Keywords
high-resolution ultrasonography - leprosy - ultrasonography
Introduction
Leprosy is a long-standing granulomatous condition caused by the immune system's reaction
to Mycobacterium leprae infection, predominantly affecting the skin and peripheral nerves. The bacterium
inhabits macrophages and Schwann cells and is distinct in its capacity to interfere
with myelin, resulting in peripheral nerve damage. Early nerve involvement is characterized
by axonal shrinkage and localized inflammation. This inflammation facilitates the
formation of a vascular route through which M. leprae migrates from the perineurium into the endoneurium,[1] which then gets engulfed by Schwann cells within the fascicle.[2] The Ridley-Jopling classification divides leprosy into five distinct types,[3] ranging from the localized, paucibacillary tuberculoid form (TT), which presents
with hypopigmented, anesthetic skin lesions, to the more widespread, multibacillary
lepromatous form. Between these two ends of the spectrum are the borderline forms:
borderline tuberculoid, borderline borderline, and borderline lepromatous, which are
immunologically unstable. These intermediate forms are particularly prone to episodic
inflammatory reactions, which can affect around 15 to 50% of patients during the disease
course and even after completion of multidrug therapy.[4] Leprosy reactions include Type 1 (reversal reaction), marked by localized inflammation
of skin lesions and nerve tenderness caused by edema-induced compression, and Type
2 (erythema nodosum leprosum), which manifests with systemic features like fever,
painful red nodules, and joint aches. While nerve involvement is a common feature
across all forms of leprosy, these reactions can precipitate acute neuritis, potentially
impairing nerve function. If not promptly managed, they may lead to permanent nerve
damage. Although clinical examination remains the primary method for detecting and
assessing nerve involvement, it is limited by considerable interobserver variability.[5] Clinical examination can be difficult due to the deep anatomical location of the
peripheral nerves.[6] In this study, clinical examination was used as the reference standard in accordance
with the World Health Organization (WHO) guidelines, which continue to recommend clinical
examination as the primary method for diagnosing leprosy, especially in peripheral
and resource-limited settings. The WHO emphasizes early clinical diagnosis based on
the cardinal signs of leprosy, including nerve involvement detectable through palpation
and sensory-motor assessment. While high-resolution ultrasonography (HRUS) has demonstrated
higher sensitivity and can detect subclinical changes, it is not widely available
in endemic regions. Thus, the use of clinical examination as the comparator reflects
both practical field relevance and WHO-endorsed diagnostic practices. HRUS offers
a precise and detailed visualization of the internal structure of peripheral nerves,
allowing for the identification of alterations in nerve thickness, echotexture, and
blood flow. Incorporating color Doppler further enhances diagnostic accuracy by highlighting
vascular changes that may indicate inflammation or other pathological processes, particularly
during Type 1 lepra reactions. Its diagnostic value has been well-documented in multiple
studies and scholarly publications.[7]
[8]
[9] It measures the maximum cross-sectional area (CSA) of nerves as an objective indicator
of enlargement and captures detailed morphological features, such as the internal
structure of fascicles, echogenicity,[10] and the condition of the perineurium and epineurium.[11] Distal nerves are normally hyperechoic, with a stippled (“honeycomb”) structure
(with hypo-anechoic fascicles on the hyperechoic background of connective tissue surrounding
them).[12] This study utilizes the capabilities of HRUS to diagnose peripheral nerve involvement
in leprosy by identifying early structural changes that may not be detectable through
clinical examination alone.
Materials and Methods
This study, approved by the institutional ethics committee, is a cross-sectional analysis
conducted on 60 newly diagnosed leprosy patients over an 18-month period, spanning
from August 2023 to January 2025. Leprosy was diagnosed when at least one of the following
cardinal signs was present (as per the 8th WHO Expert Committee on Leprosy).[13]
-
Definite loss of sensation in a pale (hypopigmented) or reddish skin patch.
-
A thickened or enlarged peripheral nerve, with loss of sensation and/or weakness of
the muscles supplied by that nerve.
-
The presence of acid-fast bacilli in a slit-skin smear.
Patients with comorbid conditions such as metabolic disorders, diabetes, uremia, hypercholesterolemia,
porphyria, and vasculitic diseases like polyarteritis nodosa (PAN), as well as those
with hypothyroidism, allergic vasculitis, history of trauma, chronic compressive neuropathy,
or hereditary neuropathies (i.e., other causes of peripheral neuropathy), were excluded
from the study. Additionally, individuals who were skin smear positive and had previously
received multidrug therapy, along with those who did not provide consent, were also
not included. After obtaining informed consent, five pairs of nerves, namely, ulnar
nerve (UN), median nerve, greater auricular nerve, posterior tibial nerve, and common
peroneal nerve, were clinically assessed. The nerves were evaluated for sensory, autonomic,
and motor functions through a comprehensive examination, which included palpation
to assess for thickening, tenderness, and consistency (such as cord-like or fibrosed).
