Carcinoma of the Gallbladder: Role ofSonography in Diagnosis and Staging

Manoj Pandey, MS,1 Bimal P. Sood, MD,2* Ram C. Shukla, MD,2 Nakul C. Aryya, MD,3

Shailesh Singh, PhD,1 Vijay K. Shukla, MCh1

1 Department of Surgery, Institute of Medical Sciences, Banaras Hindu University, Varanasi 221 005, India2 Department of Radiology, Institute of Medical Sciences, Banaras Hindu University, Varanasi 221 005, India3 Department of Pathology, Institute of Medical Sciences, Banaras Hindu University, Varanasi 221 005, India

Received 27 May 1999; accepted 15 February 2000

ABSTRACT: Purpose. In an attempt to define the sono-graphic characteristics of gallbladder cancer, we ret-rospectively analyzed the sonographic findings in 203cases of gallbladder cancer confirmed by cytology orhistopathology.

Patients and Methods. Patients with proven gall-bladder cancer presenting to a single surgical unit be-tween 1991 and 1995 were identified through a re-cords search. All patients underwent sonographicexamination followed by fine-needle aspiration (FNA),biopsy, or laparotomy for establishing the diagnosis.

Results. A mass in the gallbladder and gallbladderwall thickening (> 12 mm) were cardinal sonographicfindings of carcinoma. Liver infiltration was correctlyidentified in all patients who had it. Sonography washighly accurate for detecting mass lesions, gallstones,liver infiltration, metastasis, and ascites. However, vi-sualization of lymph nodes, common bile duct infiltra-tion, and peritoneal dissemination was poor.

Conclusions. Sonography was found to be a gooddiagnostic tool for carcinoma of the gallbladder; how-ever, its sensitivity was poor for staging nodal spreadof the disease. © 2000 John Wiley & Sons, Inc. J ClinUltrasound 28:227–232, 2000.

Keywords: gallbladder neoplasm; ultrasonography;liver infiltration

Since its first documentation by de Stoll morethan 2 centuries ago,1 carcinoma of the gall-

bladder continues to be a diagnostic and thera-peutic challenge. Gallbladder cancer is the most

common biliary-tract malignancy and the thirdmost common gastrointestinal-tract malignancyin the eastern Uttar Pradesh and Bihar regions ofIndia. It constitutes 4.4% of all malignancies and0.3% of all hospital admissions at the UniversityHospital in Varanasi.2 Although the etiology ofgallbladder cancer remains obscure, several con-tributing factors have been proposed, includingcholelithiasis,3 chronic cholecystitis,3 the pres-ence of certain blood groups,4 exposure to carcino-gens5 or lipid peroxidation products,6,7 the pres-ence of benign tumors,8 and increased amounts ofsecondary bile acids.9

Early diagnosis is the only means of improvingsurvival in gallbladder cancer. However, this goalhas remained elusive despite major advances inimaging because of the nonspecific nature of thesymptoms and the prevalence of benign gall-bladder disease in India.10 Most stage I gallblad-der tumors are detected incidentally on histo-pathologic examination of gallbladders resectedfor presumably benign disease, hence the terms“inapparent” or “subclinical” carcinoma of thegallbladder.11,12

Advanced carcinoma of the gallbladder can bedetected accurately with sonography in 70–82% ofpatients,13–15 but early carcinoma can be detectedin only 23%.16 Most of the published sonographicfindings have come from small series in areas inwhich the incidence of gallbladder cancer is low.We report here sonographic findings in a series of203 cases of gallbladder carcinoma diagnosed inthe Uttar Pradesh area of India.

PATIENTS AND METHODS

We retrospectively identified, through a recordssearch, cases of gallbladder carcinoma in which

Correspondence to: M. Pandey, Division of Surgical Oncology,Regional Cancer Centre, Medical College P.O., Trivandrum695 011, India*Present address: Department of Radiodiagnosis, Postgradu-ate Institute of Medical Sciences and Research, Chandigarh,India

© 2000 John Wiley & Sons, Inc.

VOL. 28, NO. 5, JUNE 2000 227

the patients were sonographically examined andthe disease histologically confirmed at the Ba-naras Hindu University Hospital in Varanasi be-tween 1991 and 1995. Clinical details were ob-tained from surgical and sonographic registersmaintained by 2 separate departments.

