Statins as a Medical Adjunct in the Surgical Management of Chronic Subdural Hematomas

Abstract

BACKGROUND:

By stabilizing immature leaky vessel formation in neomembranes, statin drugs have been suggested as a nonsurgical treatment option for chronic subdural hematomas (cSDH). Statin therapy seems to reduce conservatively managed cSDH volume. However, the usefulness of these medications in supplementing surgical treatment is unknown.

OBJECTIVE:

To investigate the effect of concurrent statin therapy on outcomes after surgical treatment of cSDH.

METHODS:

A retrospective single-institution cohort study of surgically managed patients with convexity cSDH between 2009 and 2019 was conducted. Patients receiving this diagnosis who underwent surgical decompression were included, and those without follow-up scans were excluded. Demographic, clinical, and radiographic variables were collected. cSDH size was defined as maximum radial thickness in millimeters on axial computed tomography of the head. Multivariable linear regression was performed to identify factors (including statin use) that were associated with preoperative to follow-up cSDH size change.

RESULTS:

Overall, 111 patients, including 36 patients taking statins on admission, were evaluated. Median time to follow-up postoperative imaging was 30 days (interquartile range, 17–42 days). Patients on statins were older (median, 75 years, range, 68–78.25 years vs. 69 years, range, 59–7 years; P = 0.006) and reported more antiplatelet use (67% vs. 28%; P < 0.001). Median change in follow-up size was 13 mm in both statin and nonstatin groups. Adjusting for other clinical covariates, statin use was associated with greater reduction in cSDH size (CE = −6.72 mm, 95% confidence interval, −13.18 to −0.26 mm; P = 0.042).

CONCLUSIONS:

Statin use is associated with improved cSDH size postoperatively. Statin drugs might represent a low-cost and low-risk supplement to the surgical management for patients with cSDH.

INTRODUCTION

Chronic subdural hematoma (cSDH) represents a high mortality and morbidity diagnosis in neurosurgery.^1,2 In the aging U.S. population, the incidence of cSDH has been approximated as up to 58 cases per 100,000 individuals.^3,4 Traditionally, treatment of symptomatic cSDH has involved surgical decompression, with few alternatives available.^5,6 Specifically, no pharmacologic intervention is recommended for monotherapy or as an adjunct to surgery. However, early evidence exists for possible therapeutic targets based on our emerging understanding of cSDH pathophysiology, and these targets have yet to be fully explored. An enhanced understanding of the clinical and surgical factors that affect clinical outcomes in this patient population would enable better management.⁷

Myriad pathophysiologic mechanisms for cSDH beyond the simple evolution of an acute subdural hematoma have been described.⁸ Animal and molecular studies have pointed to the role of inflammation from blood products and neovascularization from proangiogenesis signals.^8-10 Rebleeding, neomembrane formation with neovascularization, and a balance between coagulation and fibrinolysis all contribute to cSDH formation.^11-13

In addition to LDL reduction, HMG-CoA (hydroxymethylglutaryl coenzyme A) reductase inhibitors (statins) have a wide range of effects on several biochemical pathways, including inflammation and coagulation, which are implicated in cSDH pathophysiology.^14-17 Based on these beneficial effects, a recent randomized controlled trial (RCT) reported that statin use was associated with significantly decreased cSDH volume, improved neurologic function, and decreased need for surgical decompression for conservatively managed patients.⁵

This benefit of statins in nonsurgical patients raises the potential for benefit as an adjuvant (or neoadjuvant) therapy in surgically managed cSDH.¹⁸ To this end, the goal of this pilot study was to evaluate whether preoperative statin use is associated with improved radiologic or clinical outcomes at follow-up in surgically managed patients with cSDH. We hypothesized that preoperative statin use would be associated with improved outcomes at follow-up.

METHODS

Study Design

A retrospective cohort study was conducted evaluating patients treated surgically for convexity cSDH in the inpatient setting between 2009 and 2019 at a regional level 1 trauma center in the Southeastern United States. Approval was obtained from the institutional review board (IRB number 190599).

