Primary carcinosarcoma of the uterine cervix with somatic mutations of the ATM and NF2 genes: a case report

Abstract

Primary carcinosarcomas are a rare type of cervical tumors. A 75-year-old female patient diagnosed with stage IB2 cervical carcinosarcoma (CCS) was treated with radical surgery followed by adjuvant chemoradiotherapy. She survived without recurrence for 3 years after the surgery. Double somatic mutations of the ATM (c.802C > T [p.Glu268*], variant allele frequency [VAF] of 48.5%, and c.7181C > T [p.Ser2394Leu], VAF of 45.1%) and NF2 (c.1021C > T [p.Arg341*], VAF of 2.4%) genes were detected in the CCS via next-generation sequencing analyses. Retrospective studies using large databases should be performed to establish a consensus regarding the use of combination therapies in elderly patients with primary CCS. To the best of our knowledge, this report first revealed the presence of ATM mutations in a patient with primary CCS. Previous studies and the current data showed therapeutic possibilities for the subsets of gynecologic carcinosarcoma and provided information on the molecular mechanism of its development.

Introduction

Primary carcinosarcomas are rare cervical tumors, accounting for only 0.2%–0.6% of all cervical malignancies [1, 2]. There are no clear data on the prognostic factors, treatment options, and survival outcomes of primary cervical carcinosarcoma (CCS). This is because primary CCS, unlike other cervical cancers or uterine carcinosarcomas, is extremely uncommon [3]. Further, to the best of our knowledge, there is no study on the molecular characteristics of primary CCS. Herein, we report a 75-year-old female patient with primary CCS who did not experience recurrence for 3 years after surgery. The patient was managed with surgery, followed by adjuvant concurrent chemoradiotherapy (CCRT). In this case, molecular characterization was also performed via next-generation sequencing (NGS) analyses.

Case report

A 75-year-old, gravida 2, para 0, Japanese female patient with a history of vaginal bleeding for several months visited X clinic, where a 3-cm cervical mass was partially resected. The patient was referred to our department because of the histopathological cervical biopsy results, which revealed CCS or cervical adenosarcoma. She did not report any remarkable medical history. However, her father died of gastric cancer at the age of 76 years. A gynecological examination using colposcopy revealed a 1.5-cm fragile mass that protruded from the ectocervix (Fig. 1). The mass was then resected. Histopathological analyses revealed CCS comprising carcinomatous (adenocarcinoma, HPV-independent, and NOS) and sarcomatous (chondrosarcoma and fibrosarcoma) components (Fig. 2). The lymph vascular space was not involved. The high-risk HPV DNA test had negative results. The patient then underwent pelvic magnetic resonance imaging, which did not detect any cervical mass. Computed tomography scan did not show pelvic or para-aortic lymph node enlargement, liver masses, or lung masses. The patient underwent type C1 radical hysterectomy (RH, classified by Querleu D and Morrow CP [4]), bilateral salpingo-oophorectomy, and bilateral pelvic lymphadenectomy. The histopathological analyses did not reveal residual cervical tumor and lymph node metastasis, and lesion was determined as stage IB2 based on the International Federation of Gynecology and Obstetrics 2018 staging system. Adjuvant external pelvic radiotherapy (45 Gy: 1.8 Gy/day × 25 times) and concurrent chemotherapy (cisplatin 40 mg/m²/week × 5 courses) were completed. The patient survived without any disease for 3 years since the primary surgery.

Fig. 1.

Gynecologic examination via colposcopy revealing a 1.5-cm fragile mass protruding from the ectocervix ( →)

Fig. 2.

Histopathological analysis. a Low-magnification image of the carcinomatous and sarcomatous components (hematoxylin– & eosin [H&E] staining). The boxed areas indicate regions that were magnified in (b), (c), and (d). b Higher-magnified view of the carcinomatous components corresponding the red boxed area in (a) (adenocarcinoma; H&E staining × 40). The highly cylindrical atypical epithelium formed irregular glandular ducts that infiltrated and proliferated. c Higher-magnified view of the sarcomatous components corresponding the blue boxed area in (a) (fibrosarcoma; H&E staining × 40). The subepithelial stroma was partially densely cell-dense, and abnormal mitosis was observed. d Higher-magnified view of the sarcomatous components corresponding green boxed area in (a) (chondrosarcoma; H&E staining × 40). Juvenile cartilage-like tissues were identified. e–g Immnunohistochemical images corresponding to (b–d), respectively. Immunohistochemical examination revealing carcinomatous and sarcomatous component-positive ATM (× 20). Estrogen receptor and p16 staining had negative, results, and p53 staining had weakly positive results in a small part of the carcinomatous and sarcomatous components (p53 wild-type pattern) (data not shown). The Ki-67 scores in the carcinomatous, fibrosarcomatous, and chondrosarcomatous components were approximately 80%, 50%, and 20%–30%, respectively (data not shown)

