|

Received: 12 February 2021    Accepted: 22 April 2021

DOI: 10.1111/ene.14885

ORIGINAL ARTICLE

Prognostic value of nerve ultrasonography: A prospective

multicenter study on the natural history of chronic
inflammatory neuropathies

Johan A. Telleman1,2  | Ingrid J. T. Herraets1,2 | Hendrik Stephan Goedee2 |

Ruben P. A. van Eijk2,3  | Camiel Verhamme4 | Filip Eftimov4 | Doris Lieba-­Samal5 |
Jan Thies van Asseldonk1 | Leonard H. van den Berg2 | Willem Ludo van der Pol2 |
Leo H. Visser1

Abstract
1
Department of Neurology and Clinical
Neurophysiology, Elisabeth-­Tweesteden
Hospital, Tilburg, The Netherlands Background and objective: Nerve ultrasound is a promising new tool in chronic inflam-
2
Department of Neurology and matory neuropathies. The aim of this study was to determine its prognostic value in a
Neurosurgery, UMC Utrecht Brain
prospective multicenter cohort study including incident and prevalent patients with CIDP
Center, University Utrecht, Utrecht, The
Netherlands and MMN.
Methods: We enrolled 126 patients with CIDP, and 72 with MMN; 71 were treatment-­
3
Biostatistics & Research Support, Julius
Center for Health Sciences and Primary
Care, University Medical Center Utrecht,
naive. Patients with chronic idiopathic axonal polyneuropathy (CIAP; n = 35) were con-
Utrecht, The Netherlands sidered as disease controls. Standardized neurological examination, questionnaires, and
4
Amsterdam University Medical Center, nerve ultrasonography were obtained at time of inclusion and 1-­year follow-­up. Nerve
Amsterdam Neuroscience, University of
Amsterdam, Amsterdam, The Netherlands size development over time and correlation between nerve size and clinical outcome
5
Department of Neurology, Medical measures were determined using linear mixed effects models.
University of Vienna, Vienna, Austria
Results: Nerve size development over time was heterogeneous. Only in MMN was there
Correspondence a correlation between C5 nerve root size and deterioration of grip strength (−1.3 kPa/
Johan A. Telleman, Department of
mm2 (95% confidence interval [CI] −2.3 to −0.2). No other significant correlations be-
Neurology, Elisabeth-­Tweesteden
Hospital, Hilvarenbeekse Weg 60, 5022 tween nerve size and clinical outcome measures were found. In MMN, presence of nerve
GC, Tilburg, The Netherlands.
enlargement at inclusion predicted deterioration of grip strength, and MMN patients with
Email: [email protected]
enlargement confined to the brachial plexus seemed to have more favorable outcomes.
Funding information
No other predictive effects of sonographic nerve size were found.
Funding was received from ZonMw,
Xperiment Topzorg (project number Conclusions: The present study indicates that the natural course of nerve size develop-
842003002). The funding source had
ment in CIDP and MMN is heterogeneous, and that the prognostic value of sonographic
no involvement in study design, data
collection, analysis, interpretation of data, nerve enlargement is limited. It had some predictive effect in patients with MMN. Further
writing of the report or the decision to
research in specific subgroups of chronic inflammatory neuropathy is necessary to deter-
submit the article for publication.
mine the usefulness of nerve ultrasonography after the diagnostic phase.

KEYWORDS
chronic inflammatory demyelinating polyneuropathy, multifocal motor neuropathy, nerve
ultrasonography, prognostic value

Telleman and Herraets contributed equally to the manuscript.

The STROBE Checklist is included in Appendix S1.

© 2021 European Academy of Neurology     2327

Eur J Neurol. 2021;28:2327–2338. wileyonlinelibrary.com/journal/ene |
 |
2328       TELLEMAN et al.

I NTRO D U C TI O N Societies/Peripheral Nerve Society (EFNS/PNS) criteria (possible,

