Novel strategies for secondary stroke prevention - exploring the use of pioglitazone therapy using a mixed methods approach

Azhari, Hala Fouad (2020) Novel strategies for secondary stroke prevention - exploring the use of pioglitazone therapy using a mixed methods approach. PhD thesis, University of Glasgow.

Abstract

Pioglitazone, an oral thiazolidinedione (TZD), is used in the second- or third-line management of type 2 diabetes mellitus (T2DM). In addition, the Insulin Resistance Intervention after Stroke trial showed that pioglitazone reduced cardiovascular events in people with recent ischaemic stroke or transient ischaemic attacks who were insulin resistant but had no diabetes (relative risk reduction of 24%; 95% confidence interval (CI) 0.62-0.93; P=0.007; absolute risk reduction of 3%) versus placebo. Furthermore, with pioglitazone use, about half as many patients went on to develop diabetes. However, the net benefit of pioglitazone has been questioned because, with its use, the fracture rate increased by about a third (29% relative risk (RR); 2 to 5% absolute risk (AR)), many patients suffered substantial weight gain (55% RR; 6% AR) and there was an increase in peripheral oedema (43% RR; 10% AR). This means doctors are unsure about whether they should use pioglitazone. This is a much-discussed topic in the stroke community. This thesis addresses several areas concerning pioglitazone use after stroke. It aims to better characterise the risk of fracture, explore patient and clinician views regarding pioglitazone use and explore whether other new diabetes treatments could be similarly effective in people with strokes.
The fractures, weight gain and peripheral oedema associated with pioglitazone use are important clinical issues. However, fractures are the most concerning. I performed a systematic review and meta-analysis of the fracture risk with TZD drugs following best-practice guidance. A search of the Cochrane Register of Controlled Trials, MEDLINE, EMBASE, the Web of Science and the trial registries (from their inception to Week 36, 2017) was performed. The updated searched was done in Week 10 in 2020. The grey literature was also searched. Published and unpublished studies comparing TZDs with placebo or other anti-hyperglycaemic drugs were included if they reported data on fracture occurrence. The search identified 860 studies. A total of 78 were included in this study as they satisfied the eligibility criteria. Thirty-four included trials had a high risk of bias, eight had unclear risk and 36 low risk. These studies had an overall high-to-moderate GRADE score of evidence. They included 39,568 participants on TZDs and 25,718 on comparators with 1,917 fracture events. The fracture risk was increased with TZDs (RR 1.35; 95% CI 1.24−1.48; P<0.00001; I2=10%) versus controls. In 14 trials comparing pioglitazone versus placebo (13,451 participants; 449 fractures), pioglitazone use increased fractures (RR 1.21; 95% CI 1.01−1.45; P=0.04; I2=32%). Both non-serious (RR 1.25; 95% CI 1.03−1.51; P=0.02; I2=53%) and serious fractures (RR 1.48; 95% CI 1.10−1.98; P=0.01; I2=24%) increased with pioglitazone use. In 13 trials comparing pioglitazone to anti-hyperglycaemic comparators (11,267 participants; 99 fractures), there was no increase in fracture risk (RR 1.08; 95% CI 0.73−1.59; P=0.70; I2=15%). Fracture risk increases occurred mostly in the spine (RR 2.13; 95% CI 1.28−3.55; P=0.004; I2=66%) and the lower extremities (RR 1.85; 95% CI 1.33−2.56; P=0.0002; I2=0%) with pioglitazone use. Increased risk was more evident in females (RR 1.56; 95% CI 1.20−2.02; P=0.0008; I2=0%) than in males (RR 1.10; 95% CI 0.84−1.43; P=0.49; I2=14%) with pioglitazone use. Pioglitazone use increases fractures, but this finding was mostly driven by studies that compared its use to placebo and by an increase in women. These data may help stroke clinicians decide whether to use pioglitazone in selected patients to guide further post-stroke trials.