Patient Selection

Patients ages 18 years and older who underwent a procedure for cSDH were identified within the electronic medical record using Current Procedural Terminology codes 61314, 61312, 61313, 61154, 61108, and 61156 and International Classification of Diseases codes 432.1, I62.00, and S06.5X0A. Patients who experienced other types of intracranial hemorrhages (e.g., hemorrhagic contusions, subarachnoid hemorrhage, or epidural hemorrhage) were excluded. Patients who had a cSDH that was not in a convexity distribution (i.e., tentorial or parafalcine) were excluded, and patients who did not have follow-up brain imaging were excluded.

Patient Clinical Variables

Patient demographic information was captured including age at presentation, gender, race, ethnicity, and insurance type. Racial groups were categorized as white, African American, Asian, Pacific Islander, American Indian, or other. Insurance type was categorized as uninsured, Medicaid, Medicare, private insurance, or other. Patient comorbid conditions, neurosurgical imaging findings (presence and degree of midline shift (MLS) in millimeters and presence of postoperative pneumocephalus) and treatment modalities (burr holes, craniotomy, or subdural evacuation port system) were collected.

Details about prehospital medication were reviewed. Medications (present on admission medication reconciliation) including the current use of statins, antiplatelets, anticoagulants, and/or steroids was recorded. No patients were prescribed statins with cSDH as the indication. Pertinent historical data including history of previous head trauma, history of previous myocardial infarction, stroke, previous neurosurgical procedure, or another previous subdural hematoma were incorporated, as well as comorbid conditions reported to have an association with cSDH recurrence, including coronary artery disease, diabetes, alcoholism, metastatic cancer, chronic obstructive pulmonary disease, and sleep apnea.²

Initial imaging characteristics captured including the laterality of the surgical cSDH, measurements of the maximum thickness of the subdural hematoma, and the presence of MLS and/or brainstem compression were included. For surgical data, the date of operation and type of procedure (i.e., subdural evacuation port system, burr holes, or craniotomy) as well as use of a drain were collected. After surgery, similar imaging characteristics were captured for follow-up scans.

Clinical Outcomes

The primary study outcome was change in cSDH size in millimeters from preoperative imaging to follow-up imaging. This change was calculated by subtracting the maximum width of preoperative cross-sectional cSDH size from the maximum width cross-sectional cSDH size on follow-up scan. This method has been used in previous studies of cSDH and reflects the measurement frequently reported in radiographic interpretations when noting subdural hematoma (SDH) size.² Secondary clinical outcomes included postoperative complications (headache, seizure, delirium, stroke, and sepsis), need for repeat surgery, 90-day readmission, and the modified Rankin Scale (mRS) score. This 7-point scale is used to assess functional outcome after strokes, with 6 being death and 0 being “no symptoms.”¹⁹ Good clinical outcome was defined as an mRS score of 0–2, corresponding to “no symptoms” through “slight disability” at follow-up.

Statistical Analysis

Bivariate analyses were conducted to compare demographic, comorbidity, and clinical outcomes in patients with and without statin use. Continuous and categorical variables were analyzed using a Student t test, Mann-Whitney test, and Pearson χ² tests, respectively. Multivariable linear regression was performed to identify factors associated with change in cSDH size. The factors known to influence outcome (age,²⁰ diabetes,²¹ presence of neomembranes,²² MLS,² and postoperative pneumocephalus²), statin use (exposure of interest), and potential confounding factors (antiplatelet use, history of myocardial infarction, and time to follow-up scan) were included in the analysis. Statistical significance was set a priori at P < 0.05. Statistics were computed using R programming language (R core team, 2020) and RStudio (RStudio Team (2020) (RStudio: Integrated Development for R. RStudio, PBC, Boston, Massachusetts, USA).

RESULTS

Patient Demographics

Of 194 surgical patients, 19 had nonconvexity SDH, and 64 did not have a follow-up CT scan. Therefore, 111 patients met inclusion criteria for this study, with 36 patients (32%) taking prescribed statin drugs on admission. The demographic data are summarized in Table 1. Patients were predominantly male (n = 70, 63%) and white (n = 97, 87%). The median age was 71 years (interquartile range [IQR], 63–78). Patients using statin drugs were older (75 vs. 69 years; P = 0.018). No other significant demographic differences were found between groups.

Table 1.