probable or definite) [16,17], or (2b) a diagnosis of CIDP or MMN
Nerve ultrasonography is emerging as a low-­cost, practical and based on our previously described diagnostic criteria, but not fulfill-
widely available tool for the evaluation of peripheral nerve disor- ing the EFNS/PNS criteria, i.e., patients with a clinical phenotype of
ders [1]. Recent studies showed that it is a useful tool to diagnose CIDP or MMN, ultrasonography results suggestive of CIDP or MMN
chronic inflammatory neuropathies (CINs), such as chronic inflam- and objective treatment effect [4,5], or (2c) a diagnosis of CIAP ac-
matory demyelinating polyneuropathy (CIDP) and multifocal motor cording to previously published clinical criteria, nerve conduction
neuropathy (MMN), and that it can even improve detection of these study results and laboratory testing [18]. Patients in category 2b
potentially treatable neuropathies [1–­5]. We also found that it has were considered eligible for inclusion, as we found in our previous
low interobserver variability and that it can be used reliably in a mul- studies that clinical characteristics of these patients did not differ
ticenter setting [6]. Several exploratory studies indicated that nerve significantly from patients in category 2a [4,5]. A diagnosis of CIDP
ultrasonography could also predict disease course and treatment or MMN was made based on thorough evaluation, and patients had
response in CIN [7–­11]. To date, only few predictors of treatment re- to fulfill strict diagnostic criteria. In our clinics, these patients receive
sponse have been identified in these neuropathies, including extent follow-­up similar to that of patients in category 2a, and were incor-
of axonal damage, presence of conduction blocks and prolonged dis- porated in this study to ensure inclusion of all consecutive patients
ease duration before start of treatment [12–­15]. New sensitive prog- with CIDP and MMN.
nostic tools would be helpful, since patients with CIN often require Exclusion criteria for this study were history of polyneuropathy
long-­term, expensive treatment with immunoglobulins. However, other than CIDP, MMN and CIAP, and physical inability to undergo
large prospective studies on the prognostic value of nerve ultraso- nerve ultrasonography.
nography in prevalent and incident patients with CIDP, MMN are
lacking. Therefore, we performed a prospective multicenter cohort
study in both treatment-­naive and treated patients with CIDP and Study design
MMN. We explored the natural course of sonographic nerve size
development and investigated the prognostic value of sonographic We obtained questionnaires, and all patients underwent standard-
nerve enlargement. ized neurological examination, and nerve ultrasonography at inclu-
sion and after 1 year of follow-­up (Figure 1; Table S1). A follow-­up
visit at 6 months could be performed in treatment-­naive CIDP and
M E TH O D S MMN to document potential earlier nerve size changes after start
of treatment. Muscle strength of 14 muscle groups was graded bi-
This international prospective longitudinal multicenter cohort study laterally with the Medical Research Council (MRC) scale [4], and grip
was conducted between May 2015 and May 2018 at the Neurology strength was determined in kilopascal (kPa) with Martin Vigorimetry
outpatient clinics of two tertiary referral centers and a large general (Martin Medizintechnik). Sensory functions were tested bilaterally
teaching hospital in the Netherlands, and a tertiary referral center with the modified INCAT Sensory Sum Score (mISS). In addition,
in Austria. The study was approved by the Brabant Regional Ethics the INCAT Overall Disability Sum Score (ODSS), Rasch-­built Overall
Committee (NL50375.028.14) and the boards of all participating Disability Scale (RODS; for CIDP), [19] and modified Rankin Scale
hospitals. The study conformed with the World Medical Association (mRS; for CIAP) were obtained.
Declaration of Helsinki. All participants gave written informed Nerve ultrasonography was performed by investigators with
consent. ≥1  year of experience with nerve ultrasonography, who were
blinded to results of previous nerve ultrasonography investiga-
tions. Nerve ultrasonography was performed with a Philips EPIQ7
Inclusion and exclusion criteria (Philips Medical Instruments, Bothell, WA) at the UMC Utrecht, an
Esaote MyLabTwice (Esaote, Genoa, Italy) at the Amsterdam UMC,
Consecutive incident and prevalent patients with CIDP and MMN a Toshiba Xario XG (Toshiba, Tokyo, Japan) at the ETZ Tilburg,
were eligible for inclusion, as well as patients with chronic idiopathic and a GE Logiq E9 Platform (GE Healthcare, Chicago, USA) at the
axonal polyneuropathy (CIAP). These patients were included as a Allgemeines Krankenhaus in Vienna. All investigators used high-­
control group, as we hypothesized that CIAP generally shows no frequency probes (5–­18  MHz). Nerve cross-­sectional area (CSA)
nerve enlargement and that nerves would, therefore, not alter over was measured bilaterally within the hyperechoic rim at standard-
time. The cohort of patients investigated in this study partially con- ized sites in upper extremity nerves (Table S1): the median nerve
sisted of patients who had previously participated in our studies (including the forearm, and upper arm), ulnar nerve, and the brachial
on the diagnostic value of nerve ultrasonography in suspected CIN plexus (including the C5 nerve root). Investigators were allowed to
[4,5]. perform nerve ultrasonography on their regular ultrasound device
Inclusion criteria were: (1) age ≥18 years and (2a) a diagnosis of with the standard settings applied in their center, as a previous
CIDP or MMN according to the European Federation of Neurological study showed that a multicenter setting and the use of different
PROGNOSTIC VALUE OF ULTRASONOGRAPHY IN CIDP AND MMN |
      2329

F I G U R E 1  Flowchart. Standardized
work-­up applied in this study, Follow-up Inclusion 6 Months 1 Year
including an optional 6-­month follow-­
up visit for incident patients with
Work-up Standard visit
- Patient History
Optional Visit
- Patient History
Standard visit
- Patient History
- Questionnaires - Questionnaires - Questionnaires
chronic inflammatory demyelinating - Physical Examination - Physical Examination - Physical Examination
- Nerve Ultrasound - Nerve Ultrasound - Nerve Ultrasound
polyneuropathy (CIDP) and multifocal - Nerve Conduction

motor neuropathy (MMN). The figure Diagnosis Studies (treatment-naive

only)
additionally shows the number of included
and excluded patients, and loss to Typical N = 52
follow-­up. CIAP, chronic idiopathic axonal CIDP Treatment-naive: 18
N = 12 N = 47
polyneuropathy [Colour figure can be
viewed at wileyonlinelibrary.com] Atypical N = 74
CIDP Treatment-naive: 30
N = 21 N = 65

MMN
N = 72 N = 13 N = 67
Treatment-naive: 23

CIAP
N = 35 N = 31

Excluded: N = 4 Loss to follow-up: N = 23

Misdiagnosis: - No reaction to calls / email (N = 10)

- Benign fasciculation syndrome (N =1) - No visit due to personal reasons (N = 5)
- Progressive spinal muscular atrophy (N =1) - Deceased (N = 4): Colon carcinoma,
- Status after Guillain-Barré Syndrome (N=1) ischemic CVA, pancreatic carcinoma,
- Polyneuropathy with unknown cause (N =1) metastatic disease of unknown origin
- Comorbidity (N = 3): Dementia, myocardial
infarction, oesophageal carcinoma
- Withdrew consent (N = 1)