There are several new classes of antidiabetic drugs, some of which have been demonstrated to reduce cardiovascular risk. I assessed the effects of these therapies on the incidence of stroke via a systematic review and meta-analysis of stroke outcomes. A systematic search of the Cochrane Register of Controlled Trials, MEDLINE, EMBASE, the Web of Science and trial registries (from their inception to Week 36, 2017) was performed. The updated searched was done in Week 10 in 2020. Published and unpublished trials comparing dipeptidyl peptidase-4 inhibitors (DPP4-Is), glucagon-like peptide-1 receptor agonist (GLP1-RAs) and sodium-glucose cotransporter-2 inhibitors (SGLT2-Is) to placebo or other anti-hyperglycaemic drugs were included if they reported data on stroke outcomes. The search identified 960 trials. Of these, 256 were included in this study as they satisfied the eligibility criteria. Of these, there were no stroke events were reported in 82 studies and there were 174 trials with at least one event. These included 59,180 people treated with DPP4-Is, 36,069 with GLP1-RAs, 32,816 with SGLT2-Is and 104,570 with comparators. There were 3,914 stroke outcomes. Thirty-two studies had a high risk of bias, 19 an unclear risk and 205 a low risk. These studies had an overall moderate GRADE score of evidence except for fatal stroke with a low-to-very low quality of evidence. The use of DPP4-Is (95 trials, 109,243 participants, 1,431 stroke events) did not reduce the occurrence of stroke (RR 0.93; 95% CI 0.84−1.03; P=0.14) for either ischaemic (RR 0.95; 95% CI 0.84−1.07; P=0.38) or haemorrhagic (RR 0.68; 95% CI 0.22−2.12; P=0.51) events versus placebo or comparator therapies. The use of GLP1-RAs (33 trials, 70,443 participants, 1,637 stroke events) did reduce the occurrence of stroke (0.84; 95% CI 0.77−0.93; P=0.0005) and ischaemic events (RR 0.85; 95% CI 0.77−0.94; P=0.002) versus placebo but there was no reduction in haemorrhagic stroke rate (RR 0.50; 95% CI 0.09−2.73; P=0.42) versus placebo or comparator therapies. The use of SGLT2-Is (46 trials, 52,949 participants, 846 stroke events) did not change the risk of stroke (RR 1.01; 95% CI 0.89−1.16; P=0.84) for either ischaemic (RR 1.03; 95% CI 0.89−1.19; P=0.68) or haemorrhagic (RR 0.39; 95% CI 0.10−1.58; P=0.19) events versus placebo or comparator therapies. There was no evidence of a reduction in the risk of fatal stroke with these classes versus placebo or comparator therapies. For all classes, no evidence of statistical heterogeneity, I2=0.00. The GLP1-RAs may reduce the risk of non-fatal stroke, but there was no evidence of a reduction in stroke rate with the DPP4-1s or SGLT2-Is agents. These findings may help physicians guide treatment of T2DM in people at risk of stroke.
Next, I explored patients’ views regarding the use of pioglitazone therapy after stroke in a study called the PIOSurvey. We performed a structured survey of stroke survivors registered in the Scottish Health Research Registry. They were given a summary of the evidence for pioglitazone use and side effects (SEs) and asked a series of questions covering the following main themes: 1) whether they would be happy to take pioglitazone, 2) which potential SEs were of most concern, 3) which strategies to mitigate risk were most acceptable and 4) whether they would be willing to take part in further clinical trials. A total of 2,335 electronic invites were sent (714 were read, 145 recipients completed the survey). Of the respondents, 96 (66%) were unlikely/very unlikely to take pioglitazone and only 28 (20%) were likely/very likely to take it. Fractures were a significant concern, as was weight gain. A total of 93 respondents (65%) and 124 respondents (85%)) respectively rated fracture and weight gain as very important/important to them compared to 78 (53%) regarding recurrent stroke and myocardial infarction (MI). A total of 108 (75%) supported further research, five (3%) suggested we should not perform more research or use pioglitazone and 32 (22%) felt we should discuss using pioglitazone with patients now. Stroke patients were concerned regarding pioglitazone SEs, rating these as more important than the risk of stroke or MI recurrence. They would be unlikely to take pioglitazone but support further studies. Further trials to refine the risk/benefit ratio are needed.
