Critical Analysis of a Randomised Controlled Trial

The following essay aims to critically analyse a randomised controlled trial; ‘Effect of IV alteplase on the ischemic brain lesion at 24-48 hours after ischemic stroke’ by Mair et al.,(2018). Alteplase was investigated in order to determine whether it alters the development of ischaemic lesions on brain imaging after stroke. This intervention is a thrombolytic, tissue plasminogen activator which is currently being prescribed to patients under the age of 80 years who have suffered an ischaemic stroke. The World Health Organisation (2018)states that stroke is caused by an interruption of the blood supply to the brain that can be due to the rupturing of a vessel or, in this case, an occlusion of the vessel by a clot, rich in fibrin. The intervention in question works by binding to the fibrin in the thrombus and converts the plasminogen to plasmin, which then degrades the fibrin complex (DrugBank Database, 2018).

The authors of this trial aimed to assess the effect of the thrombolytic agent on any change in lesion appearance on CT or MRI between pre-treatment and 24-48-hour follow-up imaging. They investigated whether short-term progression of the lesion predicted long-term functional outcome after a stroke. Patients were randomly assigned to intravenous alteplase or placebo, and in follow-up imaging, acute lesion visibility was graded on an ordinal scale, reflecting the range of progressive changes in the tissue. The results of this trial showed no clinical significance but state that intravenous alteplase is effective and safe to use after an ischaemic attack. There have, however, been various studies looking at the effects of intravenous alteplase, most of which generating significant results for improvement of symptoms between 3-4.5 hours post-stroke. For example, two similar studies that recruited roughly 6000 patients found that IV alteplase was safe and effective when used within 3-4 hours of stroke onset (Wahlgren et al., 2007; Emberson et al., 2014;). A smaller, double-blind study by Hacke et al. (2008) recruited 821 patients and found that intravenous alteplase significantly improved clinical symptoms in patients with stroke. These types of studies focus on the effect the intervention has on the brain lesion in imaging and are enabling more rapid treatment for patients outside the age limit of 80 years.

This trial is a multicentre PROBE (prospective, randomised, open-label, blinded endpoint) study, the advantages among which are a lower cost and a greater similarity to clinical practice (Hansson et al., 1992). However, a prospective study is open to a bias interpretation of results which ultimately reduces their results. Patients in this trial were randomly allocated IV alteplase or placebo, this was achieved by using a randomising computer generator to reduce any selection bias by allocation concealment. Patients who were recruited were assessed for acute stroke by the National Institutes of Health Stroke Scale (NIHSS). Interpretation of this scale is as that the higher the NIHSS score, the more sever the stroke. As previously mentioned in this essay, alteplase is currently being prescribed to patients under the age of 80, however according to Stroke Center (2018), onset of stroke in the majority of the population is 65 years or more. This trial recruiting patients with no upper age limit, allowing comparison to be made between the recovery time of both elderly and younger patients.

The exclusion criteria in the research protocol is more extensive than the inclusion criteria, resulting in a rather vague participant pool (The IST-3 collaborative group, 2009). There was no minimum age limit given, however an age rangewould have provided age intervals for which future researchers could use for comparison. Nevertheless, characteristics of the control group were very similar to that of the alteplase group, which minimises the risk of any complications or anomalous data being produced. However, it was reported that out of 2,916 patients, 1,416 patients were male, which is approximately 50% of the study population. This is confirmed in the table of baseline characteristics found in table 1 of this article where it is stated that 720 males were allocated to altepalse and 696 males were allocated to the control group. However, there is no mention of the characteristics of the other 50% of patients with no explanation of whether these 50% were excluded or not included in the data, neither was there any explanation as to why this was the case. Furthermore, 276 patients were treated as part of the double-blind phase of the study. However, the authors do not state why they had a double-blind phase instead of a double-blind trial. There is no data provided that include these 276 patients either. Double-blind trials have numerous strengths in regards to reliability and bias, if this trial were to have used this method of blinding, their findings would be backed up. Informed consent was obtained for all the patients, as stated in the author’s protocol (The IST-3 collaborative group, 2009).

