Ipilimumab 3 mg/kg was the first agent to demonstrate improved survival in previously treated patients with metastatic melanoma in a phase 3 trial (MDX010-20). Ipilimumab produced a characteristic spectrum of immune-related adverse events (irAEs) of special interest, consistent with its immune-based mechanism of action.
In MDX010-20, 676 previously treated patients were randomized 3:1:1 to receive ipilimumab 3 mg/kg plus the glycoprotein 100 melanoma antigen vaccine (gp100), ipilimumab 3 mg/kg + placebo, or gp100 vaccine + placebo. For the current report, the authors conducted a detailed analysis of the time to onset and resolution of irAEs associated with ipilimumab therapy.
Grade 2 through 5 irAEs generally developed during the induction phase of treatment (0-12 weeks). Most, including grade 3/4 irAEs, were reversible when managed with treatment guidelines using vigilant monitoring and corticosteroids. The median time to resolution (to grade 1 or 0 or to the grade at baseline) of irAEs that had an onset during the induction phase was approximately 6 weeks for grade 2 through 4 irAEs and 8 weeks for grade 3 and 4 irAEs. Across the entire study duration, most grade 2 through 4 irAEs resolved within 12 weeks.
Ipilimumab, a fully human monoclonal antibody against cytotoxic T lymphocyte-associated antigen-4 (CTLA-4), was the first agent to demonstrate a survival benefit in previously treated patients with metastatic melanoma in a phase 3 trial.1 In the MDX010-20 trial, induction treatment with ipilimumab 3 mg/kg monotherapy significantly improved overall survival (OS) (hazard ratio for death compared with the glycoprotein 100 melanoma antigen [gp100] alone, 0.66; P = .003). Rates of OS in the ipilimumab group and the gp100 group were 45.6% and 25.3% at 12 months and 23.5% and 13.7% at 24 months, respectively.1 On the basis of that study, ipilimumab 3 mg/kg was approved by the US Food and Drug Administration for patients with metastatic melanoma and by the European Commission for previously treated patients with metastatic melanoma.
CTLA-4 is a negative regulator of T-cell activation and proliferation that acts to limit immune responses and dampen autoimmunity.2, 3 CTLA-4 blockade with ipilimumab inhibits the CTLA-4-mediated immunomodulatory effect to enhance tumor-specific immune responses, but it also removes CTLA-4-mediated protection from autoimmunity, resulting in increased or excessive immune activity (Fig. 1). Consequently, ipilimumab treatment typically induces drug-related adverse events (AEs) of special interest that potentially have immune-related etiologies, classified in MDX010-20 and herein as immune-related AEs (irAEs). During phase 2 trials of ipilimumab, these most commonly comprised gastrointestinal (GI), skin, endocrine, or liver toxicities.2, 4
Most irAEs are grade 1 or 2 and can be managed using protocol-specific treatment guidelines; however, early diagnosis, vigilant follow-up, and appropriate management, including the use of systemic, high-dose corticosteroids, are essential to minimize life-threatening complications of these side-effects. It is important for oncologists to become familiar with ipilimumab irAEs in both the type and the timing of their development so that published algorithms can be applied effectively. Herein, we present a detailed safety analysis of data from the phase 3 MDX010-20 study, specifically regarding patterns of development and resolution of irAEs.
MATERIALS AND METHODS
Full details of the study design, inclusion and exclusion criteria, and treatment have been published previously (clinicaltrials.gov identifier NCT00094653).1 Briefly, in total, 676 patients with unresectable stage III or IV melanoma who had received previous treatment with 1 or more regimens of dacarbazine, temozolomide, fotemustine, carboplatin, or interleukin-2 were randomized 3:1:1 to receive ipilimumab + gp100 peptide vaccine (n = 403), ipilimumab alone (n = 137), or gp100 alone (n = 136). Ipilimumab at 3 mg/kg was administered every 3 weeks for up to 4 treatments (induction). Patients could be retreated upon disease progression with their original treatment if they had stable disease of ≥6 months' duration (from baseline) or if they had a confirmed objective response, no grade 3 irAEs that precluded further ipilimumab dosing, and no grade 4 toxicity.
