Cytoreduction surgery in recurrent ovarian cancer

Decision-making based on patient selection

James Fleck, MD, PhD: Anticancerweb 16 (05), 2022


Ovarian cancer consists of a heterogeneous group of malignancies, including tumors located not only in the ovaries, but also in the fallopian tubes and peritoneum. They are predominantly (90%) epithelial tumors and the majority (70%) are classified, according to morphological, immunohistochemical and molecular criteria, as high-grade serous carcinoma (HGSC). This tumor has a high Ki67 proliferative index, diffusely express p53 and is associated with BRCA1 and BRCA2 germline mutations in up to 10% of the patients. Unfortunately, HGSC pathogenesis has not been fully clarified. Currently, the most accepted theory supports an origin in tissues embryologically derived from Müllerian ducts, leading to secondary involvement of the ovaries. This hypothesis is reinforced by the identification of serous tubal intraepithelial carcinoma in about 40% of patients diagnosed with serous ovarian carcinoma. Historical data used to support ovarian Müllerian cortical inclusion cysts (MIC) as a precursor lesion. Nowadays, the fallopian tube has begun to emerge as the most likely or possibly exclusive source of HGSC. The supposed multicentric characteristic of the tumor is the main reason for bilateral hysterosalpingo-oophorectomy (BHSO) as a radical surgical approach in the setting of a curative intent intervention. Unfortunately, most of the HGSC are diagnosed in advanced stages (III and IV), early spreading to the peritoneum and omentum, due to celomic implants. Additionally, the tumor could progress through lymphatic and blood dissemination. Nevertheless, debulking surgery including BHSO accompanied by omentectomy and resection of all macroscopic disease is still the standard operation procedure. After optimal debulking surgery, patients without known BRCA1 and BRCA2 mutations are additionally treated with combination chemotherapy (carboplatin + paclitaxel) plus concomitant and maintenance bevacizumab. When these patients harbor a homologous-repair deficient (HRD) tumor, maintenance therapy should be switched to niraparib. When HGSC expresses either somatic or germline BRCA mutation the best maintenance treatment is olaparib. Patients are followed-up using thorax, abdominal and pelvic CT plus CA-125 as a tumor marker. Unfortunately, 80 to 85% of advanced HGSC will relapse, after primary intervention. At this point, treatment algorithm is dictated mainly by the time elapsed between the last course of chemotherapy and the first sign of progressive disease, which is usually called platinum-free interval (PFI). First evidence for PFI as a prognostic and predictive factor came in a retrospective review in the early nineties. Since that time, it has been widely accepted that patients with PFI of six months or longer should be considered platinum-sensitive disease (PSD),whereas patients with PFI of less than 6 months should be considered platinum-resistance disease (PRD). Patients expressing PSD have high probability of responding again to the same previous systemic intervention, including combined platinum chemotherapy and maintenance bevacizumab, niraparib or olaparib, according to genomic profile. Ironically, these patients are also better candidates for secondary cytoreduction surgery (SCS).

A potential benefit obtained with SCS is closely related to the chance of achieving a complete gross resection (CGR). Currently, three models have been used to predict the results of surgery. The Memorial Sloan Kettering (MSK) criterion was based on a retrospective analysis of 153 patients undergoing SCS. The objective was to achieve residual disease ≤ 0.5 cm. In the MSK predictive model, the SCS recommendation is supported by a disease-free interval range (6 to >30 months), presence of single versus multiple recurrence sites, and peritoneal carcinomatosis. The AGO (Arbeitsgemeinschaft Gynäkologische Onkologie) is a European predictive model supported by the results observed in the sequential multicenter DESKTOP trials. The AGO criteria point to the prognostic importance of PSD, performance status (ECOG 0), absence of ascites in preoperative images, and no residual disease after primary surgery. The Tian model retrospectively evaluated data from nine SCS studies, published before 2009. It included MSK and AGO patients, for a total of 1075 patients. A multivariate logistic regression analysis allowed the identification of six predictive parameters of CGR, creating a risk scoring criterion. The sum of the risk scores allowed the categorization of patients into low risk (≤4.7 points) and high risk (>4.7 points). The proportion of CGR in the low-risk group was 53.4%, while in the high-risk group it was 20.1%. An external validation of the Tian model identified a sensitivity of 83.3% and a specificity of 57.6%. Tian risk model scoring criterion is shown in the table below.



