In reply: We welcome the opportunity to respond to issues raised by Hurley and Toncich. Australian, United States and European stroke guidelines based on level 1, grade A evidence recommend intravenous recombinant tissue plasminogen activator (IV-rtPA) for patients with acute ischaemic stroke (AIS) who meet specific inclusion criteria and present within 3 hours of AIS onset.1,2 Recombinant tPA is approved for this indication by the Therapeutic Goods Administration.

We obtained informed consent before thrombolysis, but recognise that this is challenging because of the urgent need to initiate treatment, rather than doubts regarding benefit.

A Cochrane systematic review of eight randomised controlled trials (RCTs), which together comprised 2955 patients, concluded that IV-rtPA within 3 hours of AIS onset was more effective in reducing death or dependency (odds ratio [OR], 0.66; 95% CI, 0.53–0.83), with no statistically significant adverse effect on death (OR, 1.13; 95% CI, 0.86–1.48).3 The Third International Stroke Trial (IST-3), which Hurley cites to support his view that sub-3-hour stroke thrombolysis remains controversial, was in fact designed to address different issues: efficacy of rtPA given 3–6 hours after symptom onset and to those aged over 80 years, as well as imaging predictors of response.

We believe Hurley’s figure from his own letter is an incorrect attempt at meta-analysis. He combines data from RCTs and selected observational studies (excluding the Safe Implementation of Thrombolysis in Stroke Monitoring Study [SITS-MOST] of 6483 patients4), a method subject to selection bias.

Although the original Cleveland study (as cited by Hurley) showed high rates of intracranial haemorrhage, 50% of patients in the study were not treated according to the protocol. The incidence of intracranial haemorrhage was subsequently reduced to the expected level following implementation of a quality improvement program to ensure strict adherence to protocols.5 Hurley suggests that projections of IV-rtPA benefits are misleading because of baseline imbalances in the NINDS trial. These concerns were addressed by reanalysis of the data by an independent committee commissioned by NINDS,6 which not only reconfirmed the original results but suggested an even greater benefit for AIS patients receiving IV-rtPA.

We believe Hurley’s statement that mortality rates are increased when stroke thrombolysis occurs in non-teaching hospitals is unsubstantiated. In the study cited, mortality rates were similar to those observed in RCTs of IV-rtPa, with no difference between teaching and non-teaching hospitals.7 Indeed, large registries in Europe (where 50% of centres had little or no prior experience)4 and in Canada8 report efficacy rates following thrombolysis similar to those from RCTs with lower rates of intracranial haemorrhage and mortality.

In our view, IV-rtPA within 3 hours of AIS does not require more RCT evidence. Due to therapeutic inertia, only a small proportion of eligible stroke patients receive this therapy. Comprehensive acute stroke protocols in the emergency department in partnership with stroke clinicians could improve delivery of this highly effective treatment.