I also conducted focus groups to assess the view of patients, caregivers and clinicians regarding the post-stroke management of diabetes (‘exploring the barriers to effective Co-morbidity Management POSt-strokE with Diabetes Focus’: The COMPOSEd study). This study employed a qualitative exploratory design using the grounded theory approach that contributed to our exploration of the use of new treatments for diabetes after stroke and the acceptability of pioglitazone treatment for post-stroke management to minimise its associated SEs and to capture and analyse the barriers to its adoption among primary and secondary stroke care services. Three focus groups were held between February and July 2018 with 19 participants from Glasgow, Scotland and Newport, Wales. Typically, the stroke survivors felt the risk of stroke to be more important than treatment-related SEs (weight gain and fractures) and seemed more willing to accept these risks rather than further disability. The professionals felt they had gaps in their knowledge regarding new drugs for post-stroke diabetes management. Although stroke patients appeared more willing to accept pioglitazone-induced SEs than a higher risk of stroke, the healthcare professionals were more concerned with the drug-associated SEs. The healthcare professionals were also likely to modify their clinical practice based on standardised recommendations or guidelines. Healthcare professionals suggested exploration of methods to reduce pioglitazone-associated SEs such as better individual selection, dose reduction, professionals' monitoring of early signs of SEs, the early use of preventive management for fractures and lifestyle measures and modification. Healthcare providers believed these might allow more treatment adherence and confidence in their decision making and practice. These findings suggest that further research is needed before widespread use of pioglitazone and that education is needed regarding use of new diabetes drugs.
In summary, pioglitazone use is associated with a clinically significant relative risk of fractures, but such fracture events may not all be serious and thus may not be as important as preventing further recurrence of stroke or MI. However, the review was limited by the reliability of the included studies, which used a broad range of fracture definitions and included little data on people with previous strokes. Compared with DPP4-1s and SGLT2-Is, the GLP1-RAs use may reduce non-fatal stroke risk. Although some views expressed in the focus groups differed from the survey findings, both patients and clinicians are reluctant to use pioglitazone to prevent recurrent stroke at present. Some guidelines suggest, however, that pioglitazone use can be considered in individual patients after considering the risks and benefits. Further research is needed to optimise the risk/benefit ratio of this use and to identify people for whom the benefits would be largest and the risk acceptable. The findings of this thesis confirm the need for such studies.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Stroke, Insulin resistance, Type 2 diabetes mellitus, Pioglitazone, Antidiabetic drugs (dipeptidyl peptidase-4 inhibitors (DPP4-Is), glucagon-like peptide-1 receptor agonist (GLP1-RAs) and sodium-glucose cotransporter-2 inhibitors (SGLT2-Is)).
• Subjects: R Medicine > R Medicine (General); R Medicine > RM Therapeutics. Pharmacology
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Cardiovascular & Metabolic Health > Cardiovascular & Metabolic Health; College of Medical Veterinary and Life Sciences > School of Cardiovascular & Metabolic Health > Stroke/Cerbovascular Disease and Care of the Elderly
• Supervisor's Name: Dawson, Professor Jesse and Haig, Dr. Caroline
• Date of Award: 2020
• Depositing User: Dr Hala Azhari
• Unique ID: glathesis:2020-81493
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 29 Jun 2020 16:04
• Last Modified: 29 Jul 2026 08:32
• Thesis DOI: 10.5525/gla.thesis.81493
• URI: https://theses.gla.ac.uk/id/eprint/81493