Patients were followed-up at 6 months to asses functional status using the Oxford Handicap Scale (OHS), a commonly used variant of the modified Rankin score. This data is only represented in table 1 of the paper, a data set in the form of an ‘Oxford Handicap Scale for outcome at 6 months by treatment group’, similar to that of the IST-3 trial as seen in figure 2 (IST-3 collaborative group, 2012) would allow for an easily interpretable set of results. T1, T2, DWI, fluid-attenuated and T2* were required pre-treatment and at follow-up, though, these images were not presented in the data. The various methods of imaging used is beneficial as it generates a clear image of the brain and so any lesions are easily visible. However, not including this data allows forthe results to become invalid, and so with the authors stating that patients were only eligible if imaging at baseline had excluded an intracranial haemorrhage, there is no proof of validation. CT was performed at follow-up with which acute lesion visibility was graded on a 3-point ordinal scale, MRI analysis used a similar scale. This scale defines grade 0 as no acute lesion visible; grade 1 as gray matter attenuation equal to that of normal white matter; and severe, or grade 2 as gray and/or white matter attenuation lower than normal white matter.

Location of the lesion was assessed using the ‘IST-3 Ischemic Lesion Score’ and the Alberta Stroke Program Early CT Score (ASPECTS). Authors stated that a positive value for the IST-3 lesion score represents imaging progression and therefore worsening of the lesion, examples of this can be found in figure e1, found in the protocol (The IST-3 collaborative group, 2009). This appears to be a sensible grading system, however very few studies have used this same scoring method and so the reliability and validity remains in question. Furthermore, in table 2the authors state that “data for the location of acute lesion and degree of swelling was not available for all patients at both time points”, therefore allowing the data to be incomplete as there was no explanation as to why this was the case (The IST3 collaborative group, 2018). The ASPECTS scoring system was well described as starting with 10 points and deducting 1 point for every area of the brain affected by the lesion. It is a tomographic system that divides the middle cerebral arterial supply into 10 regions of interest use of aspects (Pexman et al. 2001).Mair et al. (2018) used a 7-point scale to measure acute lesion swelling, the rationale behind which was heavily based on the classification system discovered by Wardlaw and Sellar (1994).This scale was based on sulcal or ventricular effacement, or midline shift with 6 representing a midline shift effacement of the basal cisterns and a 0 representing no change. Univariate tests were used to compare alteplase with the control which is more of a descriptive method of analysis rather than informative. Multivariate models were used for identifying indicators of progression, which is the more accurate method of distinguishing a relationship among multiple variables. The authors reported that they did not include the ASPECTS scores and IST-3 Lesion Scores in the same multivariate models, due to the risk of confounding, according to the authors. This would have been beneficial to the results in comparison between location and severity of the lesion.

The authors stated that their study provides Class II evidence that IV altepalse impedes the development of ischaemic lesions, however they do not explain what this means. The definition of the different classes of evidence for articles on treatment as declared by the Evidence-Based Spine-Care Journal in 2013 states that class II evidence has moderately low risk (and therefore has potential for some bias) and also has deficiencies that will not invalidate the results. The criteria for class II evidence includes an adequate sample size, equal application of co-interventions and blind or independent assessment, therefore concluding that the quality of the trial is moderate to poor and the quality of the cohort is good.

Overall the methods section of this paper was well described, and the safety of the participants was taken into account. A standard dose of IV altepalse was used and all the possible side effects and contraindication of the intervention have been listed in the full trial protocol (The IST-3 collaborative group, 2009). However, there is no data for the 276 patients involved with the double-blind phase of the study neither was this phase of the trial mentioned anywhere else in the protocol nor the paper itself.