Autonomic function was assessed by evaluating signs such as loss of sweating on the
palms and soles, hair loss, and the presence or absence of sweating as indicated by
dry-appearing skin, as well as the presence of cracks or fissures.
-
Ulnar nerve (UN): We checked for numbness and paresthesias of the fourth and fifth
digits of the hand. Both arms were examined by testing pinprick sensation at digit
5 using monofilament. Motor functions of the palmar, dorsal interossei, and adductor
pollicis were checked.[4]
-
Median nerve (MN): Pin-prick sensation in the distribution of the median nerve (in
the palm over the thenar eminence and pulp of the index finger) was checked. Motor
function of abductor pollicis brevis, opponens pollicis, flexor digitorum superficialis,
and flexor digitorum profundus (lateral half) were assessed.
-
Greater auricular nerve (GAN): Pin-prick sensation using monofilaments in the area
of skin of the auricle, skin over the parotid gland, and mastoid process were evaluated.
-
Common peroneal nerve (CPN): Pin-prick sensation using monofilaments on the lateral
part of the leg and dorsum of the foot with any evidence of foot drop was noted. Dorsiflexors
of the ankle, extensor hallucis longus and tibialis anterior muscles, were assessed.
-
Posterior tibial nerve (PTN): We assessed the sole of the foot to check for sensory
function of the nerve with assessment of the intrinsic muscles of the feet.
Sensory loss was considered present when the patient was unable to perceive 2 g of
target force on the hand and 300 g of target force on the foot by Semmes-Weinstein
monofilaments. Medical Research Council (MRC) muscle strength grading was used for
muscle weakness. Muscle weakness was present when the MRC score was 4 or less. UN,
GAN, CPN, and PTN were clinically graded after palpation as follows[4]:
-
Grade 0 is defined as a nerve not thicker than the contralateral nerve and with normal
sensation.
-
Grade 1 occurs when the affected nerve is thicker than the contralateral nerve.
-
Grade 2 is a thickening of the affected nerve with a rope-like consistency.
-
Grade 3 is a thickened nerve with a beaded or nodular feel.
Clinical grading of nerve thickening based on palpation was not performed on MN due
to its deeper location. Slit skin smear for acid-fast bacilli was sent for investigation
for bacteriological and morphological index. Before starting MDT, HRUS, and color
Doppler, evaluation of peripheral nerves was performed in the Department of Radio
Diagnosis of our hospital. A 6 to 15 linear array transducer with a frequency range
of 4.5 to 15 MHz was used in assessing peripheral nerves. Due to the superficial location
of peripheral nerves, the maximum available frequency was 15 HZ. Ultrasonography (USG)
of five pairs of peripheral nerves (UN, MN, GAN, CPN, and PTN) with color Doppler
of peripheral nerves was done; the sonographic findings were analyzed simultaneously
by two radiologists who were blinded from the clinical findings. The enlarged nerve
was visualized at the cubital tunnel for the UN, carpal tunnel for the MN, the area
of the fibular head and neck for the CPN, and the tarsal tunnel for the PTN.[8] For GAN, the scan was started at the posterior border of the sternocleidomastoid
muscle. The probe was moved cranially and caudally in the transverse view until a
structure was identified beneath the sternocleidomastoid muscle.[14] The following parameters were noted by examining nerves both transversely and longitudinally
by changing the direction of the linear array transducer.
-
CSA—In mm2, the selected normal values for cross-sectional area of ulnar nerve, median nerve,
common peroneal nerve, posterior tibial nerve, and greater auricular nerve are 8.5,
6.2, 5.9, 6.3, and 0.84 mm2, respectively.[20]
-
Echogenicity of the nerve—Normal/hypoechoic/hyperechoic.
-
Blood flow in color Doppler and change in fibrillary echotexture were estimated at
multiple sites in the transverse plane, and whether a peripheral echogenic rim was
maintained or lost.
In normal nerves, prominent arterial pulsations are typically absent, as the perineural
and intraneural blood vessels are not usually detectable on Doppler imaging due to
their minimal blood volume and slow flow velocities.[15]
The detection of blood flow signals in the perineural plexus or interfascicular vessels
on color Doppler imaging was interpreted as an indication of nerve hypervascularity.
A nerve was deemed sonologically abnormal if any one of the following criteria was
observed in the cross-sectional view.[20]
-
Increased cross-sectional area.
-
Hypoechogenicity of the nerve (fascicles and interfascicular perineurium) with or
without loss of architecture.
-
Hypervascularity.
Statistical Analysis
Categorical variables were expressed as frequency and percentage, while continuous
variables were expressed as mean and standard deviation. The agreement between clinical
examination and HRUS was assessed using Cohen's kappa statistic. The McNemar test
was used for p-value calculation, where p < 0.05 was considered significant. Sensitivity, specificity, positive predictive
values (PPVs), and negative predictive values (NPVs) for HRUS were calculated, taking
clinical examination as the gold standard. All statistical analyses were done using
Epi Info version 7.2.1.0 and Jamovi version 4.0 statistical software.