The genomic analysis was conducted using all parts of the formalin-fixed and paraffin-embedded tumor, as shown in Fig. 2. The tumor content was approximately 100% based on the histopathological findings and sequencing data. Foundation Medicine submitted a sample that should be assessed by a certified pathologist. The platform FoundationOne CDx® (Foundation Medicine, Cambridge, the USA) was used for genome analysis via targeted NGS. Her microsatellite status (MS) was MS-stable and her tumor mutation burden was 3 Muts/Mb. The genome analysis detected the c.802C > T (p.Glu268*) and c.7181C > T (p.Ser2394Leu) variants of the ATM gene and the c.1021C > T (p.Arg341*) variant of the NF2 gene in the tumor sample. The allele frequencies of these variants were 48.5%, 45.1%, and 2.4%, respectively. These variants were considered somatic mutations because they were not detected on direct DNA sequencing from the peripheral blood cell. The c.802C > T and c.7181C > T heterozygous mutations of the ATM gene were detected via direct DNA sequencing from both the micro-dissected carcinomatous and sarcomatous components. Immunohistochemical investigation revealed that the ATM (anti-ATM rabbit monoclonal antibody [Y170] ab32420, Abcom, Cambridge, the UK) tested positive for carcinomatous and sarcomatous components (Fig. 2e). The c.1021C > T heterozygous mutation of the NF2 gene was detected from the micro-dissected sarcomatous component, but not from the carcinomatous component.

Discussion

An elderly female patient with stage IB2 CCS was treated with radical surgery, followed by adjuvant CCRT. She survived without recurrence for 3 years after surgery. However, more case studies and retrospective studies using a large database are should be conducted to establish a consensus regarding the use of combination therapies in elderly patients with primary CCS. Primary cervical CCS is considered an aggressive neoplasm [2, 3]. Based on a previous study, the overall 2-year survival rates of patients with primary CCS who received primary radiotherapy and those who underwent surgery only were 17% and 68%, respectively. Further, the 2-year disease-free survival rates of patients with stage I CCS who underwent primary surgery and those who were treated with primary surgery, followed by adjuvant radiotherapy with chemotherapy, were 63% and 100%, respectively [3]. Sharma C et al. revealed that elderly women with stage IB1–IIA cervical cancer were less likely to undergo surgery (particularly RH and lymphadenectomy) and receive adjuvant radiotherapy [5]. Shimamoto K et al. showed no significant differences in terms of disease-specific survival among patients with cervical cancer treated only with standard regimens only who were aged < 65 (younger) and ≥ 65 (elderly) years. Moreover, the two patient groups who received standard treatment had better prognoses than those who received nonstandard treatment [6]. Hence, elderly patients may have better outcomes based on the type of standard treatment given, and the appropriate modality and intensity of treatment should be based on the patient’s general condition and background [6]. Information on the available treatment options for elderly women with stage I CCS in the published literature is limited. However, the elderly patient with stage I CCS in this case report was scheduled for primary radical surgery, followed by adjuvant CCRT, because she had good physical and performance status without any chronic illnesses.

