A pooled analysis of risk factors of surgically treated leiomyosarcoma of the colon in adults

Background This current systematic review aimed to evaluate the role of surgical management and risk factors by pooled cases from all identified patients with colonic leiomyosarcomas. Methods The authors searched the Ovid MEDLINE, Embase, PubMed, and Cochrane databases using the keywords “colonic,” “colon,” and “leiomyosarcoma.” Risk factors of colonic leiomyosarcoma in the pooled cohort were also evaluated. Results Between 1923 and 2019, 41 cases of colonic leiomyosarcoma were identified in 22 (53.7%) males and 19 (46.3%) females, with a mean and median age of 58.7 ± 2.2 years and 56.0 years. According to univariate analysis, smaller tumor size < 8 cm was significantly associated with longer progression-free survival (HR = 6.957, 95% CI 1.405–34.442; p = 0.017), and younger age < 60 years was trending toward better overall survival (HR = 2.765, 95% CI 0.924–8.272; p = 0.069). Conclusions Colonic leiomyosarcomas are rare neoplasms with aggressive clinical behaviors. Age < 60 years and tumor size < 8 cm were favorable factors for patients’ better survival.


Introduction
In 1923, Scott firstly reported one case of colonic leiomyosarcoma (CLMS) [1]. Gastrointestinal LMS of the colon, an uncommon condition that accounts for less than 1% of all colorectal malignancies, has a strong propensity to recur and to metastasize at distant sites (liver and lung) [2]. Surgery and adjuvant chemotherapy have been used in the treatment of CLMS patients. But given the rarity of this tumor, there was no clear information about the risk factors following therapeutic strategy. Therefore, we performed an extensive literature review, and this current systematic review aimed to evaluate the role of surgical management and risk factors by pooled cases from all identified patients with CLMS . the pertinent citations. Language was limited to English. The process was shown in Fig. 1.

Case eligibility criteria
Inclusion criteria for literature cases were as follows: (1) adult patients (≥ 18 years of age) with a diagnosis of CLMS who underwent surgery; (2) availability of overall survival (OS) data. Exclusion criteria for literature cases were as follows: (1) studies published in a language other than English; (2) unavailable patient data; (3) basic research rather than clinical report without clinical data; (4) patients have developed a distant metastasis before or at surgery; (5) patients died of other causes.

Study eligibility and data extraction
Two investigators independently and in duplicate performed title and abstract screening of the studies in the search query results. Discrepancies between the review authors over the bias in studies were resolved by discussing with a third reviewer (Q.G) when needed. The following data were extracted from each study whenever possible: author and year of published articles, patient characteristics (gender and age), tumor size, treatment strategy, primary tumor gross appearance, duration of follow-up, progression-free survival (PFS), and OS. Due to incomplete and limited patients' status for recurrence and metastasis, we defined PFS for both local and distant recurrence.

Statistical analysis
Mean values are presented with their standard deviation (SD). The primary outcome of CLMS was PFS and OS, and its pertinent adverse factors were evaluated by univariate analysis. Due to the small number of patients, we did not perform multivariate Cox regression analysis. Outcomes were PFS and OS in subgroups with significant risk factors and their pertinent estimated PFS and OS time performed using the Kaplan-Meier method. All analyses were performed using IBM SPSS (version 25.0, IBM Corp.) with significance set at p < 0.05.

Survival
In prior studies, patients' status (n = 11) about recurrence and metastasis were not available. Four patients developed recurrence, and 7 patients happened to distant metastasis. The PFS rates of 30 patients with CLMS at 1, 3, and 5 years from the time of diagnosis were 74.6%, 50.2%, and 50.2%, respectively (Fig. 2a). The mean PFS was 79.0 ± 15.9 months. The OS rates of the entire series of patients with CLMS (n = 41) at 1, 3, and 5 years from the time of the diagnosis were 81.6%, 60.8%, and 45.6%, respectively ( Fig. 2b), and the mean OS was 95.5 ± 18.6 months ( Table 2).

