Gangrenous ischemic proctitis is extremely rare because of the abundance of collateral vessels in the rectum [17,18,19,20]. The blood supply of the rectum derives from the inferior mesenteric, internal iliac, and internal pudendal arteries. Ischemic gangrenous proctitis is mainly described in patients with significant atherosclerotic and cardiac risk factors in the setting of hemodynamic instability [18,19,20]. However, our patient did not have any such risk factors. Aside from arterial occlusion, venous insufficiency can also lead to ischemia and necrosis of the tissue, since the rectum has abundant venous drainage to the portal and caval systems [17]. In earlier reports of COVID19-related gastrointestinal perforations, both upper [3] and lower gastrointestinal perforations [3,4,5,6,7,8,9,10,11,12, 21,22,23,24] have been described. There are 4 cases of rectal perforation related to COVID19 [22,23,24]. Two of these 4 patients went into septic shock on the day of hospitalization, and CT showed rectal wall thickening with free air around the mesorectum and an inflammatory stranding of perivisceral fat tissue; both these patients died without surgery within 24 h after hospitalization [23, 24]. One of the remaining 2 patients had a 1 cm diameter perforation over the anterior wall of the upper one-third of the rectum, which was repaired and a decompressing transverse loop colostomy was performed; this patient survived postoperatively [22]. In the other patient, CT showed rectal wall thickening surrounded by extraluminal free air with a small-sized fluid collection; this patient was treated conservatively and survived [24]. While conservative management is appropriate in early cases or cases with small perforations with superficial ischemia, rectal perforation cases related to COVID19 have a high mortality rate and require emergency surgery.

In our present case, the lower rectum and the surrounding mesorectum were completely gangrenous and the upper rectum was partially necrotic. The patient underwent emergency surgery and survived postoperatively. Then she underwent a resection of anal canal with residual necrotic rectal and mesorectal tissue at the reoperation and was discharged from the hospital without severe complications.

Mesenteric venous thrombosis is often caused by acquired hypercoagulable states, inherited disorders, and intra-abdominal traumatic or inflammatory conditions [25]. Many studies have reported that SARS-CoV-2 can lead to a prothrombotic state, reflecting the high cumulative incidence of associated thrombotic events, particularly in patients receiving intensive care [26, 27]. Severe COVID19 is associated with increased levels of proinflammatory cytokines, and cytokine storm initiates a prothrombic condition [15, 27]. There are some reports of mesenteric venous thrombosis related to COVID19 infection [28, 29]. In the reports of gastrointestinal perforation occurring in COVID19 patients, vascular thrombosis in the mesenteric vein was histologically recognized, and it was proposed that mesenteric venous thrombosis caused the gastrointestinal perforation [6, 10,11,12, 21].

In our case, the patient was in a highly prothrombotic state due to a cytokine storm resulting from severe COVID19 infection, which may have caused the mesenteric venous thrombosis. Histologically, thromboses of over 1 mm diameter in the mesentery veins were recognized in the rectum close to the rectal necrotic tissue. Because the distal rectum is usually spared due to its collateral blood supply, it may not be clear that the cause of rectal necrosis was limited to mesenteric venous thrombosis. Not only mesenteric venous thromboses, but also hemodynamic instability derived from severe COVID19 infection may cause ischemia of the entire rectal collateral blood supply. Otherwise, initially, a relatively small rectal perforation caused by rectal mesenteric venous thrombosis may have gradually worsened to rectal necrosis due to the delay in the clinical expression of peritonitis. However, we believe that mesenteric venous thromboses were the main trigger of rectal ischemic necrosis in this case.

