Multiple Sclerosis and Related Disorders (2015) 4, 104–111
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/msard
REVIEW
Cardiovascular autonomic dysfunction in multiple sclerosis: A meta-analysis Juan Manuel Racostaa, Luciano A. Sposatoa, Sarah A. Morrowa, Lauren Ciprianob, Kurt Kimpiskia, Marcelo Kremenchutzkya,n a
Department of Clinical Neurological Sciences, London Health Sciences Centre. Western, University, London, ON, Canada b Ivey Business School, Western University, London, ON, Canada Received 26 November 2014; received in revised form 11 February 2015; accepted 13 February 2015
KEYWORDS Multiple sclerosis; Autonomic dysfunction; Diagnosis; Autonomic nervous system
Abstract Background and objective: The definition of cardiovascular autonomic dysfunction in patients with multiple sclerosis is controversial. Thus, its true prevalence is unknown. We performed a systematic review and meta-analysis to compare the proportion of patients with multiple sclerosis that would be diagnosed with cardiovascular dysautonomia using a definition of at least one abnormal cardiac autonomic test vs. at least two abnormal studies. Methods: We searched PubMed, Embase, and Scopus from 1980 to December 2013 for publications reporting abnormal autonomic tests in patients with multiple sclerosis. We performed random-effects meta-analyses for calculating the proportion of patients diagnosed with autonomic dysfunction with both definitions. Results: We included 16 studies comprising 611 patients with multiple sclerosis, assessing Z3 cardiovascular autonomic tests. The proportion of patients with autonomic dysfunction was two-fold higher (p=0.006) when using the definition of only one abnormal autonomic test (42.1%) compared to that using at least two abnormal results (18.8%). Conclusions: We found a wide variation in the proportion of patients with multiple sclerosis diagnosed with cardiovascular dysautonomia by using the two definitions. Consensus is needed to define autonomic dysfunction in patients with multiple sclerosis. In the meantime, we encourage investigators to report results using both thresholds. & 2015 Elsevier B.V. All rights reserved.
n Correspondence to: MS Clinic, University Hospital, LHSC, 339 Windermere Road, P.O. Box 5339, London ON, Canada N6A 5A5. Tel.: +1 519 663 3697; fax: +1 519 663 3744. E-mail address: [email protected] (M. Kremenchutzky).
http://dx.doi.org/10.1016/j.msard.2015.02.002 2211-0348/& 2015 Elsevier B.V. All rights reserved.
Cardiovascular autonomic dysfunction
105
Contents 1. 2.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.
Introduction
The interactions between autonomic dysfunction and the immune system, as well as the role of autonomic dysfunction in the pathogenesis and progression of multiple sclerosis, have been a matter of increasing interest (Cosentino and Marino, 2013). Several studies attempted to relate autonomic dysfunction of patients with multiple sclerosis to demyelinating lesions within specific regions of the central nervous system (Acevedo et al., 2000). However these data are inconclusive and the nature of the pathophysiological association between autonomic dysfunction and multiple sclerosis remains unclear (Flachenecker, 2007). Identifying autonomic dysfunction in patients with multiple sclerosis is relevant both from the clinical (i.e. diagnosis and treatment) and research (i.e. identifying potential treatment targets) perspectives (Racosta and Kremenchutzky, 2014). Several types of autonomic dysfunction have been described among multiple sclerosis(MS) patients, comprising cardiovascular (Adamec and Habek, 2013), bladder, sexual (Schairer et al., 2014), and gastrointestinal involvement (el-Maghraby et al., 2005), as well as impairment of thermoregulation (Ueno et al., 2000) and pupillary dysfunction (de Seze et al., 2001). As cardiovascular autonomic dysfunction has been correlated with landmark pathophysiological processes of MS (e.g. underlying inflammation and neurodegeneration), identifying this type of autonomic dysfunction may be of critical importance (Nasseri et al., 1999; Kodounis et al., 2005; Lorberboym et al., 2008; Vita et al., 1993). One of the issues when reviewing literature on the MS population with cardiovascular autonomic dysfunction is that the definition varies across studies and to date there is no consensus on the diagnostic criteria (Nasseri et al., 1999; Kodounis et al., 2005; Lorberboym et al., 2008; Vita et al., 1993; Anema et al., 1991; McDougall and McLeod, 2003; Ferini-Strambi et al., 1995; Frontoni et al., 1996; Flachenecker et al., 1999; Acevedo et al., 2000; Merkelbach et al., 2001; Gunal et al., 2002; LabuzRoszak and Pierzchala 2007; Hale et al., 2009; Amarenco et al., 2013). Many studies report the number of abnormal autonomic tests (Nasseri et al., 1999; Kodounis et al., 2005; Lorberboym et al., 2008; Anema et al., 1991), while others define dysautonomia as at least 2 abnormal tests (Vita et al., 1993; McDougall and McLeod 2003; Ferini-Strambi et al., 1995; Frontoni et al., 1996; Flachenecker et al., 1999; Acevedo et al., 2000; Merkelbach et al., 2001; Gunal et al., 2002; LabuzRoszak and Pierzchala 2007; Hale et al., 2009; Amarenco et al., 2013).
105 106 106 107 108 110 111
Based on studies of patients with chronic renal failure, Vita et al. suggested that cardiovascular autonomic dysfunction should be diagnosed in the presence of at least 2 abnormal autonomic tests (Vita et al., 1991). In contrast, Table 1
Search strategy and results.
Search term
Number of articles
PubMed Multiple Sclerosis[Mesh]) and “Autonomic Nervous System”[Mesh] Multiple Sclerosis[Mesh]) and “Autonomic Nervous System Diseases” [Mesh] Multiple Sclerosis[Mesh]) and “Cardiovascular System” [Mesh] Multiple Sclerosis[Mesh]) and “Sympathetic Nervous System” [Mesh] Multiple Sclerosis[Mesh]) and “Parasympathetic Nervous System” [Mesh] Multiple Sclerosis[Mesh and hearth rate variability Multiple Sclerosis[Mesh] and autonomic tests Multiple Sclerosis[Mesh] and autonomic dysfunction Multiple Sclerosis[Mesh] and cardiovascular autonomic dysfunction Multiple Sclerosis[Mesh] and sympathetic skin response
1854 167
EMBASE Multiple Sclerosis and Nervous System” Multiple Sclerosis and Dysfunction” Multiple Sclerosis and System” Multiple Sclerosis and Multiple Sclerosis and “Parasympathetic” Multiple Sclerosis and
“Autonomic
983 196
“Autonomic
143
“Cardiovascular
96
“Sympathetic”
245 61
Hearth Rate
242
Scopus Multiple Sclerosis and Autonomic Dysfunction Total
108
1118 47 64
15 56 201 56 22
288 288 3125
106
J.M. Racosta et al.
