Reconsidering Dexamethasone for Antiemesis when Combining Chemotherapy and Immunotherapy

Abstract

Whether the immune suppressive action of glucocorticoid steroids, such as dexamethasone, might reduce the benefits of cancer immunotherapy has long been a concern. Observations that established tumor regressions in response to immune checkpoint inhibitors (ICIs) often persist, despite the use of steroids to mitigate ICI‐related autoimmune breakthrough, are not sufficiently reassuring, because these observations do not address the potential blunting of immune priming at the initiation of ICI therapy. With increasing indications for ICI in combination with chemotherapy, this issue merits reconsideration. Professional society guidance advises that dexamethasone should be used as first‐line prophylaxis for nausea and vomiting in patients receiving ICI and highly emetogenic chemotherapy combination regimens. Here, we review the availability of data on this subject and propose an alternative approach focused on the adoption of steroid minimization or sparing for prophylaxis of nausea until the underlying immune biology is better understood.

Short abstract

This commentary considers clinical decision‐making for the prescribing of dexamethasone as an antiemetic for patients receiving chemotherapy in combination with immune checkpoint inhibitor therapy.

Introduction

Glucocorticoids (GCs), such as dexamethasone, are used clinically as both potent antiemetics and immune suppressants. Chemotherapy‐induced nausea and vomiting (CINV) can be a debilitating side effect of common chemotherapy regimens, with a significant impact on patient quality of life [1] and high associated health care costs [2]. Dexamethasone is given in combination with 5‐hydroxytryptamine (5‐HT₃) and neurokinin‐1 (NK₁) receptor antagonists as prophylaxis [3, 4] for patients receiving highly emetogenic chemotherapy (HEC), such as carboplatin [5], and is a commonly used “as‐needed” prescription for nausea complicating HEC and other chemotherapy after treatment. HEC is now indicated in combination with immune checkpoint inhibitors (ICIs) for cancers such as non‐small‐cell lung cancer (NSCLC), small‐cell lung cancer (SCLC), and head and neck squamous cell carcinoma (HNSCC) [6, 7]. ICIs and other less emetogenic chemotherapy are also garnering increasing regulatory approval [8, 9], and it is highly likely that additional combination immunotherapy regimens will be licensed in the future, including novel immunotherapy targets [10, 11]. As a result, dexamethasone is likely to be coadministered with ICIs [4], posing the question of whether its immunosuppressive effects may lessen the efficacy of ICI. Wide national variation in adoption of antiemesis guidelines [12] has prompted the U.S. Centers for Medicare and Medicaid Services to implement 30‐day postchemotherapy acute care as a new performance measure [13]. In this perspective, we discuss considerations for clinical decision‐making around antiemetic prescribing in patients receiving combination ICI treatments.

Biological Function and Pharmacology of Glucocorticoids

GCs are a type of steroid hormone derived from cholesterol that bind to and then activate the glucocorticoid receptor (GR), which is expressed in almost all cells. The main endogenous GC in humans, cortisol, was first isolated from the adrenal cortex in 1938 [14]. The immunosuppressive role of cortisol was discovered in 1948, with the finding that injections of cortisol could treat rheumatoid arthritis (RA) [15], leading to the Nobel Prize in Medicine in 1950. Cortisol is essential for life [16] and is released according to a circadian rhythm, with peaks before waking and troughs around midnight [17]. Synthetic GCs have been developed through chemical modification of cortisol's molecular structure; examples include prednisolone, methylprednisolone and dexamethasone [18]. The major differences between GC derivatives relate to potency, duration of action, and mineralocorticoid receptor affinity. For example, dexamethasone has been modified to prolong its duration of action (using cortisol suppression as a surrogate marker) from approxiamtely 8 to 36 hours and increase its anti‐inflammatory potency 20‐fold [19]. Despite having a similar plasma half‐life to prednisolone (3–4 hours) [20], dexamethasone has a markedly longer duration of action [21], suggesting the possibility of altered binding kinetics with GR [22]. GC‐mediated activation of GR causes homodimerization and translocation to the nucleus, where the GR drives a complex transcriptional program affecting up to 20% of all genes [23], with pleiotropic effects on multiple immune subsets [24]. The dominant mechanism by which GCs drive immune suppression remains poorly understood. Recently, GR signaling was shown to promote gene expression signatures reflecting T‐cell dysfunction, associated with increased expression of inhibitory checkpoint receptors such as programmed cell death protein‐1 and increased interleukin‐10 production [25].