CAR-T therapy is a promising immunotherapeutic approach and is currently used for several hematologic malignancies [13]. CAR-T therapy carries the risk of inflammatory toxicities, such as CRS [30, 31]. In patients with progressive hypotension or hypoxia because of CRS, the IL-6 receptor antagonist tocilizumab has been the primary therapy [32, 33]. Tocilizumab has also been used for people with severe COVID19 infection [34, 35]. To date, there have been no reports of a direct relationship between CAR-T therapy and gastrointestinal perforation. However, gastrointestinal perforation has been described as a rare but potentially severe complication of tocilizumab, although the pathophysiology underlying this relation is poorly understood [36,37,38,39]. Several cases of intestinal perforations in COVID19 patients administered tocilizumab have been reported [3, 4, 39]. In regard to steroid therapy, it has been reported that corticosteroid use was associated with increased risk of gastrointestinal perforation [40,41,42]. Corticosteroids are well known to inhibit wound repair via global anti-inflammatory effects and suppression of cellular wound responses, and to increase the risk of wound infection [43]. High-dose corticosteroids such as steroid pulse therapy have been reported to decrease the clinical expression of peritonitis to the point that recognition and, therefore, treatment of gastrointestinal perforation were markedly delayed, resulting in high mortality rates [40]. Cases of intestinal perforation cases in COVID19 patients receiving steroid therapy have also been reported [3, 4, 8, 9]. Patients with baseline immunosuppression with cancer receiving immunotherapy regimen including CAR-T therapy had worse COVID19 severity and cytokine storm [14].

In our case, cytokine storm and prothrombic condition were initiated by severe COVID19 infection after CAR-T therapy. COVID19-associated coagulopathy triggered the thromboses in the mesenteric vein and eventually caused rectal ischemic necrosis. In addition, tocilizumab may have facilitated the rectal perforation. Although there was no direct relationship between the CAR-T therapy and the patient’ rectal necrosis, we speculated that the CAR-T therapy was affecting the patient’s clinical course to some extent. Moreover, the wound healing-suppressive and infectious effects of steroid therapy may have exacerbated the rectal perforation derived from rectal mesenteric venous thrombosis, while the steroid pulse therapy with high-dose corticosteroids for COVID19 pneumonia could have delayed a proper diagnosis by mitigating the clinical presentation of anus pain, eventually led to worsening of the rectal necrosis.

Rectal necrosis is a life-threatening disease and emergency surgery is the only option [18,19,20]. The surgical management of rectal necrosis is controversial and depends on the clinical status of the patient and the surgical findings. Maun et al. [18] proposed that complete proctectomy with end colostomy may be necessary as a life-saving measure in cases in which the ischemia and gangrene involve the entire rectum all the way to the dentate line, because the retention of dead rectum may create a persistent source of sepsis. Azimuddin et al. [20] reported 3 cases with ischemic gangrene of the rectum, 2 of which were surgically found to have complete gangrene of the rectum extending to the dentate line, and abdominoperineal resections (APRs) were performed. In the other patient, severe gangrene of almost the entire rectum except the last 2 cm above the dentate line was found, and because the anal canal was viable, the patient was able to avoid APR and instead underwent a low anterior resection with Hartman’s procedure. The surgeon considered that leaving behind a gangrenous rectal segment would create a source of persistent sepsis and thus the segment should be removed. On the other hand, in a report by Nassif et al. [19], a patient with necrosis of the rectum and anus with absent sphincter tone as well as extensive necrosis of the descending and sigmoid colon received a transverse colostomy and mucous fistula, leaving the gangrenous descending and sigmoid colon with the rectum in place. Nassif et al. contended that APR appears to be associated with more severe mortality in an emergency setting and would add substantial risk to the primary procedure.

In our case, if the anal canal had been viable at the emergency surgery, we could have stapled off and divided the anorectal stump, and performed a Hartman’s procedure. However, the rectum was completely necrotic, and the necrosis extended all the way to the anal canal, and we only performed the removal of rectal necrotic tissue and resection of the remaining rectum with end colostomy. After the emergency surgery, the patient’s recovery was prolonged and eventually she underwent a resection of anal canal with residual necrotic rectal and mesorectal tissue because the remaining gangrenous rectal segment had become a source of persistent purulence. It is not clear whether we should carry out APR even at the emergency surgery, but in our case the patient was in an unstable condition due to severe COVID 19 infection. Therefore, we avoided APR at emergency surgery, and the patient survived postoperatively and was finally discharged from the hospital without severe complications.