Clinicopathological Characteristics And Prognostic Factors For The Recurrence Of Abdominal Desmoid Tumors: Analysis Of 113 Patients From Two Chinses Hospitals

Background and purpose: Abdominal desmoid tumors (ADTs) are rare soft-tissue neoplasms that have a relatively high local recurrence rate. The purpose of the present study was to delineate the clinicopathologic features and explore the prognostic factors of ADTs. Methods: From January 2000 to January 2019, patients with ADTs who underwent macroscopically complete resection at the China National Cancer Center were included in the study. The clinicopathologic characteristics and follow-up data were carefully collected and reviewed. Prognostic factors such as age at presentation, sex, tumor location, tumor size and tumor proximity to nerves or vasculature were analyzed, and recurrence-free survival was analyzed with these factors. Results: A total of 113 patients with ADTs were assigned to the abdominal wall group (n = 66) or abdominal cavity group (n = 47) according to the tumor site. Abdominal wall DTs and intra-abdominal DTs demonstrated distinct clinicopathological features and prognoses. During a median 61-month follow-up period, twelve (10.2%) patients had local recurrence. According to the univariate and multivariate analyses, intra-abdominal tumors, large tumors, and positive margins were independent risk factors for poor prognosis. Conclusion: Compared with intra-abdominal DTs, abdominal wall DTs demonstrate different clinicopathological features and a better prognosis. Under the premise of ensuring negative margins during the rst surgical procedure, patients with abdominal wall DTs can obtain satisfactory prognoses through radical resection.


Introduction
Desmoid tumors (DTs), also known as aggressive bromatosis, are rare clonal proliferating tumors that may arise from mesenchymal stem cells [1,2] . Although DTs have no potential for metastasis or malignancy, the lesions show local aggressiveness to surrounding structures and exhibit a propensity to recur, leading to a high local recurrence rate and signi cant functional impairments and morbidity [3][4][5] . Consequently, surgeons choose more aggressive treatment strategies, including radical surgical resection [6][7][8] , radiotherapy [9,10] , systemic therapy [11][12][13] and neoadjuvant radiation with or without chemotherapy [14,15] . However, due to the rarity of DTs and their recognized unpredictable natural history, optimal treatment management strategies have not been well established.
DTs can occur anywhere in the body and cause different clinical symptoms. According to their distribution, they can be classi ed as extra-abdominal, abdominal wall, and intra-abdominal types. Intraabdominal DTs may cause intestinal obstruction, stulization or dysuria, and extra-abdominal DTs may result in neuropathic pain [16] . Currently, a large number of studies have suggested that age, surgical margin, tumor location, tumor size, and adjuvant radiotherapy are clinicopathologic prognostic factors associated with recurrence [17][18][19][20] . However, it has been shown that DTs located in the abdominal wall and abdominal cavity are relatively minimally invasive and have a relatively low recurrence rate [16,21,22] . Different desmoid tumors may have different biological make-ups and different genes, which indicates that abdominal DTs (ADTs) may represent a different disease from DTs in the girdle, head, neck, extremity or other parts of the body [8,16,23] . In addition, data on the management of ADTs are limited, and prognostic factors are not speci c. Therefore, we conducted a double-center study and enrolled 113 patients from two different Chinese hospitals, aiming to delineate the clinicopathologic features and determine the prognostic factors for recurrence-free survival (RFS) in ADTs after macroscopic complete surgical resection.

Patients
From January 2000 to January 2019, a total of 343 patients who underwent surgical resection and were pathologically diagnosed with DTs from two different Chinese institutions were retrospectively reviewed.
The inclusion criteria were as follows: (1) patients with DTs of the abdominal wall or abdominal cavity; (2) patients undergoing macroscopically complete surgical resection (R0 or R1); and (3) patients who received surgical resection as the initial treatment. Patients with unclear microscopic margin status of resection or those who were lost to follow-up were excluded. According to the above criteria, the remaining 113 patients were analyzed and formed the basis of the present study: 102 patients were from the Cancer Hospital Chinese Academy of Medical Sciences, and 11 patients were from Beijing Hospital.
The study protocol was approved by the ethics committee of the National Cancer Center, and all patients signed an informed consent form before the study.

