Checkpoint Inhibition and Cardiotoxicity: Cause for Concern?

Alexander M. Castellino, PhD

November 20, 2018

An extensive analysis of a World Health Organization (WHO) global database has identified myocarditis, pericardial diseases, cardiac arrhythmias, and vasculitis as significant immune-related cardiac toxicities associated with the use of immune checkpoint inhibitors (ICIs) in routine clinical practice.

These cardiotoxicities were seen more often in men than women. Pericardial diseases were more common in patients with lung cancer (56%), whereas myocarditis (41%) and vasculitis (60%) were more common in patients with advanced melanoma. Death resulting from myocarditis, pericardial disease, and vasculitis was reported to occur in 50%, 21%, and 6% of patients, respectively. Importantly, many of these events occurred early in the course of treatment with ICIs.

The study, which was published online November 12 in Lancet Oncology, is an extension of earlier published correspondence, which was reported by Medscape Medical News.

"To our knowledge, we report the largest and most extensive clinical characterisation of cardiovascular irAEs [immune-related adverse events] associated with ICIs so far through analysis of individualised reportable events from the WHO pharmacovigilance database," write senior author Javid J. Moslehi, MD, director of the cardio-oncology program at the Vanderbilt University, in Nashville, Tennessee, and colleagues.

"The results show a significant incidence of myocarditis, pericardial diseases, and vasculitis-related disorders associated with ICIs, suggesting that clinicians should be vigilant of these potential toxicities in patients," they add.

Noting that ICIs are a cornerstone of cancer treatment, the authors of an accompanying editorial indicate that the fact that these drugs are associated with some risk for cardiotoxicity suggests that these molecules can modulate autoimmunity. They point out that the occurrence of such toxicities need not come as a surprise, inasmuch as some of these toxicities have been seen in animal studies.

The editorial was written by Carlo G. Tocchetti, MD, and colleagues from the Federico II University, Naples, Italy.

"The authors should be applauded for this study and taking advantage of a large dataset to interrogate less frequent adverse events [compared to those seen in clinical studies]," lung cancer expert Justin F. Gainor, MD, director of targeted immunotherapy at Massachusetts General Cancer Center in Boston, told Medscape Medical News.

Implications for Clinical Practice

With this report, physicians are now better informed on cardiac-associated irAEs, but it is not likely to affect the use of ICIs in cancer patients, according to Moslehi and cancer experts not associated with the study.

"Although cardiotoxicity occurs rarely and is a scary side effect, I don't think my practice patterns have changed at all. I still treat patients who could potentially significantly benefit from immunotherapy and closely follow them," Michael A. Postow, MD, melanoma specialist at the Memorial Sloan Kettering (MSK) Cancer Center, New York City, told Medscape Medical News.

"I do not think patients with cardiovascular history or who may be perceived at greater risk of cardiotoxicity should be excluded from receiving immunotherapy, especially when no other great options exist," he added.

Gainor expressed similar sentiments. He said that the emergence of these toxicities would not affect his clinical practice. "However, when we are using ICIs, we have to think more broadly about patients presenting with new symptoms," he added.

Gainor explained that oncologists are accustomed to the side effects of chemotherapy, such as nausea and vomiting. "By contrast, treatment with ICIs may be different. ICIs can affect nearly any organ system, and the side effects of these toxicities may not be as clear cut," he said.

"Even though ICIs have revolutionized cancer care in our patients, we still cannot identify patients who will respond or who will be predisposed to these immune-related adverse events," Moslehi told Medscape Medical News. "Patients should get the best therapy first, but as ICIs get used routinely in clinical practice, clinicians should be more aware of these significant complications in our patients," he said.

"Diagnosis of myocarditis, pericardial diseases, and vasculitis are not always straightforward," Moslehi told Medscape Medical News. For example, a higher clinical acumen combined with appropriate testing (eg, MRI) for pericarditis may be necessary, he explained.

Because myocarditis can be difficult to diagnose clinically, the identification of concomitant immune-related AEs (eg, myositis) and cardiac presentations (ie, heart failure and arrhythmias) might help the clinician to diagnose ICI-associated myocarditis, Moslehi and colleagues note.

