Clinical characteristics and outcomes of methamphetamine-associated versus non-methamphetamine intracerebral hemorrhage

Methamphetamine use has emerged as a risk factor for intracerebral hemorrhage (ICH). We aim to investigate the clinical characteristics and outcomes of methamphetamine-associated ICH (Meth-ICH) versus Non-Meth-ICH. Patients with ICH between January 2011 and December 2017 were studied. Meth-ICH and Non-Meth-ICH were defined by history of abuse and urine drug screen (UDS). The clinical features of the 2 groups were explored. Among the 677 consecutive patients, 61 (9.0%) were identified as Meth-ICH and 350 as Non-Meth ICH. Meth-ICH was more common in Hispanics (14.6%) and Whites (10.1%) as compared to Asians (1.2%). Patients with Meth-ICH were more often younger (51.2 vs. 62.2 years, p < 0.001), male (77.0% vs. 61.4.0%, p < 0.05), and smokers (44.3% vs. 13.4%, p < 0.001). Non-Meth-ICH was more likely to have history of hypertension (72.61% v. 59%, p < 0.05) or antithrombotic use (10.9% vs. 1.6%, p < 0.05). There was no significant difference in clinical severity, hospital length of stay (LOS), rate of functional independence (29.5% vs. 25.7%, p = 0.534), or mortality (18.0% vs. 24.6%, p = 0.267) between the 2 groups. Methamphetamine use was not an independent predictor of poor outcome. Despite difference in demographics, Meth-ICH is similar to Non-Meth ICH in hospital course and outcome.

Spontaneous intracerebral hemorrhage (ICH) is a devastating type of strokes 1,2 . The reported median 30-day mortality ranges from 17% to 40%, with rates of functional independency between 12% and 39% 1 . Recent randomized controlled studies have shown no significant benefit from intensive blood pressure control and early surgical intervention [3][4][5] . Since spontaneous ICH is potentially preventable, it is therefore a research priority to investigate any emerging risk of the disease.
Methamphetamine use is a serious public health crisis, with estimated 37 million active users and 2.6 million disability-adjusted life years lost in 2010 6,7 . Caplan at al. initially described incidence of intracerebral and subarachnoid hemorrhage from methamphetamine abuse 8,9 . Ischemic stroke was subsequently reported to be associated with methamphetamine use as well [10][11][12][13][14] .
Both population-based study and forensic analysis of fatal strokes showed significant predominance of hemorrhagic strokes in young methamphetamine users 15,16 . Methamphetamine use was associated with approximately fivefold increase in hemorrhagic stroke, twice higher than the risk from either cocaine or tobacco 15 . Lappin et al. performed a systemic review of methamphetamine-related stroke in 2017 and identified 81 hemorrhagic and 17 ischemic stroke cases from case reports and single center series 17 . Both types of stroke were approximately twice as common in males.
There appears to be an alarming trend of increasing prevalence of methamphetamine-associated ICH (Meth-ICH) in the regions around the Pacific rim 17  Statistical analysis. Continuous variables were described by mean ± standard deviation (SD) or median with interquartile range (IQR) based on the results of normality testing. Categorical variables were expressed by counts with percentages. Baseline characteristics and outcomes at discharge were compared between Meth-ICH and Non-Meth-ICH groups by Wilcoxon rank-sum test for continuous variables and χ2 test for categorical variables. Univariate analyses were performed initially to assess the possible factors associated with outcomes (functional independence and mortality at discharge). The cutoff value in univariate analysis for inclusion in the multivariate logistic regression was p < 0.1. Analyses were performed using SPSS software (version 23.0). A 2-tailed value of p < 0.05 was considered statistically significant.

