|Home | About | Journals | Submit | Contact Us | Français|
A 31‐year‐old woman presented with altered bowel habits, black stools and profound weakness. Abdominal ultrasonography performed elsewhere had shown hepatic metastases. Oesophago‐gastroduodenoscopy had been reported as normal. A red blood cell labelled blood pool study had shown a slow bleed at the ileocaecal junction, caecum and/or ascending colon. Except for pallor, her general condition was good. There was no supraclavicular adenopathy or ascites. She underwent lower gastrointestinal endoscopy at our hospital and was detected to have a rectal polyp. Because the size and appearance of the polyp did not correlate with the symptoms, the possibility of a separate pathology was considered. Consequently, the patient underwent an upper gastrointestinal endoscopy, which showed friable ulcerated lesions in the third part of the duodenum. The endoscopic differential diagnosis included carcinoma, lymphoma and tuberculosis.
The rectal lesion was a juvenile polyp. The duodenal biopsy specimen showed ulceration with neutrophilic infiltrates. The lamina propria contained a neoplasm composed of confluent masses of polygonal to round cells with large, vesicular nuclei. Some of the cells possessed smudged nuclei; a few multinucleate giant cells were also present. There were moderate amounts of eosinophilic cytoplasm; some cells contained clear cytoplasm. Mitotic figures were scarce. There was no desmoplasia, haemorrhage, necrosis or calcification. The tumour cells expressed cytokeratin (1:50) strongly and diffusely and were negative for CD45 (1:50) and CD30 (1:20) (all antibodies from Dakocytomation, Carpinteria, CA, USA).
Initially, a diagnosis of a high grade undifferentiated carcinoma involving the duodenum was considered. However, the unusual clinical features of a massive gastrointestinal bleed occurring in an otherwise healthy woman and the presence of hepatic metastases without abdominal adenopathy on ultrasonography were contradictory to a histological diagnosis of a high grade epithelial neoplasm. Moreover, there were some unusual morphological features: the bulk of the neoplasm was deep in the lamina propria and there was no dysplasia in the overlying epithelium. Further, the smudged cells raised the possibility of a trophoblastic neoplasm ((figsfigs 1 and 22).). Further investigation showed a serum β‐human chorionic gonadotropin (β‐HCG) level of 11000 mIU/ml. The slides were re‐evaluated and a diagnosis of placental site trophoblastic tumour (PSTT) was made. The tumour cells expressed β‐HCG focally and human placental lactogen diffusely (both Dakocytomation, Ontario, Canada; predilute, kindly done by Dr R Chetty). The patient was re‐interviewed and we learnt that she had had an abortion recently. However, she refused further investigation or treatment and died a month later.
Primary duodenal choriocarcinoma as seen at autopsy and PSTT involving the duodenum and retroperitoneum in a known case of PSTT have been reported earlier.1,2 However, we are unaware of any reports of metastatic PSTT presenting primarily as a duodenal lesion. Choriocarcinoma usually shows haemorrhagic areas, with focal expression of human placental lactogen and strong expression of β‐HCG in the tumour cells. PSTTs possess sheets of tumour cells with diffuse human placental lactogen and focal β‐HCG expression. The moderate rise in serum β‐HCG levels in patients with PSTT, as in our case, correlates with the focal expression of β‐HCG in the tumour cells.3,4 The cells of epithelioid trophoblastic tumour are smaller than those of a PSTT and form cords and nests but no sheets; moreover, epithelioid trophoblastic tumour cells express human placental lactogen only focally. Differentiating between choriocarcinoma and PSTT is of importance because the former are exquisitely chemosensitive. PSTTs are relatively resistant to chemotherapy, though a few long term remissions to etoposide–methotrexate–actinomycin‐D or cisplatin–etoposide treatment have been reported.5
There are many lessons to be learnt from this unfortunate patient. Pathologists must always consider, especially in an unusual clinical setting, the possibility of a trophoblastic neoplasm, a potentially curable disease, before labelling a tumour as a high grade carcinoma or an incurable disease. For physicians, it underscores the fact that sound clinical judgement followed by good communication with the pathologist is the key to a correct diagnosis. Finally, it is a reminder that it is mandatory to take an obstetric history, even in apparently non‐obstetric cases.
Competing interests: None.