Issue: December 2018
December 14, 2018
14 min read

Checkpoint Checkmate: Adverse Event Management for Checkpoint Inhibitors

Issue: December 2018
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Although it might seem silly to use the game Whack-A-Mole to illustrate a point about health care — knock one down and another pops up — this seems to be a fitting comparison when dealing with immune-related adverse events associated with the immune checkpoint inhibition therapies that are proliferating in cancer treatment.

While often touted as the next “game-changer” in cancer therapy, checkpoint inhibitors have also given rise to a whole new array of complications for the patient that appear to be rheumatic in origin, a development which has brought rheumatologists into this evolving field. Though, at present, these events appear to be relatively infrequent, their infrequency also makes them difficult to spot, particularly for oncologists not familiar with the particulars of rheumatology. Consequently, little is known about how severe these complications may be, and how they need to be managed. With seven checkpoint inhibitors currently approved for a number of malignancies, and more poised to gain approval in the coming months, the field of rheumatology could be faced with a substantial new patient population, and many more “moles” to whack.

Current FDA-approved agents include anti-CTLA-4, anti-PD-1 and anti-PD-L1 drugs, which may be used in many solid tumor malignancies including melanoma, Hodgkin lymphoma, renal cell carcinoma, non-small-cell lung cancer, urothelial carcinoma and Merkel cell carcinoma, according to Cassandra Calabrese, DO, from the department of rheumatologic and immunologic disease at the Cleveland Clinic. Approvals are likely forthcoming for anti-OX40 and anti-TIM-3 monoclonal antibodies, as well as inhibitors of other immunologic checkpoints.

Although the introduction of immune checkpoint inhibitors has been a significant boon to cancer therapy, the associated complications of immune-related adverse events highlight the need for ongoing collaboration between oncology and rheumatology specialties, Anne R. Bass, MD, and other experts noted.
Although the introduction of immune checkpoint inhibitors has been a significant boon to cancer therapy, the associated complications of immune-related adverse events highlight the need for ongoing collaboration between oncology and rheumatology specialties, Anne R. Bass, MD, and other experts noted.

Source: Hospital for Special Surgery

“As one would imagine, with their increased use, more patients are going to develop immune-related adverse events (irAEs),” she told Healio Rheumatology. “A rheumatologist must be able to recognize rheumatic irAEs as a complication of checkpoint inhibitors and be aware that many clinical syndromes have been described, including inflammatory arthritis, polymyalgia rheumatica, sicca syndrome, myositis and others. We must do our best to keep up with this quickly growing field, as it has already presented many challenges.”

For Anne R. Bass, MD, from Hospital for Special Surgery and Weill Cornell Medicine, identifying and diagnosing these patients, particularly those with arthritis, is the most urgent clinical need. “We are just beginning to define these arthritides, because they are so heterogeneous,” she said. “If you see five different patients, you might see five different phenotypes.”

Underlying genetics and environmental exposures may also contribute to the heterogeneity in this population, according to Bass. “In our patients, around 20% or 25% present with what looks like rheumatoid arthritis, with both large joints and small joints affected,” she said, noting that only some of them are seropositive. “Another 20% have polymyalgia rheumatica, which causes achy shoulders and hips. Then there is a group who have a large joint arthritis, usually the knees, hips, wrists and/or shoulders. We don’t see this pattern of arthritis outside of the immunotherapy context.”

Beyond these manifestations, Bass said that there are reports of psoriatic arthritis, scleroderma and vasculitis, but they are rare. Myositis is relatively common whereas dermatomyositis and lupus — diseases that are driven by interferon — are very rare. Some patients may have a disease susceptibility that is unmasked by immunotherapy, whereas others may have an immunotherapy-specific condition. “There are just a lot of uncertainties right now.”

Professional Guidelines

The good news is that the oncology community has made strides to begin understanding these complications. As reported earlier this year in the Journal of Clinical Oncology, the American Society of Clinical Oncology has released a clinical practice guideline on management of irAEs in patients receiving checkpoint inhibitor therapy. Based on a systematic review of publications on immune checkpoint inhibitor therapy from 2000 through 2017, a multidisciplinary, multiorganizational panel of experts provided detailed recommendations for organ-specific management for skin, gastrointestinal, lung, endocrine, musculoskeletal, renal, nervous system, hematologic, cardiovascular and ocular irAEs.

Clinicians treating patients with immune checkpoint inhibitors should consider suspending these treatments in favor of corticosteroids at the sign of grade 2 toxicity. Prednisone 1 to 2 mg/kg per day or methylprednisolone 1 to 2 mg/kg per day may be used when grade 3 toxicities occur. A 4- to 6-week tapering period is recommended for patients who are transitioned to corticosteroids. It is recommended that with the exception of manageable endocrinopathies, grade 4 toxicities should result in permanent discontinuation of checkpoint inhibitor therapy.


