Efficacy and safety of minocycline on patients with acute anterior circulation ischaemic stroke undergoing mechanical thrombectomy (MIST-A): a multicentre, prospective, randomised, open-label, blinded-endpoint, phase 2 trial Authors: Xiao Zhang, Dong Wei, Yang Li, Huiping Zhang, Li Yao, Xiangjun Yuan, Wen Liu, Xingshun Ma, Bin Wang, Na Qin, Deshuai Li, Rui Shi, Zhirong Fan, Xuan Wang, Luojun Wang, Zhirong Liu, Ling Wang, Duolao Wang, Jingjing Zhao, Wen Jiang, The MIST-A Study Group Published: 2026 ABSTRACT SummaryBackgroundMinocycline has anti-inflammatory and neuroprotective properties. Prior clinical trials have shown potential benefits in acute ischaemic stroke (AIS), but its efficacy in patients who achieve successful reperfusion with mechanical thrombectomy (MT) remains unknown.MethodsThis multicentre, prospective, randomised, open-label, blinded-endpoint, phase 2 trial was conducted at eight hospitals in China. Patients with ischaemic stroke due to anterior circulation large-vessel occlusion who achieved successful recanalisation after MT were randomly assigned (1:1) via a centralised web-based system to receive either oral minocycline (100 mg twice daily for 5 days) as an adjunct to standard care or standard care alone. The primary outcome was the infarct growth ratio (day 5 infarct volume divided by baseline infarct volume), analysed in the modified intention-to-treat population (mITT). This trial was registered at ClinicalTrials.gov, NCT05487417.FindingsBetween Nov 21, 2022, and June 9, 2025, 189 patients were randomly assigned. After excluding two patients due to major protocol violations, 187 patients were included in the mITT analysis (94 minocycline, 93 control). The median day 5 infarct growth ratio was 1.8 (IQR 1.3–3.3) in the minocycline group and 1.6 (IQR 1.2–2.8) in the control group, yielding an adjusted geometric mean ratio of 1.10 (95% CI 0.84–1.45; p = 0.49). At day 90, 68.1% (64/94) in the minocycline group versus 72.0% (67/93) in the control group achieved functional independence (modified Rankin Scale score 0–2; adjusted risk ratio 1.00, 95% CI 0.70–1.43). There were no significant differences between the minocycline group and control group in the incidence of adverse events (76.6% [72/94] vs 68.8% [64/93]) or serious adverse events (25.5% [24/94] vs 25.8% [24/93]).InterpretationIn patients with AIS successfully treated with MT, adjuvant minocycline did not reduce infarct growth. The safety profile of minocycline was similar to that of standard care. These findings do not support the use of minocycline as a neuroprotective therapy in this specific population of patients.FundingNoncommunicable Chronic Diseases-National Science Technology Major Project; Shaanxi Province Special Support Program for Leading Talents in Scientific and Technological Innovation; Xijing Hospital Medical Personnel Training and Development Program; Shaanxi Provincial Health Commission Cerebrovascular Disease Scientific Research Innovation Team.Keywords: Minocycline, Acute ischaemic stroke, Mechanical thrombectomy, Anterior circulation Research in context Evidence before this study We searched PubMed from database inception to Jan 31, 2026, for randomised trials assessing the efficacy of minocycline in patients with acute ischaemic stroke, using the terms “minocycline” AND “stroke”, with no language restrictions. This literature search identified a meta-analysis of three randomised trials suggesting that minocycline significantly improved functional independence at 90 days, along with four small-scale randomised trials that reported inconsistent findings, and the recent EMPHASIS trial, a large-scale, double-blind, randomised controlled trial involving 1724 patients with acute ischaemic stroke, which showed that a 5-day course of oral minocycline significantly increased the proportion of patients achieving excellent functional outcome at 90 days compared with placebo. However, these trials predominantly enrolled patients who did not undergo mechanical thrombectomy (MT) or achieve early successful recanalisation. Whether minocycline provides additional benefit specifically in patients who have achieved successful reperfusion with MT remained unknown. Added value of this study To our knowledge, the MIST-A trial is the first randomised controlled trial to evaluate minocycline exclusively in patients with acute anterior circulation ischaemic stroke who achieved successful recanalisation after MT. In this multicentre trial involving 187 patients with large-vessel occlusion and successful reperfusion, we found that a 5-day course of oral minocycline did not reduce infarct growth compared with standard care alone. Consistently, no improvement was observed in 90-day functional outcomes, including functional independence. Safety outcomes were similar between groups. Implications of all the available evidence Earlier small-scale trials and meta-analyses suggested that minocycline may improve functional outcomes after acute ischaemic stroke. The subsequent EMPHASIS trial confirmed this benefit in a broader population of patients