Demographics of Total Cohort

	No Statin (N = 75)	Statin (N = 36)	Total (N = 111)	P Value
Demographics
Median age (years) (IQR)	69 (59–77)	75 (68–78)	71 (62.5–78)	0.018
Race				0.465
White	65 (86.7)	32 (88.9)	97 (87.4)
African American	7 (9.3)	4 (11.1)	11 (9.9)
Asian	3 (4.0)	0 (0.0)	3 (2.7)
Male sex	45 (60.0)	25 (69.4)	70 (63.1)	0.334
Insurance type				0.128
Private/other	12 (16.0)	1 (2.8)	13 (11.7)
Medicare	31 (41.3)	17 (47.2)	48 (43.2)
Uninsured/Medicaid	32 (42.7)	18 (50.0)	50 (45.0)

Values are number (%) except where indicated otherwise.

Bold values indicate statistical significant (P < 0.05).

Preoperative Clinical Characteristics

Clinical characteristics are summarized in Table 2. The median number of comorbidities for our patients was 2 (IQR, 1–4).The most common comorbidity overall was hypertension (n = 70, 63.1%), followed by coronary artery disease (n = 38, 34.2%). Forty-five patients (40.5%) were taking antiplatelet medication at the time of admission, whereas 17 patients (15.3%) were taking anticoagulants. In addition, antiplatelet use (P < 0.001) was associated with statin use. Statin users had an increased number of medical comorbidities compared with non–statin users (median, 3, IQR, 2–6 vs. 2, IQR, 1–3). In terms of specific medical comorbidities, statin users more frequently had diagnoses of congestive heart failure (n = 5; P = 0.023) and previous myocardial infarction (n = 12; P < 0.001) (Table 2). Otherwise, no significant differences in medical history existed between groups.

Table 2.

Characteristics of Statin and Non–Statin Users

	No Statin (N = 75)	Statin (N = 36)	Total (N = 111)	P Value
Medical history
Hypertension	42 (56.0)	28 (77.8)	70 (63.1)	0.026
Diabetes	17 (22.7)	12 (33.3)	29 (26.1)	0.231
Coronary artery disease				0.116
Yes	22 (29.3)	16 (44.4)	38 (34.2)
Congestive heart failure				0.023
Yes	2 (2.7)	5 (13.9)	7 (6.3)
Myocardial infarction				<0.001
Yes	6 (8.0)	12 (33.3)	18 (16.2)
Dementia				0.822
Yes	5 (6.7)	2 (5.6)	7 (6.3)
Coagulopathy				0.746
Yes	3 (4.0)	1 (2.8)	4 (3.6)
Cancer, currently receiving treatment				0.396
Yes	5 (6.7)	1 (2.8)	6 (5.4)
Liver failure				0.486
Yes	1 (1.3)	0 (0.0)	1 (0.9)
Sleep apnea				0.345
Yes	3 (4.0)	3 (8.3)	6 (5.4)
Chronic obstructive pulmonary disease				0.331
Yes	6 (8.0)	5 (13.9)	11 (9.9)
Dyspnea				0.039
Yes	0 (0.0)	2 (5.6)	2 (1.8)
Chronic kidney disease				0.396
Yes	5 (6.7)	1 (2.8)	6 (5.4)
Previous SDH				0.002
Yes	5 (6.7)	10 (27.8)	15 (13.5)
Previous neurosurgery				0.079
Yes	6 (8.0)	7 (19.4)	13 (11.7)
Number of comorbidities				<0.001
Median	2	3	2
Q1, Q3	1, 3	2, 6	1,4
Antiplatelet medication use				<0.001
Yes	21 (28.0)	24 (66.7)	45 (40.5)
Anticoagulant medication use				0.772
Yes	12 (16.0)	5 (13.9)	17 (15.3)
Presentation
Head trauma				0.890
Yes	49 (65.3)	24 (66.7)	73 (65.8)
Focal deficits				0.843
Unknown	1 (1.3)	0 (0.0)	1 (0.9)
Yes	40 (53.3)	18 (50.0)	58 (52.3)
Transient	3 (4.0)	1 (2.8)	4 (3.6)
Seizure				0.283
Unknown	1	0	1
Yes	6 (8.1)	1 (2.8)	7 (6.4)
Glasgow Coma Scale score				0.747
Median	15	15	15
Q1, Q3	14, 15	15, 15	14, 15
Acute on chronic presentation				<0.001
Yes	57 (76.0)	15 (41.7)	72 (64.9)
Imaging
Maximum thickness of SDH (mm)				0.127
Median	20	20	20
Q1, Q3	13, 25	15, 25.5	14, 25
Laterality				0.861
Left	29 (38.7)	12 (33.3)	41 (36.9)
Right	29 (38.7)	15 (41.7)	44 (39.6)
Bilateral	17 (22.7)	9 (25.0)	26 (23.4)
Midline shift (mm)				0.088
Median	8	6	8
Q1, Q3	5, 11.5	3.5, 10	4, 11
Brainstem compression				0.049
Yes	18 (24.0)	3 (8.3)	21 (18.9)
Membranes present on preoperative scan				0.820
Unknown	0	3	3
Yes	29 (38.7)	12 (36.4)	41 (38.0)
Surgery				0.592
Time to surgery (days)
Median	1	1	1
Q1, Q3	1, 2.75	1, 2	1, 2
Procedure				0.158
Subdural evacuation port system	5 (6.7)	2 (5.6)	7 (6.3)
Burr hole	43 (57.3)	28 (77.8)	71 (64.0)
Craniotomy	24 (32.0)	6 (16.7)	30 (27.0)
Craniectomy	3 (4.0)	0 (0.0)	3 (2.7)
Focal deficit resolved				0.399
Unknown	27	17	44
Yes	33 (68.8)	11 (57.9)	44 (65.7)
Postoperative symptoms worse				0.445
Yes	2 (2.7)	2 (5.6)	4 (3.6)
Postoperative complications*				0.468
Yes	26 (34.7)	10 (27.8)	36 (32.4)
Postoperative imaging
Postoperative size of SDH (mm)				0.012
Median	9	12	11
Q1, Q3	4, 13	10, 17	6.85, 14
Postoperative midline shift (mm)				0.124
Unknown	12	1	13
Yes	44 (69.8)	19 (54.3)	63 (64.3)
Postoperative pneumocephalus				0.989
Yes	40 (63.5)	21 (63.6)	61 (63.5)
Postoperative pneumocephalus size (mm)				0.636
Median	12	12	12
Q1, Q3	7, 16	9, 16.5	7.25, 16.5
Complete resolution of bleed				0.081
Yes	10 (13.3)	1 (2.8)	11 (9.9)

