Mallika Tewari, Hari S Shukla
Department of Surgical Oncology, Institute of Medical Sciences, Banaras Hindu University. Varanasi, UP, India
Received April 16th, 2009
Carcinoma; Diagnosis; Endosonography; Pancreatic Neoplasms; Tuberculosis
Although tuberculosis is endemic in this part of the world, isolated pancreatic tuberculosis is extremely rare. Thus, all lesions are first suspected as carcinoma and not tuberculosis unless the clinical features are suggestive of tuberculosis, for example, a history of miliary tuberculosis with enlarged peripancreatic nodes especially in an immunocompromised host. As our patient had none of these findings, we investigated her as one with possible periampullary carcinoma.
Currently, the cornerstone for diagnostic evaluation of a pancreatic tumor is multiphase, multidetector helical axial computed tomography (CT) performed according to a defined pancreatic protocol (i.e. triphasic cross sectional imaging and thin slices) [1, 2]. It has an accuracy of approximately 80% for assessing resectability preoperatively [1, 2]. Endoscopic ultrasound (EUS) is usually indicated as an additional imaging technique when no tumor is seen on CT and there remains a high index of suspicion of an underlying malignancy . EUS is also helpful in distinguishing benign and malignant strictures of the bile duct when no mass lesion is apparent. However, EUS is highly operator-dependent and is recommended for use only in experienced hands. Magnetic resonance cholangiopancreatography (MRCP) is another noninvasive examination which helps in the diagnosis of pancreatico-biliary tumors [1, 2, 3].
An endoscopic retrograde cholangiopancreatography (ERCP) prior to resection is routinely avoided for fear of its associated potential complications, such as pancreatitis, cholangitis, bleeding and perforation making eventual surgery difficult. However, this investigation is of value especially when CT scan findings are equivocal since fewer than 3% of patients with pancreatic carcinoma have normal findings . Though it is difficult to differentiate benign and malignant strictures or stenosis, severe stenosis and marked proximal dilatation more often indicate malignancy .
Another important point is that tissue diagnosis is not a prerequisite before routine resection. A suspicious lesion on imaging (MRCP/CT), even without a mass lesion should be treated with resection [1, 2]. Tissue diagnosis is required when a patient is not a candidate for resection or is enrolled in a neoadjuvant chemo/radiation protocol.
Our patient had features suggestive of a periampullary lesion, for example, a growth in the ampulla, irregular lower bile duct stricture with proximal dilatation and a history of jaundice. A periampullary tumor often sloughs off relieving the jaundice temporarily. A biopsy was attempted twice, once during endoscopy and another time while doing an ERCP. Again, as tuberculosis was not suspected either clinically or on histopathology of the biopsy specimens, the biopsies were thus not analyzed by polymerase chain reaction (PCR) for Mycobacterium tuberculosis. It is well known, as we have already stated in the discussion of our paper, that preoperative endoscopic biopsy is rarely diagnostic for tuberculosis . It was only after a granulomatous inflammation was revealed by histopathology of the wide local excision specimen that we decided to investigate it further by PCR.
But, indeed, if the suspicion of tuberculosis is strong, repeated aspirations guided by a skilled endosonologist may improve the diagnostic yield. But isolated lesions of the ampulla, such as the one presented in the paper, occur only rarely and tuberculosis remains a diagnosis of exclusion.
The authors have no potential