Uterine carcinosarcoma (UCS) is a rare aggressive tumor comprising carcinomatous and sarcomatous components. Current evidence indicates that the sarcomatous section dedifferentiates from the carcinomatous section [1] and UCS is postulated to be an endometrial cancer (EC) [1,2]. Mainstay treatments include surgical staging, adjuvant chemotherapy, and radiotherapy. However, the 3-year survival rate for all stages is only 42% [3]. Advanced disease stage, large tumor size, deep myometrial invasion, lymphovascular space invasion (LVSI), aggressive histology of carcinomatous and sarcomatous components, and lack of adjuvant chemotherapy have been identified as poor prognostic factors [3,4]. Furthermore, hematological parameters such as anemia, leukocytosis, and thrombocytosis, which are indicators of the inflammatory response, are reportedly associated with worse outcomes in EC [5,6].

The rate of disease progression is high, at approximately 50%. The majority of progression occurs within 2 years after treatment, with recurrence occurring outside the pelvis in most cases [3,7,8]. Tung et al. [7] reported a progression-free interval (PFI) of 8.1 months and a 5-year survival rate of 7.6% after progression in 98 patients with recurrent or progressive UCS. It was also reported that salvage therapy in combination with chemotherapy, radiotherapy, or chemoradiotherapy improved survival and that disseminated recurrence was a poor prognostic factor [7]. Although UCS is known to be an aggressive disease, there is a paucity of data on postprogression UCS to inform patients.

Therefore, in this study, we aimed to evaluate the 2-year overall survival (OS) after progressive disease and explore the clinical and prognostic factors in patients with progressive UCS after surgery with or without adjuvant therapy.

This retrospective, descriptive cohort study was approved by the Ethics Committee for Research Involving Human Subjects (REC 66-103-12-3). The requirement for informed consent was waived owing to the retrospective nature of the study, and the anonymity of the participants was maintained.

This study included patients diagnosed with UCS at all stages using the 2009 International Federation of Gynecology and Obstetrics (FIGO) staging system. The patients underwent initial surgical staging procedures with or without adjuvant therapy between January 2002 and December 2022. Patients who developed recurrent or progressive disease were included. Patients who 1) had never visited the hospital since the diagnosis of progressive disease, 2) had synchronous cancer, 3) received neoadjuvant chemotherapy or radiotherapy as primary treatment, or 4) were <18 years of age were excluded from the analysis.

The following data were extracted from electronic medical databases for analysis: patient characteristics such as age at diagnosis, weight, and height; Eastern Cooperative Oncology Group performance status; menopausal status; underlying medical disease; history of tamoxifen use; and disease characteristics including stage, tumor size >10 cm, myometrial invasion, LVSI, omental involvement, pelvic and para-aortic lymph node involvement, and optimal surgery status defined as tumor size <1 cm. Additional data included clinical aspects after recurrence, such as symptoms of progressive UCS, recurrence pattern, and pre-salvage treatment conditions such as anemia (hemoglobin levels <12 g/dL), leukocytosis (white blood cell count ≥10,000/µL), and thrombocytosis (platelet count >400,000/µL). Information on salvage treatments and the end status of the patients was also collected.

After primary treatment with or without adjuvant therapy, the patients underwent follow-up clinical and physical examinations performed by gynecologic oncologists, fellows, or residents under the supervision of the gynecologic oncology staff. The follow-up protocol involved visits every 3 months for 2 years, every 6 months until the 5 years after treatment, and annually thereafter for surveillance. Computed tomography (CT) and magnetic resonance imaging were performed for patients who exhibited abnormal symptoms on physical examination and for those with residual disease for follow-up response to treatment. Symptomatic patients were defined as those who complained of any symptoms related to progressive disease during routine follow-ups or before appointments. Asymptomatic patients were defined as those whose disease progression was detected by examination or imaging in accordance with the follow-up protocol and who did not present with abnormal symptoms.

