Ovarian squamous cell carcinoma: clinicopathological features, prognosis and immunotherapy outcomes

Ovarian squamous cell carcinoma (SCC) is a rare malignant tumor comprising less than 1% of primary ovarian malignant tumors [ 1 ]. The majority of these tumors arise from malignant transformation of teratoma, and less commonly, from the progression of ovarian endometriosis or the differentiation from other ovarian tumors, such as Brenner’s tumor, endometrioid adenocarcinoma, and mixed mesodermal tumor [ 2 ]. Compared with other epithelial ovarian cancers, the prognosis of SCC is inferior [ 3 ]. The 5-year cancer-specific survival rate was 86% for stage I, and 54.3%, 36.3%, and 2.8% for stages II, III, and IV, respectively [ 4 ]. Given the rarity of the pathologic type, most of the current literature is limited to case reports and small case series studies. There are no standardized treatment strategies for ovarian SCC, and current treatment typically follows the principles used for epithelial ovarian cancer. The typical approach involves complete debulking surgery and adjuvant chemotherapy [ 3 ]. Similar to other non-human papillomavirus SCCs, ovarian SCC reported a high frequency of mutations in the TP53 , CDKN2A , and PIK3CA genes [ 5 ], along with a high tumor mutation burden (TMB) of 13.4-25.7mutations/Mb [ 6 ]. Since the SCC of lung [ 7 ], head and neck [ 8 ], and cutaneous [ 9 ] benefit from immunotherapy, ovarian SCC with a high TMB level may also be a promising candidate for immunotherapy. We reviewed ovarian SCC cases from the past 2 decades at our institution, reporting the molecular characteristics and treatment outcomes of immune checkpoint inhibitors (ICIs).

A total of 42 patients were included in this study, with 1 patient lost to follow-up. Table 1 summarized the clinicopathological features of all patients and detailed information for each patient was provided in Table S1 . The median age at diagnosis was 51.5 years (range, 23–74), and 14.3% were nulliparous. The most common presenting symptoms were abdominal pain (54.8%), increased abdominal girth (19.0%) and abdominal bloating (14.2%). The median tumor diameter was 10 cm (range, 4–25). The elevated tumor markers mainly included carbohydrate antigen (CA 125, CA19-9, and squamous cell carcinoma antigen (SCC-Ag), with a median value of 53.8 U/mL, 40.1 U/mL, and 9.4 ng/mL, respectively. Patients were diagnosed with FIGO stage I (n=15, 35.7%), II (n=7, 16.7%), III (n=14, 33.3%), and IV (n=6, 14.3%), respectively. Teratoma was the most common precursor lesion (54.8%), followed by endometriosis (14.3%), Brenner’s tumor (4.8%) and endometrioid adenocarcinoma (2.4%). In 23.8% of patients, no clear precursor lesion was identified. Values are presented as number (percentile) or median (range). BSO, bilateral salpingo-oophorectomy; CA, carbohydrate antigen; CEA, carcinoembryonic antigen; CRS, cytoreduction surgery; CT, chemotherapy; FIGO, International Federation of Gynecology and Obstetrics; IDS, interval debulking surgery; IT, immunotherapy; NA, not applicable; NACT, neoadjuvant chemotherapy; NOS, not otherwise specified; RT, radiotherapy; SCC, squamous cell carcinoma; SCC-Ag, squamous cell carcinoma antigen; TT, targeted therapy; USO, unilateral salpingo-oophorectomy. Five patients underwent whole exome sequencing or targeted exome sequencing. The median TMB was 10.00 mutations/Mb (range, 7.28–46.86). One patient was identified with a germline heterozygous mutation in FANCC (c.1330-1G>A). All patients had at least 1 somatic pathogenic or likely pathogenic variant, including PIK3CA , BRCA2 , TP53 , TERT , ARIDIB , NF2 , PTEN , CDKN2A , FAT1 , FBXW7 , PBRM1 , and PRKDC . Somatic PIK3CA mutations were mostly noted (4/5, 80.0%), with 3 patients showing missense mutations (variant allele frequencies ranging from 5.3% to 44.1%) and 1 patient showing gene amplification (copy number of 2.95). One patient exhibited PTEN somatic copy number reduction at 10q23 (copy number 1.2). Nineteen patients (45.2%) underwent primary debulking or staging surgery at initial diagnosis. Five patients (11.9%) with advanced disease received 1–4 cycles of platinum-based neoadjuvant chemotherapy, followed by interval debulking surgery. Seventeen patients (40.5%) underwent salpingo-oophorectomy or ovarian cystectomy, with or without hysterectomy, and 12 of them (70.5%) received