ORIGINAL PAPER

Kay-Oliver Kliche Æ Konstantin Kubsch Æ Martin Raida

Rami Masri-Zada Æ Klaus Hoffken

Chronomodulated chemotherapy in metastatic gastrointestinal cancercombining 5-FU and sodium folinate with oxaliplatin, irinotecanor gemcitabine: the Jena experience in 79 patients

Received: 15 November 2001 /Accepted: 18 June 2002 / Published online: 21 August 2002� Springer-Verlag 2002

Abstract Purpose: To study efficacy and tolerability ofchronomodulated (CM)-chemotherapy in patients withmetastatic or locally advanced tumors of the GI tract.Furthermore, calcium folinate was replaced by sodiumfolinate due to better feasibility. Patients and meth-ods: We treated 79 patients with metastatic or locallyadvanced colorectal cancer (n=52), cancer of the pan-creas/biliary tract (n=14), and other malignancies(n=14) with a total of 592 courses of CM-therapy. Outof the total study population 53/79, i.e., 67.1% hadreceived prior chemotherapy. Most of the patients(77.2%) received sodium-folinate-5-FU-oxaliplatin-CM (SOFOX-CM) as first-line chronomodulatedtherapy, 20.3% received sodium-folinate-5-FU-irino-tecan-CM (SOFIRI-CM), and 2.5% (n=1) receivedsodium-folinate-5-FU-gemcitabine-CM (SOFGEM-CM). Results: We found a moderate overall toxicitywith grade 3–4 neuropathy in 7.46% of patients during atotal of 433 courses of SOFOX-CM and grade 3–4diarrhea in 10.26% of patients after 154 courses ofSOFIRI-CM. SOFOX-CM had to be stopped only inone patient due to grade 3–4 sensory neuropathy. CM-therapy led to complete response (CR) in 1.3%, partialresponse (PR) in 15.2%, stable disease (SD) in 32.9%,and progressive disease (PD) in 44.3% of all patients.For the 26 chemonaive patients remission data were asfollows: CR one patient (3.8%), PR four patients(15.4%), SD seven patients (26.9%), PD 12 patients(46.3%), lost to follow-up one patient (3.8%), and

too-early-for-analysis one patient (3.8%). The medianprogression-free-survival (PFS) was 4 months (range,0–24 months). The median PFS was also 4 months(range, 0–21 months) for those patients receivingSOFOX-CM as first CM-therapy (n=61), while it wasfound to be 0 months (range, 0–10 months) for patients(n=16) receiving SOFIRI-CM as first chronomodulatedtherapy. Conclusions: We found CM-therapy to beeffective and safe in the treatment of advanced malig-nancies of the GI tract. Sodium folinate offers superiorfeasibility and compatibility with cytostatic drugs with-out drawbacks.

Keywords Chronotherapy Æ Colorectal cancer ÆOxaliplatin Æ Irinotecan Æ Sodium folinate

Introduction

Normal cell physiology is characterized by uniform aswell as stable changes in a circadian manner. More than20 years ago it was found that these endogenousrhythms can be exploited to reduce cytotoxic treatmentinsults provided that temporal adjustments are beingmade to the chemotherapy schedule (chronotherapy)(Halberg et al. 1980; Hrushesky et al. 1982). Originally5-FU was the first cytostatic drug to be investigated forchronomodulated application, since the rate-limitingenzyme, dihydropyrimidine dehydrogenase (DPD), hadbeen reported to vary according to the time of day. Inthis context our group recently studied circadianchanges in mRNA expression of DPD in leukocytes often patients with advanced gastro-intestinal carcinomasprior to CM-therapy and in five healthy controls. Evi-dence for a circadian DPD mRNA expression period-icity was found only in the healthy controls, while nosuch pattern could be detected in patients with advancedGI tumors (Raida et al. 2001).

The diaminocyclohexane-platinum complex – oxa-liplatin (l-OHP) – is a third-generation platinum com-plex, which lacks any renal toxicity and displays

J Cancer Res Clin Oncol (2002) 128: 516–524DOI 10.1007/s00432-002-0363-0

K.-O. Kliche (&) Æ K. Kubsch Æ M. Raida Æ K. HoffkenDepartment of Internal Medicine II,Friedrich-Schiller University,Erlanger Allee 101, 07740 Jena, GermanyE-mail: kay-oliver.kliche@med.uni-jena.deTel.: +49-3641-939303Fax: +49-3641-939372

R. Masri-ZadaDepartment of Internal Medicine I,Friedrich-Schiller University,Jena, Germany

antitumor activity against colorectal cancer (Extra et al.1990). The combination of 5-FU, leucovorin (LV), andl-OHP has consistently achieved 40–67% objective re-sponse rates in patients with metastatic colorectal car-cinoma especially when these drugs were given aschronotherapy (Levi et al. 1992; Levi et al. 1994, 1997).

In several clinical trials the adaptation of drug de-livery to circadian rhythms has improved tolerabilityand has enhanced antitumor efficacy against metastasesfrom colorectal cancer (Levi et al. 1994, 1997; Levi et al.1995b, 1995a).