Results
The study included 60 participants with a mean age of 35.83 years, predominantly in
the 20- to 49-year age group (76.67%), indicating that leprosy is most common in middle-aged
individuals. A male predominance (75%) was observed; lower socioeconomic status was
prevalent in 61.67% of cases, with only 15% of patients reporting a family history
of leprosy. Using the Ridley-Jopling classification, the most common form was borderline
tuberculoid (51.67%), followed by borderline lepromatous (16.67%), while multibacillary
leprosy accounted for 76.67% of cases under the WHO classification. Slit skin smear
positivity was seen in 36.67% of cases [Table 1].
Clinical examination of major peripheral nerves ([Table 2]) revealed that among the 600 nerves examined, 90 (15%) were clinically abnormal
based on the presence of any one or more clinical signs such as thickening, sensory
loss, or motor weakness. These categories are overlapping, as some nerves exhibited
more than one abnormality simultaneously. Nerve thickening was seen in 12.2% (n = 73/600) of all nerves, with the UN being the most commonly thickened (26.7%), followed
by the CPN (15%), PTN (10%), MN (4.2%), and GAN (4.2%). Sensory loss was more frequently
observed than motor loss, affecting 11.8% of nerves, with the UN showing the highest
sensory involvement (28.3%), followed by the CPN (15.8%), PTN (10.8%), and MN (4.2%).
Motor loss was detected in 2.3% of nerves, primarily in the UN (10.8%) and PTN (0.8%)
nerves. HRUS detected abnormalities in 18.5% (n = 111/600) of all nerves, surpassing the 15% detected clinically. Among the HRUS
findings ([Table 2]), increased CSA was seen in 17% (n = 102/600) nerves, hypervascularity in 3.5% nerves (21/600), and hypoechogenicity
in 2.6% nerves (17/600). The UN again had the highest frequency of USG abnormalities
(30.8%, [Fig. 1]), followed by the CPN (22.5%, [Fig. 2]), PTN (21.6%, [Fig. 3]), MN (5.8%, [Fig. 4]), and GAN (4.2%, [Fig. 5]). [Table 3] illustrates HRUS findings of clinically abnormal and normal nerves. Among clinically
abnormal nerves, the most significant finding was that 83.3% of clinically abnormal
nerves exhibited an increased CSA, hypoechogenicity was observed in 12.2%, loss of
fascicular architecture was seen in 10%, and increased blood flow on Doppler was observed
in 17.7% of clinically abnormal nerves ([Figs. 6] and [7]). In clinically normal nerves, 5.3% showed increased CSA, hypoechogenicity in 1.2%
nerves, and increased blood flow on Doppler was detected in only 0.98% nerves. [Table 4] illustrates the relationship between clinical findings and USG abnormalities, specifically
increased CSA, increased endoneurial blood flow on Doppler, and hypoechogenicity in
600 peripheral nerves. A significant association was observed between increased CSA
and all major clinical features: nerve thickening (p < 0.001), tenderness (p < 0.001), sensory loss (p < 0.001), and motor loss (p < 0.001). Among nerves with increased CSA, 60.8% had clinical thickening, 58.8% had
sensory loss, and 12.7% showed motor weakness. Similarly, increased blood flow on
color Doppler was significantly associated with clinical thickening (p < 0.001), tenderness (p < 0.001), sensory loss (p < 0.001), and motor loss (p < 0.001), though its prevalence was lower. Hypoechogenicity was also significantly
correlated with nerve thickening (p < 0.001) and sensory loss (p < 0.001), but not with tenderness (p = 0.927) or motor loss (p = 0.072), suggesting a more selective association with inflammatory or demyelinating
changes. Grade-wise nerve thickening did not show significant correlation with any
HRUS parameters (p > 0.05), indicating that qualitative grading on palpation may not reflect the extent
of sonological abnormality. [Table 5] illustrates the diagnostic parameters of HRUS for individual nerves. Among individual
nerves, the GAN had the highest sensitivity (100%), followed by CPN (90.5%), PTN (89.5%),
UN (88.6%), and MN (60%). In terms of specificity, again GAN had the highest specificity
(100%) followed by MN (95.5%), UN (91.8%), CPN (89.9%), and PTN (89.1%). The highest
agreement was observed for the greater auricular nerve (κ = 1.0), indicating perfect concordance, though this must be interpreted cautiously
due to the small sample size (n = 5 abnormal GANs). The UN also demonstrated strong agreement (κ = 0.784), followed by the common peroneal (κ = 0.699) and posterior tibial nerves (κ = 0.659), all indicating substantial agreement. The median nerve showed moderate
agreement (κ = 0.530), possibly due to its deeper anatomical position, affecting both palpation
and ultrasonographic clarity. Increased CSA was the most reliable HRUS parameter,
showing high sensitivity (83.3%), specificity (94.7%), and substantial agreement with
clinical findings (κ = 0.740; [Table 6]). In contrast, hypoechogenicity and hypervascularity had high specificity (>98%)
but low sensitivity (12.2 and 17.8%, respectively) with only slight to fair agreement
(κ = 0.166 and 0.246, respectively). The combined use of all three parameters improved
sensitivity to 86.7% with a κ = 0.732, supporting CSA as the dominant marker while others offer adjunctive value.
