- Email: [email protected]
Auditory brainstem response findings and peripheral auditory sensitivity in adrenoleukodystrophy
Journal of the Neurological Sciences 247 (2006) 130 – 137 www.elsevier.com/locate/jns
Auditory brainstem response findings and peripheral auditory sensitivity in adrenoleukodystrophy J.P. Pillion a,b , S. Kharkar c , A. Mahmood a , H. Moser a,c,⁎, H. Shimizu d a
b
Kennedy Krieger Institute, United States Department of Physical Medicine and Rehabilitation, Johns Hopkins University School of Medicine, United States c Department of Neurology, Johns Hopkins University School of Medicine, United States d Department of Otolaryngology, Johns Hopkins University School of Medicine, United States Received 18 November 2005; received in revised form 6 April 2006; accepted 10 April 2006 Available online 5 June 2006
Abstract Measurements of the auditory brainstem response (ABR) were obtained in 96 individuals with X-linked adrenoleukodystrophy (X-ALD). The patients were divided into five diagnostic groups on the basis of neurologic diagnosis. The five groups were cerebral childhood and adolescent, pure adrenomyeloneuropathy (pure AMN), adrenomyeloneuropathy cerebral (AMN cerebral), Addison's only and symptomatic female heterozygotes. Results indicated the presence of marked ABR abnormalities for all groups most frequently involving Wave V, followed by Wave III and Wave I. Abnormalities of all interpeak latency intervals (i.e., I–III, III–V and I–V) were observed for all groups. ABR abnormalities were most frequently seen in the AMN-cerebral and pure AMN groups but were also common in the symptomatic female heterozygote group. The ABRs in the cerebral childhood and adolescent group were the least impaired of the five groups examined. Age was found to be a significant independent predictor of bilateral ABR abnormalities but VLCFA levels, MRI Loes score, and duration of symptoms were not found to be independent predictors of bilateral ABR abnormalities after adjusting for ALD phenotype. Patients with AMN were significantly more likely to have bilateral ABR abnormalities than the cerebral childhood and adolescent group after adjusting for age, duration of symptoms, EDSS score, VLCFA levels and MRI Loes scores. The prevalence of peripheral hearing loss was not found to exceed that present in age and sex matched normal control groups derived from the NHANES (1999–2000), indicating a lack of association between peripheral hearing loss and X-linked adrenoleukodystrophy. It was concluded that: (1) auditory sensitivity in X-ALD is not significantly impaired; (2) ABR abnormalities are a frequent finding and may be caused by abnormalities of fiber tracts in the region of the lateral lemniscus and inferior colliculus; and, (3) the abnormalities progress slowly and appear to be associated mainly with the AMN phenotype. © 2006 Published by Elsevier B.V. Keywords: Addison's disease; Adrenoleukodystrophy; Adrenomyeloneuropathy; Auditory brainstem response; Interpeak latency intervals; MRI Loes score; EDSS; Very long chain fatty acid
1. Introduction X-linked adrenoleukodystrophy (X-ALD) is caused by a defect in the gene ABCD1 which maps to X-q28 and codes for ALDP, a peroxisomal membrane protein [1]. More than ⁎ Corresponding author. Department of Neurogenetics, Kennedy Krieger Institute, 707 North Broadway, Baltimore, MD 21205, United States. Tel.: +1 443 923 2750; fax: +1 443 923 2775. E-mail address: [email protected] (H. Moser). 0022-510X/$ - see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.jns.2006.04.001
500 different ABCD1 mutations have been identified in XALD patients and are updated in the website (http://www. x-ald.nl). The nature of the mutation does not correlate with phenotypic expression [2]. The accumulation of saturated very long chain fatty acids (VLCFA), such as hexacosanoic acid (C26:0), in the cerebral white matter and adrenal cortex [3] and other tissues and plasma, is the main biochemical abnormality in X-ALD. Demonstration of increased VLCFA levels in plasma is the most commonly used diagnostic assay [4].
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
X-ALD has four main phenotypes. Phenotypes 1 and 2 involve the nervous system. More than 70% of male X-ALD patients have primary adrenocortical insufficiency. 1. Cerebral phenotypes occur most commonly in childhood, but may also occur in adolescence and adulthood, and are referred to as the childhood cerebral, adolescent cerebral, and the relatively rare adult cerebral phenotypes. They are rapidly progressive, are associated with an inflammatory demyelination [5] which involves the action of effector molecules [6] and may involve auto-immune mechanisms [7]. Characteristic brain MRI abnormalities [8] with several distinct patterns [9] have been reported. A method to score the severity of the MRI abnormality has been developed by Loes et al. [10]. 2. Adrenomyeloneuropathy (AMN). The basic lesion is a non-inflammatory distal axonopathy [11]. AMN is slowly progressive and involves adults mainly. In “pure” AMN abnormalities are present mainly in the spinal cord, peripheral nerves, and the corticospinal tract in the internal capsule. The degree of severity of pure AMN is assessed by the Kurtzke expanded disability scale [12] and more precisely by quantitative sensorimotor tests [13]. A scale specific for adult X-ALD has been devised [14]. Approximately 30% of AMN patients also have or develop varying degrees of inflammatory cerebral involvement. Such patients are referred to as having the AMN cerebral phenotype. Up to 50% of women who are heterozygous for X-ALD develop a somewhat milder variant of pure AMN in middle age or later [15]. The conventional MRI in pure AMN is often normal but may show changes in the internal capsule [9,16], but Mass Spectrometry [17] and Diffusion Tensor Imaging [18] show axonal pathology. Spinal cord imaging in AMN shows non-specific atrophy, which can be quantitated by Magnetization Transfer MRI [19]. There was correlation between the severity of neuroradiological abnormalities and the functional deficits assessed by quantitative sensorimotor tests or Kurtzke disability scale. Somatosensory evoked responses are delayed in AMN [20]. 3. The “Addison only” phenotype. Here there is primary adrenocortical insufficiency without clinical or radiological evidence of neurological involvement. Most patients with this phenotype later also develop neurological disabilities. 4. The asymptomatic phenotype. These individuals are asymptomatic and have a normal conventional MRI. They are identified by screening relatives of symptomatic X-ALD patients with the plasma VLCFA assay [21]. Virtually all later develop endocrine or neurological abnormalities. Therapy of X-ALD varies with the phenotype [22]. Adrenal steroid replacement therapy is mandatory for all patients with adrenal insufficiency. Dietary therapy with “Lorenzo's oil” is recommended for asymptomatic boys who
131
have a normal brain MRI and are less than 8 years old [23] and is undergoing a placebo controlled trial for in men and women with pure AMN. Hematopoietic stem cell transplant (HCT) is beneficial in patients in the early phases of the childhood and adolescent cerebral phenotypes [24]. It is not known whether HCT benefits patients with pure AMN. Auditory brainstem (ABR) abnormalities in patients with X-ALD have been reported in several previous studies each of which involved 2 to 9 patients with varying phenotypes. The most consistent finding was increased latency of waves III and or V and prolonged interwave intervals, particularly I–III, III–V, and I–V [25–29]. Auditory sensitivity has not been evaluated in detail. In one series it was impaired in two of nine patients [28], in another in one out of three [27]. Studies in women heterozygous for X-ALD showed increased I–V intervals, or increased I–III, III–V and I–V intervals in a total of five women who were obligate heterozygotes on the basis of pedigree analysis [29,30]. Taken together these studies indicate that auditory pathways are involved in X-ALD but are insufficient to assess the frequency of hearing loss, rate of progression, correlation with phenotype, radiological and biochemical abnormalities, and pathogenetic mechanisms. We now report audiological findings in 102 patients with various forms of X-ALD who have been examined at the Kennedy Krieger Institute, which has made it possible to address these issues. 2. Subjects Subjects were 102 individuals with X-linked adrenoleukodystrophy (X-ALD) who ranged in age from 5.32 to 73.44 years. The mean age was 31.8 years, S.D. 15.58. The diagnosis of X-ALD was established on the basis of clinical findings and plasma VLCFA assay [4]. The mean plasma hexacosanoic acid level was 0.565 μg/ml, S.D. 0.331 (control 0.23 ± 0.09). It was significantly increased in all subjects, except in some of the heterozygotes, where normal VLCFA have been shown to occur in approximately 20% [4]. The women with normal or equivocal plasma VLCFA levels were shown to be obligate heterozygotes on the basis of pedigree analysis or mutation analysis [31]. Table 2 shows phenotype distribution, utilizing previously described criteria [32] which are summarized in the Introduction. The AMN group was subclassified into the “pure” AMN group which showed clinical evidence of spinal cord and peripheral nerve involvement, but no clinical or radiological [9,16] inflammatory cerebral demyelination, and the “AMN cerebral” group showed varying degrees of cerebral involvement. The scale devised by Loes et al. [10] was used to grade the severity of brain MRI abnormality. With this system a score of < 1 is classified as normal, and 34 is the maximally abnormal. Abnormal scores were present in the AMN cerebral and the childhood and adolescent cerebral group. The mean Loes score in the entire 102
132
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