types of ultrasound machines did not affect interobserver reliability nerve CSA of the median nerve at the forearm and upper arm and
[6]. However, measurements were performed at a standard window C5 nerve root for analyses as these sites show relatively low inter-
depth of 2 cm (unless nerves were located deeper), and investigators observer variability, and these are obtained on a standard basis in
were not allowed to use zoom function as this may increase variabil- our short diagnostic protocol to detect CIDP and MMN (the Dutch
ity. Moreover, patients were investigated on the same ultrasound Ultrasound Protocol for Polyneuropathy [DUP-­P]) [4–­6]. We did not
device during the length of the study. include other nerve sites in our analyses as we aimed to avoid mul-
tiple testing and to determine the prognostic value of a short ultra-
sonography protocol used in routine work-­up. We compared mean
Statistics nerve CSA at all investigated nerve sites between disease groups
using Kruskal–­Wallis and post hoc Mann–­Whitney U-­tests.
We used SPSS version 25 (SPSS Inc.) for statistical analysis. Data To determine the natural course of sonographic nerve enlarge-
were analyzed for four different disease groups: typical CIDP; atypi- ment, i.e., the development of nerve size over time, linear mixed
cal CIDP (based on the clinical criteria of the EFNS/PNS, the atypical effect models (LMEs) were fitted, where nerve CSA served as the
CIDP group includes pure motor, pure sensory, asymmetrical, and outcome measure, and study duration (in months) was the fixed ef-
distal predominant variants); MMN; and CIAP [16–­18]. A distinction fect. The random part contained a random intercept and slope (for
between typical and atypical CIDP was made because we hypoth- study duration) per individual in order to correct for variability in
esized that patients with typical CIDP could have a more uniform nerve size development due to individual patient characteristics.
pathophysiological mechanism underlying nerve enlargement. Models were fitted with restricted maximum likelihood based on an
Different atypical CIDP variants were analyzed separately in a post unstructured covariance matrix.
hoc analysis. Patients with treatment-­naive CIDP and MMN were To determine whether nerve size could serve as a marker of
also analyzed separately from prevalent patients to determine poten- disease activity, we studied associations between nerve CSA and
tial effects in the early phases of treatment. Data were summarized vigorimetry (as primary outcome). ODSS (for all disease groups),
per disease group as mean (standard deviation [SD]) for normally dis- RODS score CIDP (for CIDP), mRS score (for CIAP) and mISS (for
tributed variables, median (range) for non-­normal distributed vari- CIDP and CIAP) served as secondary outcome measures. MRC
ables, and n (%) for categorical variables. The mean nerve CSA value Sum Score was omitted from these analyses because of extreme
(of right and left side combined) was used in all analyses. We used skewness of the data. The relationship between clinical outcome
2330      | TELLEMAN et al.

TA B L E 1  Baseline characteristics

Typical CIDP Atypical CIDP MMN CIAP

(n = 52) (n = 74) (n = 72) (n = 35)

Age, years 60.3 ± 14.0 59.0 ± 13.0 53.6 ± 10.7 63.5 ± 8.9

Sex: male/female, n (%) 35 (67)/17 (33) 52 (70)/22 (30) 57 (79)/15 (21) 20 (57)/15 (43)
Disease duration, months 29 (1–­360) 50 (2–­312) 72 (3–­550) 60 (10–­240)
Treatment-­naïve, n (%) 18 (35) 30 (41) 23 (32) -­
EFNS/PNS criteria, n (%)
Definite 39 (75) 64 (86) 44 (61) -­
Probable 3 (6) 2 (3) 12 (17) -­
Possible 1 (2) 0 (0) 16 (22) -­
Not fulfilled 9 (17) 8 (11) 0 (0) -­
1-­year follow-­up completed: y/n, 47 (89)/5 (11) 65 (88)/9 (12) 67 (93)/5 (7) 31 (89)/4 (11)
n (%)
Treatment received during 1-­year follow-­up period, n (%)
IVIg 22 (47) 40 (62) 66 (99) -­
Corticosteroids 5 (11) 5 (8) -­ -­
IVIg +corticosteroids 10 (21) 9 (14) -­ -­
Plasmaferesis 1 (2) 3 (5) 0 (0) -­
No treatment: in remission 6 (13) 10 (15) 1 (1) -­
No treatment: no remission 4 (9) 1 (2) 0 (0) -­
Outcome measure development over 1-­year follow-­up period: all patients/treatment-­naive patients
Vigorimetry at inclusion, kPa 58.6 ± 25.6 62.0 ± 27.0 74.2 ± 31.9 80.9 ± 32.3
58.6 ± 27.1 60.5 ± 29.4 87.9 ± 28.6 -­
Vigorimetry change, kPa +15.8 ± 24.7 +1.0 ± 20.2 −1.1 ± 16.2 1.2 ± 13.4
+12.7 ± 19.3 ±5.0 ± 13.7 −0.9 ± 20.2 -­
ODSS at inclusiona 4.1 ± 2.1 3.5 ± 1.9 3.0 ± 1.4 1.9 ± 1.3
4.6 ± 1.2 3.2 ± 2.2 2.7 ± 1.2 -­
ODSS changea −1.3 ± 2.3 −0.3 ± 1.8 −0.3 ± 1.2 0.3 ± 1.1
−1.7 ± 1.5 −0.3 ± 1.9 −0.5 ± 1.3 -­
MRC Sum Score at inclusionb 129 (56–­140) 130 (60–­140) 133 (70–­140) 140 (117–­140)
121 (85–­134) 129 (74–­140) 134 (116–­139) -­
MRC Sum Score development +2 (−13 to 75) +1 (−24 to 50) 0 (−11 to 13) 0 (−9 to 3)
+8.5 (−13 to 28) +1.5 (−16 to 50) 0.5 (−11 to 13) -­
mISS at inclusionc 9 (0–­3 0) 9 (0–­29) -­ 9 (2–­21)
10.5 (0–­3 0) 7 (0–­29) -­
mISS change −3 (−17 to 7) −0.5 (−18 to 12) -­ −0.5 (−16 to 8)
−4 (−17 to 1) 0 (−12 to 5)
RODS score CIDP at inclusiond 32 (10–­48) 37 (3–­48) -­ -­
31 (11–­43) 35 (3–­48)
RODS score CIDP change, % +4.5 (−19 to 61) +2.0 (−38 to 61) -­ -­
+10.5 (−2 to 45) 0.0 (−11 to 45)
mRS score at inclusione -­ -­ -­ 1 (0–­3)
-­
mRS score change -­ -­ -­ 0 (−1 to 1)