The PPV was 70.3%, while the NPV was 97.6% ([Table 6]).
Table 1
Demographic profile of patients
|
Age group (years)
|
N
|
Percentage
|
|
<20 y
|
5
|
8.33
|
|
20–29 y
|
16
|
26.67
|
|
30–39 y
|
18
|
30
|
|
40–49 y
|
12
|
20
|
|
50–59 y
|
7
|
11.67
|
|
60–69 y
|
2
|
3.33
|
|
Total
|
60
|
100
|
|
Mean ± SD
|
35.83 ± 12.25 y
|
|
Gender
|
N
|
Percentage
|
|
Male
|
45
|
75
|
|
Female
|
15
|
25
|
|
Total
|
60
|
100
|
|
Socioeconomic status
|
N
|
Percentage
|
|
Lower
|
37
|
61.67
|
|
Lower middle
|
14
|
23.33
|
|
Upper
|
9
|
15
|
|
Total
|
60
|
100
|
|
Duration of disease (months)
|
N
|
Percentage
|
|
≤6 mo
|
5
|
8.33
|
|
6–12 mo
|
36
|
60
|
|
>12 mo
|
19
|
31.67
|
|
Total
|
60
|
100
|
|
Mean ± SD
|
16.17 ± 13.86 mo
|
|
Family history of leprosy
|
N
|
Percentage
|
|
Present
|
9
|
15
|
|
Absent
|
51
|
85
|
|
Total
|
60
|
100
|
|
Ridley-Jopling classification
|
N
|
Percentage
|
|
Tuberculoid
|
5
|
8.33
|
|
Lepromatous
|
5
|
8.33
|
|
Borderline borderline
|
5
|
8.33
|
|
Borderline lepromatous
|
10
|
16.67
|
|
Borderline tuberculoid
|
31
|
51.67
|
|
Indeterminate
|
1
|
1.67
|
|
Neuritic
|
3
|
5
|
|
Total
|
60
|
100
|
|
WHO classification
|
N
|
Percentage
|
|
Multibacillary
|
46
|
76.67
|
|
paucibacillary
|
14
|
23.33
|
|
Total
|
60
|
100
|
|
Slit skin smear
|
N
|
Percentage
|
|
Positive
|
22
|
36.67%
|
|
Negative
|
38
|
63.33%
|
|
Total
|
60
|
100
|
Table 2
Clinical and HRUS findings of major peripheral nerves in leprosy patients
|
Clinical findings of major peripheral nerves in leprosy patients
|
|
Ulnar (
N
= 120)
|
Median (
N
= 120)
|
Common peroneal (
N
= 120)
|
Posterior tibial (
N
= 120)
|
Greater auricular (
N
= 120)
|
All nerves (
N
= 600)
|
|
Thickening
|
32 (26.7%)
|
5 (4.2%)
|
18 (15%)
|
12 (10%)
|
5 (4.2%)
|
73 (12.2%)
|
|
Grade 1
|
27 (22.5%)
|
0
|
17 (13.4%)
|
12 (10%)
|
4 (3.2%)
|
60 (10%)
|
|
Grade 2
|
3 (2.5%)
|
0
|
1 (0.8%)
|
0
|
1 (0.8%)
|
5 (0.8%)
|
|
Grade 3
|
2 (1.7%)
|
0
|
0
|
0
|
0
|
2 (0.3%)
|
|
Tenderness
|
2 (1.7%)
|
5 (4.2%)
|
3 (2.5%)
|
10 (8.3%)
|
0
|
20 (3.2%)
|
|
Sensory loss
|
34 (28.3%)
|
5 (4.2%)
|
19 (15.8%)
|
13 (10.8%)
|
0
|
71 (11.8%)
|
|
Motor loss
|
13 (10.8%)
|
0
|
0
|
1 (0.8%)
|
0
|
14 (2.3%)
|
|
HRUS findings of major peripheral nerve involvement in leprosy patients
|
|
Ulnar (
N
= 120)
|
Median (
N
= 120)
|
Common peroneal (
N
= 120)
|
Posterior tibial (
N
= 120)
|
Greater auricular (
N
= 120)
|
All nerves (
N
= 600)
|
|
Increased CSA
|
37 (30.8%)
|
7 (5.8%)
|
27 (22.5%)
|
26 (21.6%)
|
5 (4.2%)
|
102 (17%)
|
|
Hypoechoic
|
9 (7.5%)
|
3 (2.5%)
|
3 (2.5%)
|
2 (1.6%)
|
0
|
17 (2.6%)
|
|
Loss of fascicular/fibrillary architecture
|
9 (7.5%)
|
1 (0.8%)
|
2 (1.6%)
|
0
|
0
|
12 (2%)
|
|
Increased blood flow on Doppler
|
8 (6.6%)
|
5 (4.2%)
|
2 (1.6%)
|
6 (5%)
|
0
|
21 (3.5%)
|
Fig. 1 (A, D) Transverse axis view of the right UN at the level of the medial epicondyle shows
increased cross-sectional area (51 mm2 encircled) with loss of normal honeycomb pattern. (B) Longitudinal axis of the right UN shows thickening of the nerve (vertical white
line). (C) Increased vascularity of nerve as seen by red, blue color Doppler signals. (Landmark—UN
running medial to medial epicondyle [ME].)
