Abstract
Objective
To determine the relationship of cerebral degeneration with survival in amyotrophic lateral sclerosis (ALS).
Methods
Patients with probable or definite ALS underwent magnetic resonance spectroscopic imaging (MRSI) of the brain between July 1996 and May 2002, and were followed prospectively until March 2004. Creatine (Cr), choline (Cho) and the neuronal marker N‐acetylaspartate (NAA) were quantified as ratios in the motor cortex.
Results
In 63 patients compared with 18 healthy people, NAA/Cho was reduced by 13% (p<0.001), NAA/Cr was reduced by 5% (p = 0.01) and Cho/Cr was increased by 8% (p = 0.01). NAA/Cho was used for survival analysis, given its larger effect size and superior test accuracy (a sensitivity of 67% and a specificity of 83%). Median survival after MRSI was 24 months. Multivariate analysis showed reduced survival for lower NAA/Cho (hazard ratio (HR) 0.24, 95% confidence interval (CI) 0.08 to 0.72, p = 0.01), older age (HR 1.03, 95% CI 1.00 to 1.06, p = 0.04) and shorter symptom duration (HR 0.96, 95% CI 0.93 to 0.99, p = 0.01). Patients with NAA/Cho <2.11 had a reduced survival of 19.4 v 31.9 months (HR 2.05, 95% CI 1.12 to 4.03, p = 0.02).
Conclusions
Cerebral degeneration is predictive of reduced survival in ALS.
The extent to which cerebral involvement contributes to shortened life expectancy in amyotrophic lateral sclerosis (ALS) has been difficult to elucidate. The clinical evaluation of upper motor neurone (UMN) integrity as an index of motor cortex degeneration falls short owing to the possible masking of UMN signs by the loss of lower motor neurones (LMNs). Indeed, autopsy series have shown UMN degeneration in cases of LMN‐predominant motor neurone disease that would not have otherwise met the clinical criteria for ALS.1 The identification of prognostic indicators would aid in selecting patients for clinical trials who are predicted to behave in a homogeneous manner with respect to disease progression and, potentially, pathogenesis. Prognostic indicators would also have clinical value in counselling patients.
Proton magnetic resonance spectroscopy (MRS) studies in ALS have consistently shown evidence of neuronal loss or dysfunction in the motor cortex by the finding of decreased N‐acetylaspartate (NAA), a neuronal marker.2,3
We aimed to determine whether cerebral neurochemical abnormalities quantified by MRS correlated with survival.
Methods
Participants
Patients were referred from the ALS clinic of the Montreal Neurological Hospital, Montreal, Canada, for magnetic resonance spectroscopic imaging (MRSI), between July 1996 and May 2002. At the time of scanning they had a symptom duration <5 years and met the El Escorial criteria4 for probable or definite ALS; thus, all had examination findings of UMN dysfunction. Patients were followed prospectively during routine clinic visits until March 2004. Patients were not included or excluded on the basis of participation in other studies and clinical trials. Healthy patients without neurological disorders were used as controls. Informed consent was obtained from the participants and the study was approved by the local ethics committee.
Imaging
MRSI was carried out on a 1.5‐T scanner with the volume of interest placed over the central sulcus (see supplementary fig A on http://jnnp.bmjjournals.com/supplemental). The volume of interest was 20 mm thick. Anteroposterior and left–right dimensions averaged 75 mm each. A point‐resolved spectroscopic localisation scheme was used (TR/TE (repetition time/echo time) = 1750/272 ms, field of view 250×250 mm, 32×32 phase‐encoding steps, water suppression by selective excitation). Mean metabolite ratios (NAA/Cho (choline), NAA/Cr (creatine), Cho/Cr) of the primary motor cortex were calculated by averaging the ratios for voxels containing at least 50% of the precentral gyrus.
Statistics
Group differences were assessed using the Mann–Whitney U test and χ2 tests. Receiver operator characteristic analysis was carried out for metabolite ratios to determine the optimum cut‐off values with the highest accuracy (minimal false negatives and false positives) at which sensitivity and specificity were calculated. The metabolite ratio with the best accuracy profile was used for subsequent correlative and survival analyses. Spearman's rank (R) correlation coefficients were computed to determine the relationship of metabolite ratios with demographic and clinical parameters.
Time to death or tracheostomy after the MRSI was studied using the multivariate Cox proportional hazards method. The Kaplan–Meier method was used to calculate survival curves, which were compared with the log rank test. Patients were censored at the end of the observation period if still alive without a tracheostomy. Those lost to follow‐up before study termination were censored at the date of the last clinic visit or phone conversation with clinic personnel.