The combination of 5-FU, LV, and l-OHP was de-veloped at first as a chronomodulated schedule in orderto take advantage of the improved tolerability associ-ated with this mode of administration. In an initial phaseII study, this three-drug combination was given as a5-day chronomodulated infusion to 93 patients withunresectable colorectal metastases. A 58% objectiveresponse rate and a median overall survival of16 months were achieved (Levi et al. 1992). Two con-secutive multicenter phase III studies compared flatversus chronomodulated infusion of the same three-drugcombination. These trials registered a total of 278patients with previously untreated and unresectablemetastatic colorectal cancer. Chronotherapy wassignificantly less toxic and more effective (objective re-sponse rate 51% vs 29%, P =0.0001) than constant-rateinfusion. In addition, the median-dose intensity of 5-FUwas 22% higher with chronotherapy than with flatinfusion (P <0.0001) (Levi et al. 1994, 1997). The highefficacy and the good tolerability of this chronomodu-

lated regimen led to further intensification with higherresponse rates both in chemotherapy-naive patients andin previously treated ones (Levi et al. 1999; Bertheault-Cvitkovic et al. 1996). Meanwhile, further preclinicaland clinical data were reported that demonstrate thatanother cytostatic drug effective in GI-tract cancer, i.e.,irinotecan (CPT-11), can be applied in a chronomodu-lated fashion (Garufi et al. 2001).

We here report our experience in a poor-prognosticpopulation of 79 pretreated and chemonaive patientswith various metastatic or locally advanced tumors ofthe GI tract using three different CM-schedules: sodium-folinate-5-FU-oxaliplatin-CM, sodium-folinate-5-FU-irinotecan-CM, sodium-folinate-5-FU-gemcitabine-CM(SOFOX-CM, SOFIRI-CM, and SOFGEM-CM,respectively).

Patients and methods

From June 1999 to August 2001, a total of 79 patients receivedinfusional chemotherapy with three different chronomodulatedprotocols always comprising 5-FU and sodium folinate in combi-nation with oxaliplatin, irinotecan or gemcitabine (details are givenbelow, see Figs. 1, 2, and 3).

Treatment was mainly administered in an outpatient settingemploying a four-channel programmable-in-time pump (Melodie,Aguettant, Lyon, France). Patients whose condition did not allowtreatment on an outpatient basis received their therapy in hospital.

Initial procedure for study entry included physical exam, routineblood and serum chemistry, liver enzyme tests, carcino-embryonicantigen (CEA) serum level, abdominal ultrasound, and, in mostcases, abdomino-pelvic and lung computerized tomography (CT).

Fig. 1. SOFOX-CM Protocol

Fig. 2. SOFIRI-CM Protocol

517

Patients that were treated in an ambulatory setting had a visit everytwo (SOFOX-CM) or three (SOFIRI-CM) weeks. Disease statuswas assessed with X-ray (chest radiograph) as well as ultrasound orthoracic and abdomino-pelvic CT scans which were performedevery 2 to 3 months. Responses were assessed by CT scan ratherthan ultrasound in order to allow for objective comparisons. Tumorresponse was assessed according to WHO criteria. Diagnoses aswell as further characteristics of patients are listed in Table 3.

Primary and secondary objectives

The primary criterion was maximum tumor response to therapy.Tumor response was first assessed by the clinical investigator. AllCT scans as well other imaging diagnostics were reviewed by expertradiologists at our department of radiologic diagnostics and in-ternists (ultrasound). Target lesions were considered measurable ifthey were not located in any previously irradiated field or in thebones.

Antitumor activity was assessed according to the World HealthOrganisation criteria: response was assessed after every secondtreatment course and was defined as follows: (1) complete response(CR), complete disappearance of all symptoms and signs of diseasefor a minimum of 4 weeks; (2) partial response (PR), a 50% re-duction (or more) in the sum of the products of the perpendiculardiameters of measurable disease and the appearance of no newmalignant lesion for a minimum of 4 weeks; (3) stable disease (SD),no appearance of new areas of disease or less than 50% decrease orless than 25% increase in the described measurements; and(4) progressive disease (PD), more than 25% increase in themeasurements and/or appearance of new lesions (WHO 1979).Secondary objectives included toxicity assessed according toWHO-scale and progression-free-survival (PFS).

The toxicities of each course were recorded before the nextcourse was started and were graded according to the WHO criteriafor hematology, hand-foot syndrome, nausea, vomiting, diarrhea,and mucositis (WHO 1979).

Chemotherapeutic regimen and dose modifications

Patients received a 4-day course of chronomodulated, intravenousinfusion 5-FU (Medac, Hamburg, Germany) (750 mg/m2/day) andsodium folinate (Medac, Hamburg, Germany) (150 mg/m2/day),which were infused simultaneously from 22.00 hours to 10.00hours. Oxaliplatin (Sanofi-Synthelabo, Berlin, Germany) (25 mg/m2/day) was infused from 10.00 hours to 22.00 hours on days 1–4(Fig. 1). This protocol was termed SOFOX-CM. Therapy was re-initiated on day 15. In several patients therapy intervals were ex-tended to three weeks and even up to 7 weeks in one patient afterthey had received more than ten courses. The antiemetic drug on-dansetron was administered in a dosage of 8 mg prophylactically atthe beginning of each course on day 1.

