Inflammatory arthritis due to immune checkpoint inhibitors: current approaches to management

Abstract

The introduction of immune checkpoint inhibitors (ICI) has changed the landscape of treatment of cancer. As a consequence of immune activation, ICIs can cause immune-related adverse events (irAEs) affecting almost all organ systems. Several irAEs have now been described such as ICI-inflammatory arthritis (IA), sicca syndrome, polymyalgia rheumatica (PMR), myositis and vasculitis. The onset of the ICI-IA can vary from after the first infusion of ICIs to a delayed presentation a year or more after ICI initiation. Just as de novo IA is pleomorphic, several different forms of ICI-IA have been reported. Physical examination, lab and imaging tests can help the rheumatologist in determining the grade (severity) of arthritis. Corticosteroids tend to be the first line agent for treatment of ICI-IA with initial doses varying depending on the grade of the arthritis. For patients who are unable to taper steroids or who do not have a good clinical response, different forms of immunosuppression are required. Ultimately, baseline patient and tumor characteristics, the types of immunotherapies used (monotherapy versus combined therapies), pre-existing autoimmune diseases and/or other irAEs, as well as patient preferences will all shape the discussions around ICI-IA management. Engaging in multi-disciplinary discussions is needed when caring for this patient population.

Introduction:

In the past several years, introduction of immune checkpoint inhibitors (ICI) has changed the landscape of treatment of cancer. Using a novel approach involving activation of the host immune response, immunotherapy has improved prognosis and survival in several cancers [1]. ICIs are monoclonal antibodies that block regulatory immune molecules allowing T-cell activation and an enhanced immune response against cancer cells [2].

Immune checkpoints play an important role in controlling T-cell function by down-regulating T-cell activation, allowing self-tolerance, and preventing autoimmunity. Several proteins have been identified as immune checkpoints, including cytotoxic T-lymphocyte associated protein-4 (CTLA-4), programmed cell death protein-1 (PD-1) and programmed death-ligand 1 (PD-L1), while several others are being investigated [2]. Blockade of checkpoint proteins prevents inhibitory signaling and promotes a robust anti-tumor response. Ipilimumab, a monoclonal antibody against CTLA-4, was the first ICI approved for treatment of metastatic melanoma in 2011 [3]. Since then, several other therapies targeting PD-1, PD-L1, and LAG-3 have been approved for treatment of a variety of cancers, and multiple studies are ongoing evaluating other checkpoints and combination therapies.

As a consequence of immune activation, checkpoint inhibition with ICIs has been associated with immune-related adverse events (irAEs) affecting almost all organ systems. While endocrine, dermatologic, and gastrointestinal toxicities are seen most commonly, rheumatologic irAEs occur in about 5–10% of patients treated with ICIs [2]. Arthralgia is the most common reported rheumatologic irAE, up to 40% in some placebo-controlled trials [4]. Many other rheumatologic conditions have been reported. While the majority of irAEs improve with cessation of ICI therapy or therapy nonsteroidal anti-inflammatory drugs and/or glucocorticoids, some patients may be steroid refractory requiring further immunosuppression and/or go on to develop persistent symptoms. In this chapter, we focus on clinical features, evaluation and management of ICI-induced inflammatory arthritis (ICI-IA).

Clinical characteristics and epidemiology of ICI-IA:

Several rheumatic irAEs have now been described from such as ICI-IA, sicca syndrome, polymyalgia rheumatica (PMR), myositis and vasculitis. The most common of these rheumatic irAEs is inflammatory arthritis, with an incidence in one systematic review of 7% [4] although the rates of arthralgias (joint pain without swelling) are much higher. The onset of the ICI-IA can vary from after the first infusion of ICIs to a delayed presentation a year or more after ICI initiation. Cappelli et al reported that different ICI regimens might lead to development of distinct IA phenotypes [5]. In that cohort of 30 patients with ICI-IA, those treated with combination ICI therapy were more likely to present with knee arthritis, to have higher levels of C-reactive protein, and to have already had another irAE. In contrast, patients treated with ICI monotherapy were more likely to have initial small joint involvement and to have IA as their only irAE.

