Prior MI or stroke may predict diuretic use for HTN management

A prespecified secondary analysis of the Diuretic Comparison Project (DCP) shows that when deciding between chlorthalidone (CTD) and hydrochlorothiazide (HCTZ) for the management of hypertension (HTN), a history of myocardial infarction (MI) or stroke may have to be considered.

Results of the primary DCP analysis showed no difference between CTD and HCTZ on the evaluated outcomes when low doses were used. [JAMA Intern Med 2020;180:542-551] “But in this secondary analysis, there was a significant qualitative interaction between treatment assignment and the presence or absence of MI or stroke at baseline with respect to the primary outcome,” the researchers said.

“Participants with prior MI or stroke randomized to CTD had lower risk of MACE* and noncancer death than those receiving HCTZ. In participants without prior MI or stroke, the outcomes were not significantly different,” they continued.

DCP is a pragmatic randomized trial done within 72 participating Veterans Affairs healthcare systems from June 2016 to June 2021. The trial looked at 13,523 HTN patients (median age 72 years, 96.8 percent men) on HCTZ at baseline. They were randomized to either continue HCTZ or switch to CTD at pharmacologically comparable doses. Of these, 1,455 did have prior MI or stroke at baseline (n=733 on CTD; 722 on HCTZ [arm A]) and 12,068 did not (n=6,023 and 6,045, respectively [arm B]). [JAMA Netw Open 2024;7:e2411081]

In arm A, the incidence of the composite primary outcome was lower with CTD vs HCTZ (14.3 percent vs 19.4 percent; absolute risk reduction, 5.1 percent; adjusted hazard ratio [aHR], 0.74; p=0.02). The researchers attributed this effect to a reduced risk of acute heart failure (HF; 4.6 percent vs 7.2 percent; aHR, 0.64; p=0.054). The opposite was true in arm B (9.9 percent vs 8.9 percent; aHR, 1.12; p=0.054). A comparison between arms yielded an adjusted p-value of 0.004 for interaction.

Hypokalaemia, AEs

Thiazide diuretics are the currently recommended first-line agents for HTN. [Hypertension 2018;71:e13-e115] However, HCTZ is more prescribed in the US due to AEs tied to CTD (eg, hypokalaemia, defined as potassium level <3.1 mEq/L). [https://data.cms.gov/summary-statistics-on-use-and-payments/medicare-medicaid-spending-by-drug/medicare-part-d-spending-by-drug/data, accessed July 19, 2024]

Hypokalaemia is tied to an increased risk of adverse CV outcomes, and this link has been a limiting factor for use of high-dose diuretics. [Am J Med 2018;131:318.e9-318.e19; Eur Heart J Cardiovasc Pharmacother 2021;7:557-567; Eur Heart J 2018;39:1535-1542]

In arm A, the rates of nadir potassium level <3.1 mEq/L were similar between CTD and HCTZ (5.9 percent vs 5.1 percent; p=0.57), as were the incidences of hospitalization for hypokalaemia (1.9 percent vs 2.2 percent; p=0.72).

“A hypothesis for the [difference] is that the benefits of CTD were unmasked by the prevention of a nadir potassium level <3.1 mEq/L in the CTD [vs] HCTZ group and/or by the greater use of supplemental potassium in the CTD group. There may, however, be an alternative explanation,” the researchers pointed out.

In arm B, the rates of the two outcomes noted above were higher with CTD vs HCTZ (4.9 percent vs 3.4 percent; p<0.001 and 1.4 percent vs 0.9 percent; p=0.02, respectively). The researchers noted that the higher incidence of a nadir potassium level <3.1 mEq/L in the CTD group may have driven the higher rate of hospitalization for hypokalaemia.

The incidence of adverse events (AEs) was generally higher in arm A vs B (39.4 percent vs 28.1 percent; p<0.001), which may have been driven by the higher incidences of all-cause mortality and hospitalizations in arm A. In arm B, the risk of nonserious AEs was higher with CTD vs HCTZ (20.4 percent vs 17.5 percent; p<0.001), primarily due to hypokalaemia (5.9 percent vs 4.2 percent; p<0.001).

Real effect or chance finding?

“[The findings] suggest a qualitative interaction. This type of interaction is less common than a quantitative interaction, in which the treatment is beneficial (or harmful) in all subgroups, but the magnitude of effect varies among subgroups,” the researchers explained. “It is typically difficult to interpret … qualitative interactions to determine whether the effect is real or simply from chance.”

Formal tests for interaction have been done, yet the significance remained after adjustments for multiplicity have been made. The suggestion of a qualitative interaction by baseline MI or stroke status was also observed for both the primary endpoint and its individual components, including acute HF, MI, noncancer-related death, and total mortality.

Furthermore, hypokalaemic AEs differed between arms A and B. “[Hence,] it is possible that the observed interaction was due to chance alone,” they said.

It is not clear whether the benefits observed in arm A were due to baseline risk, hypokalaemia prevention, or the added potassium supplementation. “We can only conclude that the combination of all three was associated with an improved outcome in participants taking CTD compared with those on HCTZ. There may be other factors not explored here that may have led to the observed interaction,” the researchers said.

“Accounting for this subgroup knowledge when treating patients, CTD may benefit those with prior MI or stroke,” the researchers concluded.

Whether the findings are “simply a chance finding or are real and potentially mediated through either hypokalaemia prevention or potassium supplementation” in this subgroup needs to be substantiated in future in-depth investigations, they said.