Effect of Dual RAAS Blockade and Intensive BP Lowering on Risk of End-Stage Kidney Disease and Death in Autosomal Dominant Polycystic Kidney Disease: Long-Term Follow-Up of the HALT-PKD Trials

To the editor:

Autosomal dominant polycystic kidney disease (ADPKD) is a slowly progressive kidney disease, and the effect of interventions on hard outcomes such as onset of end-stage kidney disease (ESKD) may be challenging to ascertain given the short duration of most clinical trials, where the statistical power may be low.¹ We previously demonstrated a “legacy effect” of target BP interventions in patients with CKD in the Modification of Diet in Renal Disease and African American Study of Kidney Disease and Hypertension trials (which had null effects during the parent trial) after two decades of post-trial follow-up facilitated by linkage to the US Renal Data System (USRDS) and National Death Index (NDI).^2–4 Our objective was to determine the risk of ESKD, and secondarily, ESKD or death among patients previously randomized to receive intensive BP control or dual renin-angiotensin-aldosterone system (RAAS) blockade in the HALT PKD trials through linkage to national registries.

The design and results of the HALT PKD trials have been previously described (Items S1–S2).^5,6 Briefly, the HALT-PKD Trials previously randomly assigned patients with ADPKD to either dual RAAS blockade (ACE inhibitor [ACEi] plus ARB) versus ACE inhibitor alone in two related trials: HALT A (which enrolled patients with eGFR>60 mL/min/1.73 m²) and HALT B (which enrolled patients with eGFR 25–60 mL/min/1.73 m²).^5,6 Additionally, patients in HALT A trial were also randomized in 2×2 factorial design to strict (95/60 to 110/75mmHg) versus usual (120/70 to 130/80 mmHg) BP control, while all HALT B trial participants were treated to a target BP of ≤130/80 mmHg. To provide longer-term follow-up for kidney and mortality endpoints, we linked former enrollees from five HALT-PKD study sites to the USRDS (the national kidney registry)⁷ and NDI to ascertain ESKD and vital status through 12/31/2020 as in prior work.^2,3

Unadjusted and adjusted cause-specific hazards models were used to examine the effect of each intervention in the two trials separately, and also in aggregated analysis (for dual RAAS blockade intervention), allowing baseline hazard to vary by study site (see Item S1). Time in all models began at randomization and follow-up ended at ESKD, death, or December 31, 2020 (for sites with linked data) or end of trial closure (for the two sites who did not participate in long-term linkage).

Baseline characteristics of 1043 trial participants included for analysis are shown in Table S1 by randomized intervention and separately by linkage status in Tables S2–S3. During median follow-up of 10.6 years (IQR 6, 12.9), 343 patients developed ESKD and 26 deaths occurred before ESKD onset (Table S4).

In aggregated analysis of both HALT A and B trials, there was no statistically significant effect of dual RAAS blockade on the risk of ESKD in unadjusted (HR 0.92; 95% CI 0.74–1.13) or adjusted analyses (HR 0.84; 95% CI 0.67–1.04). The confidence interval admits estimates of up to a 33% reduction in the hazard of ESKD and limits the potential for harm to 4%, which suggest a higher likelihood of benefit than harm compared to estimates reported previously in the parent trials.^5,6

In HALT B trial, the effect of dual RAAS blockade trended towards benefit in unadjusted (HR 0.82; 95% CI 0.65–1.04) and adjusted analysis (HR 0.82; 95% CI 0.65–1.05), but this trend was not observed in HALT A trial. Similar findings were noted when we used Fine-Gray models, or when we examined the composite outcome of ESKD or death (Table 1 and Figure 1).

Table 1.

Risk of ESKD and the composite risk of ESKD or death in Cox proportional hazards and Fine-Gray models.

