Pancreatic cystadenocarcinoma: diagnostic problems

1982, British Journal of Radiology, 55, 558-561

Pancreatic cystadenocarcinoma: diagnostic problems By 1P. Busilacchi, 2 G. Rizzatto, 3 M. Bazzocchi, 4E. Boltro, 5F. Candiani, 6F. Ferrari, G. M. Giuseppettl, 7P. Mirk, 5 L Rubaltelli, 6 L Volterrani. and 8F. Zappasodi Universities of 1Ancona, 6Cattolica di Roma, 5Padova, 6Siena and 3Trieste and the Hospitals of 8Forli, 2 Gorizia, 4Oftalmico di Torino 1

(Received November 1981) ABSTRACT

Eleven cases of pancreatic cystadenocarcinoma were studied with respect to the differential diagnosis from that of pancreatic pseudocysts, using diagnostic ultrasound as the initial investigative technique. Most of the cases presented an ultrasonic picture that correlated well with the gross pathology, but this correlation was not reliable enough to be considered characteristic in differentiating such masses from pseudocysts. The authors divide the cystadenocarcinomas of the pancreas into four echographic classes. In all but one of the classes, diagnostic ultrasound was insufficient by itself to render an accurate diagnosis. In such cases other techniques, i.e. CT, tissue biopsy, and especially angiography, were required for a reliable diagnosis.

Most cysts of the pancreas are pseudocysts and a minority are retention cysts. Neoplastic cystic pathology, whether benign (cystadenoma) or malignant (cystadenocarcinoma), is rare, representing about 15% of all pancreatic cysts (Becker et al, 1965; Piper et al, 1962). At the Mayo Clinic, USA, only 42 cystadenocarcinomas and 20 cystadenomas were reported over a 50-year period (Piper et al, 1962; Cullen et al, 1963). Until 1965 only 65 cases of pancreatic cystadenocarcinoma had been reported in the medical literature. Data from diagnostic ultrasound is sparse, the major contribution being from Freeny et al (1978), who reported one case of cystadenoma and seven cystadenocarcinomas. MATERIALS AND METHODS

A controlled series of 11 patients with proven pancreatic cystadenocarcinoma underwent echographic examinations at various medical centres that cooperated in this study. Three were female (age range 32-75 years), and eight were male (age range 55-77 years). Eight of the patients subsequently had angiographic studies, and two also CT scans. After surgical exploration the prospective radiological diagnosis was subject to pathological confirmation. None gave a history of chronic pancreatic disease or alcoholism. At the time of hospitalization four patients complained of anorexia or digestive disturbances, five presented with a palpable abdominal mass, six showed evident weight loss and eight complained of a diffuse abdominal pain, on occasion being described as a "a belt around the upper belly". Isolated symptoms were: jaundice, diarrhoea, anaemia, left pleural effusions, sarcophobia, and hyperglycaemia. 0007-1285/82/5508-558 $2.00/0

The technique of the ultrasound examination was one "standard" for a pancreatic abdominal study (Weill, 1978). Various types of equipment were used. All scanners were static with digital grey scale display and 3.5 MHz transducers. RESULTS AND DISCUSSION

Based on our data we have divided the cystadenocarcinomas into four echographic classes. (1) An anechoic mass with enhancement of the posterior wall and marginal irregularities (Fig. 1). Three of our cases were Class 1. (2) An anechoic mass with internal non-homogeneous echoes, stratified with the patient supine and mobile with the patient in different decubitus positions (Fig. 2). Three of our cases were Class 2. (3) An anechoic mass with irregular internal vegetations protruding into the lumen, and showing no movement with the patient in the various decubitus positions (Fig. 3). Two cases were Class 3. (4) A completely echoic mass with relatively distinct limits and a non-homogeneous echo pattern (Figs. 4a, b; 5a-d). Three cases were Class 4. There was good correlation between the mass dimensions estimated by ultrasound and those found at surgery. The smallest mass was 3.5x5.5 cm, the largest 10x12 cm, the average (greater axis) being 8.2 cm. In the three cases of Class 1 (one in the pancreatic head and two in the tail) an incorrect prospective diagnosis of benign cystic mass was made. Even when all the erroneously presumed "pseudocysts" were reviewed retrospectively a correct differential diagnosis, based on the echographic marginal irregularities correlating with neoplastic infiltration, was difficult. In two of these three cases the mass could not be attributed to the pancreas; only angiography, showing pathological changes in the pancreatic vasculature, gave a correct diagnosis. The three cases of Class 2 were all seen as mixed masses. In two of these cases an incorrect diagnosis of haemorrhagic pseudocyst was suggested. In the third case the mass was thought to be malignant because concurrent liver secondary deposits were seen. The correct diagnosis in this last case was not supported by angiography, which showed an avascular pancreatic mass with displacement of the normal pancreatic vessels. There were two other cases, however, where angiography correctly disputed the preliminary echographic diagnosis. The use of CT, where angiographic and ultrasound data conflicted, did not prove conclusive in our small series.

