As of May 2012, indications for thrombolysis in acute ischaemic stroke were particularly limited [1]: patients had to be treated within 4.5 h from the onset, should be younger than 80 years old, should present with a moderate neurological deficit, and should not present important early ischaemic changes on admission head CT scan, and should not have any of the traditional contraindications for rt-PA therapy. As a consequence, it is not surprising that, despite the commendable efforts of many stroke clinicians, the number of treated patients was very small, with figures around 2–3 % of all strokes [2]. On 23 May 2012, during the European Stroke Conference in Lisbon, results of international stroke trial 3 (IST 3) were presented. The IST 3 is the largest randomised study on thrombolysis in acute ischaemic stroke ever done. The trial, whose complete results were subsequently published in Lancet on the 23 June 2012 [3], included 3,035 patients, without age or severity limits, randomised and treated within 6 h from symptoms onset. IST 3 is a non-profit trial, driven by researchers, with financial support from public agencies, randomized, with open treatment but blind assessment (PROBE design); it lasted 12 years (from 2000 to 2012). It is obvious that during this time many aspects changed in analysing and interpreting stroke trials results, and therefore some adjustments had been necessary during the trial conduct, to adequately study new information, particularly in the field of outcome evaluation. In fact, at the trial start a traditional cut off in Rankin scale (0–2 vs. 3–6) was chosen to evaluate the outcome, and from this cut off a sample size of 3,100 patients was calculated, to show an absolute benefit of 4 %. In the following years, however, two facts prompted the steering committee to add further prognostic evaluations: the first one was ECASS 3 results publication [4], which were not conventionally significant for 2/3 Rankin cut-off, but were statistically significant when using a different cut off (0–1 vs. 2–6); these results caused a wide variation in clinical practice, moving the time limit of treatment from 3 to 4.5 h. The second one was the re-evaluation of trials results by means of ordinal shift analysis [5], a statistical tool which uses all the information from the data, considering all the shifts from one to another grade in Rankin scale, and therefore is more powerful and sensitive for outcome evaluation in acute stroke. Therefore, the steering committee decided, much in advance of the database opening, to introduce these two new endpoints (Rankin cut-off 1/2 and OSA), and to publish the whole statistical analysis plan [6], a procedure not very frequent in stroke trials; in this way, anyone could know the plans for data analysis and interpretation, with the researchers still blind to the actual results. Further to the whole analysis of data, with the three aforementioned endpoints, a few hypotheses, based on clinical impressions and beliefs, had to be verified by IST 3 results: the earlier the treatment is given, the more effective it is (the concept of ‘‘time is brain’’); rt-PA therapy is less effective and more dangerous in older and more severe patients. From 2000 to 2012, 156 hospitals in 12 countries participated in IST 3; in Italy, 21 centres included 326 patients. 53 % of patients were older than 80; on the whole, 95 % did not fulfil the official licence criteria for rtPA use. At 6-month follow up, 37 % of patients treated with thrombolysis and 35 % of controls were alive and independent (Rankiin scale \3) (OR 1.13, CI 0.95/1.35, S. Cenciarelli T. Mazzoli S. Ricci (&) Unita Operativa di Neurologia, ASL 1 dell’ Umbria, Centro di Coordinamento IST 3 per l’ Italia, Ospedali di Citta di Castello e Branca, Citta di Castello, Italy e-mail: stefano.ricci@asl1.umbria.it
Read full abstract
Jun 9, 2014
A J Wright + 1
Open Access
