Abstract
Objective: To determine the ability of apolipoprotein E (APOE) genotypes to predict days of unconsciousness and a suboptimal functional outcome in traumatic brain injury (TBI) survivors.
Background: TBI is known to be associated with neuropsychological deficits and functional disability. Recent evidence indicates that APOE plays a pivotal role in CNS response to injury.
Methods: In this prospective study the authors determined the APOE genotypes and tested their ability to predict days of unconsciousness and functional outcome after at least 6 months in 69 survivors of TBI. A good functional outcome was defined as no dysarthria, behavioral abnormalities, or dysphasia; no severe cognitive abnormalities; and the ability to live independently.
Results: The odds ratio of more than 7 days of unconsciousness was 5.69 in those with the APOE-ε4 allele compared with those without the ε4 allele (95% CI, 1.69 to 20.0; p = 0.001). Only 1 of 27 subjects (3.7%) with the ε4 allele had a good functional outcome compared with 13 of 42 (31.0%) of those without the ε4 allele (p = 0.006). The OR of a suboptimal outcome (fair or unfavorable) was 13.93 for those with the ε4 allele compared with those without the allele after controlling for age and time of unconsciousness (95% CI, 1.45 to 133.97; p = 0.02).
Conclusion: The results demonstrate a strong association between the APOE-ε4 allele and a poor clinical outcome, implying genetic susceptibility to the effect of brain injury. Additional studies of TBI patients are warranted to confirm their findings.
The variability in outcome after traumatic brain injury (TBI) is only partly explained by prognostic factors such as age and estimated damage.1 Genetic factors that may influence the brain’s susceptibility to brain injury and the capacity for repair and regeneration may also aid in predicting outcome.
The gene that is responsible for the production of apolipoprotein E (apoE) is an attractive candidate for predicting outcome after TBI. apoE is one of the best-characterized apolipoproteins with respect to structure and function. The gene for apoE is located on chromosome 19 and is highly polymorphic. The three most common alleles are ε2, ε3, and ε4, which encode the main three isoforms of apoE: E2, E3, and E4.2-4 In the nervous system apoE is engaged in the redistribution of cholesterol from cells during membrane synthesis, and neuritic extension growth and repair.5,6 In various cell lines, apoE3 has been shown to increase growth and branching of neurites, whereas apoE4 was found to have the opposite effect.7-9 Furthermore, apoE4 binds more avidly than apoE3 to amyloid β peptide (Aβ) and enhances aggregation of Aβ—a cleavage product of amyloid precursor protein. Aβ accumulates in the brain as a component of the neuritic plaque leading to neuronal damage and dysfunction.10,11 This provides biological plausibility for a role of the APOE-ε4 allele in retarding repair and in leading to increased damage during the repair process after TBI.
The apoE polymorphism has been shown recently to predict both acute and chronic outcomes after brain damage. Alberts et al.12 demonstrated that APOE-ε4–carrying patients who survive intracerebral hemorrhage have a less favorable outcome than patients without the ε4 allele. Mayeux et al.13 showed that patients with the APOE-ε4 allele who survived head trauma had a 10-fold increased risk of AD than those without the ε4 allele. Furthermore, after TBI one study14 showed that in 16 patients the apoE4 genotype was associated with a higher risk of prolonged unconsciousness over 1 year, and Tardiff et al.15 showed that ε4–carrying patients undergoing cardiac surgery recover neuropsychological functions less efficiently.
In the current study we explored the association of the APOE-ε4 allele and outcome in patients who survived TBI. We hypothesized that the APOE-ε4 allele would predict those who were unconscious for at least 7 days and identify a subgroup that would be unlikely to recover satisfactorily.
Methods.
Patients.
TBI patients who participated in this study were recruited from two sources: 1) consecutive TBI patients referred to the Department of Brain Injury and Rehabilitation at the Loewenstein Rehabilitation Hospital and 2) consecutive patients seen at the outpatient rehabilitation clinic. Patients included only those with blunt trauma; those with penetrating injuries or anoxic brain damage were excluded. Informed consent for participating in the study was obtained from either the patients or their legal guardians, and the Ethical Committee of the Israeli Ministry of Health approved the study. Of the 72 patients originally recruited, 3 (2 with substantial anoxic damage and 1 with a penetrating brain lesion) were not included in the study. Analysis was performed on 69 patients (52 men and 17 women) who experienced blunt head injuries. The mean age was 35.7 ± 14.6 years (range, 18 to 73 years). On admission to the study, the Glasgow Coma Scale (GCS) score was recorded at the time of hospitalization in the neurosurgical wards and intensive care units.
