survival duration in hepa-tocellular carcinoma treated with in-traarterial yttrium-90 microspheres?Hepatogastroenterology 2001;48:338–340.

89. Szeto CC, Wong TY, Leung CB, et al.Selective internal radiation therapy byyttrium-90 microspheres for hepato-cellular carcinoma after renal trans-plantation. Clin Transplant 2001; 15:284–288.

90. Gray BN, Burton MA, Kelleher DK, etal. Selective internal radiation (SIR)therapy for treatment of liver metas-tases: measurement of response rate. JSurg Oncol 1989; 42:192–196.

91. Gray BN, Anderson JE, Burton MA, etal. Regression of liver metastasesfollowing treatment with yttrium-90microspheres. Aust N Z J Surg 1992;62:105–110.

92. Gray B, Van Hazel G, Hope M, et al.Randomized trial of SIR-Spheres pluschemotherapy vs chemotherapy alonefor treating patients with liver metas-tases from primary large bowel can-cer. Ann Oncol 2001; 12:1711–1720.

93. Stubbs RS, Cannan RJ, Mitchell AW.Selective internal radiation therapywith 90yttrium microspheres for ex-tensive colorectal liver metastases. JGastrointest Surg 2001; 5:294–302.

94. Stubbs RS, Cannan RJ, Mitchell AW.Selective internal radiation therapy(SIRT) with 90Yttrium microspheresfor extensive colorectal liver metasta-ses. Hepatogastroenterology 2001; 48:333–337.

95. Geller DA, Carr BI, Amesur N, et al.Responses in Chemo-refractory me-tastases from colon cancer to a singledose of hepatic artery Y90 micro-spheres. Presented at the AmericanSociety of Clinical Oncology AnnualMeeting; 2003; Chicago.

96. Goldstein D, Van Hazel G, PavlakisN, et al. Selective internal radiationtherapy (SIRT) plus systemic chemo-therapy with irinotecan: a phase Idose escalation study. Presented atAmerican Society of Clinical Oncol-ogy Annual Meeting; May 13–17,2005; Orlando, FL.

97. Sharma RA, Van Hazel GA, Blan-shard K, et al. Internal radiationtreatment of liver metastases fromcolorectal cancer with concomitantsystemic radiosensitising chemother-apy. Presented at the World Congresson Gastrointestinal Cancer 2006; Bar-celona, Spain.

98. Murthy R, Xiong H, Nunez R, et al.Yttrium 90 resin microspheres for thetreatment of unresectable colorectalhepatic metastases after failure ofmultiple chemotherapy regimens:preliminary results. J Vasc Interv Ra-diol 2005; 16:937–945.

99. Lim LC, Gibbs P, Yip D, et al. Aprospective evaluation of treatmentwith selective internal radiation ther-apy (SIR-spheres) in patients with un-resectable liver metastases from colo-rectal cancer previously treated with5-FU based chemotherapy. BMJ Can-cer 2005 5: 132.

100. Kennedy AS, Coldwell D, Nutting C,et al. Resin (90)Y-microspherebrachytherapy for unresectable colo-rectal liver metastases: modern USAexperience. Int J Radiat Oncol BiolPhys 2006; 65:412–425.

101. Murthy R, Oh Y, Tam A, et al. Yttri-um-90 microsphere treatment for liver

dominant hepatic metastases fromlung cancer. Presented at the Ameri-can Society of Clinical Oncology An-nual Meeting; 2006; Atlanta.

102. Lim L, Gibbs P, Yip D, et al. Prospec-tive study of treatment with selectiveinternal radiation therapy spheres inpatients with unresectable primary orsecondary hepatic malignancies. InternMed J 2005; 35:222–227.

103. Kennedy AS, W D, McNeillie P, et al.Fractionation, dose selection, and re-sponse of hepatic metastases of neu-roendocrine tumors after Y90 micro-sphere brachytherapy. Presented atthe Annual Meeting of the AmericanBrachytherapy Society; May 10–12,2006; Philadelphia.

104. Popperl G, Helmberger T, Munzing W,et al. Selective internal radiation ther-apy with SIR-Spheres in patients withnonresectable liver tumors. CancerBiother Radiopharm 2005; 20:200–208.