Figure 1 PRISMA flow diagram.
other authors observed that, among patients with chronic renal failure on hemodialysis, if only one abnormal test was considered for the diagnosis of cardiovascular autonomic dysfunction, the sensitivity was high (86%) and the positive predictive value was excellent (100%) (Sahin et al., 2006). Whether one or two abnormal tests should be used as the threshold for diagnosing cardiovascular autonomic dysfunction inpatients with multiple sclerosis is a matter of debate and the question remains unanswered (Acevedo et al., 2000; Vita et al., 1991). We aimed to assess the impact of using at least one vs. at least two abnormal tests for defining autonomic dysfunction on the proportion of dysautonomia among multiple sclerosis patients. We conducted a meta-analysis of studies reporting autonomic dysfunction in patients with multiple sclerosis and compared two potential definitions: (a) at least one abnormal test vs. (b) at least two abnormal tests. We also investigated the proportion of patients with sympathetic and parasympathetic dysfunction.
2.
Material and methods
To qualify for inclusion, studies had to fulfill the following criteria: (1) prospective cohort studies, (2) patients with multiple sclerosis, either with relapsing-remitting or progressive forms, diagnosed according to available criteria (Poser et al., 1983; McDonald et al., 2001; Polman et al., 2005; Polman et al., 2011), (3) use of non-invasive cardiovascular autonomic assessment: heart rate variability to deep breathing, handgrip test, tilt test (blood pressure and/or hearth rate variation), Valsalva maneuver, (4) clearly stating the number of patients with abnormal results in autonomic tests, and (5) written in English, Spanish or French. Studies reporting duplicated cohorts were excluded. This meta-analysis was done according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for reporting systematic reviews and
meta-analyses (Liberati et al., 2009). We searched PubMed, Embase, and Scopus from 1980 to December 30, 2013. Search terms and their results are shown in Table 1. Additionally, the reference lists were screened for qualifying studies. One author (JMR) assessed the eligibility of the 38 studies pre-selected in the initial title and abstract search (Figure 1), and collated and extracted the data from the qualifying studies. Extracted data was reviewed by a second author (LAS). The primary endpoint used for the meta-analysis was the proportion of patients with cardiovascular autonomic dysfunction, considering autonomic dysfunction as at least one vs. a minimum of two abnormal cardiac autonomic tests. For this purpose, we abstracted data about the number of patients undergoing autonomic testing (denominator) and the number of patients with abnormal results (numerator). With these data, we calculated the proportion of patients with multiple sclerosis diagnosed with autonomic dysfunction. We pooled data with random effects meta-analyses and reported the results as proportions and 95% confidence intervals (95% CI). Finally, we investigated the proportion of patients with sympathetic and parasympathetic dysfunction, defined as the presence of at least one abnormal test. For this purpose, we analyzed the rate abnormalities in the deep breathing test and changes in blood pressure during the tilt test, as markers of parasympathetic and sympathetic function, respectively.
2.1.
Analysis
The variance of the proportion, the arcsine transformation of the proportion, and the variance of the arcsinetransformed proportion for each study was calculated. For cohorts with no patients with tests abnormalities, the variance of the proportion was calculated assuming 0.5 observed diagnoses so that the studies could be included in the summary statistic. For the variance of the arcsine transformed values, we followed the analytical arcsine
Cardiovascular autonomic dysfunction
Table 2
107
Description of studies included in the meta-analysis.
Study and year
PrAD 1 Test n (%)
PrAD 2 tests n (%)
Sample size
Control sample
PentlandEwing 1987 Anema 1991 Vita 1993 Ferini-Strambi 1994 Frontoni 1996 Nasseri 1998 Flachenecker 1999 Acevedo 2000 Flachenecker 2001 Merkelbach 2001 Gunal 2002 McDougall 2003 Labuz-Roszak 2007 Lorberboym 2008 Hale 2009 Amarenco 2013
37 (74%)
50
16 (53%) 26 (65%) 2 (8%)
30 4 (10%) 40 0 (0%) 25
●
4 (25%) 9 (20%) 16 (40%)
1 (6%)
●
●
R–R at rest
●
TT (BP)
HT RRMS (%)
EDSS DD mean7SD mean7SD
●
●
●
●
●
NA
NA
● ● ●
● ●
● ● ●
● ● ●
●
3.872.1 3.371.3 3.87NA
5.475.1 4.376.2 9.67NA
● ● ●
● ● ●
●
●
●
●
43 68
2.171.4 3.471.6 3.271.4
7.470.2 10.077.0 NA
16 46 40
●
● ● ●
24 (60%) 17 (43%) 40 11 (42%) 26
● ●
● ●
● ●
● ●
● ●
●
75 74
3.672.1 2.570.9
NA 6.574.9
●
●
●
●
●
64
3.471.6
8.976.1
● ●
● ●
● ●
● ●
● ●
100 63
1.871.1 4.172.0
8.076.0 15.2711.5
●
●
●
●
●
46
4.072.0
6.175.5
●
●
●
70
4.071.8
9.674.8
● ●
● ●
● ●
71
3.071.3 5.17NA
9.478.6 NA
3 (8%)
●
DM VM TT (HR)
64 (76%) 23 (27%) 84 10 (45%) 3 (14%) 22 18 (29%) 11 (18%) 62
●
7 (29%)
5 (21%) 24
5 (50%)
4 (40%) 10
●
5 (16%)
4 (13%) 31 25 (38%) 65
●
●
●
● ●
●
DD: duration of the disease. DM: deep breathing maneuver. EDSS: expanded disability status scale. HT: handgrip test. MS: multiple sclerosis. PrAD 1 Test: proportion with autonomic dysfunction defined as only one abnormal test. PrAD 2 Tests: proportion with autonomic dysfunction defined as at least two abnormal tests. NA: data not available. RRMS (%): proportion with relapsing-remitting multiple sclerosis. TT (BP): tilt test (blood pressure). TT (HR): Tilt test (hearth rate). VM: Valsalva maneuver. ● means that the variable indicated in the heading of the column is present for a given study.
method as described by Rucker et al (Rücker et al., 2009). We estimated inter-study variance using the iterative method of Paule and Mandel (Paule and Mandel, 1982). We used the I2 statistic to assess heterogeneity in results of individual studies. Thresholds of heterogeneity were classified as follows: low (I2o25%), modest (I2 = 25–50%), and large (I2450%) (Higgins et al., 2003). We used funnel plots and Egger regression tests for the assessment of publication bias. A p value of o0.1 was deemed significant for publication bias in Egger tests (Egger et al., 1997). We performed the meta-analyses with Microsoft Excel (Microsoft Excel 2011 for Macintosh, Redmond, WA, USA). We used Student’s t test to compare the summary statistics for the two definitions of autonomic dysfunction. The analytic variance of the arcsine transformed proportion and the Paule and Mandel inter-study variance were computed using R 2.15.0 (R, Vienna, Austria). All tests were two-tailed and a p value of less than 0.05 was deemed statistically significant. For developing the funnel plots, we used a modified version of the spreadsheet for Microsoft Excel developed by Public Health England, available at http://www.wmpho.org.uk/ tools/ (Analytical Tools for Public Health: Funnel plot).