A supplementary analysis was carried out in a subset of patients who had complete datasets with an Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS, range 0–40) and forced vital capacity (FVC, % predicted) carried out within 2 weeks of the MRSI. These measures were collected retrospectively. Because the restricted sample size would prohibit a complete multivariate analysis, an automated backward stepwise approach was used to select variables for the multivariate Cox proportional hazards model.
A two‐sided value of p<0.05 was considered to be significant.
Results
Clinical features
No difference was observed with respect to age or sex in patients compared with controls. NAA/Cho was the most abnormal of the three metabolite ratios (table 1). The area under the receiver operating characteristic curve (see supplementary fig B on http://jnnp.bmjjournals.com/supplemental) was greatest for NAA/Cho, reflecting its superior sensitivity–specificity profile. Of the three metabolite ratios, NAA/Cho was used for subsequent correlative and survival analyses given its larger mean group difference and slightly better accuracy profile.
Table 1 Demographic data and spectroscopy results.
| Control | ALS | % Difference in ALS compared with control | p Value | Sensitivity %, specificity % (ROC curve cut‐off) | |
|---|---|---|---|---|---|
| n | 18 | 63 | |||
| Age (years) | 60.5 (52.0–69.0) | 61.0 (51.0–67.8) | |||
| Sex, male:female | 12:6 | 44:19 | |||
| Symptom onset, limb:bulbar | 49:14 | ||||
| Symptom duration (months) | 15 (9–25) | ||||
| NAA/Cho | 2.42 (2.24–2.60) | 2.11 (1.87–2.30) | −13 | <0.001 | 67, 83 (2.22) |
| NAA/Cr | 2.46 (2.41–2.55) | 2.34 (2.17–2.47) | −5 | 0.01 | 67, 78 (2.41) |
| Cho/Cr | 1.06 (0.96–1.10) | 1.15 (1.06–1.23) | +8 | 0.01 | 62, 83 (1.12) |
ALS, amytrophic lateral sclerosis; Cho, choline; Cr, creatine; MRSI, magnetic resonance spectroscopy imaging; NAA, N‐acetylaspartate; ROC, receiver operating characteristic.
Values are median (interquartile range).
Spectroscopy ratios are unitless.
Duration of symptoms is from onset of first symptom to time of MRSI.
In the ALS group, NAA/Cho did not vary with age, sex, symptom duration or site of symptom onset.
Survival analysis
By the end of the observation period, 40 patients had died (n = 36) or had a tracheostomy (n = 4), 14 were alive and 9 were lost to follow‐up. Median survival after the MRSI was 24 months. Those lost to follow‐up were not different from the rest of the group with respect to site of symptom onset, sex, disease duration and metabolite ratios.
Multivariate Cox proportional hazards analysis identified decreased NAA/Cho to have the greatest correlation with reduced survival (hazard ratio (HR) 0.24, 95% confidence interval (CI) 0.08 to 0.72, p = 0.01). Other predictors were older age (HR 1.03, 95% CI 1.00 to 1.06, p = 0.04) and shorter symptom duration (HR 0.96, 95% CI 0.93 to 0.99, p = 0.01). Sex and site of initial symptoms (bulbar v limb) did not correlate with survival.
Survival was compared in patients stratified at the median NAA/Cho of the ALS group (fig 1). The median survival in those with NAA/Cho <2.11 was 19.4 months compared with 31.9 months (HR 2.05, 95% CI 1.12 to 4.03, p = 0.02).
Figure 1 Kaplan–Meier survival curves in patients with amyotrophic lateral sclerosis stratified according to the median NAA/Cho (N‐acetylaspartate/choline) in the motor cortex. Survival is reduced in patients with NAA/Cho below (dotted line) compared with those above (solid line) this median value of 2.11 (p = 0.02).
In a subgroup analysis of 32 patients with an ALSFRS (median 33, range 19–39) and FVC (median 86%, range 53–109%), these variables were entered in addition to those used in the previous model in a stepwise Cox proportional hazards analysis. Only reduced NAA/Cho was retained in the final model as being predictive of survival (HR 0.14, 95% CI 0.03 to 0.68, p = 0.02).
Discussion
We used MRSI to show that impaired neuronal integrity in the motor cortex correlates with survival in ALS. MRSI was the best predictor of survival in multivariate analysis (HR 0.24). The risk of death was increased by 2.05 if motor cortex NAA/Cho was below the median value of 2.11. The magnitude of these risks is comparable with other published values for various clinical parameters in survival analyses in ALS.5,6,7,8 These results suggest that the neuronal integrity of the motor cortex is a marker for survival. Whether motor cortex degeneration shortens life span directly, and by what mechanism, cannot be elucidated from this study; a better understanding of this may be possible by further investigation of the relationship between clinical parameters and spectroscopic markers.
