Study Design

The BEACH trial (ClinicalTrials.gov identifier: NCT05020535) is a multicenter, randomized, double-blind, placebo-controlled phase 2a clinical trial conducted at 10 US clinical sites (supplementary material). All sites are comprehensive stroke centers and were selected from Clinical and Translational Science Awards Program sites. BEACH is directed by the University of Kentucky and the Johns Hopkins University BIOS Clinical Trials Coordinating Center. The Johns Hopkins Medicine IRB, serving as the single IRB, approved the study design. The trial will be conducted in accordance with the principles of the Declaration of Helsinki, the Good Clinical Practice guidelines of the International Council for Harmonisation, and all applicable regulatory requirements.

Patient Population

A total of 120 participants (60 per group) with acute spontaneous nontraumatic ICH are eligible for this study. The inclusion and exclusion criteria are detailed in Table 1. Written informed consent will be obtained from all participants or their legally authorized representatives.

Table 1 Inclusion and exclusion criteriaStreamlined Patient Screening

Traditionally, clinical sites have used linear and semiquantitative estimates of ICH and intraventricular hemorrhage (IVH) volume, such as ABC/2 [19], to assess eligibility for clinical trials. Errors in volume measurements occur without training and increase with large or irregularly shaped ICHs [20, 21]. More time-intensive methods, such as computed-tomography-based planimetry (CTP), demonstrate good accuracy within 5 mL or 20% of CTP volume. Recent research has demonstrated that novel concepts, such as the use of brain atlases, heuristic image quantitation, and machine learning, could provide increased diagnostic precision in volumetric assessments for patients with ICH and IVH. These methods yield significant time savings compared to traditional ABC/2 and CTP methods, potentially increasing enrollment rates for clinical studies of patients prescreened for ICH or IVH [21,22,23]. The BEACH trial uses the Viz RECRUIT platform, which uses artificial intelligence to expedite the screening process for ICH trials through automated scanning of patient computed tomography (CT) images in real time, allowing for precise volumetric assessments for suspected patients with ICH, identification of potential participants, and immediate notification of potential patients to research teams [24]. Currently, there is no US Food and Drug Administration (FDA) clearance for the ICH volume measurement. The FDA-cleared ICH commercial product is indicated for ICH detection and notification. Automatic segmentation and volumetric calculation are for investigational use only and are included in the separate research/clinical trial platform called RECRUIT [25].

Randomization and Blinding

Randomization will be performed by a computer-generated randomization schedule that will assign participants 1:1 to either MW189 or placebo. The allocation sequence will be created and stored in a protected electronic data capture system. All providers and nurses participating in the treatment plan will be excluded from data and outcome analyses. The observers who assess the clinical outcomes will be blinded to the randomization results. A safety review will be performed in a blinded manner unless the data warrants unblinding because of the nature of any safety concerns. When unblinding is required, requests will be directly forwarded to the study’s unblinded statistician by the site investigator. To further maintain blinding during study drug administration, the IV equipment will be covered. Once the study drug is diluted, it must be stored at ambient temperature and protected from light until administration.

Study Intervention

The intervention will start within 24 h of symptom onset in the intensive care unit or hospital ward (Fig. 1). Participants will be monitored daily throughout the treatment phase (randomization through day 5 or discharge if sooner) and will receive guideline-based usual care treatment for the duration of the study. Usual care treatment will follow American Heart Association guidelines for the management of ICH [26]. Participants will receive either MW189 (0.25 mg/kg) or saline administered IV over approximately 20 min every 12 ± 2 h for up to 5 days (10 doses) or until discharge, whichever comes first. University of Iowa Pharmaceuticals, an FDA-registered pharmaceutical contract manufacturing organization, prepared the MW189 drug product. It is a sterile concentrated solution in 0.9% sodium chloride at 2.5 mg/mL with a pH of roughly 2.4 that is diluted in saline prior to IV administration. The concentration of the diluted solution must not exceed 0.4 mg/mL, which will confirm that the minimum solution pH is 3.0. At a dose of 0.25 mg/kg and a diluted solution concentration of 0.4 mg/mL, the volume for a 50- to 120-kg adult will be 31–75 mL over a 20-min infusion time. Participants randomized to placebo will receive 0.9% sodium chloride, USP (US Pharmacopeia), at a volume equal to that of the active treatment as calculated based on the participant’s weight.