subsequent re-staging surgery. Three patients underwent fertility-sparing surgery: 1 received fertility-sparing staging surgery, and the other 2 underwent unilateral salpingo-oopherectomy with contralateral ovarian biopsy. Additionally, 27 patients (64.3%) underwent lymphadenectomy. Thirty-nine patients (92.9%) received adjuvant chemotherapy, including chemotherapy alone (64.3%), chemotherapy with radiotherapy (14.3%), chemotherapy with targeted therapy (bevacizumab) (2.4%), chemotherapy with ICIs (9.5%), and chemotherapy with both targeted therapy (bevacizumab) and ICIs (2.4%). The predominant first-line chemotherapy regimen was paclitaxel combined with platinum (29/39), followed by platinum with cyclophosphamide (6/39), carboplatin (1/39), cisplatin with vincristine and bleomycin (1/39), cisplatin with etoposide and ifosfamide (1/39), and cisplatin with fluorouracil (1/39). Table 2 summarized the molecular characterization and treatment information of patients treated with ICIs. Five patients received ICIs combined with chemotherapy as first-line postoperative adjuvant therapy. One patient with stage IC continued maintenance therapy with programmed death-1 (PD-1) inhibitor (sintilimab) after 3 cycles of chemotherapy combined with immunotherapy, and the remaining 4 patients with stage IIB–IVB received 6 cycles of chemotherapy combined with PD-1 inhibitor (tislelizumab) followed by maintenance therapy with ICIs. The median duration of maintenance therapy was 13 cycles (range, 10–24), and 2 patients were currently continuing the maintenance therapy. HRD, homologous recombination deficiency; irAE, immune-related adverse event; MSI, microsatellite instability; MSS, microsatellite stable; NA, not applicable; RCRS, re-cytoreductive surgery; TMB, tumor mutation burden. Among the 22 patients with progression or relapse, 7 (31.8%) underwent secondary cytoreductive surgery, 6 (27.3%) received systemic treatment and 1 received radiotherapy alone. Four patients with second recurrence underwent surgery, radiotherapy, surgery combined with chemotherapy, and chemotherapy combined with immunotherapy, respectively. Two patients received ICIs after relapse: One stage IA patient was treated with cisplatin, fluorouracil, and sirolimus after pelvic recurrence. After a second relapse of pulmonary metastasis, the patient received 6 cycles of chemotherapy combined with PD-1 inhibitor (tislelizumab) and 30 cycles of immune maintenance therapy, showing no signs of disease at 69.3 months of follow-up. Another stage IC patient underwent secondary cytoreductive surgery followed by 1 cycle of etiliricon, cisplatin and PD-1 inhibitor (toripalimab) treatment, but died after disease progressed, with an OS of 8.8 months. The median follow-up time was 23.2 months (range, 0.9–250.5). The median OS was 42.0 months, and the 1- and 5-year OS rates were 73.7% and 48.7%, respectively ( Fig. 1A ). Nine patients had persistent or progressive disease, and 14 patients experienced recurrence (10 relapsed once and 4 relapsed twice). The median PFS was 26.9 months, and the 1- and 5-year PFS rates were 57.5% and 43.8%, respectively ( Fig. 1B ). (A) OS. (B) PFS. OS, overall survival; PFS, progression-free survival. OS differed among patients with different prodromal lesions (p=0.030; Fig. 2A and B ). Patients with SCC associated with endometriosis or endometrioid adenocarcinoma had an inferior prognosis, with 71.4% (5/7) dying within 1 year. Patients with stage I had a better OS (p=0.022; Fig. 2C and D ). A cohort of 15 stage I patients showed superior prognosis for stage IA compared to stage IC. All 9 stage IA patients survived, with only 1 experiencing recurrence. In contrast, 3 of the 6 stage IC patients died ( Fig. 2E and F ). (A) OS according to pathological types of prodromal lesions. (B) PFS according to pathological types of prodromal lesions. (C) OS according to FIGO stage. (D) PFS according to FIGO stage. (E) OS of patients with FIGO stage I. (F) PFS of patients with FIGO stage I. FIGO, International Federation of Gynecology and Obstetrics; NOS, not otherwise specified; OS, overall survival; PFS, progression-free survival; SCC, squamous cell carcinoma. The median follow-up for the 5 patients (patient #37–40, 42) who received ICIs as first-line adjuvant treatment was 20.1 months (range, 9.5–25.1). One patient (patient #40) with stage IVB