Abstract
Intracerebral hemorrhage (ICH) caused by structural vascular lesions is associated with better outcomes than primary ICH, but this relationship is poorly understood. We tested the hypothesis that ICH from a vascular lesion has more benign hematoma characteristics compared to primary ICH. We performed a retrospective study using data from our medical center. The SMASH-U criteria were used to adjudicate the etiology of ICH. The co-primary outcomes were admission parenchymal hematoma volume and hematoma expansion at 24 hours. Linear and logistic regression analyses were performed to test associations. A total of 231 patients were included of whom 42 (18%) had a vascular lesion. Compared to primary ICH patients, those with structural vascular lesions were younger (49 vs. 68 years, p<0.001), less likely to have hypertension (29% vs. 74%, p<0.001), had lower mean admission systolic blood pressure (140 ± 23 vs. 164 ± 35, p<0.001), less frequently had IVH (26% vs. 44%, p=0.03), and had mostly lobar or infratentorial hemorrhages. The median admission hematoma volume was smaller with vascular lesions (5.9 vs. 9.7 mL, p=0.01). In regression models, ICH from a vascular lesion was associated with smaller admission hematoma volume (beta, −0.67, 95% CI, −1.29 to −0.05, p=0.03), but no association with hematoma expansion was detected when assessed as a continuous (OR, 0.93; 95% CI, −4.46 to 6.30, p= 0.73) or dichotomous exposure (OR, 1.86; 95% CI, 0.40 to 8.51, p=0.42). In a single-center cohort, patients with ICH from vascular lesions had smaller hematoma volumes than patients with primary ICH.
Keywords: Cerebral Hemorrhage, Intracranial Hemorrhage, Hypertension, Intracranial Arteriovenous Malformations
Introduction
Primary intracerebral hemorrhage (ICH) causes significant morbidity and mortality, with only about 50% surviving for 1 year, and only about half of survivors being functionally independent 1,2. Clinical and radiological factors associated with poor ICH outcomes include age, baseline hematoma volume, hematoma expansion, ICH location, presence of intraventricular hemorrhage, and antithrombotic therapy preceding ICH 3,4.
The majority of studies on ICH focuses on primary causes of ICH such as chronic hypertension or cerebral amyloid angiopathy, or ICH secondary to anticoagulation use. ICH due to structural vascular lesions are relatively less common and our current understanding of acute hematoma dynamics and management has been extrapolated from studies on primary ICH. However, emerging data suggest that ICH due to vascular lesions, particularly arteriovenous malformations, have lower mortality and favorable functional outcomes 5–7. The specific mechanism is poorly understood, and it is unclear whether hematoma characteristics in the acute period of ICH differ between a primary ICH and ICH from a structural vascular lesion. We sought to evaluate baseline hematoma volume and hematoma expansion in primary vs. vascular lesion-associated ICH. We hypothesized that favorable hematoma characteristics in vascular lesion-associated ICH may contribute to improved outcomes.
Methods
Patient Population and Measurements
The Institutional Review Board at Weill Cornell Medicine approved this study and waived the requirement for informed consent. Patients with a hemorrhagic stroke admitted to NewYork-Presbyterian Hospital/Weill Cornell Medical Center are prospectively enrolled into the American Heart Association’s Get With The Guidelines—Stroke registry. This registry provides the foundation for the Cornell AcutE Stroke Academic Registry (CAESAR), which is then supplemented through retrospective collection of additional clinical, laboratory, and radiographic data. For enrollment in the American Heart Association’s Get With the Guidelines—Stroke registry, trained hospital analysts prospectively collect data on demographics, comorbidities, and in-hospital treatments and outcomes. Additional data collected as part of the CAESAR registry included Glasgow Coma Scale on admission, hematoma location (deep, lobar, infratentorial, or purely intraventricular), and ICH etiology. A panel of neurologists adjudicated the cause of ICH using SMASH-U criteria (Structural vascular lesion, Medication-related, Amyloid angiopathy, Systemic Disease, Hypertension, Unknown) 8. Of all adult patients with ICH registered in the Cornell Acute Stroke Academic Registry from 2011 through 2017, we included in this analysis those with ICH resulting from hypertension or amyloid angiopathy (henceforth referred to as primary ICH), and those with ICH from a structural vascular lesion, the adjudication of which required a clear radiographic or operative evidence of an underlying vascular abnormality, such as an arteriovenous malformation, a cavernous malformation, a dural arteriovenous malformation, or an aneurysm. Patients with ICH due to other causes (e.g., anticoagulation) or unknown cause were excluded.