Mair et al. (2018) define an acute lesion as “any reduced tissue CT attenuation/increased T2-weighted hyperintensity, or swelling”, which is demonstrated well in figure e1 of the protocol (The IST-3 collaborative group, 2009). Out of the 2,916 patients, 1,183 had acute lesions of which 92% involve the middle cerebral artery (MCA) territory. The authors summarise their results with lesion visibility decreasing in more patients allocated to control than to alteplase. This is a relatively accurate conclusion to draw as lesion visibility increased in 64% of patients in the alteplase group, compared to 67% in the control. Although the difference is slight, this confirms a positive result in favour of the intervention. Ordinal regression analysis concluded that patients allocated to alteplase were less likely to show an increase in lesion visibility during follow-up scans. This is validated by the low p value (p < 0.001) and therefore significance of the analysis, as seen in table 4. It was also found that treatment with alteplase and an increased time between baseline and follow-up independently predicted a better outcome at 6 months, shown in table 5. However, these two variables are the only two that increase the likelihood of a better outcome, due to the fact that an odds ratio greater than 1 indicates an increased likelihood of a better outcome. Worsening of lesions on imaging at 24-48 hours, according to the authors, also independently predicted poor outcome at 6 months. This conclusion is based solely on the ordinal regression analysis and is therefore not an accurate statement to make. The odds ratio for worsening ASPECTS score was 0.92, however, both increased visibility and increased age generated an odds ratio of 0.97 meaning that these two variables would be more likely to generate a better outcome than worsening of ASPECTS. Both scores lie close to 1, indicating a relatively higher likelihood of a better outcome.

In order to further confirm their results, the authors carried out a meta-analysis to identify trials of IV altepalse that reported outcomes by imaging assessments of ischaemic lesions. However, the researchers did not state where they collected this data from, instead they report that for every trial found, PubMed was searched for any subgroup analyses (The IST-3 collaborative group, 2018). Searching a multiple of scientific databases for multiple articles would warrant a better comparison and can be shown using a flow diagram of the literature search, along with a table of characteristics of included studies, enabling clear interpretation for the readers. Only 6 articles were eventually included in their meta-analysis, which is a relatively low number of articles to show the accuracy of their gathered information. They found that altepalse impeded lesion growth, however, this statement is not strongly supported by the evidence provided. Although there were a large number of patients included overall, the number of trials and therefore the methods and various errors would be highlighted rather than disregarded. The authors’ interpretation of the meta-analysis was not conducted in grave detail, instead the final outcome was all that was described. Figure 2 shows the analysis of the results generated by the meta-analysis, from the odds ratio (OR) graph it is clear that the confidence intervals are relatively high, therefore reducing the significance of the data. There is only one study that had significance at the study level which is due to the confidence interval not crossing the midline at 1. The other studies all have confidence intervals that cross the midline therefore making them insignificant at the study level. However, on the meta-analysis level, the estimates both lie to the right of the midline, in favour of the intervention and therefore are significant. The I² value is a test of heterogeneity between the trials, the lower the I² the lower the heterogeneity and therefore less variability between the studies. The Cochrane Handbook for Systematic Reviews of Interventions (2011) reports thresholds for interpretation as; ‘0-40% might not be important; 30-60% may represent moderate heterogeneity; 50-90% may represent substantial heterogeneity; 75-100% considerable heterogeneity’. From figure 2 it is reported that the I²value was 31.5%, which represents moderate heterogeneity. 

The results were relatively easy to follow and interpret and do confirm the previously stated hypothesis of ‘determining whether alteplase alters the development of ischaemic lesions on brain imaging after stroke’ (Mair et al., 2018).

The authors state that it seemed likely for the IST-3 trial to have been underpowered to detect small changes in lesion extent but that when combining the data in a meta-analysis, however, they found that alteplase could reduce the progression of an ischaemic lesion. This statement suggests that the trial on its own, not including the meta-analysis would have been insufficient to declare the intervention clinically beneficial. Mair et al. (2018) were able to pick up on unusual findings such as the fact that pre-stroke leukoaraiosis appeared to reduce the progression of lesion visibility. They supported this statement by suggesting that this could be because lesions were seen less often in patients with abnormal white matter at baseline. Although they could not generate any reasoning for this, it is good that they were able to acknowledge abnormal results.

The protocol states a list of side effects and contraindications of alteplase with the most common appearing to be bleeding from damaged blood vessels or punctures (The IST-3 collaborative group, 2009). However, a table with the number of patients experiencing any side effects would be an informative way to present the data. This way, the public would be fully informed with which side effects are most likely to affect them and can therefore take that into consideration. A large number of people avoid needles and so intravenous administration could be a minor limitation to them. On the other hand, intravenous administration has a wide variety of advantages when compared to other routes of administration. This method would allow the drug to get into the blood stream directly and will therefore have a very fast therapeutic effect where the rate of administration is more easily controlled. Some people are not able to swallow tablets and so this method would benefit them greatly.