progressed after 12 cycles of PD-1 inhibitor (tislelizumab) maintenance therapy, while the others (stage IIB, IIIB, IIIC, and IC, respectively) showed no signs of relapse. Survival analysis indicated a non-significant trend towards better OS and PFS with ICIs (p=0.134 and p=0.202, respectively; Fig. 3A and B ). Six patients were assessed for immune-related adverse events, except for 1 who was not evaluated due to rapid progression after a dose of immunotherapy. Two patients developed grade 2 hypothyroidism and were treated with regular oral levothyroxine. No patient discontinued immunotherapy due to adverse reactions. (A) OS. (B) PFS. ICI, immune checkpoint inhibitor; OS, overall survival; PFS, progression-free survival. Univariate analysis showed that age, stage, and lymphadenectomy were prognostic predictors for OS ( Table S2 ). Multivariate analysis showed that advanced stage (hazard ratio [HR]=2.987; 95% confidence interval [CI]=1.094–8.157; p=0.033) was associated with poorer OS, and age ≥50 years (HR=0.370; 95% CI=0.141–0.966; p=0.042) was associated with better OS.

We reviewed the clinical characteristics of 42 patients with ovarian SCC, reporting the molecular features and treatment outcomes of patients receiving ICIs. Ovarian SCC is most commonly associated with teratoma malignancy, particularly showing a higher rate of malignancy among women over the age of 45 [ 10 ]. Cystic teratomas are usually detected 15–20 years earlier than malignant transformation, and prolonged exposure to carcinogens may lead to malignant transformation [ 11 ]. In rare cases, ovarian SCC has been reported in endometriosis with squamous metaplasia and malignant transformation [ 12 ]. In our study, 23.8% of patients had antecedent lesions that could not be identified. Since the ovary lacks squamous epithelium, the theory that SCC develops from other lesions is widely accepted [ 13 ]. In some samples, identifying the primary lesion may be challenging, possibly due to extensive involvement of the SCC component. Another hypothesis suggests that SCC may arise from ovarian epithelial hyperplasia and squamous cell transformation [ 14 ]. Further molecular studies are essential to clarify the evolution of ovarian SCC. Currently, there is still a lack of standardized treatment for ovarian SCC. Following the principle of management of ovarian epithelial cancer, it is considered cytoreductive surgery followed by adjuvant chemotherapy [ 15 ]. Chen et al. [ 16 ] found that patients with early stage of SCC arising from teratoma did not benefit from chemotherapy or radiotherapy, while those with advanced disease did. Hackethal et al. [ 17 ] observed that alkylating agents were more effective than other chemotherapy options in stages beyond IA. Conversely, radiotherapy may increase morbidity and negatively affect survival [ 17 ]. Our center primarily used paclitaxel combined with platinum chemotherapy, and larger sample sizes are needed to compare the efficacy of different regimens. FSS may be considered in younger patients with teratoma malignancies confined to the ovary and without local or distant invasion [ 18 ]. In our study, all 9 stage IA patients survived, and 1 patient successfully conceived and delivered after completing treatment. Patients with stage IC have a poorer prognosis than those with stage IA, making it crucial to avoid intraoperative tumor rupture. Our study found PIK3CA mutations in 80% of ovarian SCC patients, with a median TMB of 10.00 mutations/Mb, which is higher than the 4.2 mutations/Mb reported by Tamura et al. [ 19 ] for SCC arising from teratoma. PIK3CA mutations activate the PI3K/AKT/mTOR pathway, influencing tumor growth [ 20 ]. In this study, 1 patient with somatic PIK3CA mutations received sirolimus after relapse and achieved 7 months of PFS. In addition, another mutation detected in patient involves CDKN2A , a multifunctional gene that inhibits the cell cycle through the CDK4/6 regulatory mechanism [ 21 ]. A phase II clinical trial assessed the efficacy and safety of the CDK4/6 inhibitor palbociclib in patients with refractory germ cell tumors and 2 of 10 patients with malignant transformation of teratomas achieved 24 weeks of PFS [ 22 ], suggesting that CDK4/6 inhibitors may be a promising strategy for this condition [ 23 ]. Our study also detected mutations in homologous recombination repair genes like