3.
Results
We initially identified 3125 articles. A total of 38 manuscripts were analyzed in full-text format, of which 16 fulfilled the inclusion criteria (Figure 1). The assessment of the reference lists of the 38 papers did not provide any additional articles to be included in the study. We were able to identify 16 studies (611 patients) reporting at least one abnormal test and 12 studies (459 patients) reporting a minimum of two (Table 2). All the studies were done prospectively. Nine (56.2%) were case-control studies. The weighted mean age was 38.971.3 years and proportion of males was 30.4% (95%CI 25.2–35.6). All studies excluded patients with other cardiovascular illnesses or pathologies with possible dysautonomia such as diabetes, alcoholism or collagen diseases. Twelve studies were conducted in Europe (75%), 2 in Australia (12.5%), 1 in North America (6.2%), 1 in Asia (6.2%), and no studies were done in Africa or South America. We present the results based on the arcsine-transformed proportions because it has stable variance when proportions are 0 or 100%. The meta-analyses performed directly on the
108
J.M. Racosta et al.
Figure 2 Proportion of patients with autonomic dysfunction.
proportions led to the same general conclusions. The proportion of patients with autonomic dysfunction was two-fold higher (p= 0.006) when using the definition of at least one abnormal autonomic test (42.1%, 95%CI 30.9–53.8; Q= 112, p(Q)o0.001, I2 = 87.5%) than when using the definition of at least two abnormal results (18.8%, 95% CI 9.9– 28.8; Q =46, p(Q)o0.001, I2 = 78.4%) (Figure 2). Ten studies were included in each analysis of sympathetic and parasympathetic dysfunction. There were no differences (p = 0.23) when comparing the proportion of patients with deep breathing test abnormalities (25.6%, 95% CI 13.2–40.4, Q = 101, p(Q)o0.001, I2 = 92.1%) and
abnormal changes in blood pressure during the tilt test (15.8%, 95% CI9.7–23.1; Q = 25 p(Q)o0.001, I2 = 68.3%) (Figure 3). The Egger regression test was non-significant for publication bias in any of the analyses but all the funnel plots showed considerable asymmetry (Figure 4).
4.
Discussion
The lack of consensus about the threshold for the diagnosis of autonomic dysfunction in patients with multiple
Cardiovascular autonomic dysfunction
Figure 3
109
Proportion of patients with sympathetic and parasympathetic dysfunction.
sclerosis can cause problems of comparability in the literature (e.g. it is difficult to compare treatment efficacies when the patients studied are diagnosed according to different definitions). The lack of a clear definition for autonomic dysfunction has considerable clinical implications. Using a threshold of at least two abnormal tests may lead to underdiagnosing autonomic dysfunction, preventing a number of patients from accessing potentially beneficial treatments for disabling autonomic symptoms. Conversely, using a threshold of just one abnormal autonomic test may lead to overdiagnosing autonomic dysfunction, labeling some patients in a potentially detrimental way (e.g. a Valsalva index slightly below normal value does not have that same significance as a positive Tilt test but is still considered to be an abnormal result). Thus, no threshold is completely acceptable for diagnosing autonomic dysfunction, although using the definition of at least one abnormal test is the approach that appears to have the greatest sensitivity. Better understanding of the prevalence of autonomic dysfunction according to different thresholds may help to determine how autonomic dysfunction should be defined in MS patients.
In our meta-analysis of 16 studies including 611 patients, the proportion of patients with autonomic dysfunction was two-fold higher when using a threshold of one than when using a threshold of two abnormal autonomic tests (42.1% vs. 18.8%, p= 0.006). The 42.1% proportion of cardiovascular autonomic dysfunction defined as one abnormal autonomic tests is similar to the frequency reported for autonomic cardiovascular symptoms such as orthostatic dizziness (49%) and for symptoms of dysautonomia affecting other systems, such as urinary incontinence or retention (55%) and constipation (45%) in patients with multiple sclerosis (Anema et al., 1991; Acevedo et al., 2000). Acevedo et al. proposed including an abnormal response of blood pressure to orthostatic challenge as a marker of cardiovascular autonomic dysfunction (Acevedo et al., 2000). Studies using diverse orthostatic challenges to blood pressure through Tilt Testing have reported sensitivities ranging from 49% to 66% and specificities ranging from 50% to 90% to diagnose neurally mediated syncope (Kapoor et al., 1994; Kapoor and Brant 1992; Natale et al., 1995). Because looking for postural hypotension is a readily available bedside screening maneuver, with low cost for the
110
J.M. Racosta et al. Cardiovascular Autonomic Dysfunctio n Two Abnormal Tests Egger regression test’s p=0·60
Cardiovascular Autonomic Dysfunction One Abnormal Test Egger regression test’s p=0·51 50
50
Precision (1/SE)
Precision (1/SE)
40 30 20 10
0.2
0.4
0.6
0.8
1.0
0 0.0
1.2
0.4
0.6
0.8
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Sympathetic Dysfunction Egger regression test’s p=0·61
Parasympathetic Dysfunction Egger regression test’s p=0·49
50
50
40
40
30 20
30 20 10
10 0 0.0
0.2
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Precision (1/SE)
Precision (1/SE)
0 0.0
0.2
0.4
0.6
0.8
0 0.0
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Figure 4
0.2
0.4
0.6
0.8
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Funnel plots and egger tests.
health care system, it could be used to identify patients at higher risk of having autonomic dysfunction, although the absence of orthostatic hypotension should not be used to rule out dysautonomia. This meta-analysis has several limitations. First, only studies assessing cardiovascular autonomic function were included, while studies of other forms of autonomic function (e.g. genitourinary, gastrointestinal and thermoregulatory) were not included (Adamec and Habek, 2013). The inclusion of these latter forms of autonomic dysfunction may have changed our results and should be investigated in the future. Second, as only a few studies provided detailed information about the values of the tests performed, we were not able to analyze which test was most frequently abnormal. Third, significant heterogeneity was observed in fixed-effect analysis (data not shown) requiring the use of random-effects models to account for both within- and between-study variability. Fourth, despite the thorough literature search, we may have missed relevant studies. Not limiting the search to studies published in English, Spanish and French may have minimized this potential selection bias. The present meta-analysis is the largest one to date about cardiovascular autonomic dysfunction in patients with multiple sclerosis and also comprises patients from diverse geographic areas (e.g. Europe, Oceania, North America, Asia), resulting in potentially good
generalizability. Not only the high proportion of cardiovascular autonomic dysfunction as a whole, but also the relatively high frequency of impairment of the sympathetic (15.8%) and parasympathetic (25.6%) nervous systems found in this meta-analysis should also prompt systematic screening for autonomic dysfunction in routine clinical practice of multiple sclerosis. Considering the current lack of a clear definition for the diagnosis of cardiovascular autonomic dysfunction, and based on our results showing a wide variability in the proportion of multiple sclerosis patients found to have dysautonomia with different definitions, we propose that future studies should report their results by using both thresholds. This will allow for comparisons between studies and for a greater generalizability of the results. We also call for the development of a consensus on how autonomic dysfunction should be defined in future multiple sclerosis studies and trials. This consensus should gather the opinions from experts from the multiple sclerosis and autonomic nervous system fields.