Clinical data collection
In the present study, clinical data were collected based on electronic records and included age at diagnosis, sex, body mass index (BMI), comorbidities, previous abdominal surgery, admission status, tumor location, tumor size, tumor stage, surgical resection margins, and tumor proximity to important blood vessels or nerves. In addition, the surgical outcomes, including the duration of operation, intraoperative blood loss, postoperative complications, and postoperative hospital days, were also collected and reviewed. Tumor site was categorized as intra-abdominal or abdominal wall.
Retroperitoneal lesions were considered extra-abdominal DTs and were excluded. Surgical margin and tumor size were examined by two pathologists specializing in gastroenteric tumors, and the microscopic margin status was considered positive if the tumor was identi ed in the pathological specimen to be less than 0.5 cm from the edge of the inkblot. According to the patient's general condition and the radiologist's experience, adjuvant radiation with a median overall dose of 55 Gy was recommended. The patients received a follow-up survey every 1-2 years via outpatient visits or telephone until recurrence, death or December 31, 2019. Local recurrence was the main endpoint and was diagnosed by physical computed tomography (CT) scans or magnetic resonance imaging (MRI) [24,25] .

Statistical analysis
All data were analyzed using the Statistical Package for the Social Sciences (SPSS version 24.0, IBM Corp., Armonk, NY, United States). Quantitative data were expressed as the mean ± standard deviation, and the two groups were compared with paired Student's t-tests and Mann-Whitney U-tests for independent values for normally and nonnormally distributed values, respectively. Qualitative data and ordinal data are presented as the number of cases and percentages, and the groups were compared using χ2 tests or Mann-Whitney U-tests, as appropriate. The local RFS time was de ned as the time interval between the date of pathological diagnosis and recurrence. RFS rates were analyzed by the Kaplan-Meier method and were compared between the subgroups with the log-rank test. In addition, the Cox proportional hazards regression model was used to perform multivariate analysis to identify the independent prognostic factors. A P value of <0.05 was considered statistically signi cant.
According to the tumor sites, 113 patients were assigned to the abdominal wall group (n = 66) or the intra-abdominal cavity group (n = 47). Figure 2 shows the distribution of various tumor sites in the abdominal cavity, including stomach (n = 2, 4.3%), pancreas (n = 1, 2.1%), duodenum (n = 2, 4.3%), small intestine mesentery (n = 25, 53.2%), ileocecal mesentery (n = 7, 14.9%), transverse mesentery (n = 8, 17.0%) and sigmoid mesentery (n = 2, 4.3%). The proportion of female patients in the abdominal wall group was signi cantly higher than that in the intra-abdominal group (93.9% vs. 44.7%, P<0.001). In contrast, the proportion of patients aged 35 years or older was signi cantly higher in the intra-abdominal group than in the abdominal wall group (91.5% vs 63.6%, P=0.001). There were more patients with a previous history of cesarean section in the abdominal wall group than in the intra-abdominal group (34.8% vs. 6.4%, P<0.001). In addition, patients with lesions in the abdominal wall were less likely to present with clinical symptoms than patients with intra-abdominal lesions (13.6% vs. 40.4%, P = 0.001).
The number of patients in the intra-abdominal group with lesions greater than 10 cm was signi cantly higher than that in the abdominal wall group (27.7% vs. 7.6%, P = 0.011), and the lesions in the intraabdominal group were mostly adhered to important nerves or vasculature (42.6% vs. 13.6%, P = 0.001). In terms of surgical outcomes, the abdominal wall group had a signi cantly shorter operation time (98.7 ± 50.0 min vs. 194.5 ± 90.3 min, P<0.001) and a lower amount of intraoperative blood loss than the intraabdominal group (24.1 ± 41.1 ml vs. 136.0 ± 149.4 ml, P<0.001).

Survival analysis
The median follow-up interval was 61 months (range 9-250). Only 2 patients died at the last follow-up, and no patient died due to recurrence or complications caused by ADTs. During the whole follow-up period, a total of 12 patients developed recurrence after surgery with or without adjuvant radiotherapy: 2 (16.7%) cases of recurrence were found in patients presenting with intra-abdominal DTs, and 10 (83.3%) cases of recurrence were found in patients presenting with abdominal wall DTs. The estimated 5-and 10year RFS rates for the entire cohort were 92.8% and 89.8%, respectively ( Figure 3). It is worth noting that recurrence was still observed in one patient beyond 10 years. In addition, none of the 12 patients who were admitted with recurrent disease experienced recurrence during follow-up after surgical resection combined with or without radiotherapy.
According to the univariate analysis, tumor location, tumor size, and margin status signi cantly affected RFS (P<0.05). A signi cantly higher rate of local recurrence was associated with abdominal wall DTs, a tumor size >10 cm, and an R1 margin status ( Figure 4A

