Abstract
Background:
Neutrophil-mediated inflammation in the acute phase of intracerebral hemorrhage (ICH) worsens outcome in pre-clinical studies. Soluble intercellular adhesion molecule-1 (sICAM-1), an inducible ligand for integrins and cell-cell adhesion molecules, is critical for neutrophil extravasation. We aimed to determine whether serum levels of sICAM-1 are associated with worse outcomes after ICH.
Methods:
We conducted a post-hoc secondary analysis of an observational cohort using data from the FAST (Factor-VII-for-Acute-Hemorrhagic-Stroke-Treatment) trial. The study exposure was the admission serum level of sICAM-1. The co-primary outcomes were mortality and poor outcome (modified Rankin Score 4-6) at 90 days. Secondary radiological outcomes were hematoma expansion at 24 hours and perihematomal edema expansion at 72 hours. We used multiple linear and logistic regression analyses to test for associations between sICAM-1 and outcomes, after adjustment for demographics, ICH severity characteristics, change in the systolic blood pressure in the first 24 hours, treatment randomization arm, and the time from symptom onset to study drug administration.
Results:
Of 841 patients, we included 507 (60%) with complete data. Hematoma expansion occurred in 169 (33%), while 242 (48%) had a poor outcome. In multivariable analyses, sICAM-1 was associated with mortality (OR, 1.53 per SD increase; 95% CI, 1.15-2.03) and poor outcome (OR, 1.34 per SD increase; CI, 1.06-1.69). In multivariable analyses of secondary outcomes, sICAM-1 was associated with hematoma expansion (OR, 1.35 per SD increase, CI, 1.11-1.66), but was not associated with log-transformed perihematomal edema expansion at 72 hours. In additional analyses stratified by treatment assignment, similar results were noted in the recombinant activated Factor VII (rFVIIa) arm, but not in the placebo arm.
Conclusions:
Admission serum levels of sICAM-1 were associated with mortality, poor outcome, and hematoma expansion. Given the possibility of a biological interaction between rFVIIa and sICAM-1, these findings highlight the need to further explore the role of sICAM-1 as a potential marker of poor ICH outcomes.
Graphical Abstract
INTRODUCTION
Intracerebral hemorrhage (ICH) is the most disabling type of stroke and causes about 70% of all stroke-related early mortality.1-3 The hyperacute phase of ICH encompasses primary injury caused by the hematoma and its expansion, and secondary injury that results from the activation of inflammation and coagulation cascades.4-7 The pathophysiologic hallmark of early brain inflammation after ICH is neutrophil extravasation, which occurs within 4-5 hours of ICH onset and peaks at 3 days.8 Neutrophils damage the brain by producing reactive oxygen species, releasing proinflammatory proteases, affecting blood–brain barrier permeability, and promoting neuronal death.9,10 Pre-clinical studies have shown that neutrophil-mediated inflammation in the acute phase of ICH contributes to worsening of tissue injury and worse outcome.11 Therefore, molecules involved in neutrophil activation may be early prognostic markers after ICH.
Intercellular adhesion molecule-1 (ICAM-1), an inducible ligand for integrins and cell-cell adhesion molecules, is critical for neutrophil extravasation. The induction of ICAM-1 in neurons could promote the attachment of leukocytes to neurons inducing neuronal injury through direct cell-to-cell interaction and the release of cytotoxic substances.12 ICAM-1 hence serves as a gate keeper for neutrophils aiming to invade perihematomal brain tissue.13-15 ICAM-1 can be measured both in its cell-bound form through histological staining and in its soluble circulating form (soluble or sICAM-1) through antibody assays.16 Emerging clinical data suggest that sICAM-1 levels on admission are associated with poor outcomes after acute ischemic stroke.17 Although the timing of peak neutrophil activity coincides with the phase of hematoma expansion or peak perihematomal edema clinically, whether serum levels of sICAM-1 influence outcomes after ICH is poorly understood. We therefore sought to study the relationship between admission serum sICAM-1 levels and ICH outcomes after ICH using prospectively collected data from FAST (Factor VII for Acute ICH), a large randomized trial, where patients with an ICH were randomized to receive recombinant activated Factor VII (rFVIIa) or placebo.18 The trial showed that hemostatic therapy with rFVIIa reduced hematoma growth but did not improve survival or functional outcome.18 We hypothesized that higher serum sICAM-1 levels on admission are associated with poor functional outcomes and neuroimaging markers of ICH severity.
METHODS
Data Availability
Data from the FAST trial cannot be shared by the authors; however, anonymized data from the FAST trial are available by request through Novo Nordisk (https://www.novonordisk-trials.com).
Study Design and Participants
We performed a post-hoc analysis of FAST, a multicenter, randomized, double-blind, placebo-controlled trial of recombinant activated Factor VIIa (rFVIIa) for the treatment of spontaneous ICH.6 FAST eligibility criteria were adult patients with a spontaneous ICH on brain computed tomography (CT) within 3 hours of symptom onset. FAST exclusion criteria are listed in the supplementary material.
A total of 841 patients were enrolled between May 2005 and February 2007 at 122 sites across 22 countries. Institutional review board approval was obtained at each enrolling trial site. Subjects were randomly assigned to receive placebo, 20 or 80 μg/kg of recombinant factor VIIa (rFVIIa) intravenously within 4 hours of ICH onset and within 60 minutes of the baseline CT. Written informed consent was provided by each patient or their legally authorized representative for FAST participation. Given that de-identified data were used, our post-hoc study was exempt from additional local IRB approval. Results were reported in adherence to the STROBE guidelines.19
Measurements
The study exposure was the serum level of sICAM-1, which was collected at the baseline study visit, within 4 hours of ICH symptom onset. Blood samples were frozen, stored at −80°C, and analyzed in batches at a central core lab facility.20 Serum sICAM-1 levels were measured with commercially available enzyme-linked immunosorbent assay (ELISA).
