Primary Ovarian Leiomyosarcoma Is a Very Rare Entity: A Narrative Review of the Literature

Simple Summary Primary ovarian leiomyosarcoma (POLMS) is a very rare malignancy characterized by unclear management and poor survival. We reviewed all 113 cases of POLMS reported in the literature till September 2022 to identify prognostic factors and the best treatment. Most patients received surgical resection, associated with lymphadenectomy in 12.5% of cases. Only 40% of patients received chemotherapy. POLMS is usually diagnosed at an early stage. Increasing stage and number of mitoses are associated with a worse prognosis. On the contrary, surgical resection with lymphadenectomy and chemotherapy is associated with increased survival. Ultimately, 43.4% of patients relapsed, and their mean disease-free survival was 12.5 months. There is a need for an international registry for POLMS that can help collect comprehensive and reliable data from around the world so that the best treatment can be definitively identified. Abstract Background: Primary ovarian leiomyosarcoma is a very rare malignancy characterized by unclear management and poor survival. We reviewed all the cases of primary ovarian leiomyosarcoma to identify prognostic factors and the best treatment. Methods: We collected and analyzed the articles published in the English literature regarding primary ovarian leiomyosarcoma from January 1951 to September 2022, using PubMed research. Clinical and pathological characteristics, different treatments and outcomes were analyzed. Results: 113 cases of primary ovarian leiomyosarcoma were included. Most patients received surgical resection, associated with lymphadenectomy in 12.5% of cases. About 40% of patients received chemotherapy. Follow-up information was available for 100/113 (88.5%) patients. Stage and mitotic count were confirmed to affect survival, and lymphadenectomy and chemotherapy were associated with a better survival rate. A total of 43.4% of patients relapsed, and their mean disease-free survival was 12.5 months. Conclusions: Primary ovarian leiomyosarcomas are more common in women in their 50s (mean age 53 years). Most of them are at an early stage at presentation. Advanced stage and mitotic count showed a detrimental effect on survival. Surgical excision associated with lymphadenectomy and chemotherapy are associated with increased survival. An international registry could help collect clear and reliable data to standardize the diagnosis and treatment.


Introduction
Primary ovarian sarcomas are rare and account for less than 2% of all ovarian malignancies [1]. Primary ovarian leiomyosarcoma (POLMS) is an extremely rare smooth muscle neoplasm that accounts for less than 0.1% of ovarian malignancies [2]. POLMS occurs mainly in women between the ages of 45 and 60 [2] and is usually a one-sided POLMS from January 1951 to September 2022 using PubMed (https://pubmed.ncbi.nlm.nih.gov, accessed 30, September, 2022), Scopus (https://www.scopus.com/home, accessed 30, September, 2022), Web of Science (https://login.webofknowledge.com) research, and the terminologies "primary ovarian leiomyosarcoma", "primary leiomyosarcoma of the ovary", "primary sarcoma of the ovary", "primary ovarian sarcoma" and "ovarian sarcoma". We included all English papers describing POLMS such as interventional, observational, prospective and retrospective studies and case reports. Abstracts of medical conferences, editorials, preliminary studies with animal models and previous reviews were excluded. Papers reporting tumor of uncertain diagnosis, or the studies that had scant or aggregated data were excluded. Three authors [VM, AP, LC,] performed the literature review and collected data. Discrepancies were corrected in discussions with the principal investigator [VDM], and similarly correct data extraction was reviewed by the principal investigator [VDM]. A PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow chart with summary of search results is shown in Figure 1.  We identified 152 articles on PubMed, 191 articles on Scopus and 113 articles on Web of Science databases. After duplicate exclusion, 231 records underwent first-step screening of titles and abstracts. Of these, 157 record were excluded because the tumor did not originate from the ovary or because the tumor was of uncertain diagnosis. A total of 74 full texts were considered for eligibility, and after reading them, 12 articles were excluded for being unfit according to the inclusion criteria or because they presented scant or aggregated data. An amount of 62 studies were finally included in the review, for a total of 113 POLMS patients (Table 1).