Outcomes
The co-primary outcomes were baseline hematoma volume and hematoma expansion at 24 hours. We measured hematoma volumes on the baseline scan and the scan closest in time to 24 hours after baseline. Hematoma volumes were calculated using semi-automated planimetry with 3D Slicer consistent with our prior studies 9,10. 3D Slicer has been shown to be more accurate than the traditional ABC/2 method for hematomas of varying shapes differing from the ellipsoid form 11. Digital Imaging and Communications in Medicine (DICOM) data were transferred to a research computer and assessed with 3D Slicer. Hematoma volume was automatically detected slice by slice using the level tracing tool of the segment editor module. Hematoma expansion was defined as a relative increase of >33% or an absolute increase of >6 mL in the size of the hematoma from baseline to follow-up images 12.
Statistical Analysis
Patient characteristics are presented using standard descriptive statistics. Discrete variables are presented as counts (percentages [%]) and continuous variables as means (standard deviation [SD]) or medians (interquartile range [IQR]), as appropriate. We used multiple linear regression to assess the relationship between ICH etiology and admission hematoma volume. We performed natural logarithmic transformation of the admission hematoma volume to satisfy model assumptions. Covariates for the model included age, sex, race, hematoma location, and time to baseline CT scan, since these have previously been shown to be independently associated with baseline hematoma volume in a large cohort of ICH patients 13. Collinear covariates, defined by a variance inflation factor > 4, were identified and removed from the model.
We then assessed whether ICH etiology predicted hematoma expansion, which was performed in two ways. First, we modeled hematoma expansion as a continuous outcome measure using linear regression as detailed above. Second, we modeled hematoma expansion as a dichotomous variable (relative increase of >33% or an absolute increase of >6mL) using multiple logistic regression. We a priori selected known predictors of hematoma expansion and clinical outcomes as covariates, which were age, sex, hematoma location, presence of intraventricular hemorrhage, and time from symptom onset to baseline CT scan 3,4. Statistical analyses were performed using Stata (version 14.0, College Station, TX). All analyses were two-tailed with the threshold of statistical significance allowing for an alpha error of 0.05.
Results
There were a total of 492 patients with ICH, of whom 261 were excluded for the following reasons: baseline magnetic resonance imaging instead of a CT or hematoma volumes too small to calculate (n=68), primary intraventricular hemorrhage (n=13), and ICH from other SMASH-U etiologies besides hypertension, amyloid angiopathy and structural vascular lesions (n=180). Among the 231 ICH patients included in the analysis, 189 (82%) had primary ICH, while 42 (18%) had a vascular lesion. Compared to primary ICH patients, those with structural vascular lesions were younger (49 vs. 68 years, p<0.001), less likely to have hypertension (29% vs. 74%, p<0.001), had lower mean admission systolic blood pressure (140 ±23 vs. 164 ±35, p<0.001; Figure 1), less frequently had IVH (26% vs. 44%, p=0.03; Figure 2), and had mostly lobar or infratentorial hemorrhages (Table 1, Figure 2).
Figure 1:
Comparing arrival NIHSS and systolic BP in patients with primary ICH and vascular lesion related ICH. Red lines represent median, red crosses represent outliers. Box a whisker plots with individual data points superimposed. BP, blood pressure; NIHSS, NIH stroke scale.
Figure 2:
Comparing initial hematoma characteristics between primary ICH and vascular lesion related ICH. Log-transformed plot of hematoma volume, demonstrating smaller volume hematomas in the vascular lesion group (see Table 1 for formal statistics). Vascular lesion ICH were less likely to have associated IVH, and more likely to be either lobar of infratentorial than primary ICH.
Table 1.
Characteristics of Patients Stratified by ICH Etiology.