Double somatic mutations of the ATM gene (c.802C > T [p.Glu268*], variant allele frequency [VAF] of 48.5%, and c.7181C > T [p.Ser2394Leu], VAF of 45.1%) were detected in a patient with primary CCS. Previous reports and the current data indicated therapeutic possibilities for the subsets of gynecologic carcinosarcoma and provided information on the molecular mechanism of its development. ATM encodes the ataxia-telangiectasia-mutated serine/threonine protein kinase, a member of the PI3K/PI4K family that plays an essential role in DNA damage response [7]. The loss of functional ATM promotes tumorigenesis [7]. The ATM mutation known as c.802C > T (p.Glu268*) is involved in the pathogenesis of hereditary cancer-predisposing syndrome (accessed https://www.ncbi.nlm.nih.gov/clinvar/on clinvar, April 2024). Meanwhile, the ATM mutation, c.7181C > T (p.Ser2394Leu), is a possible pathogenic factor of hereditary cancer-predisposing syndrome and/or familial types of breast cancer (accessed https://www.ncbi.nlm.nih.gov/clinvar/on clinvar, April 2024). Alterations, such as those observed in the current case, may disrupt ATM function. Somatic mutations in the ATM gene have been observed in various tumors, including the subsets of squamous cell carcinoma, adenocarcinoma, mesonephric carcinoma, neuroendocrine carcinoma, malignant mixed mesonephric cervical tumor (referred COSMIC, April 2024), and uterine and ovarian carcinosarcoma [8]. The frequency of ATM mutations in uterine corpus carcinosarcoma is 5.5% (referred COSMIC, April 2024). To date, there are no reports on molecular targeted therapies specifically targeting ATM mutations in this disease. To the best of our knowledge, this report first revealed ATM mutations in primary CCS. The loss of functional ATM, which causes a defective DNA damage response and homologous recombination-mediated DNA repair, may predict poly ADP ribose polymerase (PARP) inhibitor sensitivity [9]. The clinical trials have been conducted to evaluate the efficacy of PARP inhibitors (PARPi) used as monotherapy or in combination with other agents for various cancers with ATM mutation or ATM loss (Table 1) [10–16]. The PROfound trial examined the efficacy of olaparib in patients with metastatic castration-resistant prostate cancer who had mutations in the homologous recombination repair (HRR) genes [10]. In patients with mutations in either the BRCA1, BRCA2, or ATM gene, olaparib was associated with a significant improvement in radiological progression-free survival (rPFS) (median: 7.4 vs 3.6 months) (hazard ratio [HR]: 0.34, 95% CI 0.25–0.47, p < 0.0001), overall survival (median: 19.1 vs. 14.7 months) (HR: 0.69, 95% CI 0.50–0.97, p = 0.0175), and objective response rate (33% vs 2%, p < 0.0001) compared with control therapy. Further, olaparib was associated with a significantly improvement in rPFS (median: 5.8 vs. 3.5 months) (HR: 0.49, 95% CI 0.38–0.63, p < 0.0001) compared with control therapy in patients with HRR mutations. Based on these results, in 2020, the Food and Drug Administration approved the use of olaparib in adult patients with deleterious or suspected deleterious germline or somatic HRR gene-mutated metastatic castration-resistant prostate cancer, who presented with disease progression after treatment with enzalutamide or abiraterone [10]. To date, other PARPis for patients with ATM mutations or ATM loss have not been approved by the FDA. Moreover, in a subgroup analysis of patients with ATM mutations or ATM loss from seven clinical trials including the PROfound trial, olaparib did not significantly improve overall survival compared with control therapy. Therefore, ATM mutations or ATM loss may not be a valid biomarker of PARPis (Table 1).

Table 1.

Selected completed clinical phase II–III trials† with PARP inhibitors for various cancers with the ATM mutation or ATM loss