Acknowledgements None.
Cardiovascular autonomic dysfunction
References Acevedo AR, Nava C, Arriada N, Violante A, Corona T. Cardiovascular dysfunction in multiple sclerosis. Acta Neurol Scand 2000;101:85–8. Acevedo AR, Nava C, Arriada N, Violante A, Corona T. Cardiovascular dysfunction in multiple sclerosis. Acta Neurol Scand 2000;101:85–8. Adamec I, Habek M. Autonomic dysfunction in multiple sclerosis. Clin Neurol Neurosurg 2013;115(Suppl 1):S73–8. Amarenco G, Raibaut P, Hubeaux K, Jousse M, Sheikh Ismael S, Lapeyre E. Autonomic nervous system alteration in multiple sclerosis patients with urinary symptoms. Clinical, urodynamic and cardiovascular study. Prog Urol 2013;23:1505–10. Anema JR, Heijenbrok MW, Faes TJ, Heimans JJ, Lanting P, Polman CH. Cardiovascular autonomic function in multiple sclerosis. J Neurol Sci 1991;104:129–34. Cosentino M, Marino F. Adrenergic and dopaminergic modulation of immunity in multiple sclerosis: teaching old drugs new tricks? J Neuroimmune Pharmacol 2013;8:163–79. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta– analysis detected by a simple, graphical test. Br Med J 1997;315:629–34. Ferini-Strambi L, Rovaris M, Oldani A, Martinelli V, Filippi M, Smirne S, et al. Cardiac autonomic function during sleep and wakefulness in multiple sclerosis. J Neurol 1995;242:639–43. Flachenecker P. Autonomic dysfunction in Guillain-Barre syndrome and multiple sclerosis. J Neurol 2007;254(Suppl 2: II):S96–101. Flachenecker P, Wolf A, Krauser M, Hartung HP, Reiners K. Cardiovascular autonomic dysfunction in multiple sclerosis: correlation with orthostatic intolerance. J Neurol 1999;246:578–86. Frontoni M, Fiorini M, Strano S, Cerutti S, Giubilei F, Urani C, et al. Power spectrum analysis contribution to the detection of cardiovascular dysautonomia in multiple sclerosis. Acta Neurol Scand 1996;93:241–5. Gunal DI, Afsar N, Tanridag T, Aktan S. Autonomic dysfunction in multiple sclerosis: correlation with disease-related parameters. Eur Neurol 2002;48:1–5. Hale LA, Nukada H, Du Plessis LJ, Peebles KC. Clinical screening of autonomic dysfunction in multiple sclerosis. Physiother Res Int 2009;14:42–55. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J 2003;327:557–60. Kapoor WN, Brant N. Evaluation of syncope by upright tilt testing with isoproterenol. A nonspecific test. Ann Intern Med 1992;116:358–63. Kapoor WN, Smith MA, Miller NL. Upright tilt testing in evaluating syncope: a comprehensive literature review. Am J Med 1994;97:78–88. Kodounis A, Stamboulis E, Constantinidis TS, Liolios A. Measurement of autonomic dysregulation in multiple sclerosis. Acta Neurol Scand 2005;112:403–8. Labuz-Roszak B, Pierzchala K. Difficulties in the diagnosis of autonomic dysfunction in multiple sclerosis. Auton Res 2007;17:375–7. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol 2009;62:e1–34.
111 Lorberboym M, Lampl Y, Nikolov G, Sadeh M, Gilad R. I-123 MIBG cardiac scintigraphy and autonomic test evaluation in multiple sclerosis patients. J Neurol 2008;255:211–6. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–7. McDougall AJ, McLeod JG. Autonomic nervous system function in multiple sclerosis. J Neurol Sci 2003;215:79–85. Merkelbach S, Dillmann U, Kolmel C, Holz I, Muller M. Cardiovascular autonomic dysregulation and fatigue in multiple sclerosis. Mult Scler 2001;7:320–6. Nasseri K, Uitdehaag BM, van Walderveen MA, Ader HJ, Polman CH. Cardiovascular autonomic function in patients with relapsing remitting multiple sclerosis: a new surrogate marker of disease evolution? Eur J Neurol 1999;6:29–33. Natale A, Akhtar M, Jazayeri M, Dhala A, Blanck Z, Deshpande S, et al. Provocation of hypotension during head-up tilt testing in subjects with no history of syncope or presyncope. Circulation 1995;92:54–8. Paule RC, Mandel J. Consensus values and weighting factors. J Res Natl Bur Stand 1982;87:377–85. Polman CH, Reingold SC, Edan G, Filippi M, Hartung HP, Kappos L, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the McDonald Criteria. Ann Neurol 2005;58:840–6. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011;69:292–302. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983;13:227–31. Racosta JM, Kremenchutzky M. The Role of Autonomic Dysregulation from Pathophysiology to Therapeutics of Multiple Sclerosis: A Putative Novel Treatment Target? J Neurol Neurophysiol 2014;5:212. Rücker G, Schwarzer G, Carpenter J, Olkin I. Why add anything to nothing? The arcsine difference as a measure of treatment effect in meta–analysis with zero cells Stat Med 2009;28:721–38. Sahin M, Kayatas M, Urun Y, Sennaroglu E, Akdur S. Performing only one cardiovascular reflex test has a high positive predictive value for diagnosing autonomic neuropathy in patients with chronic renal failure on hemodialysis. Ren Fail 2006;28:383–7. Schairer LC, Foley FW, Zemon V, Tyry T, Campagnolo D, Marrie RA, et al. The impact of sexual dysfunction on health-related quality of life in people with multiple sclerosis. Mult Scler 2014;5:610–6. Ueno M, Tokunaga Y, Terachi S, Gondo K, Hara T. Asymmetric sweating in a child with multiple sclerosis. Pediatr Neurol 2000;23:74–6. Vita G, Princi P, Messina C. Multivariate analysis of cardiovascular reflexes applied to the diagnosis of autonomic neuropathy. J Neurol 1991;238:251–5. Vita G, Fazio MC, Milone S, Blandino A, Salvi L, Messina C. Cardiovascular autonomic dysfunction in multiple sclerosis is likely related to brainstem lesions. J Neurol Sci 1993;120:82–6. de Seze J, Arndt C, Stojkovic T, Ayachi M, Gauvrit JY, Bughin M, et al. Pupillary disturbances in multiple sclerosis: Correlation with MRI findings. J Neurol Sci 2001;188:37–41. el-Maghraby TA, Shalaby NM, Al-Tawdy MH, Salem SS. Gastric motility dysfunction in patients with multiple sclerosis assessed by gastric emptying scintigraphy. Can J Gastroenterol 2005;19:141–5.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/msard
REVIEW
Cardiovascular autonomic dysfunction in multiple sclerosis: A meta-analysis Juan Manuel Racostaa, Luciano A. Sposatoa, Sarah A. Morrowa, Lauren Ciprianob, Kurt Kimpiskia, Marcelo Kremenchutzkya,n a
Department of Clinical Neurological Sciences, London Health Sciences Centre. Western, University, London, ON, Canada b Ivey Business School, Western University, London, ON, Canada Received 26 November 2014; received in revised form 11 February 2015; accepted 13 February 2015
KEYWORDS Multiple sclerosis; Autonomic dysfunction; Diagnosis; Autonomic nervous system
Abstract Background and objective: The definition of cardiovascular autonomic dysfunction in patients with multiple sclerosis is controversial. Thus, its true prevalence is unknown. We performed a systematic review and meta-analysis to compare the proportion of patients with multiple sclerosis that would be diagnosed with cardiovascular dysautonomia using a definition of at least one abnormal cardiac autonomic test vs. at least two abnormal studies. Methods: We searched PubMed, Embase, and Scopus from 1980 to December 2013 for publications reporting abnormal autonomic tests in patients with multiple sclerosis. We performed random-effects meta-analyses for calculating the proportion of patients diagnosed with autonomic dysfunction with both definitions. Results: We included 16 studies comprising 611 patients with multiple sclerosis, assessing Z3 cardiovascular autonomic tests. The proportion of patients with autonomic dysfunction was two-fold higher (p=0.006) when using the definition of only one abnormal autonomic test (42.1%) compared to that using at least two abnormal results (18.8%). Conclusions: We found a wide variation in the proportion of patients with multiple sclerosis diagnosed with cardiovascular dysautonomia by using the two definitions. Consensus is needed to define autonomic dysfunction in patients with multiple sclerosis. In the meantime, we encourage investigators to report results using both thresholds. & 2015 Elsevier B.V. All rights reserved.