The so-called SOFIRI-CM-protocol lasts for 5 days and usesidentical dosages for 5-FU and sodium folinate with slightlydifferent infusion times (24.00 hours–12.00 hours on days 2–5).

Instead of oxaliplatin, irinotecan (CPT-11) is given on day 1(350 mg/m2/day) from 0200 hours to 0800 hours (Fig. 2). Therapywas reinitiated on day 21 and ondansetron was administered pro-phylactically alike. Atropine was not regularly administered in aprophylactic manner, but rapidly alleviated symptoms in patientsexperiencing side effects.

Additionally, we conceptualized a chronomodulated protocolcontaining Gemcitabine (SOFGEM-CM, Fig. 3), which is – indosages and infusion times of 5-FU and sodium folinate –similar tothe SOFOX-CM-protocol. Instead of oxaliplatin, gemcitabine isinfused on day 1 from 10.00 hours to 11.00 hours (750 mg/m2/day).It has to be stated though that currently there is no clear rationalefor the gemcitabine timing, which still needs to be defined in futureinvestigations.

Treatment withdrawal

Patients were taken off study if there was no complete recoveryfrom diarrhea, vomiting, stomatitis, and/or hematologic toxicitywithin 6 weeks after the last course or grade 2 sensory neuropathy.The assigned protocol treatment was also discontinued if diseaseprogressed. Other causes of withdrawal are indicated and includedrefusal by patients, loss to follow-up, or death.

Results

All patients

A total number of 79 patients with metastatic or locallyadvanced tumors, mainly of gastrointestinal origin, weretreated with one of three chemotherapy protocols (de-tails outlined above). Forty-three were male patients(54.4%)and their median age was 56 years. Diagnosesand further characteristics of patients are specified indetail in Tables 1, 2, and 3.

Fifty-three (67.1%) of our patients had received pri-or, often extensive, chemotherapy with up to three dif-ferent protocols (see Table 3 for details). A smallnumber (8.9%) of all patients had received prior oxa-liplatin, 7.6% other platinum derivatives, i.e., cisplatinand carboplatin, and 5.1% prior irinotecan. None of ourpatients had ever received chronotherapy before.

In 52/53 (98%) of these individuals the disease wasprogressive when chronomodulated therapy was started.CM-therapy was started only in one patient with gastriccancer when she still had stable disease.

A total of 592 single courses of chronomodulatedchemotherapy were given, 459 of which were applied as

Fig. 3. SOFGEM-CM Protocol

518

the first CM-therapy, 132 represented the second type ofCM-therapy for the individual patient, i.e., SOFOX-CMor SOFIRI-CM, depending on what was given as firstline CM-therapy, and one patient suffering from biliarytract cancer even received a third type of CM-therapy.

The 132 courses of second line CM-therapy wereadministered to 28 patients (35.4% of the total popu-lation), who either did not develop disease control orshowed progressive disease after temporary response tofirst line CM-therapy. The corresponding figures of pa-tients being switched, i.e., receiving a second CM-ther-apy, according to the diagnosis are listed in Table 2.

Since maximum tumor response to therapy is theprimary objective of this analysis, treatment results ac-cording to the WHO classification were obtained. Afterfirst line CM-therapy a complete remission was achievedin one patient (1.3%), 12 patients (15.2%) reached apartial remission, disease stabilization occurred in 26patients (32.9%), while the disease was found to beprogressive in 35 individuals, i.e., in 44.3% of the study

population (see Table 1). As mentioned earlier, 28 pa-tients were treated with a second CM-therapy that re-sulted in PR in two patients (7.1%) and SD in anothernine patients (32.1%). Remarkably, among these 11patients there are two subjects with colon or pancreaticcancer, showing continuous disease control for morethan 2 years.

The median progression-free-survival for all 79 pa-tients is 4 months (range, 0–24 months, see Table 2).Figure 4 depicts PFS for all diagnoses grouped accord-ing to colorectal cancer (CRC), cancer of the pancreas/biliary tract (PB) or other malignancies (other). In caseof more than one phase of PFS the individual timeperiods were summed up.

Colorectal cancer

Patients suffering from colon cancer represent the largestgroup in our study (n=52). Thirty-seven of them

Table 1. Treatment results

Primary tumor site Colorectal Pancreas/biliary Others All patients

Number of patients 52 14 13 79 (100%)Gender Male 25 (48%) 10 (71%) 8 (61.5%) 43 (54.4%)Age (years) 61 55 56 56Prior chemotherapy Yes 37 (71.2%) 7 (50%) 9 (69.2%) 53 (67.1%)Prior oxaliplatin 7 (13.5%) 0 0 7 (8.9%)Other platinum derivatives 0 (0%) 1 (7.1%) 5 (38.5%) 6 (7.6%)Prior irinotecan 3 (5.8%) 1 (7.1%) 0 4 (5.1%)Prior chronotherapy 0 (0%) 0 (0%) 0 (0%) 0 (0%)Cycles of 1. CM-therapy 341 52 66 459Cycles of 2. CM-therapy 105 24 3 132Cycles of 3. CM-therapy 0 1 0 1Total number of SOFOX-CM cycles 311 61 61 433Total number of SOFIRI-CM cycles 135 11 8 154Total number of SOFGEM-CM cycles 0 4 0 4Treatment result after 1.CM-therapy in 79 patients