Cancers	Treatment	Phase	Population	Efficacy	References
Metastatic castration-resistant prostate cancer	Olaparib vs enzalutamide or AAP (control)	P-III	i) BRCA1-, BRCA2-, and ATM-mutated population ii) ATM-mutated population	i) Median OS (95% CI): olaparib, 19.1 (17.4–23.4) months vs. control, 14.7 (11.9–18.8) months, HR: 0.69 (95% CI: 0.50–0.97) ii) Median OS (95% CI): olaparib, 18.0 (14.4–23.4) months vs. control, 15.6 (12.1–22.0) months, HR, 0.93 (95% CI: 0.53–1.75)	[17]
Metastatic castration-resistant prostate cancer	Niraparib + AAP vs placebo + AAP (control)	P-III	i) HRR gene-mutated population ii) ATM-mutated population	i) Median rPFS: niraparib + AAP, 19.3 months vs. control, 13.8 months, HR: 0.46 (95% CI: 0.32–0.67) ii) rPFS: HR, 1.26 (95% CI: 0.74–2.13)	[28]
Metastatic castration-resistant prostate cancer	Talazoparib + enzalutamide vs. placebo + enzalutamide (control)	P-III	i) HRR gene-mutated population ii) ATM-mutated population	i) Median rPFS: talazoparib + enzalutamide, not reached vs. control, 13.8 months, HR, 0.45 (95% CI: 0.33–0.61) ii) rPFS: HR, 0.76 (95% CI: 0.30–1.94)	[29]
Advanced-stage gastric cancer	Olaparib + paclitaxel vs placebo + paclitaxel (control)	P-III	i) Overall patient population ii) ATM-negative‡ population	i) Median OS (95% CI): olaparib + paclitaxel, 8.8 (7.4–9.6) months vs. control, 6.9 (6.3–7.9) months, HR, 0.79 (97.5% CI: 0.63–1.00) ii) Median OS (95% CI): olaparib plus paclitaxel, 12.0 (7.8–18.1) months vs. control, 10.0 (6.4–13.3) months, HR, 0.73 (97.5% CI: 0.40–1.34)	[23]
Metastatic breast cancer	Olaparib	P-II	i) Germline non-BRCA1/2 HRR gene [PALB2]-mutated population ii) Somatic HRR gene [BRCA1/2]-mutated population iii) ATM-mutated population	i) ORR: 33% (9PR, 8SD, 10PD) [ORR: 82% (9PR, 2SD)] ii) ORR: 31% (8PR, 10SD, 8PD) [50% (8PR, 6SD, 2PD)] iii) 1 PR in a patient with gPALB2 and gATM mutations; 1 PR in a patient with sBRCA1 and sATM mutations; no response in 8 patients with the ATM mutations and 2 patients with the ATM and CHEK2 mutations	[15]
Advanced-stage, previously treated pancreatic ductal adenocarcinoma	Olaparib	P-II	i) BRCAness§ population ii) ATM-mutated population vs. ATM wild-type or ATM-loss⁑ population	i) Median PFS (95% CI): 3.7 (12.9–5.7) months; ORR: 4% (2PR, 33SD, 12PD); PRs in 1 patient with the gPALB2 mutation and 1 with the sATM mutation ii) Median PFS: 5.0 months vs. wild type 3.6 months, P = 0.75, vs. ATM loss 3.3 months, P = 0.60	[12]
Metastatic pancreatic ductal adenocarcinoma	Veliparib + 5-FU + oxaliplatin	P-I/II	i) HRR gene-mutated [non-mutated/unknown] population ii) ATM-mutated population	i) ORR: 44% (2CR, 6PR, 6SD, 3PD) [18% (2CR, 5PR, 25SD, 8PD)]; median PFS, 6.0 [3.6] months; median OS, 10.4 [6.9] months ii) No response in 2 patients with the ATM mutations	[26]

AAP abiraterone acetate and prednisone, OS overall survival, CI confidence interval, HR hazard ratio, HRR homologous recombination repair, rPFS radiographic progression-free survival, ORR objective response rate, PR partial response, SD stable disease, PD progressive disease, 5-FU 5-fluorouracil

^† Including a phase I/II study; ‡A clinical status of ATM-positive was assigned to cases with ≥ 25% or gastric cancer cells exhibiting nuclear staining (weak, moderate, or strong staining intensity) and an ATM-negative status to cases with < 25% nuclear staining. § Defined as previously known HRR genetic alterations, personal or family history of BRCA-associated cancers, or ATM protein loss as determined via immunohistochemistry. ⁑ Defined as < 10% neoplastic nuclear staining at any intensity in the presence of positive lymphocyte staining