Several non-profit foundations were listed in the protocol, as well as a number of research councils, that supported the trial with grants (The IST-3 collaborative group, 2009). However, during the double-blind phase of the trial, Boehringer-Ingelheim GmbH supplied the 300 patients with the alteplase and the placebo. This could be a potential explanation as to why the data from the patients in the double-blind phase were not included in the results. According to their website (Boehringer-Ingelheim, 2018), this company sells IV alteplase in the form of Actilyse. The authors should have addressed this issue as a conflict of interest as this could have enhanced sales of their product for which the intention is unclear. Not only will this reduce the validity of the results, but it will also undermine confidence in medical knowledge, especially from the public’s perspective. A similar issue is that there have been vast amounts of media coverage by newspapers such as Scottish Daily Express, The Scotsman, The Edinburgh Paper: Evening News, and many more. The bulk of these articles are in favour of the drug therefore spreading a positive impact of its efficacy. Members of the public tend to rely on media coverage as a source of knowledge from the world of science, however, the media do not mention the relatively severe side effects, as previously mentioned which misinforms the public (Mills, R., 2012; Walker, N., 2012; Cummings, L., 2012).

In conclusion, this trial was part of important research in finding novel, more rapid treatments for stroke. The results of this trial could potentially reduce costs by lowering the demand of long-term care. Although there were multiple confounders that interfered with this trial, a relatively accurate conclusion can be made that alteplase is associated with reduced short-term progression in lesion visibility. This is supported by the collective data from the meta-analysis that confirmed the authors’ work with previous studies. Furthermore, Mair et al. (2018) stated that imaging biomarkers for the use of alteplase on the ischaemic brain would be valuable for clinical practice. A small study by Yoo et al. (2013)concluded that MRI-based analysis of hemisphere growth was a suitable biomarker for oedema formation. More rigorous research should be done around this topic as this could be vital for future clinical practice.

