US-Guided Percutaneous Needle Biopsy ofthe Spleen Using 18-Gauge versus 21-GaugeNeedles

Ping Liang, MD,1 Yangyan Gao, MD,2 Yang Wang, MD,1 Xiaoling Yu, MD,1 Dejiang Yu, MD,1

Baowei Dong, MD1

1 Department of Ultrasound, Chinese PLA General Hospital, 28 Fuxing Road, Beijing 100853, China2 Department of Special Diagnosis, South Building, The General Hospital of Chinese People’s Armed Police Forces,69 Yongding Road, Beijing 100039, China

Received 31 August 2006; accepted 18 April 2007

ABSTRACT: Purpose. To compare the techniques of

sonographically (US)-guided percutaneous needle

biopsy of the spleen using 18-gauge and 21-gauge

needles.

Methods. Forty-two patients undergoing 43 spleen

biopsy procedures for focal lesions (n ¼ 27 [16 single,

11 multiple]) or diffuse splenomegaly (n ¼ 15) were

analyzed. Two groups were divided randomly accord-

ing to needle type: group 1 comprised 25 patients

biopsied with an 18-gauge cutting needle for histo-

logic examination; group 2 comprised 17 patients

biopsied with a 21-gauge needle for histologic and

cytologic examinations. Diagnostic accuracy, compli-

cation rate, and number of needle passes were com-

pared between the 2 groups.

Results. Correct histopathologic diagnosis was

obtained in 36 cases, whereas incorrect diagnosis

occurred in 6 cases. The accuracy of US-guided

spleen biopsy in this series was 85.7%, with 1 patient

(2.4%) having postprocedural hemorrhage. Com-

pared with the 21-gauge needle, the 18-gauge needle

had higher diagnostic accuracy (P < 0.05), required

fewer needle passes (P < 0.05), and there was no sig-

nificant difference in overall complication rate.

Conclusion. Because biopsy with an 18-gauge nee-

dle yields larger and unfragmented samples with

higher diagnostic rate compared with a 21-gauge nee-

dle, and no increased rate of major complication

requiring surgical intervention, it may be advanta-

geous to use an 18-gauge cutting needle in the US-

guided needle biopsy of splenic lesions. VVC 2007 Wiley

Periodicals, Inc. J Clin Ultrasound 35:477–482,

2007; Published online in Wiley InterScience (www.

interscience.wiley.com). DOI: 10.1002/jcu.20390

Keywords: spleen; biopsy; ultrasonography; core

needle biopsy; fine needle biopsy; interventional

ultrasound

Sonographically (US)-guided percutaneousneedle biopsy of solid visceral organs has

been widely used. However, the spleen, thoughfrequently involved in primary or secondary dis-eases, is rarely biopsied.1 The reason for this isthe risk of hemorrhage2,3 and the difficulty infinding a safe access to avoid injuring adjacentorgans such as the pleura, pancreatic tail, splenicflexure of the colon, and kidney. Several reportshave demonstrated that spleen biopsy under USguidance, especially with 20–22-gauge fine nee-dles, is safe and efficient.1,4–7 However, few stud-ies describing the use of 18-gauge needles havebeen reported.7 In this study, we report our expe-rience in the last 7 years of US-guided percutane-ous needle biopsy of the spleen with 18-gaugeand 21-gauge needles and compared the safetyand accuracy of both techniques.

PATIENTS AND METHODS

Forty- three US-guided spleen biopsies were per-formed between January 1998 and August 2005in 42 patients (29 men, 13 women) aged 7–75

Correspondence to: P. LiangGrant sponsors: National Science and Technology PromotionProject of China (grant no. 2004BA714B10); National ScientificFoundation Committee of China (grant no. 30371367)

' 2007 Wiley Periodicals, Inc.

VOL. 35, NO. 9, NOVEMBER/DECEMBER 2007—DOI 10.1002/jcu 477

years. Twenty-seven patients had a biopsy of asolitary lesion (n ¼ 16) or multiple focal lesions(n ¼ 11), whereas 15 patients had a biopsy for dif-fuse splenomegaly of unknown origin. Sizes ofthe focal lesions ranged from 1.5 to 7.8 cm, with amean 6 SD of 3.92 6 1.54 cm. One patient withpain in the left upper quadrant, fever, and singlesplenic lesion received repeat biopsy 1 week laterbecause of prior unsatisfactory tissue samples forhistologic examination.

