Characteristics of appendicitis after immune checkpoint inhibitor therapy among cancer patients

Immune checkpoint inhibitor (ICI) therapy has revolutionized cancer care but is associated with immune-related adverse events (irAEs). Recent case reports raised the concern that acute appendicitis may be an irAE. In this study, we sought to describe the disease course of post-ICI therapy appendicitis and its associated complications. Adult patients who had an International Classification of Diseases code for appendicitis within the first 2 years after initiating ICI therapy from January 2010 to April 2021 and who had imaging evidence of appendicitis were studied retrospectively. 13,991 patients were identified who had ICI exposure during the study period, 44 had codes for appendicitis, 10 of whom met the inclusion criteria. Their median age at the time of diagnosis was 59 years. The median time from ICI therapy initiation to appendicitis onset was 188 days. The most common presenting symptoms were abdominal pain (70%) and fever (40%). Abscesses were present in two patients, and a perforation was present in one. All 10 patients received broad-spectrum antibiotics. Five patients needed surgery or interventional radiology drainage. Nine patients had resolution of appendicitis symptoms after treatment. Post-ICI therapy appendicitis is rare but presents similarly to and has similar complications rates as conventional appendicitis. Appendectomy remains the mainstay of treatment, but its use can be limited in cancer patients. The decision to continue ICI therapy remains at the discretion of the clinician. Further studies are needed to bring awareness to and advance the understanding of this clinical entity.


Introduction
Immune checkpoint inhibitor (ICI) therapy has revolutionized cancer care and the management of advanced malignancies. Immune checkpoints, the body's innate mechanisms of regulating the immune response and preventing autoimmunity, are often exploited by tumors to escape immune surveillance. ICI therapy targets and blocks these checkpoints, namely, cytotoxic T-lymphocyte-associated antigen 4 and programmed cell death protein 1/programmed death-ligand 1, thereby boosting the host's T cell-mediated immune responses and antitumor effects (Chenfei Zhou 2019; Kim et al. 2017;Vaddepally et al. 2020;Wang et al. 2019).
Although ICI therapy is effective, it is often associated with several immune-related adverse events (irAEs) that may affect almost any organ system (Postow et al. 2018), with the skin and gastrointestinal tract being the most frequently 1 3 involved. Combinations of different types of ICIs have increased risk of irAEs. Acute appendicitis as a possible new irAE is infrequently reported in the literature. Mangas et al. (2020) reported on an otherwise healthy 40-year-old man with metastatic melanoma who received ipilimumab and nivolumab and had an excellent clinical response. He had incidental findings that raised concern for acute appendicitis based on radiological evaluation, which was followed shortly thereafter by right-sided abdominal pain that did not resolve with use of antibiotics and required a laparoscopic appendectomy. The pathological findings from his appendix demonstrated expansion of the lamina propria by a lymphoplasmacytic inflammatory cell infiltrate with frequent eosinophils as well as neutrophilic crypt microabscesses, cryptitis, and an increased number of intraepithelial lymphocytes. Since this histological pattern was atypical for conventional acute appendicitis, his presentation was attributed to ICI therapy (Mangas et al. 2020). Similarly, Papageorgiou et al. (2021) reported on a 75-year-old woman diagnosed with stage IV lung adenocarcinoma who experienced acute abdominal pain without preceding symptomatology while receiving maintenance treatment with pembrolizumab and pemetrexed. She was found to have a contained rupture of the appendix that was managed conservatively. Subsequent endoscopic and histopathological findings from biopsy specimens obtained via colonoscopy associated the clinical and imaging findings with immune-mediated colitis. Whether this appendiceal pathology is a separate entity from or an atypical presentation of ICI-associated colitis remains unclear (Papageorgiou et al. 2021).
Conventional acute appendicitis is one of the most common general surgical emergencies worldwide, with an estimated lifetime risk of 7-8% (Körner et al. 1997;Omari et al. 2014;Stewart et al. 2014). The peak incidence usually occurs in the second or third decade of life, and it is less common at both extremes of age. Most studies have shown a slight male predominance, with a male: female ratio of 1.4:1.0. The overall lifetime risk is 8.6% for male individuals and 6.7% for female individuals in the United States (Addiss et al. 1990;Lee et al. 2010). Histopathological findings of conventional appendicitis include neutrophilic infiltration of the mucosa, submucosa, and muscularis propria. Transmural inflammation, extensive ulceration, and intramural abscesses are also seen (Carr 2000).
To understand the differences in presentations and outcomes of conventional appendicitis and post-ICI therapy appendicitis, in this study, we aimed to describe the course of post-ICI therapy appendicitis and its associated complications.