AEs were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0. Characterization of AEs as immune-related required the investigator to consider the event treatment-related and for it to be included in a preassembled list of AEs with known inflammatory etiology. Protocol-defined management of irAEs included the administration of corticosteroids (orally or intravenously), a delay in a scheduled dose, or discontinuation of therapy. Doses were delayed in the case of nondermatologic grade ≥2 irAEs until the event improved to grade 1 or lower; if the event did not improve to grade 1 or lower, then treatment was discontinued permanently. Patients discontinued treatment if they experienced any nonskin-related grade ≥3 AEs considered to be related to ipilimumab treatment (with the exception of potentially reversible grade <4 inflammation or endocrinopathies in which clinical symptoms could be controlled with appropriate hormone-replacement therapy) or if they had any grade ≥4 AEs.1, 5
Analyses were performed to evaluate the time to onset and resolution of irAEs reported during the induction phase (from the day of the first dose to the day before retreatment or 70 days after the last induction dose, whichever came first) or for the entire study duration (from the day of the first dose to 70 days after the last dose, regardless of treatment phase). Here, we primarily focus on those patients who received the subsequently approved induction regimen of ipilimumab 3 mg/kg every 3 weeks for a total of 4 doses.
Incidence of Immune-Related Adverse Events
Of 643 patients who received at least 1 dose of a study drug, 105 patients who received ipilimumab monotherapy (80%) had a treatment-related AE, and 80 of those patients (61%) reported an irAE of any grade (Table 1), as previously reported by Hodi et al.1 The respective results for patients who received ipilimumab plus gp100 were 338 patients (89%) with treatment-related AEs and 221 patients (58%) with irAEs. Most irAEs were grade 1 or 2, although some life-threatening irAEs were observed, and approximately 10% to 15% of ipilimumab-treated patients reported grade 3/4 events. Among all 643 patients, it appeared that 14 deaths were related to the study drugs (2.2%), of which half were associated with irAEs, comprising bowel perforation (n = 1) and liver failure (n = 1) for ipilimumab monotherapy; and colitis and septicemia (n = 1); bowel perforation/inflammatory colitis, bowel perforation, or multiorgan failure/peritonitis (n = 3); and Guillain-Barre syndrome (n = 1) for combined ipilimumab plus gp100. Details of these 7 patients are provided below.
Table 1. Summary of Treatment-Related Adverse Events and Immune-Related Adverse Events in the Safety Population Throughout the Duration of a Phase 3 Study
Deaths associated with irAEs included deaths related to ipilimumab monotherapy (colic bowel perforation [n = 1], liver failure [n = 1]) and deaths related to ipilimumab plus gp100 (grade 3 colitis and septicemia [n = 1]; bowel perforation/inflammatory colitis, bowel perforation, or multiorgan failure/peritonitis [n = 3]; Guillain-Barre syndrome [n = 1]).
Deaths that were not associated with irAEs included deaths in the ipilimumab monotherapy group (severe infection/renal failure/septic shock [n = 1], vascular leak syndrome [n = 1]), in the gp100 group (cachexia [n = 1], septic shock [n = 1]), and in the ipilimumab plus gp100 group (sepsis [n = 1], myelofibrosis [n = 1], acute respiratory distress syndrome [n = 1]).
A woman aged 77 years had grade 3 diarrhea, colitis, fever, and neutrophilia before the planned fourth induction dose of ipilimumab monotherapy, which was postponed. After the failure of broad spectrum antibiotics to resolve the symptoms, intravenous corticosteroids were initiated, leading to a rapid improvement in symptoms. The patient was discharged 3 days later on a nontapering oral daily dose of 65 mg prednisone. Ipilimumab was discontinued because of disease progression, and the patient started treatment with temozolomide plus continued oral corticosteroid therapy. Approximately 6 weeks later, the patient was readmitted to hospital with abdominal pain, dehydration, and acute sigmoid diverticulitis, which resolved after supportive care, antibiotics, and a decrease in steroid dose. The patient started treatment with dacarbazine but was subsequently readmitted with a perforated bowel. This was 136 days after the last ipilimumab dose and 2 months after the last documented use of steroids, with no complaints of diarrhea or constipation in the meantime. The patient was given palliative support and died 3 days later.