Just as de novo IA is pleomorphic, several different forms of ICI-IA have been reported. About one half to two-third of the cases of ICI-IA have a rheumatoid arthritis-like presentation with symmetric small joint involvement [6, 7]. A unique feature of ICI-IA is presentation with impressive tenosynovitis [8] and/or enthesitis-sometimes even without arthritis. Another phenotype is the asymmetric large joint involvement that mimics spondyloarthropathy (approximately 13–15% of cases).

Physical examination and serial evaluation:

The approach to initial evaluation and management of ICI-IA is outlined in Figure 1. The first step is to identify whether the presenting symptoms and signs point to an inflammatory arthritis or exacerbation of pre-existing traumatic injury or mechanical arthritis (termed “activated” osteoarthritis). The latter entity has been recently described [9] and can be managed with conservative treatments without compromising ICI responses. Physical examination should focus on assessing the number/symmetry of joints involved with pain and swelling, whether there is tendon/entheseal involvement, and whether there is axial involvement which is rare but reported as a manifestation of ICI-IA. Laboratory assessment and imaging tests can help determine the severity of inflammation and to also rule out alternative causes of joint pain such as metastasis or septic arthritis. Unlike RA where the autoantibodies (RF, anti-CCP) are prevalent, ICI-IA is unique in usually being seronegative. In a systematic review of case reports and series of ICI-IA, only 9% of patients with ICI-IA were seropositive for RF and/or anti-CCP [6].

Figure 1:

Flow diagram indicating the approach to management when seeing a patient with question of ICI-IA in the clinic.

Physical examination, lab and imaging testing can help the rheumatologist in determining the grade of arthritis. Per the commonly used Common Terminology Criteria for Adverse Events (CTCAE) [10], grade 1 ICI-IA would be defined as mild joint pain; grade 2 as moderate joint pain that is limiting instrumental ADLs; grade 3 as one where the pain is severe, sometimes requiring urgent care or admission, there is erosive or permanent joint damage, and/or symptoms pose severe limitations to daily activities.

Use of imaging in ICI-IA evaluation:

Imaging techniques helpful in diagnosis and monitoring of patients with ICI-IA include plain radiographs, musculoskeletal US (MSUS) and magnetic resonance imaging (MRI), and positron emission tomography (PET) scans. Ultrasound and MRI imaging have been used to define patterns of musculoskeletal involvement as well as identify soft tissue inflammation. Imaging modalities can also assess for structural damage and identify aggressive disease.

Plain radiographs are the most commonly performed imaging studies in patients with inflammatory arthritis. While they are easily accessible and can identify underlying pathologies such as osteoarthritis as well as rule out mechanical processes such as fractures, plain radiographs have low sensitivity in early stages of inflammatory arthritis. Early findings such as periarticular osteopenia are non-specific, while advanced findings such as erosions may not be seen for several months or longer after onset of symptoms. Radiographs have a limited role in diagnosis of early inflammatory arthritis, but are useful in monitoring progression.

MSUS is becoming an increasingly popular modality for assessment of joint and tendon pathology. MSUS has been shown to be more sensitive than physical examination to assess joint inflammation, and more sensitive than plain radiograph to evaluate for evidence of joint damage such as erosions [11]. Synovitis was the most common finding reported in a descriptive study of 9 patients with ICI-induced arthritis, followed by tendon involvement and tenosynovitis. Effusions and enthesitis were reported most commonly in the knees [12]. In another study, synovial hypertrophy, and power doppler similar to patients with rheumatoid arthritis was reported in patients with seronegative ICI-IA. Peripheral symmetric synovitis was also noted in patients with PMR-like symptoms without clinically evident peripheral arthritis [13]. Other findings commonly reported in patients with ICI-IA include synovial hypertrophy, power doppler signal, bursitis and dactylitis. Bone erosions and cortical damage have been reported as well and are considered suggestive of aggressive disease. Additionally, early enthesophytes can be appreciated, another sign of structural changes due to ICI-IA. Figure 2 shows characteristic synovitis, tenosynovitis, joint effusions, and patellar enthesophyte from patients with ICI-IA assessed with MSUS.

Figure 2:

Images showing various features that can be observed in ICI-IA on musculoskeletal ultrasound (courtesy of Dr. Jemima Albayda).