Total days of loss of consciousness (LOC) was also recorded. The ability to obey commands was used as our operational definition of the end of coma.16,17 The ability to obey a command means that a message has been received, understood, and acted on.17 We preferred the use of LOC to that of the duration of post-traumatic amnesia (PTA), which is only rarely used routinely, despite the fact that PTA is also a useful indicator of the severity of injury.17
After a follow-up period of 6 to 8 months, a functional assessment regarding mobility and independence in activities of daily living was performed. Residual communicative disorders, either dysphasia or dysarthria, were determined by speech pathologists. Cognitive abnormalities and behavioral disturbances were evaluated by an interdisciplinary team including physiatrists trained in TBI rehabilitation, psychiatrists, psychologists, and neuropsychologists, all of whom are well trained and involved directly in TBI rehabilitation.18 Cognitive status and behavioral disturbances were grouped into four categories: none, mild, moderate, and severe. Mild cognitive disturbances were defined as deficits in attention and memory alone. Moderate cognitive disturbances were defined as additional deficits in integrated function, such as drawing conclusions and forming social judgments. Severe cognitive disturbances were problems with basic thought processes, such as perceptive deficits and apraxias, in addition to the previously mentioned disorders. Mild behavioral disturbances were defined as dysphoria, irritability, passivity, and dependency without any other overt disturbances. Moderate behavioral disturbances included outbursts of frank depression in addition to those described previously. Severe behavioral disturbances were characterized by such phenomena as confabulations, delusions, and frank maladaptive behavior.18
We considered that patients without dysarthria or dysphasia who were independent and who were without severe cognitive or behavioral abnormalities would have a good functional outcome. An unfavorable outcome included those who were fully dependent or had severe cognitive abnormalities. Patient assessment on follow-up, including cognitive and behavioral evaluations, was performed without knowledge of the apoE status of the patients.
Genetic analysis.
Genomic DNA was prepared from peripheral blood using lysis buffer and proteinase K, and it underwent phenol chloroform extraction and ethanol precipitation. PCRs for the APOE genotype were prepared in a final volume of 20 μL and contained 1 μg genomic DNA, 23 mM of each dNTP, 18.87 mM (NH4)2 SO4, 76 mM Tris (pH, 8.8), 76 mM MgCl2, 11.36 mM ditiothreitol, 193 μg/mL bovine serum albumin, 11.36% dimethyl sulfoxide, and 250 ng each of oligonucleotide primers A (5′-GAG AAG CTT GCG GCG CAG GCC CGG CTG GGC GCG-3′) and B (5′-TGA AGC TTC GCT CGG CGC CCT CGC GGG CCC GGG-3′). The reaction mixtures were heated to 95 °C for 15 minutes, and Taq polymerase (Boehringer, Mannheim, Germany) was added at 88 °C, 1 μL per reaction. DNA was amplified using a PCR thermal cycler (Perkin–Elmer Cetus, Foster City, CA). The PCR conditions were denaturation for 2 minutes at 95 °C, annealed for 2 minutes at 65 °C, and extended for 2 minutes at 72 °C. PCR was repeated for 35 cycles. The PCR products were digested with Cfo–1, and the fragments were separated on 8% polyacrylamide gel.
After electrophoresis, the gel was stained with ethidium bromide and visualized with ultraviolet light. The APOE genotype was determined by comparing the combination of fragment sizes with various controls.19
Statistical methods.
Univariate analysis was performed using the chi-square test with the Yates correction. When small numbers were present, Fisher’s exact test was used. For univariate analysis, ORs with Cornfield 95% CIs were calculated or, if not accurate, the exact CIs were substituted (Epi Info, version 3; Centers for Disease Control, Atlanta, GA). For multivariate analysis, a logistic regression model was used (SAS statistical package, version 6.03, Cary, NC). All predictor values were entered into the model and then variates that did not add significantly to the model were removed. After removal they were re-added one at a time and included in the model only if they added significance (p < 0.05).