The important “take-home message” of the American Society of Clinical Oncology practice guidelines, for Maria E. Suarez-Almazor, MD, PhD, of the department of general internal medicine at The University of Texas MD Anderson Cancer Center, is awareness that these events may occur, and that they need to be addressed promptly. She added that it is time for other organizations to follow suit. “The American College of Rheumatology has been offering educational opportunities at the annual meeting to make rheumatologists aware that these guidelines, and others from oncology groups, exist,” she said. “However, not all doctors can go to the meeting, and although there are online resources, most of these are in oncology forums.”

Uma Thanarajasingam, MD, PhD, of the department of rheumatology at the Mayo Clinic, underscored the practical nature of raised awareness of the guidelines. “Community rheumatologists need to understand that there is a place they can turn for information about these events, whether it is the guidelines or other experts who are familiar with them,” she said.

This can then lead to basic interventions early in the course of irAEs, according to Bass. “What I have learned from caring for patients with immunotherapy-induced arthritis is that it often becomes chronic, so it is important to start early with steroid-sparing agents,”she said.

By the Numbers

In what many experts are considering the first key data set in the field, Richter and colleagues retrospectively reviewed all patients from the database at the Mayo Clinic who underwent checkpoint inhibitor therapy over a recent 7-year period. Results published in Arthritis & Rheumatology demonstrated that among 1,293 eligible patients, 43 were clinically diagnosed with rheumatic irAEs. Inflammatory arthritis was reported in 2% of the cohort (n = 34), with myopathy reported in 10 patients, while 17 fell into the category of “other rheumatic syndromes,” according to the findings.

“We still lack strong epidemiologic data, but rheumatologic irAEs like inflammatory arthritis can be seen in up to 7% of patients treated with certain checkpoint inhibitors,” Laura C. Cappelli, MD, MHS, from the division of rheumatology at Johns Hopkins School of Medicine, told Healio Rheumatology. “Other rheumatologic irAEs like myositis occur in less than 1% of patients treated with checkpoint inhibitors.”

Anne R. Bass, MD
Anne R. Bass

Other findings from Richter and colleagues showed that inflammatory arthritis was polyarticular in most cases, and that 76% of these patients required glucocorticoids.

“We are really just getting a handle on the epidemiology,” Bass said. “The problem is that we see these cases, but don’t always know the denominator.”

Laura C. Cappelli, MD, MHS
Laura C. Cappelli

Bass noted that a systematic review of clinical trials by Cappelli and colleagues in Arthritis Care & Research, reported that arthritis may occur in 1% to 7% of patients treated with checkpoint inhibitors, while arthralgia rates may range from 7% to 43%. “My guesstimate, based on the available data we have, is that arthritis occurs in about 5% of patients treated with checkpoint inhibitors, and around 20% experience joint pain,” she said. “However, many of these cases are missed by oncologists for two reasons: one, because they don’t pay as much attention to joint pain as a rheumatologist would, and two, because steroids used for other adverse events masks joint pain.”

A key concern is that oncologists do not typically recognize the difference between arthritis, arthralgia and polymyalgia syndromes, according to Suarez-Almazor. “Reports vary greatly across trials, possibly because of differences in musculoskeletal adverse events characterizations,” she said.

In a presentation at the 2018 Congress of Clinical Rheumatology, Ami A. Shah, MD, MHS, of the Johns Hopkins University School of Medicine, suggested that inflammatory arthritis may develop in approximately 5% of patients treated with anti-PD1 therapy. She added that sicca, polymyalgia rheumatica, myositis, single-organ vasculitis, lupus nephritis, psoriasis and psoriatic arthritis have also occurred, along with reports of scleroderma.


“When we talk about these rheumatic syndromes, there are three primary presentations: arthritis/arthralgia, polymyalgia and myositis,” Suarez-Almazor said. “Myositis is obviously the scariest, because it is acute and can be life-threatening. Admittedly, myositis probably occurs in less than 1% of patients, but we are very concerned about it because patients can also develop myocarditis, a potentially fatal complication.”

“What is felt to be weakness from myositis may be the result of progressive intracranial metastasis, for example,” Calabrese added.

Other issues Suarez-Almazor encouraged rheumatologists to consider is whether the immunotherapy is given as adjuvant or a primary treatment, patient’s response to oncologic therapy and life expectancy, and whether there are other alternatives for oncologic treatment. “These drugs have not been widely used in patients with limited cancer yet, but that could change,” she said. “Also, new drugs and new combinations will likely come on board, so that will also change the landscape.”

Results of 1,293 patients who underwent checkpoint inhibitor therapy over a 7-year period.

