The five RADIANCE-HTN and SPYRAL-HTN randomised studies suggest that the BP lowering effect of RDN corresponds to the effect of one antihypertensive drug

Renal denervation (RDN) may be a new treatment modality for patients with hypertension. Initially, efforts to test the efficacy of RDN in lowering blood pressure (BP) have focussed on patients with apparent treatment resistant hypertension (aTRH). The SYMPLICITY HTN-2 trial [1] reported a major reduction in systolic BP with RDN in patients with aTRH using office-based BP measurement. However, using ambulatory BP, the state-of-the art technique for measuring BP in patients with aTRH [2], BP reductions were less evident [1]. Further, since poor drug adherence, which is common in aTRH [3], was not monitored in SYMPLICITY HTN-2, interpretation of the study results could be confounded by the Hawthorne effect i.e. patients started taking their drugs as prescribed in response to the attention devoted to them [4].

SYMPLICITY HTN-3 [5] included a sham control group and ambulatory BP measurements that balanced the Hawthorne and white-coat, placebo, and regression-to-the–mean effects, resulting in a BP reduction of 2 mmHg in the RDN treatment group compared to the sham control. Further, meta-analyses of the first generation of randomised controlled studies of RDN did not show BP lowering effects of RDN (Figures 1 and 2), whether or not SYMPLICITY HTN-3 was included [6], and whether or not a sham control (Figures 3 and 4) was a part of the design [7].

Figure 1. Forest plot shows six-month response of office systolic blood pressure to renal denervation (RDN) or to control treatment in the first 6 randomised studies (from ref. [6] with permission). RDN and SBP indicate renal denervation and systolic blood pressure, respectively. Solid points represent the effect size in individual studies and have a size proportional to the inverse of the variance. Horizontal lines and diamonds denote the 95% confidence intervals for individual studies and the pooled estimate, respectively. For all trials combined, the pooled within-group change is given with 95% confidence interval. P values refer to the significance of the pooled between-group estimate and Cochran’s Q test for heterogeneity.

Figure 2. Forest plot shows six-month response of 24-h systolic blood pressure to renal denervation (RDN) or to control treatment in the first 7 randomised studies (from ref. [6] with permission). RDN and SBP indicate renal denervation and systolic blood pressure, respectively. Solid points represent the effect size in individual studies and have a size proportional to the inverse of the variance. Horizontal lines and diamonds denote the 95% confidence intervals for individual studies and the pooled estimate, respectively. For all trials combined, the pooled within-group change is given with 95% confidence interval. P values refer to the significance of the pooled between-group estimate and Cochran’s Q test for heterogeneity.

Figure 3. Forest plot shows six-month response of office systolic blood pressure (SBP) to renal denervation (RDN) or to follow-up in the control group in 7 no sham controlled studies (upper panel) and in 1 sham controlled study (lower panel) (from ref. [7] with permission). Solid points represent the effect size in individual studies and have a size proportional to the inverse of the variance. Horizontal lines and diamonds denote the 95% confidence intervals for individual studies and the pooled estimate, respectively. For all trials combined, the pooled within-group change is given with 95% confidence interval. P values refer to the significance of the pooled between-group estimate and Cochran’s Q test for heterogeneity.

Figure 4. Forest plot shows six month response of 24 h systolic blood pressure (SBP) to renal denervation (RDN) or to follow-up in the control group in 7 no sham controlled studies (upper panel) and in 3 sham controlled studies (lower panel) (from ref. [7] with permission). Solid points represent the effect size in individual studies and have a size proportional to the inverse of the variance. Horizontal lines and diamonds denote the 95% confidence intervals for individual studies and the pooled estimate, respectively. For all trials combined, the pooled within-group change is given with 95% confidence interval. P values refer to the significance of the pooled between-group estimate and Cochran’s Q test for heterogeneity.

However, these disappointments [5–7] did not end the interest in RDN for many reasons. First, total abdominal sympathectomy resulting from surgical splanchnicectomy was highly effective in the treatment of severe hypertension in cohorts of patients reported in the 1930s [8] and 1950s [9,10]. Second, the meta-analyses showed that RDN did not lead to severe adverse events and could be considered safe [6,7]. Third, the role of the sympathetic nervous system in the pathophysiology of hypertension is strong [11,12]. Further, the procedural problems that contributed to the failure of early RDN trials to lower BP could be overcome [13,14]. Therefore, new protocols were designed to assess the antihypertensive efficacy of RDN. One new approach was to perform clinical studies in untreated hypertensive patients in order to avoid any interference of drug therapies. The differences in design between the first ten RCTs included in the neutral meta-analyses [6,7] and the subsequent RCTs have been so great that it is appropriate to distinguish between first and second generation of RDN-RCTs in the treatment of hypertension [15].

RADIANCE-HTN SOLO [16] tested whether endovascular ultrasound RDN could reduce BP in patients with hypertension untreated with antihypertensive drugs for 4 weeks, whereas the SPYRAL-OFF [17] and ON MED trials [18] investigated whether renal denervation achieved by intravascular delivery of radiofrequency energy could reduce BP in patients off medication [17] or in patients on stable antihypertensive medication [18]. Interestingly, similar decreases in BP were obtained in RADIANCE-HTN SOLO and the SPYRAL studies (Table 1), in patients either off or on antihypertensive treatment.