| Characteristica | Vascular Lesion (N =42) |
Primary ICH (N =189) |
P value |
|---|---|---|---|
| Age, mean (SD), y | 49.4 (18.5) | 68.1 (14.1) | <0.001 |
| Female | 14 (33.3) | 83 (43.9) | 0.21 |
| Race | 0.73 | ||
| White | 34 (81.0) | 144 (76.2) | |
| Black | 6 (14.3) | 26 (13.8) | |
| Other | 2 (4.7) | 19 (10.0) | |
| Hypertension | 12 (28.6) | 139 (73.5) | <0.001 |
| Diabetes | 6 (14.3) | 37 (19.6) | 0.43 |
| Dyslipidemia | 11 (26.2) | 69 (36.5) | 0.20 |
| Atrial fibrillation | 1 (2.3) | 10 (5.3) | 0.42 |
| Admission Systolic BP, mm Hg | 140 (22) | 164 (35) | <0.001 |
| Admission Diastolic BP, mm Hg | 81 (13) | 90 (27) | 0.06 |
| Admission NIHSS | 3 (1–8.5) | 11 (3–21) | <0.001 |
| Admission GCS | 15 (13–15) | 14 (10–15) | 0.01 |
| Admission hematoma volume, mL | 5.9 (1.0–16.1) | 9.7 (2.5–31.3) | 0.01 |
| Intraventricular hemorrhage | 11 (26.2) | 84 (44.4) | 0.03 |
| Hematoma location | <0.001 | ||
| Lobar | 27 (64.3) | 49 (25.9) | |
| Deep | 3 (7.1) | 123 (65.1) | |
| Infratentorial | 12 (28.6) | 17 (9.0) | |
| Hematoma volume at 24 hours | 7.7 (1.7–17.9) | 9.6 (2.4–28.7) | 0.44 |
| Hematoma expansion | 3/19 (15.8) | 25/123 (20.3) | 0.64 |
| Time to baseline CT scan, hours | 2.3 (1.8–5.4) | 2.8 (2.0–6.1) | 0.61 |
| Time to 24-hour CT scan, hours | 26.0 (12–42) | 22 (13–27) | 0.14 |
Abbreviations: BP, blood pressure; CT, computed tomography; GCS, Glasgow Coma Scale; ICH, intracerebral hemorrhage; NIHSS, NIH Stroke Scale; SD, standard deviation
Data are presented as number (%) unless otherwise specified.
The median admission hematoma volume was lower in vascular lesions (5.9 [1.0–16.1] vs. 9.7 [2.5–31.3] mL, p=0.01, Figure 2). In multiple linear regression models, ICH from a vascular lesion was significantly associated with smaller admission hematoma volume (beta, −0.67; 95% CI, −1.29 to −0.05; p=0.03) after adjustment of confounders (Table 2). For the analysis of hematoma expansion, we had 142 ICH patients since 89 had missing 24-hour CT scans, and as a result, our adjusted linear regression (OR, 0.93; 95% CI, −4.46 to 6.30, p= 0.73; Table 3) and logistic regression models (OR, 1.86; 95% CI, 0.40–8.51, p=0.42; Table 4) for the association between ICH from a vascular lesion and hematoma expansion were inconclusive.
Table 2.
Association between ICH Etiology and Log-Transformed Admission Hematoma Volume*.
| beta (95% CI) | p value | |
|---|---|---|
| Vascular lesion (vs. primary ICH) | −0.67 (−1.29 to −0.05) | 0.03 |
| Age, per year | −0.01 (−0.02 to 0.09) | 0.38 |
| Female sex | −0.51 (−0.94 to −0.08) | 0.02 |
| White race | −0.18 (−0.69 to 0.34) | 0.50 |
| Intraventricular hemorrhage | 1.76 (1.33 to 2.21) | <0.001 |
| Time to baseline CT scan | 0.98 (0.23 to 6.24) | 0.56 |
Abbreviations: CI, confidence interval; CT, computed tomography; ICH, intracerebral hemorrhage.
Linear regression was used.
Table 3.
Association between ICH Etiology and Hematoma Expansion (Continuous Measure)*
| beta (95% CI) | p value | |
|---|---|---|
| Vascular lesion (vs. primary ICH) | 0.93 (−4.46–6.30) | 0.73 |
| Age, per year | 0.02 (−0.89 – 0.14) | 0.44 |
| Female sex | −1.47 (−4.48–1.90) | 0.39 |
| White race | 0.25 (−3.62–4.13) | 0.89 |
| Intraventricular hemorrhage | 0.14 (−4.42–3.71) | 0.94 |
| Time to baseline CT scan | −0.02 (−0.01–0.02) | 0.47 |
Abbreviations: CI, confidence interval; CT, computed tomography; ICH, intracerebral hemorrhage.
Linear regression was used.
Table 4.
Association between ICH Etiology and Hematoma Expansion*.
| beta (95% CI) | p value | |
|---|---|---|
| Vascular lesion (vs. primary ICH) | 1.86 (0.40–8.51) | 0.42 |
| Age, per year | 1.03 (0.97–1.00) | 0.04 |
| Female sex | 1.05 (0.42–2.54) | 0.94 |
| White race | 0.69 (0.24–1.89) | 0.47 |
| Intraventricular hemorrhage | 1.67 (0.68–4.07) | 0.26 |
| Time to baseline CT scan | 0.05 (0.04–0.67) | 0.18 |
Abbreviations: CI, confidence interval; CT, computed tomography; ICH, intracerebral hemorrhage.