Neuroimaging Protocol
Patients enrolled in the FAST trial underwent a CT scan of the head at baseline, and at 24 and 72 hours after administration of the investigational product. Hematoma (intracerebral and intraventricular hemorrhage) and perihematomal edema (PHE) volumes were calculated semi-automatically using computerized planimetric methods (AnalyzeDirect, Overland Park, KS, USA) and read by two neuroradiologists who were blinded to randomization.13
Outcomes
The outcomes used in this study were similar to those ascertained in the FAST trial. The co-primary outcomes were 90-day mortality, and a composite of death or major disability (poor outcome), defined as a modified Rankin Score (mRS) 4-6 at 90 days. Secondary radiological outcomes included hematoma expansion at 24 hours and PHE expansion at 72 hours. Consistent with previous literature, hematoma expansion was defined as growth in hematoma volume by either ≥33% or by ≥6 mL from baseline to 24 hours.21 PHE expansion was defined as the difference in the PHE volumes between 72 hours and baseline.6,22
Statistical Analysis
We used standard descriptive statistics to report and compare baseline characteristics (online supplement). sICAM-1 was not normally distributed. We therefore considered sICAM-1 units per standard deviation change. In the primary analyses, we used logistic regression to evaluate the association between serum sICAM-1 levels and functional outcomes (mortality and poor outcome). The covariates were selected based on prior literature establishing associations between the covariates and outcome after ICH.23 The covariates included in the multivariable logistic regression analysis were age, sex, baseline ICH volume, ICH location, baseline Glasgow Coma Scale, presence of intraventricular hemorrhage, change in systolic blood pressure in 24 hours24, treatment randomization, and time from symptom onset to administration of the study drug. We also studied the relationship between quartiles of sICAM-1 levels and outcomes.
In secondary analyses, we studied the relationship between serum sICAM-1 levels on radiological outcomes (hematoma expansion and PHE expansion). We used multiple logistic regression to evaluate for hematoma expansion (dichotomized), after adjustment for the covariates listed in the primary analysis. We subsequently used linear regression to evaluate PHE expansion and hematoma expansion as continuous outcome measures. Hematoma expansion and PHE expansion were natural log-transformed. The multivariable analyses were adjusted for covariates similar to that of the primary analysis. We ensured that the assumptions for linear regression, regarding linearity, homoscedasticity, and multicollinearity were met. Multicollinearity, or the absence thereof, was identified using a variance inflation factor value higher than 4. The threshold for statistical significance was a 2-sided P <0.05. Statistical analyses were performed using Stata, release 16 (StataCorp LLC, College Station, TX, USA). Data analyses were performed between August 1, 2022, and February 21, 2023.
RESULTS
Cohort Characteristics
Among the 842 patients enrolled in the FAST trial, we excluded 23 due to missing outcomes variables, and 312 due to missing sICAM-1 laboratory data, resulting in a final analytical cohort of 507 patients (Figure 1, Table S1). The mean [standard deviation] age of the final cohort was 64 [13] years, and 191 [38%] patients were female. A total of 262 (47.7%) had poor outcome, and 97 (19.1%) patients died at 90 days. Patients with poor outcome (mRS 4-6 at 90 days) compared to those with good outcome were significantly older, had a higher prevalence of diabetes, had larger changes in systolic blood pressure between admission and 24-hour measurements, and had more severe ICH characteristics (Table 1).
Figure 1.
Patients Inclusion Criteria
Table 1.
Baseline Characteristics of Patients with Intracerebral Hemorrhage, Stratified by Poor Outcome
| Characteristic | Good Outcome N = 265 |
Poor Outcome N= 242 |
P value |
|---|---|---|---|
| Age, y, mean (SD) | 60.9 (11.7) | 67.3 (13.3) | <0.001 |
| Female | 103 (38.9) | 88 (36.4) | 0.56 |
| Hypertensiona | 230 (94.6) | 201 (91.4) | 0.16 |
| Hyperlipidemiaa | 82 (33.7) | 64 (20.1) | 0.28 |
| Diabetes mellitusa | 22 (9.1) | 36 (16.4) | 0.02 |
| Coronary artery diseasea | 11 (4.5) | 15 (6.8) | 0.29 |
| Atrial fibrillationa | 8 (3.3) | 9 (4.1) | 0.65 |
| Prior ischemic stroke or TIAa | 14 (5.8) | 13 (5.9) | 0.95 |
| ICH volume, baseline, mLb | 9.4 (4.4-17.2) | 27.3 (10.6-48.7) | <0.001 |
| PHE volume, baseline, mLb | 10.1 (4.8-16.8) | 22.8 (10.5-41.4) | <0.001 |
| IVH volume, baseline, mLb | 2.6 (0-5.4) | 5.3 (2.2-12.4) | 0.001 |
| Presence of IVH | 78 (29.6) | 107 (44.8) | <0.001 |
| ICH location | 0.09 | ||
| Lobar | 31 (11.7) | 41 (16.9) | |
| Nonlobar | 234 (88.3) | 201 (83.1) | |
| Hematoma expansion | 47 (17.8) | 122 (50.4) | <0.001 |
| Admission SBP, mm Hg | 177 (159-198) | 178 (160-198) | 0.35 |
| Delta SBP in 24 hours, mm Hg | 81 (62-100) | 88 (70-105) | 0.01 |
| Treatment assignment | 0.20 | ||
| 0 μg | 74 (27.9) | 85 (35.1) | |
| 80 μg | 94 (35.5) | 74 (30.6) | |
| Placebo | 97 (36.6) | 83 (34.3) | |
| sICAM-1, ng/mL | 349 (132.0) | 380 (159.5) | 0.01 |
| IL-6, ng/L | 3.6 (4.7) | 5.6 (7.7) | <0.001 |
| MMP-3, ng/L | 18.4 (12.6) | 18.2 (13.1) | 0.05 |
| MMP-9, ng/L | 764 (426) | 834 (545) | 0.36 |
| TNF-alpha, ng/L | 2.9 (1.3) | 2.8 (1.4) | 0.26 |
Abbreviations: ICH, intracerebral hemorrhage; IL-6, interleukin-6; IVH, intraventricular hemorrhage; mL, milliliters; MMP, matrix metalloproteinase; PHE, perihematomal edema; SBP, systolic blood pressure; SD, standard deviation; sICAM-1, soluble intercellular adhesion molecule-1; TIA, transient ischemic attack; TNF, tumor necrosis factor.
