To the Editor: With great interest we read the case report “Clinical Sequencing Contributes to a BRCA-Associated Cancer Rediagnosis That Guides an Effective Therapeutic Course” by Chapman et al,1 which shows the application of next-generation sequencing to cancer diagnostics. The authors present a case of a woman with an abdominal tumor that was initially diagnosed as pancreatic adenocarcinoma. However, subsequent mutational profiling and comparison with trans-cancer data from The Cancer Genome Atlas prompted the authors to rediagnose this tumor as ovarian adenocarcinoma, which was eventually confirmed through further clinicopathologic workup.
Although we find the idea of using molecular profiling to improve the accuracy of tumor diagnostics appealing, we would like to comment on the authors' statements that histologic diagnoses are often incorrect and that, in the case they present, molecular profiling was necessary to render the correct diagnosis of ovarian adenocarcinoma. Despite the suggestive radiologic findings, the small cytological image the authors provide in their paper clearly shows 3-dimensional papillary cell aggregates primarily suggestive of papillary adenocarcinoma of the lung or the ovary,2,3 whereas a pancreatobiliary primary appears less likely from a pathologic point of view. Although we contend that a definitive diagnosis would not rely on morphology alone but would require immunohistochemistry, the morphology should have prompted considering ovarian carcinoma as one of the top differential diagnoses. Further tumor typing through immunohistochemical staining of the transcription factors WT1 and PAX8 (which, according to Chapman et al,1 was performed afterward to confirm the molecularly based suspicion of ovarian origin) would then have provided strong support for ovarian and against pancreaticobiliary adenocarcinoma according to the criteria of current WHO tumor classification.3
To evaluate how a larger group of pathologists would assess this particular case, we performed a Web-based survey among pathologists at the Institute of Pathology of the Charité Medical University Berlin. In the survey, we presented the clinical information of a 54-year-old female patient with abdominal pain and the cytological image available in the paper, including the immunoprofile used at the initial diagnosis. We offered the answer choices: (a) gastric cancer, (b) pancreatico-biliary cancer, (c) colon cancer, (d) ovarian cancer, and (e) endometrial cancer. The 16 participating pathologists were unaware of the case report (ie, blinded to the context of the data provided). Twelve pathologists (75.0%) chose ovarian carcinoma as the most likely diagnosis, whereas 2 pathologists (12.5%) believed the tumor cells to be of gastric origin, and the remaining 2 (12.5%) favored a pancreatobiliary primary. Most participants commented that they would have requested further immunostains, including WT1 and PAX8. Therefore, although most pathologists had already suspected the correct diagnosis based on morphology alone, they—and also the pathologists in favor of gastric or pancreaticobiliary carcinoma—would have used additional immunostains that would have led them to the same (correct) diagnosis.
Finally, although we strongly agree with Chapman et al1 that molecular profiling, in addition to providing important clinical information on actionable mutations beyond histology, might in certain morphologically and immunohistologically ambiguous cases help classify tumors in the future using the approach they present, we would like to point out that correct tumor typing in the present case would have been rather straightforward with conventional morphological assessment and a few additional standard immunostains without any further molecular profiling. Although it is evident that molecular profiling will play an increasing role in cancer precision medicine in the coming years,4–7 we believe that these time-consuming and costly methods should be carefully integrated with the current standards of histo-/cytomorphological and immunohistochemical workup to optimize diagnostics and patient care.
Call for Correspondence
JNCCN is committed to providing a forum to enhance collaboration between academic medicine and the community physician. We welcome comments about the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines), articles published in the journal, or any other topic relating to cancer prevention, detection, treatment, supportive care, or survivorship.
Please send correspondence to JNCCN.edmgr.com.
Letters should be no more than 400 words, with no more than 5 references if included. Please include the full names, degrees, and affiliations of all letter authors and a phone number or e-mail address for contact.
Letters are considered for publication as space allows. NCCN reserves the right not to publish correspondence for any reason it deems appropriate. All letters are subject to editing and/or abridgment.
Chapman JS, Asthana S, Cade L et al.. Clinical sequencing contributes to a BRCA-associated cancer rediagnosis that guides an effective therapeutic course. J Natl Compr Canc Netw 2015;13:835–845.
Travis WD, Brambilla E, Burke AP et al.. WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart, 4th ed. Lyon, France: IARC Press: 2014.
Kurman RJ, Carcangiu ML, Herrington CS, Young RH. WHO Classification of Tumours of Female Reproductive Organs, 4th ed. Lyon, France: IARC Press: 2014.
Klauschen F, Heim D, Stenzinger A. Histological tumor typing in the age of molecular profiling. Pathol Res Pract 2015;211:897–900.
Stenzinger A, Weichert W, Lennerz JK, Klauschen F. Basket trials: just the end of the first quarter. J Clin Oncol 2015;33:2823–2524.
Heim D, Budczies J, Stenzinger A et al.. Cancer beyond organ and tissue specificity: next-generation-sequencing gene mutation data reveal complex genetic similarities across major cancers. Int J Cancer 2014;135:2362–2369.