105. Wong CY, Qing F, Savin M, et al.Reduction of metastatic load to liverafter intraarterial hepatic yttrium-90radioembolization as evaluated by[18F]fluorodeoxyglucose positronemission tomographic imaging. J VascInterv Radiol 2005; 16:1101–1106.

106. Yu M, Lewandowski RJ, Wong COW,et al. Yttrium-90 (SIR-Spheres) treat-ment for metastatic cancer to the liver:midterm results. Presented at the Soci-ety of Interventional Radiology AnnualMeeting 2006; Toronto, ON, Canada.

107. Jakobs TF, Hoffmann RT, Schmitz A,et al. Treatment of extensive meta-static liver disease with intrahepaticyttrium 90 microspheres: toxicity as-sessment. Presented at the Society ofInterventional Radiology AnnualMeeting 2006; Toronto, ON, Canada.

108. Jakobs TF, Hoffmann RT, Schmitz A,et al. Selective internal radiationtherapy with yttrium-90 microspheresin metastatic liver disease: clinical re-sults. Presented at the Society of Inter-ventional Radiology Annual Meeting2006; Toronto, ON, Canada.

109. Jakobs T, Hoffman RT, Yu M, et al.Yttrium-90 SIR-Spheres treatment formetastatic cancer to the liver: mid-term results. Presented at the WorldConference on Interventional Oncol-ogy; 2006; Cernobbio, Italy.

110. Cianni R, Saltarelli A, Notarianni E, etal. Radioembolism (yttrium 90) ofhepatic liver metastases: preliminaryexperience. Presented at the WorldConference on Interventional Oncol-ogy; 2006; Cernobbio, Italy.

111. Wang S, Ho YY, Leung F, et al. Out-comes following selective internalwith SIR-Spheres in non-resectablehepatic malignancy. Presented at theWorld Conference on InterventionalOncology; 2006; Cernobbio, Italy.

Salem and Thurston • 1591Volume 17 Number 10

Page 22: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

112. King J, Morris DL, Glenn D, et al.Pilot study of SIR-Spheres in liver me-tastases from neuroendocrine cancer(NETLM). Presented at the Third Eu-ropean Neuroendocrine Tumor Soci-ety (ENTS) Conference 2006. Prague,Czech Republic.

113. Coldwell D, Kennedy AK. Treatmentof nodular cholangiocarcinoma with yt-trium-90 microspheres. Presented at theWorld Congress on GastrointestinalCancer 2006; Barcelona, Spain.

114. Kennedy A, Coldwell D, Nutting C, etal. Liver brachytherapy for unre-sectable colorectal metastases: US re-sults 2000-2004. Presented at theAmerican Society of Clinical Oncol-ogy Gastrointestinal MalignanciesSymposium; 2005. Hollywood, FL.

115. Wong CO, Savin M, Sherpa KM, et al.Regional yttrium-90 microspheretreatment of surgically unresectableand chemotherapy refractory meta-static liver carcinoma. Cancer BiotherRadiopharm 2006 (in press).

116. Garrean S, Espat NJ. Yttrium-90 inter-nal radiation therapy for hepatic malig-nancy. Surg Oncol 2006 (in press).

117. Lewandowski RJ, Salem R. Yttri-um-90 radioembolization of hepato-cellular carcinoma and metastatic dis-ease to the liver. Semin InterventRadiol 2006; 23:64–72.

118. Georgiades CS, Salem R, GeschwindJF. Radioactive microspheres for thetreatment of HCC. In: Golzarian J, ed.Vascular embolotherapy: a compre-hensive approach. New York:Springer, 2005;141–148.

119. Salem R, Thurston KG, Geschwind JF.Yttrium-90 radioembolization for thetreatment of liver metastases. In: Gol-zarian J, ed. Vascular embolotherapy:a comprehensive approach. NewYork: Springer, 2005;149–161.

120. Mulcahy MF, Benson AB III. Bevaci-zumab in the treatment of colorectalcancer. Exp Opin Biol Ther 2005;5:997–1005.

121. Stathopoulos GP, Rigatos SK, Statho-poulos JG, et al. Efficacy and tolera-bility of oxaliplatin plus irinotecan5-fluouracil and leucovorin regimenin advanced stage colorectal cancerpatients pretreated with irinotecan5-fluouracil and leucovorin. Am J ClinOncol 2005; 28:565–569.

