Remote ischaemic conditioning in patients with supratentorial intracerebral haemorrhage (RICH-2): a multicentre, randomised, sham-controlled phase 3 trial in China.

Remote ischaemic conditioning in patients with supratentorial intracerebral haemorrhage (RICH-2): a multicentre, randomised, sham-controlled phase 3 trial in China - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice eClinicalMedicine . 2026 Apr 13;95:103900. doi: 10.1016/j.eclinm.2026.103900 Search in PMC Search in PubMed View in NLM Catalog Add to search Remote ischaemic conditioning in patients with supratentorial intracerebral haemorrhage (RICH-2): a multicentre, randomised, sham-controlled phase 3 trial in China Wenbo Zhao Wenbo Zhao a Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, China b Stroke Research Group, Department of Clinical Neuroscience, University of Cambridge, Cambridge, UK Find articles by Wenbo Zhao a, b, ∗ , Jing Wang Jing Wang a Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, China Find articles by Jing Wang a , Xuebin Yu Xuebin Yu c Department of Neurosurgery, Shaoxing Hospital, Zhejiang University School of Medicine, Zhejiang, China Find articles by Xuebin Yu c , Guang Wu Guang Wu d Department of Neurology, Nanshi Hospital of Nanyang, Henan, China Find articles by Guang Wu d , Song Yang Song Yang e Department of Neurology, The First People's Hospital of Changzhou, Third Affiliated Hospital of Soochow University, Jiangsu, China Find articles by Song Yang e , Da Lu Da Lu f Department of Neurosurgery, Tianjin Huanhu Hospital, Tianjin, China Find articles by Da Lu f , Xinjing Gao Xinjing Gao g Department of Neurosurgery, The Sixth People's Hospital of Hengshui, Hebei, China Find articles by Xinjing Gao g , Yuping He Yuping He h Department of Neurology, Zhuji People's Hospital of Zhejiang Province, Zhejiang, China Find articles by Yuping He h , Lijian Niu Lijian Niu i Department of Neurosurgery, Beijing Fengtai Youanmen Hospital, Beijing, China Find articles by Lijian Niu i , Haihang Zhou Haihang Zhou j Department of Neurosurgery, The Second Affiliated Hospital of Jiaxing University, Zhejiang, China Find articles by Haihang Zhou j , Caixia Xuan Caixia Xuan k Department of Neurology, The Second People's Hospital of Qinzhou City, Guangxi, China Find articles by Caixia Xuan k , Na Li Na Li l Department of Geriatrics, Beijing Renhe Hospital, Beijing, China Find articles by Na Li l , Mengfei Zhong Mengfei Zhong m Department of Neurology, Shengli Oilfield Central Hospital, Shandong, China Find articles by Mengfei Zhong m , Jiangtao Zhang Jiangtao Zhang n Department of Neurology, Chengde Central Hospital, Hebei, China Find articles by Jiangtao Zhang n , Qinglin Dong Qinglin Dong o Department of Emergency, Rizhao People's Hospital, Shandong, China Find articles by Qinglin Dong o , Xiaokun Geng Xiaokun Geng p Department of Neurology, Beijing Luhe Hospital, Capital Medical University, Beijing, China Find articles by Xiaokun Geng p , Qingfeng Ma Qingfeng Ma a Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, China Find articles by Qingfeng Ma a , Haiqing Song Haiqing Song a Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, China Find articles by Haiqing Song a , Yunjian Yin Yunjian Yin q Beijing Institute for Brain Disorders, Capital Medical University, Beijing, China Find articles by Yunjian Yin q , Lan Liu Lan Liu q Beijing Institute for Brain Disorders, Capital Medical University, Beijing, China Find articles by Lan Liu q , David C Hess David C Hess r Departments of Neurology, Medical College of Georgia, Augusta University, Augusta, GA, USA Find articles by David C Hess r , Craig S Anderson Craig S Anderson s The George Institute for Global Health, Faculty of Medicine, University of New South Wales, Sydney, New South Wales, Australia t Neurology Department, Royal Prince Alfred Hospital, Sydney, NSW, Australia u The Institute of Science and Technology for Brain-inspired Research Fudan University, Shanghai, China Find articles by Craig S Anderson s, t, u , Sijie Li Sijie Li v Neurocardiology Centre, Xuanwu Hospital, Capital Medical University, Beijing, China w Beijing Key Laboratory of Hypoxic Conditioning Translational Medicine, Xuanwu Hospital, Capital Medical University, Beijing, China Find articles by Sijie Li v, w, ∗∗ , Xunming Ji Xunming Ji q Beijing Institute for Brain Disorders, Capital Medical University, Beijing, China v Neurocardiology Centre, Xuanwu Hospital, Capital Medical University, Beijing, China Find articles by Xunming Ji q, v, ∗∗∗ ; RICH-2 Investigators x , for the Author information Article notes Copyright and License information a Department of Neurology, Xuanwu Hospital, Capital Medical University, Beijing, China b Stroke Research Group, Department of Clinical Neuroscience, University of Cambridge, Cambridge, UK c Department of Neurosurgery, Shaoxing Hospital, Zhejiang University School of Medicine, Zhejiang, China d Department of Neurology, Nanshi Hospital of Nanyang, Henan, China e Department of Neurology, The First People's Hospital of Changzhou, Third Affiliated Hospital of Soochow University, Jiangsu, China f Department of Neurosurgery, Tianjin Huanhu Hospital, Tianjin, China g Department of Neurosurgery, The Sixth People's Hospital of Hengshui, Hebei, China h Department of Neurology, Zhuji People's Hospital of Zhejiang Province, Zhejiang, China i Department of Neurosurgery, Beijing Fengtai Youanmen Hospital, Beijing, China j Department of Neurosurgery, The Second Affiliated Hospital of Jiaxing University, Zhejiang, China k Department of Neurology, The Second People's Hospital of Qinzhou City, Guangxi, China l Department of Geriatrics, Beijing Renhe Hospital, Beijing, China m Department of Neurology, Shengli Oilfield Central Hospital, Shandong, China n Department of Neurology, Chengde Central Hospital, Hebei, China o Department of Emergency, Rizhao People's Hospital, Shandong, China p Department of Neurology, Beijing Luhe Hospital, Capital Medical University, Beijing, China q Beijing Institute for Brain Disorders, Capital Medical University, Beijing, China r Departments of Neurology, Medical College of Georgia, Augusta University, Augusta, GA, USA s The George Institute for Global Health, Faculty of Medicine, University of New South Wales, Sydney, New South Wales, Australia t Neurology Department, Royal Prince Alfred Hospital, Sydney, NSW, Australia u The Institute of Science and Technology for Brain-inspired Research Fudan University, Shanghai, China v Neurocardiology Centre, Xuanwu Hospital, Capital Medical University, Beijing, China w Beijing Key Laboratory of Hypoxic Conditioning Translational Medicine, Xuanwu Hospital, Capital Medical University, Beijing, China ∗ Corresponding author. Department of Neurology, Xuanwu Hospital, Capital Medical University, No. 45, Changchun Street, Xicheng District, Beijing, 100053, China. [email protected] ∗∗ Corresponding author. Neurocardiology Centre, Xuanwu Hospital, Capital Medical University, No. 45, Changchun Street, Xicheng District, Beijing, 100053, China. [email protected] ∗∗∗ Corresponding author. Neurocardiology Centre, Xuanwu Hospital, Capital Medical University, No. 45, Changchun Street, Xicheng District, Beijing, 100053, China. [email protected] x Collaborators authors listed in Supplementary Material . Received 2025 Nov 20; Revised 2026 Mar 27; Accepted 2026 Mar 27; Collection date 2026 May. © 2026 The Author(s) This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). PMC Copyright notice PMCID: PMC13092074  PMID: 42011250 Summary Background Haematoma clearance is a key therapeutic target in intracerebral haemorrhage (ICH). Remote ischaemic conditioning (RIC) has shown potential to accelerate haematoma resolution in preclinical and early-phase clinical studies. We evaluated whether RIC improves functional outcomes in patients with acute supratentorial ICH not requiring surgical intervention. Methods In this randomised, sham-controlled, outcome-blinded trial at 20 centres in China, adults (aged 18–80 years) with imaging-confirmed supratentorial ICH (volume 10–30 mL, National Institutes of Health Stroke Scale score 6–20, GCS >8) and no surgical indication were randomly assigned (1:1) to receive RIC or sham intervention for 7 days after randomisation, alongside standard medical management. RIC was initiated 24–48 h after onset using an upper-limb cuff inflated to 200 mmHg (sham: 30 mmHg) for five daily cycles of 5-min inflation/deflation. The primary outcome was an mRS score of 0–2 at 90 days (favourable outcome) in the intention-to-treat population. This trial is registered with ClinicalTrials.gov , NCT04657133 . Findings Between April 22, 2021, and April 30, 2024, 458 patients were enrolled (median age 58 years [interquartile range 51–68], female 29.3%). Favourable outcome was