Fig. 2 (A) Transverse axis view of the left CPN shows increased cross-sectional area (19 mm2, encircled area) associated with hypoechogenicity. (B) The longitudinal axis of the left CPN shows thickening of the nerve (white vertical
dotted line). (Bony landmark—fibular head [FH]; CPN winds to the lateral head of fibula.)
Fig. 3 (A, C) Transverse axis of the left PTN shows increased CSA (16 mm2, encircled nerve) with hypoechogenicity of the nerve. (B) Longitudinal axis of the left PTN shows nerve enlargement (white vertical line)
with thickening of epineurium (blue arrow). (Landmark: The posterior tibial nerve,
artery, and vein run close together behind the medial malleolus [MM]. The posterior
tibial artery is positioned most anteriorly, followed by the posterior tibial vein
[blue color in C], and then the posterior tibial nerve [encircled nerve in A, C].)
Fig. 4 (A, C) Transverse view of the left MN shows increased CSA (18 mm2) with a honeycomb fascicular pattern at the wrist. (B) Longitudinal axis of the MN with increased thickness (white vertical line) with
tram track appearance. (This section was taken at the level of the wrist, where MN
travels anteriorly to the surface of the radius and is superficial to the pronator
quadratus [PQ]. In the carpal tunnel, it lies along with flexor tendons [white arrows],
deeper to the flexor retinaculum.)
Fig. 5 (A) Transverse axis view of the right GAN shows increased cross-sectional area (3 mm2, encircled) with hypoechogenicity. (B) Longitudinal axis of GAN shows thickening of nerve (vertical white line) with hyperechoic
epineurium (black arrows). (Landmark: GAN emerges along the posterior aspect of the
sternocleidomastoid [SCM] muscle and ascends vertically across the SCM muscle. This
USG image was obtained at the upper one-third level of the SCM muscle.)
Table 3
HRUS findings of clinically abnormal and normal nerves
|
USG findings of clinically abnormal peripheral nerves in leprosy patients
|
|
Ulnar (
N
= 35)
|
Median (
N
= 10)
|
Common peroneal (
N
= 21)
|
Posterior Tibial (
N
= 19)
|
Greater auricular (
N
= 5)
|
All nerves (
N
= 90)
|
|
Increased CSA
|
31 (88.6%)
|
4 (40%)
|
18 (85.7%)
|
17 (89.5%)
|
5 (100%)
|
75 (83.3%)
|
|
Hypoechoic
|
7 (20%)
|
1 (10%)
|
2 (9.5%)
|
1 (5.3%)
|
0
|
11 (12.2%)
|
|
Loss of fascicular/fibrillary architecture
|
8 (22.9%)
|
0
|
1 (4.8%)
|
0 (%)
|
0
|
9 (10%)
|
|
Increased blood flow on Doppler
|
8 (22.9%)
|
2 (20%)
|
1 (4.8%)
|
5 (26.3%)
|
0
|
16 (17.7%)
|
|
USG findings of clinically normal peripheral nerves in leprosy patients
|
|
Ulnar (N = 85)
|
Median (N = 110)
|
Common peroneal (N = 99)
|
Posterior tibial (N = 101)
|
Greater auricular (N = 115)
|
All nerves (N = 510)
|
|
Increased CSA
|
6 (7.1%)
|
3 (2.7%)
|
9 (9.1%)
|
9 (8.9%)
|
0
|
27 (5.3%)
|
|
Hypoechoic
|
2 (2.4%)
|
2 (1.8%)
|
1 (1%)
|
1 (0.99%)
|
0
|
6 (1.2%)
|
|
Loss of fascicular/fibrillary architecture
|
1 (1.2%)
|
1 (0.9%)
|
1 (1%)
|
0
|
0
|
3 (0.6%)
|
|
Increased blood flow on Doppler
|
0
|
3 (2.7%)
|
1 (1%)
|
1 (0.99%)
|
0
|
5 (0.98%)
|
Abbreviations: CSA, cross-sectional area; HRUS, high-resolution ultrasonography; USG,
ultrasonography.
