Higher Rate of Intracerebral Hemorrhage in Hispanic Patients with Cerebral Cavernous Malformation

Abstract

Cerebral cavernous malformations (CCM) are vascular malformations prone to intracerebral hemorrhage and epilepsy. Studies about the natural history and clinical presentation in the Hispanic population are lacking [7]. Retrospectively, we identified demographics and clinical features of Hispanic patients with CCM in our neurology clinic. Comparison with studies in the non-Hispanic White population with CCM was conducted.

Results

A total of 15 Hispanic patients diagnosed with CCM were identified in our neurology clinic. The majority of our patients presented with symptomatic hemorrhage (27%, n = 4) or seizure(s) (47%, n = 7). Of 15 patients, there were a total of four patients (27%) that had recurrent hemorrhages with an annual rate of recurrent hemorrhage of 8.04%. Our cohort had a higher rate of intracerebral hemorrhage after any initial mode of presentation (47%, n = 7) compared with non-Hispanic White population studies: 11% (n = 32) and 4% (n = 5). (p < 0.05)

Conclusions

Hispanic patients with CCM have a higher rate of hemorrhage during follow-up after any presentation of CCM (p<0.05) when compared to CCM in the non-Hispanic White population.

Introduction

Cerebral cavernous malformations (CCMs) are blood vessels devoid of elastic and muscular tissue. They are lined with endothelial cells that do not have intervening tight junctions, and therefore, CCMs are prone to hemorrhage [2]. CCMs appear to be the most frequently diagnosed entity among cerebrovascular malformations and are very common in Hispanics [1]. Intracerebral hemorrhage (ICH) is the most serious complication and is suspected to be more common in the Hispanic population [4]. rior ICH and multiple lesions, meaning multiple CCMs located in various locations in the central nervous system (CNS), seem to be the most important risk factors for presenting with and having recurrent intracerebral hemorrhage (ICH) [3]. The epidemiology and clinical characteristics of CCM among the Hispanic population remains largely unexplored.

Study Design and Methods

We retrospectively reviewed the Texas Tech University Health Sciences Center of El Paso (TTUHSC) Neurology clinic. We first identified our patients by searching in the Neurology clinic database for patients with the ICD-9 diagnosis code of 747.81. This is a general category for cerebrovascular anomalies that include CCM, arteriovenous malformation (AVM), arteriovenous fistula (AVF), and developmental venous anomaly (DVA). Each chart with the ICD-9 code of 747.81 was reviewed by the first author (AJ). Neuroimaging was reviewed for all 37 charts by a neurologist (AM) to specifically identify patients with CCM as opposed to AVM, AVF, or DVAs. We defined CCM according to the Standard Definition and Reporting Standards reported previously [5,6]for research concerning CCM; therefore, all hemorrhages recorded were located within the vicinity of the CCM and all were noted to therefore be symptomatic according to the Standard Definition and Reporting Standards. Neuroimaging was also reviewed for concomitant DVA but this was not part of the primary outcome. The inclusion criteria were as follows: 18 years of age or older presently, Hispanic ethnicity [7] and confirmed to have a diagnosis of CCM.

Demographics and clinical characteristics were abstracted from each medical record. Age, gender, ethnicity, family history, past medical history and comorbid conditions such as hypertension, diabetes, and cigarette smoking, mode of presentation, location, size and multiplicity of the lesions. The study was approved by the local institutional review board.

The primary objective of this study was to investigate the most frequent mode of presentation of CCM among Hispanics, including the rate of recurrent hemorrhage and compare our results with the non-Hispanic White population with CCM.

Statistical Analysis

Continuous data were summarized using mean and standard deviation (SD), while categorical data were summarized using percentages. In order to assess the association between the studies, a contingency table was generated in which Fisher’s exact test was used to determine the level of association. We computed the rate for hemorrhage after an initial presentation of ICH and the rate for hemorrhage regardless of the patient’s mode of presentation. In order to compare the rates between the studies, two sample tests of proportion were used to compare each outside study with our own. Significance was determined when p values were found to be less than 5%. All statistical analysis was performed using STATA 13.1. (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP)