achieved in 156 (68.1%) of 229 patients in the RIC group and 163 (71.2%) of 229 in the sham group at 90 days (adjusted risk ratio 0.97, 95% confidence interval 0.87–1.08; p = 0.61). Serious adverse events within 180 days occurred in 19 (8.3%) of 229 patients in the RIC group and 20 (8.7%) in the sham group, and no deaths were attributed to the RIC intervention. Interpretation In patients with acute supratentorial ICH managed without surgery, RIC did not improve 90-day functional outcomes. Further research should evaluate its efficacy in selected populations using optimised protocols. Funding National Natural Science Foundation of China; Beijing Natural Science Foundation; National Key R&D Program of China. Keywords: Intracerebral haemorrhage, Remote ischaemic conditioning, Haematoma, Randomised controlled trial Research in context. Evidence before this study To improve clinical outcomes in patients with intracerebral haemorrhage (ICH), new treatments are urgently needed. Remote ischaemic conditioning (RIC) with transient ischaemia and reperfusion of the arm or leg as an adjunct to medical management has been proposed as an adjunctive therapy to promote haematoma resolution and enhance neuroprotection. Preclinical and early-phase clinical studies have suggested that it might accelerate haematoma clearance and mitigate secondary brain injury. However, whether RIC improves functional outcomes in patients with ICH remains uncertain. We searched PubMed from database inception to October 20, 2025, using the search terms “remote ischaemic conditioning” AND “intracerebral haemorrhage”, for randomised controlled trials that compared RIC plus standard medical management with standard medical management alone in patients with supratentorial ICH. No completed randomised trials were identified that specifically evaluated the effect of RIC on functional outcomes in this population. One trial investigating RIC in patients with acute stroke included 165 patients with ICH, with 87 in the RIC group and 78 in the sham group, and results showed no significant association between RIC and improved functional outcomes (odds ratio, 1.15; 95% confidence interval, 0.65–2.04). Added value of this study To our knowledge, this is the first adequately powered, multicentre, randomised, sham-controlled, outcome-blinded trial to evaluate whether RIC improves functional outcomes in patients with spontaneous supratentorial ICH managed without surgery. In this study of 458 patients, RIC administered as an adjunct to standard medical management did not improve 90-day functional outcomes. However, the intervention was feasible, safe, and well tolerated, and was associated with a greater reduction in haematoma volume and enhanced haematoma resolution (post-hoc calculation based on pre-specified haematoma volume measurements) at 7 days, as well as a lower incidence of subsequent surgical intervention. These findings may provide important clinical and mechanistic insights into the role of RIC in ICH management. Implications of all the available evidence The RICH-2 trial provides evidence that RIC does not improve functional outcomes at 90 days in patients with supratentorial ICH who do not require surgical intervention. This study further confirms the safety and feasibility of RIC in this population, supporting further exploration of its use in other stroke subtypes and clinical settings, including prehospital care. Given the potential neuroprotective effects of RIC, future research may focus on ICH populations more likely to benefit, including patients with large haematomas or those undergoing surgical decompression. Optimisation of the treatment protocol, such as increased duration or frequency of RIC, should also be considered based on biological plausibility and insights from existing evidence. Introduction Acute spontaneous intracerebral haemorrhage (ICH), which results from rupture of a penetrating cerebral vessel, accounts for at least one-quarter of all types of new strokes worldwide, and it has devastating consequences. 1 , 2 As the haematoma volume and rate of expansion are strongly correlated with adverse clinical outcomes in ICH, both medical and surgical interventions to attenuate the haematoma are key therapeutic strategies. 3 , 4 Indeed, a minimally invasive approach to evacuate the haematoma has recently been shown to improve the outcome from ICH, especially in patients with lobar ICH. 5 Medical management remains the cornerstone of ICH treatment, primarily focussing on coordinated supportive care such as early, intensive blood pressure lowering and the control of other physiological parameters. 6 However, outcomes remain poor, motivating the search for adjunctive therapies to enhance recovery. Remote ischaemic conditioning (RIC) is a non-pharmacological intervention involving repeated cycles of brief limb ischaemia and reperfusion via cuff inflation and deflation, triggering endogenous protective pathways in distant organs, including the brain and heart. 7 , 8 Some studies have shown that RIC provides benefits to patients with ischaemic cerebrovascular disease. 9 , 10 Pre-clinical data show that RIC accelerates haematoma clearance, attenuates perihaematomal oedema, and promotes neurological recovery in animal models of ICH. 11 , 12 Mechanistically, RIC has been shown to promote haematoma resolution via AMPKα1-dependent anti-inflammatory macrophage polarisation and upregulation of CD36, a scavenger receptor critical for erythrophagocytosis. These findings suggest that RIC facilitates clot clearance through immunometabolic modulation of innate immune responses. In addition, our proof-of-concept pilot study also found that daily RIC administered for seven consecutive days was safe and well-tolerated, significantly promoting haematoma resolution and reducing perihaematomal oedema. 13 However, it remains uncertain whether this treatment approach could benefit patients by improving functional outcomes. The RICH-2 trial aimed to evaluate whether adjunctive RIC, in addition to standard medical management, when initiated between 24 and 48 h after supratentorial ICH onset and continued for seven consecutive days, could improve functional outcomes in patients managed without surgical intervention. Methods Trial design The RICH-2 trial was an investigator-initiated, multicentre, randomised, sham-controlled, parallel-group trial. It compared a 7-day course of RIC plus the standard medical management with sham RIC plus the standard medical management in adults with acute spontaneous supratentorial ICH who did not require surgical intervention. The trial protocol was approved by the ethics committees of Xuanwu Hospital, Capital Medical University, and all the participating centres. The details of the trial, protocol, protocol amendments, and statistical analysis plan are available in the Supplementary Materials . Several protocol amendments were implemented to enhance procedural clarity and ensure standardisation across trial sites. Full details are included within the Supplementary Materials . Versions 1.1 and 2.0 were finalised prior to the commencement of enrolment in April 2021. These revisions included refinements to treatment timing descriptions and further specification of the hierarchy and categorisation of pre-specified outcomes. Version 2.1 was implemented after enrolment had begun. This amendment was restricted to improving the clarity of the study objectives (not study outcomes) and standardising terminology between study objectives and outcomes measurements sections of the protocol. No changes were made to the eligibility criteria, intervention protocol, definitions or timepoints of primary or secondary outcomes, or the statistical analysis plan after enrolment commenced. For clarity, the primary outcome remained unchanged throughout the study and all secondary and exploratory analyses reported in the manuscript were pre-specified in protocol versions and finalised prior to data analysis. The trial was designed and supervised by a steering committee, which also provided an overview of the data analysis and interpretation. An independent data and safety monitoring board was responsible for the ongoing safety surveillance, ethical oversight, and trial conduct. All clinical outcomes were adjudicated by an independent clinical event adjudication committee whose members were blinded to group allocation and consisted of a neurocritical care specialist, a stroke neurologist, and a stroke neurosurgeon. The statistical analyses were performed by the blinded independent statistician. Imaging data were centrally reviewed and quality controlled by an independent neuroimaging core laboratory. Ethics This trial was approved by the Ethics Committee of Xuanwu Hospital, Capital Medical University (CR2021-002-R1), as well as the local ethics committees of all participating centres. Written informed consent was obtained from all participants or their legal representatives prior to enrolment. This trial was conducted in accordance with the ethical principles of the Declaration of Helsinki and followed the Good Clinical Practice Guidelines of the International Council for Harmonisation. Participants Patients were eligible if they were adults aged 18–80 years with imaging-confirmed supratentorial ICH and could be randomised and treated within 24–48 h of symptom onset. Key inclusion criteria were a moderate-to-severe level of neurological impairment defined by scores of 6–20 on the National Institutes of Health Stroke Scale (NIHSS), a haematoma volume of 10–30 mL (assessed by local investigators using the ABC/2 method), and a level of consciousness greater than 8 points on the Glasgow Coma Scale. Patients with ICH were excluded if surgery was performed or planned, including haematoma evacuation or decompressive craniectomy, concomitant subarachnoid haemorrhage or intraventricular haemorrhage, other conditions that precluded the use of RIC procedures (e.g. severe soft tissue injury, fracture, or peripheral vascular disease involving the upper limb), or an estimated life expectancy of <6 months due to a severe comorbid condition. The full list of inclusion and exclusion criteria is outlined in the Supplemental Methods section. Randomisation and masking Participants who fulfilled all the inclusion criteria but did not fulfil any of the exclusion criteria were randomly allocated to the RIC group or the sham RIC group in a 1:1 ratio. Randomisation was performed via a real-time, internet-based system ( https://iwrs.xhedc.com/ ), according to the stratification variables of baseline neurological severity (NIHSS scores ≤10 vs. >10) and baseline haematoma volume (≤20 mL vs. >20 mL), and a fixed block size of six. All participants, treating physicians, outcome assessors, and study investigators were blinded to the treatment assignments. The devices used for RIC and sham procedures were prepared with identical appearances, with the displayed treatment-related parameters disabled on the electronic screens. In addition, several other strategies were also implemented to ensure blinding, including informing participants during consent that they would receive RIC treatment and be made aware of the general procedure and possible mild discomfort, without explicitly disclosing group differences in cuff pressure; accommodating patients from different treatment arms in separate wards; and ensuring that trained nurses administered the interventions without interpreting patient reactions or commenting on treatment sensations. Further details are provided in the study protocol ( Supplemental Appendix ). Interventions All enrolled participants received standard medical management according to guidelines. 14 , 15 Participants assigned to the RIC group and the sham group received RIC or sham procedures, respectively, which were performed with a blood pressure cuff placed on the unilateral upper arm and programmed for 5 cycles of 5 min duration of inflation followed by 5 min of deflation (that is, a total length of 45 min per session). 16 The cuff was inflated to a pressure of 200 mmHg for RIC, whereas it was inflated to 30 mmHg for sham RIC. Both groups received the intervention within 10 min of randomisation and then once daily for seven consecutive days. All procedures were performed with the assistance of trained nurses. Outcome assessments were undertaken according to the time of randomisation. A clinical examination was performed at 24 (±6), 48 (±6), and 72 (±6) hours, and 7 days (±12 h) after randomisation. A repeat CT scan or MRI was performed 7 days after randomisation (or at discharge if earlier). When MRI was used, haematoma volume was assessed on T1-weighted sequences, which show acceptable agreement with CT for haematoma volumetry, with limited systematic bias. 17 An assessment of physical function on the modified Rankin scale (mRS) was undertaken at 30 (±7), 60 (±7), 90 (±7), and 180 (±14) days after randomisation, either in-person or by structured telephone interview. All assessments were conducted by independently blinded, certified evaluators. Local trained investigators evaluated the mRS, NIHSS, and other clinical outcomes, while all imaging data were reviewed centrally by the neuroimaging core laboratory. Outcomes The primary outcome was the proportion of participants with a favourable functional outcome, defined as a score of 0–2 on the modified Rankin Scale (mRS, range 0 [no symptoms] to 6 [death]) at 90 days post-randomisation. 18 , 19 This endpoint was evaluated in person or by telephone by certified personnel in scoring this scale. Secondary outcomes were categorised into neurological and imaging assessments. Neurological outcomes included: (1) excellent functional outcome (mRS score of 0–1) at 90 and 180 days; (2) favourable functional outcome at 180 days (mRS score of 0–2); (3) good functional outcome (mRS score 0–3) at 90 and 180 days; (4) a shift analysis across the full range of mRS scores at 90 and 180 days; and (5) the change in NIHSS within 7 days post-randomisation (assessed through serial clinical monitoring at 24 h, 48 h, 72 h, and day 7). Imaging outcomes included the intracerebral haematoma volume at baseline and at day 7 (or discharge if earlier), and the change from baseline to day 7. Peri-haematomal oedema volume at day 7 was pre-specified in the protocol but was not assessed due to technical limitations in CT-based measurement. Pre-specified safety outcomes included: (1) all-cause mortality within 180 days; (2) any adverse events (AEs) or serious adverse events (SAEs) within 180 days; (3) haematoma expansion, defined as a relative increase of ≥33% or an absolute increase of ≥6 mL within 7 days; and (4) neurological deterioration, defined as an NIHSS increase of ≥4 points or a Glasgow Coma Scale (GCS) decrease of ≥2 points within 7 days (assessed through serial clinical monitoring at 24 h, 48 h, 72 h, and day 7). Exploratory outcomes included the incidence of subsequent surgical intervention for the index ICH. Statistical analysis Based on prior clinical trials, we estimated that 30% of participants in the sham group would achieve a favourable functional outcome (mRS score 0–2) at 90 days. 20 , 21 , 22 Using data from our pilot study, we expected this proportion to be 43% in the RIC group. 13 Assuming a two-sided significance level of 0.05, a power of 80%, and a 1:1 allocation ratio between groups, a total of 215 patients per group were required. The resulting sample size of 430 patients was increased to a maximum of 452 to account for 5% of patients lost to follow-up or dropping out. For the primary efficacy outcome analysis, the data were analysed according to the intention-to-treat principle, which included all participants who underwent randomisation. The chi-squared test was used for the unadjusted analysis. In the adjusted analysis, a Poisson regression with robust error was used with adjustments for age, baseline NIHSS score, baseline haematoma volume, and baseline haematoma location. The risk ratio (RR) was calculated with the corresponding 95% confidence interval (CI). As five patients had missing 90-day outcome data, imputation procedures were used for the last known value or where this was unavailable; the worst score of 6 on the mRS was assigned to the RIC group, and the best score of 0 was assigned to the sham group. Two sensitivity analyses were performed to evaluate the robustness of the primary outcome under alternative missing data assumptions. Model I employed a complete-case analysis, excluding participants with missing 90-day functional data. Model II utilized a composite imputation approach: for participants with missing 90-day mRS but confirmed survival, the last observation carried forward method was applied; for participants with both missing functional data and unknown survival status, a worst-case scenario was assumed by assigning an mRS score of 6. Prespecified subgroup analyses of the primary efficacy outcome included age (≤60 or >60 years), sex, baseline NIHSS score (≤10 or >10), baseline haematoma volume (≤20 mL or >20 mL), prior use of antiplatelet therapy, baseline haematoma location (basal ganglia region vs. lobar region), side of clot location (left or right), and history of diabetes mellitus and hypertension. For each subgroup, treatment-by-subgroup interaction terms were tested to assess statistical evidence of effect modification. Analysis of binary secondary endpoints on the mRS was performed using a method similar to the primary efficacy analysis. A proportional odds model was used in the shift analysis of the