Typical transverse T1-weighted magnetic resonance images and illustrative examples of the cross-sectional version of the Structural Imaging Evaluation of Normalized Atrophy software (SIENAX) output of a healthy control subject (A and B, respectively) and a patient with relapsing-remitting multiple sclerosis (RRMS) (C and D, respectively) of similar age. Loss of brain volume is seen in the RRMS patient.
Box plots comparing magnetic resonance measurements of normalized brain volume (NBV) of the healthy control subjects (HCs) and patients with relapsing-remitting multiple sclerosis (RRMS) without (RRMSϵ4−; n = 58) and with (RRMSϵ4+; n = 18) the apolipoprotein E ϵ4 allele. The NBV values of RRMSϵ4+ patients are significantly (P = .01) lower than those of the HC and RRMSϵ4− groups. In each box, the horizontal line represents the median; the box length shows the range within which the central 50% of values fall, with the box edges at the first and third quartiles; and the limit lines represent the range of values.
Box plots comparing magnetic resonance measurements of normalized brain volume (NBV) of the healthy control subjects (HC) and patients with relapsing-remitting multiple sclerosis (RRMS) without (RRMSϵ4−) and with (RRMSϵ4+) the apolipoprotein E ϵ4 allele. In this case, from the whole group of RRMS patients, only those with short disease duration (<3 years) and absence of disability (Expanded Disability Status Scale score, <2) are selected (22 patients without and 8 patients with the ϵ4 allele). The NBV values of the HCs are from a subgroup of 12 subjects age matched with patient subgroups (mean age, 30 years in each group). The NBV values of the RRMSϵ4+ group are significantly (P = .04) lower than those of the HC and RRMSϵ4− groups. In each box, the horizontal line represents the median; the box length shows the range within which the central 50% of values fall, with the box edges at the first and third quartiles; and the limit lines represent the range of values.
De Stefano N, Bartolozzi ML, Nacmias B, Zipoli V, Mortilla M, Guidi L, Siracusa G, Sorbi S, Federico A, Amato MP. Influence of Apolipoprotein E ϵ4 Genotype on Brain Tissue Integrity in Relapsing-Remitting Multiple Sclerosis. Arch Neurol. 2004;61(4):536-540. doi:10.1001/archneur.61.4.536
Recent clinical and imaging studies have raised the hypothesis that patients with multiple sclerosis (MS) and the apolipoprotein E (ApoE) ϵ4 allele may have a more severe disease course than those without the ApoE ϵ4 allele. This seems to be related to more extensive tissue destruction and less efficient neuronal maintenance and repair in ApoE ϵ4 carriers.
To evaluate the influence of different ApoE genotypes on brain tissue integrity in patients with relapsing-remitting MS (RRMS).
We determined the ApoE genotype in 76 RRMS patients. Conventional T1-, T2-, and proton density–weighted magnetic resonance (MR) images were obtained for each patient and in a group of demographically matched healthy control subjects. On conventional T1-weighted MR images, an automated analysis tool was used to obtain total brain volumes normalized for head size (NBVs). Total brain lesion load was estimated on proton density– and T2-weighted MR images.
From the whole group of RRMS patients, we identified 18 with and 58 without the ϵ4 allele. Both patient groups were not significantly different in age, age of disease onset, clinical disability, and disease duration. Carriers of the ϵ4 allele showed significantly (P = .01) lower NBVs than controls and non–ϵ4 allele carriers. When a similar analysis was performed on only those patients with both very short disease duration and absence of clinical disability, NBV values were still significantly lower in RRMS patients with the ϵ4 allele than in those without it (P = .02) and in controls (P = .007). In contrast, RRMS patients with different ApoE genotypes did not show significant differences in values of total brain T2-weighted lesion volumes.
The presence of significant NBV decreases only in the group of RRMS patients with the ApoE ϵ4 genotype provides new evidence that links ApoE ϵ4–related impaired mechanisms of cell repair and severe tissue destruction in MS. Results of the present study suggest that this negative influence of the ApoE ϵ4 genotype might be active from the earliest disease stages.
Apolipoprotein E (ApoE) is a ubiquitous apolipoprotein involved in the metabolism of cholesterol.1 It has been suggested that ApoE-dependent uptake of lipoproteins may play an important role in the development, maintenance, and response to injury of the central nervous system.2,3 Apolipoprotein E is expressed in humans in 3 isoforms coded by the alleles ϵ2, ϵ3, and ϵ4 and is abundantly present in the brain. In many recent studies, the genotype ϵ4 has been consistently associated with severe neurodegeneration and seems to constitute a risk factor in several neurological disorders.4,5 This seems to be related to more extensive tissue destruction and less efficient neuronal maintenance and repair in ApoE ϵ4 carriers.
