Differences in safety profiles of panobinostat, ixazomib, elotuzumab, and daratumumab in real-world settings versus randomized controlled clinical trials

Multiple myeloma (MM) is a hematological cancer of plasma cells that mainly affects older populations, with a median age at diagnosis of 69 years. The introduction of novel therapies over the past couple of decades has improved patient outcomes and overall survival, but MM remains incurable.

In 2015, four new therapies were approved by the U.S. Food and Drug Administration (FDA) for the treatment of relapsed MM: panobinostat, ixazomib, elotuzumab, and daratumumab. Although there have been numerous systematic reviews, meta-analyses, and review papers that assess the efficacy and safety, including the side effects of these therapies based on the data from their randomized controlled clinical trials (RCTs) , the safety profiles of these treatments in real-world settings is lacking. Eric P. Borrelli and Conor G. McGladrigan published in Journal of Oncology Pharmacy Practice a comparison of safety profiles of the four new agents in the real-world settings versus RCTs.¹

Method

 The dataset was largely from the FDA Adverse Event Reporting System (FAERS) and included data from the randomized clinical trials that led to the approval of the four therapies. Any reports of adverse events (AE) were collected and a categorized into the ten most frequently reported events. This analysis included all reactions reported from October 2015–December 2017, regardless of grade and severity, and excluded any non-specific reporting.

 Drug approvals 

• Panobinostat, a potent oral pan-deacetylase inhibitor, was approved by the FDA in February 2015 to be taken in combination with bortezomib and dexamethasone for the treatment of patients with MM who have received ≥ 2 prior lines of therapy. The approval was based on the results from a subgroup of 193 patients who participated in the PANORAMA-1 (NCT01023308) trial²

• Ixazomib, an oral proteasome inhibitor, was approved by the FDA in November 2015 to be taken in combination with lenalidomide and dexamethasone for the treatment of patients with MM who have failed ≤ 1 prior treatment. The approval was based on the interim results from the pivotal phase III TOURMALINE-MM1 (NCT01564537) trial³
• Elotuzumab, a SLAMF7/CS1 targeting monoclonal antibody, was approved by the FDA in November 2015 to be taken in combination with lenalidomide and dexamethasone for the treatment of patients with MM who received 1–3 prior treatments. The approval was based on data from the phase III ELOQUENT-2 (NCT01239797) study⁴
• Daratumumab, an anti-CD38 monoclonal antibody, was approved by the FDA in November 2015 for the treatment of patients with MM who have received ≥ 3 prior lines of therapy. The approval was based on the results from the phase II SIRIUS (MMY2002; NCT01985126) study⁵

Results

 There were only a few adverse drug reactions (ADRs) among the ten most frequent AEs that were reported in both randomized controlled trials and FAERS for each agent (Table 1). These were as follows:

• Daratumumab: cough and back pain
• Elotuzumab: anemia and diarrhea
• Ixazomib: diarrhea, constipation, fatigue, nausea, and peripheral neuropathy
• Panobinostat: diarrhoea, fatigue, nausea, and constipation

 Table 1. Comparison of the ten most frequent AEs reported in clinical trials versus data extrapolated from the FAERS ¹

• The reporting odds ratio (ROR) is a measure of association between an exposure (e.g., RCT vs real-world setting) and an outcome (e.g., ADRs). The ROR represents the odds that an outcome will ensue given a particular exposure versus the odds of the outcome without that exposure⁶
• For daratumumab, the risk estimates for the top ten reported ADRs in the FAERS were statistically significantly increased compared with all other agents (Table 2)
• For elotuzumab, the risk estimates for the top ten reported ADRs in the FAERS were significantly higher in comparison with all the other treatments, with exception of diarrhea (ROR, 1.54; 95% CI, 0.73–3.25) and rash (ROR, 2.04; 95% CI, 0.84–4.93)
• For ixazomib, the risk estimate for abdominal pain the had a statistically significant decrease in comparison with all the other treatments (ROR, 0.58; 95% CI, 0.40–0.85), but all the other ADRs increased in risk
• For panobinostat, there was a statistically significant increase in the risk estimate of all of the top ten ADRs in the FAERS when compared with all other therapies

 Table 2. Most frequently reported AEs from the FAERS with a statistically significant higher ROR¹

Elotuzumab was associated with the development of deep vein thrombosis; this finding is reinforced by other publications.¹ However, as elotuzumab was only approved as part of a triplet combination with lenalidomide and dexamethasone, during this study, the deep vein thrombosis reaction could be from lenalidomide, which carries a black box warning for this AE.

Although death was reported frequently and was a statically significant ADR for all four therapies, it may be a consequence of the disease itself or related to the high-risk patient population that failed up to three lines of therapy, the indication setting for these treatments.

Conclusion

This study outlined the differences in safety of panobinostat, ixazomib, elotuzumab, and daratumumab in real-world settings versus RCTs. The findings show a difference in risk estimates between the two settings and highlight the importance of monitoring pharmacovigilance and assessing for any additional safety signals over time. Ideally, these should be performed within 2 years of the FDA approval because beyond this there is a significant decrease in ADR reporting.

It is important to note that panobinostat has a black box warning for severe diarrhea and cardiac ischemic events. The proven increase in the FAERS reporting for ADRs in relation to a black box warning might explain the higher incidence of diarrhea with panobinostat.

A limitation of the study was the reporting in the FAERS database of events such as ‘adverse event’, ‘adverse reaction’, and ‘hospitalization’, which were included in the ROR calculations and may bias results as the causality of the ADR and the treatment were not clear.  Additionally, it is impossible to know the real incidence of the reported ADRs since the total number of patients taking each medication is unknown.