n Correspondence to: MS Clinic, University Hospital, LHSC, 339 Windermere Road, P.O. Box 5339, London ON, Canada N6A 5A5. Tel.: +1 519 663 3697; fax: +1 519 663 3744. E-mail address: [email protected] (M. Kremenchutzky).
http://dx.doi.org/10.1016/j.msard.2015.02.002 2211-0348/& 2015 Elsevier B.V. All rights reserved.
Cardiovascular autonomic dysfunction
105
Contents 1. 2.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.
Introduction
The interactions between autonomic dysfunction and the immune system, as well as the role of autonomic dysfunction in the pathogenesis and progression of multiple sclerosis, have been a matter of increasing interest (Cosentino and Marino, 2013). Several studies attempted to relate autonomic dysfunction of patients with multiple sclerosis to demyelinating lesions within specific regions of the central nervous system (Acevedo et al., 2000). However these data are inconclusive and the nature of the pathophysiological association between autonomic dysfunction and multiple sclerosis remains unclear (Flachenecker, 2007). Identifying autonomic dysfunction in patients with multiple sclerosis is relevant both from the clinical (i.e. diagnosis and treatment) and research (i.e. identifying potential treatment targets) perspectives (Racosta and Kremenchutzky, 2014). Several types of autonomic dysfunction have been described among multiple sclerosis(MS) patients, comprising cardiovascular (Adamec and Habek, 2013), bladder, sexual (Schairer et al., 2014), and gastrointestinal involvement (el-Maghraby et al., 2005), as well as impairment of thermoregulation (Ueno et al., 2000) and pupillary dysfunction (de Seze et al., 2001). As cardiovascular autonomic dysfunction has been correlated with landmark pathophysiological processes of MS (e.g. underlying inflammation and neurodegeneration), identifying this type of autonomic dysfunction may be of critical importance (Nasseri et al., 1999; Kodounis et al., 2005; Lorberboym et al., 2008; Vita et al., 1993). One of the issues when reviewing literature on the MS population with cardiovascular autonomic dysfunction is that the definition varies across studies and to date there is no consensus on the diagnostic criteria (Nasseri et al., 1999; Kodounis et al., 2005; Lorberboym et al., 2008; Vita et al., 1993; Anema et al., 1991; McDougall and McLeod, 2003; Ferini-Strambi et al., 1995; Frontoni et al., 1996; Flachenecker et al., 1999; Acevedo et al., 2000; Merkelbach et al., 2001; Gunal et al., 2002; LabuzRoszak and Pierzchala 2007; Hale et al., 2009; Amarenco et al., 2013). Many studies report the number of abnormal autonomic tests (Nasseri et al., 1999; Kodounis et al., 2005; Lorberboym et al., 2008; Anema et al., 1991), while others define dysautonomia as at least 2 abnormal tests (Vita et al., 1993; McDougall and McLeod 2003; Ferini-Strambi et al., 1995; Frontoni et al., 1996; Flachenecker et al., 1999; Acevedo et al., 2000; Merkelbach et al., 2001; Gunal et al., 2002; LabuzRoszak and Pierzchala 2007; Hale et al., 2009; Amarenco et al., 2013).
105 106 106 107 108 110 111
Based on studies of patients with chronic renal failure, Vita et al. suggested that cardiovascular autonomic dysfunction should be diagnosed in the presence of at least 2 abnormal autonomic tests (Vita et al., 1991). In contrast, Table 1
Search strategy and results.
Search term
Number of articles
PubMed Multiple Sclerosis[Mesh]) and “Autonomic Nervous System”[Mesh] Multiple Sclerosis[Mesh]) and “Autonomic Nervous System Diseases” [Mesh] Multiple Sclerosis[Mesh]) and “Cardiovascular System” [Mesh] Multiple Sclerosis[Mesh]) and “Sympathetic Nervous System” [Mesh] Multiple Sclerosis[Mesh]) and “Parasympathetic Nervous System” [Mesh] Multiple Sclerosis[Mesh and hearth rate variability Multiple Sclerosis[Mesh] and autonomic tests Multiple Sclerosis[Mesh] and autonomic dysfunction Multiple Sclerosis[Mesh] and cardiovascular autonomic dysfunction Multiple Sclerosis[Mesh] and sympathetic skin response
1854 167
EMBASE Multiple Sclerosis and Nervous System” Multiple Sclerosis and Dysfunction” Multiple Sclerosis and System” Multiple Sclerosis and Multiple Sclerosis and “Parasympathetic” Multiple Sclerosis and
“Autonomic
983 196
“Autonomic
143
“Cardiovascular
96
“Sympathetic”
245 61
Hearth Rate
242
Scopus Multiple Sclerosis and Autonomic Dysfunction Total
108
1118 47 64
15 56 201 56 22
288 288 3125
106
J.M. Racosta et al.