group of subjects was 4.065 (S.D. 5.7). The 10-point Kurtzke Expanded Disability Status Scale (EDSS) [12] was used to assess the degree of functional disability. The mean score was 3.87 (S.D. 2.02), range 0 to 8. In 29/97 (29.9%) of patients the EDSS score was 6 or greater, indicative of a severe disability. At baseline, audiometric data including pure tone audiograms and word recognition measures were available for 90 patients and ABR data was available for 96 patients. Only the ABR measurements were repeated at subsequent visits. 48 of the 96 patients returned for the second visit, and 29 patients returned for the third visit. Subjects ranged in age from 5.32 to 73.44 years (mean age = 31.8, S.D. = 15.58, range = 5.32– 73.44). At the second test session, patients ranged in age from 11.18 to 68.2 years (mean age = 37.51; S.D. = 14.29). For the third test session, patients ranged in age from 19.94 to 62.65 years (mean age = 38.64; S.D. = 12.52). The interval between the initial and second test sessions ranged from 0.46 years to 4.66 years (mean = 0.77; S.D. = 0.67). The interval between the second and third test sessions ranged from 0.16 to 1.10 years (mean = 0.56; S.D. = 0.19). Subjects were placed in one of five diagnostic categories based upon established criteria [33], as shown in Table 1. The diagnostic categories and subject characteristics for the total sample and each of the five groups are shown in Table 1. All aspects of the study received prior approval by the Johns Hopkins University School of Medicine Institutional Review Board. Informed consent was received from each patient. In order to evaluate the significance of any observed hearing loss in the five clinical groups, age and sex matched normative data was derived by random selection from the National Health and Nutrition Examination Survey (1999–
2000). The cases were stratified into seven age groups. Four control cases were selected for each clinical case. The control group was selected for each phenotype separately because the phenotype groups differed significantly in age distribution. 3. Procedures 3.1. Audiometric and ABR measurements Patients were administered an audiological test battery that included measurement of air and bone conduction thresholds, speech reception thresholds, word recognition testing, tympanometry and the auditory brainstem response (ABR). Audiometric data were obtained utilizing standard audiometric techniques. ABRs were obtained for alternating polarity clicks presented at 70 dB nHL. The forehead electrode served as the active electrode and the ipsilateral mastoid as reference with the contralateral mastoid electrode as ground. Clicks were presented at a rate of 19/s with a 15 ms recording window and a passband of 30–3000 Hz. At least two recordings were made as each presentation level. The normality of ABR peaks and ABR interpeak latency intervals (IPLIs) was judged on the basis of the normative data shown below: I: <1.88 ms III: < 4.06 ms V: < 6.02 ms I–III: <2.35 ms III–V: < 2.22 ms I–V: <4.37 ms
Table 1 Percentage of ears in patients with an abnormally increased hearing threshold (>20 dB) in the five categories of ALD with hearing loss for pure tones at 250, 500, 1000, 2000, 4000 and 8000
Number of patient ears Mean, S.D., and range of the age in years of ALD patients on the date of audiometry Mean, S.D., and range of the age in years of normative groups 250 Hz Patients Controls 500 Hz Patients Controls 1000 Hz Patients Controls 2000 Hz Patients Controls 4000 Hz Patients Controls 8000 Hz Patients Controls
Cerebral (childhood + adolescent)
Pure AMN
AMN cerebral
Symptomatic female heterozygotes
Addison's only
28 15.3 S.D. = 5.1 (5.3 to 23.08)
62 35.9 S.D. = 11.7 (12.5 to 61.6)
48 33.9 S.D. = 11.4 (18.0–59.6)
34 48.6 S.D. = 11.1 (12.5 to 61.6)
8 12.6 S.D. = 3.6 (10.2 to 17.9)
25.1 S.D. = 2.8 (20 to 29) 7.1 – 0 6.2 3.6 3.6 0 8.1 10.7 17.1 25 16.2
35.3 S.D. = 10.8 (20 to 69) 14.5 – 9.6 7.0 4.8 7.0 6.4 13.6 22.6 33.3 27.4 31.6
33.6 S.D. = 10.4 (20 to 59) 20.8 – 4.2 7.9 2.1 6.9 6.3 13.2 33.3 23.2 31.2 24.9
48.9 S.D. = 9.4 (30 to 67) 8.8 – 11.7 16.4 20.5 9.0 23.5 13.4 32.3 22.4 32.3 30.6
26.3 S.D. = 2.8 (20 to 29) 0 – 0 6.2 12.5 0 0 3.1 0 15.6 25 12.5
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
If a peak could not be identified or was not repeatable upon successive recordings, the peak was classified as abnormal. When the interaural latency difference was examined as a measure in an analysis separate from the overall analysis described above, an interaural latency difference greater than 0.3 ms was considered abnormal. Sample ABR waveforms representative of a range of waveform morphology are shown in Fig. 1.
133
ABR testing respectively. The EDSS was calculated on the corresponding dates of the ABR test session. The EDSS ranged from 0 to 8 with a score of 6 or greater indicative of severe disability. Duration of symptoms was determined by chart review and was based upon the date of onset of neurological symptoms such as alterations in memory, difficulty in running, walking or climbing. 3.3. Statistical analysis
3.2. MRI, EDSS and VLFCA measurements Subjects were examined using a high-field 1.5-T signal MRI system. The standard protocol included sagittal images (repetition time/echo time [TR/TE], 600/20 ms) and axial proton-weighted images and T2 images (TR/TE, 3000/30 ms and 3000/100 ms, respectively) with slice thicknesses of 5 mm and a 2.5-mm intersection gap. MRI images were evaluated by radiologists not involved in the patients' care and were scored by using a system designed specifically for the assessment of adrenoleukodystrophy patients [10], with a range of scores of 0 through 34 based on the presence of white matter abnormalities in key brain regions. We have demonstrated inter-rater reliability coefficients for this system between 0.87 and 0.93. MRI Loes scores less than 1.5 were classified as normal whereas scores of 1.5 or greater were classified as abnormal. MRI data were obtained on 91 of the 102 patients with ALD. The levels of VLCFA in plasma and MRI Loes scores were measured on an average 6 months after (mean = 0.57 years, S.D. = 0.88 years) and 8.5 months before (mean = − 0.7 years, S.D. = 1.06 years) the
The Fisher's exact test was used to analyze the frequency of hearing loss and ABR abnormalities across the different ALD phenotypes. Regression analysis was used to analyze the association between bilaterally prolonged I–V interpeak latency intervals and ALD phenotype. Unadjusted odds ratios were calculated using univariate logistic regression. Multivariate logistic regression was performed to adjust for potential confounding by age, duration of symptoms, MRI Loes scores, EDSS scores and very long chain fatty acid (VLCFA) concentration in plasma. All possible interaction terms were tested and were found to be non-significant. The final model was selected based on odds ratio testing [34]. 4. Results 4.1. Auditory sensitivity Table 1 shows instances of mild to severe hearing loss particularly in the older subjects with the highest prevalence in the 4000–8000 Hz frequency region. For several ears, a precipitous audiometric configuration was observed at 6000–8000 Hz. The childhood cerebral and Addison only subjects, who were younger, had the highest percentage of ears with normal peripheral auditory sensitivity. Hearing loss was less prevalent from 500 to 2000 Hz for all groups. The significance of the observed hearing losses was evaluated by comparing the obtained hearing losses in the patient groups to normative data obtained in the NHANES (1999–2000). Comparisons could not be undertaken at 250 Hz because that frequency was not included in the NHANES (1999–2000). No significant difference was found in the percentage of ears with hearing loss for any of the five patient groups at any audiometric frequency when compared to the percentage of hearing loss present in age and sex matched controls. We concluded that auditory sensitivity in X-ALD is not significantly impaired. 4.2. Auditory brainstem responses (ABR) at baseline and follow-up
Fig. 1. Auditory brainstem response waveforms depicting, (A) an AMN patient with good response morphology and repeatability but prolonged I– III, III–V and I–V interpeak latency intervals; (B) an AMN patient with more marked prolongation of the I–V interpeak latency interval and an absent peak III. The summed response to three replications was utilized due to poor peak repeatability; and, (C) a patient with AMN with normal peripheral auditory sensitivity but no clearly repeatable ABR peaks.
Fig. 1 illustrates the waveform abnormalities in patients with AMN. Table 2 shows that ABR abnormalities were most prevalent for Wave V, followed by Wave III and Wave I. ABR abnormalities were present for the wave I–III, III–V and I–V interpeak latency intervals. Table 3 compares the frequency
134
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
Table 2 Number and percentage of patients in the five categories of ALD with normal or abnormal Wave I, III, and V, interaural Wave V latency difference or Wave I–III, III–V, and I–V interpeak latency intervals (total number = 96)
Number of patients (N) Mean, S.D. and range of age in years at audiometry Wave I Bilateral Unilateral Normal Wave III Bilateral Unilateral Normal Wave V Bilateral Unilateral Normal Interaural Wave V latency difference > 0.3 Interpeak latency I–III Bilateral Unilateral Normal Interpeak latency III–V Bilateral Unilateral Normal Interpeak latency I–V Bilateral Unilateral Normal Any ABR abnormality (abnormality in any of the ABR waves, either unilaterally or bilaterally)
Cerebral (childhood + adolescent)
Pure AMN
AMN cerebral
Symptomatic female heterozygote
Addison's only
20 13.9 S.D. = 4.8 (5.3 to 23.1)
31 35.8 S.D. = 12.6 (12.5 to 61.6)
25 34.3 S.D. = 11.3 (18.0 to 59.6)
15 47.4 S.D. = 9.6 (30.6 to 67.6)
5 11.5 S.D. = 3.9 (7.2 to 17.9)
5% 15% 80%
6.4% 25.8% 67.8% (NS)
20% 16% 64% (NS)
0% 6.7% 93.3% (NS)
0% 0% 5% (NS)
20% 10% 70%
71% 9.7% 19.3% ⁎⁎
80% 12% 8%⁎⁎
26.7% 40.0% 33.3% ⁎
20% 20% 60% (NS)
20% 5% 75% 20%
87.1% 6.4% 6.4%⁎⁎ 38.7%
96% 4% 0%⁎⁎ 52%
53.3% 20.0% 26.7%⁎⁎ 26.7%
20% 20% 60% (NS) 0%
20% 25% 55%
77.4% 12.9% 9.7%⁎⁎
92% 4% 4%⁎⁎
33.3% 46.7% 20.0%⁎
40% 20% 40% (NS)
15% 5% 80%
74.2% 19.4% 6.4%⁎⁎
96% 4% 0%⁎⁎
33.3% 20.0% 46.7% (NS)
20% 0% 80% (NS)
15% 15% 70% 50%
90.3% 3.2% 6.4% 93.5%⁎⁎
96% 4% 0% 100%⁎⁎
53.3% 33.3% 13.3% 100%⁎⁎
20% 20% 60% (NS) 60% (NS)
The cerebral (childhood–adolescent) category was taken as the reference category. NS—not significant; p > 0.05 by Fisher's exact test. ⁎ p < 0.05 by Fisher's exact test. ⁎⁎ p < 0.01 by Fisher's exact test.