Sonographic examinations were carried out af-ter an overnight fast with a real-time gray-scaleSonoline 250 ultrasound scanner (Siemens, Is-saquah, WA) with 3.5- or 5.0-MHz mechanicalsector transducers. The gallbladder was exam-ined for wall thickness, mass lesions, stones, andpericholecystic fluid collections. The liver was ex-amined for hepatomegaly, metastatic nodules, di-rect liver infiltration, and intrahepatic ductal di-latation. The rest of the abdomen and pelvis wereexamined for lymphadenopathy, metastases, andascites. Color Doppler sonography was not used.Longitudinal and right oblique scans were ob-tained in all patients; other scans were obtainedas required. Informed consent was obtained fromall patients.

Preoperative histologic diagnoses were madefor 165 patients (82%), most often by non-guidedfine-needle aspiration (FNA) from a palpable gall-bladder mass (98 cases) followed by sonographi-cally guided FNA (53 cases), transhepatic sono-graphically guided Tru-Cut biopsy (8 cases), andTru-Cut biopsy of a palpable mass (6 cases). Di-agnosis in the remaining 38 patients could be con-firmed only by histopathologic examination of theresected specimen.

RESULTS

We identified 203 patients diagnosed as havinggallbladder cancer. The most common clinical

presentation of gallbladder cancer in this serieswas loss of weight (201 patients, 99%) followed byloss of appetite (197 patients, 97%), pain in theright hypochondrium (143 patients, 70%), a massin the right hypochondrium (107 patients, 53%),jaundice (79 patients, 39%), and nausea and vom-iting (21 patients, 10%).

Abdominal sonography revealed a mass lesionin the gallbladder in 177 patients (87%). Themass was intraluminal in 104 patients (59%)(Figure 1) and infiltrative in 73 patients (41%)(Figure 2; Table 1). The masses had irregularmargins and were more echogenic than the liver.Unlike calculi, they did not show acoustic shad-owing; unlike sludge, they did not move when thepatient changed position. Intraluminal masseswere located in the body of the gallbladder in 56cases and in the fundus in 48 cases. Infiltrativelesions were common at the gallbladder neck (42cases) and fundus (31 cases). No infiltrative le-sions were detected in the gallbladder body.Masses at the fundus were larger (usually > 3.5cm) than those at the neck (< 2 cm).

Anechoic areas, seen within the growth in 124cases (61%), were probably due to tumor necrosisor residual bile. Gallstones were demonstrated in143 patients (70%) (Table 1) and were always ac-companied by acoustic shadowing (Figure 3). Inmost cases, the calculi were embedded in thegrowth and did not move when the patientchanged position. However, in 12 cases, calculiseen in the gallbladder fundus were mobile. Inthese latter cases, the calculi were small (< 1.5cm) and multiple. No intraductal calculi wereidentified.

Gallbladder wall thickening (> 12 mm), inho-mogeneous echoes, and ill-defined margins were

FIGURE 1. Transverse (left) and longitudinal (right) abdominal sonograms of the gallbladder show an intra-luminal solid hypoechoic mass (m).

PANDEY ET AL

228 JOURNAL OF CLINICAL ULTRASOUND

evident in 26 patients (13%) (Figure 4). The gall-bladder wall adjacent to the liver was more oftenthickened than was the wall of the rest of thegallbladder.

Hepatic infiltration, identified as hypoechoicareas extending from the gallbladder 1–4 cm intothe liver, was noted in 73 cases (36%) (Table 1).Another 13 patients (6%) had metastatic nodulesin the liver, appearing as multiple round, 1–2-cm,well-defined hypoechoic lesions involving pre-dominantly the right lobe. Intrahepatic ductal di-latation was seen in 119 cases (59%), more oftenin patients with gallbladder neck masses (69 pa-tients) than in those with gallbladder fundusmasses (20 patients). Serum bilirubin was normalin 12 (6%) of the 203 patients. The common bileduct could not be identified in any patient withintrahepatic ductal dilatation, as the obstructionat the porta hepatis was due either to an infiltra-tive growth or to a nodal mass.

Lymph node enlargement was demonstrated in39 patients (19%). The node groups most ofteninvolved were the periportal (33 cases), followedby the pancreaticoduodenal (17 cases), the para-aortic (16 cases), and, less often, the perichole-dochal (4 cases) nodes (Table 1). These nodes ap-peared as round, well-defined hypoechoic masseswith sharp margins and few internal echoes. Mostof the nodes were larger than 2.0 cm and discrete.