predominantly without early successful recanalisation and with milder stroke severity. In contrast, the MIST-A trial found no benefit of adjunctive minocycline on infarct growth or 90-day functional outcomes in patients with acute ischaemic stroke who achieved successful reperfusion after MT. Taken together, while minocycline may offer benefit in the general population of acute ischaemic stroke patients, this benefit may not extend to the specific subgroup of patients who have undergone successful MT. Further researches are needed to understand the mechanisms underlying these different results and to identify which stroke populations are most likely to benefit from anti-neuroinflammatory therapies. Introduction Mechanical thrombectomy (MT) is a primary treatment for acute ischaemic stroke (AIS) caused by large vessel occlusion (LVO), achieving successful recanalisation in 85–97% of cases.1 Nevertheless, up to half of those with successful recanalisation fail to attain functional independence, a condition defined as futile recanalisation.1, 2, 3 This discrepancy is attributed to several pathophysiological mechanisms, including microvascular no-reflow, blood–brain barrier disruption with haemorrhagic transformation, and ongoing neuronal injury despite macrovascular reperfusion.4,5 Notably, post-stroke inflammation plays an important role in driving these pathological processes.6 These findings highlight the potential of targeting post-ischaemic inflammation as a promising adjunctive strategy to improve functional recovery in AIS treated with MT. Minocycline, a tetracycline antibiotic with intrinsic anti-inflammatory properties, has demonstrated therapeutic efficacy in inflammatory conditions such as rheumatoid arthritis and acne rosacea.7 It also delays disease progression in clinically isolated syndrome and promotes functional recovery following spinal cord injury.8,9 In experimental models of cerebral ischaemia, minocycline confers neuroprotection primarily through suppression of microglial activation, a key driver of post-ischaemic neuroinflammation,10, 11, 12, 13 and inhibition of matrix metalloproteinases, thereby preserving blood–brain barrier integrity and reducing haemorrhagic transformation.7 Although several clinical trials, most notably the recent EMPHASIS trial, have suggested that minocycline improves outcomes in AIS,14, 15, 16, 17 these trials predominantly enrolled patients without early successful recanalisation. Whether minocycline confers additional benefit in patients who achieve successful reperfusion via MT therefore remains unclear. To address this gap, we conducted a multicentre, randomised clinical trial to evaluate the efficacy and safety of adjunctive minocycline in AIS patients with successful reperfusion following MT. Methods Study design and ethics approval The MIST-A trial was a multicentre, prospective, randomised, open-label, blinded-endpoint, phase 2 Trial. It was conducted across eight hospitals in Northwest China. The trial enrolled participants between November 21, 2022, and June 9, 2025, with follow-up completed by September 9, 2025. The trial protocol was approved by the Ethics Committee of Xijing Hospital (XJLL-KY20222186) on July 20, 2022. The study was subsequently registered at ClinicalTrials.gov (NCT05487417) on July 30, 2022, and conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. The study was conducted in accordance with the approved protocol (Appendix pp 29–57), and its design and rationale have been published elsewhere.18 Written informed consent was obtained from all participants or their legally authorised representatives before enrolment. Patients Patients were eligible for enrolment if they were 18–85 years of age; presented with AIS due to occlusion of the intracranial internal carotid artery (ICA), M1 or M2 segment of the middle cerebral artery (MCA), or A1 segment of the anterior cerebral artery (ACA), with occlusion confirmed by digital subtraction angiography (DSA); had a baseline National Institutes of Health Stroke Scale (NIHSS) score between 6 and 30; and were planned for MT within the appropriate time window. For patients presenting within 6 h of symptom onset, an Alberta Stroke Program Early CT Score (ASPECTS) of ≥6 was required. For those presenting between 6 and 24 h, eligibility was determined by multi-modal imaging. Key exclusion criteria included contraindications to MT, failure to achieve successful reperfusion, pre-existing disability or dementia, severe renal or hepatic impairment, known allergy to tetracyclines, ongoing infection requiring antibiotics and pregnancy. A complete list of inclusion and exclusion criteria is provided in the eTable S1. Successful recanalisation was defined as a modified Thrombolysis in Cerebral Infarction (mTICI) score of 2b or 3.19 Randomisation and masking Patients with anterior circulation AIS-LVO who had undergone successful MT were randomised in a 1:1 ratio to receive either minocycline group (adjuvant minocycline in addition to standard care) or control group (standard care alone). All patients received