of abnormal waves in 48 patients who were tested a mean of 0.46 years after the baseline evaluation and Table 4 in 29 patients who were also tested a third time. The mean interval between the second and the third tests was 0.56 years. Table 5 shows the phenotype distribution of the patients who were tested serially and Table 6 the changes in frequency in abnormalities over time. The ABR abnormalities did not appear to increase systematically for any waves interpeak latency intervals or in the interaural latency difference from session I and session II and session II to session III. 4.3. Correlation between ABR abnormalities and phenotype ABR abnormalities were most prevalent in males with cerebral and pure AMN and in the heterozygotes, and least prevalent in the childhood and adolescent cerebral and Addison disease categories of patients. The percentage of individuals with any ABR abnormality was significantly smaller in the childhood and adolescent cerebral phenotypes
as compared to the pure AMN, AMN cerebral and symptomatic heterozygote phenotypes (p < 0.01). No significant differences were present in the percentage of individuals with any ABR abnormality between the Addison's only and the childhood and adolescent cerebral phenotypes (p > 0.999). 4.4. Statistical analyses Table 7 summarizes correlations between ABR abnormalities and a variety of factors. Based on univariate logistic regression, ALD phenotype, age (odds ratio = 1.1, p < 0.001), duration of symptoms (odds ratio = 1.3, p = 0.002), EDSS (odds ratio = 1.4, p = 0.015) and MRI Loes scores (odds ratio = 0.98, p = 0.003) were found to be statistically significant predictors of bilaterally abnormal Wave I–Wave V interpeak latency while VLCFA concentration in plasma did not reach statistical significance (odds ratio = 0.3, p = 0.18).
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137 Table 3 Percentage of patients with abnormal waves in sessions 1 and 2 (number of patients who were present for both visits = 48) Visit 1
Visit 2
Wave I Bilateral Unilateral
8.3% 16.7%
12.5% 14.6%
Wave III Bilateral Unilateral
62.5% 20.8%
62.5% 20.8%
Wave V Bilateral Unilateral Interaural Wave V latency difference > 0.3
83.3% 8.3% 43.7%
83.3% 2.1% 43.7%
Interpeak latency I–III Bilateral Unilateral
79.2% 18.7%
81.2% 12.5%
Interpeak latency III–V Bilateral Unilateral
79.2% 8.3%
70.8% 12.5%
Interpeak latency I–V Bilateral Unilateral
87.5% 6.2%
83.3% 6.2%
With respect to the multivariate analyses, the difference in odds of having a bilaterally abnormal Wave I–Wave V interpeak latency interval between the cerebral (childhood and adolescent) phenotype and pure AMN was not found to Table 4 Percentage of patients with abnormal waves in sessions 1 and 3 (number of patients who were present for both visits = 29) Visit 1
Visit 3
Wave I Bilateral Unilateral
13.8% 17.2%
13.8% 24.1%
Wave III Bilateral Unilateral
72.4% 17.2%
72.4% 20.7%
Wave V Bilateral Unilateral Interaural Wave V latency difference > 0.3
82.8% 13.8% 41.4%
86.2% 6.9% 41.4%
Interpeak latency I–III Bilateral Unilateral
86.2% 13.8%
86.2% 13.8%
Interpeak latency III–V Bilateral Unilateral
82.8% 6.9%
79.3% 3.4%
Interpeak latency I–V Bilateral Unilateral
93.1% 6.9%
89.7% 6.9%
135
Table 5 Phenotype distribution of patients in each of the three visits Phenotype
Visit 1
Cerebral (childhood + adolescent) Pure AMN AMN cerebral Symptomatic female heterozygotes Addison's disease Total
20 (20.8%)
Visit 2 5 (10.4%)
Visit 3 3 (10.3%)
31 (32.3%) 25 (26%) 15 (15.6%)
18 (37.5%) 16 (33.3%) 8 (16.7%)
11 (37.9%) 11 (37.9%) 4 (13.8%)
5 (5.2%) 96
1 (2.1%) 48
0 29
be statistically significant (odds ratio = 2.8, p = 0.57) after adjusting for age, duration of symptoms, VLCFA levels, EDSS scores and MRI Loes scores. Also, the difference in odds between the cerebral (childhood and adolescent) phenotype and the symptomatic heterozygote females failed to reach statistical significance (odds ratio = 0.01, p = 0.07) after adjusting for these covariates. However, a significant difference (odds ratio = 21.2, p = 0.049) was found between the cerebral (childhood and adolescent) phenotype and the AMN-cerebral group after adjusting for the covariates listed above. Age was also found to be significantly independent predictor of a bilaterally abnormal Wave I–Wave V interpeak latency interval after adjusting for ALD phenotype, duration of symptoms, VLCFA levels, EDSS scores and MRI Loes score (odds ratio = 1.1, p < 0.001). MRI Loes score (odds = 0.8, p = 0.31), EDSS scores (odds ratio = 1.8, p = 0.09), duration of symptoms (odds ratio = 1.05, p = 0.74), and VLCFA concentration in plasma (odds = 1.2, p = 0.94) were not found to be statistically significant independent predictors in the final multivariate regression model. 5. Discussion Our studies of audiological function in 102 subjects with X-ALD indicate that auditory sensitivity is not significantly impaired, but confirm that auditory brainstem abnormalities (ABR) are present in a large proportion. There were significant correlations between ABR abnormalities and phenotype, which provide new insights into pathogenetic mechanisms. The ABR has been utilized for many years in neurodiagnostics because it reflects activity from the auditory nerve and auditory pathways in the brainstem. On the basis of Table 6 Number and percentage of people in the five categories of ALD having abnormal Wave V latencies (defined as an interaural difference in the latency of Wave V of more than 0.3 ms)
Visit 1 Visit 2 Visit 3
Cerebral (childhood and adolescent)
Pure AMN
AMN cerebral
Symptomatic female heterozygote
Addison's only
20% 40% 33.3%
38.7% 38.9% 54.5%
52% 43.7% 36.3%
26.7% 25% 25%
0% 0% –
136
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
Table 7 Predictors of bilaterally prolonged I–V interpeak latency intervals
Phenotype Cerebral childhood and adolescent Pure AMN AMN cerebral Symptomatic female heterozygotes Addison's only Age MRI Loes score Duration of symptoms EDSS Score Very long chain fatty acid (VLCFA) conc. in plasma
Univariate analysis
Multivariate analysis
OR
p-value
OR
p-value
Reference category 56 154 8.4
–
–
<0.001 <0.001 0.03
Reference category 2.8 21.2 0.01
0.57 0.049 0.07
⁎ 1.1 0.9 1.3 1.4 0.3
⁎ <0.001 0.003 0.002 0.015 0.18
⁎ 1.2 0.8 1.05 1.8 1.2
⁎ 0.01 0.31 0.74 0.09 0.94
⁎Only two patients with the “Addison's only” phenotype had complete data and hence this category could not be included in the regression model. Both of those patients had normal I–V interpeak latency intervals.
measurements obtained during operations for cranial nerve disorders in humans [35], it has been determined that the neural generator for Waves I and II in humans is the auditory nerve. Abnormalities in these waveforms were relatively uncommon. This, combined with our data that auditory sensitivity is not significantly impaired, indicates that the auditory nerve and cochlear nucleus do not appear to be involved in X-ALD. It is more difficult to attribute specific generators to the later peaks of the ABR due to the extent of parallel processing in the auditory pathway at brainstem levels; the peaks of waves subsequent to Wave II have multiple sources underlying their generation [36]. While there is evidence on the basis of intraoperative recordings that the neural generator for Wave III of the human ABR is the cochlear nucleus [37], the generation of Wave III may be more complex than originally supposed. The contralateral cochlear nucleus [38] as well as the most proximal portion of the auditory nerve may also make a contribution to the generation of Wave III [39]. Several neural sources contribute as the generators for Wave V [40]. The most positive peak of Wave V is probably generated at the termination of the fiber tract of the lateral lemniscus, whereas the following negative trough in conventionally recorded ABRs is generated by slow dendritic potentials in the inferior colliculus [40]. The main contribution to the peak of wave V appears to be from contralateral rather than ipsilateral structures on the basis of intracranial recordings in human subjects [38]. Taken together, these data suggest that the ABR abnormality is caused by lesions in fiber tracts in the lateral lemniscus and inferior colliculus. The present ABR findings can be correlated with current concepts of the pathogenesis of X-ALD. Childhood and adolescent cerebral X-ALD are associated with an intensely inflammatory demyelination that is most severe in the cerebral hemispheres. The thalamic nuclei, basal ganglia and
geniculate bodies are also involved, but the brainstem is relatively spared [5]. Table 2 shows that most ABRs were normal in the majority of boys with these phenotypes. In contrast, nearly all AMN patients show ABR abnormalities. This difference may be partially attributable to age, because there is a significant positive correlation between age and ABR abnormalities, and AMN patients are older. However, comparison of the “pure AMN” and the “cerebral AMN” provides interesting additional information. Although Table 2 shows that cerebral AMN patients are slightly younger and have a slightly higher frequency of ABR abnormalities, these differences are not statistically significant. The basic disease process in pure AMN is a non-inflammatory axonopathy, while inflammatory demyelination is the main abnormality in childhood and adolescent cerebral forms and in the cerebral component of cerebral AMN. As has been reported before, approximately 60% of women heterozygous for XALD had abnormal ABR responses. This is consistent with the report that approximately 50% of carriers develop an AMN-like syndrome in middle age or later [28]. The high frequency of ABR abnormalities in pure AMN, and their relatively low frequency in childhood and adolescent cerebral X-ALD, and the lack of significant difference in their frequency in pure AMN and cerebral AMN, suggest strongly that the ABR abnormalities are due to a non-inflammatory axonopathy, most likely in axonal tracts in the lateral lemniscus and inferior colliculus. Application of novel imaging techniques, such as diffusion tensor imaging combined with three dimensional fiber tracking [18] make it possible to define this more precisely. The likelihood that the ABR abnormalities reflect an axonopathy rather than inflammatory demyelination is also of practical significance. Hematopoietic stem cell transplantation is recommended for X-ALD patients who show early evidence of inflammatory demyelination but at this time it is not recommended for pure AMN [24]. In conclusion we find (1) that auditory sensitivity in XALD is not significantly impaired; (2) ABR abnormalities are a frequent finding and may be caused by abnormalities of fiber tracts in the region of the lateral lemniscus and inferior colliculus; (3) the abnormalities progress slowly and appear to be associated mainly with the AMN phenotype. Acknowledgement Support from the General Clinical Research Center (MO1-RR00052) from the National Center for Research Resources is acknowledged. References [1] Mosser J, Douar AM, Sarde CO, Kioschis P, Feil R, Moser H, et al. Putative X-linked adrenoleukodystrophy gene shares unexpected homology with ABC transporters. Nature 1993;361:726–30. [2] Smith KD, Kemp S, Braiterman LT, Lu JF, Wei HM, Geraghty M, et al. X-linked adrenoleukodystrophy: genes, mutations, and phenotypes. Neurochem Res 1999; 24:521–35.
