Toxicities and outcomes: do steroids matter?

Abstract

Hypophysitis is a common toxicity of ipilimumab and may be managed with either low- or high-dose steroids. In this issue of Cancer, Faje and colleagues observe that patients treated with low-dose steroids have equivalent rates of hypophysitis resolution compared to those who receive higher doses. Importantly, however, melanoma-specific outcomes are improved in the low-dose cohort. Herein, we discuss the implications of this finding and the known associations between steroids and clinical outcomes to cancer immunotherapy.

Immune therapies, particularly those targeting programmed death-1/ligand-1 (PD-1/PD-L1) and cytotoxic T lymphocyte antigen-4 (CTLA-4) have transformed the treatment outlook for a variety of cancer types.¹ Toxicities are frequently triggered, however, by inhibiting these key negative regulators of adaptive immunity. When mild, these toxic events are treated with expectant management and supportive care, but require moderate to high dose systemic glucocorticoids (usually with prednisone 1mg/kg or equivalent) when more severe.^{2, 3} As such, steroid administration remains the keystone to reversing immune therapy toxicities.

Glucocorticoids induce lymphopenia, inhibit gene transcription of proinflammatory genes, block secretion of inflammatory cytokines, and have various other immunosuppressive mechanisms.⁴ Initially, clinicians feared that the lympholytic and immune suppressive properties of steroids, when administered for toxicities, would blunt the antitumor immunity generated by immune checkpoint inhibitors. A number of subsequent studies, however, appeared to dismiss this fear. Several clinical trials and retrospective studies demonstrated that toxicities and their attendant steroids were associated with at least equivalent, if not superior clinical outcomes (in terms of tumor response, progression-free survival, and overall survival).^5–7 As such, the hesitancy for using steroids to treat toxicities has diminished, and these agents remain the mainstay of treatment algorithms for severe toxicities across organ systems.

In this issue of Cancer, Faje and colleagues report their experience treating 98 patients with ipilimumab-induced hypophysitis. As previously published in a smaller study,⁸ patients with hypophysitis had superior clinical outcomes compared with patients without this toxicity. Importantly, in a subset of patients treated with ipilimumab monotherapy, patients who received low-dose steroids had superior clinical outcomes in terms of time to treatment failure (hazard ratio 0.28, p=0.001) and overall survival (hazard ratio 0.24, p=0.002) compared to those who received high-dose. These findings persisted when integrated into multivariable models that included known prognostic factors, such as performance status and lactate dehydrogenase levels. Thus, this finding suggests that higher doses of steroids could potentially impact anti-tumor immunity and counteract the beneficial correlation between toxicity and response.

The study adds several important insights into our understanding of immune therapy treatment and toxicities. First, in contrast to most other inflammatory toxicities, hypophysitis tends to be self-limited, and may be safely treated with replacement dose steroids alone instead of high-dose steroids. Indeed, in this study, patients with low-dose vs. high-dose steroids had equivalent rates and timing of resolution of their hypophysitis, both in terms of imaging findings and clinical symptoms. Thus, as the authors point out, this patient population provided a somewhat unique opportunity to dissect the impact of toxicity from the impact of steroid treatment on tumor-specific outcomes. Second, hypophysitis tends to occur during the first four doses of ipilimumab monotherapy administration.⁹ This avoids the confounding factor of many studies that have linked toxicities (that may occur at any time while on therapy) with response: toxicities occur at higher rates when patients continue to receive therapy; but they only continue to receive therapy if they are benefiting.^{6, 7} Thus, this study provides further evidence that at least this particular toxicity (hypophysitis) is truly linked with improved outcomes. Third, still more evidence⁸ is shown that low-dose steroids (rather than high-dose) are sufficient to treat hypophysitis, and should be considered the standard of care. Although ipilimumab monotherapy use will likely continue to decline moving forward, recent studies have demonstrated excellent activity of ipilimumab in combination with anti-PD-1 agents in renal cell carcinoma and lung cancer.^10–12 Thus, hypophysitis, a condition nearly unique to ipilimumab therapy, will only continue to grow as a clinical problem for providers across disciplines.

While important, this study does present some limitations that preclude a widespread conclusion that steroids have an adverse impact on anti-tumor immunity. First, and most obviously, the low-dose group only had 14 patients. Thus, this could have been a well-performing population due to chance alone. Second, limiting the low-dose group to those who received 7.5mg of prednisone or less seems a somewhat arbitrary cutoff, since many patients received slightly over this. While this does represent a reasonable threshold for a physiologic dose, one would presume that a major difference on the immune response exists between, for example, a month-long taper of prednisone 1mg/kg and a short burst of 15mg prednisone daily. Exploratory analyses, however, showed similar findings at somewhat higher cutoffs (up to 15mg), strengthening the findings. Third, hypophysitis appears to have a somewhat distinct mechanism compared with other immune checkpoint inhibitor toxicities. Specifically, ipilimumab-induced hypophysitis seems to arise from antibody binding CTLA-4 receptors that are directly expressed by the pituitary gland, thus leading to complement activation and an inflammatory cascade.¹³ Direct T cell mediated cytotoxicity, pre-existing tissue-specific inflammation, cytokine imbalances, and autoantibodies have been implicated in other toxic conditions.^14–16 Thus, generalizing associations observed here to other toxic conditions may not be appropriate.

This study raises other, fundamental issues surrounding immune checkpoint inhibitor therapies. While some early insights have been gained, the mechanisms of immune checkpoint inhibitor toxicities remain poorly defined. At this time, we have no ability to predict which patients will experience complications from therapy, including seemingly obvious factors like toxicities with prior immune therapies, history of autoimmune disease, or pre-existing organ dysfunction.^17–21 Unraveling mechanistic insights surrounding treatment-related toxicities may help identify patients at high risk, develop more effective immunomodulators, and perhaps most importantly, design novel therapeutics to unleash antitumor immunity more specifically. Glucocorticoids are a fairly effective treatment for immune-related toxicities, but remain a blunt, non-specific way to suppress aberrant immunity. Designing inhibitors of culprit cellular populations or cytokines may combat toxicity without compromising immune therapy efficacy or promoting systemic immunosuppression.²² In addition, the data presented in this study are useful, but at this time low-dose steroids are not a viable alternative for most serious immune toxicities for safety reasons. Prospective studies to limit steroid doses examining modestly lower doses (e.g. 0.5mg/kg prednisone equivalent) and shorter tapers would be useful, but difficult to perform.

In conclusion, this study shows that low-dose steroids do not impair the resolution of ipilimumab-induced hypophysitis, and suggests that they improve melanoma-specific outcomes compared with higher doses. Developing more effective immunomodulators, and optimizing the ones currently available to preserve antitumor immunity while suppressing tissue-specific inflammation remains a critical and unmet need.