Values are mean ± SD, or median (range), unless otherwise indicated. Table shows the baseline characteristics of 233 included patients per disease
group. In addition, details on the treatment received by patients completing 1-­year follow-­up during this year, and the development of several
outcome measures over the 1-­year follow-­up period are shown.
Abbreviations: CIAP, chronic idiopathic axonal polyneuropathy; CIDP, chronic inflammatory demyelinating polyneuropathy; IVIg, intravenous
immunoglobulins; kPa, kilopascal; mISS, modified INCAT Sensory Sum Score, MMN, multifocal motor neuropathy; MRC, Medical Research Council;
mRS, modified Rankin Scale; ODSS, INCAT Overall Disability Sum Score; RODS, Rasch-­built Overall Disability Scale.
a
Score can range from 0 to 10. bScore can range from 0 to 140. cScore can range from 0 to 30.
d
Score can range from 0 to 48. eScore can range from 0 to 6
PROGNOSTIC VALUE OF ULTRASONOGRAPHY IN CIDP AND MMN |
      2331

F I G U R E 2  Nerve size at inclusion and

follow-­up. Boxplots of the median nerve
at the forearm (a) and upper arm (b) and
the C5 nerve root (c) size in mm2 per
disease group at inclusion and at 1-­year
follow-­up. Nerve size at inclusion is shown
in blue, and at 1-­year follow-­up in green.
Dotted lines represent the cut-­off value
for demyelination established in our
previously published diagnostic cohort
study (the Dutch Ultrasound Protocol
for Polyneuropathy [DUP-­P]) [5]. CIAP,
chronic idiopathic axonal polyneuropathy;
CIDP, chronic inflammatory demyelinating
polyneuropathy; CSA, cross-­sectional
area; MMN, multifocal motor neuropathy
[Colour figure can be viewed at
wileyonlinelibrary.com]
 |
2332       TELLEMAN et al.

F I G U R E 3  Development of nerve size over time. This figure shows the development of nerve size of the median nerve at the forearm and
upper arm and the C5 nerve root over time in typical chronic inflammatory demyelinating polyneuropathy (CIDP), atypical CIDP, multifocal
motor neuropathy (MMN), and chronic idiopathic axonal polyneuropathy (CIAP). Development is shown for individual patients as well as
an estimated overall nerve size development. Gray lines represent individual patients, red lines represent overall nerve size development
over time, red dotted lines represent the 95% confidence interval (CI) of the nerve size development over time. After mixed model analysis,
no significant correlation between time and nerve size was found, except for the median nerve at the forearm in atypical CIDP (slope –­
0.071 mm2/month [95% CI −0.121 to −0.020]) and MMN (slope −0.057 mm2/month [95% CI −0.100 to −0.014]). CSA, cross-­sectional area
[Colour figure can be viewed at wileyonlinelibrary.com]

measures and nerve CSA was also assessed using LMEs. Each To determine the prognostic value of nerve ultrasonography,
model contained a random intercept per individual and nerve CSA measurements were labelled as either enlarged (1) or not enlarged
as a fixed effect. (0) at the inclusion visit for the investigated nerve sites [4,20]. These
PROGNOSTIC VALUE OF ULTRASONOGRAPHY IN CIDP AND MMN |
      2333

results were entered into an LME as a fixed effect. Study duration significant increase in nerve size over time in patients with increased
(in months), and the interaction between study duration and pres- CSA at inclusion (slope 0.31 mm2/month; 95% CI 0.08–­0.54), while
ence of enlargement at inclusion were also entered as fixed effects this was not the case for patients with normal nerve CSA at inclusion.
to determine whether the development over time depended on the
presence of nerve enlargement. A random intercept and slope for
study duration were entered for patients. Treatment-­naive patients
To determine the presence of other potential prognostic fac-
tors in CIDP and MMN, patients were dichotomized as either hav- When investigating treatment-­naive patients separately, only in
ing decreased or increased (i.e. change larger than 0) vigorimetry or MMN a reduction in size of the median nerve at the forearm was ob-
ODSS at 1 year of follow-­up. Differences in clinical and sonographic served (slope −0.119 mm2/month, 95% CI −0.181 to −0.056). Nerve
variables between these groups were tested using the independent size in CIDP and at the median nerve at the upper arm and C5 nerve
t-­test (continuous, normal), the Mann–­Whitney U-­test (continuous, root in MMN did not change significantly over time.
non-­normal), chi-­squared test (categorical) or Fishers’ exact test (cat-
egorical, small sample size).
Nerve size as a marker of disease activity

R E S U LT S We found no correlation between nerve size and grip strength in (a)

typical CIDP and CIAP (Table  2). In MMN, we observed a correla-
Study population tion between an increased CSA of the C5 nerve root and deteriora-
tion of grip strength (slope −1.3 kPa/mm2 [95% CI −2.3 to −0.2]; p =
A total of 237 patients were included (Figure 1). Four patients were 0.02). This indicates that grip strength decreases with 1.3  kPa for
excluded from the final analyses because of a changed diagnosis dur- each mm2 increase in CSA at the C5 nerve root. The correlation was
ing follow-­up. The baseline characteristics of 233 patients are shown more pronounced in treatment-­naive patients (slope −3.8 kPa/mm2
in Table 1. The 1-­year follow-­up visit was completed by 210 patients [95% CI −6.7–­−0.9]; p = 0.01). It was also observed in patients with
(90.1%); 23 were lost to follow-­up (9.9%). There were no significant pure motor CIDP (n = 11, slope −4.8 kPa/mm2 [95% CI −8.3 to −1.4];
differences in age, sex, disease type, disease duration, or treatment p = 0.01). There was no significant correlation of nerve CSA of the
status between the groups that completed 1 year of follow-­up and median nerve at the forearm and upper arm with grip strength in
that were lost to follow-­up. MMN (Table 2). Additionally, we did not find any significant correla-
tion with secondary outcome measures (ODSS, RODS score, mRS
score and mISS) in (a)typical CIDP, MMN or CIAP.
Natural course of sonographic nerve size