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137 [3] Igarashi M, Schaumburg HH, Powers J, Kishmoto Y, Kolodny E, Suzuki K. Fatty acid abnormality in adrenoleukodystrophy. J Neurochem 1976;26:851–60. [4] Moser AB, Kreiter N, Bezman L, Lu S, Raymond GV, Naidu S, Moser HW. Plasma very long chain fatty acids in 3000 peroxisome disease patients and 29,000 controls. Ann Neurol 1999;45:100–10. [5] Schaumburg HH, Powers JM, Raine CS, Suzuki K, Richardson Jr EP. Adrenoleukodystrophy. A clinical and pathological study of 17 cases. Arch Neurol 1975;32:577–91. [6] Powers JM, Liu Y, Moser AB, Moser HW. The inflammatory myelinopathy of adrenoleukodystrophy: cells, effector molecules, and pathogenetic implications. J Neuropathol Exp Neurol 1992; 51:630–43. [7] Ito M, Blumberg BM, Mock DJ, Goodman AD, Moser AB, Moser HW, et al. Potential environmental and host participants in the early white matter lesion of adrenoleukodystrophy: morphologic evidence for CD8 cytotoxic T cells, cytolysis of oligodendrocytes, and CD1mediated lipid antigen presentation. J Neuropathol Exp Neurol 2001;60:1004–19. [8] Kumar AJ, Rosenbaum AE, Naidu S, Wenger L, Citrin CM, Lindenberg R, et al. Adrenoleukodystrophy: correlating MR imaging with CT. Radiology 1987;165:497–504. [9] Loes DJ, Fatemi A, Melhem ER, Gupte N, Bezman L, Moser HW, Raymond GV. Analysis of MRI patterns aids prediction of progression in X-linked adrenoleukodystrophy. Neurology 2003;61:369–74. [10] Loes DJ, Hite S, Moser H, Stillman AE, Shapiro E, Lockman L, et al. Adrenoleukodystrophy: a scoring method for brain MR observations. AJNR Am J Neuroradiol 1994; 15:1761–6. [11] Powers JM, DeCiero DP, Ito M, Moser AB, Moser HW. Adrenomyeloneuropathy: a neuropathologic review featuring its noninflammatory myelopathy. J Neuropathol Exp Neurol 2000;59:89–102. [12] Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983;33: 1444–52. [13] Zackowski K, Dubey P, Raymond GV, Mahmood A, Moser AB, Moser HW. Relating sensorimotor function and axonal integrity in adrenomyeloneuropathy. Arch Neurol 2006;63:74–80. [14] Koehler W, Sokolowski P. A new disease-specific scoring system for adult phenotypes of X-linked adrenoleukodystrophy. J Mol Neurosci 1999;13:247–52. [15] Restuccia D, Di Lazzaro V, Valeriani M, Oliviero A, Le Pera D, Colosimo C, et al. Neurophysiological abnormalities in adrenoleukodystrophy carriers. Evidence of different degrees of central nervous system involvement. Brain 1997;120(Pt 7): 1139–1148. [16] Kumar AJ, Kohler W, Kruse B, Naidu S, Bergin A, Edwin D, Moser HW. MR findings in adult-onset adrenoleukodystrophy. AJNR Am J Neuroradiol 1995;16:1227–37. [17] Dubey P, Fatemi A, Barker PB, Degaonkar M, Troeger M, Zackowski K, et al. Spectroscopic evidence of cerebral axonopathy in patients with “pure” adrenomyeloneuropathy. Neurology 2005;64:304–10. [18] Dubey P, Fatemi A, Huang H, Nagae-Poetscher L, Wakana S, Barker PB, et al. Diffusion tensor-based imaging reveals occult abnormalities in adrenomyeloneuropathy. Ann Neurol 2005;58:758–66. [19] Fatemi A, Smith SA, Dubey P, Zackowski KM, Bastian AJ, van Zijl PC, et al. Magnetization transfer MRI demonstrates spinal cord pathology in adrenomyeloneuropathy. Neurology 2005:1739–45. [20] Kaplan PW, Tusa RJ, Rignani J, Moser HW. Somatosensory evoked potentials in adrenomyeloneuropathy. Neurology 1997;48:1662–7.
137
[21] Bezman L, Moser AB, Raymond GV, Rinaldo P, Watkins PA, Smith KD, et al. Adrenoleukodystrophy: incidence, new mutation rate, and results of extended family screening. Ann Neurol 2001;49:512–7. [22] Moser HW, Raymond GV, Dubey P. Adrenoleukodystrophy: new approaches to a neurodegenerative disease. JAMA 2005;294:3131–4. [23] Moser HW, Raymond GV, Lu SE, Muenz LR, Moser AB, Xu J, et al. Follow-up of 89 Lorenzo's oil treated asymptomatic adrenoleukodystrophy patients. Arch Neurol 2005;62:1073–80. [24] Peters C, Charnas LR, Tan Y, Ziegler RS, Shapiro EG, DeFor T, et al. Cerebral X-linked adrenoleukodystrophy: the international hematopoietic cell transplantation experience from 1982 to 1999. Blood 2004;104:881–8. [25] Vercruyssen A, Martin JJ, Mercelis R. Neurophysiological studies in adrenomyeloneuropathy. A report on five cases. J Neurol Sci 1982; 56:327–36. [26] Ochs R, Markand ON, DeMyer WE. Brainstem auditory evoked responses in leukodystrophies. Neurology 1979;29:1089–93. [27] Grimes AM, Elks ML, Grunberger G, Pikus AM. Auditory brainstem responses in adrenomyeloneuropathy. Arch Neurol 1983;40:574–6. [28] Shimizu H, Moser HW, Naidu S. Auditory brainstem response and audiologic findings in adrenoleukodystrophy: its variant and carrier. Otolaryngol Head Neck Surg 1988;98:215–20. [29] Garg BP, Markand ON, DeMyer WE, Warren Jr C. Evoked response studies in patients with adrenoleukodystrophy and heterozygous relatives. Arch Neurol 1983;40:356–9. [30] Moloney JB, Masterson JG. Detection of adrenoleukodystrophy carriers by means of evoked potentials. Lancet 1982;2:852–3. [31] Boehm CD, Cutting GR, Lachtermacher MB, Moser HW, Chong SS. Accurate DNA-based diagnostic and carrier testing for X-linked adrenoleukodystrophy. Mol Genet Metab 1999;66:128–36. [32] Moser HW, Smith KD, Watkins PA, Powers J, Moser AB. X-linked adrenoleukodystrophy. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease. Eighth edition. New York: McGraw Hill; 2001. p. 3257–301. [33] Moser HW, Smith KD, Moser AB. X-linked adrenoleukodystrophy. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular basis of inherited disease. New York: McGraw Hill; 1994. p. 2325–49. [34] Hosmer DW, Lemeshow S. Applied logistic regression. Second edition. New York: John Wiley and Sons, Inc.; 2000. [35] Moller AR, Jannetta PJ, Moller MB. Neural generators of brainstem evoked potentials. Results from human intracranial recordings. Ann Otol Rhinol Laryngol 1981;90:591–6. [36] Moller AR, editor. Interoperative neurophysiologic monitoring. Luxembourg: Harwood Academic Publishers; 1995. p. 66. [37] Moller AR, Jannetta P, Moller MB. Intracranially recorded auditory nerve response in man. New interpretations of BSER. Arch Otolaryngol 1982;108:77–82. [38] Moller AR, Jho HD, Yokota M, Jannetta PJ. Contribution from crossed and uncrossed brainstem structures to the brainstem auditory evoked potentials: a study in humans. Laryngoscope 1995;105:596–605. [39] Moller AR, Jho HD. Compound action potentials recorded from the intracranial portion of the auditory nerve in man: effects of stimulus intensity and polarity. Audiology 1991;30:142–63. [40] Moller AR, Jannetta PJ. Evoked potentials from the inferior colliculus in man. Electroencephalogr Clin Neurophysiol 1982;53:612–20.