Results.
The frequency of the alleles was as follows; E2/2-3, 4.3%; E2/3-7, 10.1%; E3/3-32, 46.4%; E2/4-1, 1.4%; E3/4-25, 36.2%; and E4/4-1, 1.4%. Thus, there were 27 subjects with at least one allele of ε4. This distribution is similar to that reported previously.20
There was a significantly higher frequency of a poor GCS score (66% of all patients had a GCS score < 9) and LOC over 7 days in those subjects with the APOE-ε4 allele compared with those without the allele (table 1). All the poorer outcomes except for the presence of epilepsy were more frequent in those with the ε4 allele. Mean age (± SD) was lower in those with the ε4 allele (31.8 ± 14.1 years) compared with those without the ε4 allele (38.2 ± 14.6 years, p = 0.07).
Comparison of traumatic brain injury patients with and without the APOE-ε4 allele
Only one of 27 (3.7%) of those with the ε4 allele had excellent function compared with 13 of 42 (31.0%) of those without such an allele (p = 0.006; see table 1). However, those with unfavorable outcomes were distributed evenly between those with and those without the ε4 allele.
A logistic model was used to predict those whose outcome was suboptimal (fair or unfavorable). All independent variables were entered into the model (GCS, period of LOC, age, and education). Only LOC more than 1 week, age, and the presence of the APOE-ε4 allele added significantly to the model (table 2). There was no significant interaction between LOC and the presence of the APOE-ε4 allele (p = 0.67) or between age and the allele (p = 0.93). The OR of a suboptimal outcome was 13.93 (95% CI, 1.45 to 133.97; p = 0.02) for those with the ε4 allele compared with those without the allele. The Hosmer–Lemeshow goodness of fit test was not significant (p = 0.99), indicating an adequate fit.
Predictors of a suboptimal (fair or unfavorable) outcome
Discussion.
The main finding of our study is that the ε4 allele of apoE predicted both short-term and long-term morbidity. Patients with the ε4 allele were much more likely to remain unconscious for more than 7 days and were less likely to have a good outcome.
Understanding the role of apoE in normal functioning of the nervous system and in patients with traumatic head injury depends on the cellular expression and localization of this protein. Interest in the role of apoE in the nervous system has increased dramatically since the worldwide recognition of the APOE-ε4 allele as a major risk factor for the development of late-onset AD.21,22 Investigators have demonstrated evidence for apoE localization in glia and astrocytes of human and nonhuman primates, and it is the major apolipoprotein in the CSF.23-25 Experimental models of cerebral injury have shown that the expression of apoE increases in astrocytes after ischemic brain insult and in degeneration neurons and their processes.26-28
In confirmation of a recent publication,29 our data indicate an association between the APOE-ε4 allele and outcome in patients with TBI, implying a genetic susceptibility. However, in our study there is a selection bias because only TBI survivors referred for rehabilitation and who completed an inpatient rehabilitation program were included in the study, in contrast to the previous study29 in which patients who died were included as well. Moreover, in the current study we focused especially on predicting those who will have a good recovery, and the issues of cognition and behavioral outcome were addressed for the first time with regard to the APOE genotype. However, it should be emphasized that because moderate and minor cognitive abnormalities are nearly universal after recovery from TBI, we included such patients as having a good recovery because they were able to live independently and without dysarthria, behavioral abnormalities, or dysphasia. The proportion of those with an unfavorable outcome was nearly identical in those with and without the APOE-ε4 allele, and the major difference was between those with either an excellent or fair overall functional outcome. Finally the model selection process involved searching several independent variables, and the p values are likely to be overly optimistic and should be interpreted with caution. Thus, additional studies of TBI patients are warranted to confirm these results.
It is possible that patients with the ε4 allele have pre-existing age-related deposits of Aβ, and it is the pre-existing disease rather than either impaired repair or the acute deposition of Aβ that explains the severity of the injury. It seems likely that if this were the case, there would be an interaction with age, with the association of a poor outcome, and with the ε4 allele found predominantly in the older patients. We did not find such an interaction; in fact, in both our study and that of Teasdale et al.,29 the age of those with the ε4 allele was younger.
Acknowledgments
Supported by a grant from H.J. Leir to G.F.
References
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.