Data are represented as number (%) unless otherwise stated.
denominator=463 patients (243 with good outcome, 220 with poor outcome)
data shown as median (interquartile range)
The median [IQR] sICAM-1 level in the analytical cohort was 340 [280-414] ng/mL. Those with poor outcome had higher mean [SD] sICAM-1 levels than those with good outcome (380 [160] versus 349 [132] ng/mL, p=0.01). Baseline characteristics of the analytical cohort stratified by quartiles of sICAM-1 levels are shown in Table 2. As previously reported, patients with poor outcome had higher mean interleukin-6 serum levels on admission (5.6 versus 3.6 p<0.001) (Table 1).20
Table 2.
Baseline Characteristics of Patients with Intracerebral Hemorrhage, Stratified by Quartiles of sICAM-1
| Characteristic | Quartile 1 N = 128 |
Quartile 2 N= 129 |
Quartile 3 N =124 |
Quartile 4 N = 126 |
P value |
|---|---|---|---|---|---|
| sICAM-1 levels, ng/mL, median (interquartile range) | 240 (202-264) | 313 (300-326) | 380 (360-394) | 496 (440-600) | NA |
| sICAM-1 levels, ng/mL, mean(SD) | 220 (65) | 313 (17) | 376 (20) | 549 (151) | NA |
| Age, y, mean (SD) | 61.0 (12.9) | 65.6 (12.1) | 64.7 (13.2) | 63.8 (13.3) | 0.72 |
| Female | 44 (34.3) | 42 (32.6) | 45 (36.3) | 56 (44.4) | 0.23 |
| Hypertensiona | 106 (93.8) | 108 (90.0) | 109 (93.7) | 113 (94.6) | 0.44 |
| Hyperlipidemiaa | 29 (35.7) | 42 (35.0) | 40 (34.5) | 36 (30.3) | 0.39 |
| Diabetes mellitusa | 12 (10.7) | 14 (11.7) | 11 (9.5) | 21 (17.7) | 0.23 |
| Coronary artery diseasea | 5 (4.5) | 8 (6.7) | 7 (6.0) | 6 (5.0) | 0.88 |
| Atrial Fibrillationa | 3 (2.7) | 5 (4.2) | 4 (3.5) | 5 (4.2) | 0.91 |
| Prior ischemic stroke or TIAa | 6 (5.3) | 3 (2.5) | 13 (11.2) | 6 (5.0) | 0.04 |
| Baseline GCSb | 14 (12-15) | 14 (13-15) | 13 (13-15) | 12 (12-15) | 0.04 |
| ICH volume, baseline, mL | 17.1 (16.3) | 19.6 (17.8) | 22.4 (25.2) | 26.9 (27.1) | 0.02 |
| PHE volume, baseline, mL | 18.8 (16.9) | 20.3 (21.0) | 22.8 (25.1) | 25.3 (24.5) | 0.04 |
| IVH volume, mL | 3.3 (0.9) | 3.8 (1.1) | 4.1 (1.2) | 4.4 (1.3) | 0.03 |
| Presence of IVH | 45 (34.9) | 51 (39.8) | 44 (35.5) | 45 (35.7) | 0.84 |
| ICH location | 0.87 | ||||
| Lobar | 16 (12.5) | 20 (15.7) | 17 (13.7) | 19 (15.1) | |
| Nonlobar | 113 (87.5) | 107 (84.3) | 107 (86.3) | 107 (84.9) | |
| Treatment assignment | 0.81 | ||||
| 20 μg | 46 (35.4) | 36 (27.7) | 40 (32.0) | 38 (29.9) | |
| 40 μg | 41 (32.3) | 42 (32.3) | 44 (35.2) | 44 (34.6) | |
| Placebo | 41 (32.3) | 51 (40.0) | 40 (32.8) | 44 (35.3) | |
| Absolute hematoma growth from baseline to 24 hours, mL | 2.8 (13.7) | 4.5 (16.2) | 5.3 (11.8) | 8.9 (10.7) | 0.03 |
| Hematoma expansion | 35 (26.9) | 34 (26.2) | 48 (38.4) | 56 (44.1) | 0.004 |
| Mortality | 21 (16.2) | 22 (17.2) | 19 (15.3) | 35 (28.0) | 0.03 |
| Poor Outcome (mRS 4-6) | 56 (43.1) | 52 (40.6) | 61 (49.2) | 73 (58.4) | 0.02 |
Abbreviations: GCS, Glasgow Coma Scale; ICH, intracerebral hemorrhage; IVH, intraventricular hemorrhage; mL, milliliters; PHE, perihematomal edema; SD, standard deviation; sICAM-1, soluble intercellular adhesion molecule-1; TIA, transient ischemic attack.
