Aug 20, 2018
This JCO Podcast provides observations and commentary on the JCO article, "Impact of Baseline Steroids on Efficacy of PD-(L)1 Blockade in Patients With NSCLC" by Arbour et al. My name is Deepa Rangachari, and I am an Assistant Professor of Medicine at the Beth Israel Deaconess Medical Center and Harvard Medical School in Boston, Massachusetts, USA. My oncologic specialty is thoracic cancers.
For decades, the management of advanced non-small-cell lung cancer has relied on use of cytotoxic chemotherapies with a median overall survival not exceeding one year. As this longstanding therapeutic approach has been limited by modest efficacy, finite durability, and significant treatment-associated toxicity, evolving better tailored, more effective, and less toxic care strategies has long been an unmet need. Beginning in the mid-2000s with the identification of actionable oncogenic driver mutations in the epidermal growth factor receptor, the landscape for personalized care in advanced non-small-cell lung cancer has been permanently transformed. Today, the evidence-based standard of care for upfront therapeutic stratification in all patients with advanced stage disease relies on mandatory genomic and immunologic profiling, with emphasis on identifying subsets of patients who will experience better outcomes and less toxicity through the use of disease-specific therapies.
It is in this landscape that the use of immune checkpoint inhibitors has flourished for patients with lung cancer. Since 2015, 3 different immune checkpoint inhibitors—Nivolumab, Pembrolizumab, and Atezolizumab—have all gained approval from regulatory agencies for management of patients with advanced stage non-small-cell lung cancer on the basis of phase III studies showing improved overall survival with more durable responses, less toxicity, and better quality of life with checkpoint inhibitors when compared with chemotherapy.[1-5] This benefit has been seen in both previously treated patients regardless of tumor programmed death ligand 1 (hereafter referred to simply as “PD-L1”) status and in upfront management of patients with high tumor PD-L1, defined by a tumor proportion score >/= 50%. Combined chemoimmunotherapy for patients with non-squamous disease regardless of tumor PD-L1 status is also now a vetted approach.
With widespread use of immune checkpoint inhibitors now the standard of care for many patients with this disease, important pragmatic concerns regarding concurrent use of checkpoint inhibitors and corticosteroids have emerged. Specifically, as many patients will present with refractory anorexia, nausea, fatigue, pain, brain metastases, and/or dyspnea for which corticosteroids are often used as helpful adjuncts to cancer-directed therapy for symptomatic palliation, what if any are the consequences of concurrent use? How do we balance these symptomatic issues against the need for delivering effective cancer therapy?
Due to concerns for the immunomodulatory effects of corticosteroids on T-cell activity and function, nearly all clinical trials leading to approval of these agents have excluded patients requiring the equivalent of Prednisone 10-20mg daily or more prior to initiation of checkpoint inhibitor therapy. In contrast, there is now an emerging body of evidence suggesting the safety of subsequent corticosteroid use, once checkpoint inhibitors have been started, for management of immune-related adverse events , without jeopardizing any previously achieved therapeutic benefit.[7, 8]
But what about those patients being newly initiated on checkpoint inhibitors who have been previously maintained on palliative corticosteroids?
In the article that accompanies this podcast, Arbour and colleagues present important insights into this question. This is a retrospective analysis of 640 patients with advanced non-small-cell lung cancer with immunotherapy-naïve disease treated with checkpoint inhibitors at the Memorial Sloan Kettering and Goustave Roussy Cancer Centers from 2011-2017. Data regarding steroid dose, mode of administration, and indications for use were collected along with efficacy of checkpoint inhibitors as assessed by an independent team of radiologists using the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. At the time of checkpoint inhibitor initiation, 90/640 (14%) of patients were receiving the equivalent of Prednisone 10mg/day or more; an additional 17 patients (or 3% of the total study population) were receiving <10mg of Prednisone/day and were analyzed as part of the “non-steroid” group in subsequent assessments.
The most common indications for steroid use reflect those seen in routine clinical practice: dyspnea/respiratory symptoms (33%), fatigue (21%), and brain metastases (19%). Clinicopathologic characteristics between the steroid and no steroid groups were generally well balanced. As might be expected, however, there were more patients with poor performance status and history of brain metastases in the steroid group.
Amongst the Sloan Kettering cohort, steroid use of >/= 10mg/day was associated with statistically significant reductions in objective response rates (6% vs. 19%), median progression free survival (1.9 vs. 2.6 months), and median overall survival (5.4 vs. 12.1 months) in the steroid vs. no steroid groups, respectively. Similar findings were seen in the Gustave Roussy cohort. Pooled data from both cohorts confirmed the same detrimental impact of baseline steroid use >/= 10mg/day. On multivariable analysis, the findings of inferior response rates and worse progression free and overall survival in those receiving Prednisone >/= 10mg/day were upheld. Further, a similar decrement in checkpoint inhibitor efficacy was observed in patients receiving >/= 20mg/day of Prednisone as compared to those receiving lower doses of 10-19 mg/day.
Patients receiving and then discontinuing steroids 1-30 days prior to initiation of immune checkpoint inhibitors had intermediate progression free and overall survival as compared to those receiving steroids on the same day and those not receiving any steroids within 30 days of starting a checkpoint inhibitor.
The authors do additionally note that 6 patients achieved a partial response to checkpoint inhibitor therapy, despite steroid use at treatment initiation and with durable responses for most of these patients lasting >15 months.
Notable strengths of this investigation are the analysis of real world clinical scenarios where steroid use is often a pragmatic reality and the availability of highly relevant details regarding steroid dose, indication, duration, and initiation/discontinuation relative to checkpoint inhibitor therapy. Outcomes with immune checkpoint inhibitors seen in patients receiving <10mg/day of Prednisone are concordant with what has been seen with these agents in day-to-day practice and in the published literature to date.
Limitations of the study include its retrospective nature and overall low incidence of steroid use. With the advent of combined chemoimmunotherapy regimens, finite duration steroid use concurrently with each cycle of treatment for prevention of chemotherapy-related nausea and rash has become standard. However, as patients in this study were treated with immune checkpoint inhibitor monotherapy, the impact of these finite, repeating courses of steroids is not explored here. Further, though multivariable analysis confirmed the independent association between steroid use and inferior outcomes with checkpoint inhibitors, a mechanistic explanation for this has not yet been elucidated. Lastly, immunologic profiling results (i.e. tumor PD-L1 status and tumor mutation burden) were not reported.
In my own practice and consistent with the observations made by Arbour and colleagues in this analysis, I try to minimize or avoid the use of daily palliative steroids for patients in whom immune checkpoint inhibitor therapy is shortly anticipated. Whenever possible, steroid-sparing alternatives should be considered and used. However, in those patients who require steroids and without reasonable alternatives (i.e. those with brain metastases and intracranial edema), medically necessary steroid therapy should not be deferred or discontinued until it is clinically safe to do so.
Arbour and colleagues should be congratulated on their important contribution to this highly relevant clinical issue.
This concludes this JCO Podcast. Thank you for listening.