Ataxia telangiectasia and Rad3-related (ATR) is a serine/threonine kinase involved in DNA repair in response to DNA damage and replication stress [17]. Targeting ATR is a relevant strategy for tumors with DNA damage response defects and replication stress, including the inactivation of the ATM signaling. The preclinical models have shown that the inhibition of ATR can lead to synthetic lethality in ATM-deficient malignancies [18–20] and clinical trials have been conducted to evaluate the efficacy of ATR inhibitors (ATRis) as monotherapy or in combination with other drugs against a variety of ATM-deficient tumors (Table 2) [21–27]. Early clinical trials on the use of ATRis in combination with chemotherapy agents, such as gemcitabine and carboplatin, which can induce DNA replication stress, were often conducted [21, 28, 29]. In these clinical trials, the combination of ATRis and chemotherapy showed antitumor activity. However, it was also associated with higher rates of adverse events [21]. Several ongoing clinical trials combining ATRis with molecular targeted therapies are still awaiting results. Nevertheless, PARPis are one of the most promising candidates that can be used in combination with ATRis. The combination of PARPi and ATRi can overcome PARPi-resistance in preclinical models [30]. In a previous phase II study on the use of ceralasertib, an ATRi, in combination with olaparib, two of five patients with tumors harboring the ATM mutations achieved complete response or stable disease at 26 months [22]. In addition, another study reported that ATM loss was associated with higher rates of response and PFS [23]. Immune checkpoint inhibitors (ICIs) are another promising candidate that can be applied in combination with ATRis. ICIs have revolutionized cancer treatment for patients with a variety of solid tumors [31, 32]. However, not all patients achieve durable responses, and resistance to ICIs is common [33]. In a phase II study on advanced-stage non-small cell lung cancer after failed ICI and platinum-doublet therapy, the combination of ceralasertib and durvalumab, an ICI, had a higher response rate than other durvalumab-based combination therapy. In addition, this regimen exhibited a greater activity in the ATM-altered cohort, who has higher response rates and prolonged clinical outcomes compared with the ATM-intact cohort [26]. In another phase II study on advanced-stage gastric cancer, the combination of ceralasertib and durvalumab had a promising antitumor activity, with durable responses, particularly in patients with ATM loss and/or a high proportion of mutational signatures attributed to homologous recombination deficiency [25]. These findings indicated that the use of combined therapy with ceralasertib and olaparib or durvalumab for tumors with ATM mutations or ATM loss, where there is a significant unmet need for effective treatment options, must be further investigated.

Table 2.

Selected completed clinical phase II trials with ATR inhibitors for various types of cancers with ATM mutation or ATM loss