558

AUGUST 1982

Pancreatic cystadenocarcinoma: diagnostic problems

1. Anechoic mass in the pancreatic tail, with enhancement of the posterior wall (arrows) and marginal irregularities. Left kidney (K). FIG.

FIG. 2. Anechoic mass with internal non-homogeneous echoes (arrows), statified and mobile.

FIG. 3 (left). Anechoic mass (M) with irregular internal vegetation protruding into the lumen (arrows).

FIG. 4 (below). Completely echoic mass (M) of the pancreatic tail. Liver (L), and left kidney (K). (a) Transverse sub-xyphoid scan; (b) transverse umbilical scan.

559

VOL.

55, No. 656 P. Busilacchi et al

In all the cases of Class 2 the internal echographic solid component of the mass, showing nonhomogeneous echoes in both dimension and intensity, was seen at gross pathology to consist of squamous-type cells in a muco-gel material. The internal component of the mass found to be blood clots was longer and more echogenic than the solid component. In summary, for this Class 2, without a proper history and/or special additional procedures, a differential diagnosis between haemorrhagic pseudocysts and cystadenocarcinoma could not be made by ultrasound alone. For two of the cases in Class 3 (both in the pancreatic tail) an ultrasound diagnosis of a liquid mass with internal vegetations, probably malignant, was made. This was confirmed by angiography. In one of the two cases it was initially difficult to determine the organ of origin of the mass due both to its large size and the lack of good

FIG. 5. Solid mass (M) of pancreatic head, with relatively distinct limits. Liver (L), right kidney (K). (a) ultrasound transverse scan; (b) ultrasound longitudinal scan; (c) CT: solid-liquid mass of the pancreatic head; (d) angiogram: hypovasculized mass (arrow).

560

AUGUST 1982

Pancreatic cystadenocarcinoma: diagnostic problems echographic landmarks. Only a subsequent angiogram showed the mass to be of pancreatic origin. In Qass 3 a differentia] diagnosis between cystadenocarcinoma and haemorrhagic pseudocysts was easier; the latter presented mobile contents that appeared deeply stratified; whereas the cystadenocarcinoma showed vegetations linked to the mass wall by a pedicle, typical of papilliary neoplasms. The prospective echographic diagnosis in the cases of Class 4 was that of a solid pancreatic mass. Angiography in one case demonstrated a neoplastic circulation, yet in another it showed an avascular mass. A subsequent CT scan of this second case showed a liquid-solid mass, suggesting cystadenocarcinoma of the pancreatic head. Anatomically the echogenic solid component of the mass was referable to the cells and, especially, to the mucus of the mass. In contrast to Class 2, the mucus component filled the entire mass. An accurate diagnosis is difficult in such a class, often requiring collateral evidence, i.e. liver secondary deposits, for validity.

Only in our Class 3 of cystadenocarcinomas, rare in our study, was the echographic diagnosis accurate. A reliable diagnosis in the other three classes required supplementary diagnostic procedures, i.e. angiography, CT, tissue biopsy, etc. In some cases, due to the large size of the mass, such additional examinations were required just to localize the mass to the pancreas. Angiography was especially valuable in demonstrating a pathological circulation in cases where ultrasound suggested a pancreatic pseudocyst.

REFERENCES BECKER, W. F., WELCH, R. A. & PRATT, H. S., 1965. Cys-

tadenoma and cystadenocarcinoma of the pancreas. Annals of.Surgery, 161, 845-860. CULLEN, P. K., REMINE, W. H. & DAHLIN, D. C , 1963. A

clinicopathologic study of cystadenocarcinoma of the pancreas. Surgery, Gynecology and Obstetrics, 117, 189-195. FREENY, P. C , WEINSTEIN, C. J., TAFT, D. A. & ALLEN, F. H.,

1978. Cystic neoplasms of the pancreas: new angiographic and ultrasonographic findings. American Journal of Roentgenology, 131, 795-802.

CONCLUSIONS

Cystadenocarcinomas of the pancreas present different echographic findings. In most cases the ultrasonic picture correlates well with the gross pathology, but unfortunately this is not reliable enough to be called "characteristic".

PIPER, C. E., REMINE, W. H. & PRIESTEY, J. T., 1962. Pancreatic

cystadenoma. Report of 20 cases. Journal of the American Medical Association, 180, 648-652. WEILL, F. S., 1978. In Ultrasonography of Digestive Diseases (C. V. Mosby Company, Saint Louis).

Books received Receipt of the following books is acknowledged with thanks. They have been placed in the library of THE BRITISH INSTITUTE OF RADIOLOGY for the use of readers.

International Advances in Surgical Oncology. Vol.5. Ed. by Gerald P. Murphy. 1982. xii+404 pp. (Alan R. Liss, New York). £43.70 (DM.192.00) ISBN 0-8451-0504-3

1981 Year Book of Cancer. Ed. by R. L. Clark, R. W. Cumley and R. C. Hickey 463 pp. (Year Book Medical Publishers, Chicago/London). £31.50. ISBN 0-8151-1789-2

561