Cohort	Intervention	Outcome**	Unadjusted			Adjusted*
			(s)HR	95% CI	p-value	(s)HR	95% CI	p-value
HALT A & B (N=1,043)	Lisinopril+Telmisartan (vs. Lisinopril+Placebo)	ESKD	0.92	[0.74–1.13]	0.42	0.84	[0.67–1.04]	0.10
		ESKD or Death	0.93	[0.76–1.14]	0.50	0.85	[0.69–1.04]	0.12
		ESKD with Death as a Competing Risk	0.93	[0.75–1.15]	0.49	0.88	[0.68–1.13]	0.30
HALT A (N=558)	Lisinopril+Telmisartan (vs. Lisinopril+Placebo)	ESKD	1.14	[0.68–1.90]	0.62	1.08	[0.62–1.86]	0.79
		ESKD or Death	1.16	[0.73–1.85]	0.52	1.02	[0.63–1.67]	0.93
		ESKD with Death as a Competing Risk	1.18	[0.71–1.97]	0.52	0.97	[0.56–1.70]	0.92
	Low BP Control (vs. Standard)	ESKD	0.84	[0.50–1.40]	0.51	0.79	[0.46–1.34]	0.38
		ESKD or Death	0.78	[0.49–1.25]	0.30	0.74	[0.45–1.20]	0.22
		ESKD with Death as a Competing Risk	0.86	[0.52–1.44]	0.57	0.76	[0.44–1.34]	0.35
HALT B (N=485)	Lisinopril+Telmisartan (vs. Lisinopril+Placebo)	ESKD	0.82	[0.65–1.04]	0.10	0.82	[0.65–1.05]	0.11
		ESKD or Death	0.83	[0.66–1.05]	0.12	0.83	[0.66–1.05]	0.12
		ESKD with Death as a Competing Risk	0.83	[0.66–1.04]	0.11	0.87	[0.66–1.14]	0.32

^*Adjusted for age, sex, race/ethnicity, baseline eGFR, logarithm of albuminuria, body mass index, systolic blood pressure, BP target strategy (all patients in HALT B were treated as having standard BP control; HALT A participants were classified based on randomized assignment). We additionally adjusted models for baseline total kidney volume in the HALT A trial.

^**Cause-specific hazards models allowing baseline hazards to vary by site.

Figure 1.

Cumulative incidence curves accounting for the competing risk of death in each trial by a) dual RAAS blockade intervention in HALT A; b) dual RAAS blockade intervention in HALT B; and c) intensive BP intervention (HALT A only)

In HALT A, the effect of intensive (versus less intensive) BP lowering on the risk of ESKD was not statistically significant (unadjusted HR 0.84; 95% CI 0.50–1.40; adjusted HR 0.79; 95% CI 0.46–1.34; Table 1 and Figure 1).

Because the HALT PKD trials ended in 2014, the shorter post-trial follow-up duration and the higher eGFR of participants in HALT-PKD A, where the number of cases of ESKD was low, may have contributed to our null findings despite a median follow-up duration of approximately 10 years. Similar to prior trials, we did not find a definitive benefit of dual RAAS blockade in HALT A or B trials, though there was less signal for potential harm in HALT B trial, which differs from the results of the VA-Nephron-D trial.⁸

The null effects of intensive BP control in HALT A study are consistent with a post-trial follow-up study of SPRINT participants focused on CKD outcomes, where there was no benefit of strict BP control with regard to kidney outcomes after a median follow up of 8.3 years post randomization.⁹ Importantly, the BP targets tested in HALT PKD A are lower than the targets currently recommended by guidelines for patients with CKD.¹⁰

A few limitations should be noted. Following trial closure, any difference in BP or use of dual RAAS blockade between the two arms likely diminished over time, so that the effect sizes that we observed may be attenuated. In addition, we do not have long-term data on other outcomes such as quality of life, hyperkalemia, or cardiovascular events. The confidence intervals for the outcomes were wide, though a potential for benefit from the dual RAAS intervention on kidney failure would be possible with a higher event rate. It is possible that patients who emigrated from the US were not linked to national databases but developed an outcome of interest, though we believe this would be rare. Finally, our results may not generalize to all patients with PKD.

In conclusion, neither dual RAAS blockade nor intensive BP lowering are associated with lower risk of ESKD or death during longer term follow-up of HALT-PKD participants.