CR 1 (1.9%) 0 (0%) 0 (0%) 1 (1.3%)PR 7 (13.5%) 1 (7.1%) 4 (30.8%) 12 (15.2%)SD 20 (38.5%) 3 (21.4%) 3 (23.1%) 26 (32.9%)PD 20 (38.5%) 10 (71.4%) 5 (38.5%) 35 (44.3%)Lost-to-follow-up 2 (3.9%) 0 (0%) 1 (7.7%) 3 (3.8%)Too early 2 (3.9%) 0 (0%) 0 (0%) 2 (2.5%)

Treatment result after 2.CM-therapy in 28 patients

CR 0 (0%) 0 (0%) 0 (0%) 0 (0%)PR 1 (1.9%) 1 (7.1%) 0 (0%) 2 (7.1%)SD 8 (15.4%) 1 (7.1%) 0 (0%) 9 (32.1%)PD 12 (23.1%) 0 (0%) 1 (7.7%) 13 (46.4%)Lost-to-follow-up 0 (0%) 0 (0%) 0 (0%) 0 (0%)Too early 2 (3.9%) 1 (7.1%) 1 (7.7%) 4 (14.3%)

Table 2. Treatment results

Primary Tumor Site Colorectal Pancreas/biliary Others All patients

Number of patients 52 (65.8%) 14 (17.7%) 13 (16.5%) 79 (100%)SOFOX as first CM-therapy 39 (75.0%) 11 (78.6%) 11 (84.6%) 61 (77.2%)SOFIRI as first CM-therapy 13 (25. 0%) 1 (7.1%) 2 (15.4%) 16 (20.3%)SOFGEM as first CM-therapy 0 (0%) 2 (14.3%) 0 (0%) 2 (2.5%)Patients receiving second CM-therapy 23 (44.2%) 3 (21.4%) 2 (15.4%) 28 (35.4%)Patients receiving third CM-therapy 0 (0%) 1 (7.1%) 0 (0%) 1 (1.3%)Total number of CM-cycles applied (Median; range) 7 (0-28) 4 (2–19) 4.5 (1–14) 5 (0–28)Total time on CM-therapy (Months; median; range) 5 (0–24) 3 (1–24) 4 (1–10) 5 (0–24)Progression-free survival (Months; median; range) 4 (0–24) 2 (0–24) 4 (0–8) 4 (0–24)

519

Table 3. Treatment results

Primarytumor site

Initial TNMStaging

Site ofmetastases

Prior che-motherapy

Therapyprotocol

Result

Biliary tract Unknown Bone, locally advanced No n.a. n. a.Biliary tract Unknown Locally advanced No n.a. n. a.Biliary tract T3N1M0R0 Lung No n.a. n. a.Biliary tract T3N1M1 Liver Yes Gemcitabine PDBiliary tract T4N0M1R0 Liver, peritoneum No n.a. n. a.Biliary tract T4N1M1 Pleura, bone, skin Yes Gemcitabine, 5FU/FA, CPT11 PDBiliary tract T4N2M1 Liver, bone No n.a. n. a.Biliary tract T4NXM1 Liver Yes Chemoembolization PDBiliary tract T4NXM1 Locally advanced Yes 5 FU/FA PDBiliary tract T4NXM1 Locally advanced No n.a. n. a.Colon T2N0M0R0 Locally advanced Yes 5 FU/FA PDColon T2N1M1R2 Liver Yes 5 FU/FA, high-dose 5-FU PRColon T3M0G2R0 Liver Yes 5FU/FA PDColon T3M0G2R0 Lung, CNS Yes 5FU/FA ResponseColon T3M0G3R0 Lung, pleura No n. a. n. a.Colon T3N0M1G2 Liver, lung, CNS No n. a. n. a.Colon T3N0MXG2 Liver, lung Yes 5FU/FA PDColon T3N1M0 Liver, lung Yes FolFox – flat rate PDColon T3N1M0G3 Liver Yes 5FU/FA, CPT11,