All patients underwent abdominal sonographicand/or CT examination before biopsy showing asolitary lesion or multiple focal lesions or asplenomegaly unrelated to cirrhosis. When a defi-nite diagnosis could not be established based onimaging, biopsy of the spleen was performed. Theclinical findings and medical histories are listedin Table 1.

Patients who were candidates for spleen bi-opsy were randomly divided into 2 groups. Sizeand location of the lesion did not influence alloca-tion of the patients. Group 1 consisted of 25patients receiving core needle biopsy with 18-gauge cutting needles and a Magnum automatedbiopsy gun (Bard, Covington, GA) for histologicexamination, while group 2 comprised 17 patientswho underwent fine needle biopsy with a 21-gaugemanual biopsy needle (Hakko, Togura, Japan) forboth histologic and cytologic examinations. Spe-cial immunohistochemical stains were performedas needed by the pathologists.

A platelet count of >100,000/ll, a prothrombintime of <20 seconds, and prothrombin activity>50% of the normal control were required. Thestudy was approved by our institutional reviewboard, and informed consent was obtained fromall patients before undergoing the procedure.

Sonographic examinations were performedwith a 128XP/10 ART or a Sequoia 512 scanner(Siemens Ultrasound, Mountain View, CA) with2.5–5-MHz probe. The most direct pathway to thelesion via the least amount of splenic paren-chyma was chosen under real-time examination.Under local anesthesia with 1% lidocaine, needlebiopsies were performed under US guidanceusing an 18-gauge cutting needle with an auto-mated biopsy gun or a 21-gauge manual biopsyneedle while the patient temporarily suspendedrespiration. Under US guidance, an 18-gaugebiopy needle was inserted into the edges of thesplenic lesions, the trigger of the automated gunwas released, and the needle was withdrawn.One puncture was defined as a ‘‘pass.’’ To avoidsampling error, 2 needle passes were made foreach biopsy. A pathologist was present at thetime to evaluate the specimens for adequacy of

the core (length, fragmentation) and of the cy-tology material by microscopy (hematoxylin-eo-sin stain). If samples were insufficient or unsat-isfactory, 1 or more passes would be made. Theprocedure was considered to be complete oncethe pathologist determined that adequate sam-ples had been obtained. After the procedure,the patients were observed in the supine posi-tion to monitor vital signs and assess for imme-diate complications. Two hours later, sono-graphic examination was performed to confirmabsence of bleeding. Outpatients without bleed-ing were discharged, whereas those with sus-pected hemorrhage were admitted for 24-hourobservation. Inpatients would continue the ob-servation 2 hours later after returning to theirdepartment.

All patients were followed up with clinicalcourse and sonography or CT 1 month after bi-opsy. Final diagnosis was established via surgicalpathology (n ¼ 27), response to adjuvant chemo-therapy (n ¼ 8), antituberculosis therapy (n ¼ 1),anti-inflammatory therapy (n ¼ 2), or the combi-nation of clinical course and imaging follow-upfrom 6 months to 3 years (n ¼ 4).

Final pathology results and patient medicalrecords were reviewed for analysis of biopsy accu-racy and complications. A biopsy was consideredsuccessful if a specific benign or malignant histo-logic diagnosis was made at subsequent surgeryor if clinical follow-up confirmed the results ofthe biopsy.

Diagnostic accuracy, complication rate, andnumber of needle passes were compared betweenthe 2 groups using the Fisher exact of X2 test.Statistical analysis was performed using Statasoftware version 7.0 (Stata Corporation, CollegeStation, TX). A p value of less than 0.05 was con-sidered statistically significant.

TABLE 1

Clinical Findings/History and Patterns of Splenic Lesions on

Imaging Prior to US-Guided Needle Biopsy in 42 Patients

Clinical Findings/History

Splenic Lesions

Diffuse

Focal

Single

Focal

Multiple

Abdominal dull pain 4 5 3

Fever of unknown cause 1 1 4

Left upper quadrant pain and fever 2 2 0

Tumor in other parts of the body 1 3 3

Suspected hematologic diseases 6 0 0

Lesion found incidentally 0 5 1

Swollen extremities and weight loss 1 0 0

Total 15 16 11

LIANG ET AL

478 JOURNAL OF CLINICAL ULTRASOUND—DOI 10.1002/jcu

RESULTS

Using the 18-gauge cutting needle and automatedbiopsy gun, 1 sample was obtained in 2–3 seconds,while with the 21-gauge manual needle, 1 passlasted 3–5 seconds. Of the 42 patients, all 8 outpa-tients were discharged 2 hours after the biopsy.One patient underwent a repeat biopsy because ofprior unsatisfactory pathologic samples.