Fig. 1

Schematic of study design. ICH, intracerebral hemorrhage; IV, intravenously

Follow-Up Time

At each study site, trained research coordinators blinded to treatment assignments will perform follow-up visits and calls. Follow-up clinic visits will be done at 30 ± 3 days and 180 ± 14 days after drug last dose. Follow-up phone calls will be done at 90 ± 3 days after drug last dose.

Study Outcomes

The primary study outcome is the proportion of deaths from all causes during the first 7 days post randomization. Secondary outcomes include all-cause mortality at 30 days, hematoma expansion and recurrent ICH, brain infection, standard pharmacokinetic parameters (time to maximum plasma concentration [Tmax], Cmax, area under the curve, half-life [t1/2], clearance [CL], volume of distribution during terminal phase [Vz], apparent terminal elimination rate constant [λz]) determined at different times after MW189 administration, radiographic measures of edema during treatment and at the end of treatment for days 1–5, and measures of plasma and cerebrospinal fluid inflammatory and neuronal injury biomarkers during treatment and at the end of treatment for days 1–5. We will use a semiautomated threshold approach (range 44–100 Hounsfield units) to determine hematoma volume. Significant hematoma expansion will be defined as an increase between baseline and any follow-up CT scan of > 6 mL or > 33%. IVH volume will not be included in hematoma expansion. Recurrent ICH refers to the emergence of a new ICH following an initial episode of spontaneous ICH. It is subsequent spontaneous bleeding in the brain distinct from the initial event, typically occurring after a time frame of around 30 days or more. This is different from ICH expansion, which involves the enlargement of the initial bleed without the occurrence of a completely new bleeding event. Brain infection refers to infection present within the brain tissue and is evaluated based on the clinical symptoms, laboratory test results, and imaging findings. Radiographic measures of edema refer to the quantitative assessment of PHE and will be analyzed via both CT and magnetic resonance imaging (MRI) scans using semiautomated volumetrics with OsiriX software to compare the volume of edema between the two treatment groups. The primary imaging outcomes are absolute PHE and relative PHE and edema extension distance measured on brain CT scans performed at admission and at day 5 of the infusion. PHE measurements on MRI will be performed on the subset with MRI at baseline and at the end of the study. Group differences in edema volumes will be compared.

Exploratory end points include clinical and functional outcomes collected at days 30, 90, and 180 (assessed using dichotomized modified Rankin Scale scores of 0–3 vs. 4–6, the National Institutes of Health Stroke Scale, the Montreal Cognitive Assessment, the Barthel Index, the Stroke Impact Scale, and the Center for Epidemiologic Studies-Depression scale) and participant status at 180 days from ICH, including serious AEs, AEs of special interest, hemorrhage expansion and recurrent ICH, infections, and mortality.

Data and Safety Monitoring

We have established an independent Data and Safety Monitoring Board (DSMB) to ensure the safety of enrolled participants. The DSMB review will occur every 6 months or independently after enrollment of 12, 30, 60, and 120 participants. Safety events that reach or exceed the prespecified threshold will trigger “suspend recruitment and review” by the DSMB to further investigate any presumed cause and impact of any events. Termination or suspension circumstances include determining unexpected or unacceptable risks, demonstrating significant efficacy, inadequate compliance to the protocol, incomplete or inconclusive data, meeting the primary end point, or establishing futility. The trial also includes monitoring events of special interest, such as phlebitis and infection. If 75% of all participants experience one or more of these events, the DSMB will be notified for review. This assessment will be conducted after the enrollment of the first ten participants. All clinical trial data, including all AEs, will be recorded on the case report form through day 5 or discharge, whichever occurs first.