Ascites, present in 11 cases (5%), appearedsonographically as clear areas with no septae(Table 1). Finally, a compression effect on the por-tal vein was evident in 72 patients (35%), all ofwhom had large growths that had replaced al-most the entire gallbladder. Most of these pa-tients also had dilated intrahepatic ducts andliver invasion.

The sensitivity of the techniques used for pre-operative tissue diagnosis are compared in Ta-ble 2.

Of the 104 patients who underwent laparoto-my, all were found to have mass lesions in thegallbladder, 98 were found to have gallstones, and47 were found to have liver infiltration. Laparot-omy and sonography findings are compared inTable 3. Sonography was highly accurate in dem-onstrating gallbladder masses, gallstones, liverinfiltration, and ascites. However, sonographywas not accurate for visualizing common bile ductinfiltration and lymph node involvement, espe-cially of the pericholedochal and peripancreaticnodes, and did not detect any of the 6 cases ofperitoneal dissemination.

DISCUSSION

Our analysis of 203 cases of gallbladder cancerpresenting to a university hospital in India over a5-year period suggests that sonography should bethe primary imaging modality for patients withsuspected gallbladder cancer.

Until the mid-1970s, oral cholecystographywas the standard diagnostic imaging examinationfor suspected gallbladder disease.17,18 At thattime, gray-scale sonography began to be used as acomplement to cholecystography19,20 for patients

TABLE 1

Sonographic Findings in 203 Patients with Carcinoma of

the Gallbladder

Finding No. Patients (%)

Gallbladder mass lesion 177 (87)Intraluminal 104 (51)

Body 56 (28)Fundus 48 (23)

Infiltrative 73 (36)Fundus 31 (15)Neck 42 (21)

Gallbladder wall thickening 26 (13)Gallstones 143 (70)

Embedded in mass 131 (65)Mobile 12 (6)

Direct hepatic infiltration 73 (36)# 2 cm 46 (23)> 2 cm 27 (13)

Hepatic metastatic nodules 13 (6)Intrahepatic ductal dilatation 119 (59)Enlarged lymph nodes 39 (19)

Periportal 33 (16)Pancreaticoduodenal 17 (8)Para-aortic 16 (8)Pericholedochal 4 (2)

OtherAscites 11 (5)Portal vein compression 72 (35)

FIGURE 2. Longitudinal abdominal sonogram shows an intraluminalgallbladder mass infiltrating the adjacent liver.

CARCINOMA OF THE GALLBLADDER

VOL. 28, NO. 5, JUNE 2000 229

in whom the gallbladder could not be visualizedby that method, for pediatric or pregnant patientsto reduce their exposure to radiation, or foracutely ill patients who could not tolerate chole-cystography. Technologic refinements and im-

provements in the diagnostic accuracy of sonog-raphy have led to its becoming the primarymethod for diagnosing gallbladder diseases be-cause of its ease of performance, lack of ionizingradiation, and ability to image the entire abdo-men at the time of examination.21,22

Gallbladder carcinoma typically appears on so-nography as a mass with inhomogeneous echoesin the gallbladder with or without anechoic areasand calculi. Our finding of anechoic areas withinthe gallbladder is similar to that of Dalla Palmaet al,14 who suggested that these areas were dueto necrosis or residual bile within the gallbladder.We found that the site of the lesion within thegallbladder was reflected in the clinical presenta-tion. Mass lesions at the gallbladder fundus arethought to present later in the disease processthan do those at the neck, presumably becauseneck masses would infiltrate the common bileduct and produce biliary obstruction earlier in thecourse of the disease.23,24

In our series, large gallbladder masses oftenappeared on sonography as subhepatic masses.These masses were easily distinguished from

FIGURE 3. Two sonograms along slightly different planes show a gallbladder (GB) mass containing calculi(CAL) with acoustic shadowing (AS).

FIGURE 4. Longitudinal abdominal sonogram shows a small mass atthe gallbladder fundus and significant gallbladder wall thickening.

TABLE 2

Sensitivity of Preoperative Diagnostic Procedures in 203

Patients with Gallbladder Cancer

ProcedureNo. Tests

Performed*No. PositiveTest Results Sensitivity

FNA 132 98 74%Sonographically

guided FNA 59 53 90%Tru-Cut biopsy 7 6 86%Sonographically

guided Tru-Cut biopsy 9 8 89%

Abbreviation: FNA, fine-needle aspiration.*Four patients in whom FNA was inconclusive subsequently under-

went sonographically guided FNA.