Gainor agreed. "Typical tools may not be applicable," he said. In their experience at the Massachusetts General Hospital (MGH), for example, for many patients who presented with myocarditis, echocardiograms and ejection fraction values were normal.

"As our experience with using ICIs increases, the oncologist will have to wear a clinician's hat when treating patients," Gainor commented. "Treating patients with ICIs requires multidisciplinary care involving dedicated specialists connected with oncologists," Gainor said.

Indeed, Moslehi indicated that as a cardio-oncologist, his focus is on cancer patients and cardiac problems in cancer patients. "Our data suggest that we should be more aware of these situations when we treat our patients with ICIs," he told Medscape Medical News.

Observations From Two Melanoma Experts

"We've definitely had some cases [of cardiac toxicity], including severe/fatal cases at our institution [MSK]," Postow said. "My colleagues and I have also had nonfatal cases of mild to moderate cardiotoxicity that has fortunately resolved. Almost all of the cardiotoxicity we've seen is myocarditis. I can't recall any significant pericarditis or vasculitis, as per the Lancet Oncology article, but I'm sure it exists," he added.

Postow indicated that although these cardiac toxicities are known to occur, he and his colleagues at MSK are not deterred from using ICIs, including the combination of ipilimumab (Yervoy, Bristol-Myers Squibb) and nivolumab (Opdivo, Bristol-Myers Squibb).

A melanoma expert, Alexander Menzies, MD, PhD, from the Melanoma Institute Australia, University of Sydney, who was not associated with the study, told Medscape Medical News that cardiotoxicity is very rare at his institution and is often diagnosed late, owing to the fact that the signs and symptoms are subtle until marked inflammation and cardiac dysfunction become established. "Therefore, patients are often managed late with resultant morbidity/mortality," he said.

"We even treat patients with preexisting, stable cardiac comorbidities like a remote myocardial infarction or other known history of coronary artery disease with ICIs, recognizing they may be at higher risk," Postow said. "If a patient has active cardiac disease already, I may be more inclined to treat them with single-agent anti-PD-1 [anti–programmed cell death protein–1] therapy instead of the combination of ipilimumab + nivolumab," he added.

"I do not feel that myocarditis is frequent enough to specifically dissuade me from treatment, including in those with cardiac disease (ischemic, valvular, etc)," Menzies told Medscape Medical News. "Cardiac transplantation is clearly another situation entirely, although not all transplants are the same (renal vs hepatic)," he added.

Regarding whether some patients, such as those at cardiac risk at baseline, may not be appropriate candidates for therapy, Postow said: "Certain diseases and situations vary. In BRAF wild-type metastatic melanoma, there are no really great alternatives to ICIs, so I proceed with checkpoint inhibition except in extremely rare, unusual situations (ie, a patient with a solid organ transplant on immunosuppression if we can use surgery or RT [radiotherapy] to manage their metastatic disease). Otherwise, I give immunotherapy and hope for the best.

"The alternative of progressing malignancy is likely to be invariably fatal," Postow said.

The alternative of progressing malignancy is likely to be invariably fatal Dr Michael Postow

Menzies also indicated that these toxicities do not deter them from treating with ICIs. "Toxicity, both mild or frequent and rare or serious, is a consideration for all patients, with differing impact based on the patient's priorities (efficacy vs toxicity), their underlying health conditions, and treatment context (adjuvant/metastatic)," Menzies told Medscape Medical News.

Postow and Menzies both noted that decisions regarding the use of ICIs in the adjuvant treatment of melanoma are more complex. "I would not routinely recommend adjuvant PD-1 [programmed cell death protein–1] immunotherapy for patients with resected melanoma who have severe autoimmune diseases," Postow said.

Observations From a Lung Cancer Expert

Gainor, who was not associated with the study, pointed out that a published report on myocarditis in 35 patients included a sizeable cohort from the MGH.

"As a community, we did not fully appreciate these toxicities. As our experience with using ICIs continues, we may begin to see relatively rare adverse events, such as myocarditis," he said. "With respect to the onset of this toxicity, the MGH experience is similar to that reported in the Lancet Oncology article — it occurs early," Gainor said. He added that high levels of immunosuppression are required to control it.