Values are number (%) except where indicated otherwise.

Bold values indicate statistical significant (P < 0.05).

SDH, subdural hematoma; Q, quartile.

^*Included headache, seizure, delirium, stroke, and sepsis.

On clinical presentation, 73 patients overall (65.8%) reported recent head trauma. A total of 9 patients presented with an international normalized ratio >1.7, and 5 patients presented with a Glasgow Coma Scale (GCS) score <8. Statin users had a similar GCS score at presentation (median, 15) and symptoms compared with non–statin users, but they were less likely to have an acute component to the bleed (P < 0.001). These were the only significant differences between groups.

On imaging at presentation, the median maximum thickness (mm) of SDH was 20 mm (IQR, 14–25 mm). The only difference between groups for imaging variables was the reduced frequency of brainstem compression for the statin group, which was seen in only 3 patients (8.3%) using statins compared with 18 patients (24%) who did not use statins (P = 0.049). Procedure choice was similar between statin and non–statin groups (Table 2).

Clinical Outcomes

Fifty-nine patients (53%) were discharged home rather than to postacute care, and 45 patients (41%) were readmitted within 90 days. Median time to follow-up imaging was 30 days (IQR, 17–42) (Table 3). The rate of postoperative complications was similar between statin and non–statin groups (27.8% and 34.7%, respectively; P = 0.61). Statin use was not associated with need for repeat surgery (P = 0.69, n = 5 [14%] vs. n = 7 [9%]) or readmission within 90 days (P = 0.98, n = 14 [39%] vs. n = 31 [41%]). Five of the 14 (36%) readmissions in the statin group were for a neurosurgical reason, compared with 17 of 31 (55%) for the non–statin group (P = 0.39). mRS scores ranged from 0 to 5 at follow-up and were missing in 20 patients. Fifty patients (55%) had a good functional outcome at follow-up, defined as an mRS score of 0–2 (Table 3).

Table 3.