Thirty-six cases had a correct pathologic diag-nosis, including 17 malignant and 19 benign(Figure 1A and 1B). The accuracy of US-guidedspleen biopsy in this series was 85.7% (Table 2).In group 1, a correct diagnosis was established in19 cases. Nine biopsy targets were near the sple-nic hilum (Figure 2A and 2B), and 9 casesrequired additional immunohistochemical stains.In group 2, a correct diagnosis was obtained in 17cases. Five targets were near the splenic hilum,and 8 cases required additional immunohisto-chemical stains. The diagnosis was incorrect in 6cases (1 in group 1 and 5 in group 2). Only 1 caseof hemangioma was near the hilum, and the bi-opsy was performed with a 21-gauge needle. In10/13 patients with lymphoma, immunohisto-chemical stain was performed and allowed a sub-type diagnosis of lymphoma. False-negativeresults occurred in 2 patients who underwent a21-gauge needle biopsy without immunohisto-chemical staining.

Of the 42 patients, 37 had a spleen biopsyalone. Four patients underwent concomitant

liver biopsy (2 in group 1 and 2 in group 2). Onepatient in group 1 underwent concomitant kidneybiopsy. The additional biopsies were performedbecause of uncertainty as to whether the spleniclesion and lesions in other organs were of thesame origin. Of the 4 patients who had concomi-tant liver biopsies, 1 with hepatosplenomegalyhad non-Hodgkin’s lymphoma involving the liverand spleen. The spleen shrunk significantly afterchemotherapy. One patient underwent hepaticresection and splenectomy, the final diagnosisbeing hepatocellular carcinoma with spleen me-tastases. Focal lesions in the liver and spleen inthe remaining 2 patients were hepatosplenic me-tastases from lung cancer and skin melanoma,both of which were eventually confirmed by clini-cal course. One patient with a left renal lesionand splenomegaly was diagnosed as splenic pro-liferation and renal acidophilic cell adenoma,which was consistent with surgery result.

Each patient in our series underwent 2 or 3passes. Average needle passes were 2.3 in group 1and 2.8 in group 2. There was a statistically signif-icant difference in the number of needle passes,with 18-gauge needle biopsy requiring fewer needlepasses to obtain satisfactory specimens (p< 0.05).

Although biopsy with a 21-gauge needle yieldedboth histologic and cytologic results, the diagnos-tic accuracy (12/17) was significantly lower thanthat of the 18-gauge core needle (24/25), whichallowed histologic examination only (p< 0.05).

TABLE 2

Results of US-Guided Needle Biopsy of the Spleen with 18-Gauge or 21-Gauge Needle in 42 Patients

Final Diagnosis

18-Gauge 21-Gauge

Correct Incorrect Correct Incorrect

Malignant

Non-Hodgkin’s lymphoma 5 4 2

Hodgkin’s lymphoma 1 1

Chronic granulocytic leukemia 1 1

Metastases of hepatocellular carcinoma 1

Metastases of melanoma of thoracic skin 1

Metastases of small-cell carcinoma of the lung 1

Hemangiosarcoma 1

Benign

Splenic parenchymal proliferation 3 2

Hemangioma 2 1 2

Inflammatory infiltration 1 1

Hamartoma 2 1

Congestion 2

Lymphangioma 1

Inflammatory pseudotumor 1

Infarct 1

Tuberculosis 1

Abscess 1

Castleman’s disease 1

Total 24 1 12 5

The diagnostic accuracy of biopsies with 18-gauge needle was superior to that of biopsies with 21-gauge needles (p < 0.05).

US-GUIDED NEEDLE BIOPSY OF THE SPLEEN

VOL. 35, NO. 9, NOVEMBER/DECEMBER 2007—DOI 10.1002/jcu 479

One patient, a 62-year-old man, suffered frompostprocedural bleeding. This patient had beenreferred to our hospital for a low-grade fever ofunknown origin. CT examination revealed diffuseenlargement of the spleen with multiple, small,ill-defined areas of low density suspicious for lym-phoma. Sonographic examination revealed asmall amount of ascites. The patient’s coagula-tion profile was normal. US-guided biopsy of thespleen was performed using an 18-gauge cuttingneedle with 3 passes. The amount of ascites didnot change on the follow-up examination after 2hours of observation. However, the patient com-plained of fullness of the abdomen 8 hours afterthe biopsy. One thousand milliliters of bloody as-cites were drained, and the patient recoveredwith a 400-ml blood transfusion. The biopsyshowed congestion of the spleen with neutro-philic granulocyte infiltration, which was non-diagnostic. The diagnosis of multicentric Castle-

man’s disease was established via subsequentcervical lymph node and bone marrow biopsiesand laboratory results.