In the MGH-authored report, the prevalence of myocarditis was 1.14%, and the median time to onset after starting ICIs was 34 days. Over 102 days of follow-up, 16 patients developed major adverse cardiac events (MACEs) and 38% of MACEs occurred in patients whose ejection fraction values were normal. Steroids were administered to 89% of patients.

"Currently, we are seeing the sickest patients," Gainor noted. "The fulminant cases of myocarditis are being reported," he said. "However, there may be subclinical forms of these toxicities that are still not recognized," he added.

Gainor was less certain about cardiac toxicity being associated with ICIs used for specific diseases. "This is an observation that needs to be further evaluated," he said. He explained that the time frame of the study captured more patients who were undergoing treatment for melanoma and lung cancer, which were the earliest indications for which ICIs were approved.

Gainor is reassured that existing guidelines will evolve and that clinicians will be informed regarding these emerging toxicities as the guidelines are updated.

"What is important at this time is to get more data on how these toxicities [cardiac irAEs] present and the best management approach," he said.

Details of the Pharmacovigilance Study

The observational, retrospective, pharmacovigilance study was a "disproportionality" analysis based on adverse events reported in VigiBase, the WHO database of global deduplicated individual case study reports. VigiBase includes data from reports from more than 130 countries, as well as the US Food and Drug Administration's pharmacovigilance database.

Using VigiBase, the researchers used novel, statistical tools that identified these new signals in cancer patients treated with ICIs alone or in combination. "Disproportionality analysis allows us to see events that are more than expected or that have a higher than expected chance of occurring and which are reported as RORs [reporting odds ratios]," Moslehi explained.

The ROR was with reference to the entire database, which catalogued the adverse event across the entire spectrum of diseases.

There were a total of 12,455,401 adverse events in the database that occurred during the period in which adverse events associated with the use of ICIs were reported. There were 31,321 adverse events reported with all ICIs (9667 from clinical trials; 21,654 from clinical practice).

ICI treatment was associated with a higher reporting for myocarditis (5515 for the entire database vs 122 for ICIs; ROR: 11.21), pericardial diseases (12,800 for the entire database vs 95 for ICIs; ROR: 3.80), supraventricular arrhythmias (68,597 for the entire database vs 222 for ICIs; ROR: 1.72), and vasculitis (33,289 for the entire database vs 82 for ICIs; ROR: 1.56).

"No other cardiac-related issues (eg, heart failure, heart attack, stroke) were found to be significant," Moslehi told Medscape Medical News.

The majority of these events were associated with settings other than clinical trials. This suggested to the researchers that these reports were a result of an expansion of the ICI indications and increased awareness from healthcare professionals during postmarketing surveillance rather than an increase in monitoring from clinical trial settings.

The editorialists note that the occurrence of myocarditis with ICIs has been reported by other groups. Although mechanisms for myocarditis have not been elucidated, preclinical studies have found that deletion of PD-L1 and treatment with PD-L1-inhibiting antibodies can cause lethal myocarditis in rats, Tocchetti and colleagues point out.

The fact that more cases of pericardial diseases occurred in patients with lung cancer may be due to synergy between radiotherapy and immunotherapy that patients receive, Moslehi and colleagues note. Patients who receive ICI after undergoing thoracic radiotherapy may be prone to develop pericardial diseases, exposing potential shared antigens to recognition by T cells, Moslehi and colleagues suggest. Tocchetti and colleagues note that pericardial diseases are common complications of cancer.

As to ICIs causing an increase in cases of vasculitis in melanoma patients, "it is likely that the use of ICIs might indeed reproduce an immune environment favourable for the development of vasculitis," the editorialists suggest.

That cardiac-associated irAEs occur more frequently in men than in women may relate to the fact that more men are treated with ICIs in clinical trial settings. Another possibility is that men may more frequently have preexisting cardiovascular conditions or risk factors than women, Tocchetti and colleagues indicate.

Several of the authors and editorialists have disclosed relevant financial relationships, which are listed in their respective articles.

Lancet Oncol. Published online November 12, 2018, 2018. Abstract, Editorial


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.