Recently, clinical and imaging studies have raised the hypothesis that in multiple sclerosis (MS), as in many neurological disorders, patients with the ApoE ϵ4 allele may have a more severe disease course than those without the allele. More specifically, the presence of the ϵ4 allele in the ApoE genotype of patients with MS seems to be related to poorer recovery after relapses,6 faster disease progression,7 higher frequency of severe disease form,8 and higher T1-weighted lesion load and axonal damage seen on magnetic resonance (MR) images.9,10 In MS, in contrast to what has been described in patients with other neurological disorders,4 the presence of the ϵ4 allele seems to have little or no effect on the age of disease onset, and its deleterious effect seems to be evident only late in the course of the disease.4,11 However, whether significant differences in brain tissue damage between MS patients with and without the ϵ4 allele can be detected at early disease stages has not been fully evaluated by previous studies.
Cerebral tissue loss can be quantitatively assessed from conventional T1-weighted MR images with the aid of software allowing automatic or semiautomatic computed measurements of total brain volume.12- 17 With the use of these methods, visually undetectable brain atrophy can be found and seems to be relevant in MS patients from the earliest disease stages.18 Because brain atrophy in MS should be interpreted as the consequence of destructive pathological changes (ie, demyelination, gliosis, and axonal damage) that occur in lesions and normal-appearing brain,18,19 measures of cerebral volumes normalized for head size can represent a reliable marker of adverse outcome of this disease. Thus, measurements of brain atrophy can discriminate the real disease-modifying effect of the ApoE genotype in MS.
With this background, we performed conventional brain MR imaging on a group of patients with relapsing-remitting MS (RRMS), assessed their ApoE genotype, and used an automated method for computing analysis of total brain volumes (Structural Imaging Evaluation of Normalized Atrophy [SIENA]17) to evaluate the influence of the ApoE genotype on tissue destruction in the whole group of patients and in a subgroup of them with minimal disease duration and no clinical disability.
We studied 76 consecutive patients (50 women and 26 men aged 18-55 years [mean ± SD age, 34.5 ± 8 years]) with RRMS.20 When MR images were obtained, none of the RRMS patients had been treated with corticosteroids for at least 1 month, and 7 of the 76 were being treated with interferon beta. A neurological evaluation (which included the rating of disability using the Expanded Disability Status Scale [EDSS]21) was performed by an experienced observer on each patient within 24 hours of the performance of the MR examination. A population of 22 demographically matched healthy control subjects (HCs; 14 women and 8 men aged 25-56 years [mean ± SD age, 34.9 ± 8 years]) was also studied for comparisons. The HCs were recruited from laboratory and hospital workers and included in the study if they had a negative history for neurological disorders and no abnormalities on conventional brain MR images. Before study initiation, local ethical committee approval and written informed consent were obtained from all the subjects.
The ApoE genotype was determined in all RRMS patients through standard polymerase chain reaction and restriction analyses using the method previously described.22
All RRMS patients and HCs were examined using the same MR protocol. We used a transverse dual-echo, turbo spin-echo sequence (repetition time/echo time 1/echo time 2, 2075/30/90 milliseconds; 256 × 256 matrix; 1 signal average; 250-mm field of view) yielding proton density (PD)– and T2-weighted images with 50 contiguous 3-mm slices acquired parallel to the line connecting the anterior and posterior commissures. Subsequently, a T1-weighted sequence (repetition time/echo time, 35/10 milliseconds; 256 × 256 matrix; 1 signal average; 250-mm field of view) was performed. This sequence yielded image volumes of 50 slices, 3-mm thick, oriented to match the PD/T2-weighted acquisition exactly.
Classification of T2-weighted lesion volume (LV) was performed in each patient by a single observer by means of a user-supervised thresholding technique. The observer was unaware of the subjects' identities. Lesion borders were determined primarily on PD-weighted images, but information from T2- and T1-weighted images were also considered, as the software used (MEDx; Sensor Systems Inc, Sterling, Va) offered the ability to toggle between the PD-, T2-, and T1-weighted images. This provided the operator with convenient access to the information in both data sets while defining lesions and facilitating the discrimination of cerebrospinal fluid from periventricular plaques. The value of total brain LV was calculated by multiplying lesion area by slice thickness and was reproducible to about 5% in serial measurements.