guideline-based management prior to randomisation, including intravenous alteplase and MT, as determined by the treating physician.20 Randomisation was performed after confirmation of successful recanalisation with MT. An independent statistician generated the allocation sequence using SAS 9.4 software, with permuted block randomisation using a fixed block size of 4, stratified by study centre. The allocation sequence was concealed within a centralised, web-based randomisation system (http://www.epiedc.com/kf.php). At each site, an independent investigator accessed this system to obtain the treatment assignment for eligible participants, ensuring allocation concealment. The allocation was unblinded to the physicians and participants. However, all investigators involved in the adjudication of the primary imaging endpoint (central laboratory), the assessment of clinical secondary endpoints (functional outcome assessors), and the statistical analysis were blinded to the group allocation throughout the trial. Procedures All enrolled patients underwent MT in accordance with current guideline recommendations.20 The procedure was performed using stent retrievers, aspiration devices, or a combination of both, at the discretion of the local neurointerventional team. Adjunctive procedures, including balloon angioplasty or stent implantation for underlying atherosclerotic stenosis, were permitted. In minocycline group, patients were prescribed with oral minocycline (Hanhui Pharmaceutical Co., Shanghai, China; 100 mg per capsule) at a dosage of 200 mg as a first dose, administered within 1 h after recanalisation, followed by 100 mg twice daily for five consecutive days. For patients who were unable to swallow due to impaired consciousness or dysphagia, the study drug was administered via nasogastric tube. In control group, patients were prescribed with standard medical care alone. Both the interventional groups received standard care following guidelines as detailed in the previously published trial protocol.18,20 Data collection Before the initiation of MT, data on demographics, medical history, physical examination findings, vital signs, laboratory tests, and scores on the modified Rankin Scale (mRS), Barthel Index (BI), and NIHSS were recorded, along with magnetic resonance imaging (MRI) results (T1-weighted imaging, T2-weighted imaging, and diffusion-weighted imaging [DWI]). At 24 h after MT, a non-contrast computed tomography (CT) scan was performed to assess for intracranial Haemorrhage (ICH). At day 5, a follow-up MRI, including T1-weighted, T2-weighted, and fluid-attenuated inversion recovery (FLAIR) sequences, was conducted to determine the final infarct volume. Laboratory tests were repeated on day 3 to monitor for safety events, and additionally at any time if considered necessary by the physician. The NIHSS, mRS, and BI scores were collected on days 1, 3, 5, 7, and 90 after stroke onset to monitor clinical outcomes. Adverse events (AEs) and serious adverse events (SAEs) were recorded throughout the hospitalisation and follow-up periods. MRI analysis Imaging analysis methods have been described previously.18 Briefly, for the segmentation of lesion areas in the FLAIR-weighted and DWI-weighted MRI images of patients with cerebral infarction, the Lesion Prediction Algorithm was utilised. This method is a supervised learning algorithm that combines high-dimensional statistical models with Markov Chain Monte Carlo (MCMC) algorithms.21 All MRI assessments were conducted and analysed at a central facility by personnel blinded to the group assignments. Outcomes The primary efficacy outcome was the evolution of infarct volume from baseline DWI to day 5 FLAIR. This change was specifically defined as the relative growth in infarct volume, calculated as the ratio of the final infarct volume at day 5 (V5) to the baseline infarct volume (V0) for each patient, as previously reported.22 Secondary efficacy outcomes included distribution of scores on mRS, functional independence (mRS 0–2), excellent recovery (mRS 0–1), BI at day 90, changes in NIHSS scores from baseline to day 7, any ICH detected on 24-h post-procedural CT, all-cause mortality at day 90, final infarct volume at day 5, and length of hospital and intensive care unit stay. Safety outcomes comprised all AEs and SAEs. The post-hoc secondary efficacy outcomes were primary efficacy outcome without imputation and NIHSS score at day 7. Further details are provided in the study protocol (see Appendix pp 29–57). Statistical analysis Preclinical studies have shown that minocycline can reduce the infarct volume by 41%–45% within 2–28 days after cerebral infarction.11, 12, 13 In light of previous literature,22 the academic committee has designed the study with sufficient statistical power to detect a mean logarithmic reduction in the relative growth of infarct volume from baseline of 0.40 (one-third decrease in relative growth) with a standard deviation (SD) of 0.9.22 The preliminary estimation of the sample size was determined with an alpha level (α) of 0.05 and a power of 1-β (beta) of 80%. Accounting for a 10% loss to follow-up rate, the final sample size was calculated to be 180 cases. The primary efficacy endpoint was quantified as the log-relative growth, calculated as ln (V5/V0), where V0 and V5 represent the infarct volumes at baseline and day 5, respectively. The primary analysis was performed using a linear mixed-effects model in the modified intention-to-treat (mITT) population. The mITT population was defined as all randomised patients except those with major eligibility violations, those who did not receive the assigned treatment, or those with no post-randomisation data. The model included treatment group as a fixed effect, study centre as a random effect. A prespecified per-protocol (PP) analysis was conducted as a sensitivity analysis to assess the robustness of the primary findings. Subgroup analyses were performed exploratively for the primary endpoint across the following baseline factors: age (≥65 vs. <65 years), sex, baseline infarct volume (>median vs. ≤median), baseline NIHSS score (>median vs. ≤median), intravenous thrombolysis (yes vs. no), and time from onset to puncture (>median vs. ≤median). These same factors were included as covariates in all adjusted analyses. Missing data for the primary endpoint were handled using multiple imputation. A MCMC method with a fixed random seed of 123 was used to generate ten imputed datasets, and estimates were combined using Rubin's rules. The ordinal analysis of mRS at day 90 used a proportional-odds mixed model to calculate common odds ratio. Binary secondary outcomes were analysed using Poisson generalised linear mixed models to calculate risk ratio (RR). Continuous secondary endpoints were analysed with linear mixed-effects models to estimate geometric mean ratio (GMR). All models accounted for study centre as a random effect. In analyses of functional outcomes, patients who died were assigned the worst possible scores (NIHSS = 42, BI = 0, mRS = 6) at all subsequent time points. Safety analyses were performed on the safety population. AEs were summarised and compared between treatment groups. No interim analysis was planned or performed in this trial. All statistical tests were two-sided, with a significance level of α = 0.05. Analyses were conducted using SPSS (version 25.0) and R (version 4.4.1). A detailed description of the statistical methods is provided in the statistical analysis plan (see Appendix pp 77–87). Role of the funding source The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Patients and the public were also not involved in the study design. Results A total of 543 patients were assessed for eligibility between November 21, 2022 and June 9, 2025, of whom 189 underwent randomisation (eTable S2, eFigure S1). After excluding two patients who had major protocol violations of the key eligibility criteria, 187 patients were included in the final modified mITT analysis, consisting of 94 in the minocycline group and 93 in the control group (Fig. 1). The demographic and clinical characteristics at baseline were similar between the two groups in both the mITT population (Table 1) and the PP population (eTable S3). Missing data for the primary endpoint were imputed for 19 patients (14 without day 5 FLAIR imaging and five with FLAIR images of insufficient quality). Trial profile. Table 1Baseline characteristics (modified intention-to-treat population).Control (N = 93)Minocycline (N = 94)Age, years70.0 (61.0–76.0)66.0 (58.0–74.0)Sex Male69 (74.2%)57 (60.6%) Female24 (25.8%)37 (39.4%)BMI, kg/m224.2 (22.2–25.9)23.9 (21.9–26.0)Medical history Atrial fibrillation36 (38.7%)45 (47.8%) Hypertension59 (63.4%)58 (61.7%) Diabetes21 (22.5%)16 (17.0%) Coronary heart disease24 (25.8%)22 (23.4%) Previous ischaemic stroke15 (16.1%)14 (14.8%) Ever smoker35 (37.6%)31 (32.9%) Alcohol use13 (13.9%)11 (11.7%)NIHSS score at baseline12.0 (9.0–16.0)12.0 (9.0–16.0)Infarct volume at baseline, mL9.0 (3.1–26.3)10.5 (5.1–28.5)Occlusion site Internal carotid artery20 (21.5%)26 (27.6%) Anterior cerebral artery (A1)1 (1.0%)1 (1.0%) Middle cerebral artery (M1)63 (67.7%)55 (58.5%) Middle cerebral artery (M2)1 (1.0%)3 (3.1%) Tandem lesion8 (8.6%)9 (9.5%)Intravenous thrombolysis24 (25.8%)36 (38.3%)Time from stroke onset, mins To hospital292.0 (148.0–558.0)238.0 (120.0–478.0) To puncture410.0 (260.0–657.0)373.0 (262.0–630.0) To randomisation630.0 (400.0–832.0)540.0 (396.0–820.0)mTICI after MT mTICI 381 (87.1%)86 (91.5%) mTICI 2b12 (12.9%)8 (8.5%)Data are median (IQR) or n (%), unless otherwise indicated.BMI = Body Mass Index, NIHSS = National Institutes of Health Stroke Scale, mTICI = Modified Thrombolysis in Cerebral Infarction, MT = Mechanical Thrombectomy. Baseline characteristics (modified intention-to-treat population). Data are median (IQR) or n (%), unless otherwise indicated. BMI = Body Mass Index, NIHSS = National Institutes of Health Stroke Scale, mTICI = Modified Thrombolysis in Cerebral Infarction, MT = Mechanical Thrombectomy. In the mITT population, the NIHSS score was comparable between groups, with a median of 12.0 (IQR, 9.0–16.0) in