Figure 1 PRISMA flow diagram.
other authors observed that, among patients with chronic renal failure on hemodialysis, if only one abnormal test was considered for the diagnosis of cardiovascular autonomic dysfunction, the sensitivity was high (86%) and the positive predictive value was excellent (100%) (Sahin et al., 2006). Whether one or two abnormal tests should be used as the threshold for diagnosing cardiovascular autonomic dysfunction inpatients with multiple sclerosis is a matter of debate and the question remains unanswered (Acevedo et al., 2000; Vita et al., 1991). We aimed to assess the impact of using at least one vs. at least two abnormal tests for defining autonomic dysfunction on the proportion of dysautonomia among multiple sclerosis patients. We conducted a meta-analysis of studies reporting autonomic dysfunction in patients with multiple sclerosis and compared two potential definitions: (a) at least one abnormal test vs. (b) at least two abnormal tests. We also investigated the proportion of patients with sympathetic and parasympathetic dysfunction.
2.
Material and methods
To qualify for inclusion, studies had to fulfill the following criteria: (1) prospective cohort studies, (2) patients with multiple sclerosis, either with relapsing-remitting or progressive forms, diagnosed according to available criteria (Poser et al., 1983; McDonald et al., 2001; Polman et al., 2005; Polman et al., 2011), (3) use of non-invasive cardiovascular autonomic assessment: heart rate variability to deep breathing, handgrip test, tilt test (blood pressure and/or hearth rate variation), Valsalva maneuver, (4) clearly stating the number of patients with abnormal results in autonomic tests, and (5) written in English, Spanish or French. Studies reporting duplicated cohorts were excluded. This meta-analysis was done according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for reporting systematic reviews and
meta-analyses (Liberati et al., 2009). We searched PubMed, Embase, and Scopus from 1980 to December 30, 2013. Search terms and their results are shown in Table 1. Additionally, the reference lists were screened for qualifying studies. One author (JMR) assessed the eligibility of the 38 studies pre-selected in the initial title and abstract search (Figure 1), and collated and extracted the data from the qualifying studies. Extracted data was reviewed by a second author (LAS). The primary endpoint used for the meta-analysis was the proportion of patients with cardiovascular autonomic dysfunction, considering autonomic dysfunction as at least one vs. a minimum of two abnormal cardiac autonomic tests. For this purpose, we abstracted data about the number of patients undergoing autonomic testing (denominator) and the number of patients with abnormal results (numerator). With these data, we calculated the proportion of patients with multiple sclerosis diagnosed with autonomic dysfunction. We pooled data with random effects meta-analyses and reported the results as proportions and 95% confidence intervals (95% CI). Finally, we investigated the proportion of patients with sympathetic and parasympathetic dysfunction, defined as the presence of at least one abnormal test. For this purpose, we analyzed the rate abnormalities in the deep breathing test and changes in blood pressure during the tilt test, as markers of parasympathetic and sympathetic function, respectively.
2.1.
Analysis
The variance of the proportion, the arcsine transformation of the proportion, and the variance of the arcsinetransformed proportion for each study was calculated. For cohorts with no patients with tests abnormalities, the variance of the proportion was calculated assuming 0.5 observed diagnoses so that the studies could be included in the summary statistic. For the variance of the arcsine transformed values, we followed the analytical arcsine
Cardiovascular autonomic dysfunction
Table 2
107
Description of studies included in the meta-analysis.
Study and year
PrAD 1 Test n (%)
PrAD 2 tests n (%)
Sample size
Control sample
PentlandEwing 1987 Anema 1991 Vita 1993 Ferini-Strambi 1994 Frontoni 1996 Nasseri 1998 Flachenecker 1999 Acevedo 2000 Flachenecker 2001 Merkelbach 2001 Gunal 2002 McDougall 2003 Labuz-Roszak 2007 Lorberboym 2008 Hale 2009 Amarenco 2013
37 (74%)
50
16 (53%) 26 (65%) 2 (8%)
30 4 (10%) 40 0 (0%) 25
●
4 (25%) 9 (20%) 16 (40%)
1 (6%)
●
●
R–R at rest
●
TT (BP)
HT RRMS (%)
EDSS DD mean7SD mean7SD
●
●
●
●
●
NA
NA
● ● ●
● ●
● ● ●
● ● ●
●
3.872.1 3.371.3 3.87NA
5.475.1 4.376.2 9.67NA
● ● ●
● ● ●
●
●
●
●
43 68
2.171.4 3.471.6 3.271.4
7.470.2 10.077.0 NA
16 46 40
●
● ● ●
24 (60%) 17 (43%) 40 11 (42%) 26
● ●
● ●
● ●
● ●
● ●
●
75 74
3.672.1 2.570.9
NA 6.574.9
●
●
●
●
●
64
3.471.6
8.976.1
● ●
● ●
● ●
● ●
● ●
100 63
1.871.1 4.172.0
8.076.0 15.2711.5
●
●
●
●
●
46
4.072.0
6.175.5
●
●
●
70
4.071.8
9.674.8
● ●
● ●
● ●
71
3.071.3 5.17NA
9.478.6 NA
3 (8%)
●
DM VM TT (HR)
64 (76%) 23 (27%) 84 10 (45%) 3 (14%) 22 18 (29%) 11 (18%) 62
●
7 (29%)
5 (21%) 24
5 (50%)
4 (40%) 10
●
5 (16%)
4 (13%) 31 25 (38%) 65
●
●
●
● ●
●
DD: duration of the disease. DM: deep breathing maneuver. EDSS: expanded disability status scale. HT: handgrip test. MS: multiple sclerosis. PrAD 1 Test: proportion with autonomic dysfunction defined as only one abnormal test. PrAD 2 Tests: proportion with autonomic dysfunction defined as at least two abnormal tests. NA: data not available. RRMS (%): proportion with relapsing-remitting multiple sclerosis. TT (BP): tilt test (blood pressure). TT (HR): Tilt test (hearth rate). VM: Valsalva maneuver. ● means that the variable indicated in the heading of the column is present for a given study.
method as described by Rucker et al (Rücker et al., 2009). We estimated inter-study variance using the iterative method of Paule and Mandel (Paule and Mandel, 1982). We used the I2 statistic to assess heterogeneity in results of individual studies. Thresholds of heterogeneity were classified as follows: low (I2o25%), modest (I2 = 25–50%), and large (I2450%) (Higgins et al., 2003). We used funnel plots and Egger regression tests for the assessment of publication bias. A p value of o0.1 was deemed significant for publication bias in Egger tests (Egger et al., 1997). We performed the meta-analyses with Microsoft Excel (Microsoft Excel 2011 for Macintosh, Redmond, WA, USA). We used Student’s t test to compare the summary statistics for the two definitions of autonomic dysfunction. The analytic variance of the arcsine transformed proportion and the Paule and Mandel inter-study variance were computed using R 2.15.0 (R, Vienna, Austria). All tests were two-tailed and a p value of less than 0.05 was deemed statistically significant. For developing the funnel plots, we used a modified version of the spreadsheet for Microsoft Excel developed by Public Health England, available at http://www.wmpho.org.uk/ tools/ (Analytical Tools for Public Health: Funnel plot).
3.
