Get started
Work

One more piece of the puzzle? Adenosquamous carcinoma in a pancreatic mass.

Year: 2011
Type: article
Abstract: Adebajo and colleagues present a rare case of what appears to be a pancreatic metastasis from a primary adenosquamous carcinoma of the colon and, in doing so, provide a piece of the puzzle that is gastrointestinal tract adenosquamous carcinoma.1 Interestingly, the pancreatic mass—which was diagnosed via fine-needle aspiration on endoscopic ultrasound as well as a subsequent core biopsy—developed in a patient who had long-standing ulcerative colitis and in whom a T1N0M0 colonic adenocarcinoma had been removed 3 years earlier via total proctocolectomy. The pancreatic tumor was unresectable, but it responded to palliative chemotherapy consisting of gemcitabine hydrochloride (Gemzar, Lilly), docetaxel (Taxotere, Sanofi Aventis), and capecitabine (Xeloda, Hoffmann La Roche). The case study's reference section provides a broad sampling of articles on these rare tumors. Based on the cited literature, as well as tumor marker studies performed on the index patient, Adebajo and associates conclude that the pancreatic mass and clinically positive lymph nodes were metastatic from the previous, albeit early-stage, primary colorectal cancer.1 Several pieces of evidence support this conclusion: the proportionately high incidence of colorectal cancer as a source of pancreatic metastasis, the rarity of primary pancreatic squamous-cell cancer, the clinical finding of immunohistochemical similarities between the pancreatic and colonic tumors, and the retrospective identification of foci of squamous differentiation in the colonic tumor.2,3 However, in colorectal cancer, squamous and adenosquamous histologies are also quite uncommon, with a reported incidence of 0.1% in over 60,000 cases.4 Fewer than 100 squamous and adenosquamous colorectal cancers have been reported in the literature.5 In the largest series of pancreatic adenosquamous carcinomas to date, microscopic evaluation of Armed Forces Institute of Pathology (AFIP) specimens uniformly demonstrated dual differentiation toward adenocarcinoma and squamous-cell carcinoma, and all of the specimens were immunoreactive with keratin CK1 and AE1:AE3.6 Other keratin markers in the AFIP series were variably expressed, including CK5/6 (88%) and CK20 (26%). In the case report by Adebajo and coworkers, immunohistochemical staining of the patient's tumor was strongly positive for CK5/6 but negative for CK7 and CK20.1 Adebajo and colleagues provided a brief but concise summary of the known pathophysiologic hypotheses for the development of gastrointestinal squamous and adenosquamous tumors.1 These theories include the proliferation of uncommitted, or basal, cells in response to mucosal injury, as well as squamous metaplasia of glandular epithelium as a sequela of chronic inflammation. A further hypothesis was that squamous differentiation may arise within adenomas, based on the observation of squamous differentiation in a small number of benign colonic adenomatous polyps.7 The report by Adebajo and colleagues provides an important addition to the literature surrounding these rare tumors, as well as to the reporting of immunohistochemical and serum-based tumor markers.1 Of particular interest is p16 (CDKN2A)-positive staining of the primary and presumed metastatic tumor, as well as the initially high level of serum CA 19-9, which returned to normal after chemotherapy. Re-analysis of the original T1N0M0 colonic tumor showed foci with features of squamous differentiation and immunohistochemical positivity for p16 (CDKN2A), a cell-cycle checkpoint protein that binds to cyclin-dependent kinases, resulting in cell-cycle arrest at the G1/S checkpoint.8 In keeping with the favorable response to chemotherapy noted in the report by Adebajo and associates, a recent meta-analysis showed a somewhat tenuous, but favorable, prognostic significance for p16 in pancreatic tumors that is consistent with its tumor suppressor function; however, the authors of the meta-analysis were quick to point out that the small number of eligible studies precludes any meaningful conclusions regarding p16 expression as a reliable marker of prognosis.1,9 Conversely, in a more robust database of 902 colorectal cancers from 2 independent cohort studies, p16 (CDKN2A) promoter hypermethylation was present in only 30% of tumors. Unlike the data available for pancreatic tumors, and despite p16's well-established role in carcinogenesis, p16 (CDKN2A) staining in colorectal cancer was not independently associated with patient prognosis.10 In addition to the data on p16, data on CA 19-9 are interesting. None of the articles cited by Adebajo and coworkers discussed CA 19-9.1 In the previously mentioned AFIP series, immunohistochemistry was positive for CA 19-9 in 89% of cases; however, serum CA 19-9 data were not present. It was noted that the aggressive behavior of adenosquamous carcinoma observed in the AFIP study was analogous to that of anaplastic carcinoma, leading the authors to hypothesize that these 2 entities are related and represent different degrees of ductal adenocarcinoma differentiation.6 Another observation relevant to the report by Adebajo and associates is the conclusion made by the AFIP study authors; after noting that some pathologists require that 30% of a specimen contain cellular elements, the AFIP study authors concluded that primary tumors of the pancreas that show any degree of definitive, malignant, squamous-cell differentiation on routine sectioning should be considered adenosquamous carcinoma.1,6 By such a definition, presumably the pancreatic tumor in the report by Adebajo and associates would be considered adenosquamous, but the colonic tumor would not.1 Although the report by Adebajo and coworkers focuses on the question of whether the pancreatic tumor arose de novo from the pancreas or was metastatic from a T1M0N0 primary colonic tumor, there is, of course, a third possibility: a biliary-tract primary tumor.1 Elevated preoperative serum CA 19-9 levels have been associated with a high risk of recurrence after curative resection of biliary-tract cancer.11 The recurrence rate was significantly higher when the baseline CA 19-9 serum level was at least 55 U/mL (hazard ratio, 3.282; 95% confidence interval, 1.684–6.395; P<.001).11 Finally, the role of chronic inflammation and the development of adenosquamous tumors should be discussed. As noted by Adebajo and colleagues, the incidence of squamous and adenosquamous tumors was 0.1% in a large series of colorectal tumors; in patients with preexisting ulcerative colitis, the incidence increased to 9.5%.1,12 The progression of squamous metaplasia to squamous-cell carcinoma has been documented in the setting of ulcerative colitis.13 In the setting of chronic pancreatitis, squamous metaplasia of the pancreatic ductal epithelium is noted in adjacent ducts in 4% of adenocarcinomas.14 Unlike anal cancer, there appears to be no association between human papillomavirus infection and squamous-cell carcinoma of the colon and upper rectum (>8 cm above the dentate line).15 In summary, the report by Adebajo and colleagues raises more questions than it answers, as should all such case reports.1 From what organ and cell lineage did the tumor in question arise? What is the role of preexisting chronic inflammation? Are either p16 or CA 19-9 pathogenetic? Is there a prognostic significance to any of the studied tumor markers? Does the organ of origin make a difference in determining treatment? It is fair to conclude that Adebajo and colleagues have not provided any definitive answers; however, they have contributed one more important piece of the puzzle.1
Language: English
Field: Medicine
Open Access status: green
OpenAlex ID: W101108581
FWCI:
Cites: 3
Cited by: 1
Related to: 10