both the minocycline and control groups. The baseline infarct volume was 9.0 mL (IQR, 3.1–26.3) in the minocycline group and 10.5 mL (IQR, 5.1–28.5) in the control group. Intravenous thrombolysis was administered prior to MT in 36 (38.3%) patients in the minocycline group and 24 (25.8%) in the control group. Regarding the occlusion site, the most frequently affected vessel was the M1 segment of the MCA, observed in 55 patients (58.5%) in the minocycline group and 63 (67.7%) in the control group. The second most common occlusion site was the ICA, with 26 (27.6%) in the minocycline group and 20 (21.5%) in the control group (Table 1). Details regarding MT are provided in the eTable S4. In the minocycline group, the initial loading dose was administered orally in 84 patients (89.4%) and via nasogastric tube in ten patients (10.6%); further details on study drug administration are provided in eTable S5. As shown in Table 2, the median day 5 infarct growth ratio was 1.8 (IQR 1.3–3.3) in the minocycline group and 1.6 (IQR 1.2–2.8) in the control group, with the GMR of 1.07 (95% confidence interval [CI] 0.82–1.40; p = 0.63). The adjusted analysis also indicated that the application of minocycline did not inhibit the infarct growth ratio, with the GMR of 1.10 (95% CI 0.84–1.45; p = 0.49) (Table 2). The sensitivity analysis of the infarct growth ratio in PP population yielded a consistent result with the mITT analysis (eTable S6). A post-hoc sensitivity analysis of infarct growth ratio performed on the mITT population without imputation for missing data was also consistent with the mITT analysis (eTable S7). Fig. 2 demonstrated no significant difference in the infarct growth ratio in terms of each prespecified subgroup between minocycline group and control group. Notably, in the PP population, the final infarct volume at day 5 was larger in the minocycline group than in the control group (adjusted GMR 1.71, 95% CI 1.14–2.55; eTable S6). Table 2Efficacy and safety outcomes (modified intention-to-treat population).Control (n = 93)Minocycline (n = 94)Measure of effectUnadjusted effect (95% CI)Adjusted effecta (95% CI)Primary outcome Infarct volume growth ratio at day 5b,c1.6 (1.2–2.8)1.8 (1.3–3.3)Geometric mean ratio1.07 (0.82–1.40)1.10 (0.84–1.45)Secondary outcomes mRS score at day 90dCommon odds ratio1.13 (0.68–1.88)1.02 (0.59–1.75) mRS score 0–1 at day 90e50/93 (53.7)48/94 (51.0)Risk ratio0.95 (0.64–1.41)0.97 (0.64–1.48) mRS score 0–2 at day 90e67/93 (72.0)64/94 (68.1)Risk ratio0.95 (0.67–1.33)1.00 (0.70–1.43) NIHSS score at day 7c4.0 (2.0–9.0)5.0 (1.0–11.0)Geometric mean ratio1.06 (0.81–1.38)1.01 (0.81–1.26) BI score at day 90c100.0 (75.0–100.0)97.5 (70.0–100.0)Geometric mean ratio0.93 (0.68–1.28)0.94 (0.69–1.29) Infarct volume at day 5, mLc14.8 (5.3–53.4)23.9 (9.8–59.9)Geometric mean ratios1.65 (1.10–2.46)1.28 (0.92–1.77) Hospital stay, daysc11.0 (8.0–14.0)11.0 (8.2–14.0)Geometric mean ratio1.03 (0.91–1.17)0.99 (0.88–1.12) ICU stay, daysc4.0 (3.0–8.0)5.0 (3.0–8.0)Geometric mean ratio1.13 (0.91–1.41)1.12 (0.91–1.38) Any ICH at 24he,f20/91 (21.9)18/94 (19.1)Risk ratio0.87 (0.46–1.65)0.80 (0.42–1.55) Mortality at day 90e3/93 (3.2)5/94 (5.3)Risk ratio1.65 (0.39–6.90)1.59 (0.33–7.71)Safety outcomes AEs at day 90e64/93 (68.8)72/94 (76.6)Risk ratio1.11 (0.80–1.56)1.08 (0.76–1.54) SAEs at day 90e24/93 (25.8)24/94 (25.5)Risk ratio0.99 (0.56–1.74)0.91 (0.50–1.64)Data are n/N (%) or median (IQR), unless otherwise indicated.mRS = Modified Rankin Scale, NIHSS = National Institutes of Health Stroke Scale, BI = Barthel Index, ICU = Intensive Care Unit, ICH = Intracerebral Haemorrhage, AE = Adverse Event, SAE = Serious Adverse Event, DWI = Diffusion-Weighted Imaging, FLAIR = Fluid Attenuated Inversion Recovery, GLMM = Generalised Linear Mixed Model.aFully adjusted models included age, sex, baseline infarct volume, baseline NIHSS, intravenous thrombolysis, and onset-to-puncture time as covariate. For the primary endpoint event, since the observed values are calculated as V5/V0, the baseline infarct volume has already been accounted for. Therefore, baseline infarct volume is not included as a covariate for the primary endpoint event.bThe “Infarct volume growth ratio at day 5” is defined as the FLAIR infarct volume on Day 5 divided by the baseline DWI volume. Inferential analyses were performed on log-transformed ratios (ln [V_Day5/V_Day0]).cCentre-random-intercept linear mixed model, treatment group as study variable, and study centre as random effect. To meet the normality assumption of the linear mixed-effects model, continuous variables were natural log-transformed.dProportional-odds mixed model on day 90 mRS grades (0–6) with centre random effect.ePoisson GLMM with log link and centre random effect.fAny intracranial haemorrhage detected on 24-h post-procedural CT. Efficacy and safety outcomes (modified intention-to-treat population). Data are n/N (%) or median (IQR), unless otherwise indicated. mRS = Modified Rankin Scale, NIHSS = National Institutes of Health Stroke Scale, BI = Barthel Index, ICU = Intensive Care Unit, ICH = Intracerebral Haemorrhage, AE = Adverse Event, SAE = Serious Adverse Event, DWI = Diffusion-Weighted Imaging, FLAIR = Fluid Attenuated Inversion Recovery, GLMM = Generalised Linear Mixed Model. Fully adjusted models included age, sex, baseline infarct volume, baseline NIHSS, intravenous thrombolysis, and onset-to-puncture time as covariate. For the primary endpoint event, since the observed values are calculated as V5/V0, the baseline infarct volume has already been accounted for. Therefore, baseline infarct volume is not included as a covariate for the primary endpoint event. The “Infarct volume growth ratio at day 5” is defined as the FLAIR infarct volume on Day 5 divided by the baseline DWI volume. Inferential analyses were performed on log-transformed ratios (ln [V_Day5/V_Day0]). Centre-random-intercept linear mixed model, treatment group as study variable, and study centre as random effect. To meet the normality assumption of the linear mixed-effects model, continuous variables were natural log-transformed. Proportional-odds mixed model on day 90 mRS grades (0–6) with centre random effect. Poisson GLMM with log link and centre random effect. Any intracranial haemorrhage detected on 24-h post-procedural CT. Subgroup analyses of the primary outcome (modified intention-to-treat population). §Centre-random-intercept linear mixed model. ∗Dichotomized by median. yr = year. NIHSS = National Institutes of Health Stroke Scale. DWI = Diffusion-Weighted Imaging. GMR = Geometric Mean Ratio. The proportion of patients achieving functional independence (mRS 0–2) was 68.1% (64/94) in the minocycline group versus 72.0% (67/93) in the control group (adjusted RR 1.00, 95% CI 0.70–1.43). The incidence of excellent outcome (mRS 0–1) was 51.0% (48/94) in the minocycline group versus 53.7% (50/93) in the control group (adjusted RR 0.97, 95% CI 0.64–1.48) (Table 2). The distribution of day 90 mRS scores is shown in Fig. 3, which did not demonstrate a favourable shift in mRS among patients treated with minocycline compared with the control group (adjusted common odds ratio 1.02, 95% CI 0.59–1.75) (Table 2). Analyses of functional outcomes in the PP population were consistent with those in the mITT population (eFigure S2, eTable S6). Distribution of mRS Scores at day 90 (modified intention-to-treat population). A score of 0 on the mRS indicates no symptoms, a score of 1 indicates no clinically significant disability, a score of 2 indicates slight disability, a score of 3 indicates moderate disability, a score of 4 indicates moderately severe disability, a score of 5 indicates severe disability, and a score of 6 indicates death. mRS = modified Rankin Scale. No significant differences were observed between the minocycline and the control groups in the incidence of ICH at 24 h (19.1% [18/94] vs 21.9% [20/91]; adjusted RR 0.80, 95% CI 0.42–1.55) or all-cause mortality at 90 days (5.3% [5/94] vs 3.2% [3/93]; adjusted RR 1.59, 95% CI 0.33–7.71), with similar findings in the PP population (Table 2, eTable S6). The changes in NIHSS scores from baseline to day 7 are presented in eFigure S3. Over the first 7 days following enrolment, neurological function improved significantly in both treatment groups (p < 0.001 for time effect); however, no significant between-group difference was observed (p = 0.85 for group effect). At day 7, the median NIHSS score was 4.0 (IQR, 2.0–9.0) in the control group and 5.0 (IQR, 1.0–11.0) in the minocycline group, with no significant difference between groups (adjusted GMR 1.01, 95% CI 0.81–1.26; Table 2). These findings were consistent in the PP population (eFigure S3, eTable S6). For other secondary outcomes, including the BI at day 90 and the length of hospital and intensive care unit stay, no statistically significant differences were observed between groups in either the mITT or PP populations (Table 2 and eTable S6). The incidence of any AEs (76.6% [72/94] vs 68.8% [64/93]; adjusted RR 1.08, 95% CI 0.76–1.54) and any SAEs (25.5% [24/94] vs 25.8% [24/93]; adjusted RR 0.91, 95% CI 0.50–1.64) was similar between the minocycline group and control group (Table 2; eTable S6). The details of AEs and SAEs are displayed in eTable S8. Discussion In the MIST-A trial, adjuvant minocycline did not reduce the infarct growth in patients with acute anterior circulation stroke who had undergone successful MT. Unexpectedly, the PP analysis revealed that minocycline was associated with larger final infarct volume at day 5. The incidences of any AEs or SAEs did not differ between groups. The futile recanalisation continues to affect up to half of patients undergoing MT, where inflammation plays an important role in the pathological processes.1, 2, 3, 4 Minocycline, a promising neuroprotective candidate through its potential anti-inflammatory effects, has been widely investigated. Recently, the EMPHASIS trial reported that minocycline significantly increased the proportion of excellent functional recovery at 90 days in patients with AIS (52.6% vs 47.4%; adjusted RR 1.11, 95% CI 1.03–1.20).17 However, the MIST-A trial did not show