Potential beneficial effect of low‐dose danazol in combination with renin–angiotensin system inhibitors in diabetic macular oedema

Although intra-ocular antivascular endothelial growth factor (VEGF) injections are the standard of care for patients with diabetic macular oedema (DME), some fail this treatment. This begets the need for alternative therapies for these patients. Multiple basic science studies have demonstrated low concentrations of danazol significantly enhance endothelial barrier function and lower vascular permeability, making it a prime candidate therapy for DME (Thomas et al. 2012; Thomas et al. 2007). The results of a subsequent Phase 2 randomized controlled trial (RCT) indicated danazol was safe and its efficacy was dependent on body mass index (BMI) (Yau et al. 2012). We are now reporting an interesting observation from a second Phase 2 RCT, where we have identified a synergistic interaction between danazol and renin–angiotensin system (RAS) inhibitors. These data come from a multicenter, double-blind, RCT evaluating the efficacy and safety of two low doses of oral danazol (0.5 mg/BMI/day or 1 mg/BMI/day) in adults with DME, compared to placebo. Twenty-one ophthalmology clinics across the United States were included in the trial. The conduct of this trial was in accordance with the Declaration of Helsinki. The primary and secondary outcomes were change from baseline to the end of the double-blind phase in best-corrected visual acuity (BCVA) and central retinal thickness (CRT, measured by optical coherence tomography, μm). Three observations (one in each group) were excluded due to being more than three standard deviations away from the group mean. Overall, there were 425 eyes included in the trial, equally distributed amongst the three treatment groups. The groups were similar at baseline, with 53% having failed a previous anti-VEGF therapy. We confirmed the interaction of danazol with BMI. The second BMI quartile (27.7–31.3 BMI) had the most significant response to danazol; all results are presented for this quartile (n = 107). More interestingly, we also discovered a second, synergistic interaction with RAS inhibitors (Fig. 1). Seventy per cent of eyes in the second BMI quartile were on a concomitant RAS inhibitor. The 0.5 mg danazol group showed a significant 5.8-letter increase in BCVA from baseline at 12 weeks (p < 0.001); this was significantly larger (p = 0.01) than the 0.8-letter, non-significant increase from baseline observed in the placebo group. Concomitant RAS inhibitor use was not shown to modify the effect of 1 mg danazol/BMI on BCVA. In the remaining 30% of eyes not on a concomitant RAS inhibitor, all treatment groups did not show a significant increase from baseline in BCVA. It was only when the 0.5 mg danazol/BMI treatment was combined with RAS inhibitors that significant improvements from baseline were seen in both BCVA and CRT at every visit (Fig. 1). In subjects on a concomitant RAS inhibitor, both the 0.5 mg and 1 mg danazol/BMI treatment groups showed significant decreases from baseline in CRT (−49.0 μm, p = 0.01; −39.3 μm, p = 0.03), and the decrease in the 0.5 mg danazol/BMI group was significantly greater than the +10.2 μm increase seen in the placebo group (p = 0.03). It is important to note the RAS inhibitor interaction was examined across all BMI quartiles. Only the second BMI quartile showed a significant interaction, again suggesting the 0.5 mg/BMI/day dosage was optimal in this BMI range. The combination of oral RAS inhibitors and oral low-dose danazol might be a painless, safe and efficacious therapy for a subset of patients suffering from DME. Combination therapies addressing multiple disease pathways are likely to be more resilient to physiological compensatory mechanisms than therapies addressing single pathways. With an estimated one to two million people in the United States suffering from DME and on a path towards permanent blindness, alternate therapies to intra-ocular injections must be explored, particularly for patients refractory to available therapies.(Das et al. 2015; Bar-Or et al. 2012).