Extensor tenosynovitis (▲) and synovitis (★) in a wrist on A) long view and B) transverse view C) Large effusion at a knee D) Enthesophytes at the quadriceps tendon (arrow)

MRI is highly sensitive imaging study for evaluation of peripheral as well as axial inflammatory arthritis. Tenosynovitis and synovitis were reported as a common features in hands and wrists observed soon after onset of symptoms, while joint effusions and synovial thickening was seen more commonly in large joints [14]. Bone marrow edema was noted as early as 4 weeks after onset of symptoms [15] suggestive of early disease, with likely higher risk of disease progression. Another study describing MRI findings in 10 patients with ICI-IA described 3 patterns of musculoskeletal involvement – predominant joint involvement, prominent “periarticular” involvement and myofasciitis [16]. Presence of fasciitis in all 3 patterns was a notable finding of this study. This is a key difference to traditional forms of IA like RA and spondyloarthritis. In 3 patients presenting with knee pain with no objective arthritis, myofasciitis was reported in surrounding muscles suggesting periarticular tissue infiltration.

FDG PET scans are commonly used to evaluate tumor burden and assess response to therapy. Increased metabolic activity has been reported in patients with ICI-IA and PET scans may have a role in imaging, though mild metabolic uptake has also been reported in non-inflammatory conditions [13]. In patients with tumor burden, it may be challenging to differentiate joint disease from tumor burden depending on the involved sites. FDG PET imaging likely plays a role in evaluation and monitoring of inflammatory arthritis in patient populations. Many patients may obtain serial scans for cancer monitoring, so could also be used to monitor ICI-IA presence and activity.

Corticosteroid regimens:

Similar to other irAEs, corticosteroids are the first line treatment for grade 2 or higher ICI-IA as outlined in guidelines from the NCCN, ASCO, SITC, and ESMO [17–20]. Initial corticosteroid doses can vary widely. For grade 2 ICI-IA, 10–20 mg prednisone/daily or equivalent is a recommended initial dose. If grade 1 symptoms do not improve with NSAIDs or intra-articular symptoms, they should be treated as grade 2. For grade 3 ICI-IA, defined as arthritis symptoms that lead to severe pain, limit instrumental ADLs, or lead to permanent joint damage (e.g., erosions on imaging), prednisone 0.5–1 mg/kg daily or equivalent is recommended. Guidelines published by oncology organizations typically recommend a step-up approach with steroid dosing. There are no published data comparing initial higher dose (0.5–1 mg/kg) to lower dose (10–20 mg) steroids in terms of efficacy, effects on tumor responses, or resultant cumulative steroid dose. For grade 2 or higher symptoms, referral to rheumatology is recommended so rheumatologists will have a key role in choosing the dose of initial steroids as well as the rapidity of the steroid taper.

There are several possible outcomes from corticosteroid treatment. A first question is when to assess a patient’s response to corticosteroids. Waiting anywhere from 2–6 weeks after initiation of steroids before determining response is generally accepted. Patients may be steroid responsive, steroid unresponsive where they have no clinical benefit from steroids, steroid resistant where there is incomplete improvement, or steroid dependent where symptoms recur when steroids are tapered leading to the inability to stop the steroids [21]. Definitions for these terms were recently proposed in a consensus paper from the Society of Immunotherapy for Cancer [21]. If the patient improves, the rheumatologist must decide how quickly to taper the steroids. Tapers may last 4–8 weeks with a goal of decreasing doses to prednisone 10 mg daily or lower. At these low doses of prednisone, most oncologists will feel comfortable restarting ICI therapy. If patients have severe symptoms (grade 3) and are deriving no benefit after 2 weeks of corticosteroids, then considering early DMARD/biologic therapy is indicated.

Other forms of corticosteroids:

Intra-articular steroids are useful if 1–3 medium/large joints are affected. In Oncology guidelines, intra-articular steroids are only recommended for mild ICI-IA, but in practice they can be used concurrently with systemic steroids or DMARDs/biologics. There is no literature about using intramuscular (IM) injections of corticosteroids for ICI-IA instead of an oral taper. IM corticosteroids may be an option for patients who have difficulty with oral medications or to minimize cumulative steroid dose, but more evaluation of this approach is needed.