Fig. 6 Longitudinal view of the nerves: (A) Thickening of the epineurium of the ulnar nerve (black arrows) with a tram track
appearance of the nerve, with slight hypervascularity at the level of the medial epicondyle.
(B) Hypervascularity of the median nerve with increased thickness at the level of the
flexor retinaculum (white vertical line).
Fig. 7 Posterior tibial nerve thickening (white vertical line) with hypervascularity at
the level of the medial malleolus with loss of normal tram track pattern on longitudinal
axis.
Table 4
Clinical findings in relation to HRUS parameters (cross-sectional area, increased
blood flow on Doppler, and hypoechogenicity; N = 600)
|
Clinical findings in relation to USG findings of increased cross-sectional area in
all nerves (
N
= 600)
|
|
Clinical findings
|
|
Increased CSA(
N
= 102)
|
Normal CSA(
N
= 498)
|
p
-Value
|
|
N
|
%
|
N
|
%
|
|
Nerve thickening
|
Yes
|
62
|
60.78
|
10
|
2
|
<0.001 (S)
|
|
No
|
40
|
39.22
|
488
|
98
|
|
Thickening grade
|
Grade 1
|
55
|
53.92
|
10
|
3.54
|
0.523
|
|
Grade 2
|
5
|
4.90
|
0
|
0
|
|
Grade 3
|
2
|
1.96
|
0
|
0
|
|
Tenderness
|
Yes
|
16
|
15.69
|
4
|
0.8
|
<0.001 (S)
|
|
No
|
86
|
84.31
|
494
|
99.2
|
|
Sensory loss
|
Yes
|
60
|
58.82
|
11
|
2.2
|
<0.001 (S)
|
|
No
|
42
|
41.18
|
487
|
97.8
|
|
Motor loss
|
Yes
|
13
|
12.75
|
1
|
0.2
|
<0.001 (S)
|
|
No
|
89
|
87.25
|
497
|
99.8
|
|
Clinical findings in relation to USG findings of increased blood flow on Doppler in
all nerves (
N
= 600)
|
|
Clinical findings
|
|
Increased blood flow (N = 21)
|
Normal blood flow (N = 579)
|
p-Value
|
|
N
|
%
|
N
|
%
|
|
Nerve thickening
|
Yes
|
9
|
42.86
|
63
|
10.88
|
<0.001 (S)
|
|
No
|
12
|
57.14
|
516
|
89.12
|
|
Thickening grade
|
Grade 1
|
8
|
38.10
|
57
|
9.84
|
0.193
|
|
Grade 2
|
0
|
0
|
5
|
0.86
|
|
Grade 3
|
1
|
4.76
|
1
|
0.17
|
|
Tenderness
|
Yes
|
8
|
38.10
|
12
|
2.07
|
<0.001 (S)
|
|
No
|
13
|
61.90
|
567
|
97.93
|
|
Sensory loss
|
Yes
|
14
|
66.67
|
57
|
9.84
|
<0.001 (S)
|
|
No
|
7
|
33.33
|
522
|
90.16
|
|
Motor loss
|
Yes
|
5
|
23.81
|
9
|
1.55
|
<0.001 (S)
|
|
Clinical findings in relation to USG findings of hypoechogenicity in all nerves (
N
= 600)
|
|
Clinical findings
|
|
Increased hypoechogenicity (N = 17)
|
Normal hypoechogenicity(N = 583)
|
p-Value
|
|
N
|
%
|
N
|
%
|
|
Nerve thickening
|
Yes
|
9
|
52.94
|
63
|
10.81
|
<0.001 (S)
|
|
No
|
8
|
47.06
|
520
|
89.19
|
|
Thickening grade
|
Grade 1
|
8
|
47.06
|
57
|
9.78
|
0.193
|
|
Grade 2
|
0
|
0
|
5
|
0.86
|
|
Grade 3
|
1
|
5.88
|
1
|
0.17
|
|
Tenderness
|
Yes
|
1
|
5.88
|
19
|
3.26
|
0.927
|
|
No
|
16
|
94.12
|
564
|
96.74
|
|
Sensory loss
|
Yes
|
10
|
58.82
|
61
|
10.46
|
<0.001 (S)
|
|
No
|
7
|
41.18
|
522
|
89.54
|
|
Motor loss
|
Yes
|
2
|
11.76
|
12
|
2.06
|
0.072
|
Abbreviations: CSA, cross-sectional area; HRUS, high-resolution ultrasonography; USG,
ultrasonography.