Auditory brainstem response findings and peripheral auditory sensitivity in adrenoleukodystrophy J.P. Pillion a,b , S. Kharkar c , A. Mahmood a , H. Moser a,c,⁎, H. Shimizu d a
b
Kennedy Krieger Institute, United States Department of Physical Medicine and Rehabilitation, Johns Hopkins University School of Medicine, United States c Department of Neurology, Johns Hopkins University School of Medicine, United States d Department of Otolaryngology, Johns Hopkins University School of Medicine, United States Received 18 November 2005; received in revised form 6 April 2006; accepted 10 April 2006 Available online 5 June 2006
Abstract Measurements of the auditory brainstem response (ABR) were obtained in 96 individuals with X-linked adrenoleukodystrophy (X-ALD). The patients were divided into five diagnostic groups on the basis of neurologic diagnosis. The five groups were cerebral childhood and adolescent, pure adrenomyeloneuropathy (pure AMN), adrenomyeloneuropathy cerebral (AMN cerebral), Addison's only and symptomatic female heterozygotes. Results indicated the presence of marked ABR abnormalities for all groups most frequently involving Wave V, followed by Wave III and Wave I. Abnormalities of all interpeak latency intervals (i.e., I–III, III–V and I–V) were observed for all groups. ABR abnormalities were most frequently seen in the AMN-cerebral and pure AMN groups but were also common in the symptomatic female heterozygote group. The ABRs in the cerebral childhood and adolescent group were the least impaired of the five groups examined. Age was found to be a significant independent predictor of bilateral ABR abnormalities but VLCFA levels, MRI Loes score, and duration of symptoms were not found to be independent predictors of bilateral ABR abnormalities after adjusting for ALD phenotype. Patients with AMN were significantly more likely to have bilateral ABR abnormalities than the cerebral childhood and adolescent group after adjusting for age, duration of symptoms, EDSS score, VLCFA levels and MRI Loes scores. The prevalence of peripheral hearing loss was not found to exceed that present in age and sex matched normal control groups derived from the NHANES (1999–2000), indicating a lack of association between peripheral hearing loss and X-linked adrenoleukodystrophy. It was concluded that: (1) auditory sensitivity in X-ALD is not significantly impaired; (2) ABR abnormalities are a frequent finding and may be caused by abnormalities of fiber tracts in the region of the lateral lemniscus and inferior colliculus; and, (3) the abnormalities progress slowly and appear to be associated mainly with the AMN phenotype. © 2006 Published by Elsevier B.V. Keywords: Addison's disease; Adrenoleukodystrophy; Adrenomyeloneuropathy; Auditory brainstem response; Interpeak latency intervals; MRI Loes score; EDSS; Very long chain fatty acid
1. Introduction X-linked adrenoleukodystrophy (X-ALD) is caused by a defect in the gene ABCD1 which maps to X-q28 and codes for ALDP, a peroxisomal membrane protein [1]. More than ⁎ Corresponding author. Department of Neurogenetics, Kennedy Krieger Institute, 707 North Broadway, Baltimore, MD 21205, United States. Tel.: +1 443 923 2750; fax: +1 443 923 2775. E-mail address: [email protected] (H. Moser). 0022-510X/$ - see front matter © 2006 Published by Elsevier B.V. doi:10.1016/j.jns.2006.04.001
500 different ABCD1 mutations have been identified in XALD patients and are updated in the website (http://www. x-ald.nl). The nature of the mutation does not correlate with phenotypic expression [2]. The accumulation of saturated very long chain fatty acids (VLCFA), such as hexacosanoic acid (C26:0), in the cerebral white matter and adrenal cortex [3] and other tissues and plasma, is the main biochemical abnormality in X-ALD. Demonstration of increased VLCFA levels in plasma is the most commonly used diagnostic assay [4].
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
X-ALD has four main phenotypes. Phenotypes 1 and 2 involve the nervous system. More than 70% of male X-ALD patients have primary adrenocortical insufficiency. 1. Cerebral phenotypes occur most commonly in childhood, but may also occur in adolescence and adulthood, and are referred to as the childhood cerebral, adolescent cerebral, and the relatively rare adult cerebral phenotypes. They are rapidly progressive, are associated with an inflammatory demyelination [5] which involves the action of effector molecules [6] and may involve auto-immune mechanisms [7]. Characteristic brain MRI abnormalities [8] with several distinct patterns [9] have been reported. A method to score the severity of the MRI abnormality has been developed by Loes et al. [10]. 2. Adrenomyeloneuropathy (AMN). The basic lesion is a non-inflammatory distal axonopathy [11]. AMN is slowly progressive and involves adults mainly. In “pure” AMN abnormalities are present mainly in the spinal cord, peripheral nerves, and the corticospinal tract in the internal capsule. The degree of severity of pure AMN is assessed by the Kurtzke expanded disability scale [12] and more precisely by quantitative sensorimotor tests [13]. A scale specific for adult X-ALD has been devised [14]. Approximately 30% of AMN patients also have or develop varying degrees of inflammatory cerebral involvement. Such patients are referred to as having the AMN cerebral phenotype. Up to 50% of women who are heterozygous for X-ALD develop a somewhat milder variant of pure AMN in middle age or later [15]. The conventional MRI in pure AMN is often normal but may show changes in the internal capsule [9,16], but Mass Spectrometry [17] and Diffusion Tensor Imaging [18] show axonal pathology. Spinal cord imaging in AMN shows non-specific atrophy, which can be quantitated by Magnetization Transfer MRI [19]. There was correlation between the severity of neuroradiological abnormalities and the functional deficits assessed by quantitative sensorimotor tests or Kurtzke disability scale. Somatosensory evoked responses are delayed in AMN [20]. 3. The “Addison only” phenotype. Here there is primary adrenocortical insufficiency without clinical or radiological evidence of neurological involvement. Most patients with this phenotype later also develop neurological disabilities. 4. The asymptomatic phenotype. These individuals are asymptomatic and have a normal conventional MRI. They are identified by screening relatives of symptomatic X-ALD patients with the plasma VLCFA assay [21]. Virtually all later develop endocrine or neurological abnormalities. Therapy of X-ALD varies with the phenotype [22]. Adrenal steroid replacement therapy is mandatory for all patients with adrenal insufficiency. Dietary therapy with “Lorenzo's oil” is recommended for asymptomatic boys who
131
have a normal brain MRI and are less than 8 years old [23] and is undergoing a placebo controlled trial for in men and women with pure AMN. Hematopoietic stem cell transplant (HCT) is beneficial in patients in the early phases of the childhood and adolescent cerebral phenotypes [24]. It is not known whether HCT benefits patients with pure AMN. Auditory brainstem (ABR) abnormalities in patients with X-ALD have been reported in several previous studies each of which involved 2 to 9 patients with varying phenotypes. The most consistent finding was increased latency of waves III and or V and prolonged interwave intervals, particularly I–III, III–V, and I–V [25–29]. Auditory sensitivity has not been evaluated in detail. In one series it was impaired in two of nine patients [28], in another in one out of three [27]. Studies in women heterozygous for X-ALD showed increased I–V intervals, or increased I–III, III–V and I–V intervals in a total of five women who were obligate heterozygotes on the basis of pedigree analysis [29,30]. Taken together these studies indicate that auditory pathways are involved in X-ALD but are insufficient to assess the frequency of hearing loss, rate of progression, correlation with phenotype, radiological and biochemical abnormalities, and pathogenetic mechanisms. We now report audiological findings in 102 patients with various forms of X-ALD who have been examined at the Kennedy Krieger Institute, which has made it possible to address these issues. 2. Subjects Subjects were 102 individuals with X-linked adrenoleukodystrophy (X-ALD) who ranged in age from 5.32 to 73.44 years. The mean age was 31.8 years, S.D. 15.58. The diagnosis of X-ALD was established on the basis of clinical findings and plasma VLCFA assay [4]. The mean plasma hexacosanoic acid level was 0.565 μg/ml, S.D. 0.331 (control 0.23 ± 0.09). It was significantly increased in all subjects, except in some of the heterozygotes, where normal VLCFA have been shown to occur in approximately 20% [4]. The women with normal or equivocal plasma VLCFA levels were shown to be obligate heterozygotes on the basis of pedigree analysis or mutation analysis [31]. Table 2 shows phenotype distribution, utilizing previously described criteria [32] which are summarized in the Introduction. The AMN group was subclassified into the “pure” AMN group which showed clinical evidence of spinal cord and peripheral nerve involvement, but no clinical or radiological [9,16] inflammatory cerebral demyelination, and the “AMN cerebral” group showed varying degrees of cerebral involvement. The scale devised by Loes et al. [10] was used to grade the severity of brain MRI abnormality. With this system a score of < 1 is classified as normal, and 34 is the maximally abnormal. Abnormal scores were present in the AMN cerebral and the childhood and adolescent cerebral group. The mean Loes score in the entire 102
132
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
