This case demonstrated a drawback of conversion surgery for rectal resection after LP. We also showed the potential effectiveness of LP therapy promoting the transformation of unresectable advanced, chemo-resistant endometrial cancer to resectable cancer.

Rectal anastomotic leakage is a fatal complication in colorectal surgery. Previous reports of colon cancer have reported an incidence rate of rectal anastomotic leakage of 3.2–7.3% in colorectal cancer and 2.9–4.2 % in ovarian cancer [8,9,10]. The known risk factors for colorectal anastomotic leakage after colorectal surgery are low anastomotic levels, malnutrition, anemia, liver comorbidity, additional surgery, and long operation time [8,9,10]. However, the pre-operative administration of TC therapy is not considered a risk factor for anastomotic leakage [9]. In the present case, a long operation time (over 8 h) and additional surgeries (such as hysterectomy and pelvic peritoneal stripping) might increase the risk of anastomotic leakage. However, the patient had normal nutritional status, good liver function, and no severe complications during the operation. Besides, the phase 3 trial of KEYNOTE-775, which compared treatment with LP therapy versus conventional chemotherapy in patients with advanced or recurrent endometrial cancer, information on the risks of bowel resection after LP therapy was not reported. Although LP therapy for advanced endometrial cancer and renal cell carcinoma was approved by Food and Drug Administration since 2021, there is only a case report of conversion surgery without bowel resection for renal cell carcinoma before LP therapy (Table 1) [7]. We have described three cases of conversion surgery with complete tumor resection after lenvatinib with or without immune checkpoint inhibitors for thyroid, hepatocellular, and renal cancers. Although all these patients had a 7-day-withdrawal before conversion surgery, mild bile leak was reported in only one case of partial hepatic resection after lenvatinib and the anti-PD-1 antibody camrelizumab for hepatocellular carcinoma (Table 1). However, none of the patients underwent gastrointestinal or colorectal resection. In our case, with lenvatinib and pembrolizumab, a longer preoperative lenvatinib withdrawal period (12 days) was used; however, postoperative colorectal anastomotic leakage occurred (Table 1).

Lenvatinib is an oral multi-kinase inhibitor that selectively inhibits vascular endothelial growth factor (VEGF) receptors 1–3, fibroblast growth factor (FGF) receptors 1–4, and other proangiogenic tyrosine kinases and has a potent antiangiogenic property [11]. The inhibition of VEGF receptors, through their antiangiogenic properties, reduces the capillary bed in tissues by 22–59% [12]. FGF signaling plays a critical role in wound healing by promoting processes such as cell proliferation, angiogenesis, and tissue remodeling. FGF signaling stimulates the migration of fibroblasts to the wound site, enhances collagen production, and supports the formation of new blood vessels, all of which are essential for effective tissue repair and regeneration [13]. Therefore, the surgical decisions after administration of antiangiogenic drugs like lenvatinib must be taken with caution because of the ability of the drug to delay wound healing [14]. However, the surgery of the low anterior resection of the rectum after lenvatinib therapy has not been reported, though some reports have demonstrated that lenvatinib increased the risk of gastrointestinal perforation [3, 14, 15]. The half-life of lenvatinib is approximately 28 h [11]. In surgeries for other types of cancer after lenvatinib treatment, presurgical pauses in medication ranging 4–10 days have been reported [5, 6, 16]. Based on these considerations, we performed conversion surgery after a 12-day suspension of lenvatinib administration. However, an anastomotic leakage was diagnosed on POD 11. Anastomotic leakage typically occurs within the first 7 days, postoperatively [17]. In this case, no signs of infection (which are common signs to anastomotic leakage) were observed and the leakage occurred later than the typical timeframe after colorectal surgery. The antiangiogenic effects of lenvatinib may have resulted in reduced blood flow to the anastomotic site, thereby delaying the healing process. Although generalizing from a single case is not possible, these findings suggest the need for greater caution in intestinal surgery after lenvatinib administration. Moreover, bevacizumab which is an antiangiogenic antibody, is one of the risk factors for anastomotic leakage [18]. The appropriate interval between the last dose of bevacizumab and elective surgery is unknown; however, the half-life of bevacizumab is estimated to be 20 days. Thus, the National Comprehensive Cancer Network guidelines suggest withholding bevacizumab at least 6 weeks before surgery [18]. Owing to the VEGF inhibitory effect, the capillary density in normal tissues may remain reduced even after lenvatinib discontinuation and the subsequent decline in blood concentration, potentially disrupting normal tissue architecture. Currently, the evidence is limited regarding the appropriate preoperative discontinuation period for lenvatinib and the time needed for the tissues to recover to a safe state for surgery remains unclear. However, considering the antiangiogenic mechanism of lenvatinib, a preoperative withdrawal period close to that recommended for bevacizumab may be advisable.