Therapeutic steroid sparing agents for ICI induced inflammatory arthritis:

As stated previously, for patients who are unable to taper steroids or who do not have a good clinical response, different forms of immunosuppression are required. Another consideration of steroid sparing agents in the treatment of ICI-IA is the evidence of steroids diminishing the effectiveness of ICI anti-tumor efficacy. This has been well documented in studies showing that a dose of 10mg or more of prednisone prior and at the initiation of ICI impacted anti-tumor response [22, 23]. Some studies, however, have shown that steroids do not have this deleterious effect if used later during the ICI treatment. Steroids used to manage irAE have less impact on progression free survival (PFS) than at baseline, but these patients still tend to have shorter PFS than patients without any steroid treatment [23]. The current guidelines advise to use the steroid sparing agents when patients are steroid dependent or resistant. It is advised to use csDMARDs first [24]. Available evidence is based on retrospective observational data mostly excluding patients with preexisting autoimmune diseases (Table 1).

Table 1.

Studies investigating the effect of the main steroid-sparing agents on inflammatory arthritis.

DMARD studied	Author (Reference)	N	Effect on cancer progression studied?	Comparative study?
Methotrexate	Bass et al [31] Leipe et al [28]	72 6	Yes Yes	Yes No
Hydroxychloroquine	Roberts et al [29]	11	Yes	No
Anti-TNF	Bass et al [31] De la Fuente et al [41] Braaten et al [40]	33 5 11	Yes Yes Yes	Yes No No
Anti-IL6R	Bass et al [31] De la Fuente et al [41] Fa’ak et al [42]	42 13 67	Yes Yes Yes	Yes No No

DMARD: Disease modifying anti-rheumatic drug.

Often the first option considered for ICI-IA and the most widely used is methotrexate [6, 25–27] as it is the first line treatment in the management of other rheumatic diseases. Methotrexate has been shown to be efficient and well tolerated [28]. Other options include hydroxychloroquine. which has been reported to be used in monotherapy or as a steroid sparing agent in 11 patients, with 7 patients responding [29]. Sulfasalazine is less commonly used and has been shown in a small case series to cause a drug hypersensitivity reaction in combination with anti PD-1 [30].

Biologic DMARDs may be considered to be used in severe irAE, when there is insufficient response to csDMARDs, or when a more rapid response is desired due to limited life expectancy or the need in the future to restart cancer therapy. Thus far anti-TNF and anti-IL-6R have been preferred options [24].

To guide the clinician to decide which treatment is effective and is less detrimental on the anti-tumor response a large (147 patients) retrospective comparative study comparing MTX with anti-TNF and anti-IL-6R in the treatment of steroid dependent ICI associated IA and PMR has been performed [31]. In the anti-TNF group and after adjustment on time from ICI initiation to DMARD initiation the time to cancer progression was significantly shorter compared to the MTX group while there was no significant difference between MTX anti anti-IL-6. On the other hand, anti-TNF therapy was able to control arthritis faster. The retrospective nature of this study led to imbalance in the maximal dose of steroids with higher doses in the bDMARD groups compared to the MTX. Other biologics have not been studied on a large scale with reports on the efficacy of anti-IL 17 in two patients [32] but a case of cancer progression [33]. JAK inhibitors have recently been shown to cause an increased risk of cancer in rheumatoid arthritis patients compared to anti-TNF HR 1.48 (1.04–2.09) and are thus not a preferred option [34]. In the future, anti-IL-12/23 agents may be considered as they have been used with some success in dermatologic and gastrointestinal irAEs [35, 36]. The limited literature from translational studies of ICI-IA, suggests targeting Th17 or Th1 associated cytokines would be beneficial [36]. Ultimately a better understanding of pathophysiology will lead to targeted therapy for ICI-IA.

Co-management with other irAEs:

Commonly irAEs do not occur in isolation, and patients may experience more than one irAE over the course of ICI treatment and in the months after ICI cessation. In prior studies of ICI-IA, 40–50% of patients have had at least one other irAE [6, 37]. Some common irAEs that may present prior to, concurrently with, or after ICI-IA include colitis, skin rashes (psoriasis and others), pneumonitis, and endocrine irAEs. As outlined in NCCN guidelines [20], corticosteroid dosing may differ for some of these irAEs as compared to ICI-IA. With colitis, pneumonitis, and hepatitis, for example, recommended corticosteroid dosing is significantly higher, as high as 2 mg/kg in grade 3 cases. As a result of both these differences in steroid dosing and differences in time to onset, patients may first present with ICI-IA when they have weaned off their steroids or other immunosuppression for another irAE.