Patient selection
This was a retrospective study of adult patients at The University of Texas MD Anderson Cancer Center from January 2010 to April 2021 who had an International Classification of Diseases code for appendicitis (code 10: K35.80 or K35.89; code 9: 540.9). Patients were included if their appendicitis was diagnosed within the first 2 years after initiation of ICI therapy and if they had confirmed imaging evidence of appendicitis with or without its associated complications. The study was approved by the MD Anderson Institutional Review Board (PA18-0472). Informed Consent was waived for this study, and waiver of consent was granted.

Patient characteristics
The following information was obtained via review of the patients' medical records and pharmacy databases: demographic variables (e.g., age, sex, race), smoking status, medical comorbidities, cancer type and stage according to the seventh edition of the American Joint Committee on Cancer staging manual, type and duration of ICI therapy, chemotherapy within 3 months of ICI use, and other reported irAEs. Patients' vital statuses and survival durations were recorded at their last follow-up visit.

Characteristics, treatment, and outcome of appendicitis
The following appendicitis-related information was collected from the patients' medical records: presenting symptoms, time to onset of symptoms from ICI therapy initiation, imaging findings, histological features, treatment received, and outcomes. Treatment of appendicitis was consisted of antibiotics, surgery, and/or interventional radiology (IR) drainage. Outcomes of treatment were classified as persistent symptoms or response (e.g., improvement in symptoms, resolution of symptoms).

Baseline patient characteristics
Of the 13,991 patients who received ICIs at MD Anderson in the study period, 44 had International Classification of Diseases diagnostic codes for appendicitis. Our final sample comprised 10 patients who met the inclusion criteria. The patient selection details are shown in Fig. 1. Their median age at the time of appendicitis diagnosis was 59 years (interquartile range [IQR], 55-60 years). Seven patients were male. The most common malignancies were melanoma (n = 4) and genitourinary cancers (n = 3). Nine patients had stage IV cancer. Most of the patients received anti-programmed cell death protein 1/programmed deathligand 1 monotherapy (60%) ( Table 1). The median followup duration from the onset of appendicitis was 28.5 months (IQR, 14.0-35.5 months).

Appendicitis characteristics and disease course
The median time from ICI therapy initiation to appendicitis onset was 188 days (IQR, 46-386 days), and the median number of ICI doses received was 4 (IQR, 2-15). The most common presenting symptoms were abdominal pain (70%) and fever (40%). Abscesses were present upon imaging in two patients, and a perforation was present in one patient. No patients had evidence of concurrent colitis. All 10 patients received broad-spectrum antibiotics. Five patients needed a surgical appendectomy or IR drainage ( Table 2). Details of the antibiotics administered are listed in Supplemental  Table S1.
Nine patients had resolution of appendicitis symptoms after treatment. ICI therapy was permanently discontinued  Table 1 Baseline demographic characteristics of the study patients (n = 10) a Gastrointestinal adverse events were defined according to the Common Terminology Criteria for Adverse Events (version 5.0). Some patients had multiple irAEs CTLA-4 cytotoxic T-lymphocyte-associated antigen 4, PD-1 programmed cell death protein 1, PD-L1 programmed death-ligand 1 1 3 in two patients due to appendicitis and cancer progression. One of these patients resumed non-ICI cancer therapy. All 10 patients required hospitalization, with a median length of stay of 9 days (IQR, 2-11 days).