Making the Diagnosis

Calabrese outlined the varied challenges in making a diagnosis. “This is a very heterogeneous patient population, with many patients having received chemotherapy, radiation and/or surgery prior to checkpoint inhibitor therapy, and they may have progression of their tumor at the time of evaluation,” she said. “It is important to keep other etiologies on the differential when evaluating a patient for rheumatic irAEs.”

Cassandra Calabrese, DO
Cassandra Calabrese

It is also important to note that the rheumatic irAEs described thus far frequently have atypical features compared to de novo disease, according to Calabrese. “For example, they may lack characteristic autoantibodies or require higher doses of prednisone than would be expected,” she said. “Also, unlike all other irAEs to date, rheumatic manifestations appear to be the only irAE with inflammation that persists, even despite discontinuation of the checkpoint inhibitor.”

This persistence of rheumatologic irAEs, even several years after checkpoint inhibitor therapy has been stopped, is another prominent concern for rheumatologists, according to Cappelli. “We have patients who are 3 years out from receiving immune checkpoint inhibitors who continue to have issues with inflammatory arthritis,” she said. However, she believes that diagnosis is improving. “It is still a problem for oncology and other nonrheumatology providers. There is a knowledge gap in the range of rheumatologic irAEs like polymyalgia rheumatica, myositis, vasculitis and inflammatory arthritis.”

Cappelli added that the musculoskeletal examination is not taught in depth in all medical training programs, making it difficult for providers to determine if joints are swollen, warm or have a decreased range of motion when evaluating for inflammatory arthritis. “In the area of myositis, sometimes routine labs will show an elevated aspartate aminotransferase, which can come from muscle in the setting of myositis, but clinicians may go down the path of investigating liver pathology,” she said.

For Suarez-Almazor, referral to a rheumatologist at the first sign of any such syndrome is critical. “Not only will a rheumatologist be able to make a diagnosis, but they will be familiar with the medications necessary to treat these patients,” she said. “We have seen that these syndromes are different from other immune-related arthritis in that they sometimes can have an insidious onset and can linger for weeks or months, even after discontinuing immunotherapy.”

Treatment Complications

As with other aspects of these syndromes, many clinicians remain mostly in the dark about treatment complications. In their study in the Journal for ImmunoTherapy of Cancer, Moseley and colleagues conducted a retrospective case series for patients referred to endocrinology or rheumatology for bone disorders after treatment with therapies targeting PD-1, CTLA-4, or both. Though there were only six events overall, they suggested that it may portend skeletal adverse events of immune checkpoint therapy.


“The main challenge in managing patients with rheumatologic irAEs is the lack of evidence-based medicine to guide treatment,” Cappelli said. “IrAEs are a relatively new clinical issue and more research is needed to determine the most effective and safe treatments. The concern in treating irAEs is that immunosuppression may undermine the tumor immunologic response and have a negative impact on cancer progression.”

That said, Johnson and colleagues examined a cohort of 75 patients with immune-related enterocolitis who had undergone treatment with checkpoint inhibition. They hypothesized that the addition of infliximab (Remicade, Janssen) to front-line corticosteroid therapy may improve these events. Results published in the Journal for ImmunoTherapy of Cancer demonstrated that diarrhea resolved in 3 days among patients in the infliximab arm and 9 days among those treated only with corticosteroids (P < .001). Time to steroid titration was also shorter in patients in the infliximab arm, 4 vs. 13 days (P < .001).

Thanarajasingam offered a general point on treating these patients. “Rheumatologists understand better than oncologists how aggressive we can be,” she said. However, she echoed Cappelli’s point that evidence-based strategies are lacking, and that communication with other physicians is critical.

Additional findings from Richter and colleagues showed that the mean treatment duration was 18 weeks (SD = 18 weeks). Disease-modifying drugs were reported in 15% of their cohort, while 9% required discontinuation of immune checkpoint therapy. Nine patients required permanent discontinuation of immune checkpoint therapy after treatment of myopathy with glucocorticoids. They also found cases of connective tissue diseases, vasculitis, PMR-like syndrome and flared pre-existing rheumatic disease. Clinicians used immunosuppression in 71% of these cases. However, 12% required discontinuation of checkpoint inhibition.

Uma Thanarajasingam, MD, PhD
Uma Thanarajasingam

“Any time we go beyond prednisone, we have to partner with oncologists to see how much damage we might be doing to the immune system,” Thanarajasingam said. “We have to understand the status of the underlying cancer, whether the patient is responding really well to checkpoint inhibitor therapy even if they are having a great deal of arthritic pain. Sometimes we have to take a hiatus from the treatment we are using, or switch courses.”