Results

From a total of 1,928 patients seen in the neurology clinic since January 1, 2008 to May 31, 2012, 37 patients were identified using the ICD-9 code of 747.81. Sixteen were found to have CCM (as opposed to the other diagnosis that fit under the ICD-9 code including AVM, AVF, and DVA). One was excluded due to non-Hispanic ethnicity, and therefore, a total of 15 patients met our inclusion criteria. The mean age at diagnosis was 42.35 ± 15 years of age, and 67% were females (n = 10). Half of our population (53.3%, n = 8) had hypertension, 7% (n = 1) had type two diabetes mellitus, and 7% (n = 1) was a current cigarette smoker. The most common modes of presentations were as follows: seizure 47% (n = 7, p = 0.002), intracerebral hemorrhage 27% (n = 4), progressive focal neurological deficits 13% (n = 2), headache 7% (n = 1), and incidental finding in 7% (n = 1). The median length of follow-up was 6.25 years. Multiplicity of lesions was seen in 73% (n = 11). Two patients had spinal cord involvement and one had an intraventricular lesion.

Recurrent hemorrhage occurred in one of the patients that initially presented with a symptomatic hemorrhage (25%) one month later. Out of 11 patients who did not initially present with hemorrhage, six patients (55%) had hemorrhages with three patients (27%) having more than one recurrent hemorrhage (8 months later, 6 years later, and 23 years later) during the median follow up period. Therefore, of 15 patients, there were a total of four patients (27%) who had recurrent hemorrhages with an annual rate of recurrent hemorrhage of 8.04%.

We compared our results with two large studies of CCM in the non-Hispanic White population in North America and Europe, referred as the “Mayo Clinic (retrospective) study” [3] and the “Scottish (prospective) study” respectively [2]. (Table 2) The Mayo Clinic [3] study had a median length of follow-up of 7.3 years and the Scottish study [2] had a median length of follow-up of 5 years. We found a similar frequency in the initial mode of presentation of CCM in the Hispanic and the non-Hispanic White population. However, our cohort had a higher rate of intracerebral hemorrhage after any initial mode of presentation (47%, n = 7) compared with the Mayo Clinic and Scottish studies: 11% (n = 32) and 4% (n = 5), respectively (p<0.05). Two-thirds of our Hispanic CCM population presented with multiple CCM lesions in the CNS as compared to only 18% and 17% in the Mayo Clinic and Scottish studies, respectively (p<0.001).

Table 2. Comparison of Variables Between Our Study and the Mayo Clinic and Scottish Studies.

		Our Study	Mayo Clinic Study	Scottish Study
		N = 15	Compared to Our Study	Compared to Our Study
			N = 292	N = 134
	ICH	27% (n = 4)	25% (n = 74)	13% (n = 17)
	Seizure	47% (n = 7)	31% (n = 91)	26% (n = 35)
Mode of presentation	FND	13% (n = 2)	6% (n = 17)	15% (n = 21)
	Incidental	7% (n = 1)	38% (n = 110)	46% (n = 61)
	Headache	7% (n = 1)	N/A	N/A
Hemorrhage after initial presentation of seizure, FND, or incidental		55% (n = 6)	6% (n = 13)*	1.7% (n = 2)*
Hemorrhage after initial presentation of ICH		25% (n = 1)	26% (n = 19)	18% (n = 3)
Hemorrhage after any mode of presentation		47% (n=7)	11% (n = 32)*	4% (n = 5)*
Total recurrent hemorrhage		27% (n = 4)	10% (n = 28)	3% (n = 4)*
Multiplicity of lesions		74% (n = 11)	18% (n = 55)**	17% (n = 23)**

Notes: ICH = Intracerebral Hemorrhage, FND=Focal Neurological Deficit, CCM=Cerebral Cavernous Malformation,

^*p<0.05,

^**p<0.001

Discussion

CCM is the most common vascular malformation of the central nervous system that can cause neurologic disability due to headaches, motor weakness, cognitive decline, and symptomatic seizures. It is well known that CCM is frequent in people of Hispanic origin. The most feared complication of CCM is intracerebral hemorrhage. Compared with studies in non-Hispanic White populations, we found a similar mode of presentation but a significantly higher rate of hemorrhage after any initial presentation [2,3]. The cause of the higher rate of recurrent hemorrhage in our study is not clear, possibly due to the nature of the sample and the retrospective type of analysis. We analyzed the influence of cardiovascular risk factors and multiplicity of lesions in the rate of hemorrhage recurrence.

