Cardiovascular disease modifies blood pressure response to mandibular advancement therapy versus CPAP in obstructive sleep apnea: a prespecified subgroup analysis of the CRESCENT trial

Background

Patients with atherosclerotic cardiovascular disease (ASCVD) benefit from blood pressure (BP) control. Obstructive sleep apnea (OSA) is an important cause of hypertension. In the CRESCENT trial, mandibular advancement device (MAD) was non-inferior to CPAP in reducing 24-hour mean BP. In this prespecified analysis, we evaluated whether the relative effectiveness of MAD vs CPAP differs between patients with and without ASCVD.

Methods

CRESCENT is an investigator-initiated, non-inferiority trial that randomized 220 patients with hypertension and elevated cardiovascular risk to MAD or CPAP (1:1). All the patients were diagnosed to have OSA (AHI ≥15 event/hour) The primary endpoint was change in mean arterial BP from baseline to 6 months, using 24-hour ambulatory BP monitoring. This prespecified subgroup analysis evaluated BP responses stratified by baseline ASCVD status (coronary artery disease and/or stroke). Within-group changes were assessed using paired analyses, and between-group differences using ANCOVA adjusted for baseline BP. Treatment × ASCVD interaction terms were used to assess effect modification.

Results

Of the 220 participants, 145 (66%) had ASCVD and 75 (34%) did not. The two groups were similar in age (61 [56–65] vs 62 [55–65] years, P=0.981) and baseline AHI (35.3 [23.8–51.5] vs 41.4 [30.1–56.9] events/hour, P=0.060). Compared with the non-ASCVD group, the ASCVD group were more likely to be male (93% vs 71%, P < 0.001) and had lower BMI (27.1 [25.2–29.7] vs 28.6 [26.4–31.7] kg/m², P=0.004).

In the ASCVD group, 71 patients were randomized to MAD and 74 to CPAP, with median nightly use of 5.6 and 5.0 hours, respectively. In the MAD group, reductions in ambulatory BP from baseline to 6 months were observed, including 24-hour mean BP (95 to 93 mmHg, P=0.011), 24-hour systolic BP (124 to 121 mmHg, P=0.025), and asleep mean BP (91 to 90 mmHg, P=0.017). In contrast, no significant BP reductions were observed in the CPAP group. In between-group analyses, MAD was associated with greater BP reductions than CPAP, including 24-hour mean BP (−2.50 mmHg; 95% CI −4.90 to −0.11; P=0.028), 24-hour systolic BP (−3.41 mmHg; 95% CI −6.56 to −0.37; P=0.040), asleep mean BP (−3.97 mmHg; 95% CI −7.31 to −0.62; P=0.021), and asleep systolic BP (−5.13 mmHg; 95% CI −9.36 to −0.91; P=0.018). In the non-ASCVD group, 39 patients were randomized to MAD and 36 to CPAP, with median nightly use of 5.5 and 4.8 hours, respectively. No significant BP reductions were observed in either treatment, and no significant between-group differences were found. A significant treatment × ASCVD interaction was observed for asleep systolic BP (β = −6.85, P=0.034).

Conclusions

BP-lowering effects of OSA treatment differ according to ASCVD status. In patients with OSA and established ASCVD, MAD, compared with CPAP, achieved greater reductions in BP. This analysis provides insights that support a personalized approach on BP control.Figure. 1