Imaging and histopathology
All 10 patients underwent computed tomography (CT) that confirmed the diagnosis of appendicitis. Common radiological features included appendiceal dilation, appendiceal wall thickening, periappendiceal inflammation, and fat stranding. Two patients had radiographic evidence of contained perforation, and one patient had abscess formation reported as a "rim-enhancing partially walled-off collection adjacent to the cecum" (Fig. 2). Pathology of the patients' surgically resected appendices included findings of markedly increased apoptosis in crypt epithelium and transmural infiltration of neutrophils with abscess formation and serositis (Fig. 3).

Characteristics of patients with appendicitis-related complications
Two patients had complications related to appendicitis, both of whom received anti-cytotoxic T-lymphocyte-associated antigen 4 regimens (Table 3). One patient developed an abscess and had perforation of the appendix. Since this patient was a poor candidate for appendectomy and interventional radiology was unable to place a drainage tube, the patient received antibiotic therapy alone (piperacillin-tazobactam and metronidazole). Serial CT scans revealed improvement in inflammation and abscess dimensions. Subsequently, the patient was discharged from the hospital in stable condition while continuing antibiotic therapy. The other patient had a history of metastatic melanoma and relapsed acute myeloid leukemia treated with stem cell transplantation. Acute appendicitis was diagnosed in the setting of neutropenic fever. An appendectomy was not performed due to neutropenia, and the patient was treated with antibiotic therapy alone (daptomycin and ertapenem). Repeat CT scans showed an appendiceal perforation, but again, surgical management was deferred given the patient's comorbidities. The patient had three more hospital admissions for this episode of appendicitis because of persistent symptoms. The patient ultimately completed 8 weeks of treatment with broad-spectrum antibiotics with added antifungal coverage. Interval CT scans revealed recurrence of appendiceal inflammation, and the patient continued to experience clinical deterioration over the ensuing months, including ICI pneumonitis, which ultimately led to death.

Discussion
Although ICI therapy has changed the landscape of cancer treatment, irAEs continue to pose a significant challenge to its use. Case reports have suggested that immune-mediated appendicitis is one such adverse event. However, it is still poorly understood, and whether this condition is pathologically or histologically different from conventional appendicitis is unclear (Mangas et al. 2020;Papageorgiou et al. 2021). Additionally, whether it is a separate entity from ICI colitis or lies on a spectrum of disease remains unclear. In this study, we addressed this knowledge gap by describing the characteristics and disease course of post-ICI therapy appendicitis and its associated complications.
Researchers have extensively studied ICI-mediated colitis over the past 5 years. It often presents with diarrhea, abdominal pain, rectal bleeding, and fever (Abu-Sbeih and Wang 2020; Wang et al. 2018a, b, c;Wang et al. 2018a, b, c). Whereas some of these symptoms also occur with post-ICI therapy appendicitis as described herein, diarrhea, a predominant symptom of colitis, did not occur in our cohort. The