Compounds	Cancers	Population	Efficacy	References
I) Chemotherapy combination
Berzosertib + gemcitabine vs. gemcitabin (control)	Platinum-resistant high-grade serous ovarian cancer	i) Overall patient population ii) ATM-negative or ATM-low † population [‡]	i) Median OS (90% CI): berzosertib + gemcitabine – 59.4 (33.7–86.6) weeks vs. control – 43.0 (34.4–67.9) weeks, HR – 0.79 (90% CI: 0.52–1.20); Median PFS: berzosertib + gemcitabine – 22.9 (17.9–72.0) weeks vs. control – 14.7 (9.7–36.7), HR: 0.57 (90% CI: 0.33–0.98) ii) OS: HR – 0.50 (90% CI: 0.23–1.08) [HR: 0.32 (90% CI: 0.14–0.73) ‡]	[28]
II) PARP inhibitor combination
Ceralaserib + olaparib	Advanced-stage cancers harboring HRR alterations	i) Overall patient population ii) ATM-mutated population	i) ORR: 8.3% (1CR, 1PR, 13SD, 10PD) ii) 1 CR and 1 SD ongoing at 24 + months in 5 patients with ATM mutations	[22]
Ceralaserib + olaparib	Triple-negative advanced-stage breast cancer	i) Overall patient population ii) ATM loss § [ATM intact] population‖	i) ORR: 17.1% (2CR + 10PR/70) ii) ORR: 21.4% (3/14) [13.8% (4/17)], median PFS (IQR): 3.4 (1.4–10.2) months [2.5 (1.9–10.0) months]	[23]
Ceralaserib + olaparib vs olaparib	Relapsed or refractory small-cell lung cancer	i) Overall patient population ii) ATM loss population	i) ORR: ceralaserib + olaparib 3.8% (1PR, 10SD, 11PD, 4NE) vs. olaparib 6.7% (1PR, 4SD, 9PD, 1NE); median PFS (95% CI): ceralaserib + olaparib 7.2 (6.0–10.2) months vs. olaparib 8.1 (5.1–16.5) months ii) Ceralaserib + olaparib: 2SD, 1NE vs. Olaparib: 1PR, 3PD, 1NE	[24]
III) Immune checkpoint inhibitor combination
Ceralasertib + durvalumab	Previously treated advanced-stage gastric cancer	i) Overall patient population ii) ATM loss ⁑ or sig.HRDHi vs. ATM intact and sig.HRDLO	i) Median PFS (95% CI): 3.0 (2.1–3.9) months; median OS (95% CI): 6.7 (3.8–9.6) months; ORR: 22.6% (7PR, 11SD, PD12, NE1) ii) Median PFS: ATM loss or sig.HRDHi 5.60 months vs. ATM intact and sig.HRDLO 1.65 months, HR: 0.13 (95% CI: 0.045–0.39)	[25]
Ceralasertib + durvalumab	Non-small-cell lung cancer	i) Overall patient population ii) ATM biomarker-matched # [nonmatched primary resistance; nonmatched acquired resistance] population iii) ATM-mutated [wild-type] population	i) Median PFS (80% CI): 5.8 (4.6–7.4) months; median OS (80% CI): 17.4 (14.1–20.3); ORR: 22.6% (7PR, 11SD, PD12, NE1) ii) ORR: 26.1% (6PR, 11SD, 4PD, 2NE) [13.0% (3PR, 9SD, 8PR, 3NE); 6.1% (2PR, 17SD, 8PD, 6NE)]; PFS (80% CI): 8.4 (6.0–9.7) months [4.6 (1.9–6.8) months; 4.0 (3.0–5.8)]; OS (80% CI): 22.8 (12.6–29.9) months [12.0 (10.5–18.8) months; 19.1 (15.8–20.3) months] iii) PFS (80% CI): 8.4 (6.0–9.5) months [4.6 (3.0–6.8) months]; OS (80% CI): 24.1 (12.6–29.9) months [14.1 (11.3–19.1) months]	[26]
Ceralasertib + durvalumab	Advanced-stage/metastatic melanoma that was not successfully managed with anti-PD-1 therapy	i) Overall patient population ii) ATM loss ⁑ [ATM intact] population	i) ORR: 30% (9PR, 10SD, 10PD, 1NE); median PFS (95% CI): 7.1 (3.6–10.6) months; median OS (95% CI): 14.2 (9.3–19.1) months ii) ORR: 33.3% (1PR/3) [36.8% (7PR/19)]	[27]

IHC immunohistochemistry, OS overall survival, CI confidence interval, HR hazard ratio, PFS progression-free survival, ORR objective response rate, CR complete response, PR partial response, SD stable disease, PD progressive disease, IQR interquartile range, NE not evaluable, sig.HRD^Hi high proportion of mutational signature attributed to homologous repair deficiency, sig.HRD^LO low proportion of mutational signature attributed to homologous repair deficiency

^†, defined as the presence of nuclear ATM staining in < 50% of tumor cells. ‡, defined as patients who died and who were excluded from the analysis. §, defined as an Hscore ≤ 10. ‖, defined as patients without a germline BRCA1/2 mutation. ⁑, defined as a nuclear expression in ≤ 20% of the stained cells. #, defined as patients with detectable aberrations in ATM on next-generation sequencing or low ATM protein expression

Initial treatment with radical surgery, followed by CCRT, was successful in an elderly female patient with CCS. However, more cases and retrospective studies using large databases are required to establish a consensus regarding the use of combination therapies in elderly patients with primary CCS. CCS is associated with both nonsense and missense somatic mutations of the ATM gene. To the best of our knowledge, this report first showed ATM mutations in primary CCS. However, somatic mutations in the ATM gene have been reported in the subsets of gynecological carcinosarcoma [8]. Taken together, our report showed therapeutic possibilities for the subsets of gynecologic carcinosarcoma and provided information on the molecular mechanism of its development.

Author contributions

All authors contributed to the study conception and design. TI, YH, JM and MI prepared the materials, colleceted and analyzed the data. ST drafted the manuscript. All authors reviewed and edited the manuscript. All authors approved the submitted version.

Funding

This research did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets shown in this report are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors declare that they have no conflicts of interest.

Ethical approval

This case report did not need to be reviewed by the relevant ethical commitees.

Informed consent

The patient provided informed consent.