122. Herscher LL, Cook JA, Pacelli R, et al.Principles of chemoradiation: theoret-ical and practical considerations. On-cology (Williston Park) 1999;13(10suppl 5):11–22.

123. Florman S, Toure B, Kim L, et al.Liver transplantation for neuroendo-crine tumors. J Gastrointest Surg 2004;8:208–212.

124. Lehnert T. Liver transplantation formetastatic neuroendocrine carcinoma:an analysis of 103 patients. Transplan-tation 1998; 66:1307–1312.

125. Norton JA. Surgical treatment ofneuroendocrine metastases. Best PractRes Clin Gastroenterol 2005; 19:577–583.

126. Norton JA, Warren RS, Kelly MG, etal. Aggressive surgery for meta-static liver neuroendocrine tumors.Surgery 2003;134: 1057–1063.

127. Dejong CH, Parks RW, Currie E, et al.Treatment of hepatic metastases ofneuroendocrine malignancies: a 10-year experience. J R Coll Surg Edinb2002; 47:495–499.

128. Grazi GL, Cescon M, Pierangeli F, etal. Highly aggressive policy of he-patic resections for neuroendocrineliver metastases. Hepatogastroenter-ology 2000; 47:481–486.

129. Sarmiento JM, Heywood G, Rubin J,et al. Surgical treatment of neuroen-docrine metastases to the liver: a pleafor resection to increase survival. JAm Coll Surg 2003; 197:9–37.

130. Jaeck D, Oussoultzoglou E, BachellierP, et al. Hepatic metastases of gas-troenteropancreatic neuroendocrinetumors: safe hepatic surgery. WorldJ Surg 2001; 25:689–692.

131. Berber E, Flesher N, Siperstein AE.Laparoscopic radiofrequency ablationof neuroendocrine liver metastases.World J Surg 2002; 26:985–990.

132. Seifert JK, Cozzi PJ, Morris DL.Cryotherapy for neuroendocrine livermetastases. Semin Surg Oncol 1998;14:175–183.

133. Cozzi PJ, Englund R, Morris DL.Cryotherapy treatment of patients withhepatic metastases from neuroendo-crine tumors. Cancer 1995; 76:501–509.

134. Gupta S, Johnson MM, Murthy R, etal. Hepatic arterial embolizationand chemoembolization for the treat-ment of patients with metastatic neu-roendocrine tumors: variables affect-ing response rates and survival.Cancer 2005; 104:1590–1602.

135. Kress O, Wagner HJ, Wied M, et al.Transarterial chemoembolization ofadvanced liver metastases of neu-roendocrine tumors: a retrospectivesingle-center analysis. Digestion 2003;68: 94–101.

136. Delanoit T, Rubin J, Neczyporenko F,et al. Somatostatin analogues in thetreatment of gastroenteropancreaticneuroendocrine tumors. Mayo ClinProc 2005; 20:502–506.

137. di Bartolomeo M, Bajetta E, BuzzoniR, et al. Clinical efficacy of oct-reotide in the treatment of metastaticneuroendocrine tumors. A study by

the Italian Trials in Medical OncologyGroup. Cancer 1996; 77:402–408.

138. Angeletti S, Corleto VD, Schillaci O, etal. Single dose of octreotide stabilizemetastatic gastro-entero-pancreaticendocrine tumours. Ital J Gastroen-terol Hepatol 1999; 31:23–27.

139. Oberg K. Interferon-alpha versus so-matostatin or the combination of bothin gastro-enteropancreatic tumours. Di-gestion 1996;57(suppl 1):81–83.

140. Murthy R, Habbu AA, Richey SL, etal. Implications of patterns of metas-tases of colorectal cancer on liver-di-rected therapy in patients undergoingcombination systemic treatment withirinotecan and cetuximab. Presented atthe World Conference on Interven-tional Oncology; 2006; Cernobbio, Italy.

141. Kennedy AS, Nutting C, Coldwell D,et al. Pathologic response and mi-crodosimetry of (90)Y microspheres inman: review of four explanted wholelivers. Int J Radiat Oncol Biol Phys2004; 60:1552–1563.

142. Bruix J, Sherman M. Management ofhepatocellular carcinoma. Hepatol-ogy 2005; 42:1208–1236.

143. Sato K, Lewandowski RJ, Bui JT, et al.Treatment of unresectable primaryand metastatic liver cancer with yttri-um-90 microspheres (TheraSphere(R)):assessment of hepatic arterial emboliza-tion. Cardiovasc Intervent Radiol 2006;29:522–529.