The analysis of the OS curves of patients subjected to different treatments showed that surgery improved the prognosis of POLMS ( Figure 3A)(Surgery: HR = 0.27, 95% CI = 0.08-0.96, p = 0.042, Surgery + CHT: HR = 0.21, 95% CI = 0.05-0.80, p = 0.022); however, the choice of treatment was dependent on the tumor stage ( Figure 3B) (p = 0.007), while no significant association was observed with mitotic count ( Figure 3C).   Focusing on the patients who underwent surgery, we registered a significant ferent effect of treatments on OS (p = 0.016, Figure 4A) and risk of death (p = 0.032, 4B). In particular, patients treated with surgery including lymphadenectomy com with chemotherapy showed a better prognosis, and no events of death were regi during follow-up. In this case, the choice of treatment was dependent on the tumor (p = 0.026, Figure 4C) but was also influenced by the mitotic count (p = 0.046, Figur In fact, both these variables were significantly associated with the risk of death in P patients ( Figure 4E,F). Focusing on the patients who underwent surgery, we registered a significantly different effect of treatments on OS (p = 0.016, Figure 4A) and risk of death (p = 0.032, Figure 4B). In particular, patients treated with surgery including lymphadenectomy combined with chemotherapy showed a better prognosis, and no events of death were registered during follow-up. In this case, the choice of treatment was dependent on the tumor stage (p = 0.026, Figure 4C) but was also influenced by the mitotic count (p = 0.046, Figure 4D). In fact, both these variables were significantly associated with the risk of death in POLMS patients ( Figure 4E,F).
A multivariate analysis confirmed that the stage was the principal prediction factor for the risk of death in these patients, independently from the treatment choice. (Table 3).   Figure 4A) and on the risk of death (p = 0.032, Figure 4B) of POLMS patients. Impact of stage (C) and mitotic count (D) on the choice of treatment of POLMS patients. Impact of stage (E) and mitotic count (F) on the risk of death in POLMS patients.
A multivariate analysis confirmed that the stage was the principal prediction factor for the risk of death in these patients, independently from the treatment choice. (Table 3).  Figure 4A) and on the risk of death (p = 0.032, Figure 4B) of POLMS patients. Impact of stage (C) and mitotic count (D) on the choice of treatment of POLMS patients. Impact of stage (E) and mitotic count (F) on the risk of death in POLMS patients. Subsequently, we considered only patients treated with chemotherapy and compared the OS of patients surgically treated with or without lymphadenectomy ( Figure 5A). Interestingly, a very significant difference (p = 0.0011) in OS was observed, confirming the efficacy of lymphadenectomy in improving the prognosis of POLMS patients. Similarly, the comparison between patients who underwent surgery with lymphadenectomy in combination or not with chemotherapy ( Figure 5C) confirmed the significantly improved OS in Subsequently, we considered only patients treated with chemotherapy and compared the OS of patients surgically treated with or without lymphadenectomy ( Figure 5A). Interestingly, a very significant difference (p = 0.0011) in OS was observed, confirming the efficacy of lymphadenectomy in improving the prognosis of POLMS patients. Similarly, the comparison between patients who underwent surgery with lymphadenectomy in combination or not with chemotherapy ( Figure 5C) confirmed the significantly improved OS in patients who received adjuvant treatment (p = 0.0045). Interestingly, in these subgroups of patients, the stage did not influence the choice of the most suitable treatment ( Figure 5B,D).