All patients Prognostic value of nerve enlargement at baseline

Nerve CSA of the median nerve at the forearm, upper arm and at Presence of enlargement of the median nerve at the upper arm at
the C5 nerve root was significantly higher in CIDP and MMN pa- inclusion predicted deterioration of grip strength in MMN (Figure 4,
tients than in CIAP patients, both at inclusion and 1-­year follow-­up Table S2). This predictive effect was more pronounced in treatment-­
(Figure 2). We observed a decrease in nerve size over time of only the naive patients: slope 1.12  kPa/month (95% CI 0.06–­2.19) without
median nerve at the forearm in atypical CIDP (slope –­0.071  mm2/ enlargement versus −0.90  kPa/month (95% CI −1.83–­0.03) with
month, 95% confidence interval [CI] −0.121 to −0.020) and MMN enlargement (p = 0.008). This indicates that patients without nerve
(slope −0.057  mm2/month, 95% CI −0.100 to −0.014). This corre- enlargement of the median nerve at the upper arm have higher grip
sponds with an average decrease of nerve CSA of 0.852  mm2 and strength after 1 year of follow-­up than patients with nerve enlarge-
0.684 mm2 per year at these sites. Nerve size in patients with typical ment (an increase of +13.44  kPa/year compared to a decrease of
CIDP and CIAP did not change significantly over time (Figure 3). The −10.80 kPa/year). No significant effect of the presence of nerve en-
decrease in nerve size over time in atypical CIDP was attributable to largement on grip strength was observed at other nerve sites or in
2
distal predominant CIDP (n = 35; slope −0.107 mm /month, 95% CI other disease groups (Figure 4).
−0.195 to −0.018), but not to pure motor CIDP (n = 11), pure sensory Presence of C5 nerve root enlargement at inclusion predicted
CIDP (n = 11) and asymmetrical variants of CIDP (n = 21). a significantly improved ODSS over time in treatment-­naive MMN
Separate LMEs were fitted for patients with increased nerve size patients: slope 0.00 points/month (95% CI −0.06 to 0.06) without
at inclusion and patients with normal nerve size at inclusion. Results enlargement versus −0.12 points/month (95% CI −0.19 to −0.04)
were comparable to those of the LMEs fitted for the entire disease with enlargement (p = 0.02). It also predicted significantly improved
groups at all nerve sites for all disease groups, with the exception of RODS score over time in typical CIDP: slope 0.28% (95% CI −0.23 to
the median nerve at the upper arm in atypical CIDP, which showed a 0.79) without enlargement versus 1.03% (95% CI 0.52 to 1.55) with
 |
2334       TELLEMAN et al.

TA B L E 2  Correlation of vigorimetry
Mean grip Correlation of grip strength and
and nerve size
Nerve site strength, kPa nerve size, kPa/mm2 (95% CI) p

Typical CIDP Median nerve: 70.2 −0.4 (−1.6–­0.9) 0.57

forearm
Median nerve: 66.6 0.0 (−0.9–­0.9) 0.99
upper arm
C5 nerve root 70.8 −0.5 (−1.5–­0.6) 0.38
Atypical CIDP Median nerve: 60.1 0.4 (−0.5–­1.4) 0.35
forearm
Median nerve: 57.1 0.4 (−0.1–­1.0) 0.10
upper arm
C5 nerve root 65.2 −0.1 (−0.8–­0.7) 0.87
MMN Median nerve: 74.9 −0.1 (−1.6–­1.4) 0.92
forearm
Median nerve: 86.2 −0.8 (−1.9–­0.2) 0.10
upper arm
C5 nerve root 85.0 −1.3 (−2.3–­− 0.2) 0.02
CIAP Median nerve: 81.4 −0.1 (−2.8–­2.7) 0.95
forearm
Median nerve: 52.9 2.9 (−0.6–­6.4) 0.10
upper arm
C5 nerve root 83.2 −0.5 (−4.1–­3.1) 0.79

Correlation of grip strength (in kPa) and nerve size (in mm2) of the median nerve at forearm and
upper arm and the C5 nerve root per disease group are shown. Results obtained by the fitted linear
mixed effect models are shown, including the mean grip strength (intercept) and average increase/
decrease in grip strength per mm2 in nerve size (slope) including a 95% CI and p value of the slope.
Abbreviations: CI, confidence interval; CIDP, chronic inflammatory demyelinating polyneuropathy;
MMN, multifocal motor neuropathy; kPa, kilopascal.
Bold values are statistically significant.

enlargement (p = 0.04). This positive effect of presence of enlarge- [4,5], nerve CSA does not seem to be a useful marker of disease ac-
ment at the C5 nerve root was also observed for vigorimetry, ODSS tivity in these neuropathies, as there is a poor correlation between
and mISS in typical CIDP, and for vigorimetry and ODSS in the entire nerve size and clinical outcome measures. Our study shows that
group of patients with MMN, although these results were not signif- the sonographic nerve size has limited prognostic value. MMN is a
icant (Table S2). Additional analyses showed that patients with MMN possible exception, since larger nerve size of the median nerve at
with nerve enlargement confined to the brachial plexus had a more the upper arm at inclusion was associated with lower grip strength
favorable outcome (i.e., improvement of grip strength) than patients after 1 year of follow-­up. Moreover, patients with MMN who had
with more generalized enlargement. exclusive enlargement of the brachial plexus fared better than those
with a more generalized pattern of nerve enlargement. However, the
large heterogeneity in observations limits prognostic value in indi-
Other prognostic factors in CIDP and MMN vidual patients.
Previous studies on the value of nerve ultrasonography as a fol-
Prognostic effects of previously identified clinical factors were low-­up tool showed promising results by suggesting a correlation
tested in our multicenter cohort [11–­14]. Shorter disease duration between decreasing nerve size and improved outcome. Improved
to treatment, a subacute start of complaints (nadir ≤6 weeks), and grip strength was associated with normalization of nerve size in a
lower age were all associated with improved vigorimetry and/or cohort of 23 patients with CIDP [11], and a decrease in sonographic
ODSS in both typical CIDP and MMN (p values all <0.05). score for nerve enlargement (the Ultrasound Pattern Sum Score) and
in intra-­nerve variability ratio were associated with an improved clin-
ical outcome [7–­9]. In the present study we could not replicate these
DISCUSSION findings. Retrospective designs, small sample size, and inclusion of
predominantly prevalent patients may have affected results in pre-
This large prospective study shows that sonographic nerve size vious studies, and it is less likely that the present prospective study
and its development over time are heterogeneous in CIN. Although including a large group of incident patients was subject to compa-
nerve enlargement is an important indicator of the presence of CIN rable bias. On the other hand, in the present study we analyzed
PROGNOSTIC VALUE OF ULTRASONOGRAPHY IN CIDP AND MMN |
      2335