Data are represented as number (%) unless otherwise stated.
denominator=468 patients
Association Between Baseline sICAM-1 Levels and Functional Outcome
In the unadjusted logistic regression analysis (Table 3), admission sICAM-1 serum levels were associated with poor outcome (odds ratio [OR], 1.24; 95% confidence interval [CI], 1.03-1.49), and with mortality at 90 days (OR, 1.38; CI, 1.13-1.70). In multivariable analyses, sICAM-1 was associated with poor outcome (OR, 1.34 per SD increase; CI,1.06-1.69) and with mortality (OR, 1.53 per SD increase; CI, 1.15-2.03). In an additional analysis of quartiles of sICAM-1, we found that patients in the 4th quartile had a nearly 2-fold increased odds of hematoma expansion (Figure 2A) and poor outcome (Figure 2B) compared to those in the first quartile. Furthermore, there was a dose-response relationship between sICAM-1 quartiles and outcomes. Analyses of sICAM-1 per ng/L increase are shown in Table S2.
Table 3.
Multivariable Analyses of sICAM-1 and ICH Outcomes
| Covariate | Modified Rankin Score 4-6 | Mortality | ||
|---|---|---|---|---|
| OR (95% CI) | P value | OR (95% CI) | P value | |
| sICAM-1, unadjusted | 1.24 (1.03=−1.49) | 0.02 | 1.38 (1.13-1.70) | 0.002 |
| sICAM-1, per SD# | 1.34 (1.06-1.69) | 0.013 | 1.53 (1.15-2.03) | 0.03 |
| P value for interaction between sICAM-1 and treatment assignment = 0.97 | P value for interaction between sICAM-1 and treatment assignment = 0.45 | |||
| Recombinant Factor VII Arm (N= 348) | ||||
| sICAM-1, unadjusted | 1.30 (1.04-1.64) | 0.02 | 1.49 (1.16-1.92) | 0.002 |
| sICAM-1, per SD* | 1.36 (1.01-1.82) | 0.04 | 1.66 (1.18-2.32) | 0.003 |
| Placebo Arm (N= 159) | ||||
| sICAM-1, unadjusted | 1.14 (0.84-1.55) | 0.38 | 1.20 (0.85-1.70) | 0.31 |
| sICAM-1, per SD* | 1.26 (0.86-1.83) | 0.24 | 1.39 (0.79-2.46) | 0.25 |
Abbreviations: CI, confidence interval; sICAM-1, soluble intercellular adhesion molecule-1; ICH, intracerebral hemorrhage; IVH, intraventricular hemorrhage; OR, odds ratio; PHE, perihematomal edema; SD, standard deviation.
Adjusted for age, sex, baseline ICH volume, baseline IVH volume, ICH location, baseline Glasgow Coma Scale, change in systolic blood pressure in 24 hours, treatment randomization, and time from symptom onset to administration of the study drug.
Adjusted for age, sex, baseline ICH volume, baseline IVH volume, ICH location, baseline Glasgow Coma Scale, and time from symptom onset to administration of the study drug.
Figure 2.
Odds of hematoma expansion and poor outcome by sICAM-1 admission level.
Odds of hematoma expansion between admission and 24-hour scan (panel A), and odds of poor outcome (modified Rankin score 4-6) at 90 days by sICAM-1 quartiles (panel B). The first quartile serves as the reference in each figure. ORs are plotted on a logarithmic scale. Hematoma expansion defined as growth in hematoma volume of >=33% or >=6ml. sICAM-1, soluble intercellular adhesion molecule-1.
Association Between Baseline sICAM-1 Levels and Radiological Outcomes
In unadjusted logistic and linear regression analyses, baseline sICAM-1 serum levels were associated with hematoma expansion (dichotomized, OR, 1.34 per SD increase; CI, 1.11-1.61), but not with PHE expansion (β, 0.14; CI, 0.04-0.23) (Table 3). Hematoma expansion as continuous outcome variable showed similar associations with sICAM-1 (β, 2.64; CI, 1.26-4.01) levels as dichotomized hematoma expansion. After adjustment of covariates, sICAM-1 levels remained associated with hematoma expansion (OR, 1.35 per SD increase; CI, 1.11-1.66) and with hematoma expansion analyzed as continuous outcome variable (β; 3.83; CI, 1.46-6.22), but not with PHE expansion (Table 4). An analysis stratified by sICAM-1 quartiles demonstrated a more than 2-fold increased odds of hematoma expansion in the 4th quartile (Figure 2A). Analyses of sICAM-1 per ng/L increase are shown in Table S3.
Table 4.
Multivariable Analyses of sICAM-1 and Radiological Characteristics
| Covariate | Hematoma Expansion# (Dichotomous Outcome) |
Hematoma Expansion# (Continuous Outcome) |
PHE Expansion# | |||
|---|---|---|---|---|---|---|
| OR (95% CI) | P value |
Beta (95% CI) | P value | Beta (95% CI) | P value | |
| sICAM-1, per SD, unadjusted | 1.34 (1.11-1.61) | 0.002 | 2.64 (1.26-4.01) | <0.001 | 0.14 (0.04 to 0.23) | 0.22 |
| sICAM-1, per SD# | 1.35 (1.11-1.66) | 0.003 | 3.83 (1.45-6.22) | <0.001 | 0.08 (−0.02 to 0.19) | 0.14 |
| P value for interaction between sICAM-1 and treatment assignment = 0.98 | P value for interaction between sICAM-1 and treatment assignment = 0.33 | P value for interaction between sICAM-1 and treatment assignment = 0.29 | ||||
| Recombinant Factor VII Arm (N= 348) | ||||||
| sICAM-1, per SD, unadjusted | 1.41 (1.12-1.79) | 0.003 | 2.33 (1.05-4.20) | 0.007 | 0.14 (−0.01 to 0.29) | 0.07 |
| sICAM-1, per SD* | 1.36 (1.1-1.73) | 0.01 | 2.45 (1.10-4.23) | 0.007 | 0.11 (−0.01 to 0.24) | 0.09 |
| Placebo Arm (N= 159) | ||||||
| sICAM-1, per SD, unadjusted | 1.26 (0.92-1.71) | 0.15 | 3.46 (1.07-5.85) | 0.005 | −0.04 (−0.27 to 0.19) | 0.71 |
| sICAM-1, per SD* | 1.40 (0.96-2.06) | 0.07 | 3.97 (1.63-6.32) | 0.001 | 0.05 (−0.13 to 0.23) | 0.62 |
Abbreviations: CI, confidence interval; sICAM-1, soluble intercellular adhesion molecule-1; ICH, intracerebral hemorrhage; IVH, intraventricular hemorrhage; OR, odds ratio; PHE, perihematomal edema; SD, standard deviation; SE, standard error.