Similar to other studies, we found that older age6,7 and briefer symptom duration8,9 reduced the probability of survival. Unlike the subjects of other studies, our patients with bulbar‐onset ALS did not have reduced survival compared with those with limb‐onset symptoms.10,11 This may have been a consequence of selection bias in this study against patients with significant bulbar disease who have difficulty tolerating magnetic resonance examinations, or an insufficient sample size.
That NAA/Cho correlated with life expectancy when the ALSFRS and FVC did not would suggest that it is more robust at prognostication. However, this was an exploratory analysis on a small subset of patients in whom the ALSFRS and FVC data were collected retrospectively. Other limitations include the uneven follow‐up of patients, as they were followed in the setting of clinic visits, and patient selection that may not be representative of the general ALS population. Some patients were lost to follow‐up; however, they had characteristics similar to the rest of the group. The effects on survival of riluzole,12 non‐invasive ventilation,13 nutritional supplementation through percutaneous endoscopic gastrostomy7 and correlations with functional disease measures such as the ALSFRS14,15 and FVC9 could not be adequately evaluated as documentation of such interventions and scales was not complete.
NAA was reduced relative to both choline and creatine, in keeping with numerous other studies.2,3,16,17,18 NAA/Cho was reduced more than NAA/Cr, consistent with the results of others who have measured both ratios simultaneously.18,19,20 This may be owing to an increase in choline concomitant to decreased NAA, as has been noted by some.16,18,21 Although increased choline is not substantiated by other investigators,17,22 in ALS it could be reflective of cerebral astrogliosis23 or cellular membrane degeneration.24
To the best of our knowledge, this is the first report of a relationship of motor cortex degeneration with survival and of the utility of in vivo brain spectroscopy in prognostication in ALS. This study supports the use of such metabolites as surrogate markers of cerebral pathology. The ability to prognosticate in ALS is an important step towards the identification of subgroups of patients that may be subject to varying neurobiological processes, and it could potentially help in counselling patients and their care givers. The extent to which motor cortex degeneration independently contributes to disability and mortality is relevant with respect to the development of novel treatments that should target appropriate cells and regions of the nervous system.
A prospective study is required to compare the prognostic ability of MRS with clinical measures, other objective indicators of UMN health such as transcranial magnetic stimulation,3 measures reflecting LMN function, and rates of change of brain metabolites and clinical indices.25
Acknowledgements
We thank the patients and their families for their participation and Dr Richard Camicioli for his helpful comments.
Abbreviations
ALS - amyotrophic lateral sclerosis
ALSFRS - Amyotrophic Lateral Sclerosis Functional Rating Scale
NAA - N‐acetylaspartate
FVC - forced vital capacity
LMN - lower motor neurone
UMN - upper motor neurone
MRS - proton magnetic resonance spectroscopy
MRSI - magnetic resonance spectroscopic imaging
NAA - N‐acetylaspartate
Footnotes
Funding: This study was partly funded by the Amyotrophic Lateral Sclerosis Association of America.
Competing interests: None.
Ethical approval: This study was approved by the Montreal Neurological Institute and Hospital Ethics Committee.