DISCUSSION

Most cases of focal splenic lesions and splenomeg-aly (other than cirrhosis) are asymptomatic andare found incidentally on physical examination.Compared with focal lesions of the liver, focalsplenic lesions are rare. Although color Dopplerand contrast-enhanced sonographic examinationof splenic nodules have shown promisingresults,8,9 differentiation of benign lesions (eg,hemangioma, hamartoma, and lymphangioma)and malignant tumors remains difficult via imag-ing.9–12

In 1985, Lindgren et al13 used a 14-gauge nee-dle to perform spleen biopsy, but 13% of the

FIGURE 1. A 7-year-old boy with a history of cervical synoviosar-

coma and a lesion in the spleen discovered 6 months after resection

of the neck tumor. (A) Transverse sonogram shows an isoechoic

rounded solid mass in the spleen (arrows). (B) Sonogram obtained

during US-guided biopsy shows the 18-gauge needle (arrowheads)

in the lesion (arrows). The histopathologic result was consistent

with lymphangioma confirmed by clinical course and on imaging

follow-up.

FIGURE 2. A 52-year-old woman with a single splenic lesion. (A)

Sonogram shows the hypoechoic lesion (arrows) located at the sple-

nic hilum near splenic vessels (arrowheads). (B) Sonogram obtained

during US-guided biopsy shows the 18-gauge needle (arrowheads)

within the lesion (arrows). The histopathologic result was consistent

with inflammatory pseudotumor subsequently confirmed on sple-

nectomy.

LIANG ET AL

480 JOURNAL OF CLINICAL ULTRASOUND—DOI 10.1002/jcu

patients experienced major complications. In thelast decade, however, many publications reportedthat US- or CT-guided percutaneous spleen bi-opsy with 20–23-gauge needles is safe and effi-cient.1,4–7,14–18 Civardi et al5 reported the datafrom a multicenter Italian study of US-guidedfine needle aspiration or core needle biopsy of thespleen in 398 patients, yielding an accuracy rateof 84.9–90.9% with a major complication rate of<1%. However, biopsy of the spleen with 18-gauge needles has rarely been reported.1,7,18

Muraca et al7 reported 30 biopsies in children,13 of which were performed with 18-gauge nee-dles. Accurate diagnosis was achieved in all ofthem without any complication. Several cases ofbiopsy perfomed with 18-gauge core needles werealso reported in another 2 studies.1,18 In ourseries, 25 patients underwent a biopsy with an18-gauge needle, and in all but 1 patient was adefinite diagnosis obtained safely.

Some reports have claimed that fine needleaspiration (FNA) of the spleen is helpful in mak-ing a diagnosis.4,5,14–16 Diagnostic yield of FNAranges from 30% to 85%, and when FNA is com-bined with histologic examination, the diagnosticaccuracy increases by approximately 3–5%.4,5,15

In our series, the accuracy of FNA in differentiat-ing benign from malignant lesions was only52.9% (8/17). In 2 patients with lymphoma and afalse-negative diagnosis via 18-gauge core biopsy,FNA did not help.

Most complications of splenic are related withhemorrhage, although uncontrolled bleedingresulting in death or requiring splenectomy hasbeen rare.4,5,13,18 An interesting number ofpasses is a significant risk factor of complica-tions.19 With 18-gauge cutting needles, fewerpasses are needed, thus reducing the risk of com-plications. As long as coagulation parameters arenormal and no contraindication for biopsy exists,biopsy with an 18-gauge needle could be as safeas with a 21-gauge needle. In our series, no com-plication occurred in the 17 cases performed withthe fine needle, while 1 of the 25 cases performedwith the 18-gauge needle had an intraperitonealhemorrhage. In this case, the small amount of as-cites may have prevented coagulation at thepuncture site, resulting in the hemorrhage.

As recommended by others,1,3 we selected theshortest pathway to the target lesion to minimizethe amount of normal splenic paranchyma trav-ersed by the needle. Fourteen lesions were adja-cent to the hilum. They were all biopsied withoutcomplication, most likely because real-time so-nographic monitoring allowed avoidance of thelarge vessels. Because no major complication oc-

curred, biopsy of lesions near the hilum shouldnot be regarded as a contraindication.

