A study in which ambulance paramedics treated patients suspected of having an acute stroke with transdermal glyceryl trinitrate (GTN) to lower their blood pressure found that “there was a strong tendency for GTN to be associated with a worse outcome” than placebo. Philip Bath from the University of Nottingham in the UK, who presented the results of RIGHT-2 (Rapid intervention with glyceryl trinitrate in hypertensive stroke trial-2) in a late-breaking session at the International Stroke Conference (ISC; 5–8 February, Honolulu, USA), concluded: “We cannot recommend the use of GTN in the ambulance.”
According to Bath, while high blood pressure predicts poor outcome in acute stroke, the best way to lower blood pressure and improve outcome in acute stroke is not yet known. GTN, a nitric oxide donor, was found to be safe when given to patients with acute stroke in the ENOS (Efficacy of nitric oxide in stroke) trial. Accordingly, in the RIGHT pilot trial and in an early subgroup in the ENOS trial, transdermal GTN given in ultra-acute/hyper-acute stroke improved functional outcomes in both ischaemic and intracerebral stroke. Thus, RIGHT-2 sought to investigate whether administering GTN earlier after stroke onset would improve outcomes.
Bath described the RIGHT-2, which was a multicentre, parallel-group, prospective, randomised, blinded-endpoint controlled trial, carried out from May 2015 to April 2018 in adults who presented to paramedics in the context of a 999 ambulance call for “stroke”. Patients were included in the trial if they had a face-arm-speech-time (FAST) score of ≥2, presented within 4 hours of stroke onset and had a systolic blood pressure (SBP) greater than 120mmHg.
Patients were randomised in the ambulance, with a paramedic applying either a GTN or sham transdermal patch, with three more patches applied in hospital on successive days. The trial was planned for 850 patients from more than five ambulance services and over 30 associated acute hospital stroke centres across the UK. The primary outcome was the seven-level modified Rankin Scale (mRS) at 90 days, of which the study investigators assessed by central telephone follow-up with masking to treatment.
The trial was planned as a standard intention-to-treat analysis but the rate of stroke mimics (non-stroke/non-transient ischaemic attack [TIA]) was higher than predicted at 26% vs. 12%, meaning that more than a third of patients were not the target for treatment. This required an increase in the sample size to 1,050, while the protocol was changed to a hierarchical analysis, with primary analysis in the target population (stroke or TIA patients). If the latter was found to be positive, a pragmatic analysis with intention-to-treat was subsequently carried out.
The RIGHT-2 results
Of the 1,149 patients in the trial, 568 received GTN and 581 received sham patches. Adherence with the first patch was >99%, and with the first two patches was only 57%, mainly because many patients were discharged once they got to hospital with a TIA or mimic; 96% of patients had primary outcome and 98% vital status.
The average age of patients was 72 years (52% male), with an onset to randomisation of 71 minutes with a patch applied two minutes later. Patients were later diagnosed with ischaemic stroke (52%), intracerebral haemorrhage (13%), TIA (9%) and mimic (26%). Additionally, the authors reported that systolic blood pressure was approximately 162mmHg in 25% of patients, while the FAST score was a maximum of three in 60% of patients.
According to Bath, the reduction in blood pressure with GTN was lower than expected, with systolic blood pressure reducing by 5.7mmHg and diastolic blood pressure by 2.6mmHg at hospital admission. On day two, systolic blood pressure reduced by 5.3mmHg, while diastolic blood pressure reduced by 2.1mmHg in patients in the GTN group compared with the shame group. However, the difference in blood pressure between the groups then disappeared by days three and four.
In terms of the primary outcome in stroke/TIA, Bath said it “was poor for GTN”. According to the time of randomisation, the effect of GTN on mRS was negative in patients treated within one hour, neutral with a negative tendency in those treated within one to two hours, and neutral with a positive tendency in those treated beyond two hours.
Moreover, Bath concluded that in the stroke/TIA target population: “There was a neutral effect on the mRS, but there was a strong tendency for GTN to be associated with a worse outcome”. He further put forward that “the difference between the results of the RIGHT pilot study and RIGHT-2 is due to time before treatment, with a reduction from 264 minutes to 71 minutes”.
Addressing the ISC audience, Bath discussed this discrepancy: “The problem largely lies with intracerebral haemorrhage. In that very first hour [of time before treatment] we are of course breaking the law we learnt at medical school, which is the first part of haemostasis is spasm… and we gave an antispasmodic—a vasodilator.
“But this is just speculation”, Bath added, as he ultimately concluded: “There is no indication for the use of GTN in the ambulance.”