Three prospective randomized trial, evaluating the benefit of SCS, have been published. In November, 2019, The New England Journal of Medicine reported the GOG-213 trial, randomly assigning 485 patients to receive or not receive SCS. Both arms included patients with recurrent PSD, who have undergone complete response to front-line chemotherapy and have been classified as an investigator-determined resectable disease (CGR). Unfortunately, some accrued patients considered not fit for CGR were enrolled in the chemotherapy-only arm. Platinum-based chemotherapy either with gemcitabine or paclitaxel were used at the discretion of the investigator. Prognostic impact of concomitant and maintenance bevacizumab was also an objective of the study design. CGR was obtained in 64% of the SCS randomized patients. The median overall survival was 50.6 months for the SCS arm and 64.7 months for the chemotherapy-only arm, indicating no survival benefit associated to SCS (HR = 1.29 P = 0.08). Additionally, 9% of patients in the SCS arm experience surgical morbidity. In March, 2021, The New England Journal of Medicine published an interim analysis of the Chinese SOC-1 trial, where 357 PSD-recurrent patients were randomly assigned to receive SCS + chemotherapy or chemotherapy-alone. All patients have potential resectable disease based on risk model and PET-CT imaging.  CGR was obtained in 76.7% of the SCS randomized patients. The median progression-free survival was 17.4 months in the SCS arm and was significantly longer than the 11.9 months observed in the chemotherapy-alone arm (HR = 0.58). The median follow-up at the time of the publication was only 36 months, which made the data immature to address overall survival differences. In December, 2021 The New England Journal of Medicine published the results of the DESKTOP-III trial, which randomly assigned 407 recurrent-PSD patients to SCS + platinum-based chemotherapy or platinum-based chemotherapy alone. Patients have to achieve all AGO good prognostic factors to be eligible. A CGR was achieved in 74.5 % of the patients in the SCS arm. The median overall survival was 53.7 months in the SCS arm versus 46 months in the chemotherapy-only arm (HR = 0.75, P = 0.02). A benefit from surgery was observed in all prognostic subgroups. A recently published meta-analysis of the three prospective randomized trials pointed to a potential benefit associated with SCS based on superior PFS and increased overall survival particularly seen in the CGR-SCS subpopulation. Ultimately, in January, 2022, the Journal of Clinical Oncology published a meta-analyses of 80 eligible studies, including 2805 patients who underwent SCS. There were important limitations. Fourteen  studies were located outside the funnel plot, there was strong heterogeneity (P < 0.0001 and I2 = 86%) and a long period elapsing between publications, ranging from 1995 to 2021. The results of the univariable linear regression model showed that higher proportion of complete or optimal cytoreduction, more recent publication and more advanced age were significantly associated to longer overall survival. The meta-analysis only confirms what has historically been learned as a benefit of SCS. But it definitely does not indicate who are the best candidate patients for secondary surgical approach. We are still struggling with patient selection.


References:

1.     Marjanka JJM Mingels, Maaike APC van Ham, Ineke M de Kievit, et al: Müllerian precursor lesions in serous ovarian cancer patients: using the SEE-Fim and SEE-End protocol, Modern Pathology 27, 1002–1013, 2014

2.     Caixia Jiang and Zhengyu Li: Prediction Models for Complete Resection in Secondary Cytoreductive Surgery of Patients with Recurrent Ovarian Cancer, Frontiers in Oncology, 23rd September, 2021

3.     Coleman RL, Spirtos NM, Enserro D: Secondary Surgical Cytoreduction for Recurrent Ovarian Cancer, N Engl J Med. 381(20):1929, 2019

4.     Shi T, Zhu J, Feng Y, et al: Secondary cytoreduction followed by chemotherapy versus chemotherapy alone in platinum-sensitive relapsed ovarian cancer (SOC-1): a multicentre, open-label, randomised, phase 3 trial, Lancet Oncol 22(4):439, 2021

5.     Harter P, Sehouli J, Vergote I, et al: Randomized Trial of Cytoreductive Surgery for Relapsed Ovarian Cancer, N Engl J Med 385(23):2123, 2021

6.     Marchetti C, Fagotti A, Tombolini V, Scambia G, De Felice F. The Role of Secondary Cytoreductive Surgery in Recurrent Ovarian Cancer: A Systematic Review and Meta-Analysis. Ann Surg Oncol, 28(6):3258–63, 2021

7.     Min-Hyun Baek, Eun Young Park, Hyeong In Ha, et al: Secondary Cytoreductive Surgery in Platinum-Sensitive Recurrent Ovarian Cancer: A Meta-Analysis, J Clin Oncol, February 21st, 2022

8.     Photo by Andrea De Santis on Unsplash