Pembrolizumab is one of the immune checkpoint inhibitors and has the potential to induce immune-related adverse events of bowel diseases such as colitis and intestinal perforation [19]. However, a meta-analysis investigating the effects of a combination of immune checkpoint inhibitors with neoadjuvant chemotherapy for the preoperative treatment of locally advanced esophageal cancer demonstrated that the immune checkpoint inhibitors did not increase the incidence of postoperative anastomotic leakage [20]. Additionally, in a phase 1 study of neoadjuvant nivolumab monotherapy for resectable gastric cancer, the incidence of anastomotic leakage reported in 2 out of 30 patients, and the frequency was not higher than that after neoadjuvant chemotherapy with conventional chemotherapeutic agents [21]. As indicated by these studies, these reports suggest that immune checkpoint inhibitors are not directly associated with an increase in anastomotic leakage. In this case, pembrolizumab was discontinued 1 month before conversion surgery, in accordance with the protocol reported in previous studies [20, 21]. Therefore, we believe that the preoperative administration of pembrolizumab is unlikely to be a risk factor for anastomotic leakage in this patient.

Regarding the diagnosis of the left ovarian tumor, based on the Scully’s criteria, the possibility of a double primary cancer cannot be entirely excluded [22]. However, in this case, endometriosis was not observed around the ovarian lesions. We have previously investigated synchronous primary corpus and ovarian cancers and reported that endometriosis is a significant characteristic of synchronous primary cancers [23]. Considering this along with the overall clinical picture, we concluded that this represents ovarian metastasis of endometrial cancer.

Antitumor efficacy of conversion surgery after LP therapy for unresectable uterine cancer is still unknown. In the KEYNOTE-775 trial, the median PFS for LP therapy in the proficient mismatch repair population was reported to be 6.6 months [3]. In this case, the patient mostly responded to LP therapy at 6 months; however, a slight increase was noted in tumor mass in some areas suggesting imminent LP therapy resistance. Although the KEYNOTE-775 trial did not include patients in which conversion surgery was performed after LP, we established that surgical treatment at this timing should be appropriate in our case. Pathological examination of the resected specimen showed that a significant number of viable tumor cells were intact, suggesting the limited efficacy of LP (Fig 1C). The left ovarian tumor, which had increased in size, exhibited fewer tumor-infiltrating CD4-positive and CD8-positive cells compared to the uterine tumor, suggesting different therapeutic responses between the two lesions (Fig. 2B). The lower presence of both CD4-positive and CD8-positive T cells in the growing left ovarian tumor than in the uterine tumor suggests that immune tolerance after LP therapy may have occurred, potentially leading to resistance to LP therapy in the left ovarian tumor.

At present, there is no evidence to support neoadjuvant setting with LP as a precursor to conversion surgery in advanced endometrial cancer. If more cases where surgery become possible after LP and has favorable outcomes are observed and reported, clinical trials using LP therapy prior to surgery in such cases could be explored.

In conclusion, we conducted a conversion surgery after LP therapy for unresectable advanced endometrial cancer. However, postoperative vigilance for anastomotic leakage is warranted.