The recurrence patterns were classified into three categories: local, distant, or combined. Local recurrence was defined as vaginal cuff and/or pelvic recurrence below the para-aortic vessel level. Distant recurrence was defined as recurrence outside the pelvis and/or above the para-aortic region. Combined recurrence was defined as recurrence occurring in both the local and distant areas.

Treatment after progression depends on the site of recurrence, primary treatment, and the performance status of the patient. Salvage treatments included surgery, chemotherapy, radiotherapy, and a combination of both. Palliative treatment was defined as supportive treatment for symptoms. After salvage treatment, a follow-up protocol was followed for all patients.

OS after progression was calculated from the date when recurrent or progressive disease was diagnosed through imaging or pathological examination to the date of death or last follow-up. The PFI was calculated from the date of surgery as the primary treatment to the date of progressive disease detection by imaging or histological examination. The follow-up time was calculated from the date of progressive disease diagnosis to the date of death or censoring at the last follow-up.

The sample size was calculated using the log-rank test and analyzed under the assumption of proportional hazards. Based on a study by Tung et al. [7], the hazard ratio (HR) for salvage treatment with chemotherapy was 0.41, with an alpha of 0.05 and power of 80%. The total sample size required for the formula was 46. Because of the rarity of the disease, all patients who met the inclusion criteria were included.

The demographic data of patients with progressive UCS were subjected to descriptive analyses and are presented as percentages. Clinicopathological factors and recurrence patterns were compared using Fisher’s exact test or the chi-squared test. OS after progressive UCS was analyzed using the Kaplan-Meier method and compared using the log-rank test. Variables with P<0.2 were selected for the multivariate Cox proportional hazards model. HRs with 95% confidence intervals (CIs) were calculated and statistical significance was set at P<0.05. Statistical analyses were performed using the “rms” package in R software version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria).

During the study period, 106 patients were diagnosed as having UCS. Of these, one patient received neoadjuvant radiotherapy followed by hysterectomy with bilateral salpingo-oophorectomy. Furthermore, 61 patients developed progressive disease; of these, seven were excluded owing to the loss of hospital contact since their diagnosis of progressive disease. Thus, 54 patients were eligible for inclusion in this study.

Table 1 summarizes the demographic data of the patients with progressive UCS. The mean patient age was 61 years (range, 36-83). After primary surgery, 87.04% of the patients received adjuvant therapy: 37 underwent chemotherapy, three received radiotherapy, and seven received combined chemotherapy and radiotherapy. The chemotherapy regimen is detailed in Supplementary Table 1. The median radiotherapy dose was 67 gray (Gy) (interquartile range [IQR], 50-71). The median PFI was 10.5 months (IQR, 6-18). Of the 54 patients, 43 developed progressive disease within the first 2 years. Among the 54 patients, the recurrence pattern was “local” in 14 patients, “distant” in 19 patients, and “combined” in 21 patients. Overall, 64.81% of the patients presented with abnormal symptoms prior to recurrence detection. Tumor progression was observed in 30 patients through physical examination and in 24 patients through follow-up imaging. Among the clinicopathological variables (Table 2), only LVSI was associated with a combined recurrence pattern (P=0.045).

Regarding salvage treatment, 40.74% of the patients received chemotherapy, 20.37% underwent combined treatment, and 9.26% received radiotherapy alone for local recurrence (Table 3). The chemotherapy regimens are detailed in Supplementary Table 1. Of the 16 patients who underwent salvage radiotherapy with or without chemotherapy, 12 received palliative salvage radiotherapy. The median radiotherapy dose was 45 Gy (IQR, 27.5-65.25).

The median follow-up time after progressive disease was 18 months (IQR, 12.25-27.72). Of the 54 patients, 44 died within a median time of 6 months (IQR, 1.25-12). The 2- and 5-year OS rates after disease progression were 20% and 11%, respectively (Fig. 1). Table 4 presents the results of the univariate analysis of 2-year OS after progressive UCS. Optimal debulking surgery as primary treatment for the tumors (HR, 0.31; 95% CI, 0.13-0.72; P=0.006) and salvage treatment with combined chemotherapy and radiotherapy (HR, 0.23; 95% CI, 0.08-0.7; P=0.009) were associated with improved survival after progression. During pre-salvage treatment, leukocytosis was significantly associated with poor prognosis (HR, 5.98; 95% CI, 2.33-15.3; P<0.001) (Table 5).