Demographics of patients with Meth-ICH and Non-Meth-ICH.
A total of 677 consecutive patients with spontaneous ICH between January 1, 2011 and December 31, 2017 were included. The flow chart for identification of Meth-ICH and Non-Meth-ICH is shown in Fig. 1. Thirty-two patients had recorded history of methamphetamine use. Among them, 30 had a UDS and 21 (70%) tested positive, indicative of methamphetamine use within the last 3 days 21 . Of the 645 patients who denied history of methamphetamine use, 379 (58.8%) had a UDS and 29 of them (7.7%) tested positive for amphetamine. Therefore, a total of 61 patients in the entire cohort (9%) were identified to have Meth-ICH (highlighted in pink color in Fig. 1). Given the fact that 7.7% of the patients who denied history of methamphetamine use had a positive UDS, the rate of Meth-ICH could be higher if all patients had a UDS.
As highlighted in blue color in Fig. 1, the 350 patients who denied a history of methamphetamine use and also had a negative UDS were classified as Non-Meth-ICH.

Scientific RepoRtS |
(2020) 10:6375 | https://doi.org/10.1038/s41598-020-63480-z www.nature.com/scientificreports www.nature.com/scientificreports/ Although Meth-ICH was more often treated with surgical interventions, likely due to younger ages, there was no significant difference in requirement for ventilator support and LOS in the ICU or hospital between the 2 groups.
There was also no significant difference in either functional independence (OR = 1.21, 95% CI 0.66-2.20; p = 0.534) or mortality rate at hospital discharge (OR = 0.68, 95% CI 0.34-1.36; p = 0.267). Methamphetamine use per se was not a predictor of poor outcomes in patients with ICH.

Predictors of outcomes.
Factors associated with functional independency and mortality at hospital discharge were analyzed using univariate and multivariate models. Overall, lower NIHSS and higher GCS score were associated with functional independence at hospital discharge per multivariate analysis ( Table 2). Higher NIHSS, ICH score, ventilator support, and absence of surgical interventions were independent predictors of mortality (Table 3).