Table 5
Diagnostic parameters of USG for diagnosis of nerve abnormality (individual nerves)
|
Parameter
|
Sensitivity
|
Specificity
|
PPV
|
NPV
|
Diagnostic accuracy
|
Cohen's kappa
|
|
Ulnar
|
88.6%
|
91.8%
|
81.6%
|
95.1%
|
90.8%
|
0.784
|
|
Median
|
60%
|
95.5%
|
54.6%
|
96.3%
|
92.5%
|
0.530
|
|
Common peroneal
|
90.5%
|
89.9%
|
65.5%
|
97.8%
|
80%
|
0.699
|
|
Posterior tibial
|
89.5%
|
89.1%
|
60.71%
|
97.8%
|
89.2%
|
0.659
|
|
Greater auricular
|
100%
|
100%
|
100%
|
100%
|
100%
|
1
|
|
All nerves
|
86.7%
|
93.5%
|
70.3%
|
97.6%
|
92.5%
|
0.732
|
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; USG,
ultrasonography.
Table 6
Diagnostic parameters of USG for the diagnosis of nerve abnormality
|
Parameter
|
Sensitivity
|
Specificity
|
PPV
|
NPV
|
Diagnostic accuracy
|
Cohen's kappa
|
|
Increased cross-sectional area
|
83.3%
|
94.7%
|
73.5%
|
97%
|
93%
|
0.740
|
|
Hypoechogenicity
|
12.2%
|
98.8%
|
64.7%
|
86.5%
|
85.8%
|
0.166
|
|
Hypervascularity
|
17.8%
|
99%
|
76.2%
|
87.2%
|
86.8%
|
0.246
|
|
Overall (combined)
|
86.7%
|
93.5%
|
70.3%
|
97.6%
|
92.5%
|
0.732
|
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; USG,
ultrasonography.
Discussion
The earliest documented use of USG in leprosy was in 1987 in France, where a 5-MHz
linear array real-time probe, along with a stand-off pad, was employed to detect swelling
in the lateral popliteal nerve.[16] Following a period of minimal use lasting over 10 years, the advent of HRUS with
11- to 15-MHz probes reignited interest in its utilization for diagnosing and assessing
leprosy.[7] In the short-axis view, healthy nerves exhibit a characteristic “honeycomb” pattern,
consisting of continuous hypoechoic (dark) neuronal fascicles encased by echogenic
(bright) perineurium and epineurium. When viewed in the long axis, the nerve displays
a “tram-track” appearance.[17] On static cross-sectional ultrasound images, nerves and tendons can appear similar,
particularly when positioned close to each other, for example, the median nerve and
flexor tendons in the carpal tunnel. However, nerve fascicles are typically thicker
and less numerous than the finer, more abundant fibrils seen in tendons. A distinguishing
characteristic is that tendons display greater anisotropy, meaning their echogenicity
varies more noticeably with changes in the angle of the ultrasound probe, often causing
them to appear more hypoechoic than adjacent nerves.[18] While HRUS is useful for confirmation of nerve enlargement in all clinical types
of leprosy, it is most valuable in pure neural leprosy, which is a diagnostic challenge
in view of the absence of visible skin lesions and negative skin smears. HRUS can
aid in distinguishing leprosy from other neuropathies, such as ulnar nerve entrapment,
where nerve enlargement typically occurs at the sulcus or just above the elbow. In
contrast, leprosy-related nerve thickening is usually located more proximally, ∼3
to 4 cm above the elbow.[19] In regions where peripheral nerves are located close to blood vessels, color Doppler
imaging proves valuable in distinguishing nerves from nearby vascular structures.
This study provided key insights into the utility of HRUS combined with color Doppler
in assessing leprosy-affected nerves. The increasing significance of this imaging
method stems from its ability to detect early, subclinical nerve involvement and its
cost-effectiveness, making it an important tool in the prevention of disability and
deformities.
Our study compared the findings of clinical examination with ultrasound findings.
Cross-sectional area, echogenicity of nerve, and blood flow on color Doppler were
the parameters noted on HRUS. In our study, out of 600 nerves, 90 (15%) were clinically
involved and 111 (18.5%) were sonologically abnormal. Increased cross-sectional area
was the most frequently associated sonological parameter. In a study conducted by
Venugopal et al,[20] out of the 320 nerves 71 (22.18%) were clinically abnormal, while 63 nerves (19.7%)
were sonologically abnormal. In the study conducted by Ashwini et al,[21] 86 out of 210 nerves (41%) showed thickening on clinical examination, while sonographic
evaluation identified thickening in 138 nerves (65.7%). In our study, 33/600 (5.5%)
nerves were clinically normal but sonologically abnormal. In a study by Kumaran et
al,[22] 41 out of 240 nerves (17.1%) that were clinically normal were found to be enlarged
on sonography, which was comparatively higher as compared with our study and indicated
that HRUS was able to diagnose subclinical nerve involvement. In our study, a significant
correlation was observed between clinical features such as nerve thickening, sensory
impairment, motor loss, and sonological findings like increased CSA, endoneurial blood
flow, and hypoechogenicity. However, the clinical grade of nerve thickening did not
show any association with HRUS parameters. Jain et al observed a strong correlation
between clinical indicators such as the degree of nerve thickening, sensory deficits,
and muscle weakness and sonographic abnormalities, including altered nerve echotexture,
endoneurial blood flow, and increased CSA.[4]
In our study, sensitivity and specificity were highest for the greater auricular nerve
(100%) and sensitivity was lowest for the median nerve (60%), while specificity remained
above 89% for all nerves, ensuring minimal false positives. PPV was highest for the
greater auricular nerve (100%) but lower for the median nerve (54.6%), suggesting
that ultrasound was more effective in correctly identifying ulnar and common peroneal
nerve abnormalities than median nerve abnormalities. NPV was consistently high (>95%).
