Carcinoma of unknown primary: Molecular tumor board-based therapy

A 36-year-old woman presented to an outside hospital with abdominal pain in July of 2021. A computerized tomography (CT) scan of the abdomen and pelvis showed multiple, hypodense lesions within the left hepatic lobe that were concerning for metastatic disease. The largest liver lesion (3.1 × 2.2 cm) was biopsied, and pathology was consistent with metastatic, moderately differentiated adenocarcinoma. Immunohistochemistry (IHC) stains performed for further characterization of the tumor were positive for cytokeratin 7 (CK7) and CDX2 and negative for CK20, HepPar-1, Napsin A, and GATA-3, consistent with adenocarcinoma. The differential diagnosis included primary cholangiocarcinoma versus metastatic carcinoma of upper gastroesophageal or pancreaticobiliary origin. Of note, it was also suspected that the patient had amoebiasis, and she was on treatment with antibiotics. She later transferred her care to our university hospital. A follow-up CT scan of the abdomen and pelvis performed a few weeks after the initial presentation demonstrated hypodense liver lesions involving the right and left lobes as well as peripancreatic and gastrohepatic lymph nodes, which were suspicious for malignancy. Because the liver biopsy could not identify the primary tumor conclusively, subsequent endoscopic ultrasound and fine-needle aspiration of the peripancreatic lymph node were performed almost a month after the initial presentation.

The treatment paradigm of advanced CUP has remained the same for several years now. Patients with CUP are offered platinum-based and/or paclitaxel-based cytotoxic therapies at most oncology centers in a frontline setting, but the median survival remains 6–15 months.¹⁹ Attempts to treat based on tissue-of-origin identification in general have not yielded a survival advantage compared with empiric chemotherapy.⁵³

CUP is almost an ideal tumor for the incorporation of NGS-based therapeutic matching. With this strategy, genomics is the diagnosis.⁵⁴ Importantly, prior studies have shown that tailored combinations of drugs matched to a majority of the patients' genomic, transcriptomic, and immunomic alterations can be given safely (generally by using initial dose reductions and titrating the doses to tolerance) and that enhanced degrees of matching correlate with improvements in all outcome parameters across multiple tumor types, including CUP.^{3, 5, 6} Even patients, such as ours, in whom tissue is not available for NGS can have their cancer's genomic status interrogated by evaluating NGS on ctDNA derived from a small vial of blood. Indeed, the incorporation of blood-based NGS has further transformed this field and opened doors to more accessible and less time-consuming diagnostic tools that were almost unimaginable a decade ago

Overall, CUP is a heterogenous group of cancers that harbor distinct, characterizable molecular alterations, many of which may be pharmacologically tractable. In the case presented, based on input from experts on our Molecular Tumor Board, a patient with advanced CUP was treated with the FGFR inhibitor pemigatinib and the checkpoint blockade (anti–PD-1 agent) pembrolizumab because her cancer harbored molecular FGFR2 fusion, amplification, and rearrangement and PD-L1 positivity. She has tolerated therapy well and has an ongoing partial response at 7 months. Further prospective trials of NGS-based strategies for CUP are warranted. ■

Aditya V. Shreenivas reports research funding from Natera and serves on the advisory board of Taiho Oncology, all outside the submitted work. Shumei Kato reports contracts with ACT Genomics, Sysmex, Konica Minolta, OmniSeq, and Personalis; speaker's fees from Bayer and Roche; and personal fees from CureMatch, Foundation Medicine, Medpace, NeoGenomics, and Pfizer outside the submitted work. Jason Sicklick reports grants and research funding from Amgen and Foundation Medicine; personal fees from Deciphera; honoraria from Bayer, Deciphera, Foundation Medicine, Hoffman-La Roche, and Merck; membership in the Foundation Medicine, QED Therapeutics, and Roche speakers' bureaus; and has stock and other ownership interests in Personalis, all outside the submitted work. Razelle Kurzrock reports research funding from Biological Dynamics, Boehringer Ingelheim, Debiopharm, Foundation Medicine, Genentech, Grifols, Guardant, Incyte, Konica Minolta, Medimmune, Merck Serono, Omniseq, Pfizer, Sequenom, Takeda, and TopAlliance; personal consulting, speaking, and/or advisory board fees from Actuate Therapeutics, AstraZeneca, Bicara Therapeutics, Biological Dynamics, Daiichi Sankyo Inc., EISAI, EOM Pharmaceuticals, Iylon, Merck, NeoGenomics, Neomed, Pfizer, Prosperdtx, Roche, TD2/Volastra, Turning Point Therapeutics, and X-Biotech; equity interest in CureMatch Inc., CureMetrix, and IDbyDNA; serves on the Boards of CureMatch and CureMetrix; and is a cofounder of CureMatch, all outside the submitted work. Jingjing Hu and Catherine Skefos made no disclosures.