BRCA2 and FANCC , indicating that targeting the homologous recombination pathway may be a direction for future research. A retrospective analysis indicated that patients with high TMB (≥10 mutations/Mb) in head and neck SCC experienced extended survival (>2 years) with ICIs [ 24 ]. In our study, the median TMB reached 10 mutations/Mb, which is higher than that observed in ovarian serous carcinoma [ 25 ] and testicular/ovarian germ cell tumors [ 26 27 ]. Tumors with higher TMB carry a greater neoantigenic load and are more likely to benefit from immunotherapy [ 28 ]. Five patients of ovarian SCC arising from teratoma treated with ICIs have been reported, and 4 of them showing potential benefit from ICIs [ 29 30 31 32 33 ]. In our cohort, the proportion of patients in advanced stages treated with ICIs was higher than in the overall population (60% vs. 47.6%). Despite the higher proportion of advanced-stage patients, no deaths and only 1 recurrence were observed in the immunotherapy cohort during a median follow-up of 20.1 months. It is important to note the very poor prognosis of SCC associated with endometriosis. Acién et al. [ 34 ] reviewed 17 patients with SCC associated with endometriosis; 13 died within 1 year (typically between 3 to 6 months), and only 2 survived beyond 2 years. In our immunotherapy cohort, 2 patients with SCC associated with endometriosis (stages IIIC and IVB) were followed up for 20.1 and 19.3 months, respectively. One patient achieved tumor-free survival, and the other, after recurrence, responded well to chemotherapy combined with immunotherapy. Due to the rarity of the disease, the number of patients was limited. Additionally, the follow-up time for patients who received immunotherapy was relatively short. Although the survival analysis did not reveal statistical significance, patients receiving ICIs appear to have a better prognosis compared with the overall cohort and literature reports, warranting further investigation with additional molecular and immunotherapy data. In our study, the 1- and 5-year OS rates were 73.7% and 48.7%, respectively, slightly exceeding the 61.6% and 44.9% reported by Zhang and Ma [ 35 ] Additionally, the study demonstrated that the survival curve for ovarian SCC declined rapidly in the first 2 years after diagnosis and then leveled off. This trend was also observed in our cohort, indicating a propensity for early relapse in this type of cancer. Late-stage disease and younger age were identified as prognostic factors for poor prognosis. Other studies have reported advanced age, residual tumor, tumor rupture, tumor dissemination, and elevated level of SCC-Ag and CA125 were poor prognostic factors [ 36 ]. The difference in age-related findings may be due to the limited number of cases in our cohort and the inclusion of SCC associated with endometriosis, which is more common in younger patients and linked to a poor prognosis. Patients with SCC associated with endometriosis had poorer OS compared to those arising from teratoma [ 34 ]. The strength of this study lay in the collection of rare disease patients over 20 years at a single institution. It included comprehensive data analysis and long-term follow-up, particularly gathering molecular pathology and treatment information of ICIs. However, our study had several limitations. The retrospective and single-center design inherently introduced potential confounders and bias. Due to the limited number of patients, we could not assess the efficacy of different chemotherapeutic regimens and the impact of residual tumors on prognosis. Differences in the duration of follow-up between groups also introduced an inherent bias. Subsequently, incorporating multicenter data is essential to expand the sample size for this rare disease. Additionally, molecular characterization of more patients is needed to identify relevant targets and analyze the tumor immune signatures. Continued follow-up of these patients is necessary, along with the enrollment of additional potential beneficiaries, to evaluate the effectiveness of immunotherapy. In conclusion, ovarian SCC is a rare malignancy with a poor prognosis. The FIGO stage and patient age are prognostic predictors for survival. This carcinoma is characterized by a high frequency of PIK3CA mutations and a higher TMB. There was a trend toward improved prognosis with ICIs.