FolFox – flat ratePD

Colon T3N1M0R0 Liver, lung Yes 5FU/FA PDColon T3N1M1 Liver, lung Yes 5FU/FA PDColon T3N1MX Liver Yes 5FU/FA AdjuvantColon T3N2M0 Liver No n. a. n. a.Colon T3N2M0R0 Liver Yes 5FU/FA PDColon T3N2M1 Liver Yes 5FU/FA, CPT11 PDColon T3N2M1G3 Liver, lung Yes n. a. PDColon T3N2M1G3 Liver, lung No 5 FU/FA, high-dose 5-FU n. a.Colon T3N2M1R2 Liver No n. a. n. a.Colon T3N2MX Liver, bone Yes 5FU/FA PDColon T3N2R1G3 Lung Yes 5FU/FA PDColon T3N3G2R0 Liver Yes 5 FU/FA, FolFox – flat rate PDColon T4M1G3R0 Liver, lung, bone Yes 5FU/FA, Capecitabine, CPT-11 PDColon T4M1R2 Liver No n. a. n. a.Colon T4N1M1 Liver, lung, bone Yes FolFox – flat rate SDColon T4N2L0V0 Liver No n. a. n. a.Colon T4N2M1 Liver, CNS No n. a. n. a.Colon T4N2M1R0 Liver No n. a. n. a.Colon T4N2M1R2 Liver No n. a. n. a.Colon T4N2MXG3 Liver Yes 5FU/FA AdjuvantCUP (Adeno-carcinoma)

Unknown Liver No n.a. n. a.

CUP (Adeno-carcinoma)

Unknown Peritoneum Yes Paclitaxel, carboplatin PD

Pancreas M1 Liver No n. a. n. a.Pancreas T2M0G1R0 Liver, pleura Yes Gemcitabine PDPancreas T2N1M1 Liver Yes Gemcitabine, cisplatin, 5 FU/FA,

chemoembolizationPD, PD,Response

Pancreas T4NXM1R2 Liver, locally advanced Yes Gemcitabine PDPleura Unknown Lung No n. a. n. a.Rectum Unknown Lung, locally advanced Yes 5FU/FA PDRectum Unknown Liver, lung Yes 5FU/FA PDRectum T1NXM1 Liver No n. a. n. a.Rectum T2N1M0 lung Yes 5FU/FA ResponseRectum T3M1G2 Liver Yes 5FU/FA PDRectum T3M1R2G2 Liver, lung No n. a. n. a.Rectum T3N0G3 Liver, locally advanced Yes 5FU/FA PDRectum T3N0M0 Lung, locally advanced Yes 5FU/FA ResponseRectum T3N0M0R0 Lung Yes 5FU/FA ResponseRectum T3N0M1R0 Liver Yes 5FU/FA PDRectum T3N0R0 Bone, lung No n. a. n. a.Rectum T3N1M1R0 Liver No n. a. n. a.Rectum T3N1M1R2 Liver, lung Yes FolFox – flat rate PDRectum T3N2M0G2 Lung Yes 5FU/FA PDRectum T3N2M0R0 Liver, CNS Yes 5FU/FA PDRectum T3N3M1R2 Liver Yes 5FU/FA CR

520

(71.2%) had received prior chemotherapy, which in thevast majority did not lead to either an objective response(i.e., CR + PR) or at least disease control (i.e., SD, seeTable 3).

Seven patients (13.5%) had received prior oxaliplatin,which led to SD in one patient and left six patients withprogressive disease. Within this study five patients werereexposed to SOFOX-CM, which achieved SD in threepatients, while two patients remained progressive. Theremaining two of seven patients received SOFIRI-CM asfirst CM-therapy, which led to PR in the one and PD inthe other patient. Among the chemo-naive patients(n=15) six achieved CR (1 x) and PR (5 x) (40%) andanother three individuals (20%) showed a stabilizationof their disease. Out of the entire group, i.e., chemo-naive and pretreated, 28 (53.9%) patients achieved a

disease stabilization or better after first line CM-therapy,which in 75.0% of all cases was SOFOX-CM. Anothernine out of 23 patients (40%) could be scored as SD orbetter after second CM-therapy, which in this case fre-quently was SOFIRI-CM. Tables 1 and 2 as well asFig. 4 provide more detailed information. It has to bementioned that out of these 52 patients at least threeindividuals have been considered for surgical removal ofresidual disease after successful CM-therapy.

Pancreatic/biliary tract cancer

Within this group there were four patients with pan-creatic cancer and ten with cancer of the biliary tract.Among the four patients with pancreatic cancer there

Table 3. (Continued)

Primarytumor site

Initial TNMStaging

Site ofmetastases

Prior che-motherapy

Therapyprotocol

Result

Rectum T3NXM1G2 Lung Yes 5FU/FA, radiotherapy ResponseRectum T4M0G2R2 Locally advanced Yes Mitomycin, 5 FU/FA PDRectum T4M1R2G2 Liver Yes 5 FU/FA, FolFox – flat rate PDRectum T4N1M1R0 Liver, lung Yes FolFox – flat rate, Irinotecan,

tomudexPD

Rectum T4N2M0 Liver, lung Yes 5FU/FA PDRectum T4NxMx Locally advanced Yes 5FU/FA PDStomach T2 M0 R0 Locally advanced Yes ELF, EM PDStomach T2N1M0 Locally advanced Yes ELF SDStomach T2N1M0G3 Liver Yes FLAP PDStomach T3N1M1 Peritoneum Yes PLF, ELF, FEM ResponseStomach T3N3M0R0 Locally advanced No n. a. n. a.Stomach T3N3M1 Peritoneum Yes ELF PDStomach T3N3M1G3 Locally advanced Yes FLAP PDStomach T4N0M0 Locally advanced Yes FLAP, ELF ResponseStomach T4N2M0G3 Locally advanced No n. a. n. a.Stomach T4N2M0R0 Liver Yes ELF, EM PD

Fig. 4. PFS of all patientsshown separately for the CRC,PB, and other groups. Themedian PFS for all patients was4 months (range, 0–24 months)

521

are two remarkable responders to SOFOX-CM therapy.While one patient is currently scored as SD after sixcourses of first line CM-therapy another patient onlyentered PR after failing all preceding therapies includingchemoembolization of liver metastases as well as fourcourses of the SOFGEM regimen. Only after startingher on SOFOX-CM as second line CM-therapy did sheachieve PR which is now ongoing for 24 months. Thepatient is currently off therapy for 6 months and underclose follow-up control.