group of subjects was 4.065 (S.D. 5.7). The 10-point Kurtzke Expanded Disability Status Scale (EDSS) [12] was used to assess the degree of functional disability. The mean score was 3.87 (S.D. 2.02), range 0 to 8. In 29/97 (29.9%) of patients the EDSS score was 6 or greater, indicative of a severe disability. At baseline, audiometric data including pure tone audiograms and word recognition measures were available for 90 patients and ABR data was available for 96 patients. Only the ABR measurements were repeated at subsequent visits. 48 of the 96 patients returned for the second visit, and 29 patients returned for the third visit. Subjects ranged in age from 5.32 to 73.44 years (mean age = 31.8, S.D. = 15.58, range = 5.32– 73.44). At the second test session, patients ranged in age from 11.18 to 68.2 years (mean age = 37.51; S.D. = 14.29). For the third test session, patients ranged in age from 19.94 to 62.65 years (mean age = 38.64; S.D. = 12.52). The interval between the initial and second test sessions ranged from 0.46 years to 4.66 years (mean = 0.77; S.D. = 0.67). The interval between the second and third test sessions ranged from 0.16 to 1.10 years (mean = 0.56; S.D. = 0.19). Subjects were placed in one of five diagnostic categories based upon established criteria [33], as shown in Table 1. The diagnostic categories and subject characteristics for the total sample and each of the five groups are shown in Table 1. All aspects of the study received prior approval by the Johns Hopkins University School of Medicine Institutional Review Board. Informed consent was received from each patient. In order to evaluate the significance of any observed hearing loss in the five clinical groups, age and sex matched normative data was derived by random selection from the National Health and Nutrition Examination Survey (1999–
2000). The cases were stratified into seven age groups. Four control cases were selected for each clinical case. The control group was selected for each phenotype separately because the phenotype groups differed significantly in age distribution. 3. Procedures 3.1. Audiometric and ABR measurements Patients were administered an audiological test battery that included measurement of air and bone conduction thresholds, speech reception thresholds, word recognition testing, tympanometry and the auditory brainstem response (ABR). Audiometric data were obtained utilizing standard audiometric techniques. ABRs were obtained for alternating polarity clicks presented at 70 dB nHL. The forehead electrode served as the active electrode and the ipsilateral mastoid as reference with the contralateral mastoid electrode as ground. Clicks were presented at a rate of 19/s with a 15 ms recording window and a passband of 30–3000 Hz. At least two recordings were made as each presentation level. The normality of ABR peaks and ABR interpeak latency intervals (IPLIs) was judged on the basis of the normative data shown below: I: <1.88 ms III: < 4.06 ms V: < 6.02 ms I–III: <2.35 ms III–V: < 2.22 ms I–V: <4.37 ms
Table 1 Percentage of ears in patients with an abnormally increased hearing threshold (>20 dB) in the five categories of ALD with hearing loss for pure tones at 250, 500, 1000, 2000, 4000 and 8000
Number of patient ears Mean, S.D., and range of the age in years of ALD patients on the date of audiometry Mean, S.D., and range of the age in years of normative groups 250 Hz Patients Controls 500 Hz Patients Controls 1000 Hz Patients Controls 2000 Hz Patients Controls 4000 Hz Patients Controls 8000 Hz Patients Controls
Cerebral (childhood + adolescent)
Pure AMN
AMN cerebral
Symptomatic female heterozygotes
Addison's only
28 15.3 S.D. = 5.1 (5.3 to 23.08)
62 35.9 S.D. = 11.7 (12.5 to 61.6)
48 33.9 S.D. = 11.4 (18.0–59.6)
34 48.6 S.D. = 11.1 (12.5 to 61.6)
8 12.6 S.D. = 3.6 (10.2 to 17.9)
25.1 S.D. = 2.8 (20 to 29) 7.1 – 0 6.2 3.6 3.6 0 8.1 10.7 17.1 25 16.2
35.3 S.D. = 10.8 (20 to 69) 14.5 – 9.6 7.0 4.8 7.0 6.4 13.6 22.6 33.3 27.4 31.6
33.6 S.D. = 10.4 (20 to 59) 20.8 – 4.2 7.9 2.1 6.9 6.3 13.2 33.3 23.2 31.2 24.9
48.9 S.D. = 9.4 (30 to 67) 8.8 – 11.7 16.4 20.5 9.0 23.5 13.4 32.3 22.4 32.3 30.6
26.3 S.D. = 2.8 (20 to 29) 0 – 0 6.2 12.5 0 0 3.1 0 15.6 25 12.5
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
If a peak could not be identified or was not repeatable upon successive recordings, the peak was classified as abnormal. When the interaural latency difference was examined as a measure in an analysis separate from the overall analysis described above, an interaural latency difference greater than 0.3 ms was considered abnormal. Sample ABR waveforms representative of a range of waveform morphology are shown in Fig. 1.
133
ABR testing respectively. The EDSS was calculated on the corresponding dates of the ABR test session. The EDSS ranged from 0 to 8 with a score of 6 or greater indicative of severe disability. Duration of symptoms was determined by chart review and was based upon the date of onset of neurological symptoms such as alterations in memory, difficulty in running, walking or climbing. 3.3. Statistical analysis
3.2. MRI, EDSS and VLFCA measurements Subjects were examined using a high-field 1.5-T signal MRI system. The standard protocol included sagittal images (repetition time/echo time [TR/TE], 600/20 ms) and axial proton-weighted images and T2 images (TR/TE, 3000/30 ms and 3000/100 ms, respectively) with slice thicknesses of 5 mm and a 2.5-mm intersection gap. MRI images were evaluated by radiologists not involved in the patients' care and were scored by using a system designed specifically for the assessment of adrenoleukodystrophy patients [10], with a range of scores of 0 through 34 based on the presence of white matter abnormalities in key brain regions. We have demonstrated inter-rater reliability coefficients for this system between 0.87 and 0.93. MRI Loes scores less than 1.5 were classified as normal whereas scores of 1.5 or greater were classified as abnormal. MRI data were obtained on 91 of the 102 patients with ALD. The levels of VLCFA in plasma and MRI Loes scores were measured on an average 6 months after (mean = 0.57 years, S.D. = 0.88 years) and 8.5 months before (mean = − 0.7 years, S.D. = 1.06 years) the
The Fisher's exact test was used to analyze the frequency of hearing loss and ABR abnormalities across the different ALD phenotypes. Regression analysis was used to analyze the association between bilaterally prolonged I–V interpeak latency intervals and ALD phenotype. Unadjusted odds ratios were calculated using univariate logistic regression. Multivariate logistic regression was performed to adjust for potential confounding by age, duration of symptoms, MRI Loes scores, EDSS scores and very long chain fatty acid (VLCFA) concentration in plasma. All possible interaction terms were tested and were found to be non-significant. The final model was selected based on odds ratio testing [34]. 4. Results 4.1. Auditory sensitivity Table 1 shows instances of mild to severe hearing loss particularly in the older subjects with the highest prevalence in the 4000–8000 Hz frequency region. For several ears, a precipitous audiometric configuration was observed at 6000–8000 Hz. The childhood cerebral and Addison only subjects, who were younger, had the highest percentage of ears with normal peripheral auditory sensitivity. Hearing loss was less prevalent from 500 to 2000 Hz for all groups. The significance of the observed hearing losses was evaluated by comparing the obtained hearing losses in the patient groups to normative data obtained in the NHANES (1999–2000). Comparisons could not be undertaken at 250 Hz because that frequency was not included in the NHANES (1999–2000). No significant difference was found in the percentage of ears with hearing loss for any of the five patient groups at any audiometric frequency when compared to the percentage of hearing loss present in age and sex matched controls. We concluded that auditory sensitivity in X-ALD is not significantly impaired. 4.2. Auditory brainstem responses (ABR) at baseline and follow-up
Fig. 1. Auditory brainstem response waveforms depicting, (A) an AMN patient with good response morphology and repeatability but prolonged I– III, III–V and I–V interpeak latency intervals; (B) an AMN patient with more marked prolongation of the I–V interpeak latency interval and an absent peak III. The summed response to three replications was utilized due to poor peak repeatability; and, (C) a patient with AMN with normal peripheral auditory sensitivity but no clearly repeatable ABR peaks.
Fig. 1 illustrates the waveform abnormalities in patients with AMN. Table 2 shows that ABR abnormalities were most prevalent for Wave V, followed by Wave III and Wave I. ABR abnormalities were present for the wave I–III, III–V and I–V interpeak latency intervals. Table 3 compares the frequency
134
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
Table 2 Number and percentage of patients in the five categories of ALD with normal or abnormal Wave I, III, and V, interaural Wave V latency difference or Wave I–III, III–V, and I–V interpeak latency intervals (total number = 96)
Number of patients (N) Mean, S.D. and range of age in years at audiometry Wave I Bilateral Unilateral Normal Wave III Bilateral Unilateral Normal Wave V Bilateral Unilateral Normal Interaural Wave V latency difference > 0.3 Interpeak latency I–III Bilateral Unilateral Normal Interpeak latency III–V Bilateral Unilateral Normal Interpeak latency I–V Bilateral Unilateral Normal Any ABR abnormality (abnormality in any of the ABR waves, either unilaterally or bilaterally)
Cerebral (childhood + adolescent)
Pure AMN
AMN cerebral
Symptomatic female heterozygote
Addison's only
20 13.9 S.D. = 4.8 (5.3 to 23.1)
31 35.8 S.D. = 12.6 (12.5 to 61.6)
25 34.3 S.D. = 11.3 (18.0 to 59.6)
15 47.4 S.D. = 9.6 (30.6 to 67.6)
5 11.5 S.D. = 3.9 (7.2 to 17.9)
5% 15% 80%
6.4% 25.8% 67.8% (NS)
20% 16% 64% (NS)
0% 6.7% 93.3% (NS)
0% 0% 5% (NS)
20% 10% 70%
71% 9.7% 19.3% ⁎⁎
80% 12% 8%⁎⁎
26.7% 40.0% 33.3% ⁎
20% 20% 60% (NS)
20% 5% 75% 20%
87.1% 6.4% 6.4%⁎⁎ 38.7%
96% 4% 0%⁎⁎ 52%
53.3% 20.0% 26.7%⁎⁎ 26.7%
20% 20% 60% (NS) 0%
20% 25% 55%
77.4% 12.9% 9.7%⁎⁎
92% 4% 4%⁎⁎
33.3% 46.7% 20.0%⁎
40% 20% 40% (NS)
15% 5% 80%
74.2% 19.4% 6.4%⁎⁎
96% 4% 0%⁎⁎
33.3% 20.0% 46.7% (NS)
20% 0% 80% (NS)
15% 15% 70% 50%
90.3% 3.2% 6.4% 93.5%⁎⁎
96% 4% 0% 100%⁎⁎
53.3% 33.3% 13.3% 100%⁎⁎
20% 20% 60% (NS) 60% (NS)
The cerebral (childhood–adolescent) category was taken as the reference category. NS—not significant; p > 0.05 by Fisher's exact test. ⁎ p < 0.05 by Fisher's exact test. ⁎⁎ p < 0.01 by Fisher's exact test.