Follow-Up

	No Statin (N = 75)	Statin (N = 36)	Total (N = 111)	P Value
Time to follow-up computed tomography (days)	34	25	31	0.034
Follow-up size of subdural hematoma (mm)				0.202
Median	7	10	7.5
Q1, Q3	4, 12	5, 17	4, 13
Follow-up midline shift				0.627
Unknown	4	1	5
Yes	23 (32.4)	13 (37.1)	36 (34.0)
Readmission				0.806
Yes	31 (41.3)	14 (38.9)	45 (40.5)
Required second surgery				0.469
Yes	7 (9.3)	5 (13.9)	12 (10.8)
Modified Rankin Scale score				0.578
0	9 (15)	5 (14)	14
1	16 (26)	8 (22)	24
2	9 (15)	3 (8)	12
3	7 (11)	5 (14)	12
4	15 (24)	8 (22)	23
5	6 (10)	0 (0)	6

Values are number (%) except where indicated otherwise.

Bold values indicate statistical significant (P < 0.05).

Table 4 is a multivariable logistic regression model for clinical outcome at follow-up using the mRS. Among age >75 years, GCS score, number of comorbidities, and statin use, only age >75 years was associated with good functional outcome (P = 0.034).

Table 4.

Multivariable Logistic Regression Modeling for Good Functional Outcome, Defined as a Modified Rankin Scale Score of 0–2

Variables	Adjusted Odds Ratio (95% Confidence Interval)	P Value
Age >75 years	0.38 (0.15–0.93)	0.034
Glasgow Coma Scale	1.04 (0.86–1.25)	0.569
Number of comorbidities	0.88 (0.7–1.1)	0.267
Statin	1.39 (0.48–4.04)	0.542

Bold values indicate statistical significant (P < 0.05).

Figure 1 shows the change in cSDH size for statin users compared with non–statin users. The median change in cSDH size was 13 mm in both groups (Figure 1). Two of 36 of statin users (5.6%) had hemorrhage expansion at follow-up (defined as size increase ≥1 mm), compared with 7 of 75 non–statin users (9.3%) (P = 0.76). In a multivariable linear regression model (Table 5) for the change in SDH size (in millimeters), statin use (P = 0.042), presence of membranes (P = 0.009), and degree of MLS (in millimeters) (P = 0.009) were associated with a greater reduction in cSDH size. Conversely, diabetes (P = 0.027) and postoperative pneumocephalus size (in millimeters) (P = 0.012) were associated with an increased cSDH size.

Figure 1.

Change in chronic subdural hematoma size from preoperative scan to follow-up computed tomography scan, for statin and non–statin users.

Table 5.

Multivariable Linear Regression Model for Change in Subdural Hematoma Size (mm) on Follow-Up Scan

Variables	Coefficient (95% Confidence Interval)	P (t Test)
Age	−0.18 (−0.41 to 0.04)	0.112
Antiplatelet	−5.07 (−11.53 to 1.39)	0.12
Diabetes	6.37 (0.77 to 11.97)	0.027
Membranes	−6.95 (−12.07 to −1.83)	0.009
Midline shift (preoperative) (mm)	−0.69 (−1.23 to −0.14)	0.015
History of myocardial infarction	6.22 (−2.75 to 15.19)	0.168
Postoperative pneumocephalus size (mm)	0.53 (0.12 to 0.94)	0.012
Statin	−6.72 (−13.18 to −0.26)	0.042
Time to follow-up scan (days)	–0 (–0.12 to 0.12)	0.999

Bold values indicate statistical significant (P < 0.05).

Table 6 is a multivariable logistic regression model for functional outcome at follow-up, using follow-up hematoma size as well as age >75 years, GCS score, and number of comorbidities as independent variables. In this model, age >75 years (P = 0.035) and follow-up size (P = 0.028) were associated with decreased odds of good functional outcome (i.e., mRS score 0–2).

Table 6.

Multivariable Logistic Regression Modeling for Good Functional Outcome, Defined as a Modified Rankin Scale Score of 0–2

Variables	Adjusted Odds Ratio (95% Confidence Interval)	P Value
Age >75 years	0.37 (0.14–0.93)	0.035
Follow-up hematoma size (mm)	0.93 (0.87–0.99)	0.018
Glasgow Coma Scale	1.11 (0.92–1.35)	0.268
Time to follow-up scan (days)	1.001 (0.978–1.025)	0.900

Bold values indicate statistical significant (P < 0.05).