In a patient with colitis and ICI-IA, TNF-inhibitors should be considered if steroids are ineffective given efficacy for both irAEs. Vedolizumab, though effective for colitis as an irAE, does not have the same efficacy for ICI-IA. Additionally, there are reports of using ustekinumab for refractory colitis [38] and this may benefit IA symptoms as well.

Patients with psoriasis or a psoriasiform rash may benefit also from TNF-inhibitors. IL-12/23 inhibitors have been used for skin disease due to ICIs [36] and for traditional psoriatic arthritis, though the benefit in ICI-IA is less well documented.

Ongoing cancer therapy considerations:

Since ICI-IA is not a life-threatening irAE, the oncologist may choose to continue the ICIs even in the setting of ongoing symptomatic IA. Indeed, patient preference is often to continue ICIs as long as possible due to fear of stopping this life extending therapy [39]. There is some evidence for using steroid sparing agents like TNF-inhibitors or methotrexate while patients are actively getting ICIs. Patients may also receive 10 mg prednisone equivalent daily or less to decrease symptoms while receiving ICIs. Concerns about impairing tumor response due to immunosuppression are covered in another chapter (The Impact of traditional inflammatory arthritis therapies on cancer response to immune checkpoint inhibitors) of this edition.

ICIs are being given more commonly with chemotherapy. This may affect both the presentation and treatment of ICI-IA. If steroids are given as premedication for chemotherapy, patients may actually feel better in the days after the infusions. Also receiving active chemotherapy may limit the types of immunosuppression that may be safely used for ICI-IA.

Chronic IA management:

ICI-IA, like other irAEs, can persist after ICI cessation for months to years [21, 37, 40]. As outlined in consensus definitions, irAEs can be chronic and active (requiring immunosuppression) or chronic and inactive (e.g. endocrine irAEs where only hormone replacement is required [37]. For patients with chronic active irAEs like ICI-IA in particular, it is important to find steroid sparing immunosuppression that is effective given the side effects of long-term steroid use.

Several risk factors for persistence of ICI-IA greater than 6 months after ICI cessation have been identified. In the first study to evaluate this risk, longer duration of ICI use, receiving combination ICI therapy, and having multiple other immune-related adverse events were all independent risk factors for persistence [40]. In a follow up study, combination anti-CTLA-4/PD-1 therapy was again noted to be associated with higher risk of persistent ICI-IA as well was steroid use at baseline rheumatology evaluation [37].

There is naturally concern about long term immunosuppression and risk of cancer progression. This is detailed in more length in another chapter of this addition (The Impact of traditional inflammatory arthritis therapies on cancer response to immune checkpoint inhibitors). Specifically for ICI-IA, in a multicenter observational study of treatment of ICI-IA with DMARDs biologics like TNF-inhibitors and IL-6R inhibitors were associated with shorter time to cancer progression as compared to methotrexate [31]. These groups were not compared to a corticosteroid alone group so it is not clear if there is any more risk with steroid sparing immunosuppression than there would be for those who received only steroids.

Conclusions:

In summary, the management approach to a patient with ICI-IA requires detailed attention to several features as ‘no one size fits all’. Baseline patient and tumor characteristics, the types of immunotherapies used (monotherapy versus combined therapies), pre-existing autoimmune diseases and/or other irAEs, as well as patient preferences will all shape the discussions around ICI-IA management. Another important aspect of caring for these patients is the need for multidisciplinary discussions and dialogue between the oncologist and rheumatologists so that the patients are not burdened with different opinions from their different physicians. Although the current level of evidence for several therapeutic decisions involved in the care of these patients is low, multiple registries, multi-institutional collaborations, and clinical trials (for example, NCT04375228) are underway and have the potential to pave the way for improved guidelines and evidence in this arena.

Key points:

• In the past several years, introduction of immune checkpoint inhibitors (ICI) has changed the landscape of treatment of cancer.

• Using a novel approach involving activation of the host immune response, immunotherapy has improved prognosis and survival in several cancers.

• ICIs are monoclonal antibodies that block regulatory immune molecules allowing T-cell activation and an enhanced immune response against cancer cells.