Sample Size Determination

The primary outcome of the study is to measure the proportion of participant deaths from any cause within the first 7 days after randomization. We anticipate that the mitigation of inflammation and edema will lead to fewer deaths in the MW189 arm during the acute phase. However, with an abundance of caution, we assume the placebo arm will have a similar number of deaths as compared to the MW189 arm. Therefore, we conducted a noninferiority power analysis using previously collected data from the Minimally Invasive Surgery Plus Alteplase for Intracerebral Hemorrhage Evacuation (MISTIE) III [27] and Antihypertensive Treatment of Acute Cerebral Hemorrhage II (ATACH-2) trials [28] to approximate the safety events that will be seen in the trial.

Considering a power of 80% and a one-sided 90% confidence interval, to test the null hypothesis of noninferiority with a tolerance limit of 12% and assuming a “worst-case” safety assumption of 10% rate of death in each arm, the sample size per group was estimated to be 57. We assume a low overall withdrawal rate in the first 7 days, and therefore 120 participants (60 per group) are required to evaluate the safety of MW189. This number also meets the estimated sample size for an 80% power and two-sided α to detect various assumed differences in PHE volume and maximum observed effect (Emax) values in the interleukin-10 plasma cytokine response between treatment arms.

Statistical Analyses

The intention-to-treat analysis set (all randomized participants) will be used for efficacy analysis. All participants who receive any study treatment will be included in the safety analyses. Continuous variables of normal distribution will be expressed as mean ± standard deviation, whereas continuous variables of nonnormal distribution will be expressed as median and interquartile range. Categorical variables will be represented as frequencies and percentages. We anticipate no missing data for the primary outcome, as all-cause mortality within 7 days of treatment initiation is verifiable.

The primary outcome will involve an estimate of the difference in the proportion of all-cause mortality within 7 days of treatment initiation using a 90% one-sided confidence interval on the upper bound of this difference. The null hypothesis will be rejected if this confidence interval is less than 10%. In addition, we will also estimate the treatment difference in the proportion of 7-day all-cause mortality using a multivariable logistic regression model. This model will consider age, sex, race and ethnicity, ICH volume at stability, Glasgow Coma Scale (GCS) score, presence of IVH, withdrawal of life-sustaining therapy, and ICH location (deep versus lobar). This trial encompasses infratentorial ICH, which comprises brainstem ICH classified as deep location and cerebellar ICH classified as lobar location. Age, sex, and race and ethnicity data are collected at the screening visit. The GCS score is measured at baseline. Presence of IVH and ICH location are determined based on the diagnostic CT scan. The ICH volume assessed on the stability CT scan determines the need for additional imaging to establish hematoma at 6-h intervals until the end of the enrollment period. Withdrawal of life-sustaining therapy information is collected at any time during the study period. The schedule of activities is detailed in Table 2.

Table 2 Schedule of activities

Secondary and exploratory safety outcomes will be analyzed using appropriate linear or logistic regression models and will account for treatment group, age, GCS score, presence of IVH, ICH volumes, the timing of drug administration, and hematoma location. No imputation of missing safety end points will be used. Instead, sensitivity analyses will be performed, whereby participants with missing secondary safety data will be assumed to have had a safety event and then not to have had a safety event. The analyses will account for the variability in dosing and treatment duration, which can range from 2 to 10 doses depending on the date of discharge. This variability will be addressed through various approaches, such as categorizing participants into dose groups, treating the number of doses as a continuous variable, and calculating area under the curve that considers the dose and duration of the doses given. The point estimate of the proportion of total safety events and corresponding two-sided 95% confidence intervals will be calculated under both extreme scenarios to provide estimates of uncertainty in the presence of these missing data. SAS, STATA 17, and/or R software will be used for statistical analysis. Statistical significance will be set at p < 0.05.