Rectal GIST is relatively rare, accounting for only 5% of GISTs, our study showed rectal GISTs were 4.2% of all GISTs (138/3316), similar to the Rebecca report . The median size of rectal GISTs was 5 cm [9, 10], but the median size obtained in our study was 6.18 ± 3.02 cm, which was slightly larger than that of the above studies. The clinical symptoms of rectal GISTs are related to the size of the tumor. When the tumor is less than 2cm, no clinical symptoms are observed, and they are often found in the physical examination occasionally. However, as the tumor gradually grows into the intestinal cavity, a series of symptoms occur, including stool trait change, defecation habits change, anal discomfort or tumor rupture caused by blood in stool, and late invasion of the surrounding organs can also be manifested as hemuria and vaginal bleeding . Our study showed that rectal bleeding, change in bowel of stool, change in bowel habit and anal discomfort, were 19.8%,14.9%, 17.8% and 16.8%, respectively. It is worth mentioning that 17.8% of the patients did not have any symptoms, but were found initally durning the rectal palpation. Therefore, its clinical manifestations have no specific manifestations compared to other rectal diseases. However, rectal GISTs tends to occur in the lower segments [9, 10], in our group, 78.2% (79/101) patients had tumors located in the lower segments similar to most studies, and 69.3% (70/101) of them could be found by rectal palpation. Therefore, rectal palpation plays an important role in the early discovery and differential diagnosis of rectal GISTs.
The pathological diagnosis of rectal GISTs is mainly based on histological and immunohistochemical results . The cell morphology was divided into spindle, epithelioid, and mixed cell type, which accounted for 97.0%, 96.2% and 82.1% in our study, respectively, similar to GISTs in other sites. In terms of immunohistochemistry, CD117 and DOG-1 had the most diagnostic value, and CD34 was very significant for its diagnosis. Miettinen et al. reported that the expression rates of CD117 and CD34 in 96 cases of rectal mesenchymal tumors were 100% and 94%, respectively . In our study, the positive rates of CD117 and DOG-1 were 97.0% (98/101) and 96.2%(75/78), and CD34 also reached 82.1%(83/101), thus, CD117 was similar to Miettinen’s report, but CD34 was slightly lower. According to Miettinen’ report, KIT exon 11 mutation are common in rectal GISTs, followed by exon 9 mutation and wild-type mutation, but PDGFRA mutations are rare in rectal GISTs. Similar results were obtained in the present study, where KIT exon 11 (38/47), KIT exon 9 (4/47) and wild-type (4/47) were detected, however, no PDGFRA mutation patientswere found in our study. Rectal GISTs often has a layer of pseudo-parcapsule on the surface, rarely infiltrate along the intestinal wall, and rarely has lymph node metastasis. Hence,lymph node dissection is not necessary . Among the 101 patients in the present study,76 lymph nodes were dissected, but no metastasis occurred, thus confirming the above view.
Rectal GIST is a disease with high recurrence rate. Surgical resection is still the mostimportant treatment. The surgical principle is to complete excision, maintain the integrity of the capsule, and avoid rupture. The malignant risk of rectal GIST is higher than that ofstomach and is closer to that of intestinal GIST. Yasui, et al. reported that the proportion of rectal high risk GISTs was 45%, while the MSKCC single center reported 72.3% [4, 15]. For the 70 patients that underwent direct surgery, 45 cases were at high risk after surgery, accounting for 64.3%, which was higher than the result obtained by Yasui. However, considering that 31 patients with large diameter could not be evaluated by modified NIH criteria after IM neoadjuvant therapy, our ratio of high risk would be higher than this value, supporting the data of MSKCC. Cox regression analysis was performed on gender, age, diameter, invasiveness, ECOG Performance Score, surgical type, surgical margin, surgical scope, mitotic index, risk classification, post-operative IM adjuvant therapy of seventy patients, the results showed risk grade and post-operative IM adjuvant therapy affected DFS and OS. Therefore, for patients with intermediate and high risk rectal GISTs, postoperative adjuvant treatment with imatinib is particularly important for improving their prognosis.