Results
We initially identified 3125 articles. A total of 38 manuscripts were analyzed in full-text format, of which 16 fulfilled the inclusion criteria (Figure 1). The assessment of the reference lists of the 38 papers did not provide any additional articles to be included in the study. We were able to identify 16 studies (611 patients) reporting at least one abnormal test and 12 studies (459 patients) reporting a minimum of two (Table 2). All the studies were done prospectively. Nine (56.2%) were case-control studies. The weighted mean age was 38.971.3 years and proportion of males was 30.4% (95%CI 25.2–35.6). All studies excluded patients with other cardiovascular illnesses or pathologies with possible dysautonomia such as diabetes, alcoholism or collagen diseases. Twelve studies were conducted in Europe (75%), 2 in Australia (12.5%), 1 in North America (6.2%), 1 in Asia (6.2%), and no studies were done in Africa or South America. We present the results based on the arcsine-transformed proportions because it has stable variance when proportions are 0 or 100%. The meta-analyses performed directly on the
108
J.M. Racosta et al.
Figure 2 Proportion of patients with autonomic dysfunction.
proportions led to the same general conclusions. The proportion of patients with autonomic dysfunction was two-fold higher (p= 0.006) when using the definition of at least one abnormal autonomic test (42.1%, 95%CI 30.9–53.8; Q= 112, p(Q)o0.001, I2 = 87.5%) than when using the definition of at least two abnormal results (18.8%, 95% CI 9.9– 28.8; Q =46, p(Q)o0.001, I2 = 78.4%) (Figure 2). Ten studies were included in each analysis of sympathetic and parasympathetic dysfunction. There were no differences (p = 0.23) when comparing the proportion of patients with deep breathing test abnormalities (25.6%, 95% CI 13.2–40.4, Q = 101, p(Q)o0.001, I2 = 92.1%) and
abnormal changes in blood pressure during the tilt test (15.8%, 95% CI9.7–23.1; Q = 25 p(Q)o0.001, I2 = 68.3%) (Figure 3). The Egger regression test was non-significant for publication bias in any of the analyses but all the funnel plots showed considerable asymmetry (Figure 4).
4.
Discussion
The lack of consensus about the threshold for the diagnosis of autonomic dysfunction in patients with multiple
Cardiovascular autonomic dysfunction
Figure 3
109
Proportion of patients with sympathetic and parasympathetic dysfunction.
sclerosis can cause problems of comparability in the literature (e.g. it is difficult to compare treatment efficacies when the patients studied are diagnosed according to different definitions). The lack of a clear definition for autonomic dysfunction has considerable clinical implications. Using a threshold of at least two abnormal tests may lead to underdiagnosing autonomic dysfunction, preventing a number of patients from accessing potentially beneficial treatments for disabling autonomic symptoms. Conversely, using a threshold of just one abnormal autonomic test may lead to overdiagnosing autonomic dysfunction, labeling some patients in a potentially detrimental way (e.g. a Valsalva index slightly below normal value does not have that same significance as a positive Tilt test but is still considered to be an abnormal result). Thus, no threshold is completely acceptable for diagnosing autonomic dysfunction, although using the definition of at least one abnormal test is the approach that appears to have the greatest sensitivity. Better understanding of the prevalence of autonomic dysfunction according to different thresholds may help to determine how autonomic dysfunction should be defined in MS patients.
In our meta-analysis of 16 studies including 611 patients, the proportion of patients with autonomic dysfunction was two-fold higher when using a threshold of one than when using a threshold of two abnormal autonomic tests (42.1% vs. 18.8%, p= 0.006). The 42.1% proportion of cardiovascular autonomic dysfunction defined as one abnormal autonomic tests is similar to the frequency reported for autonomic cardiovascular symptoms such as orthostatic dizziness (49%) and for symptoms of dysautonomia affecting other systems, such as urinary incontinence or retention (55%) and constipation (45%) in patients with multiple sclerosis (Anema et al., 1991; Acevedo et al., 2000). Acevedo et al. proposed including an abnormal response of blood pressure to orthostatic challenge as a marker of cardiovascular autonomic dysfunction (Acevedo et al., 2000). Studies using diverse orthostatic challenges to blood pressure through Tilt Testing have reported sensitivities ranging from 49% to 66% and specificities ranging from 50% to 90% to diagnose neurally mediated syncope (Kapoor et al., 1994; Kapoor and Brant 1992; Natale et al., 1995). Because looking for postural hypotension is a readily available bedside screening maneuver, with low cost for the
110
J.M. Racosta et al. Cardiovascular Autonomic Dysfunctio n Two Abnormal Tests Egger regression test’s p=0·60
Cardiovascular Autonomic Dysfunction One Abnormal Test Egger regression test’s p=0·51 50
50
Precision (1/SE)
Precision (1/SE)
40 30 20 10
0.2
0.4
0.6
0.8
1.0
0 0.0
1.2
0.4
0.6
0.8
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Sympathetic Dysfunction Egger regression test’s p=0·61
Parasympathetic Dysfunction Egger regression test’s p=0·49
50
50
40
40
30 20
30 20 10
10 0 0.0
0.2
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Precision (1/SE)
Precision (1/SE)
0 0.0
0.2
0.4
0.6
0.8
0 0.0
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Figure 4
0.2
0.4
0.6
0.8
Arcsin Transformation of the Square Root of the Proportion with Autonomic Dysfunction
Funnel plots and egger tests.
health care system, it could be used to identify patients at higher risk of having autonomic dysfunction, although the absence of orthostatic hypotension should not be used to rule out dysautonomia. This meta-analysis has several limitations. First, only studies assessing cardiovascular autonomic function were included, while studies of other forms of autonomic function (e.g. genitourinary, gastrointestinal and thermoregulatory) were not included (Adamec and Habek, 2013). The inclusion of these latter forms of autonomic dysfunction may have changed our results and should be investigated in the future. Second, as only a few studies provided detailed information about the values of the tests performed, we were not able to analyze which test was most frequently abnormal. Third, significant heterogeneity was observed in fixed-effect analysis (data not shown) requiring the use of random-effects models to account for both within- and between-study variability. Fourth, despite the thorough literature search, we may have missed relevant studies. Not limiting the search to studies published in English, Spanish and French may have minimized this potential selection bias. The present meta-analysis is the largest one to date about cardiovascular autonomic dysfunction in patients with multiple sclerosis and also comprises patients from diverse geographic areas (e.g. Europe, Oceania, North America, Asia), resulting in potentially good
generalizability. Not only the high proportion of cardiovascular autonomic dysfunction as a whole, but also the relatively high frequency of impairment of the sympathetic (15.8%) and parasympathetic (25.6%) nervous systems found in this meta-analysis should also prompt systematic screening for autonomic dysfunction in routine clinical practice of multiple sclerosis. Considering the current lack of a clear definition for the diagnosis of cardiovascular autonomic dysfunction, and based on our results showing a wide variability in the proportion of multiple sclerosis patients found to have dysautonomia with different definitions, we propose that future studies should report their results by using both thresholds. This will allow for comparisons between studies and for a greater generalizability of the results. We also call for the development of a consensus on how autonomic dysfunction should be defined in future multiple sclerosis studies and trials. This consensus should gather the opinions from experts from the multiple sclerosis and autonomic nervous system fields.