Our review highlights the uncertainties in the treatment of POLMS. In fact, although the guidelines recommend treating POLMS as a ULMS, most POLMSs are treated as EOCs. It should be kept in mind, however, that there are many differences between the two types of tumors despite the same site of onset. Our review also highlighted how the stage, the number of mitoses and the type of treatment are factors that can influence the prognosis of POLMS. Although POLMS tumors are treated according to EOC guidelines, several differences have emerged between these two ovarian cancers. EOCs are usually diagnosed at an advanced stage in women with a median age of 63 years [67]. In our review, women affected by POLMS had a median age of 53 years old (range: 12-84 years, SD +/− 17.6) and were for the most part Caucasian (77%, 84/109). Different from a previous study [4], POLMS was found at early stage in 67.4% (60/89) of cases. Usually, this type of tumor presents as a unilateral (93.6%, 59/63) bulky mass (15.1 cm, range 3.3-35 cm, SD +/− 6.9 cm), generally at the right ovary. The most common symptom is abdominal/pelvic pain followed by other symptoms due to the space-occupying mass such as abdominal distension, anorexia, weight loss, constipation and urinary disorders.
Preoperative POLMS diagnosis is difficult because ultrasound may be ambiguous, showing both benign and malignant features [14] (Figure 6A,B).
A recent report described imaging findings of a POLM correlated with histopathologic features [64]. Contrast-enhanced computed tomography (CT) showed a pelvic multilocular mass with heterogeneous enhancement due to colliquative intralesional areas surrounded by solid peripheral components ( Figure 7A-C).
Coronal image reconstruction with maximum intensity projection (MIP) was useful to evaluate the vascular origin of the mass by studying the complete course of the vessel [64]. Magnetic resonance imaging (MRI) showed an isointense mass to muscle on T1-weighted images and heterogeneously hyperintense on T2-weighted images, surrounded by a subtle perilesional fluid. Diffusion-weighted images (b = 800 s/mm 2 ) and apparent diffusion coefficient (ADC) map revealed restriction of the diffusion with minimum ADC value of 0.81 × 10 −3 mm 2 /s and mean ADC value of 1.24 × 10 −3 mm 2 /s. Histological diagnosis of ovarian leiomyosarcoma can also be complex [45]. This is based on knowledge of ULMS [73]. POLMS diagnosis is made when at least two of the three diagnostic criteria (coagulative necrosis, cellular atypia and mitotic index > 10 per high-power fields (HPF)) are present ( Figure 8A) [20]. However, POLMS may also be diagnosed when the mitotic index is <10 per HPF and there is no necrosis but pronounced atypia is present [15,45,46] (Figure 8B).
According to the findings of Yuksel et al. [65], POLMS patients with a mitotic count > 10 had a worse prognosis than patients with fewer mitoses (median OS: 35 vs. 63 months, respectively, p = 0.0097) (Figure 2). Commonly, mitotic count is evaluated by pathologists in making a diagnosis of cancer, and to grade malignancy, informing prognosis [74]. Mitosis is a process of cell cycle in which replicated chromosomes are divided into two new nuclei producing genetically identical cells retaining their chromosomes number [75]. Mitotic count is an indicator of the cell proliferation rate and hence the aggressiveness of several cancers such as breast cancers, ULMS, astrocytoma, gastrointestinal stromal tumours, etc. [76][77][78]. Cancers 2023, 15, x FOR PEER REVIEW 24 of 33 Figure 6. Ultrasound appearance of POLMS: a huge pelvic mass of 20 × 15 × 16 cm occupying the pelvis going beyond the transverse umbilical, it appears as a solid mass with irregular margins (red arrow), dishomogeneous echostructure due to the presence of anechoic, poorly vascularized cystic areas at color doppler (yellow star) (previously unpublished, original photos).
According to the findings of Yuksel et al. [65], POLMS patients with a mitotic count >10 had a worse prognosis than patients with fewer mitoses (median OS: 35 vs. 63 months, respectively, p = 0.0097) ( Figure 2). Commonly, mitotic count is evaluated by pathologists in making a diagnosis of cancer, and to grade malignancy, informing prognosis [74]. Mitosis is a process of cell cycle in which replicated chromosomes are divided into two new nuclei producing genetically identical cells retaining their chromosomes number [75]. Mitotic count is an indicator of the cell proliferation rate and hence the aggressiveness of several cancers such as breast cancers, ULMS, astrocytoma, gastrointestinal stromal tumours, etc. [76][77][78].