For patients with large tumors, prone to intra-operative bleeding, and tumors close to the anal margin, IM neoadjuvant therapy can be considered, and this treatment will result in obvious tumor descent effect, improve the anal preservation rate, reduce the positive rate of surgical margin, and improve the prognosis of patients [16–19]. At present, the time of IM neoadjuvant therapy is appropriate within 6–12 months according to NCCN and ESMO guidelines [16, 17]. Each guideline recommends that the duration of neoadjuvant treatment be defined as the maximum response to medication. The maximum response time was defied as two consecutive enhanced CT or MRI scans indicate no remission of the tumor. At this time, surgical resection should be performed immediately given an opportunity for surgery . With prolonged drug treatment time, secondary mutations may occur during treatment. Bednarski et al., in a retrospective study of 93 patients treated pre-operatively, showed that the neoadjuvant treatment time of > 365 days was associated with an increased progression rate . Therefore, blindly prolonging the IM neoadjuvant treatment time to maximize is highly likely to lead to drug resistance and then miss the best operative timing . In our present study, 31 patients were treated for 3–15 months (6.1 ± 2.5 months). During IM neoadjuvant treatment, the patients were generally followed up and evaluated dynamically every 3 months timely to understand the effect of neoadjuvant therapy and accurately determine the timing of surgery. In the present study, the ORR was 83.9%, and the DCR reached as high as 96.8%. This result was similar to the Kanedo report involving 6 retrospective studies in 118 patients with neoadjuvant rectal stromal tumors with a response rate of 70.3% and a control rate of 99.2% .
Considering that the rectal GIST is located in the narrow pelvis with special anatomical structure and adjacent to important structures such as the reproductive and urinary systems, diameter of the tumor is the most important indicator for us to consider IM neoadjuvant therapy. Thus, we conducted a subgroup analysis of the diameter, and found that no factors, including IM neoadjuvant therapy, affecting DFS and OS in patients with a diameter of ≤ 5 cm. But, considering the small sample size, whether IM neoadjuvant therapy in patients with a diameter of ≤ 5 cm can provide survival benefit for rectal GISTs needs to be further verified. However, for patients with tumor diameter > 5 cm, univariate analysis (P = 0.002) and multivariate analysis (P = 0.028) both indicated neoadjuvant therapycan improve DFS. So, it was the same as Vallilas’s report that IM neoadjuvant therapy for specific sites or large tumors can improve the prognosis . Meanwhile, it is worth mentioning that in our study, whether the surgical margin was positive or not did not affect the prognosis and there was no need for further surgical resection, which was consistent with Cavnar’s and Gronchi's report [24, 25]. In summary, for patients with a diameter ≤ 5 cm, considering that the composition of IM neoadjuvant therapy was relatively low, with only 2 cases, the data analysis might be biased. Whether neoadjuvant therapy in patients with a diameter ≤ 5 cm can provide survival benefit for rectal GISTs needs to be further verified in a multi-center, large-sample prospective study. However, for rectal GISTs with diameter > 5 cm, our univariate and Cox regression analysis both showed that neoadjuvant treatment could improve the prognosis of patients. It could improve 3-year RFS and 5-year RFS and OS. Therefore, for rectal GIST patients with a diameter of > 5 cm, we recommend IM neoadjuvant therapy and then receive operation in order to tumor’s descending and improve the prognosis.
Moreover, from subgroup analysis of the diameter, our research showed that the neoadjuvant therapy of diameter > 5cm could improve safety of the surgery, preserve the anus, decrease possibilities of enterostomy and improve the post-operative quality of life of patients. At present, how to shrink the tumors that are difficult to be completely resected and how to improve the anal preservation rate of patients with low rectal GISTs have been widely focused on . It has been reported IM neoadjuvant therapy can reduce bleeding and improve safety of the surgery, which may be attributed to the reduction of tumor volume and the fibrosis, hyaline degeneration, and toughening of tumor texture caused by drugs, making it less prone to rupture and bleeding during surgery. Our study also confirmed the value of neoadjuvant therapy with shorter postoperative hospital stay(P = 0.001) and less bleeding (P = 0.022). Moreover, neoadjuvant therapy can significantly reduce the tumor diameter, and this condition is conducive to the implementation of organ preservation surgery [27, 28]. The MSKCC single-center study showed that IM neoadjuvant adjuvant therapy significantly increased the rate of anal preservation (92% vs. 48%). Our study showed that neoadjuvant therapy could increase the anal retention rate (93.1% vs. 72.2%, P = 0.031) and reduce the rate of enterostomy (10.3% vs. 33.3%, P = 0.037), demonstrating the value of IM neoadjuvant therapy in preserving organ function.
In conclusion, rectal GIST is a disease with special location, high malignancy and recurrence rate. Post-operative IM adjuvant treatment can reduce the recurrence and metastasis rate of the intermediate and high risk patients. IM neoadjuvant therapy can reduce tumor volume, protect organ structure and function, and improve prognosis of patients with a diameter > 5cm. However, the neoadjuvant treatment of rectal GISTs remains lacking standard at domestic and abroad. In the future, more prospective multi-center studies are needed to further explore and draw conclusions.