Rheumatic adverse events such as inflammatory arthritis continue to accrue with the burgeoning use of immune checkpoint inhibitors (ICIs) in the treatment of cancer, presenting challenges in management for both rheumatology and oncology.
At the annual meeting of the Florida Society of Rheumatology in July, Andrew Ostor, MD, of Cabrini Medical Center in Melbourne, Australia, predicted an ever-increasing number of these complications.
"It's likely that the rheumatic complications of these drugs will become more common than rheumatoid arthritis itself," he stated.
The ICIs block co-stimulatory molecules on T-cells, antigen presenting cells, and tumor cells, which results in unchecked T-cell activation and the upregulation of tumor targeting by the immune system -- the opposite of the immunosuppression that is the goal of much treatment for conditions such as rheumatoid arthritis and lupus.
The agents currently available include ipilimumab (Yervoy), which targets cytotoxic lymphocyte antigen 4 (CTLA-4); nivolumab (Opdivo) and pembrolizumab (Keytruda), which target programmed cell death protein 1 (PD-1); and atezolizumab (Tecentriq), durvalumab (Imfinzi), and avelumab (Bavencio), which target programmed death ligand 1 (PD-L1). Numerous other ICIs are currently being developed.
The growing number of immune-related adverse events associated with ICIs have included colitis, rash, pancreatitis, and pneumonitis, and rheumatic conditions such as arthritis, myositis, vasculitis, sicca syndrome, and polymyalgia rheumatica. Guidelines have been established for managing colitis and pneumonitis, but only preliminary recommendations are available for the management of the rheumatic events.
The Mayo Cohort
A recent publication in from the Mayo Clinic in Rochester, Minnesota, reported on 61 cases of rheumatic syndromes in patients being treated with ICIs from 2011 to 2018. When asked if the incidence of these complications will likely continue to rise, lead author Uma Thanarajasingam, MD, told 鶹ý, "We don't have a definite answer to this as yet, but I would expect that as the number of checkpoint inhibitors approved for clinical use expands, as well as the clinical indications for their use, we will be seeing a greater number of rheumatic toxicities."
The most common event in this series was inflammatory arthritis, which developed in 34 patients, whose mean age was 59. The condition was polyarticular in two-thirds. The inflammatory arthritis was most common among patients treated with combination anti-CTLA-4/PD-1 treatment (4%), compared with 2% of those who were given anti-PD-1 or anti-PD-L1 (2%) or anti-CTLA-4 monotherapy (1%).
The treatment was prednisone alone in 62% of patients, with a mean starting dose of 30.6 mg/day and mean time on steroids being 18 weeks. An additional 15% received a disease-modifying antirheumatic drug along with the steroid, and 24% also were given nonsteroidal anti-inflammatory drugs or intra-articular steroids. The arthritis resolved completely in 47% and partially in 53%.
10 of the patients in the Mayo cohort developed myopathy, which was the most severe type of rheumatic adverse event. Treatment for these patients consisted of prednisone alone in half of the patients, with mean starting doses of 74 mg/day. The remainder required intravenous methylprednisolone, and complete resolution was seen in 70%. Two patients died from complications associated with myocarditis and bulbar myopathy.
Treatment for these patients currently is based on expert opinion, according to Thanarajasingam. "Prospective trial-based data for the treatment of rheumatic toxicities are lacking at present, and will likely require multicenter collaboration and study, which I hope will be achieved in the near future," she said.
The Hopkins Cohort
Another recent series included 30 patients from Johns Hopkins University in Baltimore reported in .
The patients, whose median age was 59, had been seen from 2013 to 2017. A total of 40% were women. The most common types of tumor were non-small cell lung cancer, in 40%, and melanoma, in 23.3%. All had been treated with monotherapy with an anti-PD1 or anti-PD-L1 agent (n=16) or combination therapy with an anti-PD-1 plus a CTLA-4 agent (n=14). 10 had experienced a complete tumor response.
More than half of patients presented with involvement of one or both knees, which was particularly common in the combination therapy group, whereas patients receiving monotherapy more often had initial symptoms in the small joints of the hands. Time to onset of arthritis varied from 1 to 24 months (mean 6.2 months), suggesting a significant delay in diagnosis, particularly for those with small joint involvement, the researchers said.
Two-thirds of patients had additional immune-related adverse events -- most commonly colitis but also pneumonitis and skin rash. Patients receiving combination therapy more often had colitis, suggesting involvement of the Th17 pathway.
Systemic corticosteroids were used for 24 patients, and 10 required additional immunosuppression, particularly those who had been given combination therapy. Seven patients were treated with tumor necrosis factor (TNF) inhibitors, and all showed improvement of their arthritis. Four of the TNF inhibitor recipients had complete tumor responses at the time of arthritis treatment initiation.
This study was the first to assess the clinical pattern of arthritis according to cancer treatment regimen, noted the researchers, led by Laura Cappelli, MD.
"This is critical information, not just for rheumatologists as they try to recognize subgroups in ICI-induced inflammatory arthritis and diagnose patients with this new entity, but also for oncology providers who are usually first to encounter patients with ICI-induced inflammatory arthritis and subsequently refer patients to rheumatology," the team wrote.
Outcomes and Impact
Some of these toxicities resolve with stopping the checkpoint inhibitor and administering steroids or other treatments, but others have persisted long after the cancer treatment is withdrawn. "These are particularly challenging cases to manage, and it is my suspicion that at least a subset of cases are actually an unmasked pre-existing tendency to autoimmunity," Thanarajasingam said.
"For many patients, the symptoms eventually resolve with the cessation of the immune checkpoint inhibitor," Cappelli told 鶹ý. "But there is a subset of patients who go on to have chronic inflammatory arthritis. We're working to study this question in depth and to see how many patients develop this. Our best guess now is probably 20% to 25%."
As to the potential impact of treatment of the adverse events on cancer treatment, "the jury is still out," Thanarajasingam noted. In the Hopkins cohort, none of the patients who received anti-TNF treatment for their arthritis have thus far experienced a loss of cancer response.
"But there are conflicting reports in the literature as to the impact of steroids (the current mainstay of therapy) on overall survival and progression-free survival," Thanarajasingam said. Some retrospective studies have found no impact, whereas others found a slight negative influence on survival. "We need more prospective longitudinal data from large patient cohorts to better answer this critical question," she said.
And as to whether there might be a potential adverse impact of the arthritis treatment on cancer, that is an active area of research, Cappelli explained in an interview.
"It turns out that it might be the opposite. If you look at melanoma and lung cancer, people who developed some kind of immune-related adverse event actually had better responses to certain immune checkpoint inhibitors than those who didn't develop immune-related adverse events," she said. "So those issues may actually be a marker of who's going to respond well in terms of their cancer therapy."