Logistic regression was used.
The analytical cohort excluded anticoagulation-associated ICH since only ICH from primary and structural lesions were included (SMASH-U criteria)
Discussion
In this retrospective observational study, secondary ICH due to a structural vascular lesion was associated with lower admission hematoma volumes compared to patients with a primary ICH. We found no significant association with hematoma expansion, though our results were not conclusive due to smaller numbers with available follow-up data.
Several prior studies have shown lower mortality and better functional outcomes in ICH related to vascular lesions compared to primary ICH 5–7,14,15. While acute hematoma characteristic influence ICH outcomes 3,4, whether they differ between primary and secondary ICH remains poorly understood. It is therefore unclear if extrapolating management strategies effective in primary ICH to those with vascular lesions is the correct strategy. A prior retrospective study reported that ICH due to vascular malformations resulted in smaller hematoma volumes compared to ICH from all other etiologies combined, but this included anticoagulant-related hemorrhages which have a strong propensity to expand 16. In this context, our study provides important findings about the relationship between ICH from vascular lesions and small baseline ICH volume. There was no significant association between ICH etiology and hematoma expansion, which may be attributable to the small sample size since fewer patients in our study had comparable 24-hour neuroimaging. An alternate explanation may be that ICH from structural vascular lesions, in comparison with primary ICH patients 10, tend to have lower mean systolic blood pressures in the first 72 hours after admission which is arguably the most significant factor linked to hematoma expansion 17,18. Other reasons for better outcomes in ICH from vascular lesions have been speculated to include (1) younger age and fewer comorbid medical conditions, (2) limited tissue injury in vascular lesions as most of the hematoma is within the lesion nidus which might already be less eloquent brain, and (3) bleeding occurring from lower pressure system such as venous structures, which might limit the initial hematoma size or subsequent expansion 19,20.
Limitations and Generalizability
This study has several limitations. First, this cohort was from a single tertiary center resulting in selection bias which hampers the generalizability of our findings to the general ICH population. Second, the small sample size of the study precluded a comprehensive multivariate analysis, and may partly explain the lack of association between ICH from a vascular lesion and hematoma expansion. Given the availability of resources at our center, many of these patients underwent follow up imaging with MRI as opposed to repeat CT imaging. We felt that inclusion of these patients would confound estimates of hematoma expansion given the enhanced spatial resolution afforded on MRI, which further limited our sample size. The CAESAR database also lacks long-term outcomes, or detailed qualitative neuroradiologic assessments of hematoma descriptions that might also differ between ICH subtypes. Finally, our study used the SMASH-U criteria 8 to define hematoma etiology. As such, vascular lesion cases included all arteriovenous malformations, cavernous malformations, dural arteriovenous malformations, and aneurysms that resulted in ICH. These lesions are morphologically heterogeneous with varying hemodynamic and anatomic factors that may affect the risk of bleeding and clinical outcomes. We lacked power to study individual subtypes of vascular lesions.
Conclusions
In this single center retrospective observational study, we found that patients with ICH due a vascular lesion had smaller hematomas than patients with primary ICH. Smaller baseline hematoma volumes may partly explain the favorable outcomes in this patient cohort.
Highlights.
Secondary ICH from structural vascular lesions is more often involves the lobar and infratentorial locations, compared to primary ICH.
Intraventricular hemorrhage is less common with ICH from structural vascular lesions.
Secondary ICH due to a structural vascular lesion is associated with lower admission hematoma volumes compared to patients with a primary ICH.
Funding/Support:
The study was funded by the National Institutes of Health (NIH) (K23NS105948; PI: Dr. Murthy). The funding entities had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the article; and decision to submit the article for publication.