F I G U R E 4  Prognostic value of nerve enlargement on grip strength. Effect of presence of nerve enlargement at inclusion at the median nerve
at the forearm and upper arm, and the C5 nerve root on the development of grip strength over time in typical chronic inflammatory demyelinating
polyneuropathy (CIDP), atypical CIDP and multifocal motor neuropathy (MMN). Estimated slopes, obtained from linear mixed effect models, for
patients without enlargement (green) and with enlargement (red) are shown with 95% confidence interval (CI; dotted lines). Gray lines represent
individual patients. A significant effect was found only for the median nerve at the upper arm in MMN (slope 0.33 kPa/month [95% CI −0.19 to
0.86]) without enlargement vs. −0.36 kPa/month [95% CI −0.80 to 0.07] with enlargement; p = 0.04). The lowest box shows plots for treatment-­
naive patients with MMN. Results were only significant at the upper arm (slope 1.12 kPa/month [95% CI 0.06–­2.19] without enlargement vs.
−0.90 kPa/month [95% CI −1.83 to 0.03] with enlargement; p = 0.008). kPa, kilopascal [Colour figure can be viewed at wileyonlinelibrary.com]

nerve size only, and this feature was strikingly heterogeneous in common pathway of pathophysiological processes underlying CIDP
CIDP and MMN. The heterogeneity may be explained by the as- and MMN, its reversal is not crucial for nerve function improvement.
sumption that, despite the fact that nerve enlargement is the final Onion bulb formation, inflammatory cell infiltrates and endoneurial
 |
2336       TELLEMAN et al.

edema, interstitial accumulation of amorphous substances and fi- relatively small, which could have affected the prognostic value of ul-
brosis can all cause nerve enlargement [21–­23], but their relation to trasonography in the early stages of treatment. Also, follow-­up visits
clinical symptoms may differ. It remains to be shown whether nerve in our study were planned irrespective of the time interval between
ultrasonography variables other than CSA are better predictors of the last course of immunoglobulins. As clinical complaints may vary
outcome. We investigated nerve size at three nerve sites in the pres- markedly, this could have affected results on the correlation between
ent study, as the included sites have low interobserver variability nerve size and clinical outcome measures. However, the results on
and were the most specific for CIN in our previous studies on the the prognostic value of nerve enlargement at inclusion are likely less
diagnostic value of nerve ultrasonography (the DUP-­P) [4–­6]. An biased, as these represent long-­term effects. Nevertheless, treatment
analysis using the sum of nerve CSA at these three sites combined of CIN is often required for longer periods of time, and this study will
did not show improved performance. A sum score of sonographic continue for another year to ensure that we will not miss potential
measurements at multiple measurement sites of upper and poten- prognostic effects after the 1-­year follow-­up period.
tially lower extremity nerves may provide a wider scope of disease Nerve ultrasonography is becoming increasingly important to
activity. However, extending an ultrasonography protocol further establish a diagnosis of CIDP or MMN, and allows reliable identi-
for prognostic purposes or using a sum score may come with several fication of potentially treatment-­responsive patients [1–­5]. In this
downsides. The additional investigated sites may be less specific in large prospective multicenter study we found that initial sono-
reflecting disease activity, and the protocol may be less reproducible graphic abnormalities remain present over time, which suggests that
(with higher interobserver variability), more difficult to interpret and nerve ultrasonography is a useful diagnostic tool even in case of di-
more time-­consuming, which all could hamper applicability in rou- agnostic delay. However, the value of sonographic nerve size as a
tine clinical practice. Nevertheless, the prognostic value of a more marker of disease activity or as a follow-­up tool seems more limited.
elaborate ultrasonography protocol deserves further investigation. Some prognostic value was found only in MMN. Overall, nerve ul-
We analyzed data on nerve size only, as we only rarely observed trasonography did not detect changes in nerve sizes that reflected
hypervascularization and as it was not possible to perform reliable treatment efficacy, and basing treatment decisions on alterations in
post hoc classification of nerves based on nerve echogenicity due sonographic nerve CSA at selected sites only should not be encour-
to the use of different sonographic devices in this multicenter study. aged currently. Further prospective studies, incorporating additional
Some relatively small studies reported that differences in echo- sonographic variables, such as standardized quantitative assessment
genicity (determined on a qualitative base) correlated with clinical of nerve echogenicity, should be performed to determine the useful-
outcome in patients with CIDP, with patients showing hyperechoic ness of nerve ultrasonography after the diagnostic phase.
nerves having a worse outcome [8,21,24]. In addition, fascicle size
was associated with clinical outcome in one study [8]. The value of C O N FL I C T O F I N T E R E S T S
additional sonographic variables may thus deserve further attention J. A. Telleman, I. J. T. Herraets, R. P. A. van Eijk, D. Lieba-­S amal,
as well [8,21,24–­27]. J. T. van Asseldonk –­ disclosures: none. H. S. Goedee receives re-
Nerve size correlated poorly with clinical outcome measures in search support from the Prinses Beatrix Spierfonds, and has re-
the present study, and it did not reflect treatment response. As such, ceived a travel grant and speakerfee from Takeda. C. Verhamme
it seems not to be a suitable marker of disease activity. Despite the serves on clinical advisory board for Inflectis. F. Eftimov reports
limited level of correlation and limited prognostic value, a few of our grants from Prinses Beatrix Spierfonds, Netherlands Organization
observations may still be helpful in clinical practice. Patients with for Health Research and Development, and a consulting fee from
MMN showing nerve enlargement of the brachial plexus only had CSL Behring, UCB Pharma and Aserta Pharma that was paid to his
higher grip strength than patients with more generalized peripheral institution, outside the submitted work. As coordinating investi-
nerve enlargement. Differences in the pattern of nerve enlargement gator of INCbase, an international CIDP registry, he has received
may reflect variations in underlying pathophysiological processes funding for INCbase from CSL Behring, Takeda Pharmaceutical
or different stages of the disease. Although additional studies are Company, Kedrion and Terumo BCT. L. H. van den Berg has served
needed, involvement of the brachial plexus only may thus be a prog- on scientific advisory boards for Orion, Orphazyme, Calico and
nostically beneficial factor in addition to previously identified clinical Cytokinetics, received an educational grant from Takeda, serves
and nerve conduction study prognostic factors [12,15]. A previous on the editorial boards of Amyotrophic Lateral Sclerosis and
study on MRI of the brachial plexus in MMN did not find any prog- Frontotemporal Degeneration and the Journal of Neurology,
nostic value, but this study did not investigate coinciding peripheral Neurosurgery, and Psychiatry, and receives research support from
nerve involvement [28]. Further studies combining MRI and nerve the Prinses Beatrix Spierfonds, Netherlands ALS Foundation,
ultrasonography assessment of peripheral nervous system involve- and the Netherlands Organization for Health Research and
ment could shed additional light on this potential prognostic factor. Development (Vici Scheme, JPND [SOPHIA, STRENGTH,
This study had several limitations. First, the follow-­up duration of ALSCare]). W. L. van der Pol serves on the scientific advisory
1 year was relatively short, and although we included a large group board for SAB SMA Europe, is a member of the Branaplam data
of patients with CIDP, subgroups of patients with clinical subtypes monitoring committee (DMC) for Novartis, provides ad hoc con-
of CIDP were small. The subgroup of treatment-­naive patients was sultancy for Biogen and Avexis, and receives research support
PROGNOSTIC VALUE OF ULTRASONOGRAPHY IN CIDP AND MMN |
      2337