the similar benefit of minocycline in patients with AIS after successful MT, and even suggested an increase in the final infarct volume. The inconsistency between the trials might be explained by differences in study design and patient population. The EMPHASIS trial predominantly enrolled patients without early successful recanalisation, allowed enrolment up to 72 h after onset, and included patients with mild-to-moderate stroke severity (median NIHSS 5). Nevertheless, MIST-A exclusively enrolled patients with successful reperfusion after MT, and included patients with more severe neurological deficits (median NIHSS 12). Thrombectomy and severe cerebral infarction may lead to more complex pathophysiological mechanisms after stroke, potentially altering the effect of minocycline. Specifically, several factors may have influenced the results of the MIST-A trial. First, the futile recanalisation is driven by multiple interconnected mechanisms beyond neuroinflammation alone, including microvascular no-reflow, early re-occlusion, poor collateral status, haemorrhagic transformation, and impaired autoregulation.6 Although inflammation contributes to several of these processes, an anti-inflammatory intervention alone may be insufficient to modify such a complex, multifactorial pathology. Consistent with this, the MARVEL trial, which tested adjunctive methylprednisolone in similar population, also did not improve overall disability.23 The parallel findings of two anti-inflammatory strategies in the same clinical context suggest that anti-inflammatory approaches alone may not yield measurable benefit after successful recanalisation. Second, the net effect of microglial modulation in the context of successful reperfusion or severe stroke is uncertain. Preclinical studies indicate that minocycline can shift microglia from a pro-inflammatory M1 towards an anti-inflammatory M2 phenotype in experimental stroke,12 but this has not been confirmed clinically. In a randomised trial of traumatic brain injury, minocycline altered microglial activity yet was associated with increased neurodegeneration and worse functional recovery.24 Whether a similar phenomenon occurs after successful thrombectomy or severe stroke remains unknown. Beyond these mechanistic considerations, several aspects of the trial design may also have affected the likelihood of detecting a treatment effect. In our trial, the first dose of minocycline was administered after successful recanalisation, not before MT. Moreover, enrolment was allowed up to 24 h after onset in patients selected by multi-modal imaging. Consequently, the study population included a substantial proportion of slow progressors with favourable collateral circulation. The ESCAPE-NEXT trial of nerinetide, which reported a neutral outcome,25 cited delayed treatment initiation and the inclusion of slow progressors as potential contributors to its neutral findings. These factors may therefore also be relevant when interpreting the findings of MIST-A. Several aspects of the study merit attention. First, the minocycline regimen selected for this trial was a 200 mg loading dose administered orally followed by 100 mg twice daily for 5 days. This regimen is consistent with that used in previous positive trials of minocycline in AIS,14, 15, 16, 17 and the 5-day treatment duration was chosen because infarct volume typically reaches its maximum by day 5 after stroke onset.22 Second, regarding outcome selection, we chose infarct growth as the primary endpoint. This decision was based on well-established evidence linking smaller infarct volumes to improved functional outcomes26,27 and on the acceptance of infarct volume as a surrogate endpoint in neuroprotection trials.22 In our study, the point estimates for both the primary imaging endpoint and the secondary clinical outcomes were neutral and numerically favoured the control group, although none of the differences reached statistical significance. The consistency across imaging and functional measures supports the internal validity of the primary result. However, it should be noted that infarct growth is a surrogate endpoint and may not fully substitute for clinical outcomes. Moreover, the sample size was determined based on the surrogate endpoint and was not powered for definitive assessment of functional recovery. Third, the rate of intravenous thrombolysis in our trial was relatively low, with only 60 patients (32%). The longer time window for enrolment than for thrombolysis resulted in a high proportion of enrolled patients being ineligible for thrombolysis. Additionally, some had contraindications, while others declined after weighing the risks against the benefits. To our knowledge, the MIST-A study is the first randomised trial to assess minocycline exclusively in AIS patients who achieved successful reperfusion after MT, thereby providing crucial complementary evidence to the existing literature on this agent. Nevertheless, several limitations should be acknowledged. First, the open-label design may have introduced bias in routine care or patient