Ten patients with biliary tract cancer received a totalof 47 courses of CM-therapy. One patient achieved PRand another two were found to have controlled disease.The remaining seven, i.e., 70%, experienced progressivedisease and soon died.

The patient having PR after first line SOFOX-CMfor 6 months was able to be stabilized again by secondline SOFIRI-CM for another 6 months. So far he is theonly patient receiving a third line intended CM-therapy,which contained paclitaxel as a 3-h flat infusion followedby 5-FU and sodium folinate in a chronomodulatedfashion. Unfortunately, this approach was unable toavoid disease progression. For more detailed informa-tion on PFS for the entire group see Table 2 and Fig. 4.

Other malignancies

This group comprises 13 patients with a large subgroupof ten patients with gastric cancer, two patients withcarcinoma of unknown primary syndrome (CUP) andone patient suffering from pleuramesothelioma. Thefigures for the whole group are provided in Tables 1 and2 and Fig. 4.

A total of ten patients with the cancer of the stomachreceived 59 courses of CM-therapy with a median of 5.5(1–14) courses and a median total time on CM-therapyof 4.5 (1–10) months (see Table 2). This resulted in sevenpatients being scored as SD and PR. The two chemo-naive patients having received SOFOX-CM are aliveand have controlled disease after an observation time of7 months up to now. Among these patients there are atleast two candidates in whom another surgical approachaiming for complete macroscopic removal of residualdisease has to be considered.

Two patients with carcinoma of CUP and one patientsuffering from pleuramesothelioma received a total often CM-therapy courses. All patients showed rapiddisease progression.

Toxicity

In general, both CM-schedules were found to have onlymoderate toxicity (see Table 4 for detailed information).FolFox-CM-therapy had to be discontinued only in onecase due to a grade 3–4 sensory neuropathy. This patientwas switched to FolFiri-CM, but unfortunately showedprogressive disease and soon died.

The most frequent side effect of SOFOX-CM wassensory neuropathy in 7.46% of all patients, who re-ceived at least one course of this regimen, correspondingto 1.15% of all courses. It was followed by grade 3–4diarrhea in 4.48% of all patients and mucositis andnausea/vomiting with 3% each, respectively.

The so-called ‘‘Hand-Foot Syndrome’’ was remark-ably rare with only one patient (1.49%) out of 67 during

Table 4. Treatment resultsSOFOX-CMTherapy

SOFIRI-CMTherapy

SOFGEM-CMTherapy

n=433; Pat=67 n=154; Pat=39 n=4; Pat=2

Grade 3–4 mucositis 2 0 0% of patients 3.00% 0.00% 0.00%% of courses 0.46% 0.00% 0.00%Grade 3–4 sensory neuropathy 5 0 0% of patients 7.46% 0.00% 0.00%% of courses 1.15% 0.00% 0.00%Grade 3–4 diarrhea 3 4 0% of patients 4.48% 10.26% 0.00%% of courses 0.69% 2.6% 0.00%Grade 3–4 angina pectoris 0 1 0% of patients 0.00% 2.56% 0.00%% of courses 0.00% 0.65% 0.00%Grade 3–4 hand-foot syndrome 1 0 0% of patients 1.49% 0.00% 0.00%% of courses 0.23% 0.00% 0.00%Grade 3–4 Nausea/vomiting 2 1 0% of patients 3.00% 2.56% 0.00%% of courses 0.46% 0.65% 0.00%Grade 3–4 Psychosyndrome 1 0 0% of patients 1.49% 0.00% 0.00%% of courses 0.23% 0.00% 0.00%As first CM-therapy 14 2 0As second CM-therapy 0 4 0

522

one course (0.23%)from a total of 433 courses of SO-FOX-CM.

SOFIRI-CM caused severe (i.e., WHO grade III–IV)diarrhea in four patients (10.26%) and one episode ofangina pectoris as well one severe case of nausea/vom-iting in 39 patients treated.

Discussion

The relevance of the chronotherapy principle has beensuccessfully tested since its introduction into clinicalmedicine in the early 1980s (Halberg et al. 1980;Hrushesky et al. 1982; Sothern et al. 1989). In thisperiod, more than 1,500 patients with gastrointestinalcancer were enrolled in phase I, II, and III trials (Leviet al. 1997, 1999; Focan et al. 1999; Giacchetti et al.1999; Cure et al. 2000; Garufi et al. 2000). Chrono-modulated infusions of 5-FU, generally given withcalcium folinate or as in our case sodium folinate,form the basis of most chronotherapy schedules so far.Since there are no pharmacological interactions be-tween sodium folinate and 5-FU, patients do not needtwo-chamber ports. Thus, the use of sodium folinateoffers considerable advantages over calcium folinate inthis regard.