A man aged 62 years who had widely metastatic disease to the spleen, liver, stomach, lymph nodes, and soft tissue was admitted to hospital 24 days after receiving his first dose of ipilimumab monotherapy with worsening liver and renal function, pain, and dehydration. He was treated for sepsis associated with multiorgan failure and possible tumor lysis syndrome. Despite broad spectrum antibiotics, intubation, and intensive support, he died the next day.
Shortly after his third dose of ipilimumab plus gp100, a man aged 70 years developed diarrhea that was unresponsive to atropine plus diphenoxylate and oral budesonide. Three weeks later, acute colitis was confirmed by biopsy. The patient was admitted to hospital with ongoing diarrhea and dehydration, and intravenous corticosteroids were initiated. After being discharged after 5 days with oral prednisone and budesonide, the patient was readmitted 2 days later with recurrent bloody diarrhea and findings suggestive of a lacunar stroke. Intravenous corticosteroids were reinitiated; and, after receiving treatment for steroid-induced complications and testosterone deficiency, the patient was discharged. Less than 2 weeks later, the patient had grade 3 diarrhea and was subsequently readmitted with fever and lethargy because of Escherichia coli septicemia. He was treated with antibiotics and supportive care and died approximately 1 week later.
A woman aged 58 years experienced grade 1 diarrhea and vomiting after receiving 4 doses of ipilimumab plus gp100. The diarrhea was treated with atropine and diphenoxylate; however, the patient presented to an emergency room approximately 3 weeks later with rectal bleeding. Although intravenous corticosteroids and appropriate fluid and blood resuscitation were provided, the patient did not inform the attending staff of her involvement in the study; therefore, protocol-specific irAE management was not initiated. Five days after hospitalization, acute colitis and pneumoperitoneum were confirmed by colonoscopy and biopsy. The patient underwent total colectomy and ileostomy for pancolitis, perforation, and acute peritonitis; and an ulcerated melanoma lesion invading the retroperitoneal muscles was resected. However, the patient further deteriorated and died 3 days later.
A man aged 59 years who already was hospitalized and was being treated for pain and constipation with dexamethasone and gabapentin became acutely ill with fever, vomiting, and acute renal failure shortly after receiving his second dose of ipilimumab plus gp100. The patient remained hypotensive despite fluid replacement and had free air identified under the diaphragm on chest x-ray. He died the following day as a result of bowel perforation.
A man aged 69 years was receiving symptomatic treatment for diarrhea, having completed his induction therapy with ipilimumab plus gp100. Afterward, he presented with dysphagia and 5 or 6 bloody movements per day. Although immediate hospitalization was refused, the patient was admitted to hospital the next day, and rehydration and intravenous dexamethasone were initiated. An emergency laparotomy identified purulent, malodorous ascites and peritoneal carcinomatosis with involvement of the sigmoid colon and ileum associated with ischemic colon and a sigmoid perforation. Sigmoid pathology revealed ulcerating colitis and perforation and infiltrates of an epitheloid cell neoplasm on the serous membrane. The patient died the following day of multiorgan failure because of peritonitis.
After completing his induction treatment with ipilimumab plus gp100, a man aged 62 years developed progressive neurologic symptoms. After hospitalization, he developed dyspnea and, on electromyography, demonstrated distal and proximal axonal disturbances consistent with Guillain-Barre syndrome. High-dose intravenous immunoglobulin was initiated; but, after a transient improvement, dyspnea worsened, and the patient subsequently died from respiratory failure.
Types of Immune-Related Adverse Events
Overall, among all ipilimumab-treated patients, irAEs most frequently affected the skin and GI tract. Skin-related irAEs of any grade were reported in 209 of 511 patients (41%) and most commonly included pruritus, rash, and erythema. GI disorders were reported by 160 of 511 patients (31%) and most commonly included diarrhea and colitis. Endocrine disorders were the third most commonly reported irAE, with 25 of 511 ipilimumab-treated patients (5%) having an irAE of any grade. Of these, 5 patients who received ipilimumab monotherapy had 1 or more grade 3/4 event, the details of which are provided in Table 2.