UCS is a highly aggressive form of EC. The present study showed a high disease progression rate of 57.55% compared with disease progression rates of 47.35-58.62% in previous studies [7,9,10]. In previous reports, most patients were diagnosed with progressive disease within 2 years of primary treatment, with a PFI of 8-12 months [7,9,10]. Similarly, our study showed that 79.63% of patients had progressive disease within 2 years (median PFI, 10.5 months).

Currently, molecular classification has the potential to predict EC behavior and prognosis. The proactive molecular risk classifier for endometrial cancer trial revealed that the protein 53 null/missense (p53 abn) was a poor prognostic factor; p53 abn tumors exhibited aggressive pathological features (80.2% non-endometrioid histology and 61.3% LVSI) and advanced stage, and were significantly associated with worse oncological outcomes [11]. Notably, the prevalence of p53 abn in UCS is high (74%) [12]. Patients diagnosed with EC with p53 abn develop progressive disease more frequently than those with polymerase-ɛ (POLE) mutation, with recurrence often occurring outside the pelvis and a median survival of 10 months post-progression [13]. These reports on p53 abn may explain the high disease progression rate and 2-year OS of only 20% after progression observed in the present study. Moreover, we determined that the majority of patients experienced recurrence outside the pelvis, with a distant recurrence pattern in 35.18% of patients and a combined pattern in 38.89%; these recurrence patterns were associated with the presence of LVSI, with 72.5% of patients developing distant and combined patterns (P=0.045). Previous studies have shown that LVSI is associated with an increased risk of distant metastasis and poor outcomes [14,15], although its role in UCS remains unclear [16]. According to the latest 2023 FIGO staging for EC, molecular characterization, especially p53 abn and POLE mutations, is a precise prognostic factor for survival outcomes and treatment decisions [2]. However, we did not provide molecular data to support these findings in the present study.

Tung et al. [7] reported a 5-year survival rate of 7.6% after disease progression in 98 patients with progressive UCS. Furthermore, they revealed that salvage therapy was significantly associated with improved survival, chemotherapy (HR, 0.41; 95% CI, 0.24-0.72; P=0.002), radiotherapy (HR, 0.27; 95% CI, 0.10-0.071; P=0.008), and chemoradiotherapy (HR, 0.33; 95% CI, 0.15-0.75; P=0.008), as well as a disseminated recurrence pattern had an impact on poor outcomes (HR, 3.94; 95% CI, 1.67-9.31; P=0.002) [7]. The present study showed that optimal surgery as the primary treatment (HR, 0.31; 95% CI, 0.13-0.72; P=0.006) and salvage treatment with chemoradiation (HR, 0.23; 95% CI, 0.08-0.7; P=0.009) improved survival after progression. However, leukocytosis (HR, 5.98; 95% CI, 2.33-15.3; P<0.001) was an unfavorable outcome.

Previous studies have shown that complete cytoreduction significantly improves OS in patients with UCS by reducing the residual tumor burden and enhancing the response to adjuvant therapy [17,18]. Harano et al. [17] conducted a multicenter retrospective study on 225 patients with UCS stages III-IV (optimal surgery, 76%) and reported median progression-free survival and OS of 11.5 vs. 8.1 months (P<0.0001) and 37.9 vs. 18 months (P<0.0001), respectively, compared with sub-optimal surgery (residual tumor >1 cm). Similarly, Gracia et al. [18] performed a sub-analysis of the SARCUT-European study with 283 UCS cases and determined that incomplete cytoreductive surgery (HR, 4.02; 95% CI, 2.68-6.18; P<0.001) was an independent factor for reduced OS. Even in a progressive setting, salvage surgery with complete cytoreduction positively affects OS [19]. Unfortunately, we could not illustrate this effect on survival, as none of the patients in this study underwent salvage surgery. In our study, we reported an optimal surgical rate of 74.07%. When comparing the extent of disease between optimal and suboptimal surgeries, we identified significant associations between suboptimal surgery and stage 4, omental involvement, and pelvic node metastasis (P<0.001, P=0.013, and P=0.02, respectively; Supplementary Table 2). Because of the low sensitivity of UCS to chemoradiotherapy, residual gross tumors significantly affect the rapid progression of the disease and adversely affect survival after disease progression.