This retrospective study was approved by the Ethics Committee of Peking Union Medical College Hospital (reference number I-23PJ1911). All procedures were carried out following the Declaration of Helsinki. The written informed consent for this retrospective study was waived due to the retrospective nature of the study. All personal information was de-identified to protect the privacy of patients. The study selection followed these criteria: Inclusion criteria involved: 1) Patients with pathologically confirmed ovarian SCC who underwent surgery at Peking Union Medical College Hospital between January 2000 and September 2023; 2) Medical records were available. Exclusion criteria involved: 1) Uncertainty of primary site or ovary is the site of metastasis; 2) No surgical intervention performed; 3) Concurrent malignancies in other systems; and 4) Pathological review not consistent with ovarian SCC. Clinical data were obtained from the Hospital Information System. Tumor staging followed the 2014 International Federation of Gynecology and Obstetrics (FIGO) classification for ovarian tumors. Two experienced pathologists (X.S. and H.W.) reviewed the pathology to confirm the diagnosis. Since 2019, 5 ovarian SCC patients have undergone genetic testing. Formalin-fixed paraffin-embedded samples and paired peripheral blood were collected for whole exome sequencing or targeted exome sequencing (520- or 561-gene panel) and further assessed for TMB status. Primary surgical procedures included tumor resection, staging surgery, and cytoreductive surgery, following the principles of epithelial ovarian cancer treatment. Fertility-sparing surgery (FSS) was performed for young patients wishing to preserve fertility. Complete surgical staging was defined as peritoneal cytology, multiple peritoneal biopsies, omentectomy, or omental biopsy with or without lymph node dissection. Complete surgical cytoreduction with no macroscopic residue tumor was classified as R0 resection, and postoperative residual tumor diameters of 0.1–1 cm or >1 cm were defined as R1 and R2 resection, respectively. FSS was defined as the preservation of at least 1 ovary and the uterus. Platinum-based intravenous chemotherapy was administered and the cycles were decided according to the treatment response and principle of epithelial ovarian cancer. If the clinician deemed optimal cytoreduction unachievable or life-threatening, neoadjuvant chemotherapy was recommended as the initial treatment. The decision to administer bevacizumab was based on evidence from first-line treatment for advanced ovarian cancer and cervical SCC, as well as the decision-making of the supervising clinician. Other targeted therapies were determined based on the results of molecular testing. The choice of ICIs was informed by the patient’s molecular testing results, evidence from SCC at other locations, and insights from previous case reports. ICIs were given along with chemotherapy, followed by maintenance therapy every 3 weeks. The total duration of treatment was 2 years unless disease progression, intolerable toxicity, or patient refusal. Patients were followed up through outpatient visits or telephone interviews until March 20, 2024. Overall survival (OS) was defined as the time from initial diagnosis to death or last follow-up. Progression-free survival (PFS) was defined as the time from initial diagnosis to recurrence, disease progression, or death. Recurrence was defined as imaging findings suggestive of recurrence with or without elevated serum tumor markers, or elevated tumor markers in the absence of imaging evidence. Disease progression was established by Response Evaluation Criteria in Solid Tumors (RECIST 1.1) and serum tumor markers. The follow-up time was censored if the patient was lost during follow-up period. The OS and PFS were estimated using the Kaplan-Meier method. Univariate and multivariate statistical analysis were done according to the Cox proportional hazards model. Two-sided p-values less than 0.05 were considered statistically significant. Statistical Package for the Social Sciences software version 26.0 (IBM Corp., Armonk, NY, USA) was used for statistical analysis.