Table 2. Symptoms and Management of Severe (Grade ≥3) Endocrine Immune-Related Adverse Events Related to Ipilimumab Monotherapy in the Phase 3 Trial MDX010-20
Oral hydrocortisone 10 mg twice daily and levothyroxine 0.75 mg daily; decrease in corticotrophin levels improved to grade 2 within 2 d; ongoing at grade 2 with maintenance hydrocortisone therapy
Time to Onset and Resolution of Immune-Related Adverse Events
The vast majority of irAEs, including grade 3/4 events, developed during the induction period, ie, within 12 weeks of initial dosing, with a median time to resolution of 6 to 8 weeks (Table 3). Among 94 patients who remained alive 2 years after the start of on-study treatment (54, 24, and 16 patients in the ipilimumab plus gp100, ipilimumab monotherapy, and gp100 monotherapy groups, respectively) 5 patients (9.3%), 1 patient (4.2%), and 1 patient (6.3%), respectively, reported an irAE >70 days after their last dose of study drug. All were grade 1 or 2 in severity (vitiligo, hypothyroidism, diarrhea, proctitis, low blood testosterone, and hypogonadism), with the exception of 1 patient in the ipilimumab plus gp100 group who experienced grade 3 colitis on day 147 that recovered with residual effects.
Table 3. Patterns of Development and Resolution of Immune-Related Adverse Events From Ipilimumab Monotherapy
Induction phase: From the day of first dose to the day before reinduction or 70 days after the last induction dose, whichever came first
Rate, no. (%)
Time to onset: Median [95% CI], wk
Time to resolution of irAEs with onset during induction phase
Grade 2-4 resolution, no./total no. (%)
Median [95% CI], wk
Grade 3/4 resolution, no./total no. (%)
Median [95% CI], wk
Entire study duration: From the day of first dose to 70 day after the last dose, regardless of treatment phase
Rate, no. (%)
Time to onset: Median [95% CI], wk
Time to resolution
Grade 2-4 resolution, no./total no. (%)
Median [95% CI], wk
Grade 3/4 resolution, no./total no. (%)
Median [95% CI], wk
The frequency of irAEs observed among 40 retreated patients in the MDX010-20 trial was similar to that observed during their initial treatment. Overall, irAEs were reported by 15 of 29 patients (51.7%) in the ipilimumab plus gp100 arm, 7 of 9 patients (77.8%) in the ipilimumab monotherapy arm, and 1 of 2 patients (50%) in the gp100 arm.
Management of Immune-Related Adverse Events
Most irAEs, including grade 3/4 events, were managed successfully according to protocol treatment guidelines, including careful monitoring and the early receipt of corticosteroids. Nearly all treated patients (625 of 643; 97.2%) received at least 1 concomitant medication during the study (Table 4). Overall, analgesics were reported as concomitant medication for ≥50% of patients in each treatment group. Among the commonly reported concomitant medications, antiemetics and antacids were received by a similar proportion of patients across treatment groups, whereas antibacterials, antihistamines, corticosteroids, and antidiarrheals were received by more patients in the ipilimumab-treated groups than patients in the gp100-treated groups. The overall rate of receipt of systemic corticosteroids in the 2 ipilimumab treatment groups was approximately 10% greater than in the gp100 group. Likewise, antihistamines were frequently received for the management of skin symptoms, such as rash and pruritis.
Table 4. Concomitant Medications Used by at Least 10% of Treated Patients on the MDX010-20 Trial
A post hoc clinical review of patients with grade ≥3 GI or liver irAEs was performed, comprising 35 patients (24 who received ipilimumab plus gp100, 10 who received ipilimumab monotherapy, and 1 who received gp100) who had at least 1 GI irAE and 8 patients (4 who received ipilimumab plus gp100, 1 who received ipilimumab monotherapy, and 3 who received gp100) who had at least 1 hepatic event. Among those who had a GI event, 75% received steroid therapy within 10 days of event onset, 15% received steroid treatment more than 10 days after event onset, and approximately 10% (4 patients) received no steroid treatment. Only 5 patients required the addition of infliximab, an antibody against tumor necrosis factor-α, to manage the GI irAE, and all 5 demonstrated some improvement upon intervention. The rate of steroid treatment was lower for hepatic events than for GI events, and only 3 of 8 patients received steroid treatment (all within 10 days). However, the presence of liver metastases at baseline in 4 of these patients may have complicated decision-making regarding the use of steroids for hepatic events.
