Systemic Therapy for Non-Serous Ovarian Carcinoma

Yanin Chávarri-Guerra; Eduardo González-Ochoa; Héctor De-la-Mora-Molina; Enrique Soto-Perez-de-Celis


Chin Clin Oncol. 2020;9(4):52 

In This Article

Mucinous Epithelial Ovarian Carcinoma

Mucinous epithelial ovarian carcinoma (mEOC) accounts for 3% of EOC.[3] Median age at diagnosis is 53 years, and tobacco smoking is considered an associated clinical risk factor.[12] Between 65–80% of patients present at localized stages, since these tumors are usually very large and cause symptoms early. The 5-year overall survival (OS) for stage I tumors is 92%, while the median OS in advanced stage ranges from 12–33 months.[8] These tumors evolve in a stepwise fashion from benign epithelium to a pre-invasive lesion, to carcinoma, similar to the development of colorectal cancer. In addition, they commonly have mutations in various genes, including KRAS, HER2 and TP53, among others.[19,20] Mucinous neoplasms generally display complex glandular arrangements with areas of stromal invasion, expressing gastrointestinal markers such as CK7, CK20, CDX2, which makes it difficult to differentiate them from metastatic tumors from the gastrointestinal tract.[21] Clinical suspicion of a mucinous metastasis, rather than a primary mEOC, includes tumor size <10 cm, bilateral tumors, peritoneal spread, evidence of metastatic lesions, or a combination of these findings.[12] In these cases, a comprehensive workup must be performed to rule out an occult gastrointestinal cancer, including colonoscopy, upper gastrointestinal endoscopy, and endoscopic ultrasonography.[22]

The management of mEOC is primarily surgical. In stage I, fertility-sparing surgery is appropriate for young patients, including unilateral salpingo-oophorectomy with peritoneal staging procedures (cytology, peritoneal biopsies, and omentectomy).[23] The risk of recurrence is lower than that reported for women with stage I serous cancers (6% vs. 20%).[24] In older patients, bilateral salpingo-oophorectomy is preferred. In advanced stages, debulking surgery with the objective of a macroscopically complete resection should be attempted, followed by the administration of systemic treatment.

In contrast with that seen at early stages, the prognosis of women with advanced mEOC is worse than that of patients with other common subtypes [hazard ratio (HR) for death 2.81; 95% CI, 2.47–3.21],[25] due to a worse response to chemotherapy. A case-control study comparing the effectiveness of platinum-based regimens between patients with mEOC and those with other EOC subtypes found that response rates (RR) were significantly lower in mEOC (26.3% vs. 64.9%, P=0.01) with a significantly shorter OS of 12 vs. 37 months (P<0.001).[26] Alexandre et al.[27] retrospectively collected data from five randomized clinical trials (RCT) including 1,118 patients with advanced EOC [International Federation of Obstetrics and Gynecology (FIGO) stages IIB–IV] mostly treated with paclitaxel-carboplatin based chemotherapy, of which 54 (5%) had mEOC. Compared with non-mucinous tumors, RR were found to be lower (60% vs. 80%, P<0.001), and progression free survival (PFS) and OS were shorter (11 vs. 17.5 months, P=0.002; and 21.6 vs. 47.2 months, P<0.001 respectively). A worse prognosis was also observed among patients with FIGO stage IV and optimally debulked stage IIB–IIIC disease.[27] In contrast, an exploratory subgroup analysis of the ICON3 RCT[28] that enrolled 2,074 patients with advanced ovarian cancer (7% mEOC), reported no difference in OS/PFS for paclitaxel-carboplatin versus either carboplatin or cisplatin-cyclophosphamide-doxorubicin between mEOC and other more common types of EOC.

Given the histologic similarities between primary mEOC and gastrointestinal carcinomas, an attractive option is the use of chemotherapy regimens proven to work in gastrointestinal malignancies. The Gynecology Oncology Group (GOG) RCT 0241[13] included exclusively patients with mEOC and evaluated four different treatment regimens (paclitaxel plus carboplatin or oxaliplatin plus capecitabine, with or without bevacizumab) in 50 patients with FIGO stage II–IV disease or with recurrence after treatment for a stage I tumor. After a median follow-up of 59 months, median PFS was 16.4 months, with no difference between treatment arms. Median OS was 27.8 months (33.9 for capecitabine/oxaliplatin vs. 27.7 months for paclitaxel/carboplatin, HR 0.77, P=0.48), with no benefit added by bevacizumab.[29]

Based on this evidence, the National Comprehensive Cancer Network (NCCN) guidelines recommend surgery alone for stage IA or IB mEOC and adjuvant platinum-based chemotherapy (either paclitaxel/carboplatin or oxaliplatin with fluorouracil or capecitabine) for more advanced disease.[14]

Genomic profiling studies have shown that the mutational landscape of mEOC is different from serous carcinoma.[18] Mucinous neoplasms are not associated with BRCA mutations or defects in homologous recombination, making them unlikely to benefit from poly ADP ribose polymerase (PARP) inhibitors. However, they frequently display mutations or amplifications that might be targetable. Epidermal growth factor receptor (EGFR) amplification or mutations in BRAF, FGFR, or STK11 have been detected in human epidermal growth factor receptor-2 (HER2) negative tumors with wild-type KRAS, suggesting that these tumors may be responsive to targeted therapies.[12,30] Just as the absence of a KRAS mutation identifies a subset of patients with colorectal cancer who are more likely to benefit from EGFR-inhibition, preclinical studies have shown that cetuximab inhibits proliferation in mEOC cell lines with wild type KRAS, whereas it has no antitumor effect in a model of KRAS-mutated mEOC.[31]

About 15–20% of mEOC is associated with defects in DNA mismatch-repair system resulting in high microsatellite instability (MSI-h).[32] These tumors have a high mutational burden and dense immune infiltrates, and characteristically respond to immune checkpoint inhibition. Although there are no specific studies of immunotherapy in mEOC, 15 patients with MSI-h ovarian cancer (representing 6.4% of the entire patient population) were included in the phase II KeyNote-158 trial[33] that assessed the efficacy of pembrolizumab in non-colorectal MSI-h tumors. In these patients, the RR to pembrolizumab was 33% (3 complete and 2 partial responses), with a median PFS of 2.3 months (95% CI, 1.9–6.2 months).[34]