Antiemetic Effects of Glucocorticoids

Exogenous GCs, in particular dexamethasone, are routinely used for the preventative treatment of CINV [26], and benefit is enhanced in combination with 5‐HT₃ receptor antagonists such as granisetron and NK₁ receptor antagonists such as aprepitant [27, 28]. This is consistent with evidence that CINV is mediated by abdominal vagal afferents and neurons in the area postrema, both of which express 5‐HT₃ and NK₁ receptors [29]. Although dexamethasone is the most common exogenous GC used for antiemesis, response rates to methylprednisolone are broadly comparable [30, 31]. The antiemetic mechanism of action of GC is poorly understood and may involve reduction prostaglandin production [32], direct noncompetitive modulation of the 5‐HT₃ receptor [33] or compensation for acute adrenal hypofunction induced by platinum‐based chemotherapy [34]. In comparison with other synthetic GCs, dexamethasone has a prolonged duration of action, which could partly explain its efficacy for CINV.

Immune Suppressive Effects of Glucocorticoids

Endogenous GC secretion is inversely related to inflammation [35, 36]. Exogenous GCs are prescribed routinely for their immunosuppressive effects, across a range of indications including transplantation, RA, systemic lupus erythematosus, asthma, eczema, and septic shock. Prednisolone treatment is sufficient to prevent rejection of a genetically distinct (allogenic) transplant [37]. It has been recently demonstrated that the transcriptional immune profile of tumors responding to cancer immunotherapy shares the same activation profile and integrated immune response as rejecting renal allografts, including accumulation of T, natural killer, and B cells and the formation of tertiary lymphoid structures [38]. Furthermore, ICI treatment can promote both allograft rejection and tumor regression in the same patient [39], consistent with convergent immunological responses. Considering that, like allografts, tumors express non‐self antigens that derive from accumulation of coding genetic alterations (neoantigens) [40], it is conceivable that GC signaling would promote immunological tolerance to tumors and so curb the effectiveness of ICI. Consistent with this hypothesis, moderate elevations in systemic GCs are sufficient to cause suppression of T‐cell activation signatures and failure of anti–programmed death ligand‐1 antibody and CXCR4 antagonist treatment in an in vivo spontaneous genetically engineered model of pancreatic cancer [41].

Clinical Studies Relating to Coadministration of Glucocorticoids and Immunotherapy

There are three frequent situations in which GCs could be coadministered with ICIs (Fig. 1): baseline (for example, patients with chronic autoimmune conditions), antiemetic prophylaxis for HEC, and secondary immune modulation in patients with immune‐related adverse effects (irAEs), such as colitis [42]. irAEs are strongly associated with beneficial responses [43, 44, 45, 46], likely reflecting systemic immune competence [47], and in this subset there is minimal evidence that exogenous GCs blunt response to treatment. In contrast, baseline administration of supraphysiological doses of GCs is associated with adverse clinical outcomes in melanoma, NSCLC, and glioblastoma [48, 49, 50]. For example, Arbour et al. reported that baseline GC equivalent to more than 10 mg prednisolone per day was associated with significantly poorer overall survival on multivariate analysis in patients with NSCLC (hazard ratio, 1.66; p < .001) [48]. Genetic variants partly explain the risk of autoimmune conditions that are treated with systemic steroids, such as RA [51], and in turn, such variants are associated with enhanced responsiveness to ICI [52]. This potentially confounding issue may partly explain contradictory findings regarding the association between baseline GC and ICI response, which has not been consistently demonstrated in all patient cohorts [53]. Consistent with this, an analysis of ICI‐treated patients with NSCLC found that the subset of patients taking GCs for underlying cancer‐unrelated indications, such as RA, had noninferior survival outcomes [54].

Figure 1.

Coadministration of glucocorticoids (GCs) with ICIs can be classified into three phases: prophylaxis coinciding with day 1 for patients treated in combination with HEC drugs such as carboplatin, secondary immune modulation coinciding with irAEs or baseline chronic treatment in patients with chronic autoimmune disease or organ transplants. irAEs are associated with an active therapeutic response, with unmasking of preexisting anti‐tumoral immunity by ICIs. Approximate doses are shown as prednisolone equivalents in milligrams, with 12 mg dexamethasone approximately equivalent to 80 mg prednisolone.Abbreviations: HEC, highly emetogenic chemotherapy; ICI, immune checkpoint inhibitor; irAE, immune‐related adverse effect.