PHE volumes were natural log-transformed.
Adjusted for age, sex, baseline ICH volume, baseline IVH volume, ICH location, baseline Glasgow Coma Scale, change in systolic blood pressure in 24 hours, treatment randomization, and time from symptom onset to administration of the study drug.
Adjusted for age, sex, baseline ICH volume, baseline IVH volume, ICH location, baseline Glasgow Coma Scale, change in systolic blood pressure in 24 hours, and time from symptom onset to administration of the study drug.
Post hoc analyses stratified by FAST treatment assignment
Given that many patients received rFVIIa raising the possibility of unknown interactions between sICAM-1 and rFVIIa, we stratified the analyses based on treatment assignment, i.e. treatment with rFVIIa (patients who received high and low dose rFVIIa were pooled) vs. placebo. In these subgroups, ORs and point estimates, respectively, were overall similar compared to the analysis that included the entire cohort. However, the fully adjusted multivariable models showed associations only between sICAM-1 levels and poor functional outcome, mortality, and hematoma expansion in the rFVIIa group, similar to that of the primary analysis. The remaining associations in the rFVIIa group, and all multivariable models in the placebo arm, did not show significant associations with our primary and secondary outcomes (Tables 3 and 4).
DISCUSSION
In this post-hoc analysis of the FAST trial in patients with spontaneous ICH, we found that serum levels of sICAM-1 on admission were independently associated with mortality and poor outcome at 90 days, and hematoma expansion at 24 hours, but not with PHE expansion. In stratified analyses by treatment assignment, while similar results were noted in the rFVIIa group, there was no relationship between sICAM-1 and outcomes in the placebo arm, with the exception of hematoma expansion.
Serum levels of sICAM-1 have been correlated with an increased risk of future cardiovascular events including death from coronary artery disease, non-fatal myocardial infarction, and ischemic stroke.25-28 Furthermore, among patients with an acute ischemic stroke, higher serum sICAM-1 levels influenced poor outcomes.17 A similar relationship in patients with ICH had previously only been reported in case series and small cohort studies including up to 60 patients with ICH. These studies have suggested an association between sICAM-1 levels in peripheral blood or cerebrospinal fluid and poor functional outcome29,30; however, the low statistical power precluded adjustment for covariates that influenced outcome. In this context, our findings from a well-powered, carefully adjusted pre-specified post hoc analysis of data from the FAST trial shed more light on the relationship between sICAM-1 and ICH outcomes. We observed a nearly 2-fold increased odds of poor outcome, particularly in patients in the highest sICAM-1 quartile, suggesting a dose-response relationship, and an association with hematoma expansion.
Several factors may explain the association of sICAM-1 levels with poor outcome. First, higher sICAM-1 levels were associated with larger baseline PHE volumes and higher rates of hematoma expansion, both of which can independently lead to poor functional outcome.18,31,32 Second, sICAM-1 activation may reflect a global activation of inflammation pathways resulting in surges in the levels of cytokines and chemokines which subsequently lead to sICAM-1 activation.7,10,33,34 However, this phenomenon fails to explain why prior analyses of several cytokines and chemokines, with the exception of matrix metalloproteinase, have not shown an association with hematoma expansion.20,34-36 In fact, a recent post hoc analysis of the FAST trial reported poor outcomes with higher baseline interleukin (IL)-6 levels, but there was no association with hematoma expansion.20 In that regard, our finding of higher rates of hematoma expansion with sICAM-1 is important since sICAM-1 may be an early biomarker for hematoma expansion, arguably one of the most important prognostic factors after ICH.24 A plausible explanation for this finding may be that sICAM-1 is a fairly accurate surrogate for neutrophil activation, the timing of which coincides with the phase of peak hematoma and PHE volume expansion.37 A prior study has shown an inverse relationship between serum neutrophil counts and hematoma expansion, in line with our speculation of increased neutrophil extravasation mediated by increased sICAM-1-assisted neutrophil transendothelial migration.38 Third, poor outcomes with sICAM-1 may also highlight an underlying systemic inflammatory response syndrome which is associated with infection, another marker for poor outcome after ICH.39,40 The dataset made available to us lacked information about infections and we could therefore not evaluate this relationship in more detail.
In our post-hoc analysis, stratified by FAST treatment versus placebo arms, ORs and point estimates were similar compared to the main analysis, indicating internal consistency of our data across treatment groups. However, the associations with the primary and with the secondary outcomes were not statistically significant in the placebo group. Because this cannot solely be attributed to lower statistical power in the segmented data set (n=159 observations were sufficient for the co-variates in the models), our results need to be reproduced in future studies to confirm their clinical significance.
