This case highlights two important clinical issues with respect to malignant CVFs. First, a covered colonic SEMS may be useful for nonoperative treatment of a malignant CVF. Second, MRI can be valuable for determining the available treatment strategies for a malignant CVF.

In this case, a covered colonic SEMS was useful for closing a malignant CVF in a patient who required palliative care. The treatment of a malignant CVF is widely dependent on the progression of malignant disease and the general condition of the patient. In patients fit for major surgery, the standard surgical strategy including minimally invasive surgery consists of resection of the involved bowel tract containing the fistula, primary or delayed anastomosis, and closure of the bladder [1, 2, 5]. In patients unfit for major surgery, however, a proximal defunctioning stoma may be the only option that improves the patient’s quality of life [1]. However, this strategy rarely results in fistula closure, and recurrence frequently occurs. In addition, these patients may still be susceptible to recurrent UTI [6]. In the present case, a covered colonic SEMS was placed, because the patient had terminal-stage cancer, his life expectancy was short, and he was unable to manage the stoma on his own. Because the UTI was controlled immediately after stenting and stoma creation was avoided, covered colonic SEMS placement for a malignant CVF may be an alternative option for such surgically unsuitable patients.

A covered SEMS consists of inner and outer nitinol wire meshes with a hydrophobic polytetrafluoroethylene (PTFE) membrane between them and was developed to prevent tumor ingrowth into the stent lumen [7, 8]. In our case, the PTFE membrane acted as a new barrier between the intestinal lumen and the fistula [3, 4], and the stent’s radial force compressed the fistula; these two effects contributed to closure of the fistula. Although there are no clear data on the duration of CVF closure, a PTFE membrane may contribute to long-term fistula closure because of its strong physical and chemical resistance and biocompatibility [4, 7]. A major concern when using covered stents is the risk of stent migration [7, 8]. Stent migration did not occur in this case because in addition to the stenosis as a tumor-related factor, the following features of the stent may have helped prevent migration: the stent was uncovered at both ends (15 mm each), the outer wire meshes were firmly anchored to the tumor, and the oral side had a flared shape [7].

MRI is valuable for determining the treatment strategies for malignant CVFs. In a review of patients with CVFs including benign diseases, computed tomography was considered the gold standard in detecting CVFs, showing diagnostic accuracy up to 90–100%, whereas MRI was suitable for obtaining detailed information of the fistula-forming tissue [1]. In our patient, MRI clearly showed simple morphology with an ~ 15-mm fistula passing through the center of the tumor on T1- and T2-weighted images. Furthermore, T2-weighted images indicated no active inflammation of the surrounding tissue. In addition to these findings, colonoscopy readily provided a frontal view of the tumor, and we determined that fistula closure was technically feasible by routine colonic stenting [9, 10]. Finally, the covered colonic SEMS was successfully placed. However, few reports have focused on the usefulness of covered SEMS for treatment of CVFs [3, 4], and the latest European Society of Gastrointestinal Endoscopy (ESGE) guidelines for colonic stenting do not mention covered SEMS for CVFs [11]. Therefore, it is necessary to accumulate further cases to elucidate the usefulness of covered SEMS for CVFs.

The indication for colonic stenting during antiangiogenic therapy such as BV is controversial. For patients receiving a chemotherapy regimen that includes antiangiogenic therapy, avoidance of colonic stenting is weakly recommended in the ESGE guidelines [11]. In patients already receiving BV, SEMS insertion was shown to be a significant risk factor for complications requiring surgery (hazard ratio, 5.687; 95% confidence interval, 2.372–13.637; P < 0.001) [12]. However, Lee et al. [13] reported that the perforation rate was not higher in the BV group (n = 104) than in the non-BV group (n = 95) (0.9% vs. 3.2%, respectively). Because antiangiogenic therapy plays an important role in preventing increased microvascular density and facilitating the delivery of anticancer agents to the tumor [14], there is a need for more detailed studies of the administration of antiangiogenic therapy to patients with obstructive stage IV colorectal cancer who require colonic stenting.

Fistula closure using covered SEMS has been reported in various gastrointestinal fields. Many such reports are related to the upper gastrointestinal tract, especially the esophagus [15,16,17], and the ESGE guidelines strongly recommend esophageal covered SEMS placement for sealing malignant tracheoesophageal or bronchoesophageal fistulas [18]. In addition, the usefulness of SEMS placement for colovaginal fistulas has been reported. Lamazza et al. [19] found that covered SEMS placement was effective in rectovaginal fistulas, with fistula closure possible in 12 of 14 (85.7%) patients. By contrast, no reports have demonstrated the utility of covered SEMS placement for fistula formation caused by diverticulitis or Crohn’s disease. Therefore, although covered SEMS placement may be a useful technique for fistula closure, careful patient selection, including detailed imaging studies, is vital for its application because stent-related complications such as perforation are often life-threatening [11, 20,21,22].