Disclosures:
Dr. Merkler is supported by the American Heart Association (18CDA34110419), the Leon Levy Foundation, and reports personal fees for medicolegal consulting on stroke and neurological disorders. Dr. Gupta is supported by the NIH for research outside of the work related to this manuscript. Dr. Kamel reports serving as co-PI for the NIH-funded ARCADIA trial (NINDS U01NS095869) which receives in-kind study drug from the BMS-Pfizer Alliance for Eliquis® and ancillary study support from Roche Diagnostics, serving as Deputy Editor for JAMA Neurology, serving as a steering committee member of Medtronic’s Stroke AF trial (uncompensated), serving on an endpoint adjudication committee for a trial of empagliflozin for Boehringer-Ingelheim, and having served on an advisory board for Roivant Sciences related to Factor XI inhibition. Dr. Murthy reports personal fees for medicolegal consulting on stroke and neurological disorders.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.van Asch CJ, Luitse MJ, Rinkel GJ, van der Tweel I, Algra A, Klijn CJ. Incidence, case fatality, and functional outcome of intracerebral haemorrhage over time, according to age, sex, and ethnic origin: a systematic review and meta-analysis. Lancet Neurol 2010;9(2):167–76. [DOI] [PubMed] [Google Scholar]
- 2.Poon MTC, Fonville AF, Al-Shahi Salman R. Long-term prognosis after intracerebral haemorrhage: systematic review and meta-analysis. J Neurol Neurosurg Psychiatry 2014;85(6):660–7. [DOI] [PubMed] [Google Scholar]
- 3.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(4):891–7. [DOI] [PubMed] [Google Scholar]
- 4.Brouwers HB, Chang Y, Falcone GJ, et al. Predicting Hematoma Expansion After Primary Intracerebral Hemorrhage. JAMA Neurol 2014;71(2):158–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rosenow F, Hojer C, Meyer-Lohmann C, et al. Spontaneous intracerebral hemorrhage. Prognostic factors in 896 cases. Acta Neurol Scand 1997;96(3):174–82. [PubMed] [Google Scholar]
- 6.van Beijnum J, Lovelock CE, Cordonnier C, et al. Outcome after spontaneous and arteriovenous malformation-related intracerebral haemorrhage: population-based studies. Brain 2009;132(Pt 2):537–43. [DOI] [PubMed] [Google Scholar]
- 7.Murthy SB, Merkler AE, Omran SS, et al. Outcomes after intracerebral hemorrhage from arteriovenous malformations. Neurology 2017;88(20):1882–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Meretoja A, Strbian D, Putaala J, et al. SMASH-U: a proposal for etiologic classification of intracerebral hemorrhage. Stroke 2012;43(10):2592–7. [DOI] [PubMed] [Google Scholar]
- 9.Gusdon AM, Gialdini G, Kone G, et al. Neutrophil-Lymphocyte Ratio and Perihematomal Edema Growth in Intracerebral Hemorrhage. Stroke 2017;48(9):2589–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lin J, Piran P, Lerario MP, et al. Differences in Admission Blood Pressure Among Causes of Intracerebral Hemorrhage. Stroke 2020;51(2):644–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Xu X, Chen X, Zhang J, et al. Comparison of the Tada formula with software slicer: precise and low-cost method for volume assessment of intracerebral hematoma. Stroke 2014;45(11):3433–5. [DOI] [PubMed] [Google Scholar]
- 12.Dowlatshahi D, Demchuk AM, Flaherty ML, Ali M, Lyden PL, Smith EE. Defining hematoma expansion in intracerebral hemorrhage. Neurology 2011;76(14):1238–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Falcone GJ, Biffi A, Brouwers HB, et al. Predictors of hematoma volume in deep and lobar supratentorial intracerebral hemorrhage. JAMA Neurol 2013;70(8):988–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Choi JH, Mast H, Sciacca RR, et al. Clinical outcome after first and recurrent hemorrhage in patients with untreated brain arteriovenous malformation. Stroke 2006;37(5):1243–7. [DOI] [PubMed] [Google Scholar]
- 15.Hartmann A, Mast H, Mohr JP, et al. Morbidity of intracranial hemorrhage in patients with cerebral arteriovenous malformation. Stroke 1998;29(5):931–4. [DOI] [PubMed] [Google Scholar]
- 16.Dinc N, Won S-Y, Brawanski N, et al. Differences in bleeding patterns and outcome after intracerebral hemorrhage due to vascular malformations. PLoS One 2019;14(5):e0217017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Leasure AC, Qureshi AI, Murthy SB, et al. Association of Intensive Blood Pressure Reduction With Risk of Hematoma Expansion in Patients With Deep Intracerebral Hemorrhage. JAMA Neurol 2019;76(8):949–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Li Q, Warren AD, Qureshi AI, et al. Ultra-Early Blood Pressure Reduction Attenuates Hematoma Growth and Improves Outcome in Intracerebral Hemorrhage. Ann Neurol 2020;88(2):388–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Spetzler RF, Hargraves RW, McCormick PW, Zabramski JM, Flom RA, Zimmerman RS. Relationship of perfusion pressure and size to risk of hemorrhage from arteriovenous malformations. J Neurosurg 1992;76(6):918–23. [DOI] [PubMed] [Google Scholar]
- 20.Kader A, Young WL, Pile-Spellman J, et al. The influence of hemodynamic and anatomic factors on hemorrhage from cerebral arteriovenous malformations. Neurosurgery 1994;34(5):801–7; discussion 807–8. [DOI] [PubMed] [Google Scholar]