Any invasive procedure in the spleen bears ahigher risk of hemorrhage compared with otherorgans. Prolonged observation may be necessaryeven in patients with normal coagulation status.The risk of hemorrhage may also be influencedby the pathology of the lesion, with a higher riskassociated with angiosarcomas and angio-mas.18,20

In conclusion, 18-gauge core biopsy needlesyield bigger and unfragmented samples withhigher diagnostic rate and no major complicationrequiring surgical intervention and should beused for biopsy when histopathologic diagnosis ofsplenic lesions is needed.

REFERENCES

1. Keogan MT, Freed KS, Paulson EK, et al. Imag-ing-guided percutaneous biopsy of focal spleniclesions: update on safety and effectiveness. AJRAm J Roentgenol 1999;172:933.

2. Solbiati L, Bossi C, Vellotte E, et al. Focal lesionsin the spleen: sonographic patterns and guided bi-opsy. AJR Am J Roentgenol 1983;140:59.

3. Quinn SF, van Sonnenberg E, Casola G, et al.Interventional radiology in the spleen. Radiology1986;161:289.

4. Venkataramu NK, Gupta S, Sood BP, et al. Ultra-sound guided fine needle aspiration biopsy of sple-nic lesions. Br J Radiol 1999;72:953.

5. Civardi G, Vallisa D, Berte R, et al. Ultrasound-guided fine needle biopsy of the spleen: high clini-cal efficacy and low risk in a multicenter Italianstudy. Am J Hematol 2001;67:93.

6. Suri R, Gupta S, Gupta SK, et al. Ultrasoundguided fine needle aspiration cytology in abdomi-nal tuberculosis. Br J Radiol 1998;71:723.

7. Muraca S, Chait PG, Connolly BL, et al. US-guided core biopsy of the spleen in children. Radi-ology 2001;218:200.

8. Abbott RM, Levy AD, Aguilera NS, et al. From thearchives of the AFIP: primary vascular neoplasmsof the spleen: radiologic-pathologic correlation.Radiographics 2004;24:1137.

9. Gorg C, Gorg K, Bert T, et al. Colour Dopplerultrasound patterns and clinical follow-up of inci-dentally found hypoechoic, vascular tumours ofthe spleen: evidence for a benign tumour. Br JRadiol 2006;79:319.

10. Mohan V, Jones RC, Drake AJ 3rd, et al. Littoralcell angioma presenting as metastatic thyroid car-cinoma to the spleen. Thyroid 2005;15:170.

11. Skillings JR, Bramwell V, Nicholson RL, et al.A prospective study of magnetic resonanceimaging in lymphoma staging. Cancer 1991;67:1838.

US-GUIDED NEEDLE BIOPSY OF THE SPLEEN

VOL. 35, NO. 9, NOVEMBER/DECEMBER 2007—DOI 10.1002/jcu 481

12. Dennhardt N, Gorg C, Neubauer A. Frequencypattern and differential diagnosis of echogenicsplenic changes: sonographic follow-up study.Ultraschall Med 2000;21:151.

13. Lindgren PG, Hagberg H, Eriksson B, et al. Exci-sion biopsy of the spleen by ultrasound guidance.Br J Radiol 1985;58:853.

14. Caraway NP, Fanning CV. Use of fine-needle aspi-ration biopsy in the evaluation of splenic lesions ina cancer center. Diagn Cytopathol 1997;16:312.

15. Lieberman S, Libson E, Maly B, et al. Imaging-guided percutaneous splenic biopsy using a 20- or22-gauge cutting-edge core biopsy needle for thediagnosis of malignant lymphoma. AJR Am JRoentgenol 2003;181:1025.

16. Delacruz V, Jorda M, Gomez-Fernandez C, et al.Fine-needle aspiration diagnosis of angiosarcoma

of the spleen: a case report and review of theliterature. Arch Pathol Lab Med 2005;129:1054.

17. O’Malley ME, Wood BJ, Boland GW, et al. Percuta-neous imaging-guided biopsy of the spleen. AJRAm J Roentgenol 1999;172:661.

18. Lucey BC, Boland GW, Maher MM, et al. Percuta-neous nonvascular splenic intervention: a 10-yearreview. AJR Am J Roentgenol 2002;179:1591.

19. van der Poorten D, Kwok A, Lam T, et al. Twenty-year audit of percutaneous liver biopsy in a majorAustralian teaching hospital. Intern Med J 2006;36:692.

20. Drinkovic I, Brkljacic B. Two cases of lethal com-plications following ultrasound-guided percutane-ous fine-needle biopsy of the liver. CardiovascIntervent Radiol 1996;19:360.

LIANG ET AL

482 JOURNAL OF CLINICAL ULTRASOUND—DOI 10.1002/jcu