reporting, although imaging and clinical outcomes were assessed by blinded independent researchers. Second, we excluded patients with baseline NIHSS scores above 30, which limits the generalisability of our findings to the most severe strokes. Third, all participants were recruited from Northwest China, a region with distinct demographic and lifestyle characteristics; whether our findings apply to other populations requires further investigation. Fourth, infarct growth is a surrogate endpoint and cannot fully substitute for clinical outcomes, and the modest sample size may have limited power to detect smaller treatment effects. Fifth, the assessment of infarct volume at day 5 may not fully capture the complete evolution of tissue injury, as later time points could reveal additional changes. Finally, we did not collect pharmacokinetic data or inflammatory biomarkers, which prevents a definitive mechanistic explanation for the findings. Future trials should include such measurements to better explore the biological effects of minocycline in this setting. In conclusion, adjunctive oral minocycline did not reduce infarct growth in patients with acute anterior-circulation ischaemic stroke who achieved successful reperfusion after MT. The present findings do not support extending the clinical use of minocycline as a neuroprotective adjunct in this specific setting. Larger trials with optimised dosing regimens are needed to further evaluate any potential benefit in this population. Contributors WJ conceived and designed the study. WJ and JZ provided the main administrative, technical, and material support. XZ and YL were the trial managers, contributing to the promotion of the study, the training of personnel, and acquisition and interpretation of data. DW, HZ, LY, WL, BW, NQ, XY, and XM were the regional principal investigators who contributed to study design, technical and material support, enrolment of patients, data collection, and integrity of the data at each site. All authors had access to all the data in the study. WJ, ZL, JZ, and ZF verified the data. WJ was responsible for the decision to submit the manuscript. LingW and DWa gave support to statistical analysis and interpretation of the data. XZ, XW, RS, DL, LW, and WJ drafted the manuscript with input from all authors and with no external writing assistance. All authors have approved the final version of the manuscript to be released. All authors have consented to be accountable for all aspects of the work and in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Data sharing statement Data from the MIST-A trial are not publicly available but are planned to be made available in the future. A complete de-identified data set will be made accessible. Requests for access to the data can be made by sending an email together with a research plan to jiangwen@fmmu.edu.cn. Declaration of interests Minocycline for this trial was provided by Xijing Hospital. The manufacturer had no role in study design, data collection, data analysis, data interpretation, or writing of the report. WJ reports grant support from the Noncommunicable Chronic Diseases—National Science and Technology Major Project (Grant Number: 2024ZD0527800), Shaanxi Province Special Support Program for Leading Talents in Scientific and Technological Innovation (Wen Jiang), Shaanxi Provincial Health Commission Cerebrovascular Disease Scientific Research Innovation Team (2023TD-07); ZL reports grant support from Xijing Hospital Medical Personnel Training and Development Program (Grant Number: XJZT24LY20). All other authors declare no competing interests. Acknowledgements We thank the investigators and research staff at all participating sites for their dedication and hard work. We sincerely acknowledge all patients and their families for their participation in this study. This research was supported by grants from the Noncommunicable Chronic Diseases—National Science and Technology Major Project (Grant Number: 2024ZD0527800), Shaanxi Province Special Support Program for Leading Talents in Scientific and Technological Innovation (Wen Jiang), Shaanxi Provincial Health Commission Cerebrovascular Disease Scientific Research Innovation Team (2023TD-07), and Xijing Hospital Medical Personnel Training and Development Program (Grant Number: XJZT24LY20). During the preparation of this work, the authors used AI-assisted tools to improve language and readability. No generative AI was used to generate data, interpret results, or formulate conclusions. The authors take full responsibility for the final content. Footnotes Translation: For the Chinese translation of the abstract see the Supplementary Materials section. Supplementary data related to this article can be found at https://doi.org/10.1016/j.lanwpc.2026.101898. Contributor Information Jingjing Zhao, Email: zhaojing0110@fmmu.edu.cn. Wen Jiang, Email: jiangwen@fmmu.edu.cn. Supplementary data Supplementary Tables and Figures Translated Abstract Protocol References Associated Data Supplementary Materials Supplementary Tables and Figures Translated Abstract Protocol