Immune-related adverse events are mechanism-based events that occur with ipilimuma; they most commonly affect the skin and GI tract, and, less commonly, they affect the liver, endocrine system, and nervous system. Most irAEs in MDX010-20 were low grade, although some were life-threatening. Of 14 treatment-related deaths, 7 were considered attributable to immune-based mechanisms and were characterized as irAEs.1 With the exception of case 2, which was difficult to clearly attribute to ipilimumab, and the patient with Guillain-Barre syndrome, all patients who died after receiving ipilimumab had severe GI events; most commonly colitis.
There is no evidence that bowel involvement in patients with melanoma increases the likelihood that colitis, especially dose-limiting colitis, will occur after receiving ipilimumab. Major issues that predispose patients to a worse outcome when colitis occurs, and must be considered by the treating oncologist, include prolonged time to reporting of the symptoms, exacerbation of grade 1 or 2 diarrhea after each dose, not holding a scheduled dose when grade 2 diarrhea is still present, and noncompliance with antidiarrheal instruction and regimens. The key to handling ipilimumab-associated colitis is rapid and timely intervention mediated by a high level of communication between staff and patients.
In at least 2 of the deaths resulting from colitis, there was a delay of 3 weeks from the onset of symptoms to the initiation of steroid treatment. In a retrospective analysis of 836 patients who received ipilimumab at doses of 3 mg/kg or 10 mg/kg within clinical trials, the initiation of steroid treatment within 5 days of the onset of enterocolitis symptoms led to faster resolution than when steroid treatment was delayed for more than 5 days.6 These findings highlight the importance of 1) monitoring patients for GI signs and symptoms from the moment ipilimumab treatment is initiated, 2) careful and continuous monitoring for life-threatening side effects beyond the initial treatment period, and 3) involving a multidisciplinary team of specialists, including gastroenterologists, to consult and advise on appropriate patient management.
Immune-related adverse events that affect the endocrine system also are worthy of discussion, because they pose significant difficulties in terms of management. Hypophysitis is the second most common high-grade or dose-limiting irAE. It can result in profound symptomatology and is easily reversible with proper treatment. However, endocrinopathies often masquerade as nonspecific conditions ascribed to other causes. In these patients, appropriate medical intervention may be delayed or misguided. Therefore, the inclusion of an endocrinologist in the multidisciplinary management of patients who receive ipilimumab is warranted.
Most irAEs related to ipilimumab were manageable using protocol-specific treatment guidelines, which rely on prompt intervention and the use of oral and/or parenteral corticosteroids. Steroids do not seem to interfere with ipilimumab efficacy once an irAE has occurred7-9; however, the reasons for this are not well understood. Steroids are potent immunosuppressive drugs that are generally withheld from cancer patients who are receiving immunotherapy because of concerns that they may inhibit immune-mediated tumor regression.10, 11 When a glucocorticoid known to deplete lymphocytes was administered to mice receiving immunotherapy, however, it had no effect on T-cell-mediated tumor regression.10 Steroids have diverse effects involving different components of the immune system; including some that are immune-enhancing, such as the inhibition of immune-activated cell death.10 Therefore, although steroid treatment could abrogate or down-modulate an immune response to ipilimumab, it is possible the response will not be entirely inhibited.10, 12
Onset of the majority of irAEs occurred during the 12-week induction phase of treatment, consistent with the results from other ipilimumab studies (Tables 3, 5; Fig. 2). Previous work suggests that irAEs may have a characteristic pattern of onset, which is consistent among patients who received ipilimumab 3 mg/kg and ipilimumab 10 mg/kg (Table 5, Fig. 2).13, 14 In patients who received ipilimumab 10 mg/kg, irAEs affecting the skin typically were observed after an average of 3 to 4 weeks, effects involving the GI tract and liver were observed after 6 or 7 weeks, and endocrine events were observed after 9 weeks.14, 15 Skin and endocrine disorders appear to be the most persistent, whereas GI symptoms are the quickest to resolve despite also being the most likely to result in complications if uncontrolled (through bowel perforation and peritonitis or obstruction). The late onset of some irAEs compared with others is currently unexplained. A recent study with ipilimumab indicated an association between clinical response, a dermatologic irAE, and clonal expansion of Melan-A–specific T cells.16 Because Melan-A is an antigen shared by melanoma cells and normal melanocytes, and because the naive T-cell repertoire in humans contains Melan-A–specific clonal precursors, this may explain why skin toxicity arises before other irAEs and supports the hypothesis that ipilimumab breaks peripheral tolerance to both tumor and self-antigens.