Ki-67 is a nuclear marker closely related to tumor cell proliferation and growth ( Figure 9). It correlates with tumor stage and metastasis, and its expression is significantly higher in tumors with poorly differentiated cells [48]. In a previous paper, Mayerhofer reported a Ki-67 expression of 30% [41]; in our review, Ki-67 expression ranged between 10% and 50% with a mean value of 26.3%. In the aforementioned study, Mayerhofer et al. [41] found POLMS positivity for matrix metalloproteinases MMP1 and MMP2, which may be linked to tumor aggressiveness. Bodner et al. found a high tumor positivity to B-cell lymphoma/leukemia-2 gene (BCL2) [12]. Members of the BCL-2 protein family play an important role in the control of apoptosis, an overexpression of pro-survival BCL-2 proteins or a reduction of pro-apoptotic BCL-2 proteins, both resulting in inhibition of apoptosis, with ensuing increased cell replication even in the absence of growth factors [79]. However, in our review we were unable to find a correlation between immunohistochemical markers and POLMS prognosis.
According to GCIG consensus, routine use of chemotherapy is not recommended in POLMS limited to the ovary [12]. Nevertheless, in our review 19.2% (10/52) of POLMS patients at first stage received chemotherapy [10,15,25,27,32,45,53,55,65] and in 1 patient radiotherapy was also administered [22]. Considering only patients treated with chemotherapy and comparing the OS of patients surgically treated with or without lymphadenectomy ( Figure 5A), we observed a very significant difference (p = 0.0011) in OS, confirming the efficacy of lymphadenectomy in improving the prognosis of POLMS patients. As suggested in a recent paper [65], lymphadenectomy may improve DFS. Differently from this recent review [65], we did not analyze pelvic lymph node dissection separately from lumbo-aortic lymph node dissection, but we obtained similar findings. In our analysis, the comparison between patients who underwent surgery with lymphadenectomy in combination or not with chemotherapy ( Figure 5A) confirmed the significantly improved OS in patients who received it (p = 0.0045). Notably, in these subgroups of patients, the stage did not influence the choice of the most suitable treatment ( Figure 5B,D).
It is well known that EOC should be treated in high-volume referral centers both to ensure optimal treatment with a multidisciplinary team and to ensure the best possible prognosis [85][86][87]. Since POLMS is an even rarer tumor, POLMS should necessarily be centralized both to guarantee the best possible treatment and to build databases useful for dissolving the doubts that still remain on the management of this tumor.
Although our review includes cases collected over an extensive period of time (70 years) with the risk of including cases with no verified diagnosis and subjected to non-homogeneous treatments, unfortunately with a very rare tumor such as POLM, such a long-term review would seem, to date, the only solution to obtain indications on management in the absence of better-quality data. Including as many patients as possible and analyzing as many diagnostic and therapeutic features as possible is an attempt to increase knowledge about POLMS treatment.

Conclusions
POLMS is a very rare neoplasm, and very few case reports and case series are available in the English literature. Our review underlined the difference between POLMS and EOC (age, unilaterality of lesion, increased aggressiveness, little or no increase in tumor markers) but did not resolve the dilemma about the surgical approach, i.e., should we treat POLMS as a ULMS or as an EOC? Should we perform staging lymph node dissection as in early stage EOC or should we only remove enlarged lymph nodes as in advanced EOC and ULMS? Moreover, we confirmed that advanced stage and high number of mitoses have a negative effect on survival and that the type of treatment may also influence survival. Both lymphadenectomy and chemotherapy could improve survival of patients with POLMS. However, even today there are too few data in the literature to identify the right diagnosis and to clarify the treatment of POLMS. Only an international POLMS registry could help collect clear and reliable data to standardize the diagnosis and treatment of this extremely rare and aggressive cancer.