Limitations
Our study has several limitations. First, this was an observational study that did not provide evidence of a causal role of sICAM-1 leading to the radiographic or functional outcomes assessed here. It cannot be excluded that sICAM-1 is merely a bystander in a molecular cascade driven by other factors. Although there was no statistically significant interaction between treatment assignment and sICAM-1 for outcomes, and the point estimates were similar in the rFVIIa and placebo groups, the smaller sample size of the latter likely contributed to the lack of statistical significance of associations between sICAM-1 and outcomes when the placebo group was analyzed separately. In addition, the possibility of a biologic interaction between rFVIIa and sICAM-1 cannot be ruled out, warranting cautious interpretation of our results. Second, the selection of only those participants with complete clinical and laboratory data (~60% of the patients enrolled in the FAST trial) possibly introduced selection bias, although there were no significant baseline differences between patients with and without missing data on sICAM-1. Third, the data made available to us did not include the time interval between symptom onset to admission CT, but instead the time between symptom onset and administration of the investigational product which is only a proxy of the former, but may have led to residual confounding of both initial PHE volumes and hematoma expansion rates. Fourth, information on antiplatelet use prior to the index ICH was not available.24,41 Fifth, the majority of patients in the FAST trial only had a single measurement of sICAM, and it is possible that changes in the levels over time may have played a role in the evolution of PHE and longer term clinical outcomes. While serial sICAM-1 measurements have been conducted in patients with subarachnoid hemorrhage42,43, we are not aware of similar studies in patients with ICH. Though serial sICAM-1 data were available, the majority of patients had missing data after the baseline sICAM-1 level ascertainment, which precluded any meaningful analysis. Sixth, data on sICAM-1 in cerebrospinal fluid was not available, which may have provided further mechanistic insight. Lastly, our findings may not be generalizable to the larger ICH population since patients enrolled in FAST were subject to rigorous inclusion and exclusion criteria.
Conclusions
We found that serum levels of sICAM-1 on hospital admission for ICH in patients enrolled in the FAST trial, were associated with increased odds of poor functional outcome and less favorable radiological outcomes including larger baseline PHE volume and higher rates of hematoma expansion. These findings suggest that admission serum sICAM-1 levels may be a marker of poor outcomes in the hyperacute phase of ICH. Further studies are needed to shed light on the underlying mechanisms linking sICAM-1 with radiological and clinical outcomes to determine whether sICAM-1 might be a viable therapeutic target to benefit patients with an ICH, and to exclude the possibility of a biological interaction between rFVIIa and sICAM-1.
Supplementary Material
Sources of Funding
The study was funded by the NIH (K23NS105948) to Dr. Murthy.
Disclosures
Dr. Witsch reports fees for medicolegal consulting and service as an editorial team member of Neurology Resident & Fellow section for the American Academy of Neurology. Dr Roh reports funding from Portola Pharmaceuticals, NIH, U.S. Department of Defense, and National Blood Foundation, outside this work. Dr. Iadecola serves on the Scientific Advisory Board of Broadview Ventures. Dr. Diaz-Arrastia serves on the Scientific Advisory Board, has received grant support, holds stock options in BrainBox Solutions, is a consultant for and has received in-kind research support from MesoScale Discoveries, is a Consultant for and holds stock options from NovaSignal and Nia Technologies, and has a research collaboration with NeurAegis, all outside this work. Dr Kasner receives or has received grant support from Bayer, WL Gore, Bristol-Myers Squibb, Medtronic, Genentech, Remedy; personal compensation from diamedica, consulting fees from Bristol-Myers Squibb, AstraZeneca; and royalties from UpToDate; all outside this work. Dr Mayer serves as the medical safety monitor for the NIH–funded FASTEST trial (Factor VIAAa for Intracerebral Hemorrhage Treatment at Earliest Possible Timepoint; U01 NS110772, which receives in-kind study drug from Novo Nordisk, and compensation from MaxQAI for consultant services. Dr. Murthy reports fees for medicolegal consulting, and is supported by the NIH (K23NS105948). The remaining authors report no disclosures.
Non-Standard Abbreviations
- FAST
Factor VII for Acute ICH
- ICH
Intracerebral hemorrhage
- rFVIIa
recombinant activated factor VII
- PHE
Perihematomal edema
- sICAM-1
soluble Intercellular adhesion molecule-1
REFERENCES