144. Voboril R. Blood supply of meta-static liver tumors: an experimentalstudy. Int Surg 2005; 90 (2):71–77.

145. El-Serag HB. Hepatocellular carci-noma and hepatitis C in the UnitedStates. Hepatology 2002;36(5 suppl 1):S74–S83.

146. Kennedy AS, Dezern W, McNeillie P,et al. Dose selection of resin y90 mi-crospheres for liver brachytherapy: asingle center review. Presented at theAnnual Meeting of the AmericanBrachytherapy Society; May 10–12,2006; Philadelphia.

147. Liu DM, Salem R, Bui JT, et al. An-giographic considerations in patientsundergoing liver-directed therapy. JVasc Interv Radiol 2005; 16:911–935.

148. Murthy R, Nunez R, Szklaruk J, et al.Yttrium-90 microsphere therapy forhepatic malignancy: devices, indica-tions, technical considerations, andpotential complications. Radiograph-ics 2005;25(suppl 1):S41–S55.

149. Nutting C, Coldwell D, Kennedy AK.Coil embolization prevents gastroin-testinal ulcers during yttrium-90 he-patic radioembolization. Presented atthe Society of Interventional Radiol-ogy Annual Meeting 2004; Phoenix,AZ.

1592 • Radioembolization with 90Y Microspheres for Liver Malignancies October 2006 JVIR

Page 23: Radioembolization with Yttrium-90 Microspheres: A State-of ... · Radioembolization with Yttrium-90 Microspheres: A State-of-the-Art Brachytherapy Treatment for Primary and Secondary

CME TEST QUESTIONSThe following questions are based on the three-part series on radioembolization for liver malignancies found in theAugust, September, and current (October) 2006 issues of JVIR. Examination available at http://directory.sirweb.org/jvircme

1. You are an authorized user who plans to treat a hepatocellular carcinoma with TheraSphere radioembolization.Your goal is to treat the right lobe of the liver, which has a calculated volume of 900 mL. The patient has cirrhosiswith borderline liver function and you plan to deliver approximately 80 Gy to the target volume. Assume that thelung shunt fraction is 0%. What is the radioactivity required to deliver the desired dose to the liver?a. 0.48 GBqb. 1.48 GBqc. 2.53 GBqd. 5.24 GBq

2. In question 1, assume instead that the lung shunt fraction is 50%. Which of the following statements isfalse?a. A history of chronic obstructive pulmonary disease should be taken into account for treatment planning.b. The patient is at risk for pulmonary toxicity if the planned 80 Gy dose is delivered in a single treatment

session.c. Dose fractionation is a treatment option.d. Coil embolization of the hepatocellular carcinoma is a reasonable step before embolization.

3. You are an authorized user of SIR-spheres and would like to treat colorectal cancer metastases to theliver. The whole liver volume is 2000 cm3. The tumor volume is 350 cm3. According to the dosimetrymodel based on burden, what is the radioactivity required to treat the patient?a. 1.0 GBqb. 2.0 GBqc. 2.5 GBqd. 3.0 GBq

4. Which of the following statements regarding complications of 90Y radioembolization is false?a. Postembolization syndrome is more common with SIR-spheres than with TheraSphere.b. Patients who develop lymphopenia following radioembolization are at high risk for opportunistic

infection.c. 90Y-induced ulcers can occur with both TheraSphere and SIR-spheres and may prove refractory

to medical therapy.d. Liver failure may occur due to the irradiation of non-neoplastic liver parenchyma.

5. There is existing literature supporting all of the following conclusions regarding 90Y radioemboli-zation except:a. Okuda stage I and II patients with unresectable hepatocellular carcinoma exhibit increased

survival compared with historical controls following TheraSphere radioembolization.b. Factors associated with early (3-month) mortality include liver replacement by tumor of 70% or

more, albumin less than 3 g/dL, and bilirubin elevation of 2 mg/dL or more.c. SIR-spheres radioembolization for colorectal cancer has been shown to improve tumor re-

sponse and time to progression when combined with 5-fluorouracin.d. Treatment-related liver toxicity decreases with increased pretreatment total bilirubin levels.

Salem and Thurston • 1593Volume 17 Number 10


Recommended