There was 53.3% (n = 8) of our Hispanic patients with hypertension compared to 17.8% (n = 52) in the Mayo Clinic study [3]. Diabetes and smoking status were not reported in the other studies. Risk factors for stroke, namely hypertension and diabetes, have a significant impact on the frequency and severity of cerebrovascular disease and among the population of the Hispanic population [12,15]. Furthermore; there are possible socioeconomic factors in the US-Mexico border region that potentially could raise the rate of hemorrhage in our population. Two studies found barriers to healthcare in Mexican Americans [13,14] that could lead to (undiagnosed and uncontrolled vascular risk factors) to an increased burden of cerebrovascular disease in this particular population.

Another reason for higher rates of hemorrhage could be the multiplicity of lesions, meaning more than one CCM found. In our population, there were almost three quarters of our patients with multiple CCMs, which greatly surpasses the multiplicity of lesions found in the Mayo Clinic [3] and the Scottish studies, [2] 18% and 17%, respectively (p<0.001). The mere fact of having more lesions could undoubtedly increase the chances of having more hemorrhages.

Some CCMs are known to have a genetic component and those could behave differently as compared to the nonfamilial cases [10]. There is an autosomal dominant CCM1 gene, also known as the “Common Hispanic Mutation” [2,11,15]. In our Hispanic sample of patients with CCM, about a quarter (26.7%, n = 4) had documented a positive family history and carried the CCM1 gene mutation. Familial forms of the disease are well known for having a more aggressive course in part related to the multiplicity of lesions throughout the CNS as well as extra CNS manifestations, including the retina and the skin. It is possible that the frequency of genetic mutation in the Hispanic population might be higher compared to the non-Hispanic White population.

Several limitations can be identified in our study. First, it was a retrospective design with a small sample size. Unfortunately, CCM does not have a specific ICD-9 code; the ICD-9 code is for “cerebrovascular anomalies” that includes AVM, AVF, CCM, and DVAs. It is possible that our selection of patients could underestimate the total number of patients with CCMs in our population. Some of the patients could have been coded as either intracerebral hemorrhages or seizures. Another important limitation is the nature of the sample. We limited our analysis to a specific CCM population that is assisted in a specialized care (Neurology clinic in an academic center). It is possible that the spectrum of Hispanic population with CCM represented in our study reflect only the more severe cases. Our results might not be including the Hispanic CCM populations that remain asymptomatic or minimally symptomatic. However, our population is comparable to the two main published studies (one registry and the other retrospective) that address CCM in the non-Hispanic White population.

Conclusion

Hispanic individuals with cerebral cavernous malformations might have a higher frequency of intracerebral hemorrhage due to multiple, unproven yet, reasons. A larger study is needed to confirm our findings.

Table 1. Mode of Presentation by Case.

Case #	Mode of presentation	Gender	Age at Dx	Location	Risk factors	Any hemorrhage	Recurrent (>1) hemorrhage	F/U period (yrs)
1	Seizures	M	55	Multiple bilateral supratentorial	_	Y	N	7.6
2	Seizures	F	28	Single Supratentorial	_	N	N	0.8
3	Seizures	F	43	Multiple bilateral supratentorial and infratentorial	HTN	Y	Y	16.3
4	Seizures	F	26	Multiple bilateral supratentorial, infratentorial, and interventricular	_	Y	Y	24.3
5	Seizures	M	51	Multiple bilateral Supratentorial and Infratentorial	_	N	N	0.5
6	Seizures	M	42	Multiple Bilateral supratentorial, infratentorial, and spinal cord involvement	HTN	Y	N	10.1
7	Seizures	F	35	Multiple bilateral supratentorial and infratentorial	_	Y	Y	7.4
8	Intracerebral hemorrhage	F	58	Multiple bilateral supratentorial and infratentorial	_	Y	N	1.8
9	Intracerebral hemorrhage	F	39	Multiple bilateral supratentorial and infratentorial	HTN	Y	N	24.3
10	Intracerebral hemorrhage	F	33	Multiple unilateral supratentorial	HTN	Y	Y	15.1
11	Intracerebral hemorrhage	M	36	Multiple bilateral supratentorial and infratentorial	_	Y	N	6.1
12	Focal neurological defect	F	63	Multiple bilateral supratentorial, infratentorial, and spinal cord Involvement	DM and HTN	N	N	0.7
13	Focal neurological defect	F	63	Single supratentorial	HTN	N	N	0.2
14	Isolated headaches	F	50	Single supratentorial	HTN	N	N	2.8
15	Incidental findings	M	12	Single supratentorial	smokes	Y	N	6.2

Notes: Y = Yes; N = No; HTN = Hypertension; M = Male F = Female; DM = Diabetes Mellitus Type 2; F/U= Follow-up