Fig. 3 Pathological features of appendicitis and complications. a, b
Markedly increased apoptosis in the epithelium of the appendiceal crypts. c, d Transmural acute inflammation with infiltration of neu-trophils throughout the full thickness of the appendiceal wall and abscess formation, indicating impending perforation absence of this symptom may help differentiate ICI colitis from post-ICI therapy appendicitis clinically. Radiological patterns of appendicitis can be pathognomonic for the disease, markedly distinct from the appendiceal wall thickening commonly described in colitis cases. Moreover, the mainstay of medical management of conventional appendicitis is antibiotics, which drastically differs from the immunosuppressive therapy used for ICI colitis. In fact, antibiotic therapy has had a negative impact on the course of ICI-mediated colitis and the long-term survival of patients with ICI colitis (Abu-Sbeih et al. 2019). In our cohort, no patients received steroids or immunosuppressive agents. Also within our cohort, post-ICI therapy appendicitis appeared to typically occur in isolation following ICI therapy, although authors have reported the co-existence of appendicitis and colitis (Papageorgiou et al. 2021). Despite their similar clinical presentations and adjacent anatomical locations, we suspect that these two conditions are different entities, and further investigation to distinguish between them and guide appropriate treatment is warranted. Conventional acute appendicitis is one of the most common general surgery emergencies worldwide, with an estimated lifetime risk of 7-8% (Körner et al. 1997;Omari et al. 2014;Stewart et al. 2014). The peak incidence is usually in the second or third decade of life. However, the patients with post-ICI therapy appendicitis in our cohort had a median age of 59 years, with a range of 55-60 years. This trend of advanced age at onset was also seen in previous post-ICI therapy appendicitis case reports (Mangas et al. 2020;Patel et al. 2021). One explanation for this discrepancy may be the advanced age of cancer patients receiving ICIs. This difference aside, post-ICI therapy appendicitis in this study had presenting symptoms similar to those of conventional appendicitis, consisting of abdominal pain, fever, nausea, and vomiting. Radiological findings of appendiceal dilation, appendiceal wall thickening, periappendiceal inflammation, and fat stranding were common to both types of appendicitis. Moreover, pathological analysis of gross surgical specimens obtained in our study commonly revealed comparable findings between post-ICI therapy appendicitis and conventional appendicitis: acute appendicitis with transmural inflammation and serositis. Of note, the rate of complications associated with appendicitis (perforation or abscess) was about 20% in our cohort, similar to that for conventional appendicitis (Omari et al. 2014). However, post-ICI therapy appendicitis may foreshadow future irAEs as seen in four of our patients who resumed ICI therapy. This is not a risk in patients with conventional appendicitis.
The management of both conventional appendicitis and post-ICI therapy appendicitis is broadly classified as surgical or nonsurgical treatment. Historically, conventional appendicitis has been managed using appendectomy in the general population. Recently, nonoperative management has become a feasible alternative for uncomplicated appendicitis; in complicated cases, appendectomy within 24 h of presentation is recommended (Di Saverio et al. 2020;Teng et al. 2021). With this in mind, cancer patients are typically considered poor surgical candidates due to the clinical complexity of their underlying cancer as well as other comorbid conditions. Limited studies have shown that both nonsurgical management and early appendectomy are suitable strategies for cancer patients (Park et al. 2021;Santos et al. 2016). Within our cohort of post-ICI therapy appendicitis patients, nonsurgical management with antibiotics or IR drainage was employed more often than was surgical intervention. We observed no significant differences in outcome of surgical and nonsurgical management within our cohort. However, the significance of this finding could certainly be limited due to the small sample size.
As the most applied medical treatment of appendicitis, broad-spectrum antibiotics are usually used for coverage for gram-positive, gram-negative, and anaerobic organisms. A meta-analysis by Talan et al. (2019) described the conservative management of uncomplicated conventional appendicitis with antibiotics. In that study, the researchers found that most treatments included an initial parenteral regimen of a second or third-generation cephalosporin plus metronidazole, amoxicillin-clavulanate, piperacillin-tazobactam, or a carbapenem. This was followed by administration of an oral regimen of a fluoroquinolone, an advanced-generation cephalosporin plus metronidazole, or amoxicillin-clavulanate upon hospital discharge. Of note, whereas GI infections are typically caused by gram-negative organisms and anaerobes, gram-positive organisms may also be present, albeit to a lesser extent, necessitating broad-spectrum coverage (Leigh et al. 1974;Merlin et al. 2010;Rautio et al. 2000;Song et al. 2018). The antibiotic regimens used in our study were unique to each patient. This can be attributed to each patient's distinctive circumstances in terms of their