Evaluation of apremilast in chronic pruritus of unknown origin: A proof-of-concept, phase 2a, open-label, single-arm clinical trial.

School of Medicine, Saint Louis University, St. Louis, Missouri Division of Dermatology, Department of Medicine, Washington University School of Medicine, St. Louis, Missouri Center for the Study of Itch and Sensory Disorders, Washington University School of Medicine, St. Louis, Missouri Department of Anesthesiology, Washington University School of Medicine, St. Louis, Missouri Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri


| Study design and treatment
. Key exclusion criteria included chronic pruritus due to a primary dermatologic or other underlying medical disorder, topical treatments within 1 week of baseline, systemic immunomodulating agents within 4 weeks of baseline, and prior treatment with apremilast. The following medications were prohibited during the study: topical and oral steroids, leukotriene inhibitors, calcineurin inhibitors, allergen immunotherapy, phototherapy, tanning beds, live vaccines, and CYP450 inducers.
While there was no formally stated statistically powered a priori hypothesis for this study, the target enrollment of n = 10 subjects was based on the relative uniformity of the disease severity of the population (ie, severe itch only), and on the fact that we have previously observed relevant differences in populations of CPUO patients with only n = 5 to 6 patients per group in response to treatment. 4,5 Ten patients with CPUO were enrolled and received 16 weeks of treatment with apremilast 30 mg tablet twice daily (BID).

| Assessment
The primary endpoint analysis of this study was absolute reduction in 24-hour and 1-week NRS itch score at week 16 from baseline in patients who received apremilast 30 mg BID for 16 weeks. We chose timepoint with agents employed to treat atopic dermatitis. 6 The key secondary endpoint was absolute reduction in Dermatology Life Quality Index (DLQI) at week 16 from baseline. Safety and tolerability were assessed by monitoring the type, frequency, duration, and severity of adverse events (AEs) throughout the duration of the study by nonsystematic assessment and self-reporting by patients at each study visit. The NRS itch score is a single-question assessment tool with a scale of 0 (no itch) to 10 (worst imaginable itch). 7 Patients reported their worst level of itch over the prior 24-hour and 1-week period at each study visit. Change from baseline in DLQI was also measured to assess patient quality of life (QoL) improvement. 8 Patients were assessed at baseline and weeks 2, 4, 8, 10, 12, and 16 for these endpoints as well as for vital signs including respiratory rate, pulse, blood pressure, and temperature, and a targeted symptom-directed physical exam was conducted. Laboratory tests were performed at baseline and at week 16, which included a complete blood count and a comprehensive metabolic profile.

| Statistics
All patients were included in the intent-to-treat efficacy analysis. Given the unexpectedly high dropout rate and inability to draw any systematic conclusions (see below), we performed a last observation carried forward (LOCF) to week 16 analysis with missing data inferred for the 24-hour and 1-week NRS itch scores and DLQI score, in a post hoc manner. All efficacy data points are shown at each individual assessment. Differences in NRS and DLQI scores were assessed via Wilcoxon Signed-Rank nonparametric tests for non-normally distributed data. Differences were considered statistically significant if a two-tailed P < .05. Statistical analysis was performed using GraphPad Prism 8.0 software.   Table 1.

| Efficacy
The data were analyzed in an intent-to-treat manner with key primary and secondary endpoints measured as an absolute reduction in NRS itch and DLQI scores, respectively, at week 16 from baseline. In total, 3/10 patients completed the study, which did not allow for meaningful intent-to-treat statistical analysis. As an alternative approach, we undertook a post hoc LOCF analysis by carrying forward to week 16.
By this analysis, we observed no statistically significant reduction in 24-hour or 1-week NRS itch scores at week 16 ( Figures 1A and S1).
Further, we similarly observed no significant reduction in DLQI at week 16 from baseline ( Figure 1B).
Given that 70% of the patients did not complete the study, we sought to examine the reasons for patient dropout. Strikingly, 50% of the patients dropped out due to experiencing an AE. One additional patient opted to discontinue the study due to resolution of itch symptoms, while another desired to use a prohibited medication for itch relief. Two of the three patients who completed the study demonstrated absolute reduction in 24-hour NRS itch scores from 8 and 9.5 to 0 ( Figure 1A). Further, these same subjects also demonstrated a reduction in DLQI score from 26 and 10 to 4 and 0, respectively ( Figure 1B). Lastly, one patient completed the study who did not demonstrate a reduction in NRS itch score but a minimal reduction in DLQI score from 5 to 3.

| Safety
For this cohort of patients, apremilast was not well tolerated, and several mild to moderate AEs were reported. There were 11 total AEs reported by 5 (50%) patients in the study; all were considered treatment related and resembled AEs reported in the prescribing information for apremilast. The specific AEs and incidences are described in Table 2. Gastrointestinal (GI) dysfunction was reported by 5 (50%) patients, with nausea and diarrhea being the most common. Complaints of nervous system dysfunction were also reported by 2 (20%) patients, which included headaches and presyncope. All AEs were mild to moderate in terms of severity and ceased within 48 hours of stopping the medication. All five patients who experienced AEs dropped out of the study due to the AEs  were no clinically significant changes as determined by the PI in vital signs, physical exam, or laboratory parameters for any of the patients during the course of the study.

| DISCUSSION
In the current study, apremilast demonstrated poor tolerability in this population of patients with CPUO, which resulted in high dropout.
Because of this, a meaningful intent-to-treat statistical analysis was not possible, and we were unable to evaluate both the primary and secondary endpoints of efficacy based on NRS itch and DLQI scores.
Therefore, we performed a post-hoc LOCF analysis, and we did not observe any inferred efficacy. Collectively, we conclude that due to the unexpectedly high rate of AEs in this population, when designing future studies, power analyses would benefit from accounting for a high dropout rate.
Our statistical analysis plan assumed a mean NRS itch score of 8.8 with a SD of 1.1 based on a sampling of patients from our specialty itch clinic. Based on these values and our recent experience in treating small cohorts of patients with CPUO, 4,5 our target sample size was n = 10. Unfortunately, we did not anticipate such a high dropout rate. Notwithstanding this, two of the three patients who completed the study did demonstrate marked reduction of itch from severe (NRS itch score of 8 and 9.5) to no itch (NRS itch score of 0). This was associated with respective improvement in QoL as measured by the DLQI.
However, given that this is not a placebo-controlled study, we cannot determine whether this is due to a placebo effect or even a direct response to apremilast. As the placebo response in CPUO was recently reported to be surprisingly high in phase 2 clinical trials with other agents such as serlopitant (NCT03841331), studies in this condition will likely require much larger sample sizes than originally anticipated in the design of this study.