DISCUSSION

The aim of the present study was to evaluate whether statin drugs are associated with enhanced postsurgical outcomes for cSDH. Our results suggest that preoperative use of statin medications is associated with greater reductions in cSDH size postoperatively. Specifically, statin use was found to be associated with an additional 6.72 mm reduction in cSDH compared with nonusers, accounting for major factors known to be associated with outcome in cSDH. These results indicate that statin drugs may represent a safe and efficacious supplement to surgical management of cSDH.

Evidence from Conservative Management Studies

The incidence of SDHs is projected to increase by approximately 78% from 2020 to 2040, particularly among the increasing elderly population.²³ The projected increase in neurosurgeons during this period is disproportionately less, suggesting that future neurosurgeons may not be able to meet the surgical demand for decompression of cSDHs.²³ Considering this potential treatment gap, it is imperative for neurosurgery to pursue continuing research on conservative management of this disease. Although preexisting literature on this topic is limited, previous studies have made similar conclusions that statin use for patients with cSDHs may be a beneficial option for clinical management.²⁴ In an RCT conducted by Jiang et al. in 2018,⁵ atorvastatin was shown to significantly reduce conservatively managed cSDH on follow-up scan. Our study differed from this trial in design (surgical retrospective cohort vs. nonsurgical RCT). Despite these differences, the results are similar in that statin use is associated with a reduction in cSDH on follow-up scan for patients undergoing surgical decompression. In light of the recurrence rate (5%–30%) and morbidity after surgical treatment for cSDH, consideration should be given to conservative treatment options.^25,26

Statins in Surgical Patients with cSDH

Perioperative statin therapy was shown to improve both short-term and long-term surgical outcomes, regardless if the surgical intervention is cardiac.^27,28 For patients undergoing elective or emergent noncardiac surgery, Chan et al. in 2008²⁷ reported that perioperative statin therapy is associated with decreased perioperative mortality and morbidity, suggesting that statins may be underused in noncardiac patients. Lindenauer et al.²⁸ conducted a study in 2004 of a large retrospective cohort of patients undergoing major noncardiac surgery, finding that statin use in the perioperative period was associated with a 1% absolute reduction in hospital mortality and a 38% reduction in odds of in-hospital mortality, with the results found to be independent of type of surgery or cardiac risk factors. As previously mentioned, Tang et al.¹⁸ found a reduced recurrent risk in surgical patients with cSDH taking statins.

Although these results indicate a benefit of statins with respect to the change in the cSDH size, there were no associated differences in other clinical outcomes such as resolution or worsening of neurologic deficits, complication rate, or readmission. Unlike the Tang et al. study,¹⁸ in which a difference in reoperation rate was seen (4.8% for statin users compared with 15% for non–statin users), no difference was seen in the present study (14% vs. 9%). In addition, a logistic regression model did not find a relationship between statin and good clinical outcome defined by the mRS. This finding might suggest that the improved cSDH size reduction is not clinically meaningful, but it is likely that the higher number of comorbidities in the statin group made postoperative recovery more challenging, masking any effect from statins in these variables. This theory is supported by the fact that most readmissions (64%) within 90 days were for nonneurosurgical reasons for statin users.

Significantly more patients in the statin group had a history of previous SDH compared with patients in the non–statin group (Table 2). The median maximum diameter cSDH at index admission for patients with previous SDH and who used statins was 19.5 mm, whereas in the group of patients with previous SDH but no statin use, it was 25 mm (P = 0.36). Future studies may find a statistically significant difference, which might suggest that statins mitigate the reaccumulation of cSDH. Furthermore, the proportion of patients in the statin group who needed craniotomy or craniectomy (17%) was less than the proportion in the non–statin group (36%) (P = 0.06). Although this finding did not reach significance, this difference, and corresponding increase rate of burr-hole use, is clinically relevant. In addition, differential procedure selection may explain the larger median postoperative cSDH diameter in patients using statins.

Although a greater improvement in cSDH size is not in itself important, there is evidence that faster resolution of the bleed can reduce local inflammation and allow for earlier brain expansion.^12,29,30 There is some evidence for this theory in our data, which showed an inverse relationship between follow-up size and functional outcome. Although this finding does not necessarily mean that statins improve functional outcome, the relationship between size and functional outcomes places further importance on therapies that enhance the hemorrhage evacuation.