6. Telleman JA, Herraets IJT, Goedee HS, et al. Nerve ultrasound: a

from the Prinses Beatrix Spierfonds, Vriendenloterij and Stichting reproducible diagnostic tool in peripheral neuropathy. Neurology.
Spieren voor Spieren. L. H. Visser receives research support from 2019;92:e443-­e 450.
the Prinses Beatrix Spierfonds and the Netherlands Organization 7. Fisse AL, Pitarokoili K, Trampe N, et al. Clinical, sonographic, and
electrophysiologic longitudinal features of chronic inflammatory
for Health Research and Development.
demyelinating polyneuropathy. J Neuroimaging. 2019;29:223-­232.
8. Härtig F, Ross M, Dammeier NM, et al. Nerve ultrasound predicts
AU T H O R C O N T R I B U T I O N treatment response in chronic inflammatory demyelinating poly-
Johan Telleman: Conceptualization (lead); Data curation (equal); radiculoneuropathy –­ a prospective follow-­up. Neurotherapeutics.
Formal analysis (lead); Funding acquisition (supporting); Investigation 2018;15:439-­451.
9. Kerasnoudis A, Pitarokoili K, Gold R, Yoon MS. Nerve ultrasound and
(lead); Methodology (lead); Project administration (equal); Writing
electrophysiology for therapy monitoring in chronic inflammatory
–­ original draft (lead). Ingrid Herraets: Conceptualization (lead); demyelinating polyneuropathy. J Neuroimaging. 2015;25:931-­939.
Data curation (equal); Formal analysis (equal); Funding acquisi- 10. Rattay TW, Winter N, Décard BF, et al. Nerve ultrasound as fol-
tion (supporting); Investigation (lead); Methodology (lead); Project low-­up tool in treated multifocal motor neuropathy. Eur J Neurol.
2017;24:1125-­1134.
administration (equal); Writing –­ original draft (lead). H. Stephan
11. Zaidman CM, Pestronk A. Nerve size in chronic inflammatory de-
Goedee: Conceptualization (equal); Funding acquisition (lead); myelinating neuropathy varies with disease activity and therapy re-
Investigation (equal); Methodology (equal); Writing –­ review sponse over time: a retrospective ultrasound study. Muscle Nerve.
and editing (equal). Ruben P. A. van Eijk: Formal analysis (lead); 2014;50:733-­738.
12. Kuitwaard K, Hahn AF, Vermeulen M, Venance SL, van Doorn PA.
Methodology (equal); Writing –­ review and editing (equal). Camiel
Intravenous immunoglobulin response in treatment-­naïve chronic
Verhamme: Conceptualization (equal); Investigation (equal); inflammatory demyelinating polyradiculoneuropathy. J Neurol
Methodology (equal); Writing –­ review and editing (equal). Filip Neurosurg Psychiatry. 2015;86:1331-­1336.
Eftimov: Investigation (equal); Writing –­ review and editing (equal). 13. Van Asseldonk JT, van den Berg LH, Kalmijn S, et al. Axon loss is an
important determinant of weakness in multifocal motor neuropa-
Doris Lieba-­Samal: Investigation (equal); Writing –­ review and
thy. J Neurol Neurosurg Psychiatry. 2006;77:743-­747.
editing (equal). Jan-­
Thies H. van Asseldonk: Conceptualization 14. Cats EA, van der Pol WL, Pipers S, et al. Correlates of outcome and
(equal); Methodology (equal); Supervision (equal); Writing –­ review response to IVIg in 88 patients with multifocal motor neuropathy.
and editing (equal). L. H. Van den Berg: Conceptualization (equal); Neurology. 2010;75:818-­825.
15. Van den Berg-­Vos RM, Franssen H, Wokke JH, van Es HW, van
Methodology (equal); Supervision (equal); Writing-­review & editing
den Berg LH. Multifocal motor neuropathy: diagnostic criteria that
(equal). W. Ludo van der Pol: Conceptualization (equal); Methodology
predict the response to immunoglobulin treatment. Ann Neurol.
(equal); Supervision (equal); Writing –­ review and editing (equal). 2000;48:919-­926.
Leo H. Visser: Conceptualization (lead); Funding acquisition (lead); 16. Joint Task Force of the EFNS and the PNS. European Federation
Investigation (equal); Methodology (equal); Project administration of Neurological Societies/Peripheral Nerve Society Guideline on
management of chronic inflammatory demyelinating polyradiculo-
(equal); Supervision (lead); Writing –­review and editing (lead).
neuropathy: report of a joint task force of the European Federation