- 1.Lioutas VA, Beiser AS, Aparicio HJ, Himali JJ, Selim MH, Romero JR, Seshadri S. Assessment of Incidence and Risk Factors of Intracerebral Hemorrhage Among Participants in the Framingham Heart Study Between 1948 and 2016. JAMA Neurol. 2020;77:1252–1260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Chen Y, Wright N, Guo Y, Turnbull I, Kartsonaki C, Yang L, Bian Z, Pei P, Pan D, Zhang Y, et al. Mortality and recurrent vascular events after first incident stroke: a 9-year community-based study of 0.5 million Chinese adults. Lancet Glob Health. 2020;8:e580–e590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Witsch J, Falcone GJ, Leasure AC, Matouk C, Endres M, Sansing L, Woo D, Sheth KN. Intracerebral Hemorrhage with Intraventricular Extension Associated with Loss of Consciousness at Symptom Onset. Neurocrit Care. 2021;35:418–427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shoamanesh A, Katsanos AH. Combatting Secondary Injury From Intracerebral Hemorrhage With Supplemental Antioxidant Therapy. Stroke. 2021;52:1182–1184. [DOI] [PubMed] [Google Scholar]
- 5.Witsch J, Roh DJ, Avadhani R, Merkler AE, Kamel H, Awad I, Hanley DF, Ziai WC, Murthy SB. Association Between Intraventricular Alteplase Use and Parenchymal Hematoma Volume in Patients With Spontaneous Intracerebral Hemorrhage and Intraventricular Hemorrhage. JAMA Netw Open. 2021;4:e2135773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Murthy SB, Moradiya Y, Dawson J, Lees KR, Hanley DF, Ziai WC, Collaborators V-I. Perihematomal Edema and Functional Outcomes in Intracerebral Hemorrhage: Influence of Hematoma Volume and Location. Stroke. 2015;46:3088–3092. [DOI] [PubMed] [Google Scholar]
- 7.Ziai WC. Hematology and inflammatory signaling of intracerebral hemorrhage. Stroke. 2013;44:S74–78. doi: 10.1161/STROKEAHA.111.000662 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Nguyen HX, O'Barr TJ, Anderson AJ. Polymorphonuclear leukocytes promote neurotoxicity through release of matrix metalloproteinases, reactive oxygen species, and TNF-alpha. J Neurochem. 2007;102:900–912. [DOI] [PubMed] [Google Scholar]
- 9.Joice SL, Mydeen F, Couraud PO, Weksler BB, Romero IA, Fraser PA, Easton AS. Modulation of blood-brain barrier permeability by neutrophils: in vitro and in vivo studies. Brain research. 2009;1298:13–23. [DOI] [PubMed] [Google Scholar]
- 10.Wang J, Dore S. Inflammation after intracerebral hemorrhage. J Cereb Blood Flow Metab. 2007;27:894–908. [DOI] [PubMed] [Google Scholar]
- 11.Sansing LH, Harris TH, Kasner SE, Hunter CA, Kariko K. Neutrophil depletion diminishes monocyte infiltration and improves functional outcome after experimental intracerebral hemorrhage. Acta neurochirurgica Supplement. 2011;111:173–178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gong C, Hoff JT, Keep RF. Acute inflammatory reaction following experimental intracerebral hemorrhage in rat. Brain research. 2000;871:57–65. [DOI] [PubMed] [Google Scholar]
- 13.Lyck R, Enzmann G. The physiological roles of ICAM-1 and ICAM-2 in neutrophil migration into tissues. Curr Opin Hematol. 2015;22:53–59. [DOI] [PubMed] [Google Scholar]
- 14.Gearing AJ, Hemingway I, Pigott R, Hughes J, Rees AJ, Cashman SJ. Soluble forms of vascular adhesion molecules, E-selectin, ICAM-1, and VCAM-1: pathological significance. Annals of the New York Academy of Sciences. 1992;667:324–331. [DOI] [PubMed] [Google Scholar]
- 15.Margraf A, Germena G, Drexler HCA, Rossaint J, Ludwig N, Prystaj B, Mersmann S, Thomas K, Block H, Gottschlich W, et al. The integrin-linked kinase is required for chemokine-triggered high-affinity conformation of the neutrophil beta2-integrin LFA-1. Blood. 2020;136:2200–2205. [DOI] [PubMed] [Google Scholar]
- 16.Seth R, Raymond FD, Makgoba MW. Circulating ICAM-1 isoforms: diagnostic prospects for inflammatory and immune disorders. Lancet. 1991;338:83–84. [DOI] [PubMed] [Google Scholar]
- 17.Wang L, Chen Y, Feng D, Wang X. Serum ICAM-1 as a Predictor of Prognosis in Patients with Acute Ischemic Stroke. BioMed research international. 2021;2021:5539304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kuohn LR, Witsch J, Steiner T, Sheth KN, Kamel H, Navi BB, Merkler AE, Murthy SB, Mayer SA. Early Deterioration, Hematoma Expansion, and Outcomes in Deep Versus Lobar Intracerebral Hemorrhage: The FAST Trial. Stroke. 2022;53:2441–2448. [DOI] [PubMed] [Google Scholar]
- 19.von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, Initiative S. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147:573–577. [DOI] [PubMed] [Google Scholar]
- 20.Leasure AC, Kuohn LR, Vanent KN, Bevers MB, Kimberly WT, Steiner T, Mayer SA, Matouk CC, Sansing LH, Falcone GJ, et al. Association of Serum IL-6 (Interleukin 6) With Functional Outcome After Intracerebral Hemorrhage. Stroke. 2021;52:1733–1740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dowlatshahi D, Demchuk AM, Flaherty ML, Ali M, Lyden PL, Smith EE. Defining hematoma expansion in intracerebral hemorrhage: relationship with patient outcomes. Neurology. 2011;76:1238–1244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Murthy SB, Urday S, Beslow LA, Dawson J, Lees K, Kimberly WT, Iadecola C, Kamel H, Hanley DF, Sheth KN, et al. Rate of perihaematomal oedema expansion is associated with poor clinical outcomes in intracerebral haemorrhage. J Neurol Neurosurg Psychiatry. 2016;87:1169–1173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hemphill JC 3rd, Bonovich DC, Besmertis L, Manley GT, Johnston SC. The ICH score: a simple, reliable grading scale for intracerebral hemorrhage. Stroke. 2001;32:891–897. [DOI] [PubMed] [Google Scholar]
- 24.Morotti A, Boulouis G, Dowlatshahi D, Li Q, Shamy M, Salman RA, Rosand J, Cordonnier C, Goldstein JN, Charidimou A. Intracerebral haemorrhage expansion: definitions, predictors, and prevention. Lancet Neurol. 2023;22:159–171. [DOI] [PubMed] [Google Scholar]