Acknowledgements None.
Cardiovascular autonomic dysfunction
References Acevedo AR, Nava C, Arriada N, Violante A, Corona T. Cardiovascular dysfunction in multiple sclerosis. Acta Neurol Scand 2000;101:85–8. Acevedo AR, Nava C, Arriada N, Violante A, Corona T. Cardiovascular dysfunction in multiple sclerosis. Acta Neurol Scand 2000;101:85–8. Adamec I, Habek M. Autonomic dysfunction in multiple sclerosis. Clin Neurol Neurosurg 2013;115(Suppl 1):S73–8. Amarenco G, Raibaut P, Hubeaux K, Jousse M, Sheikh Ismael S, Lapeyre E. Autonomic nervous system alteration in multiple sclerosis patients with urinary symptoms. Clinical, urodynamic and cardiovascular study. Prog Urol 2013;23:1505–10. Anema JR, Heijenbrok MW, Faes TJ, Heimans JJ, Lanting P, Polman CH. Cardiovascular autonomic function in multiple sclerosis. J Neurol Sci 1991;104:129–34. Cosentino M, Marino F. Adrenergic and dopaminergic modulation of immunity in multiple sclerosis: teaching old drugs new tricks? J Neuroimmune Pharmacol 2013;8:163–79. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta– analysis detected by a simple, graphical test. Br Med J 1997;315:629–34. Ferini-Strambi L, Rovaris M, Oldani A, Martinelli V, Filippi M, Smirne S, et al. Cardiac autonomic function during sleep and wakefulness in multiple sclerosis. J Neurol 1995;242:639–43. Flachenecker P. Autonomic dysfunction in Guillain-Barre syndrome and multiple sclerosis. J Neurol 2007;254(Suppl 2: II):S96–101. Flachenecker P, Wolf A, Krauser M, Hartung HP, Reiners K. Cardiovascular autonomic dysfunction in multiple sclerosis: correlation with orthostatic intolerance. J Neurol 1999;246:578–86. Frontoni M, Fiorini M, Strano S, Cerutti S, Giubilei F, Urani C, et al. Power spectrum analysis contribution to the detection of cardiovascular dysautonomia in multiple sclerosis. Acta Neurol Scand 1996;93:241–5. Gunal DI, Afsar N, Tanridag T, Aktan S. Autonomic dysfunction in multiple sclerosis: correlation with disease-related parameters. Eur Neurol 2002;48:1–5. Hale LA, Nukada H, Du Plessis LJ, Peebles KC. Clinical screening of autonomic dysfunction in multiple sclerosis. Physiother Res Int 2009;14:42–55. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. Br Med J 2003;327:557–60. Kapoor WN, Brant N. Evaluation of syncope by upright tilt testing with isoproterenol. A nonspecific test. Ann Intern Med 1992;116:358–63. Kapoor WN, Smith MA, Miller NL. Upright tilt testing in evaluating syncope: a comprehensive literature review. Am J Med 1994;97:78–88. Kodounis A, Stamboulis E, Constantinidis TS, Liolios A. Measurement of autonomic dysregulation in multiple sclerosis. Acta Neurol Scand 2005;112:403–8. Labuz-Roszak B, Pierzchala K. Difficulties in the diagnosis of autonomic dysfunction in multiple sclerosis. Auton Res 2007;17:375–7. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol 2009;62:e1–34.
111 Lorberboym M, Lampl Y, Nikolov G, Sadeh M, Gilad R. I-123 MIBG cardiac scintigraphy and autonomic test evaluation in multiple sclerosis patients. J Neurol 2008;255:211–6. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–7. McDougall AJ, McLeod JG. Autonomic nervous system function in multiple sclerosis. J Neurol Sci 2003;215:79–85. Merkelbach S, Dillmann U, Kolmel C, Holz I, Muller M. Cardiovascular autonomic dysregulation and fatigue in multiple sclerosis. Mult Scler 2001;7:320–6. Nasseri K, Uitdehaag BM, van Walderveen MA, Ader HJ, Polman CH. Cardiovascular autonomic function in patients with relapsing remitting multiple sclerosis: a new surrogate marker of disease evolution? Eur J Neurol 1999;6:29–33. Natale A, Akhtar M, Jazayeri M, Dhala A, Blanck Z, Deshpande S, et al. Provocation of hypotension during head-up tilt testing in subjects with no history of syncope or presyncope. Circulation 1995;92:54–8. Paule RC, Mandel J. Consensus values and weighting factors. J Res Natl Bur Stand 1982;87:377–85. Polman CH, Reingold SC, Edan G, Filippi M, Hartung HP, Kappos L, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the McDonald Criteria. Ann Neurol 2005;58:840–6. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011;69:292–302. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, et al. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 1983;13:227–31. Racosta JM, Kremenchutzky M. The Role of Autonomic Dysregulation from Pathophysiology to Therapeutics of Multiple Sclerosis: A Putative Novel Treatment Target? J Neurol Neurophysiol 2014;5:212. Rücker G, Schwarzer G, Carpenter J, Olkin I. Why add anything to nothing? The arcsine difference as a measure of treatment effect in meta–analysis with zero cells Stat Med 2009;28:721–38. Sahin M, Kayatas M, Urun Y, Sennaroglu E, Akdur S. Performing only one cardiovascular reflex test has a high positive predictive value for diagnosing autonomic neuropathy in patients with chronic renal failure on hemodialysis. Ren Fail 2006;28:383–7. Schairer LC, Foley FW, Zemon V, Tyry T, Campagnolo D, Marrie RA, et al. The impact of sexual dysfunction on health-related quality of life in people with multiple sclerosis. Mult Scler 2014;5:610–6. Ueno M, Tokunaga Y, Terachi S, Gondo K, Hara T. Asymmetric sweating in a child with multiple sclerosis. Pediatr Neurol 2000;23:74–6. Vita G, Princi P, Messina C. Multivariate analysis of cardiovascular reflexes applied to the diagnosis of autonomic neuropathy. J Neurol 1991;238:251–5. Vita G, Fazio MC, Milone S, Blandino A, Salvi L, Messina C. Cardiovascular autonomic dysfunction in multiple sclerosis is likely related to brainstem lesions. J Neurol Sci 1993;120:82–6. de Seze J, Arndt C, Stojkovic T, Ayachi M, Gauvrit JY, Bughin M, et al. Pupillary disturbances in multiple sclerosis: Correlation with MRI findings. J Neurol Sci 2001;188:37–41. el-Maghraby TA, Shalaby NM, Al-Tawdy MH, Salem SS. Gastric motility dysfunction in patients with multiple sclerosis assessed by gastric emptying scintigraphy. Can J Gastroenterol 2005;19:141–5.