PANDEY ET AL

230 JOURNAL OF CLINICAL ULTRASOUND

masses arising from adjacent organs if the gall-bladder lumen containing bile could be visualized,as was true for most patients in another series,25

or if the portal vein was compressed. The pres-ence of other features like involved periportalnodes, ascites, and liver infiltration also aided thediagnosis; otherwise, the origin of the mass wasdifficult to ascertain. A few of our patients hadonly gallbladder wall thickening. In these cases,the walls were typically more than 12 mm thickand showed inhomogeneous echoes. Althoughcholecystitis can also produce wall thickening(range, 3–10 mm), it can be distinguished fromcarcinoma on sonography by the presence of peri-cholecystic fluid collections or edema of the gall-bladder wall, which often gives the appearance ofa double wall.24 Empyema of the gallbladder canbe confused with carcinoma.13 None of our pa-tients had empyema. Color Doppler scanning canbe useful in distinguishing a vascular tumor fromavascular sludge.

Gallstones were seen in 70% of our patientswith gallbladder cancer. These multiple smallstones were seen as hyperechoic foci either withinthe mass or free in the lumen with acoustic shad-owing. Previous studies have showed gallbladdercancer to be associated with calculi in 40–100% ofcases.2,3,26–28

The accuracy of sonography in recognizing thetumor, its nature, its site of origin, and any ex-tension in our study was almost 100%, similar tothat of earlier reports.9,15,16,24 Detection of earlygallbladder cancer is difficult as these lesions aremucosal and do not necessarily give rise to appre-ciable wall thickening. If these lesions are associ-

ated with stones in a small, contracted gallblad-der, they tend to mimic cholecystitis, and theirpresence at histologic examination comes as asurprise.11,12 We earlier reported 2 cases of ad-vanced gallbladder cancer detected in asymptom-atic women undergoing routine sonographic ex-amination for pregnancy.10

Ascites and peritoneal dissemination (carcino-matosis peritonei) both occur late in the course ofthe disease. Ascites can be due to disseminatedmalignancy or nutritional causes; many patientswith ascites have hypoproteinemia. Sonographywas highly sensitive for detecting ascites, al-though it was not possible to comment on the na-ture of the ascites. On the other hand, all of ourcases of peritoneal dissemination were missed onsonographic examination.

In cases of advanced gallbladder cancer, othersonographic findings such as obstruction at theportal fissure or common bile duct, dilated intra-hepatic biliary ducts, direct infiltration of theliver, and lymphadenopathy further confirm thediagnosis. In our study, sonography was found tobe 100% accurate in predicting liver involvement.Of the 104 patients who subsequently underwentlaparotomy, 5 were found to have extrahepaticbiliary obstruction, and none of these cases wasdetected on sonography. Sonography was accu-rate in identifying involved periportal and para-aortic nodes but was inaccurate for visualizinginvolved pericholedochal and peripancreaticnodes.

In conclusion, sonography was highly accuratein establishing a diagnosis of gallbladder carci-noma in symptomatic patients, but it was foundto be a poor tool for staging. Routine use of sonog-raphy for screening may not be effective given thedifficulty of detecting early lesions. We suggestthat sonography be used for diagnosis of gallblad-der cancer in symptomatic patients and forscreening in those patients undergoing abdomi-nal sonography, for any reason, who live in high-incidence areas.

REFERENCES

1. de Stoll M. Rationis medendi in nosocomino prac-tico unindobonensi. Part I. Vienna: Beranrdi; 1777.

2. Shukla VK, Khadelwal C, Roy SK, et al. Primarycarcinoma of the gallbladder. A review of 16-yearperiod at University Hospital. J Surg Oncol 1985;28:32.

3. Piehler JM, Crichlow RW. Primary carcinoma ofthe gallbladder. Surg Gynecol Obstet 1978;147:929.

4. Pandey M, Gautam A, Shukla VK. ABO and rhesus

TABLE 3

Sonography versus Laparotomy Findings in 104 Patients

with Gallbladder Cancer

Finding

No. with Positive Findings

AccuracyLaparotomy Sonography

Gallbladder mass 104 104 100%Gallstones 98 97 99%Liver infiltration 47 47 100%

# 2 cm 43 41 95%> 2 cm 4 6 67%

Liver metastases 1 1 100%Common bile duct

infiltration 3 0 0%Porta hepatis infiltration 2 0 0%Involved lymph nodes 33 25 76%

Pericholedochal 17 4 23%Periportal 28 23 82%Peripancreatic 3 0 0%Para-aortic 1 1 100%Pancreaticoduodenal 16 12 75%

Ascites 3 3 100%Carcinomatosis peritonei 6 0 0%

CARCINOMA OF THE GALLBLADDER

VOL. 28, NO. 5, JUNE 2000 231

blood groups in patients with cholelithiasis andcarcinoma of the gallbladder. BMJ 1995;310:1639.