Discussion
CLMSs are extremely rare neoplasms, and most of them have been described as case reports. In the past, LMS of the colon's prognosis has been generally considered to be a benign tumor that displayed optimism with a low propensity for recurrence and distant metastasis [1,30]. Later on, literature reported that frequent recurrences and distant metastasis have been observed in the CLMS [31]. Due to the paucity of data about CLMS, the information regarding its clinical characteristics and specific treatment was still unclear. Based on prior studies, we identified influencing risk factors for PFS and OS after surgical treatment, and aimed to increase the better understanding of this type of tumor.
LMS of the colon is slightly more frequent in males. Rao BK et al. reviewed 42 cases with CLMS that female dominance was found in his study [11]. The mean age in this study at the time of diagnosis was 56 years old, which is older than that in a literature review that reported a mean age at diagnosis of 50 years old [13]. Meanwhile, we found that older people had a decreased OS.
Based on the only complains and physical examinations, it is difficult to make an identified diagnosis of CLMS because preoperative symptoms, such as pain, diarrhea, and constipation, are insufficient evidence to make a diagnose of CLMS [12]. LMS could be exactly confirmed by the expression of smooth muscle actin and lack of CD117 [32].
Warkel et al. reported that survival of patients with the CLMS was not associated with the tumor size, but with mitotic activity [31]. In contrast, our study indicated that larger tumor size was associated with worsened PFS. One previous study consistent with our study advocated significant association between large tumor size and poor survival [33]. Unfortunately, with unavailable and incomplete data regarding mitotic activity, we failed to identify the relationship between mitotic activity and survival.
Surgery had been generally considered as the first line treatment for patients with CLMS. EH Ng et al. [34] published a review of 191 patients treated with surgery, and those who underwent complete resection has 25 months longer OS (the median time) than those with incomplete. In our series, due to the lack of details about the type of resection in the operative report of resection, we were unable to find significant difference in PFS or OS rates among patients who received different surgical treatments.
One finding of concerns in this study was the extremely high metastasis rate, and the most frequent sites of distant locations mainly in the liver, lung, peritoneum, humerus, and viscera. Even, many reviewers reported that CLMS with an aggressive clinical behavior that tends to high recurrence and metastasis after radical surgery [33,35]. Seven patients developed distant metastasis in this current review study, in addition, many patients have developed metastasis before they underwent surgery. Patients with a distant metastasis in general had a poor survival. Therefore, adjuvant treatment after surgery might be recommended in patients with malignant tumors. To the Best of Our Knowledge, however, no postoperative radiotherapy for LMS of the colon has  been reported yet. Adjuvant chemotherapy has been described in only two studies, with mixed results. Yaren A et al. [20] reported that a 66-year-old female with CLMS underwent adjuvant chemotherapy with ifosfamide plus doxorubicin after surgery, and no evidence of disease was observed during his follow-up time. Kiran P et al. [23] reported that a 54-year old male with LMS of the colon received postoperative chemotherapy with ifosfamide and doxorubicin for six cycles, but then he developed a recurrence after a disease-free period of half a  year. After a surgery for recurrence, he was still alive well without disease. With the two better results described, adjuvant chemotherapy following surgery might be optimal for patients with large LMS of the colon. However, we could not definitely confirm the role of it because of inadequate follow-up time and limited cases. Longer follow-up could be performed to identify the effect of adjuvant treatment.

Limitations
The limitations of this study were as follows: (1) the major was its retrospective nature, and selection bias always played a role; (2) the assessment methods of surgical management were undetailed described among studies, and we were unable to define which type of surgery did favor to increased survival; 3) we did not give an identified answer to a question whether better survival was beneficial from adjuvant therapy.

Conclusions
CLMS are rare neoplasms with aggressive clinical behaviors, with a mean OS of 95.5 ± 18.6 months. Some potential risk factors were associated with worse survival; younger age ≥ 60 years and tumor size ≥ 8 cm were associated with patients' decreased survival. Surgery followed by chemotherapy is recommended as the optimal treatment for CLMS. Given the rarity of this tumor, a prospective multiple-center randomized control trial should be performed.