Salvage treatment for progressive UCS is personalized depending on the site of metastasis, prior treatment, and patient performance status. Because of the aggressive nature of the tumor and the predominance of distant and disseminated recurrence patterns, chemotherapy with or without radiotherapy appears effective [7]. Our results indicated that salvage treatment with combined chemotherapy and radiotherapy had an effect on survival after progression. However, in frontline treatment, even when patients receive adjuvant therapy, they still have a high rate of disease progression. Therefore, it is unclear whether chemotherapy or radiotherapy results in a good response to UCS. Based on recent molecular studies, immunotherapy and targeted therapy may improve the survival of patients with UCS [12]. However, data are lacking, and enrollment of these patients for clinical trials is required.

Over the past decade, research has highlighted the potential impact of cancer-related inflammation on oncological outcomes in solid organs. This theory posits that tumor-related leukocytosis is influenced by the tumor microenvironment, which produces tumor-derived granulocyte colony-stimulating factor (G-CSF) and G-CSF-mediated interleukin-6, which are inflammatory cytokines that contribute to leukocytosis in EC. In addition, tumor-derived G-CSF leads to increased myeloid-derived suppressor cell infiltration, which suppresses T-cell activity. These factors promote tumor progression and chemoresistance [20,21]. Consistent with clinical studies, leukocytosis has been associated with worse oncological outcomes in EC [22,23].

The UCS surveillance program aims to detect early recurrence. Zola et al. [24] conducted a randomized, pragmatic, parallel-group, multicenter trial to determine the effectiveness of intensive and minimal follow-up in patients with EC at all stages. The intensive follow-up program included vaginal cytology, serum cancer antigen (CA)-125 levels, and chest and abdominal CT scans. Minimal follow-up was performed only for general and gynecological examinations. The results of the study illustrated that an intensive follow-up with no impact on OS and concluded that vaginal cytology, CA-125, or imaging was unnecessary for the follow-up protocol. Similarly, the National Comprehensive Cancer Network guidelines version 1.2024 (NCCN, Plymouth Meeting, PA, USA) [25] recommend surveillance through physical examination of the pelvis every 3-6 months for 2-3 years, every 6 months up to the 5 years, and annually thereafter, without routine serum CA-125 measurement and imaging, as indicated for suspected progressive disease. Our study showed that 64.81% of patients presented with abnormal symptoms before progressive disease detection, and 55.56% had progressive disease detected on physical examination, which are consistent with the findings of Nikolopoulos et al. [26]. Therefore, physicians should conduct intensive follow-ups during clinical and physical examinations to detect disease progression.

This study had some limitations. First, its retrospective design necessitated a prolonged data collection period, introducing potential biases and issues such as missing data, cause-specific survival, and loss to follow-up after primary treatment, or patients who did not return after being diagnosed with progressive disease. Second, our institute is a tertiary center in South Thailand; however, only a small number of patients were included because of the rarity of the disease. Finally, the lack of essential molecular data, such as p53 and POLE, may have limited our ability to precisely define prognostic factors and survival outcomes in patients with UCS. However, our study was conducted at a single referral center with nearly homogeneous follow-up and treatment. Owing to the insufficient data on progressive UCS, we are hopeful that our results will contribute to furthering this knowledge.

In conclusion, patients with progressive UCS have a poor survival rate. This study provides valuable insights into the OS of these patients and identifies the key prognostic factors. Optimal surgery as the primary treatment and salvage therapy with chemoradiotherapy improves survival following progression. However, the detection of leukocytosis during disease recurrence is a poor prognostic factor. Further investigations with a good study design, a large number of patients, and molecular factors should be conducted in progressive UCS settings.