Several studies found CM-therapy to be significantlyless toxic than a comparable flat-rate, i.e., constant-rateinfusion of the same drugs (Levi et al. 1994, 1997). Inaccordance with this we found in our series of 592courses of SOFOX or SOFIRI-CM-therapy grade 3–4sensory neuropathy as the major side effect of SOFOX-CM in 7.46% of the patients. The most frequent sideeffect of SOFIRI-CM was grade 3–4 diarrhea whichoccurred in 10.26% of all patients. CM-therapy had tobe discontinued and switched from SOFOX-CM toSOFIRI-CM in only one of our 79 patients treated dueto grade 3–4 sensory neuropathy. There was no treat-ment-related death in our cohort of patients.

Thus, CM-therapy proved to produce a remarkablylow toxicity in our hands. This holds true for the majortoxicity of the FOLFOX-regimen, which is sensoryneuropathy, as well as for diarrhea following theFOLFIRI-regimen. In comparison, the incidence ofgrade III–IV diarrhea under irinotecan and 5-FU is farhigher with conventional protocols (22.7%) (Saltz et al.2000).

In our experience CM-therapy demonstrated a re-markable treatment effect in patients who had under-gone extensive prior chemotherapy. Most of our patients(67.1%) were pretreated and about 98% had to bescored PD when CM-therapy was started. Nevertheless,nearly half of these patients achieved a stabilization oftheir disease or even CR/PR results.

Moreover, a second line CM-therapy approach in-duced 39.2% of SD/PR in our patient cohort. Amongthose patients we find numerous examples of long-last-ing disease control of up to 24 months which is onlyachieved after a second line CM-therapy.

We extended the use of CM-therapy to diseases suchas pancreatic, gastric or biliary tract cancer and madesome interesting observations. SOFOX-CM producedremarkably good results in two patients with metastaticpancreatic cancer as well as in seven patients with gastriccancer. In our patients suffering from stomach andpancreas cancer oxaliplatin demonstrated considerableand long-lasting effects, which favors further investiga-tion of this drug for these diseases.

Our results in colorectal cancer indicate that drugresistance against oxaliplatin may be overcome, if theapplication is changed from the ‘‘flat-rate mode’’ to the‘‘chrono-mode’’. However, the corresponding numbersare small and it should be emphasised that the vastmajority (86.5%) of our pretreated patients had onlyreceived various 5-FU-containing regimens without ox-aliplatin. This interesting issue will have to be addressedin a separate trial. From a clinician’s point-of-view it hasto be stressed that most of our patients could be treatedon outpatient-basis, which certainly increases quality oflife.

Regarding the limitations of our current results it hasto be stressed that our study group comprises patientswith different diagnoses in different stages having haddifferent therapies. Due to this several parameters, suchas overall survival, were omitted, because the time in-tervals from diagnosis until metastasis and the beginningof CM-therapy vary considerably. On the other hand, allpatients suffer from metastatic solid tumors originatingfrom the GI tract, which opens up a certain degree ofcomparability. Thus, while some parameters make nosense in our group of patients and were not analyzed,such as overall survival, toxicity as well as treatmentresults can reliably be assessed. Chronotherapy startedout with colorectal cancer, which clearly is the best ex-amined disease in this regard (Levi et al. 1999; Giacchettiet al. 2000; Misset and Levi 2000; Adam et al. 2001; Leviet al. 2001).

From the results described in this article we believethat it might be justified to extend CM-therapy to thetreatment of other diseases using other drugs. There arepreclinical data as well as some first clinical data onnewer drugs such as taxanes (Granda et al. 2001). At thesame time we have some knowledge regarding olderdrugs (e.g., doxorubicin) that still awaits further con-sideration (Halberg et al. 1980).

As a prerequisite for the development of clinicalprotocols exact chronopharmacological data on theappropriate way of how to apply a given drug isneeded. Some drugs are tolerated best during night-time whereas other drugs show exactly the oppositebehaviour with maximum tolerability during hours ofdaylight. With these data at hand further randomizedlarge-scale clinical trials have to be designed compar-ing chronomodulated therapy to flat-line-infusiontherapy.

Taken together our report is the first that investigatesthe potential of CM-therapy in patients with extensivepretreatment. Most of the available data on CM-therapy

523

reported during the last two decades describes results inpatients with no chemotherapy in the metastatic setting.