- 25.Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med. 2000;342:836–843. [DOI] [PubMed] [Google Scholar]
- 26.Ridker PM, Hennekens CH, Roitman-Johnson B, Stampfer MJ, Allen J. Plasma concentration of soluble intercellular adhesion molecule 1 and risks of future myocardial infarction in apparently healthy men. Lancet. 1998;351:88–92. [DOI] [PubMed] [Google Scholar]
- 27.Luc G, Arveiler D, Evans A, Amouyel P, Ferrieres J, Bard JM, Elkhalil L, Fruchart JC, Ducimetiere P, Group PS. Circulating soluble adhesion molecules ICAM-1 and VCAM-1 and incident coronary heart disease: the PRIME Study. Atherosclerosis. 2003;170:169–176. [DOI] [PubMed] [Google Scholar]
- 28.Tanne D, Haim M, Boyko V, Goldbourt U, Reshef T, Matetzky S, Adler Y, Mekori YA, Behar S. Soluble intercellular adhesion molecule-1 and risk of future ischemic stroke: a nested case-control study from the Bezafibrate Infarction Prevention (BIP) study cohort. Stroke. 2002;33:2182–2186. [DOI] [PubMed] [Google Scholar]
- 29.Kraus J, Oschmann P, Leis S, Neundorfer B, Heckmann JG. High concentrations of sVCAM-1 and sICAM-1 in the cerebrospinal fluid of patients with intracerebral haemorrhage are associated with poor outcome. J Neurol Neurosurg Psychiatry. 2002;73:346–347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wang HC, Lin WC, Lin YJ, Rau CS, Lee TH, Chang WN, Tsai NW, Cheng BC, Kung CT, Lu CH. The association between serum adhesion molecules and outcome in acute spontaneous intracerebral hemorrhage. Crit Care. 2011;15:R284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Appelboom G, Bruce SS, Hickman ZL, Zacharia BE, Carpenter AM, Vaughan KA, Duren A, Hwang RY, Piazza M, Lee K, et al. Volume-dependent effect of perihaematomal oedema on outcome for spontaneous intracerebral haemorrhages. J Neurol Neurosurg Psychiatry. 2013;84:488–493. [DOI] [PubMed] [Google Scholar]
- 32.Roh D, Sun CH, Murthy S, Elkind MSV, Bruce SS, Melmed K, Ironside N, Boehme A, Doyle K, Woo D, et al. Hematoma Expansion Differences in Lobar and Deep Primary Intracerebral Hemorrhage. Neurocrit Care. 2019;31:40–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Wang XM, Zhang YG, Li AL, Long ZH, Wang D, Li XX, Xia JH, Luo SY, Shan YH. Expressions of serum inflammatory cytokines and their relationship with cerebral edema in patients with acute basal ganglia hemorrhage. European review for medical and pharmacological sciences. 2016;20:2868–2871. [PubMed] [Google Scholar]
- 34.Ziai WC, Parry-Jones AR, Thompson CB, Sansing LH, Mullen MT, Murthy SB, Mould A, Nekoovaght-Tak S, Hanley DF. Early Inflammatory Cytokine Expression in Cerebrospinal Fluid of Patients with Spontaneous Intraventricular Hemorrhage. Biomolecules. 2021;11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Li N, Liu YF, Ma L, Worthmann H, Wang YL, Wang YJ, Gao YP, Raab P, Dengler R, Weissenborn K, et al. Association of molecular markers with perihematomal edema and clinical outcome in intracerebral hemorrhage. Stroke. 2013;44:658–663. [DOI] [PubMed] [Google Scholar]
- 36.Silva Y, Leira R, Tejada J, Lainez JM, Castillo J, Davalos A, Stroke Project CDGotSNS. Molecular signatures of vascular injury are associated with early growth of intracerebral hemorrhage. Stroke. 2005;36:86–91. [DOI] [PubMed] [Google Scholar]
- 37.Ohashi SN, DeLong JH, Kozberg MG, Mazur-Hart DJ, van Veluw SJ, Alkayed NJ, Sansing LH. Role of Inflammatory Processes in Hemorrhagic Stroke. Stroke. 2023;54:605–619. [DOI] [PubMed] [Google Scholar]
- 38.Morotti A, Phuah CL, Anderson CD, Jessel MJ, Schwab K, Ayres AM, Pezzini A, Padovani A, Gurol ME, Viswanathan A, et al. Leukocyte Count and Intracerebral Hemorrhage Expansion. Stroke. 2016;47:1473–1478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Boehme AK, Comeau ME, Langefeld CD, Lord A, Moomaw CJ, Osborne J, James ML, Martini S, Testai FD, Woo D, et al. Systemic inflammatory response syndrome, infection, and outcome in intracerebral hemorrhage. Neurol Neuroimmunol Neuroinflamm. 2018;5:e428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Murthy SB, Moradiya Y, Shah J, Merkler AE, Mangat HS, Iadacola C, Hanley DF, Kamel H, Ziai WC. Nosocomial Infections and Outcomes after Intracerebral Hemorrhage: A Population-Based Study. Neurocrit Care. 2016;25:178–184. [DOI] [PubMed] [Google Scholar]
- 41.Murthy S, Roh DJ, Chatterjee A, McBee N, Parikh NS, Merkler AE, Navi BB, Falcone GJ, Sheth KN, Awad I, et al. Prior antiplatelet therapy and haematoma expansion after primary intracerebral haemorrhage: an individual patient-level analysis of CLEAR III, MISTIE III and VISTA-ICH. J Neurol Neurosurg Psychiatry. 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Mack WJ, Mocco J, Hoh DJ, Huang J, Choudhri TF, Kreiter KT, Lozier A, Opperman M, Poisik A, Yorgason J, et al. Outcome prediction with serum intercellular adhesion molecule-1 levels after aneurysmal subarachnoid hemorrhage. J Neurosurg. 2002;96:71–75. [DOI] [PubMed] [Google Scholar]
- 43.Mocco J, Mack WJ, Kim GH, Lozier AP, Laufer I, Kreiter KT, Sciacca RR, Solomon RA, Mayer SA, Connolly ES Jr. Rise in serum soluble intercellular adhesion molecule-1 levels with vasospasm following aneurysmal subarachnoid hemorrhage. J Neurosurg. 2002;97:537–541. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data from the FAST trial cannot be shared by the authors; however, anonymized data from the FAST trial are available by request through Novo Nordisk (https://www.novonordisk-trials.com).