Table 5. Time to Onset of Grade 2 to 5 Immune-Related Adverse Events and Resolution of Grade 2 to 4 Immune-Related Adverse Events in Patients who Received Ipilimumab 3 mg/kg
Type of Immune-Related Adverse Event
Median Time to Onset, wk
Median Time From Onset to Resolution, wk
Abbreviations: NR, not reported.
The etiology for GI toxicity is less clear. It has been demonstrated that ipilimumab produces a unique pattern of dysregulation of GI mucosal immunity distinct from that reported in Crohn disease or ulcerative colitis, although the precise mechanisms for this are unknown.17 Given the well documented involvement of regulatory T cells in gut immune homeostasis and the role of CTLA-4 in regulatory T-cell function in the GI microenvironment,18 it is possible that, by blocking CTLA-4, ipilimumab also inhibits the tolerogenic effects of regulatory T cells in the GI tract, resulting in immune-mediated inflammation. This may explain the longer time to onset of GI irAEs compared with skin toxicity as well as the more rapid resolution.
Less frequent events (for example, endocrine irAEs) tend to take longer to develop and then persist. In an adjuvant trial of 75 patients with melanoma who received with ipilimumab, 11 patients had grade 2 or 3 hypophysitis that required replacement corticosteroids. Six patients required treatment for at least 2 years, including 1 patient who remained symptom-free with 5 years of corticosteroid therapy. It is noteworthy that 3 of those patients were weaned from systemic steroids, indicating that a requirement for steroid therapy resulting from ipilimumab-related endocrinopathies need not be permanent.19
Although most patients develop only mild-to-moderate irAEs, careful follow-up and early intervention are advised to address any symptoms and to prevent potentially serious complications. When historic safety data from early clinical studies of ipilimumab were compared with data from later studies, for example, the incidence of bowel perforation decreased from 0.9% to 0.5%, and the decline was attributed to the implementation and consistent use of treatment guidelines for irAE management and to the increasing experience of practitioners.15, 20
Although the data from this analysis fit with observations from phase 2 and other studies, they require confirmation from larger analyses and/or clinical observation during more widespread use of ipilimumab. Physicians, practitioners, and oncology nurses must exercise vigilance to ensure that patients who receive ipilimumab are well informed and educated so they can report symptoms promptly; this is essential to prevent potentially life-threatening complications from irAEs.
The MDX010-20 trial (NCT00094653) was sponsored by Bristol-Myers Squibb. Editorial and writing assistance was provided by StemScientific, which was funded by Bristol-Myers Squibb.
CONFLICT OF INTEREST DISCLOSURES
J.S.W. has received honoraria from Bristol-Myers Squibb. R.D. has received research funding from and has served as an advisor for AstraZeneca, Novartis, Cephalon, Merck, Transgene, Bristol-Myers Squibb, Roche, GlaxoSmithKline, and Bayer, and he also has served as an advisor for Genta and Spirig Pharma. V.d.P. is an employee of Bristol-Myers Squibb. C.L. has received honoraria from Bristol-Myers Squibb. F.S.H. has served as a nonpaid consultant and has received clinical research support from Bristol-Myers Squibb.