Discussion
DTs are clinically rare soft-tissue neoplasms of clonal myo broblastic tumors that originate from musculoaponeurotic structures, fascial planes and ligaments throughout the body [1,2] . Although DTs have no potential for metastasis or malignancy, the lesions show local aggressiveness to surrounding structures and exhibit a propensity to recur, and the local recurrence rate can be as high as 17.6%-30.7% [16,21,23,27] . According to the distribution of DTs, lesions can be classi ed as extra-abdominal, abdominal wall, and intra-abdominal types. The biological behaviors of DTs in different parts of the body are different, and previous studies have demonstrated that the prognosis of ADTs is signi cantly better than that of extra-abdominal DTs [25,21,23,27] . To further delineate the clinicopathologic features and prognostic factors of this rare tumor, we conducted a double-center retrospective study to provide a reference for clinical work.
We further divided ADT patients into an intra-abdominal DT group and an abdominal wall DT group to compare the clinicopathological features and prognosis factors of the two groups. Consistent with previous literature [8,23,28] , the present study revealed that compared to patients with intra-abdominal DTs, the majority of patients with abdominal wall DTs were young women with a history of cesarean section. Our study also found that the tumor size of patients with intra-abdominal DTs was signi cantly larger than that of patients with abdominal wall DTs, and most of these patients presented with corresponding clinical symptoms upon rst diagnosis. This may be due to the insidious nature of intra-abdominal DTs and the absence of obvious special clinical symptoms at the initial stage. As the tumor gradually increases and compresses the surrounding organs, it causes abdominal pain, ileus, hydronephrosis and other clinical symptoms [29] . Furthermore, we also found that the lesions in the intra-abdominal group were more likely to adhere to important nerves or vasculature than the lesion in the abdominal wall group, which may also be due to the insidious nature of intra-abdominal DTs, and the local aggressiveness of the tumor was more serious at the time of diagnosis.
Prognostic factors affecting the survival of DT patients have been previously reported. He et al. reported that younger age, extra-abdominal sites, a large tumor size and a close or positive margin status are independent prognostic factors for RFS in patients with DTs [27] . In addition, Mullen et al. reported that margin status was a signi cant prognostic factor for postoperative relapse in DT patients undergoing surgical resection [21] . Our study targeted ADTs, and the results showed that an intra-abdominal site, a large tumor size and R0 resection were all independent predictors for RFS in DT patients. Previous studies have demonstrated that the long-term prognosis of patients with abdominal wall DTs is signi cantly better than that of patients with intra-abdominal DT [28,30] . Wilkinson et al. reported the prognosis of 50 patients with abdominal wall DTs who underwent surgical resection. Within a median follow-up period of 5 years, the local recurrence rate was only 8% (4/50). Of these 46 disease-free patients, 13 were pregnant without complications due to abdominal mesh repair or tumor recurrence [28] . Consistent with previous literature reports, this study revealed that the RFS rate of patients with abdominal wall DTs was signi cantly better than that of patients with intra-abdominal DTs, and only 3% (2/66) of patients with abdominal wall DTs had local recurrence during follow-up. However, the local recurrence rate of patients with intra-abdominal DTs was 23.1% (10/47), and the tumor size was greater than 5 cm in all 10 patients with recurrence, of whom 4 patients had tumors larger than 10 cm. In addition, of the 10 patients with recurrence in the intra-abdominal DT group, 5 had positive margins after surgery. We believe that unlike abdominal wall DTs located on the body surface, due to the rarity of DTs and the absence of special clinical symptoms in the initial stage, lesions in the abdominal cavity are already large at the time of diagnosis and are closely related to the surrounding important nerves or vasculature. To preserve organ function as much as possible, it is di cult to achieve radical resection, which results in a signi cantly higher local recurrence rate. Furthermore, the muscular and fascial defects of the abdominal wall caused by radical resection can be repaired with synthetic or biological mesh, providing a reliable technical guarantee for R0 resection in abdominal wall DTs [31,32] .
There were several limitations to the present study. First, it was a retrospective study; thus, the bias from patient selection and information collection was unavoidable. Second, the period of our study was within a span of nearly 20 years, the treatment strategies for DTs have been changing, and the early pathological reports are not as normative as they are now. This led to a lack of vital information, such as data regarding SMA, β-catenin, Desmin, and Ki-67, in some patients and made it di cult to evaluate their value in predicting prognosis.

Conclusion
Compared with extra-abdominal DTs, ADTs have a better prognosis because only 9.7% of patients experience recurrence within 10 years after macroscopically complete surgical resection. Moreover, abdominal wall DTs and intra-abdominal DTs demonstrate distinct clinicopathological features and prognoses. Under the premise of ensuring negative margins during the rst surgical procedure, patients with abdominal wall DTs can obtain satisfactory prognoses through radical resection. Mortality 0 (0) 0 (0) 0 (0) / Postoperative hosptial stay (day, mean ± SD) 3.9 ± 3.8 3.2 ± 2.4 5.9 ± 6.3 0.002