functional outcomes after confirming that it was preferred over a multinomial model according to the Akaike information criterion. To analyse changes in NIHSS scores and haematoma volumes, group differences in the unadjusted analysis were assessed using t-tests, and adjusted analyses were conducted using multivariable linear regression. The adjusted analysis for all secondary outcomes used the same variables as those in the primary efficacy outcome analysis. Sensitivity analyses were performed in the per-protocol population for all efficacy outcomes. The per-protocol population comprised participants who met the eligibility criteria, received the assigned treatment, and did not undergo surgery, in accordance with the prespecified protocol. Participants with missing outcome data were excluded from the respective analyses, given the low frequency of missingness. Safety outcome analysis was performed in the safety set, which included participants who underwent at least one cycle of RIC or sham procedures. The number of participants with at least one serious adverse event and the incidence of serious adverse events are shown by treatment group according to the Medical Dictionary for Regulatory Activities (MedDRA) system organ class. These data were compared using risk difference and chi-squared tests. As a post-hoc analysis not prespecified in the study protocol, a derived measure of haematoma resolution rate was calculated based on pre-specified haematoma volume measurements. The haematoma resolution rate was calculated using the following formula: (baseline volume—follow-up volume)/baseline volume × 100%. Several other supplementary analyses were conducted as post-hoc analyses to further assess the robustness of the findings. These included: residual diagnostics and robust regression using Huber's M estimator for the linear models of NIHSS and imaging outcomes; a linear mixed-effects repeated-measures model evaluating NIHSS trajectories; assessment of the proportional-odds assumption together with a partial proportional-odds sensitivity analysis for the mRS shift outcome; and cumulative survival probabilities were estimated using Kaplan–Meier curves, with differences between the RIC and sham groups evaluated by the log-rank test. Interim analyses were not performed. No adjustments were made for multiplicity in the analysis of the secondary outcomes. No imputation was performed for secondary and exploratory outcomes; analyses were conducted using an available cases approach, as the proportion of missing data for these outcomes was low. Ethnicity was recorded at enrolment based on patient self-report in the case report form. All participants were of Han Chinese ethnicity; therefore, ethnicity-based analyses were not performed. All statistical analyses were conducted using R software, version 4.4.1 (R Foundation for Statistical Computing, www.r-project.org ). Statistical significance was set at p ≤ 0.05. The RICH-2 trial was prospectively registered with ClinicalTrials.gov as NCT04657133 . Role of the funding source The funders of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. The co-corresponding authors had full access to all the data in the study and had final responsibility for the decision to submit for publication. Results Patients Between April 22, 2021, and April 30, 2024, 458 participants were enrolled across 20 centres in China and randomised in a 1:1 ratio to receive either RIC (n = 229) or sham RIC (n = 229). The flow of participants is shown in Fig. 1 . The intervention was completed according to the protocol in 441 participants (223 in the RIC group and 218 in the sham group). At the 90-day follow-up, 221 patients in the RIC group and 215 patients in the sham group completed the intervention according to the protocol and were included in the per-protocol analysis. At the 180-day outcome assessment, 219 and 213 patients in the RIC and sham groups, respectively, completed the intervention as per protocol. The reasons for the non-completion of the intervention are shown in Fig. 1 . Fig. 1. Open in a new tab Trial profile . Baseline characteristics The baseline personal, clinical, and imaging characteristics were comparable between the two treatment groups ( Table 1 ). Baseline imaging data were missing for one participant in the RIC group; all other baseline data were complete. The median age of the 458 participants was 58 years (interquartile range [IQR], 51–68 years), with 324 men (70.7%) and 134 women (29.3%). The ICH was located in the basal ganglia in 419 patients (91.5%), in the lobar region in 36 (7.9%), and in both regions in three (0.6%). The median haematoma volume was 12.4 mL (IQR 8.3–17.8). The median Glasgow Coma Scale score was 14 (IQR, 13–15), and the median NIHSS score was 9 (IQR, 7–13). The baseline characteristics of the per-protocol population were also similar between the two trial groups ( eTable 1 in Supplemental Appendix ). Table 1. Baseline characteristics of participants. Characteristics RIC group N = 229 Sham group N = 229 Age, median (IQR), years 57 (51–67) 60 (51–70) Sex, no. (%) Male sex 161 (70.3) 163 (71.2) Female sex 68 (29.7) 66 (28.8) Medical history, no. (%) Hypertension 156 (68.1) 146 (63.8) Diabetes mellitus 23 (10.0) 24 (10.5) Dyslipidaemia 9 (3.9) 10 (4.4) History of coronary heart disease 9 (3.9) 12 (5.2) History of cerebral infarction 26 (11.4) 26 (11.4) History of intracerebral haemorrhage 14 (6.1) 9 (3.9) History of arrhythmia 6 (2.6) 4 (1.7) Smoking, no. (%) Current smoking 54 (23.6) 53 (23.1) Previous antiplatelet therapy, no. (%) a Mono-antiplatelet therapy 12 (5.2) 19 (8.3) Dual antiplatelet therapy 0 1 (0.4) Current use of antihypertensive drugs, No. (%) 58 (25.3) 53 (23.1) Statin use prior to admission, No. (%) 13 (5.7) 14 (6.1) Pre-stroke mRS = 1, no. (%) 20 (8.7) 16 (7.0) Baseline Glasgow Coma scale score, median (IQR) 14 (13–15) 14 (13–15) Baseline NIHSS score Median (IQR) 10 (7–13) 9 (6–12) Distribution, no. (%) ≤10 130 (56.8) 137 (59.8) >10 99 (43.2) 92 (40.2) Intracerebral haemorrhage characteristics on brain imaging b Baseline haematoma volume, median (IQR), ml 13.0 (8.5–18) 11.6 (7.9–17.3) ≤20 mL, No. (%) 184/228 (80.7) 191/229 (83.4) >20 mL, No. (%) 44/228 (19.3) 38/229 (16.6) Location of haemorrhage, no. (%) Basal ganglia 206 (90) 213 (93) Lobar regions 21 (9.2) 15 (6.6) Lobar and basal ganglia 2 (0.9) 1 (0.4) Left hemisphere site of haematoma, no. (%) 120 (52.4) 120 (52.4) Blood pressure at randomisation Median SBP (IQR)-mm Hg 160 (145–175) 160 (143–174) Median DBP (IQR)-mm Hg 94 (85–103) 92 (81.2–102) Median glucose at admission (IQR)-mmol/L 5.8 (5.1–7.0) 6.1 (5.3–7.0) Time from symptom onset to randomisation Median interval (IQR), hours 35.7 (29.1–41.3) 34.4 (29.4–40) Time from symptom onset to treatment Median interval (IQR), hours 35.8 (29.2–41.4) 34.4 (29.5–40.1) Open in a new tab RIC, remote ischaemic conditioning; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; SBP, systolic blood pressure; DBP, diastolic blood pressure. a Taking antiplatelet drugs within 1 week before the present symptom onset. b Data were missing for one patient in the RIC group. Primary outcome The primary efficacy outcome, defined as a favourable functional outcome (mRS score of 0–2) at 90 days, was achieved in 156 patients (68.1%) in the RIC group and 163 patients (71.2%) in the sham group. No significant difference was observed between the two treatment groups (adjusted risk ratio [RR] 0.97, 95% confidence interval [CI] 0.87–1.08; p = 0.61) ( Table 2 and Fig. 2 ). Table 2. Primary and secondary efficacy outcomes. Endpoint RIC group N = 229 Sham group N = 229 Measure of effect Unadjusted RR/cOR/MD (95% CI) p value Adjusted RR/cOR/MD (95% CI) a p value a Primary efficacy endpoint Functional independence (mRS score of 0–2) at 90 days 156/229 (68.1) 163/229 (71.2) RR 0.96 (0.85–1.08) 0.48 0.97 (0.87–1.08) 0.61 Secondary efficacy endpoint mRS at 90 days b Median score, median (IQR) 1 (1–3) 1 (1–3) cOR 1.03 (0.74–1.34) 0.88 0.99 (0.71–1.40) 0.94 Excellent outcome (mRS score 0–1), No. (%) 124/229 (54.1) 121/229 (52.8) RR 1.02 (0.86–1.22) 0.78 1.03 (0.89–1.20) 0.66 Good outcome (mRS score 0–3), No. (%) 188/229 (82.1) 190/229 (83.0) RR 0.99 (0.91–1.08) 0.81 0.99 (0.92–1.07) 0.80 mRS at 180 days b Median score, median (IQR) 1 (0–3) 1 (0–2) cOR 1.02 (0.73–1.31) 0.93 1.02 (0.73–1.45) 0.90 Excellent outcome (mRS score 0–1), No. (%) 144/229 (62.9) 148/229 (64.6) RR 0.97 (0.85–1.12) 0.70 0.98 (0.86–1.11) 0.70 Functional independence (mRS score of 0–2), No. (%) 169/229 (73.8) 175/229 (76.4) RR 0.97 (0.87–1.07) 0.52 0.97 (0.88–1.07) 0.57 Good outcome (mRS score 0–3), No. (%) 198/229 (86.5) 197/229 (86.0) RR 1.01 (0.93–1.08) 0.89 1.00 (0.94–1.08) 0.90 NIHSS score change from baseline c 24 h after randomisation, median (IQR) 0.0 (−1.0 to 0.0) 0.0 (−1.0 to 0.0) MD −0.12 (−0.49 to 0.25) 0.52 −0.12 (−0.50 to 0.26) 0.53 72 h after randomisation, median (IQR) −2.0 (−4.0 to 0.0) −1.0 (−3.0 to 0.0) MD −0.28 (−0.82 to 0.27) 0.32 −0.20 (−0.75 to 0.35) 0.48 7 d after randomisation or at early discharge, median (IQR) −3.0 (−6.0 to −1.0) −3.0 (−5.0 to −1.0) MD −0.17 (−0.85 to 0.51) 0.63 −0.01 (−0.69 to 0.66) 0.97 Imaging Endpoints d N = 203 N = 208 Imaging interval, median (range), days 7 (4–9) 7 (5–9) MD 0.13 (−0.07 to 0.32) 0.21 – – Intracerebral haematoma volume at follow-up, ml e 5.6 (3.4–8.2) 6.7 (4.0–9.6) MD −1.40 (−2.29 to −0.50) 0.002 −1.80 (−2.42 to −1.17) <0.001 Intracerebral haematoma volume reduction from baseline to follow-up, ml f 6.9 (4.7–9.9) 4.4 (3.1–6.8) MD 2.32 (1.57–3.07) <0.001 2.04 (1.45–2.62) <0.001 Haematoma resolution rate, % g 58.1 (47.4–66.6) 40.9 (33.7–50.5) MD 13.96 (11.09–16.82) <0.001 14.07 (11.21–16.94) <0.001 Open in a new tab RIC, remote ischaemic conditioning; mRS, modified Rankin Scale; IQR, interquartile range; NIHSS, National Institutes of Health Stroke Scale; RR, risk ratio; MD, mean difference; cOR, crude odds ratio. a The analysis was adjusted for age, baseline NIHSS, baseline haematoma volume, and baseline haematoma location. b Two patients in the RIC group and 3 patients in the sham group lost to 90-day follow up, and 4 patients in the RIC group and 5 patients in the sham group lost to the 180-day follow up. c One NIHSS value in the sham group was missing at 24 h; five values were missing at 72 h (two in the RIC group); and eight values were missing at 7 days after randomisation or at early discharge (three in the RIC group). d Twenty-six patients in the RIC group and 21 patients in the sham RIC group were excluded in the imaging analysis. In the RIC group 16 patients did not perform follow-up CT/MRI imaging, 8 patients experienced haematoma expansion, and 2 patients underwent surgery. In the sham group, 9 patients did not perform follow-up CT/MRI imaging, 3 patients experienced haematoma expansion, and 9 patients underwent surgery. e All baseline haematoma volumes were assessed on non-contrast CT performed within 6 h prior to randomisation. f MRI was used for follow-up in 31 patients (13 in the RIC group and 18 in the sham RIC group); the remaining follow-up scans were assessed on CT. g Haematoma resolution rate was calculated by the following formula (baseline volume − follow-up volume)/baseline volume × 100%. Fig. 2. Open in a new tab Functional outcome at 90 and 180 days . RIC, remote ischaemic conditioning. A neutral effect was confirmed in the sensitivity analysis of the 90-day mRS in the intention-to-treat population ( eTable 2 , eFigs. 1 and 2 ). Subgroup analyses of the primary outcomes in the intention-to-treat population demonstrated consistent findings across most categories ( Fig. 3 ). Similar results were observed in further sensitivity analyses (Supplemental Results, eFigs. 3 and 4 ). Fig. 3. Open in a new tab Subgroup analysis on the primary outcome in the intention-to-treat population . RIC, remote ischaemic conditioning; NIHSS, National Institutes of Health Stroke Scale. ˆImaging data were missing for one patient in the RIC group. ∗Three patients who had intracerebral hemorrhage in both lobar and basal ganglia were not included, including two patients in the RIC group and one patient in the sham group. #Data of one patient in the sham group were unclear. In the per-protocol analyses of the primary outcome, no significant differences were detected ( eTable 3 and eFig. 5 ), and similar results were observed in the subgroup analyses of the primary outcome in the per-protocol population ( eFig. 6 ). Neurological outcomes The secondary efficacy outcomes are summarised in Table 2 . At 90 days, the mRS score was 1 (IQR, 1 to 3) in both the RIC group and the sham group, with no significant difference observed (adjusted crude odds ratio [cOR] 0.99, 95% CI 0.71 to 1.40; p = 0.94). The proportional-odds assumption held for the treatment effect, and a partial proportional-odds model relaxing the assumption for age produced similar results (adjusted cOR 1.07; 95% CI, 0.77–1.50), supporting the robustness of the analysis. The proportion of patients achieving an excellent outcome (mRS score 0–1) was 54.1% in the RIC group compared to 52.8% in the sham group (adjusted RR 1.03, 95% CI, 0.89–1.20; p = 0.66). Good outcomes (mRS score 0–3) were achieved by 82.1% and 83.0% of patients in the respective groups (adjusted RR 0.99, 95% CI, 0.92–1.07; p = 0.80). At 180 days, there were no differences in mRS outcomes. Median scores remained 1 (IQR, 0–3 in the RIC group and 0–2 in the sham group; adjusted cOR 1.02, 95% CI, 0.73–1.45; p = 0.90). Excellent outcomes were reported in 62.9% and 64.6% of patients, respectively (adjusted RR 0.98, 95% CI, 0.86–1.11; p = 0.70), while functional independence (mRS score 0–2) was achieved in 73.8% vs. 76.4% (adjusted RR 0.97, 95% CI, 0.88–1.07; p = 0.57). Good outcomes (mRS score 0–3) were observed in 86.5% in the RIC group and 86.0% in the sham group (adjusted RR 1.00, 95% CI, 0.94–1.08; p = 0.90). In addition, changes in NIHSS scores from baseline at 24 h, 72 h, and 7 days or at discharge were also similar between the groups, with no statistically significant differences in the adjusted analyses. Robust regression yielded nearly identical effect estimates, confirming the robustness of these findings ( eTables 4 and 5 ; eFig. 7 ). Additionally, the results of the per-protocol analysis of secondary outcomes were similar to those of the primary intention-to-treat analyses ( eTable 3 and eFig. 8 ). Imaging outcomes The imaging outcomes are summarised in Table 2 . While baseline haematoma volumes were initially estimated using the ABC/2 method by local investigators to determine eligibility, all imaging data included in the final analysis were assessed centrally by the core imaging laboratory. A total of 411 patients (89.7%) were included in the imaging analysis. The median interval between baseline and follow-up imaging was 7 days in both the RIC group (range 4–9) and the sham group (range 5–9), with no significant difference between groups. All baseline haematoma volumes were measured from non-contrast CT scans. MRI was used for follow-up imaging in 31 patients (13 in the RIC group and 18 in the sham RIC group), while the remaining patients underwent follow-up CT. At follow-up, the median haematoma volume was 5.6 mL (IQR, 3.4–8.2) in the RIC group and 6.7 mL (IQR, 4.0–9.6) in the sham group, which was significantly different (adjusted mean difference −1.80 mL, 95% CI, −2.42 mL to −1.17 mL, p < 0.001). The haematoma volume reduction from baseline to follow-up was 6.9 mL (IQR, 4.7 to 9.9) in the RIC group and 4.4 mL (IQR, 3.1–6.8) in the sham group, which was significantly greater in the RIC group (adjusted mean difference 2.04 mL, 95% CI, 1.45–2.62 mL, p < 0.001). Additionally, the haematoma resolution rate (post-hoc calculation based on pre-specified haematoma volume measurements) was significantly higher in the RIC group, with a median of 58.1% (IQR, 47.4%–66.6%), compared to 40.9% (IQR, 33.7%–50.5%) in the sham group (adjusted mean difference, 14.07%; 95% CI, 11.21%–16.94%; p < 0.001). These results remained consistent in robust regression analyses, with effect estimates similar to those of the primary models ( eTable 4 ; eFig. 9 ). Safety outcomes and adverse events The safety outcomes are summarised in Table 3 and detailed in eTables 6 and 7 . At 90 days, mortality occurred in 3.1% of patients in the RIC group and 2.7% in the sham group (adjusted risk difference [RD] 0.3%, 95% CI, −2.7%–3.3%; p = 0.84). At 180 days, the mortality rates were 3.1% and 4.0%, respectively (adjusted RD –1.0%, 95% CI –4.4%–2.4%; p = 0.57). Kaplan–Meier survival curves are presented in eFig. 10 . Haematoma expansion was observed in 7 (3.1%) patients in the RIC group and 12 (5.2%) in the sham group, with no statistically significant difference between groups (adjusted RD –0.15%, 95% CI –5.0 to 2.0; p = 0.41). However, the incidence of subsequent surgical intervention was significantly lower in the RIC group (0.9%) compared with the sham group (3.9%) (adjusted RD –3.3%, 95% CI –6.1 to −0.5; p = 0.02). Serious adverse events within 90 and 180 days were comparable between groups, with no statistically significant differences observed. The rate of neurological deterioration at 24, 48, and 72 h, as well as at 7 days, were similar between the two treatment groups, with no significant differences in the adjusted analyses. Table 3. Safety outcomes. Outcome RIC group (N = 229) Sham group (N = 229) Measure of effect Unadjusted RD (95% CI) p value Adjusted RD (95% CI) b p value b Deathˆ Within 90 days, no. (%) 7/227 (3.1) 6/226 (2.7) RD 0.4 (−2.7 