On T1-weighted MR images, normalized volumes of the whole of the brain parenchyma were measured using a method for brain volume measurement (the cross-sectional version of the SIENA software17 [SIENAX]) (Figure 1). SIENAX uses a method to extract the brain and skull from the MR images, as previously described.23 A tissue segmentation program24 is then used to segment the extracted brain image into brain tissue, cerebrospinal fluid, and background, yielding an estimate of total brain tissue volume. The original MR images are registered to a canonical image in a standardized space (using the skull image to provide the scaling cue), a procedure that provides a spatial normalization factor for each subject. The estimate of brain tissue volume for a subject is then multiplied by the normalization factor to yield the normalized brain volume (NBV).
We compared measures relative to the whole RRMS patient group with those of the HC group using the nonparametric Mann-Whitney test. Differences among RRMS patients with different ApoE genotypes and HCs were assessed using analysis of variance followed by pairwise post-hoc comparison using the Tukey Honestly Significant Difference procedure to account for multiple comparisons. The same statistical procedure was also used after grouping the RRMS patients with different ApoE genotypes by short disease duration (<3 years) and absence of clinical disability (EDSS score, <2). Values were considered significant at the level of .05 or less. Unless otherwise indicated, data are expressed as mean ± SD.
In agreement with the results of previous studies,17,19,25 mean NBV values of the whole group of RRMS patients were significantly lower than those of the HC group (1574 ± 65 cm3 [RRMS group] vs 1610 ± 37 cm3[HCs]; P = .01) (Figure 1).
In the group of 76 RRMS patients, we identified 18 carriers of the ϵ4 allele (14 with the genotype ApoE ϵ3/ϵ4, 3 with the genotype ApoE ϵ4/ϵ4, and 1 with the genotype ApoE ϵ2/ϵ4) and 58 subjects without the ϵ4 allele (10 with the genotype ApoE ϵ2/ϵ3 and 48 with the genotype ApoE ϵ3/ϵ3). The 2 patient groups showed no significant differences in age, age at disease onset, disease duration, and EDSS score (Table 1).
In the RRMS patients grouped by their ApoE genotype, we found that MS patients with the ϵ4 allele showed significantly (P = .01) lower values of NBV than HCs and patients without the ϵ4 allele (1553 ± 50 cm3 [RRMS ϵ4-positive group] vs 1581 ± 68cm3 [RRMS ϵ4-negative group]vs 1610 ± 37 cm3 [HCs]) (Figure 2). In contrast, the 2 groups of MS patients with different ApoE genotypes did not show significant differences in values of total brain T2-weighted LV (3.6 ± 2.8 cm3 [RRMS ϵ4-positive group] vs 4.4 ± 3.8 cm3[RRMS ϵ4-negative group]; P = .54).
Similar analyses were also performed selecting a subgroup of MS patients with very short disease duration (<3 years) and absence of clinical disability (EDSS, <2) from the whole patient group. When patients with different ApoE genotypes of these subgroups (30 subjects, 8 with the ϵ4 allele and 22 without the ϵ4 allele) were compared with 12 age-matched HCs (mean age, 30 years in each group), NBV values were still significantly lower in RRMS patients with the ϵ4 allele than in MS patients without the ϵ4 allele (P = .02) and HCs (P = .007) (1561 ± 41 cm3 [RRMS ϵ4-positive group] vs 1609 ± 56 cm3 [RRMS ϵ4-negative group] vs 1618 ± 42 cm3 [HCs]) (Figure 3).
Although there is lack of general consensus on this matter,8 several recent clinical studies have reported on the disease-modifying effect of the ApoE genotype in MS.6- 8,26,27 This is based on the frequent observation that the presence the ϵ4 allele in the genotype of MS patients is associated with greater severity and faster progression of the disease. In agreement with this are recent MR studies showing increased "black holes"9 and decreased values of N-acetylaspartate10 (both reliable surrogate markers of tissue/axonal damage or loss) in brains of MS patients who carry the ϵ4 allele compared with those who do not have the ϵ4 allele in their ApoE genotype. The present study, by showing that cerebral volumes are significantly lower in RRMS patients with the ϵ4 allele than in those without the ϵ4 allele and HCs, adds to previous work reporting further evidence that links ApoE genotypes to the degree of tissue damage in brains or MS patients. This suggests that the ϵ4 allele may have a real, deleterious effect on MS prognosis.