References
- 1.Ince P G, Evans J, Knopp M.et al Corticospinal tract degeneration in the progressive muscular atrophy variant of ALS. Neurology 2003601252–1258. [DOI] [PubMed] [Google Scholar]
- 2.Pioro E P, Antel J P, Cashman N R.et al Detection of cortical neuron loss in motor neuron disease by proton magnetic resonance spectroscopic imaging in vivo. Neurology 1994441933–1938. [DOI] [PubMed] [Google Scholar]
- 3.Kaufmann P, Pullman S L, Shungu D C.et al Objective tests for upper motor neuron involvement in amyotrophic lateral sclerosis (ALS). Neurology 2004621753–1757. [DOI] [PubMed] [Google Scholar]
- 4.Brooks B R, Miller R G, Swash M.et al El Escorial revisited: revised criteria for the diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord 20001293–299. [DOI] [PubMed] [Google Scholar]
- 5.Tysnes O B, Vollset S E, Larsen J P.et al Prognostic factors and survival in amyotrophic lateral sclerosis. Neuroepidemiology 199413226–235. [DOI] [PubMed] [Google Scholar]
- 6.Haverkamp L J, Appel V, Appel S H. Natural history of amyotrophic lateral sclerosis in a database population. Validation of a scoring system and a model for survival prediction. Brain 1995118707–719. [DOI] [PubMed] [Google Scholar]
- 7.Chio A, Mora G, Leone M.et al Early symptom progression rate is related to ALS outcome: a prospective population‐based study. Neurology 20025999–103. [DOI] [PubMed] [Google Scholar]
- 8.del Aguila M A, Longstreth W T, Jr, McGuire V.et al Prognosis in amyotrophic lateral sclerosis: a population‐based study. Neurology 200360813–819. [DOI] [PubMed] [Google Scholar]
- 9.Stambler N, Charatan M, Cedarbaum J M. Prognostic indicators of survival in ALS. ALS CNTF Treatment Study Group. Neurology 19985066–72. [DOI] [PubMed] [Google Scholar]
- 10.Tysnes O B, Vollset S E, Aarli J A. Epidemiology of amyotrophic lateral sclerosis in Hordaland county, western Norway. Acta Neurol Scand 199183280–285. [DOI] [PubMed] [Google Scholar]
- 11.Magnus T, Beck M, Giess R.et al Disease progression in amyotrophic lateral sclerosis: predictors of survival. Muscle Nerve 200225709–714. [DOI] [PubMed] [Google Scholar]
- 12.Lacomblez L, Bensimon G, Leigh P N.et al Dose‐ranging study of riluzole in amyotrophic lateral sclerosis. Amyotrophic Lateral Sclerosis/Riluzole Study Group II. Lancet 19963471425–1431. [DOI] [PubMed] [Google Scholar]
- 13.Aboussouan L S, Khan S U, Meeker D P.et al Effect of noninvasive positive‐pressure ventilation on survival in amyotrophic lateral sclerosis. Ann Intern Med 1997127450–453. [DOI] [PubMed] [Google Scholar]
- 14.Cedarbaum J M, Stambler N. Performance of the Amyotrophic Lateral Sclerosis Functional Rating Scale (ALSFRS) in multicenter clinical trials. J Neurol Sci 1997152(Suppl 1)S1–S9. [DOI] [PubMed] [Google Scholar]
- 15.Kaufmann P, Levy G, Thompson J L.et al The ALSFRS predicts survival time in an ALS clinic population. Neurology 20056438–43. [DOI] [PubMed] [Google Scholar]
- 16.Bowen B C, Pattany P M, Bradley W G.et al MR imaging and localized proton spectroscopy of the precentral gyrus in amyotrophic lateral sclerosis. Am J Neuroradiol 200021647–658. [PMC free article] [PubMed] [Google Scholar]
- 17.Sarchielli P, Pelliccioli G P, Tarducci R.et al Magnetic resonance imaging and 1H‐magnetic resonance spectroscopy in amyotrophic lateral sclerosis. Neuroradiology 200143189–197. [DOI] [PubMed] [Google Scholar]
- 18.Suhy J, Miller R G, Rule R.et al Early detection and longitudinal changes in amyotrophic lateral sclerosis by 1H MRSI. Neurology 200258773–779. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Block W, Karitzky J, Traber F.et al Proton magnetic resonance spectroscopy of the primary motor cortex in patients with motor neuron disease: subgroup analysis and follow‐up measurements. Arch Neurol 199855931–936. [DOI] [PubMed] [Google Scholar]
- 20.Pohl C, Block W, Karitzky J.et al Proton magnetic resonance spectroscopy of the motor cortex in 70 patients with amyotrophic lateral sclerosis. Arch Neurol 200158729–735. [DOI] [PubMed] [Google Scholar]
- 21.Schuff N, Rooney W D, Miller R.et al Reanalysis of multislice H MRSI in amyotrophic lateral sclerosis. Magn Reson Med 200145513–516. [DOI] [PubMed] [Google Scholar]
- 22.Gredal O, Rosenbaum S, Topp S.et al Quantification of brain metabolites in amyotrophic lateral sclerosis by localized proton magnetic resonance spectroscopy. Neurology 199748878–881. [DOI] [PubMed] [Google Scholar]
- 23.Urenjak J, Williams S R, Gadian D G.et al Proton nuclear magnetic resonance spectroscopy unambiguously identifies different neural cell types. J Neurosci 199313981–989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Miller B L, Chang L, Booth R.et al In vivo 1H MRS choline: correlation with in vitro chemistry/histology. Life Sci 1996581929–1935. [DOI] [PubMed] [Google Scholar]
- 25.Traynor B J, Zhang H, Shefner J M.et al Functional outcome measures as clinical trial endpoints in ALS. Neurology 2004631933–1935. [DOI] [PubMed] [Google Scholar]