to 3.5) 0.79 0.3 (−2.7 to 3.3) 0.84 Within 180 days, no. (%) 7/225 (3.1) 9/224 (4.0) RD −0.9 (−4.3 to 2.5) 0.61 −1.0 (−4.4 to 2.4) 0.57 Haematoma expansion c , no. (%) 7/229 (3.1) 12/229 (5.2) RD −2.2 (−1.5 to 5.8) 0.24 −0.15 (−5.0 to 2.0) 0.41 Surgical intervention 2/229 (0.9) 9/229 (3.9) RD −3.1 (−6.3 to −0.4) 0.03 −3.3 (−6.1 to −0.5) 0.02 Neurological deterioration a , no. (%) Within 24 h 5/229 (2.2) 2/229 (0.9) RD 1.3 (−0.9 to 3.6) 0.25 1.4 (−0.9 to 3.7) 0.23 Within 48 h 6/229 (2.6) 4/229 (1.7) RD 0.9 (−1.8 to 3.6) 0.52 0.9 (−1.8 to 3.6) 0.53 Within 72 h 2/229 (0.9) 2/229 (0.9) RD 0.0 (−1.7 to 1.7) >0.99 0.0 (−1.7 to 1.8) 0.97 Within 7 days 3/229 (1.3) 1/229 (0.4) RD 0.9 (−0.8 to 2.6) 0.32 1.1 (−0.6 to 2.8) 0.21 RIC-related adverse events, no. (%) Skin petechiae 38/229 (16.6) 0/229 (0.0) RD 16.6 (11.8–21.4) <0.001 16.8 (12.0–21.6) <0.001 Transient pain 12/229 (5.2) 1/229 (0.4) RD 4.8 (1.8–7.8) 0.002 4.8 (1.9–7.7) 0.001 Serious adverse events within 90 days, no. (%) 16/229 (7.0) 12/229 (5.2) RD 1.8 (−2.7 to 6.1) 0.44 1.7 (−2.7 to 6.0) 0.45 Serious adverse events within 180 days, no. (%) 19/229 (8.3) 20/229 (8.7) RD −0.4 (−5.6 to 4.7) 0.87 −0.3 (−5.4 to 4.8) 0.91 Open in a new tab RIC, remote ischaemic conditioning; RD, Risk Difference. a Neurological deterioration is defined as an increase of 4 points or more on the NIHSS or a decline of 2 points or more on the Glasgow Coma Scale within 7 days post-randomisation. b The analysis was adjusted for age, baseline NIHSS, baseline haematoma volume, and baseline haematoma location. c Haematoma expansion was defined as a 33% relative increase or an absolute increase of ≥6 mL. Adverse events related to RIC were more frequent in the intervention group. Skin petechiae were reported in 16.6% of patients receiving RIC, compared with none in the sham group (adjusted RD 16.8%, 95% CI, 12.0%–21.6%; p < 0.001). Transient pain was reported in 5.2% of patients in the RIC group vs. 0.4% in the sham group (adjusted RD 4.8%, 95% CI, 1.9%–7.6%; p = 0.001). None of the participants withdrew from the study because of the adverse events. Discussion This phase 3 clinical trial aimed to investigate the effects of RIC in patients with supratentorial ICH who did not undergo surgical interventions. While RIC did not significantly improve functional or other clinical outcomes, the procedure was proven to be feasible, safe, and well tolerated in this patient population, and was associated with a significant reduction in haematoma volume and an increased haematoma resolution rate at the 7-day follow-up. Additionally, it was associated with a lower rate of subsequent surgical intervention therapy. However, its effects on haematoma clearance and surgical need should be interpreted as exploratory. Although RIC was originally developed for ischaemic diseases, several studies have investigated it in animal model of ICH, and the results show that it can accelerate haematoma resolution, attenuate perihaematomal oedema, and promote neurological recovery. Early research demonstrated that ischaemic preconditioning could reduce ICH-induced brain oedema, likely through activation of p44/42 MAPKs and upregulation of haem oxygenase-1, transferrin, and transferrin receptor, which may enhance iron metabolism and reduce oxidative injury. 11 More recently, a study employing once-daily bilateral RIC starting 2 h post-ICH for 7 days demonstrated enhanced haematoma clearance and neurological recovery, mediated by AMPK-dependent macrophage polarisation and increased erythrophagocytosis. 12 These findings provide mechanistic support for the use of RIC in ICH, with effects distinct from its originally proposed role in ischaemic injury. Consistent with previous pre-clinical studies and our pilot clinical trial, 12 , 13 this trial further confirmed the effects of RIC in promoting haematoma resolution in patients with ICH. Both the absolute reduction in haematoma volume and the clearance rate reflected this process, with the latter representing a normalised proportion of volume reduction. Despite these promising results, however, our trial did not demonstrate a functional benefit, consistent with the results of a previous randomised study. The Remote Ischaemic Conditioning in Patients with Acute Stroke Trial (RESIST) investigated the effect of RIC when initiated in the prehospital setting and continued in the hospital on functional outcome in patients with acute stroke, and the results also showed no improvement in functional outcomes in the 165 participants with ICH. 23 Several factors may explain the neutral findings in functional outcomes observed in this trial. First, most of the participants had less severe clinical and radiological profiles within the spectrum of ICH, with a median NIHSS score of 9 and a median haematoma volume of 12.4 mL at baseline. The majority achieved a favourable outcome even with standard medical management alone, as reflected by approximately 70% achieving functional independence and over half of them attaining an excellent outcome at the 90-day follow-up. The favourable prognosis might greatly reduce the trial's power to detect any additional benefits from RIC. Second, the regimen of RIC, administered unilaterally once daily for 7 consecutive days, was rather pragmatic but differed from protocols used in ischaemic stroke trials, where bilateral RIC is commonly applied twice daily for up to 14 days and has shown potential efficacy in improving functional outcomes. 24 , 25 Moreover, in a preclinical study where RIC promoted neurological recovery, it was initiated 2 h after haematoma induction and applied bilaterally to the lower extremities, 12 a protocol that also differed from the one used in our trial. The proportion of favourable functional outcome in the sham group was substantially higher than anticipated at the time of trial design. Whereas the sample-size calculation assumed a 30% rate of functional independence (mRS 0–2), approximately 70% of participants achieved this outcome under standard management, reflecting a relatively mild ICH population. This higher baseline event rate reduced the absolute effect size the trial was powered to detect and increased the risk of a type II error. Nevertheless, interpretation of the primary endpoint should be informed by the precision of the effect estimates. The 95% confidence interval for the primary outcome risk ratio (0.87–1.08), together with the observed absolute difference of −3.1%, does not support a clinically meaningful functional benefit of RIC in this population and argues against a masked treatment effect due solely to limited statistical power. The dissociation observed in this trial between enhanced haematoma resolution and neutral functional outcomes merits further consideration. RIC is thought to facilitate haematoma clearance mainly by modulating secondary injury processes, such as neuroinflammation and microglial phagocytosis, which may limit delayed tissue damage but cannot reverse irreversible primary injury sustained at haemorrhage onset, particularly when critical neural structures or white-matter tracts are involved. 26 In addition, the median time to randomisation was approximately 35 h after symptom onset, by which time perihaematomal oedema and inflammatory cascades may have already caused substantial injury, potentially limiting the functional impact of later haematoma clearance. Finally, although haematoma volume was significantly reduced at day 7, the absolute magnitude of reduction may not have been sufficient to translate into a clinically meaningful shift on the mRS, especially in a population with relatively mild baseline severity and high rates of favourable outcomes under standard care. The proportion of functional independence in this trial was higher than initially anticipated and exceeded the rates reported in previous trials of medical treatment for ICH, which are generally below 50%. 20 , 21 , 22 Several factors may account for this discrepancy. Compared with studies such as CHANT (Cerebral Haematoma and NXY Treatment trial) and GAIN (Glycine Antagonist in Neuroprotection for Patients with Acute Stroke), the present study enrolled patients with smaller haematoma volumes and benefited from advances in stroke care over the past two decades. While the haematoma volume and study period were more comparable to the i-DEF (Intracerebral Haemorrhage Deferoxamine) trial, that study included more than one-third of patients with intraventricular haemorrhage, a strong predictor of poor functional outcome. 27 In contrast, patients with intraventricular haemorrhage were excluded in this trial. Furthermore, this trial enrolled patients 24–48 h after symptom onset, a relatively stable phase during which the risk of haematoma expansion is lower, whereas prior studies recruited patients within 24 h or even within 6 h of symptom onset, when clinical deteriorate due to haematoma expansion is more likely. 