Table 1. Comparisons of the main characteristics (approximate average for the patients who were randomised to renal denervation and to SHAM control) and the net results of 5 new studies of intravascular renal denervation (refs. [16–18,20,21]).

Variable	RADIANCE-HTN SOLO	SPYRAL HTN-OFF	SPYRAL HTN-ON	SPYRAL OFF PIVOTAL	RADIANCE-HTN TRIO
No. randomised	146	80	80	331	136
Age (yrs.)	54	54	53	53	53
% men	60	70	84	66	80
BMI (kg/m^2)	30	30	32	31	33
Office SBP (mmHg)	154	162	164	163	163
Office DBP (mmHg)	100	101	101	102	104
24-h SBP (mmHg)	143	152	152	151	145
24-h DBP (mmHg)	88	99	98	99	89
Net results (mmHg) expressed as difference in reductions between renal denervation and control (SHAM):
Change Office-SBP	−6.5	−7.7	−6.8	−6.7	−5.0
Change Office-DBP	−4.1	−4.9	−3.5	−4.1	−4.0
Change 24 h-SBP	−4.1	−5.0	−7.4	−4.1	−5.6
Change 24 h-DBP	−1.8	−4.4	−4.1	−2.9	−3.0
Change daytime SBP	−6.3		−6.1	−4.0	−5.0
Change daytime DBP	−2.6		−4.1	−4.0	−2.9

BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure.

However, the Hawthorne effect was not completely ruled out in RADIANCE-HTN SOLO [16], as investigators relied on patient-reported antihypertensive drug use. Reported medication use was more common in the sham groups in the RADIANCE-HTN trial [16] and in SPYRAL ON-MED [18], with a mean drug adherence of 60%. In addition, hidden use of drugs was not evaluated by direct measurements in blood or urine. Thus, assuming that hidden drug use and variations in drug intake were not more common in the treatment group, RADIANCE-HTN SOLO [16] together with the SPYRAL-OFF [17] and ON MED trials [18] may provide the first true trial evidence that RDN lowers BP. The designs, patient inclusion and exclusion criteria, baseline characteristics and ambulatory and office BP lowering effects were similar in all 3 trials (Table 1). The reasons for these standardizations of the protocols are likely that the same investigators were involved. Procedures were standardised and constantly improved in many aspects using the French DENERHTN study [19] as a pioneer and example to follow. The DENERHTN study also showed a 6 mmHg BP reduction compared to control [19].

RADIANCE-HTN SOLO, and the SPYRAL-HTN OFF and ON MED trials thus open a new chapter in RDN research [16–18]. Together, these trials suggest that a RDN class effect may exist. Renal denervation achieved by intravascular delivery of radiofrequency energy or ultrasound energy may produce comparable BP lowering, and RADIANCE-HTN SOLO, together with the SPYRAL trials, take the RDN field into a new era. Recently SPYRAL-HTN OFF MED PIVOTAL [20], with a larger number of included patients than SPYRAL-HTN OFF MED [17], and RADIANCE-HTN TRIO [21], in which drugs for aTRH were standardised in a 3-component single-pill (valsartan, amlodipine, hydrochlorothiazide) confirmed significant BP reductions with RDN compared to SHAM (Table 1).

What is the clinical benefit? The reduction in ambulatory systolic BP of approximately 6 mmHg reported in the new studies [16–18,20,21] and confirmed in the Global SYMPLICITY Registry corresponds to the effect of one antihypertensive drug [22,23]. As shown in RADIANCE-HTN SOLO [16] there may be responders and non-responders, and some patient subgroups may benefit more than others. In that respect, RDN does not differ from an antihypertensive drug, as the BP response to a new BP lowering drug may be unpredictable. However, the percentage of previously untreated hypertensive patients reaching target BP with RDN remains unknown and a majority of patients still need antihypertensive drugs to achieve therapeutic goals. Further research is needed to identify patients most likely to benefit from RDN [24]. A recent analysis of the HTN-OFF MED study [25] found reductions of ∼ 10 mmHg in office SBP and ∼ 6 mmHg in ambulatory SBP in patients with baseline heart rate ≥70 beats/min, but no BP lowering effects in patients with heart rate <70 beats/min. Further research also needs to confirm a sustained long-term BP lowering effect of RDN [26], which could then make the procedure cost-effective, at least for aTRH [27]. Long-term safety beyond the initial 6 months post-procedure must also be established. Currently we have 3 year safety data from the Global SIMPLICITY Registry [28] and 7 year safety data from the Oslo RDN study [29] suggesting no deterioration in renal function. Most importantly, future research needs to show that BP reduction following RDN protects the brain, heart, kidneys and large arteries and lowers cardiovascular morbidity and mortality. RADIANCE-HTN and the SPYRAL studies inject new optimism into this field of research, delivering energy related to the favourable effects of endovascular renal denervation. After overcoming multiple problems, with a dramatic fall in expectations from extremely high [1] down to almost nothing [5–7], we may again be on an uphill spiral [16–21,30].

Disclosure statement

SEK, KN, MB and SO are editors of Blood Pressure and report no conflict of interest to disclose related to this editorial review.