immunocompromise, comorbidities, and severity of presenting illness. Nonetheless, all patients received some combination of antibiotics that are routinely used to treat conventional appendicitis, with occasional addition of less frequently used antibiotics like vancomycin and aztreonam or antifungals such as caspofungin (Di Saverio et al. 2020;Teng et al. 2021). Treatment of uncomplicated conventional appendicitis can last anywhere from 4 to 15 days, with the most common duration being 7-10 days (Becker et al. 2018;Di Saverio et al. 2020;Talan et al. 2019). Following an appendectomy, current guidelines recommend 1-2 weeks of postoperative antibiotic therapy, whereas a few studies have suggested that 3 days is sufficient (de Wijkerslooth et al. 2019;Mazuski et al. 2017;van den Boom et al. 2020;van Rossem et al. 2014). Given the complicated medical history of the patients in our study, the durations of antibiotic therapy exhibited great variation, ranging from 5 days to 6 weeks. Longer treatment durations were utilized when surgical appendectomy was not achievable. Overall, the management of conventional appendicitis and that of post-ICI therapy appendicitis have some striking similarities. However, given the complexity of patients presenting post-ICI therapy appendicitis, a more individualized approach to treatment than that of conventional appendicitis may be necessary regarding the decision to operate, antibiotic selection, and treatment duration. This is especially true considering the additional concerns faced by patients receiving ICIs. For instance, studies have shown that the use of antibiotics, especially those with anaerobic activity, after ICI therapy is associated with an increased risk of severe ICI colitis, likely through disruption of the gut microbiome (Abu-Sbeih et al. 2019). Although the fundamental treatment remains the same for both types of appendicitis, more factors must be considered with the immune-related form of the disease. As such, the optimal treatment strategy for post-ICI therapy appendicitis requires further investigation.
Recently, investigators have used next-generation sequencing to examine the appendiceal microbiome in pediatric patients with appendicitis. Findings of this analysis revealed an increased abundance of fusobacteria and certain other oral cavity flora (Jackson et al. 2014;Zhong et al. 2014). Additionally, studies have demonstrated that the intraluminal microbial composition differed significantly depending on the degree of inflammation in pediatric patients with acute appendicitis (Schülin et al. 2017). Another study demonstrated differences in the composition of the intestinal microbiota of appendicitis patients and healthy individuals; specifically, the diversity of the phyla Firmicutes, Actinobacteria, Fusobacteria, and Verrucomicrobia were lower in the patients with appendicitis (Peeters et al. 2019). Microbiome analysis of patients with post-ICI therapy appendicitis has yet to be performed, but based on our current knowledge of the microbiome in patients with conventional appendicitis, certain microbiological signatures may be discovered. Furthermore, microbiota analysis has greatly increased our understanding of ICI-mediated colitis and could be another way of differentiating post-ICI therapy appendicitis from conventional appendicitis or ICI-mediated colitis (Abu-Sbeih et al. 2019;Dubin et al. 2016;Wang et al. 2018a, b, c). The distinction of post-ICI therapy appendicitis from ICI-mediated colitis is a particular concern, as therapies for ICI-mediated colitis, including fecal microbiota transplantation, have had documented success (Wang et al. 2018a, b, c). If these two entities lie on a spectrum, then effective treatments of ICI-mediated colitis would also be beneficial for post-ICI therapy appendicitis.
To the best of our knowledge, our sample is one of the largest of post-ICI therapy appendicitis cases to date. Nonetheless, our study had certain limitations that reduce its applicability. First, this is a retrospective, single-center, small-scale study. Therefore, selection bias and all limitations inherent in this study design are present. Second, given the overlap of nonspecific GI symptoms of appendicitis and other GI diseases, the frequent use of antibiotics in cancer patients, some cases may have been missed, leading to underestimation of the real incidence of post-ICI therapy appendicitis. Third, the complexity of cancer patients and their comorbidities limited the surgical options for select patients and may have confounded their outcomes. Fourth, the rarity of post-ICI therapy appendicitis makes drawing conclusions from the data difficult; at best, only inferences can be made, and even then, making any generalizations would be difficult considering the variations in the cases.
Appendicitis directly attributed to ICI therapy is rare. Compared with conventional appendicitis, it manifests at an older age but with similar clinical presentations and comparable complication rates. The management strategies for post-ICI therapy appendicitis overlap those for conventional appendicitis, with appendectomy being the mainstay of treatment, although it is often limited by the complexity of cancer patients. The appropriateness of continuing ICI therapy after an episode of appendicitis has yet to be delineated. Further studies are needed to bring awareness to this clinical entity and advance understanding of its management.