28 These factors may collectively explain both the higher-than-expected rate of functional independence and the relatively rapid haematoma clearance observed in our study. The results of our trial have several implications for future research on RIC in patients with acute stroke. This trial further confirmed the safety of RIC in patients with ICH, which may provide additional safety evidence for research on RIC in the field of acute stroke (including ischaemic and haemorrhagic stroke), especially in prehospital scenarios. With ongoing advances in surgical and minimally invasive techniques, as well as a gradual broadening of operative indications, some patients with spontaneous ICH who might previously have been treated conservatively are now increasingly considered for surgical intervention. As a result, focussing exclusively on non-surgical ICH populations for RIC research may become more challenging or less generalisable in future studies. Given the observed effects of RIC on haematoma resolution, further research should explore its potential benefits in patients with large haematoma volume, including those treated with decompressive craniectomy, 29 or even those with sizeable haematomas that are not suitable for surgical therapy. The rationale for this focus is that in patients with larger haematomas, where haematoma volume strongly drives secondary injury and mass effect, the pro-resolution effect of RIC is more likely to translate into meaningful functional improvement than in those with smaller haematomas, in whom ceiling effects may limit detectable benefit. In addition, adopting a RIC protocol that has been shown to be effective in other studies would seem to be a logical choice; that is, an intervention in both arms, twice daily for two weeks or longer. 30 Another proof-of-concept trial involving 20 patients with ICH is in progress to test the hypothesis that RIC can accelerate the reabsorption of haematoma, 31 and a multicentre randomised controlled trial, the SERIC-ICH trial ( NCT05609110 ), is being conducted to determine its effects on clinical outcomes by enrolling 2000 patients with ICH. 32 These trials utilise different study designs and inclusion/exclusion criteria, but will also make important contributions to the assessment of RIC for ICH. Our study had several limitations. First, this trial was conducted exclusively in China in an ethnically homogeneous population (all participants were of Han Chinese ethnicity). Pattern and management of ICH may differ from those in Western populations, and patients with either severe or minor neurological impairment were excluded. These factors may limit the generalisability of the results. Second, although this trial recruited patients consecutively, less than one-third of the participants were female, which may have impacted the interpretation of the results. Third, this trial used the ABC/2 formula to estimate haematoma volume during enrolment. Although simple and widely used, this method tends to overestimate volumes, particularly for irregular haematomas, leading to the inclusion of patients with actual volumes below the target range of 10–30 mL and consequently influencing the observed treatment effects. Fourth, the sham-controlled design used a cuff pressure of 30 mmHg for the control group, which may not have perfectly matched the sensory experience of the 200 mmHg active RIC and could potentially have led to partial unblinding of participants or clinical staff, introducing performance bias. Future RIC trials should investigate whether utilising a higher sham pressure can better preserve blinding integrity while remaining sub-ischaemic. Finally, information on haematoma expansion prior to enrolment was not systematically collected; therefore, we cannot exclude the possibility that early expansion before randomisation may have influenced baseline haematoma characteristics and the effects of RIC. In patients with acute supratentorial ICH managed without surgical intervention, RIC was safe, well-tolerated, and promoted early haematoma clearance with a reduced need for subsequent surgical intervention; however, these mechanistic benefits did not translate into improved functional outcomes at 90 or 180 days. Further studies should investigate extended RIC regimens and their effectiveness in specific ICH patient subgroups. Contributors WZ, SL, and XJ obtained funding, designed the study, and were responsible for the overall principal leadership for the study. JW, XY, GW, SY, DL, XJG, YH, LN, HZ, CX, NL, MZ, JZ, QD, XKG, QM, HS collected the data. WZ, JW, YJ were involved in data cleaning and verification. The manuscript was written by WZ, SL, XJ, YY, and LL. Statistical analyses and drawing of figures were performed by the blinded statistician LL. Several authors, particularly DH and CSA, made critical revisions to the manuscript text. All authors have seen and approved the final text. WZ, SL, and XJ had accessed and verified the underlying raw data. XJ was responsible for the decision to submit the manuscript for publication, and attested that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. Data sharing statement Data collected for the study, including deidentified individual participant data and a data dictionary defining each field in the set, can be made available to researchers on reasonable request and after signing appropriate data sharing agreements. Please send data access requests to the corresponding author. Requests for data access should be directed to the corresponding author and must be approved by the relevant ethics committees and data custodians. The code used to process the study data can also be provided to researchers upon reasonable request. The analyses were conducted using the publicly available R software. Declaration of interests The patent of the RIC device (patent no. ZL200820123637.X) used in this trial belongs to Xuanwu Hospital, Capital Medical University. RICH-2 was an investigator-initiated trial, and none of the authors has any competing interests related to the submitted work. WBZ and SJL report receiving research grants from the Beijing Natural Science Foundation, National Natural Science Foundation of China, and Beijing Municipal Science & Technology Commission, paid to their institution. QFM reports receiving research grants from the National Natural Science Foundation of China and the Ministry of Science and Technology of China, paid to his institution. XMJ reports receiving research grants from the Beijing Municipal Education Commission, Beijing Municipal Bureau of Finance, National Natural Science Foundation of China, and Ministry of Science and Technology of China, paid to his institution. LL served as Data and Safety Monitoring Board member for a trial sponsored by Medtronic. D.C. H. reports receiving research grants from the U.S. National Institutes of Health (NIH). C.S.A. reports receiving research grants from the National Health and Medical Research Council of Australia and the Medical Research Foundation of the UK, and is a consultant to Auzone Biotechnology Shanghai, Astra Zeneca and Bayer. He is President-elect of the World Stroke Organisation, Editor-in-Chief of Cerebrovascular Diseases and Chairs several data and safety monitoring boards of investigator-initiated studies. All other authors declare no competing interests. Acknowledgements We thank all participants, their relatives or carers, and their primary care practitioners; imaging adjudicators, outcome event adjudicators, the trial steering committee, and the data monitoring committee. The RICH-2 trial was supported by the National Natural Science Foundation of China of China (82422024), the Beijing Natural Science Fund for Outstanding Young Scholars (JQ22020), the National Natural Science Foundation of China (82001257), Beijing Nova Program (Z201100006820143), the National Key R&D Program of China (2022YFC2408800), and the Outstanding Young Talents Program of Capital Medical University (1220090115). The funders of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. Footnotes Appendix A Supplementary data related to this article can be found at https://doi.org/10.1016/j.eclinm.2026.103900 . Contributor Information Wenbo Zhao, Email: [email protected]. Sijie Li, Email: [email protected]. Xunming Ji, Email: [email protected]. RICH-2 Investigators: Xunming Ji , Zhiying Zhao , Jianli Niu , Lin Zhao , Song Yang , Na Li , Yanbo Cheng , Xuebin Yu , Yuping He , Haihang Zhou , Mengfei Zhong , Ming Wei , Xinjing Gao , Guang Wu , Jiangtao Zhang , Chaixia Xuan , Qingling Dong , Junzhao Sun , Jianguo Wang , and Xiaokun Geng Appendix ASupplementary data Supplementary Data mmc1.pdf (1.7MB, pdf) Protocol Amendments by version mmc2.docx (39.2KB, docx) SAP and its amendments mmc3.pdf (480.7KB, pdf) Trial protocol mmc4.pdf (1.2MB, pdf) References 1. Global, regional, and national burden of stroke and its risk factors, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Neurol. 