References

Adam R, et al (2001) Five-year survival following hepatic resectionafter neoadjuvant therapy for nonresectable colorectal. AnnSurg Oncol 8:347–353

Bertheault-Cvitkovic F, et al (1996) Biweekly intensified ambula-tory chronomodulated chemotherapy with oxaliplatin, fluoro-uracil, and leucovorin in patients with metastatic colorectalcancer. J Clin Oncol 14:2950–2958

Cure H, et al (2000) Phase II trial of chronomodulated infusion of5-fluorouracil and folinic acid in metastatic colorectal cancer.Anticancer Res 20:4649–4653

Extra JM, et al (1990) Phase I study of oxaliplatin in patients withadvanced cancer. Cancer Chemother Pharmacol 25:299–303

Focan C, et al (1999) Continuous delivery of venous 5-fluorouraciland arterial 5-fluorodeoxyuridine for hepatic metastases fromcolorectal cancer: feasibility and tolerance in a randomizedphase II trial comparing flat versus chronomodulated infusion.Anticancer Drugs 10:385–392

Garufi C, et al (2000) Overcoming resistance to chronomodulated5-fluorouracil and folinic acid by the addition of chronomod-ulated oxaliplatin in advanced colorectal cancer patients.Anticancer Drugs 11:495–501

Garufi C, et al (2001) Irinotecan and chronomodulated infusion of5-fluorouracil and folinic acid in the treatment of patients withadvanced colorectal carcinoma: a phase I study. Cancer 91:712–720

Giacchetti S, et al (1999) Long-term survival of patients with un-resectable colorectal cancer liver metastases following infu-sional chemotherapy with 5-fluorouracil, leucovorin,oxaliplatin and surgery. Ann Oncol 10:663–669

Giacchetti S, et al (2000) Phase III multicenter randomized trial ofoxaliplatin added to chronomodulated fluorouracil-leucovorinas first-line treatment of metastatic colorectal cancer. J ClinOncol 18:136–147

Gornet JM, et al (2000) Dramatic tumor response of bulky livermetastases following treatment with CPT-11 and a chrono-modulated 4-day infusion of 5-fluorouracil, folinic acid andoxaliplatin every 2 weeks in a colorectal cancer patient. Anti-cancer Drugs 11:263–268

Granda TG, et al (2001) Experimental chronotherapy of mousemammary adenocarcinoma MA13/C with docetaxel anddoxorubicin as single agents and in combination. Cancer Res61:1996–2001

Halberg F, et al (1980) Chronotherapy of mammary cancer in rats.Int J Chronobiol 7:85–99

Hrushesky WJ, et al (1982) Circadian stage dependence of cis-diamminedichloroplatinum lethal toxicity in rats. CancerRes 42:945–949

Levi F (1999) Cancer chronotherapy. J Pharm Pharmacol51:891–898

Levi F, et al (1992) A chronopharmacologic phase II clinical trialwith 5-fluorouracil, folinic acid, and oxaliplatin using an am-bulatory multichannel programmable pump. High antitumoreffectiveness against metastatic colorectal cancer. Cancer69:893–900

Levi FA, et al (1994) Chronomodulated versus fixed-infusion-ratedelivery of ambulatory chemotherapy with oxaliplatin, fluoro-uracil, and folinic acid (leucovorin) in patients with colorectalcancer metastases: a randomized multi-institutional trial. J NatlCancer Inst 86:1608–1617

Levi, F., S. Giacchetti, et al (1995a) Chronomodulation of che-motherapy against metastatic colorectal cancer. InternationalOrganization for Cancer Chronotherapy. Eur J Cancer31A:1264–1270

Levi F, et al (1995b) A phase I-II trial of five-day continuous in-travenous infusion of 5-fluorouracil delivered at circadianrhythm modulated rate in patients with metastatic colorectalcancer. J Infus Chemother 5 [3 Suppl 1]:153–158

Levi F, et al (1997) Randomised multicentre trial of chronotherapywith oxaliplatin, fluorouracil, and folinic acid in metastaticcolorectal cancer. International Organization for CancerChronotherapy. Lancet 350:681–686

Levi F, et al (1999) A multicenter evaluation of intensified,ambulatory, chronomodulated chemotherapy with oxaliplatin,5-fluorouracil, and leucovorin as initial treatment of patientswith metastatic colorectal carcinoma. International Organiza-tion for Cancer Chronotherapy. Cancer 85:2532–2540

Levi F, et al (2001) Chronotherapy of colorectal cancer metastases.Hepatogastroenterology 48:320–322

Misset JL, Levi F (2000) Chronotherapy with 5-fluorouracil andother drugs in gastrointestinal malignancies. ChronotherapyGroup of the European Organization for Research andTreatment of Cancer. Semin Oncol 27 [5 Suppl 10]:78–82

Raida M, et al (2002) Circadian variation of dihydropyrimidinedehydrogenase mRNA expression in leukocytes and serumcortisol levels in patients with advanced gastrointestinal carci-nomas compared to healthy controls. J Cancer Res Clin Oncol128:96–102

Saltz LB, et al (2000) Irinotecan plus fluorouracil and leucovorinfor metastatic colorectal cancer. Irinotecan Study Group. NEngl J Med 343:905–914

Sothern RB, et al (1989) Control of a murine plasmacytoma withdoxorubicin-cisplatin: dependence on circadian stage of treat-ment. J Natl Cancer Inst 81:135–145

WHO (1979) WHO Handbook for Reporting Results of CancerTreatment. WHO, Geneva, Switzerland

524