Discussion
This study represents the largest single center cohort of Meth-ICH yet reported. It demonstrates that methamphetamine is an important risk factor for ICH in young males, smokers, Hispanic and White populations in Southern California. The rate of Meth-ICH was 9% in our large cohort, as compared to 13% and 16.4% reported previously 18,19 . Of note, of the 645 patients who denied history of methamphetamine use, only 379 patients (58.8%) had a UDS and 29 (7.7%) were urine positive for amphetamine. The actual rate of Meth-ICH could be significantly higher if all ICH patients had a UDS.
The observed age disparities in patients with Meth-ICH versus Non-Meth-ICH were consistent with previous reports 18,19 .
Although ordering UDS at admission was routine practice at our Stroke Center, the decision to obtain the test was at the discretion of the ED and on-call stroke team. Previous report showed ethnic disparities in ordering drug screens in patient with ICH, with young African Americans having more UDS 25 . We examined the ordering   26 . Of the 596 ICH patients, 357 (60%) had a UDS. A UDS was more likely to be obtained in patients who were younger (age < 45), male, smokers, self-reported methamphetamine users, with more severe deficit at presentation (NIHSS > 4), without diabetes, or not taking anticoagulant. There was no significant difference in UDS among different races (p = 0.319) 26 . Therefore, the higher rates of Meth-ICH in Hispanics and Whites were not due to bias in ordering UDS. In addition, hypertensive ICH is very prevalent in Asians 2 . However, only 1.2% (2/166) Asians in this cohort had Meth-ICH. These findings may have significant public health implication and may guide targeted community education for the prevention of Meth-ICH.
The majority of Meth-ICH was located in the basal ganglia/thalamus, suggesting deep white matter small-vessel injuries from direct toxicity of methamphetamine or increased sympathetic system activation 15,17,27 . The proportion of lobar hemorrhage was significantly higher in Non-Meth-ICH (38.0% vs 16.4%), likely due to older age and history of hypertension, ischemic stroke, or cerebral amyloid angiopathy [28][29][30] .
In our current study, there was no significant difference in ICH severity, ventilator support, LOS in the ICU or hospital, rate of functional independence or mortality at hospital discharge between Meth-ICH and  www.nature.com/scientificreports www.nature.com/scientificreports/ Non-Meth-ICH. The severity of neurologic deficits at admission was the main predictor of functional independence and mortality. Methamphetamine use and demographic features including age were not independently associated with outcomes as they had been in previous studies 17,29,30 .
Of note, our findings are significantly different from what was previously reported. Swor et al. showed that patients with Meth-ICH had significantly longer LOS in the ICU (10 ± 8 vs 5 ± 5 days) and hospital (18 ± 27 vs 8 ± 8 days) than Non-Meth-ICH 19 . They also showed higher mortality rates (29% vs 34%), as compared to our study (18% vs 25%). Several factors may account for lack of difference in LOS between the two groups and lower mortality rates in our study. First, although the Meth-ICH patients had higher NIHSS score, they were younger with less comorbidities as compared with Non-Meth-ICH patients. Second, there was no significant difference in highest blood pressures, GCS, ICH score and ventilator support between Meth-ICH and Non-Meth-ICH groups. Third, all ICH patients were initially managed in the dedicated Neuroscience ICU with standard ICH order-set and clinical pathway by board-certified neurointensivits at our center. Management of ICH by neurocritical care specialists have been shown to be associated with shorter LOS and reduced mortality [31][32][33] .
The mechanisms by which methamphetamine may cause ICH remains unclear. In 1970, Citron et al. studied 14 patients with polysubstance abuse, with almost all admitting methamphetamine use, and found necrotizing angiitis in medium-sized and small arteries of the brain and other organs 34 . Rumbaugh et al. analyzed the angiographic features of patients with methamphetamine abuse and described beaded arteries, segmental changes in vessel caliber, and regions of slow flow 35 . In monkeys receiving intravenous methamphetamine for two weeks, angiography showed similar beading and segmental changes, and necropsy revealed small cerebral hemorrhages, zones of infarction, and microaneurysms 36 . Davis et al. reported 4.9% incidence of methamphetamine intoxication in all autopsies in San Diego County from 1 January 1987 to 31 December 1993 and suggested possible propagation and rupture of berry aneurysms from methamphetamine use 37 . Of note, methamphetamine is a potent sympathomimetic. ICH may also occur secondary to methamphetamine-induced hypertensive surge in the absence of pre-existing cerebrovascular disease 38 . Chronic use of methamphetamine may cause long-term systemic hypertension and vessel damage 13,17,27,39 . Further studies are warranted to investigate the mechanisms of Meth-ICH.
To the best of our knowledge, this is the largest single-center cohort to compare the clinical characteristics and outcomes of patients with Meth-ICH versus Non-Meth-ICH. Our study demonstrates that there is no significant difference in severity of ICH, LOS, favorable outcome and mortality between Meth-ICH and Non-Meth-ICH. In addition, we found considerable proportion of patients who denied history of methamphetamine use turned out to have positive urine drug test, indicating possible higher prevalence of methamphetamine use in patients with ICH. The strengths of this study are the comprehensive comparison of the clinical features and outcomes of Meth-ICH verse non-Meth-ICH. Our findings provide better understanding of Meth-ICH vs Non-Meth-ICH and may help develop strategies for the effective treatment and prevention of Meth-ICH.
Our study has a few limitations. First, this is a single-center study performed in Southern California. Our findings may not be generalizable to other regions or populations. Second, despite denying drug use history and negative urine drug test, it is still possible that there were undetected meth users in the Non-Meth-ICH group.  Table 3. Factors associated with hospital mortality. Abbreviations: SBP, systolic blood pressure; DBP, diastolic blood pressure. NIHSS, National Institutes of Health Stroke Scale; GCS, Glasgow Coma Scale. ICH, intracerebral hemorrhage.