2021;20(10):795–820. doi: 10.1016/S1474-4422(21)00252-0. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Montaño A., Hanley D.F., Hemphill J.C. Hemorrhagic stroke. Handb Clin Neurol. 2021;176:229–248. doi: 10.1016/B978-0-444-64034-5.00019-5. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Magid-Bernstein J., Girard R., Polster S., et al. Cerebral hemorrhage: pathophysiology, treatment, and future directions. Circ Res. 2022;130(8):1204–1229. doi: 10.1161/CIRCRESAHA.121.319949. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Wilkinson D.A., Keep R.F., Hua Y., Xi G. Hematoma clearance as a therapeutic target in intracerebral hemorrhage: from macro to micro. J Cereb Blood Flow Metab. 2018;38(4):741–745. doi: 10.1177/0271678X17753590. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Pradilla G., Ratcliff J.J., Hall A.J., et al. Trial of early minimally invasive removal of intracerebral hemorrhage. N Engl J Med. 2024;390(14):1277–1289. doi: 10.1056/NEJMoa2308440. [ DOI ] [ PubMed ] [ Google Scholar ] 6. Greenberg S.M., Ziai W.C., Cordonnier C., et al. 2022 guideline for the management of patients with spontaneous intracerebral hemorrhage: a Guideline from the American Heart Association/American Stroke Association. Stroke. 2022;53(7):e282–e361. doi: 10.1161/STR.0000000000000407. [ DOI ] [ PubMed ] [ Google Scholar ] 7. Zhao W., Li S., Ren C., Meng R., Jin K., Ji X. Remote ischemic conditioning for stroke: clinical data, challenges, and future directions. Ann Clin Transl Neurol. 2019;6(1):186–196. doi: 10.1002/acn3.691. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Xu Y., Wang Y., Ji X. Immune and inflammatory mechanism of remote ischemic conditioning: a narrative review. Brain Circ. 2023;9(2):77–87. doi: 10.4103/bc.bc_57_22. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Chen H.S., Cui Y., Li X.Q., et al. Effect of remote ischemic conditioning vs usual care on neurologic function in patients with acute moderate ischemic stroke: the RICAMIS randomized clinical trial. JAMA. 2022;328(7):627–636. doi: 10.1001/jama.2022.13123. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Zhao W., Yu W., Li S., Ren C., Ji X. Remote ischemic conditioning for stroke: where we are and where to go. Conditioning medicine. 2021;4(4):185–191. [ Google Scholar ] 11. He Y., Karabiyikoglu M., Hua Y., Keep R.F., Xi G. Ischemic preconditioning attenuates brain edema after experimental intracerebral hemorrhage. Transl Stroke Res. 2012;3(1 Suppl 1):180–187. doi: 10.1007/s12975-012-0171-z. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Vaibhav K., Braun M., Khan M.B., et al. Remote ischemic post-conditioning promotes hematoma resolution via AMPK-dependent immune regulation. J Exp Med. 2018;215(10):2636–2654. doi: 10.1084/jem.20171905. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Zhao W., Jiang F., Li S., et al. Safety and efficacy of remote ischemic conditioning for the treatment of intracerebral hemorrhage: a proof-of-concept randomized controlled trial. Int J Stroke. 2022;17(4):425–433. doi: 10.1177/17474930211006580. [ DOI ] [ PubMed ] [ Google Scholar ] 14. Liu L., Chen W., Zhou H., et al. Chinese stroke association guidelines for clinical management of cerebrovascular disorders: executive summary and 2019 update of clinical management of ischaemic cerebrovascular diseases. Stroke Vasc Neurol. 2020;5(2):159–176. doi: 10.1136/svn-2020-000378. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Chinese Society of Neurology, Chinese Stroke Society Chinese guidelines for diagnosis and treatment of acute intracerebral hemorrhage 2019. Chin J Neurol. 2019;52(12):994–1005. [ Google Scholar ] 16. Hou C., Lan J., Lin Y., et al. Chronic remote ischaemic conditioning in patients with symptomatic intracranial atherosclerotic stenosis (the RICA trial): a multicentre, randomised, double-blind sham-controlled trial in China. Lancet Neurol. 2022;21(12):1089–1098. doi: 10.1016/S1474-4422(22)00335-0. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Schlunk F., Kuthe J., Harmel P., et al. Volumetric accuracy of different imaging modalities in acute intracerebral hemorrhage. BMC Med Imaging. 2022;22(1):9. doi: 10.1186/s12880-022-00735-3. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Wilson J.T., Hareendran A., Hendry A., Potter J., Bone I., Muir K.W. Reliability of the modified Rankin Scale across multiple raters: benefits of a structured interview. Stroke. 2005;36(4):777–781. doi: 10.1161/01.STR.0000157596.13234.95. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Weimar C., Kurth T., Kraywinkel K., et al. Assessment of functioning and disability after ischemic stroke. Stroke. 2002;33(8):2053–2059. doi: 10.1161/01.str.0000022808.21776.bf. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Lees K.R., Asplund K., Carolei A., et al. Glycine antagonist (gavestinel) in neuroprotection (GAIN International) in patients with acute stroke: a randomised controlled trial. GAIN International Investigators. Lancet. 2000;355(9219):1949–1954. doi: 10.1016/s0140-6736(00)02326-6. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Lyden P.D., Shuaib A., Lees K.R., et al. Safety and tolerability of NXY-059 for acute intracerebral hemorrhage: the CHANT trial. Stroke. 2007;38(8):2262–2269. doi: 10.1161/STROKEAHA.106.472746. [ DOI ] [ PubMed ] [ Google Scholar ] 22. Selim M., Foster L.D., Moy C.S., et al. Deferoxamine mesylate in patients with intracerebral haemorrhage (i-DEF): a multicentre, randomised, placebo-controlled, double-blind phase 2 trial. Lancet Neurol. 2019;18(5):428–438. doi: 10.1016/S1474-4422(19)30069-9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Blauenfeldt R.A., Hjort N., Valentin J.B., et al. Remote ischemic conditioning for acute stroke: the RESIST randomized clinical trial. JAMA. 2023;330(13):1236–1246. doi: 10.1001/jama.2023.16893. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Zhao W., Hausenloy D.J., Hess D.C., Yellon D.M., Ji X. Remote ischemic conditioning: challenges and opportunities. Stroke. 2023;54(8):2204–2207. doi: 10.1161/STROKEAHA.123.043279. [ DOI ] [ PubMed ] [ Google Scholar ] 25. An J.Q., Cheng Y.W., Guo Y.C., et al. Safety and efficacy of remote ischemic postconditioning after thrombolysis in patients with stroke. Neurology. 2020;95(24):e3355–e3363. doi: 10.1212/WNL.0000000000010884. [ DOI ] [ PubMed ] [ Google Scholar ] 26. Zhao W., Wu C., Stone C., Ding Y., Ji X. Treatment of intracerebral hemorrhage: current approaches and future directions. J Neurol Sci. 2020;416 doi: 10.1016/j.jns.2020.117020. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Hinson H.E., Hanley D.F., Ziai W.C. Management of intraventricular hemorrhage. Curr Neurol Neurosci Rep. 2010;10(2):73–82. doi: 10.1007/s11910-010-0086-6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Li Z., You M., Long C., et al. Hematoma expansion in intracerebral hemorrhage: an update on prediction and treatment. Front Neurol. 2020;11:702. doi: 10.3389/fneur.2020.00702. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. Beck J., Fung C., Strbian D., et al. Decompressive craniectomy plus best medical treatment versus best medical treatment alone for spontaneous severe deep supratentorial intracerebral haemorrhage: a randomised controlled clinical trial. Lancet. 2024;403(10442):2395–2404. doi: 10.1016/S0140-6736(24)00702-5. [ DOI ] [ PubMed ] [ Google Scholar ] 30. Ji X., Zhao W., Boltze J., et al. Clinical practice guidelines of remote ischemic conditioning for the management of cerebrovascular diseases. Cond Med. 2019;2(5):225–241. [ Google Scholar ] 31. Jarrahi A., Shah M., Ahluwalia M., et al. Pilot study of remote ischemic conditioning in acute spontaneous intracerebral hemorrhage. Front Neurosci. 2022;16 doi: 10.3389/fnins.2022.791035. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Guo Z.N., Qu Y., Abuduxukuer R., et al. Safety and efficacy of remote ischemic conditioning for spontaneous intracerebral hemorrhage (SERIC-ICH): a multicenter, randomized, parallel-controlled clinical trial study design and protocol. Eur Stroke J. 2024;9(1):259–264. doi: 10.1177/23969873231201712. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Data mmc1.pdf (1.7MB, pdf) Protocol Amendments by version mmc2.docx (39.2KB, docx) SAP and its amendments mmc3.pdf (480.7KB, pdf) Trial protocol mmc4.pdf (1.2MB, pdf) Articles from eClinicalMedicine are provided here courtesy of Elsevier ACTIONS View on publisher site PDF (929.1 KB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top