of Neurological Societies and the Peripheral Nerve Society—­First
DATA AVA I L A B I L I T Y S TAT E M E N T Revision. J Peripher Nerv Syst. 2010;15:1-­9.
The data that support the findings of this study will be available from 17. Joint Task Force of the EFNS and the PNS. European Federation
of Neurological Societies/Peripheral Nerve Society Guideline on
the corresponding author upon reasonable request.
management of multifocal motor neuropathy. Report of a Joint
Task Force of the European Federation of Neurological Societies
ORCID ant the Peripheral Nerve Society—­first revision. J Peripher Nerv
Johan A. Telleman  https://orcid.org/0000-0002-4065-9999 Syst. 2010;15:295-­3 01.
18. Visser NA, Notermans NC, Linssen RS, van den Berg LH, Vrancken
Ruben P. A. van Eijk  https://orcid.org/0000-0002-7132-5967
AF. Incidence of polyneuropathy in Utrecht, the Netherlands.
Neurology. 2015;84:259-­264.
REFERENCES 19. Van Nes SI, Vanhoutte EK, van Doorn PA, et al. Rasch-­built Overall
1. Telleman JA, Herraets IJT, Goedee HS, van Asseldonk JT, Visser LH. Disability Scale (R-­ODS) for immune-­mediated peripheral neuropa-
Ultrasound scanning in the diagnosis of peripheral neuropathies. thies. Neurology. 2011;25:337-­3 45.
Pract Neurol. 2021;21:186–­195. https://doi.org/10.1136/pract​neuro​l-­​ 20. Goedee HS, van der Pol WL, van Asseldonk JH, et al. Nerve so-
20​20-­0 02645 nography to detect peripheral nerve involvement in vasculitis syn-
2. Telleman JA, Grimm A, Goedee HS, Visser LH, Zaidman CM. Nerve dromes. Neurol Clin Pract. 2016;6:293-­3 03.
ultrasound in polyneuropathies. Muscle Nerve. 2018;57:716-­728. 21. Padua L, Granata G, Sabatelli M, et al. Heterogeneity of root and
3. Goedee HS, van der Pol WL, van Asseldonk JT, et al. Diagnostic nerve ultrasound pattern in CIDP patients. Clin Neurophysiol.
value of sonography in treatment-­naive chronic inflammatory neu- 2014;125:160-­165.
ropathies. Neurology. 2017;88:143-­151. 22. Di Pasquale A, Morino S, Loreti S, Bucci E, Vanacore N, Antonini G.
4. Herraets IJT, Goedee HS, Telleman JA, et al. Nerve ultrasound im- Peripheral nerve ultrasound changes in CIDP and correlations with
proves detection of treatment-­responsive chronic inflammatory nerve conduction velocity. Neurology. 2015;84:803-­8 09.
neuropathies. Neurology. 2020;94:e1470-­e1479. 23. Matsuda M, Ikeda S, Sakurai S, Nezu A, Yanagisawa N, Inuzuka T.
5. Herraets IJT, Goedee HS, Telleman JA, et al. Nerve ultrasound for Hypertrophic neuritis due to chronic inflammatory demyelinating
diagnosing chronic inflammatory neuropathy: a multicenter valida- polyradiculoneuropathy (CIDP): a post-­mortem pathological study.
tion study. Neurology. 2020;95:e1745-­e1753. Muscle Nerve. 1996;19:163-­169.
 |
2338       TELLEMAN et al.

24. Fisse AL, Pitarokoili K, Motte J, et al. Nerve echogenicity and intra-
nerve CSA variability in high-­resolution nerve ultrasound (HRUS) S U P P O R T I N G I N FO R M AT I O N
in chronic inflammatory demyelinating polyneuropathy. J Neurol. Additional supporting information may be found online in the
2019;266:468-­475. Supporting Information section.
25. Jain S, Visser LH, Praveen TL, et al. High-­resolution sonography: a
new technique to detect nerve damage in leprosy. PLoS Negl trop
Dis. 2009;3:e498.
26. Martinoli C, Derchi LE, Bertolotto M, et al. US and MR imaging of
peripheral enrves in leprosy. Skeletal Radiol. 2000;29:142-­150. How to cite this article: Telleman JA, Herraets IJ, Goedee HS,
27. Chaduvula MV, Visser LH, Suneetha S, et al. High-­resolution sonog- et al. Prognostic value of nerve ultrasonography: A
raphy as an additional diagnostic and prognostic tool to monitor prospective multicenter study on the natural history of
disease activity in leprosy: a two-­year prospective study. Ultrachall
chronic inflammatory neuropathies. Eur J Neurol.
Med. 2016;39:80-­89.
28. Jongbloed BA, Boos JW, Rutgers D, van der Pol WL, van den Berg 2021;28:2327–2338. https://doi.org/10.1111/ene.14885
LH. Brachial plexus magnetic resonance imaging differentiates be-
tween inflammatory neuropathies and does not predict disease
course. Brain Behav. 2017;7:e00632.