Other Adjunct Medical Treatments

Other medical therapies to treat cSDH in addition to statins have also been recently suggested, such as the use of steroids, tranexamic acid, and platelet-activating factor receptor antagonists.^31-33 Compared with these other medications, statin drugs have been shown to have relatively low cost, making them more accessible for underinsured populations.^34,35 Further, many studies have shown the low risk of adverse effects by use of statins compared with other drugs such as steroids.³⁶ Steroids, in particular, have been shown to have a large range of side effects, particularly if used long-term or in the postoperative period.^6,10,37-39 A recent RCT comparing dexamethasone and placebo showed that dexamethasone was associated with worse functional outcome and an increased rate of adverse events.³⁹ As many of the elderly population in the United States are prescribed statin drugs (≤18% as of 2016), the demographic of patients with cSDH is most similar to that of typical statin users, suggesting that their use might be particularly safe and efficacious.⁴⁰

Other Associations with Change in cSDH Size

Certain comorbidities, such as diabetes mellitus type II, were shown to be associated with increase in cSDH size. This result is unsurprising, because diabetic patients have been shown extensively in the literature to have worsened clinical outcomes in a multitude of diseases, including cSDH.^21,41,42 The size of postoperative pneumocephalus was found to be associated with increased cSDH on postoperative scan as well. Although the role of postoperative pneumocephalus in previous studies had unknown association with clinical outcomes, our study showed that the relative size of this pneumocephalus may be of more clinical importance than previously believed.⁴³ Our results indicated that presence of neomembranes and degree of MLS were associated with a greater decrease in cSDH size at follow-up. Although the possible mechanisms are unclear, greater MLS may simply be related to increased preoperative cSDH size and, therefore, greater room for improvement in follow-up cSDH size. It is also possible that greater MLS is associated with better brain expansion postoperatively. For neomembranes, their presence may indicate a bleed that is further along the time course toward resolution. This association held when craniotomy/craniectomy was substituted into the model, indicating that this association held for patients undergoing craniotomy/craniectomy as well as patients undergoing burr-hole surgery (i.e., regardless of ability to remove the membranes intraoperatively).

Study Limitations

This study is limited by the retrospective cohort design, which makes nuanced interpretation of these data challenging, especially considering the heterogeneity of our patient cohort with respect to comorbid medical conditions. The exposure to statins was determined by review of documented home medications during each patient’s hospital admission. We cannot guarantee that patients were taking their prescribed statin medications at time of admission. In addition, apart from myocardial infarction, other medical indications for statin use were not accounted for in the regression model. Also, the exposure in this cohort study was preoperatively prescribed statin therapy; these study results do not directly support the initiation of statin therapy postoperatively. Nonetheless, the evidence for greater improvement in cSDH size supports the notion that future investigations should consider adjuvant statin therapy.

This study did not find an association between statin use and functional outcome. However, it is likely that the study was underpowered to investigate this association. For this reason, the findings in this study point to the need for prospective studies that are designed to assess the relationship between statin use and functional outcome.

Our study used two-dimensional measurements to account for changes in cSDH. Volume of cSDH may be more correlated with clinical outcomes and may also introduce additional error in measurements. In addition, the number of patients included in our study was limited by the fact that some patients were lost to follow-up, because fewer patients insured by Medicaid or who were uninsured had repeat scans. This limitation may introduce selection bias into our study. Data on the specific dose and type of statin medication, as well as duration or adherence, were not available, further limiting analysis of these results.

Additional single-center or multicenter studies are needed to corroborate the finding that statins may improve postoperative reduction in cSDH size as well as to investigate the possible effect of statin therapy on functional outcome. Rates of comorbid conditions with an indication for statin therapy among patients with cSDH have not been thoroughly described. Sufficient cohort and case-control evidence for this adjuvant treatment is needed to justify a more formal RCT of statin therapy in patients with surgically treated cSDH.

CONCLUSIONS

Statin use is associated with improved cSDH size postoperatively. Statin drugs might represent a low-cost and low-risk supplement to the surgical management for patients with cSDH. Future prospective trials exploring the effects and dosages of statin use on this patient population are needed to better understand this effect.

Abbreviations and Acronyms

cSDH

Chronic subdural hematoma

GCS

Glasgow Coma Scale

IQR

Interquartile range

MLS

Midline shift

mRS

Modified Rankin Scale

RCT

Randomized controlled trial

SDH

Subdural hematoma