of abnormal waves in 48 patients who were tested a mean of 0.46 years after the baseline evaluation and Table 4 in 29 patients who were also tested a third time. The mean interval between the second and the third tests was 0.56 years. Table 5 shows the phenotype distribution of the patients who were tested serially and Table 6 the changes in frequency in abnormalities over time. The ABR abnormalities did not appear to increase systematically for any waves interpeak latency intervals or in the interaural latency difference from session I and session II and session II to session III. 4.3. Correlation between ABR abnormalities and phenotype ABR abnormalities were most prevalent in males with cerebral and pure AMN and in the heterozygotes, and least prevalent in the childhood and adolescent cerebral and Addison disease categories of patients. The percentage of individuals with any ABR abnormality was significantly smaller in the childhood and adolescent cerebral phenotypes
as compared to the pure AMN, AMN cerebral and symptomatic heterozygote phenotypes (p < 0.01). No significant differences were present in the percentage of individuals with any ABR abnormality between the Addison's only and the childhood and adolescent cerebral phenotypes (p > 0.999). 4.4. Statistical analyses Table 7 summarizes correlations between ABR abnormalities and a variety of factors. Based on univariate logistic regression, ALD phenotype, age (odds ratio = 1.1, p < 0.001), duration of symptoms (odds ratio = 1.3, p = 0.002), EDSS (odds ratio = 1.4, p = 0.015) and MRI Loes scores (odds ratio = 0.98, p = 0.003) were found to be statistically significant predictors of bilaterally abnormal Wave I–Wave V interpeak latency while VLCFA concentration in plasma did not reach statistical significance (odds ratio = 0.3, p = 0.18).
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137 Table 3 Percentage of patients with abnormal waves in sessions 1 and 2 (number of patients who were present for both visits = 48) Visit 1
Visit 2
Wave I Bilateral Unilateral
8.3% 16.7%
12.5% 14.6%
Wave III Bilateral Unilateral
62.5% 20.8%
62.5% 20.8%
Wave V Bilateral Unilateral Interaural Wave V latency difference > 0.3
83.3% 8.3% 43.7%
83.3% 2.1% 43.7%
Interpeak latency I–III Bilateral Unilateral
79.2% 18.7%
81.2% 12.5%
Interpeak latency III–V Bilateral Unilateral
79.2% 8.3%
70.8% 12.5%
Interpeak latency I–V Bilateral Unilateral
87.5% 6.2%
83.3% 6.2%
With respect to the multivariate analyses, the difference in odds of having a bilaterally abnormal Wave I–Wave V interpeak latency interval between the cerebral (childhood and adolescent) phenotype and pure AMN was not found to Table 4 Percentage of patients with abnormal waves in sessions 1 and 3 (number of patients who were present for both visits = 29) Visit 1
Visit 3
Wave I Bilateral Unilateral
13.8% 17.2%
13.8% 24.1%
Wave III Bilateral Unilateral
72.4% 17.2%
72.4% 20.7%
Wave V Bilateral Unilateral Interaural Wave V latency difference > 0.3
82.8% 13.8% 41.4%
86.2% 6.9% 41.4%
Interpeak latency I–III Bilateral Unilateral
86.2% 13.8%
86.2% 13.8%
Interpeak latency III–V Bilateral Unilateral
82.8% 6.9%
79.3% 3.4%
Interpeak latency I–V Bilateral Unilateral
93.1% 6.9%
89.7% 6.9%
135
Table 5 Phenotype distribution of patients in each of the three visits Phenotype
Visit 1
Cerebral (childhood + adolescent) Pure AMN AMN cerebral Symptomatic female heterozygotes Addison's disease Total
20 (20.8%)
Visit 2 5 (10.4%)
Visit 3 3 (10.3%)
31 (32.3%) 25 (26%) 15 (15.6%)
18 (37.5%) 16 (33.3%) 8 (16.7%)
11 (37.9%) 11 (37.9%) 4 (13.8%)
5 (5.2%) 96
1 (2.1%) 48
0 29
be statistically significant (odds ratio = 2.8, p = 0.57) after adjusting for age, duration of symptoms, VLCFA levels, EDSS scores and MRI Loes scores. Also, the difference in odds between the cerebral (childhood and adolescent) phenotype and the symptomatic heterozygote females failed to reach statistical significance (odds ratio = 0.01, p = 0.07) after adjusting for these covariates. However, a significant difference (odds ratio = 21.2, p = 0.049) was found between the cerebral (childhood and adolescent) phenotype and the AMN-cerebral group after adjusting for the covariates listed above. Age was also found to be significantly independent predictor of a bilaterally abnormal Wave I–Wave V interpeak latency interval after adjusting for ALD phenotype, duration of symptoms, VLCFA levels, EDSS scores and MRI Loes score (odds ratio = 1.1, p < 0.001). MRI Loes score (odds = 0.8, p = 0.31), EDSS scores (odds ratio = 1.8, p = 0.09), duration of symptoms (odds ratio = 1.05, p = 0.74), and VLCFA concentration in plasma (odds = 1.2, p = 0.94) were not found to be statistically significant independent predictors in the final multivariate regression model. 5. Discussion Our studies of audiological function in 102 subjects with X-ALD indicate that auditory sensitivity is not significantly impaired, but confirm that auditory brainstem abnormalities (ABR) are present in a large proportion. There were significant correlations between ABR abnormalities and phenotype, which provide new insights into pathogenetic mechanisms. The ABR has been utilized for many years in neurodiagnostics because it reflects activity from the auditory nerve and auditory pathways in the brainstem. On the basis of Table 6 Number and percentage of people in the five categories of ALD having abnormal Wave V latencies (defined as an interaural difference in the latency of Wave V of more than 0.3 ms)
Visit 1 Visit 2 Visit 3
Cerebral (childhood and adolescent)
Pure AMN
AMN cerebral
Symptomatic female heterozygote
Addison's only
20% 40% 33.3%
38.7% 38.9% 54.5%
52% 43.7% 36.3%
26.7% 25% 25%
0% 0% –
136
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137
Table 7 Predictors of bilaterally prolonged I–V interpeak latency intervals
Phenotype Cerebral childhood and adolescent Pure AMN AMN cerebral Symptomatic female heterozygotes Addison's only Age MRI Loes score Duration of symptoms EDSS Score Very long chain fatty acid (VLCFA) conc. in plasma
Univariate analysis
Multivariate analysis
OR
p-value
OR
p-value
Reference category 56 154 8.4
–
–
<0.001 <0.001 0.03
Reference category 2.8 21.2 0.01
0.57 0.049 0.07
⁎ 1.1 0.9 1.3 1.4 0.3
⁎ <0.001 0.003 0.002 0.015 0.18
⁎ 1.2 0.8 1.05 1.8 1.2
⁎ 0.01 0.31 0.74 0.09 0.94
⁎Only two patients with the “Addison's only” phenotype had complete data and hence this category could not be included in the regression model. Both of those patients had normal I–V interpeak latency intervals.
measurements obtained during operations for cranial nerve disorders in humans [35], it has been determined that the neural generator for Waves I and II in humans is the auditory nerve. Abnormalities in these waveforms were relatively uncommon. This, combined with our data that auditory sensitivity is not significantly impaired, indicates that the auditory nerve and cochlear nucleus do not appear to be involved in X-ALD. It is more difficult to attribute specific generators to the later peaks of the ABR due to the extent of parallel processing in the auditory pathway at brainstem levels; the peaks of waves subsequent to Wave II have multiple sources underlying their generation [36]. While there is evidence on the basis of intraoperative recordings that the neural generator for Wave III of the human ABR is the cochlear nucleus [37], the generation of Wave III may be more complex than originally supposed. The contralateral cochlear nucleus [38] as well as the most proximal portion of the auditory nerve may also make a contribution to the generation of Wave III [39]. Several neural sources contribute as the generators for Wave V [40]. The most positive peak of Wave V is probably generated at the termination of the fiber tract of the lateral lemniscus, whereas the following negative trough in conventionally recorded ABRs is generated by slow dendritic potentials in the inferior colliculus [40]. The main contribution to the peak of wave V appears to be from contralateral rather than ipsilateral structures on the basis of intracranial recordings in human subjects [38]. Taken together, these data suggest that the ABR abnormality is caused by lesions in fiber tracts in the lateral lemniscus and inferior colliculus. The present ABR findings can be correlated with current concepts of the pathogenesis of X-ALD. Childhood and adolescent cerebral X-ALD are associated with an intensely inflammatory demyelination that is most severe in the cerebral hemispheres. The thalamic nuclei, basal ganglia and
geniculate bodies are also involved, but the brainstem is relatively spared [5]. Table 2 shows that most ABRs were normal in the majority of boys with these phenotypes. In contrast, nearly all AMN patients show ABR abnormalities. This difference may be partially attributable to age, because there is a significant positive correlation between age and ABR abnormalities, and AMN patients are older. However, comparison of the “pure AMN” and the “cerebral AMN” provides interesting additional information. Although Table 2 shows that cerebral AMN patients are slightly younger and have a slightly higher frequency of ABR abnormalities, these differences are not statistically significant. The basic disease process in pure AMN is a non-inflammatory axonopathy, while inflammatory demyelination is the main abnormality in childhood and adolescent cerebral forms and in the cerebral component of cerebral AMN. As has been reported before, approximately 60% of women heterozygous for XALD had abnormal ABR responses. This is consistent with the report that approximately 50% of carriers develop an AMN-like syndrome in middle age or later [28]. The high frequency of ABR abnormalities in pure AMN, and their relatively low frequency in childhood and adolescent cerebral X-ALD, and the lack of significant difference in their frequency in pure AMN and cerebral AMN, suggest strongly that the ABR abnormalities are due to a non-inflammatory axonopathy, most likely in axonal tracts in the lateral lemniscus and inferior colliculus. Application of novel imaging techniques, such as diffusion tensor imaging combined with three dimensional fiber tracking [18] make it possible to define this more precisely. The likelihood that the ABR abnormalities reflect an axonopathy rather than inflammatory demyelination is also of practical significance. Hematopoietic stem cell transplantation is recommended for X-ALD patients who show early evidence of inflammatory demyelination but at this time it is not recommended for pure AMN [24]. In conclusion we find (1) that auditory sensitivity in XALD is not significantly impaired; (2) ABR abnormalities are a frequent finding and may be caused by abnormalities of fiber tracts in the region of the lateral lemniscus and inferior colliculus; (3) the abnormalities progress slowly and appear to be associated mainly with the AMN phenotype. Acknowledgement Support from the General Clinical Research Center (MO1-RR00052) from the National Center for Research Resources is acknowledged. References [1] Mosser J, Douar AM, Sarde CO, Kioschis P, Feil R, Moser H, et al. Putative X-linked adrenoleukodystrophy gene shares unexpected homology with ABC transporters. Nature 1993;361:726–30. [2] Smith KD, Kemp S, Braiterman LT, Lu JF, Wei HM, Geraghty M, et al. X-linked adrenoleukodystrophy: genes, mutations, and phenotypes. Neurochem Res 1999; 24:521–35.