The differences in brain tissue damage between MS patients who carry and do not carry the ϵ4 allele do not coincide, in our study, with significant differences in age at disease onset, disease duration, and clinical disability between the 2 patient groups. As reported by previous studies, MS patients with the ApoE ϵ4 allele do not seem to have a higher risk for development of the disease than those carrying the ApoE ϵ3 allele,4 and the deleterious effect of the ApoE ϵ4 allele seems to become evident later, when the cerebral abnormality develops and impaired repair efficiency results in more pronounced loss of brain function.4 However, on the basis of our results and those of previous9 MR studies, we can speculate that in ϵ4 carriers with MS, the potential functional impairment due to a more prominent tissue destruction is not initially clinically evident. Instead, it is fully compensated by the adaptive mechanisms of reorganization. Clinical differences between ϵ4 and ϵ3 carriers could become evident later, when compensatory resources of the central nervous system are exhausted.28,29 Results indicating that neuronal plasticity is more impaired in transgenic mice possessing the human ApoE ϵ4 allele than in those having the ApoE ϵ3 allele30 seem to confirm this hypothesis. On the other hand, by restricting the analysis to RRMS patients with minimal disease duration (<3 years) and absence of clinical disability (EDSS score, <2), our data suggest that, although clinical differences are still absent, the ϵ4 allele links with more pronounced brain tissue damage, even at the very early stages of MS.
Cerebral atrophy occurs in most neurodegenerative disorders with a mechanism that is mostly driven by neuronal and axonal loss.31 The pathological correlates of brain atrophy in MS imply the presence of inflammation and focal demyelination and the consequent loss of brain components such as myelin and glial cells.18,19 A relevant component of atrophy, however, appears to be independent of focal demyelinating lesions and is probably the expression of the widespread neuroaxonal damage due to inflammation and/or primitive neurodegeneration.18,25,32,33 This might explain why, in our study, NBV decreases in ApoE ϵ4 carriers were not associated with more pronounced total brain T2-weighted LV and why, in a previous MS study,9 the LV was not significantly higher in ϵ4 carriers than in ϵ3 carriers.
Although our study has limitations owing to the cross-sectional design and the relatively small sample of patients, this is, to our knowledge, the first study attempting quantitative measurements of cerebral volumes in MS patients with different ApoE genotypes. Similar measurements exist in patients with Alzheimer disease with discrepant results on differences in total brain volumes between patients with different ApoE genotypes,34- 37 but with homogeneous findings on differences in brain regional volumes (ie, hippocampus and amygdala) between ϵ3 and ϵ4 carriers.35,37- 39 These results, which have been confirmed by a recent postmortem study,40 suggest a regional specificity of the effect of the ApoE ϵ4 allele in the brains of patients with Alzheimer disease. This regional specificity is very unlikely in MS (where most of the normal-appearing brain is affected by the disease32,33), making a marker such as the NBV more appropriate than other regional or tissue specific markers to assess evidence of tissue damage.
Our data add to previous evidence of the adverse effect of ApoE ϵ4 genotype on MS patients showing the presence of more pronounced structural brain damage in RRMS patients who carry the ϵ4 allele than in those who do not carry this allele. This structural brain damage (which is present in MS patients with the ApoE ϵ4 allele, even at the earliest disease stages) might be clinically silent while mechanisms of adaptation can compensate, but is probably responsible for the faster progression of disability observed in MS patients carrying the ApoE ϵ4 allele when central nervous system compensatory resources are exhausted. Characterization of the ApoE genotype can be helpful to explain some of the variance in disease progression in clinical studies and, perhaps, should be considered before deciding which patients should begin early treatment in MS.
Corresponding author and reprints: Nicola De Stefano, MD, Department of Neurological and Behavioral Sciences, Viale Bracci 2, 53100, Siena, Italy (e-mail: firstname.lastname@example.org).
Accepted for publication December 3, 2003.
Author contributions: Study concept and design (Drs De Stefano, Bartolozzi, Guidi, Sorbi, and Amato); acquisition of data (Drs De Stefano, Bartolozzi, Nacmias, Zipoli, Mortilla, Guidi, Siracusa, and Federico); analysis and interpretation of data (Drs De Stefano and Zipoli); drafting of the manuscript (Drs De Stefano, Nacmias, Zipoli, Mortilla, and Siracusa); critical revision of the manuscript for important intellectual content (Drs De Stefano, Bartolozzi, Guidi, Sorbi, Federico, and Amato); obtained funding (Dr Federico); administrative, technical, and material support (Drs Bartolozzi, Mortilla, Guidi, and Federico); study supervision (Drs De Stefano, Sorbi, and Amato).
This study was supported by grants from the University of Siena (Progetto Ateneo di Ricerca) and the Associazione Italiana Sclerosi Multipla (Dr De Stefano), and by a grant from MURST, Rome, Italy (Dr Federico).
We thank Sridar Narayanan for thoughtful discussion.