Approximately 40% of ICI clinical trials specifically preclude the use of baseline exogenous GC [55]. This includes trials demonstrating the benefit of combination ICI regimens in NSCLC, SCLC, and HNSCC [8, 56, 57, 58, 59, 60, 61], making it difficult to extrapolate these findings to patients on long‐term GC treatment. There is a paucity of data relating the implications of short‐term dexamethasone as antiemetic prophylaxis in patients receiving ICI in combination with HEC. Trials examining ICI‐HEC combinations vary between recommending [56], permitting [59], or explicitly advising against dexamethasone‐based prophylaxis [11, 57, 58, 60, 61]. Consequently, from the available data, it is not possible to assess what proportion of patients in these trials received dexamethasone prophylaxis and whether it was associated with inferior clinical outcomes. In clinical trials of agonistic CD40 monoclonal antibodies as immune therapy of cancer, biomarkers of immune response at 2–3 days after infusion are highly blunted if the patient receives dexamethasone either prophylactically or as needed for monoclonal antibody infusional reactions [62]. Altogether, the totality of molecular, preclinical, and clinical data suggest that supraphysiological GC treatment preceding ICI may well reduce effectiveness. It is likely that these findings can be extrapolated to dexamethasone‐based emesis prophylaxis, despite absence of specific data.

Conclusion and Perspectives

Available evidence supports a model in which steroids, akin to their role in transplant medicine, can drive immunological tolerance to the tumor and so undermine the action of ICIs. When ICIs were first approved as single‐agents this issue was less concerning, but with increasing number of indications of ICI and chemotherapy, it becomes important to better understand. We raise the question of whether the clinical benefits of ICI and chemotherapy may be suboptimal in the setting of GCs given prior or during chemotherapy administration. Recent guidance from the American Society of Clinical Oncology (ASCO), however, states that “there is no clinical evidence to warrant omission of dexamethasone from guideline‐compliant prophylactic antiemetic regimens when…[ICIs] are administered to adults in combination with chemotherapy” [4]. We propose consideration of an alternative approach focused on the adoption of dexamethasone‐sparing prophylaxis, with the aim of administering the minimum dose of dexamethasone that is sufficient to relieve symptoms.

Current national guidelines from National Comprehensive Cancer Network and ASCO recommend 8–12 mg dexamethasone once daily on days 1–4 of each cycle of HEC [4, 63]. In light of the prolonged duration of action of dexamethasone (>32 hours), it is likely that supraphysiological GR signaling may persist for days or even weeks after a 4‐day course of dexamethasone. Indeed, a 4‐day course of dexamethasone is at least as effective as a 4‐week course of prednisolone for primary immune thrombocytopenia [64], whereas the efficacy of a single dose of dexamethasone is broadly comparable to a 4‐day course of prednisolone for acute asthma exacerbations [65, 66]. Recently, a randomized controlled trial (RCT) in patients receiving HEC demonstrated that, when in combination with 5‐HT₃ and NK₁ antagonists, a single dose of 8 mg dexamethasone on day 1 is noninferior to a 3‐day course [67]. A recent RCT in cisplatin‐treated patients demonstrated the addition of 5 mg olanzapine once daily on days 1–4 to standard emesis prophylaxis (including dexamethasone) substantially increased the proportion of patients without vomiting from 66% to 79% [68].

We fully acknowledge the importance of minimizing side‐effects from treatments for patients and, therefore, of high‐quality prophylaxis of nausea and vomiting during HEC treatment. However, we and other colleagues [69] advocate for consideration of alternatives to high doses of dexamethasone, until more is understood regarding potentially negative effects of dexamethasone with ICI and chemotherapy. We recommend the lowest effective steroid dose, tailored to the patient, and suggest reconsideration of the broad use of standard operating procedures in oncology infusion suites that call for high‐dose dexamethasone administration as prophylaxis.

Disclosures

Robert H. Vonderheide: Medimmune, Verastem (C/A, H) Fibrogen, Janssen, Eli Lilly & Co(RF), inventor on a licensed patent relating to cancer cellular immunotherapy and cancer vaccines (IP), Children's Hospital Boston (Other‐royalties for a licensed research‐only monoclonal antibody). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board