J.P. Pillion et al. / Journal of the Neurological Sciences 247 (2006) 130–137 [3] Igarashi M, Schaumburg HH, Powers J, Kishmoto Y, Kolodny E, Suzuki K. Fatty acid abnormality in adrenoleukodystrophy. J Neurochem 1976;26:851–60. [4] Moser AB, Kreiter N, Bezman L, Lu S, Raymond GV, Naidu S, Moser HW. Plasma very long chain fatty acids in 3000 peroxisome disease patients and 29,000 controls. Ann Neurol 1999;45:100–10. [5] Schaumburg HH, Powers JM, Raine CS, Suzuki K, Richardson Jr EP. Adrenoleukodystrophy. A clinical and pathological study of 17 cases. Arch Neurol 1975;32:577–91. [6] Powers JM, Liu Y, Moser AB, Moser HW. The inflammatory myelinopathy of adrenoleukodystrophy: cells, effector molecules, and pathogenetic implications. J Neuropathol Exp Neurol 1992; 51:630–43. [7] Ito M, Blumberg BM, Mock DJ, Goodman AD, Moser AB, Moser HW, et al. Potential environmental and host participants in the early white matter lesion of adrenoleukodystrophy: morphologic evidence for CD8 cytotoxic T cells, cytolysis of oligodendrocytes, and CD1mediated lipid antigen presentation. J Neuropathol Exp Neurol 2001;60:1004–19. [8] Kumar AJ, Rosenbaum AE, Naidu S, Wenger L, Citrin CM, Lindenberg R, et al. Adrenoleukodystrophy: correlating MR imaging with CT. Radiology 1987;165:497–504. [9] Loes DJ, Fatemi A, Melhem ER, Gupte N, Bezman L, Moser HW, Raymond GV. Analysis of MRI patterns aids prediction of progression in X-linked adrenoleukodystrophy. Neurology 2003;61:369–74. [10] Loes DJ, Hite S, Moser H, Stillman AE, Shapiro E, Lockman L, et al. Adrenoleukodystrophy: a scoring method for brain MR observations. AJNR Am J Neuroradiol 1994; 15:1761–6. [11] Powers JM, DeCiero DP, Ito M, Moser AB, Moser HW. Adrenomyeloneuropathy: a neuropathologic review featuring its noninflammatory myelopathy. J Neuropathol Exp Neurol 2000;59:89–102. [12] Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 1983;33: 1444–52. [13] Zackowski K, Dubey P, Raymond GV, Mahmood A, Moser AB, Moser HW. Relating sensorimotor function and axonal integrity in adrenomyeloneuropathy. Arch Neurol 2006;63:74–80. [14] Koehler W, Sokolowski P. A new disease-specific scoring system for adult phenotypes of X-linked adrenoleukodystrophy. J Mol Neurosci 1999;13:247–52. [15] Restuccia D, Di Lazzaro V, Valeriani M, Oliviero A, Le Pera D, Colosimo C, et al. Neurophysiological abnormalities in adrenoleukodystrophy carriers. Evidence of different degrees of central nervous system involvement. Brain 1997;120(Pt 7): 1139–1148. [16] Kumar AJ, Kohler W, Kruse B, Naidu S, Bergin A, Edwin D, Moser HW. MR findings in adult-onset adrenoleukodystrophy. AJNR Am J Neuroradiol 1995;16:1227–37. [17] Dubey P, Fatemi A, Barker PB, Degaonkar M, Troeger M, Zackowski K, et al. Spectroscopic evidence of cerebral axonopathy in patients with “pure” adrenomyeloneuropathy. Neurology 2005;64:304–10. [18] Dubey P, Fatemi A, Huang H, Nagae-Poetscher L, Wakana S, Barker PB, et al. Diffusion tensor-based imaging reveals occult abnormalities in adrenomyeloneuropathy. Ann Neurol 2005;58:758–66. [19] Fatemi A, Smith SA, Dubey P, Zackowski KM, Bastian AJ, van Zijl PC, et al. Magnetization transfer MRI demonstrates spinal cord pathology in adrenomyeloneuropathy. Neurology 2005:1739–45. [20] Kaplan PW, Tusa RJ, Rignani J, Moser HW. Somatosensory evoked potentials in adrenomyeloneuropathy. Neurology 1997;48:1662–7.
137
[21] Bezman L, Moser AB, Raymond GV, Rinaldo P, Watkins PA, Smith KD, et al. Adrenoleukodystrophy: incidence, new mutation rate, and results of extended family screening. Ann Neurol 2001;49:512–7. [22] Moser HW, Raymond GV, Dubey P. Adrenoleukodystrophy: new approaches to a neurodegenerative disease. JAMA 2005;294:3131–4. [23] Moser HW, Raymond GV, Lu SE, Muenz LR, Moser AB, Xu J, et al. Follow-up of 89 Lorenzo's oil treated asymptomatic adrenoleukodystrophy patients. Arch Neurol 2005;62:1073–80. [24] Peters C, Charnas LR, Tan Y, Ziegler RS, Shapiro EG, DeFor T, et al. Cerebral X-linked adrenoleukodystrophy: the international hematopoietic cell transplantation experience from 1982 to 1999. Blood 2004;104:881–8. [25] Vercruyssen A, Martin JJ, Mercelis R. Neurophysiological studies in adrenomyeloneuropathy. A report on five cases. J Neurol Sci 1982; 56:327–36. [26] Ochs R, Markand ON, DeMyer WE. Brainstem auditory evoked responses in leukodystrophies. Neurology 1979;29:1089–93. [27] Grimes AM, Elks ML, Grunberger G, Pikus AM. Auditory brainstem responses in adrenomyeloneuropathy. Arch Neurol 1983;40:574–6. [28] Shimizu H, Moser HW, Naidu S. Auditory brainstem response and audiologic findings in adrenoleukodystrophy: its variant and carrier. Otolaryngol Head Neck Surg 1988;98:215–20. [29] Garg BP, Markand ON, DeMyer WE, Warren Jr C. Evoked response studies in patients with adrenoleukodystrophy and heterozygous relatives. Arch Neurol 1983;40:356–9. [30] Moloney JB, Masterson JG. Detection of adrenoleukodystrophy carriers by means of evoked potentials. Lancet 1982;2:852–3. [31] Boehm CD, Cutting GR, Lachtermacher MB, Moser HW, Chong SS. Accurate DNA-based diagnostic and carrier testing for X-linked adrenoleukodystrophy. Mol Genet Metab 1999;66:128–36. [32] Moser HW, Smith KD, Watkins PA, Powers J, Moser AB. X-linked adrenoleukodystrophy. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease. Eighth edition. New York: McGraw Hill; 2001. p. 3257–301. [33] Moser HW, Smith KD, Moser AB. X-linked adrenoleukodystrophy. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular basis of inherited disease. New York: McGraw Hill; 1994. p. 2325–49. [34] Hosmer DW, Lemeshow S. Applied logistic regression. Second edition. New York: John Wiley and Sons, Inc.; 2000. [35] Moller AR, Jannetta PJ, Moller MB. Neural generators of brainstem evoked potentials. Results from human intracranial recordings. Ann Otol Rhinol Laryngol 1981;90:591–6. [36] Moller AR, editor. Interoperative neurophysiologic monitoring. Luxembourg: Harwood Academic Publishers; 1995. p. 66. [37] Moller AR, Jannetta P, Moller MB. Intracranially recorded auditory nerve response in man. New interpretations of BSER. Arch Otolaryngol 1982;108:77–82. [38] Moller AR, Jho HD, Yokota M, Jannetta PJ. Contribution from crossed and uncrossed brainstem structures to the brainstem auditory evoked potentials: a study in humans. Laryngoscope 1995;105:596–605. [39] Moller AR, Jho HD. Compound action potentials recorded from the intracranial portion of the auditory nerve in man: effects of stimulus intensity and polarity. Audiology 1991;30:142–63. [40] Moller AR, Jannetta PJ. Evoked potentials from the inferior colliculus in man. Electroencephalogr Clin Neurophysiol 1982;53:612–20.