Guidelines for the design and conduct of human clinical trials on ingestion-time differences - chronopharmacology and chronotherapy - of hypertension medications

Abstract

Current hypertension guidelines fail to provide a recommendation on when-to-treat, thus disregarding relevant circadian rhythms that regulate blood pressure (BP) level and 24 h patterning and medication pharmacokinetics and pharmacodynamics. The ideal purpose of ingestion-time (chronopharmacology, i.e. biological rhythm-dependent effects on the kinetics and dynamics of medications, and chronotherapy, i.e. the timing of pharmaceutical and other treatments to optimize efficacy and safety) trials should be to explore the potential impact of endogenous circadian rhythms on the effects of medications. Such investigations and outcome trials mandate adherence to the basic standards of human chronobiology research. In-depth review of the more than 150 human hypertension pharmacology and therapeutic trials published since 1974 that address the differential impact of upon-waking/morning versus at-bedtime/evening schedule of treatment reveals diverse protocols of sometimes suboptimal or defective design and conduct. Many have been "time-of-day," i.e. morning versus evening, rather than circadian-time-based, and some relied on wake-time office BP rather than around-the-clock ambulatory BP measurements (ABPM). Additionally, most past studies have been of too small sample size and thus statistically underpowered. As of yet, there has been no consensual agreement on the proper design, methods and conduct of such trials. This Position Statement recommends ingestion-time hypertension trials to follow minimum guidelines: (i) Recruitment of participants should be restricted to hypertensive individuals diagnosed according to ABPM diagnostic thresholds and of a comparable activity/sleep routine. (ii) Tested treatment-times should be selected according to internal biological time, expressed by the awakening and bed times of the sleep/wake cycle. (iii) ABPM should be the primary or sole method of BP assessment. (iv) The minimum-required features for analysis of the ABPM-determined 24 h BP pattern ought to be the asleep (not "nighttime") BP mean and sleep-time relative BP decline, calculated in reference to the activity/rest cycle per individual. (v) ABPM-obtained BP means should be derived by the so-called adjusted calculation procedure, not by inaccurate arithmetic averages. (vi) ABPM should be performed with validated and calibrated devices at least hourly throughout two or more consecutive 24 h periods (48 h in total) to achieve the highest reproducibility of mean wake-time, sleep-time and 48 h BP values plus the reliable classification of dipping status. (vii) Calculation of minimum required sample size in adherence with proper statistical methods must be provided. (viii) Hypertension chronopharmacology and chronotherapy trials should preferably be randomized double-blind, randomized open-label with blinded-endpoint, or crossover in design, the latter with sufficient washout period between tested treatment-time regimens.

Keywords: Ambulatory blood pressure monitoring; blood pressure dipping; hypertension chronotherapy; hypertension medications; sleep-time blood pressure; trial design recommendations and guidelines.

Figure 1.

24 h SBP profile (dashed thick lines) of four male subjects, respectively, dipper (upper left), extreme-dipper (upper right), nondipper (lower left) and riser pattern (lower right), having the same 24 h SBP mean of 124.5 mmHg. The SBP profiles are plotted with respect to circadian time-specified tolerance limits for normal SBP (continuous thin lines), calculated from a reference population of normotensive individuals of comparable rest-activity cycle and male sex.

Figure 2.

Adjusted HR of CVD events in the Hygia Project cohort entailing 21,963 individuals. Participants were categorized into four nonoverlapping groups according to the level (normal or elevated) of the ABPM-derived awake and asleep SBP/DBP means. The ABPM-derived awake SBP/DBP means were considered normal if <135/85 mmHg and elevated otherwise. The asleep SBP/DBP means were considered normal if <120/70 mmHg and elevated otherwise. Adjustments were applied, if significant, for sex, age, diabetes, CKD and history of previous major CVD event.

Figure 3.

Adjusted HR of CVD events in the Hygia Project cohort entailing 21,963 individuals. Participants were categorized into four nonoverlapping groups according to the level (normal or elevated) of the ABPM-derived asleep SBP/DBP means and extent of the sleep-time relative SBP decline. The ABPM-derived asleep SBP/DBP means were considered normal if <120/70 mmHg and elevated otherwise. Participants were designated as dipper when the sleep-time-relative SBP decline was ≥10% and as nondippers when <10%, using data sampled at 20–30 min intervals by ABPM for 48 consecutive hours. Adjustments were applied, if significant, for sex, age, diabetes, CKD and history of previous major CVD event.

Figure 4.

24 h SBP profile (dashed thick lines) of a normotensive dipper man plotted with respect to circadian time-specified tolerance limits for normal SBP (continuous thin lines) calculated from a reference population of normotensive individuals of comparable rest-activity cycle and male sex. The same ABP data are represented as a function of clock time (left) and biological time, i.e. hours from bedtime (right).

Figure 5.

Clock hour distribution of wake-up (left) and bed times (right) at the baseline ABPM evaluation of the MAPEC Study cohort entailing 3,344 individuals.

Figure 6.

Left: Distribution of duration of sleep determined by wrist actigraphy at the baseline ABPM evaluation of the MAPEC Study cohort entailing 3,344 individuals. Right: Correlation between duration of sleep and age for the same MAPEC Study clinical cohort.

Figure 7.

Difference of ~4.4% in the calculated sleep-time relative decline for SBP versus DBP of the same arterial hypertensive individual assessed by around-the-clock ABPM, demonstrating the differential and misleading sleep-time relative decline dipping classification when improperly based on DBP measurements (resulting in conclusion of dipper categorization) rather than when properly based on SBP measurements (resulting in conclusion in nondipper categorization).

Figure 8.

Bland-Altman plots assessing agreement of the asleep SBP mean calculated utilizing the original data sampled by ABPM every 20–30 min for 48 consecutive hours versus that estimated from the (i) modified time series composed of data sampled every 1 h for 48 consecutive hours (top) and (ii) original sampling rate of every 20–30 min for the first 24 h, only (bottom). Dotted horizontal line of each graph represents the average of the differences across the entire studied population. Dashed lines represent the values of the average difference ±1.96SD (assumingly containing 95% of the individual values).