• Boehringer-Ingelheim., (2018). Actilyse® [online]. Boehringer-Ingelheim. [Viewed 24 November 2018]. Available from: https://www.boehringer-ingelheim.com/products/actilyse
• Cummings, L., (2012). Breakthrough after Capital team’s stroke drug trial. The Edinburgh Paper: Evening News. 23 May. p. 8.
• Emberson, J., Lees, K.R., Lyden, P., Blackwell, L., Albers, G., Bluhmki, E., Brott, T., Cohen, G., Davis, S., Donnan, G., Grotta, J., Howard, G., Kaste, M., Koga, M., Kummer, R.V., Lansberg, M., Lindley, R.I., Murray, G., Olivot, J.M., Parsons, M., Tiley, B., Toni, D., Toyoda, K., Wahlgren, N., Wardlaw, J., Whitelet, W., Zoppo, G.L.D., Baignet, C., Sandercock, P. and Hacke, W., (2014). Effect of treatment delay, age, and stroke severity on the effects of intravenous thrombolysis with alteplase for acute ischaemic stroke: a meta-analysis of individual patient data from randomised trials. The Lancet [online]. 384(9958), pp.1929-1935. [Viewed 24 November 2018]. Available from: https://doi.org/10.1016/S0140-6736(14)60584-5
• Evid. Based Spine Care J., (2014). Definition of Classes of Evidence (CoE) and Overall Strength of Evidence (SoE). Evidence-Based-Spine-Care Journal [online]. 5(1), p. 71. [Viewed 24 November 2018]. Available from: https://doi.org/10.1055/s-0034-1373841
• Hacke, W., Kaste, M., Bluhmki, E., Brozman, M., Davalos, A., Guidetti, D., Larrue, V., Lees, K.R., Medeghri, Z., Machnig, T., Schneider, D., Kummer, R.V., Wahlgren, N and Toni, D., (2008). Thrombolysis with Alteplase 3 to 4.5 Hours after Acute Ischemic Stroke. The New England Journal of Medicine [online]. 359(13), pp.1317-1329. [Viewed 24 November 2018]. Available from: https://doi.org/10.1056/NEJMoa0804656
• Hansson, L., Hedner, t. and Dahlof, B., (1992). Prospective randomised open blinded end-point (PROBE) study. A novel design for intervention trials. Prospective Randomised Open End-Point. Blood Press. 1(2), pp.113-119.
• Higgins, J.P.T. and Green, S. (eds.), (2011) Cochrane Handbook for Systematic Reviews of Interventions [Online]. [Viewed 25 November 2018]. Available from: http://training.cochrane.org/handbook
• The Internet Stroke Center, (2018). Stroke Statistics [online]. The Internet Stroke Center. [Viewed 25 November 2018]. Available from: http://www.strokecenter.org/patients/about-stroke/stroke-statistics/
• The IST-3 collaborative group, (2009) Research protocol of the IST-3 trial [online]. [Viewed 25 November 2018]. Available from: http://www.dcn.ed.ac.uk/ist3/
• The IST-3 collaborative group, (2012). The benefits and harms of intravenous thrombolysis with recombinant tissue plasminogen activator within 6 h of acute ischaemic stroke (the third international stroke trial [IST-3]): a randomised controlled trial. The Lancet [online]. 379(9834), pp.2352-2363. [Viewed 24 November 2018]. Available from: https://doi.org/10.1016/S0140-6736(12)60768-5
• Mair, G., Kummer, R.V., Morris, Z., Heijne, A.V., Bradey, N., Cala, L., Peeters, A., Farrall, A.J., Adami, A., Potter, G., Sandercock, P.A.G., Lindley, R.I. and Wardlaw, J.M., (2018) Effect of IV alteplase on the ischemic brain lesion at 24-48 hours after ischemic stroke. Neurology [online]. [Viewed 24 November 2018]. Available from: https://doi.org/10.1212/WNL.0000000000006575
• Mills, R., (2012). New drug trial brings hope to patients. Scottish Daily Express. 24 May. p. 25.
• Pexman, J.H.W., Barber, P.A., Hill, M.D., Secivk, R.J., Demchuk, A.M., Hudon, M.E., Hu, W.Y. and Buchan, A., (2001). Use of the Alberta Stroke Program Early CT Score (ASPECTS) for Assessing CT Scans in Patients with Acute Stroke. American Journal of Neuroradiology. 22(8), pp.1534-1542.
• Wahlgren, N., Ahmed, N., Davalos, A., Ford, G.A., Grond, M., Hacke, W., Hennerici, M.G., Kaste, M., Kuelkens, S., Larrue, V., Lees, K.R., Rione, R.O., Sionne, L., Toni, D. and Vanhooren, G., (2007). Thrombolysis with alteplase for acute ischaemic stroke in the Safe Implementation of Thrombolysis in Stroke Monitoring Study (SITS-MOST): an observational study. The Lancet [online]. 369(9558),pp.275-282. [Viewed 24 November 2018]. Available from: https://doi.org/10.1016/S0140-6736(07)60149-4
• Walker, N., (2012) Clot-busting drug helps recovery from stroke. The Scotsman. 24 May. p. 7.
• Wardlaw, J.M. and Sellar, R., (1994). A Simple Practical Classification of Cerebral Infarcts on CT and Its Interobserver Reliability. American Journal of Neuroradiology. 15(10), pp.1933-1939.
• WHO., (2018). Stroke, Cerebrovascular accident [online]. The World Health Organisation. [Viewed 24 November 2018]. Available from: http://www.who.int/topics/cerebrovascular_accident/en/
• Yoo, A.J., Sheth, K.N., Kimberly, W.T., Chaudhry, Z.A., Elm, J.I., Jacobson, S., Davis, S.M., Donnan, G.A., Albers, G.W., Stern, B.J. and Gonzales, R.G., (2013). Validating Imaging Biomarkers of Cerebral Edema in Patients with Severe Ischemic Stroke. Journal of Stroke and Cerebrovascular Diseases [online]. 22(6), pp.742-749. [Viewed 25 November 2018]. Available from: https://doi:10.1016/j.jstrokecerebrovasdis.2012.01.002