5. Klamer TW, Max MH. Carcinoma of the gallblad-der. Surg Gynecol Obstet 1983;156:641.

6. Shukla VK, Shukla PK, Pandey M, et al. Lipid per-oxidation products in bile from patients with car-cinoma of the gallbladder. A preliminary study. JSurg Oncol 1994;54:258.

7. Pandey M, Khatri AK, Dubey SS, et al. Erythrocytemembrane fatty acid profile in patients with pri-mary carcinoma of the gallbladder. J Surg Oncol1995;59:31.

8. Aldridge MC, Bismuth H. Gallbladder cancer: thepolyp-cancer sequence. Br J Surg 1990;77:363.

9. Shukla VK, Tiwari SC, Roy SK. Biliary bile acids incholelithiasis and carcinoma of the gallbladder.Eur J Cancer Prev 1993;2:155.

10. Pandey M, Khatri AK, Sood BP, et al. Cholecysto-sonographic evaluation of the prevalence of gall-bladder disease. A university hospital experience.Clin Imaging 1996;20:269.

11. Yamaguchi K, Tsuneyoshi M. Subclinical gallblad-der carcinoma. Am J Surg 1992;163:382.

12. Shirai Y, Yoshida K, Tsukada K, et al. Inapparentcarcinoma of the gallbladder. An appraisal of radi-cal second operation after simple cholecystectomy.Ann Surg 1992;215:326.

13. Yeh HC. Ultrasonography and computed tomogra-phy of carcinoma of the gallbladder. Radiology1979;133:227.

14. Dalla Palma L, Rizzatto G, Pozzi-Mucelli RS, et al.Grey scale ultrasonography in the evaluation ofcarcinoma of the gallbladder. Br J Radiol 1980;53:662.

15. Ruiz R, Teyssou H, Fernandez N, et al. Ultrasonicdiagnosis of primary carcinoma of the gallbladder:a review of 16 cases. J Clin Ultrasound 1980;8:489.

16. Tsuchiya Y. Early carcinoma of the gallbladder:microscopic and ultrasonic findings. Radiology1991;179:171.

17. Graham EA, Cole WH, Copher GH. Visualizing thegallbladder by sodium salt of tetra bromophenol-phthalein. JAMA 1924;82:1077.

18. Burhenne HJ, Obata WG. Single visit oral chole-cystography. N Engl J Med 1975;292:627.

19. Leopold GR, Amberg J, Gosink BB, et al. Grayscale ultrasonic cholecystography: a comparisonwith conventional radiographic technique. Radiol-ogy 1976;121:445.

20. Lawson TL. Gray scale cholecystosonography: di-agnostic criteria and accuracy. Radiology 1977;122:247.

21. Cooperberg PL, Gibney RG. Imaging of the gall-bladder, 1987. Radiology 1987;163:605.

22. Cooperberg PL, Burhenne HJ. Real-time ultraso-nography. Diagnostic technique of choice in calcu-lus gallbladder diseases. N Engl J Med 1980;302:1277.

23. Crade M, Taylor KJW, Rosenfield AT, et al. Surgi-cal and pathological correlation of cholecystosonog-raphy and cholecystography. AJR Am J Roentgenol1978;131:327.

24. Yum HY, Fink AH. Sonographic findings in pri-mary carcinoma of the gallbladder. Radiology1980;134:693.

25. Weill F, Eisenscher A, Zeltner F, et al. Aspecte ul-trasonore des cancer de la vesicule billare. Ann Ra-diol (Paris) 1979;22:17.

26. Adson MA, Farnell MB. Hepatobiliary cancer. Sur-gical considerations. Mayo Clin Proc 1981;56:686.

27. Strauch GO. Primary carcinoma of the gallbladder:presentation in 70 cases from the Rhode Islandhospital and cumulative review of last 10 years ofAmerican literature. Surgery 1960;47:368.

28. Shukla VK, Pandey M, Kumar M, et al. Ultrasoundguided fine needle aspiration cytology of malignantgallbladder masses. Acta Cytol 1997;41:1654.

PANDEY ET AL

232 JOURNAL OF CLINICAL ULTRASOUND