Bass suggested that oncologists who deal with melanoma are more familiar with immune-related adverse events because they have been using immunotherapy the longest. “For my part, I try to understand the nature of the malignancy and its prognosis, and how immunosuppressive therapy may impact the patient,” she said. For example, she noted that Merkel cell carcinoma is immunologically-sensitive and occurs primarily in immune-suppressed individuals; treating immune-related adverse events in such a patient may be much less safe than in someone with melanoma. “You have to understand that the treatment approach for every patient and every tumor is different. I would love to be able to offer concrete recommendations that always apply. The key is to work together with the oncologist and the patient to discuss risks and benefits, and to learn from each other.”

A small but growing area of research deals with specific drug-drug interactions, according to Shah’s presentation. She suggested that currently available information indicates that corticosteroids and short-term TNF inhibitor use may be tolerable without affecting melanoma responses to both ipilimumab (Yervoy, Bristol-Myers Squibb) and nivolumab (Opdivo, Bristol-Myers Squibb).


“These interactions are still quite an unknown area,” Thanarajasingam said. “We get a sense that certain combinations may have more or less toxicity than others, and there are data emerging for ipilimumab and nivolumab, for example.”

When more malignancies are added to the mix, the difficulties are likely to be compounded, according to Thanarajasingam. “Obviously, because these drugs were first approved in melanoma, the toxicities reported in that type of cancer are greatest,” she said. “As usage expands, it will be interesting to see how the toxicity profile changes. For example, lymphoma is a completely different type of cancer, and the immune cells are part of the cancer, so what are the implications for adverse events? It’s still such a new area, so it’s hard to say.”

Working Alongside Oncologists

Cappelli stressed basics when it comes to working with oncologists. “Communication is the most important thing to successfully care for patients with rheumatologic irAEs,” she said. “The oncologist and rheumatologist must reach consensus about ways to manage each patient.”

Even logistics are not to be taken lightly, according to Cappelli. “We contact the patient’s oncologist on the day of their clinic visit to us so we can confirm the plan,” she said. “It is also important for the patient to be aware of the role each physician plays in their care. Often patients ask us about plans for their cancer therapy and we need to remind them that their oncologist will guide their cancer treatment and we will manage their side effects, in collaboration with their oncologist.”

Calabrese stressed basics. “If there are symptoms of grade 2 or higher based on CTCAE grading, they should be referred to a rheumatologist,” she said. “Long-term management of these patients requires multidisciplinary care that addresses the patients’ rheumatic manifestations while at the same time treating their cancer.”

Maria E. Suarez-Almazor, MD, PhD
Maria Suarez-Almazor

Management may need to involve other specialties, according to Suarez-Almazor. “We need to promptly recognize when others need to be involved, for instance cardiologists when suspecting myocarditis,” she said. “The benefits and harms of therapies need to be understood quickly, along with how the various therapies are interacting with each other.”

Shifting the Field

There is no doubt that this patient population will change the field of rheumatology, according to Bass. The question is how much. “We have to take every opportunity to learn, because we are going to see more and more of these patients,” she said. “It is critical that we have people who understand and can manage these patients and the problems they confront.”

Bass believes that this also provides an opportunity to learn. “In most cases of RA, the development of antibodies to disease onset can usually be 3 to 5 years,” she said. “In immunotherapy patients, we are seeing arthritis develop within 3 to 4 months of starting treatment. This means we can study this disease and biomarkers over a shorter period of time, and they will likely teach us something about rheumatic diseases that we see in other contexts.”


Learning how to predict which patients receiving checkpoint inhibitors are prone to rheumatic complications will also allow researchers to study these events prospectively, Bass noted. “This may be a first step toward learning how to prevent arthritis, and could also be applied to the preclinical phases of RA.”

For Suarez-Almazor, having mechanisms in place to identify and manage these fast-developing patients is critical. “Many of these are patients who cannot wait to be seen by a rheumatologist for weeks or months, as decisions about their cancer treatment are critical,” she said.

“At this point, there are so many more questions than answers,” Calabrese said. “We have much to learn about risk factors, biomarkers, pathogenesis and optimal treatment of rheumatic irAEs, as well as the impact that treatment for irAEs may have on cancer outcomes. I am most interested in identifying risk factors for rheumatic irAEs and will be looking at autoimmune serology and family history prior to checkpoint inhibitor initiation and trying to see if we can identify serological biomarkers or other signs that a patient may be predisposed to developing a rheumatic irAE.”

Thanarajasingam is optimistic that the rheumatology community is responding appropriately. “Over the last two ACR meetings, the field went from a couple of posters at the back of the presentation hall to several talks in larger sessions,” she said. “We are reaching more people, including pharma, which is going into the community to let providers know that this is an issue.” – by Rob Volansky

Disclosures: Bass, Calabrese, Suarez-Almazor and Thanarajasingam report no relevant financial disclosures. Cappelli reports receiving research funding from Bristol-Myers Squibb, and consulting for Regeneron/Sanofi.