Cohen's kappa showed strong agreement between clinical and USG findings, particularly
for the greater auricular (1.0) and ulnar nerves (0.784) and lowest for the median
nerve (0.530).
In comparison to our findings, Venugopal et al[20] reported that HRUS demonstrated the highest sensitivity for the right median nerve
(100%), while the lowest sensitivity was observed for both the right and left posterior
tibial nerves (40% each). The highest specificity was noted for the right and left
common peroneal nerves (100% each), whereas the right UN showed the lowest specificity
at 73.1%. Dugad et al[23] conducted a detailed evaluation of each ultrasonographic parameter for individual
nerves and reported high diagnostic performance. For the UN, echogenicity showed a
sensitivity of 94.74%, a PPV of 100%, and an accuracy of 94.74%. Similarly, for other
nerves, the sensitivity, PPV, and accuracy were all above 80%: for MN (88.88%, 100%,
and 88.88%), lateral popliteal nerve (80.00%, 100%, and 80.00%), and posterior tibial
nerve (85.71%, 100%, and 85.71%), respectively. Since the study focused solely on
positive findings, specificity and NPV were not calculated by them.
The overall sensitivity of HRUS when combining all three parameters was 86.7%, specificity
was 93.5%, the PPV was 70.3%, and the NPV was 97.6%. The diagnostic accuracy of HRUS
was 92.5%, confirming that it is a highly reliable tool for detecting nerve involvement
in leprosy, which is quite in alignment with the findings of Venugopal et al,[20] where he reported overall sensitivity of HRUS as compared with clinical examination
as 63% and specificity as 92.7%, PPV of 71.4%, and NPV of 89.9%.
In this study, a nerve was considered abnormal on HRUS if it exhibited any one of
three parameters: increased CSA, hypoechogenicity, or hypervascularity, following
the diagnostic approach used by Venugopal et al.[20] While CSA alone accounted for the majority of HRUS positive findings and demonstrated
the highest sensitivity (83.3%) and agreement (κ = 0.740), we included hypoechogenicity and hypervascularity to ensure a more comprehensive
evaluation. Although their individual sensitivities were low (12.2%, κ = 0.166) and (17.8%, κ = 0.246, respectively), these features may capture specific pathological changes,
particularly inflammatory activity or early nerve damage, which is not always associated
with nerve enlargement. Thus, the combined definition, while CSA-dominant, may provide
additive diagnostic value. However, we acknowledge this approach could be refined
in future studies, potentially through a weighted scoring system. The overall Cohen's
kappa value (κ = 0.732) indicates a substantial agreement between HRUS and clinical examination,
suggesting that both methods largely correlate in detecting nerve abnormalities.
Limitations
While some studies incorporated control groups to determine baseline ultrasound parameters,
our study did not include a control group due to a limited sample size. The GAN showed
100% sensitivity, specificity, and diagnostic accuracy; however, this result is based
on only five clinically abnormal GANs. Such a small sample size limits the statistical
reliability and may lead to overestimation. These perfect values should therefore
be interpreted with caution. We acknowledge this as a methodological limitation, and
larger studies are needed to validate these findings. Another key limitation was the
absence of formal interobserver agreement analysis between the two radiologists. While
findings were reviewed jointly to resolve discrepancies in real time, future studies
should include statistical evaluation of interobserver reliability using kappa or
intraclass correlation coefficients to enhance methodological rigor. Additionally,
there are no universally established cutoff or reference values for normal nerve CSAs;
hence, cutoff values in our study were adopted from previously published research.
Follow-up HRUS after initiating treatment for leprosy was not done.
Conclusion
HRUS was able to identify abnormalities in several nerves that appeared normal on
clinical examination. Conversely, some nerves that were clinically abnormal showed
no sonological changes. This indicates that HRUS alone cannot replace clinical evaluation
in assessing peripheral nerve involvement in leprosy. Instead, it would be beneficial
to use HRUS as a complementary tool alongside clinical examination, wherever the technology
is accessible. Doing so may enhance diagnostic accuracy in more patients, especially
since peripheral nerve involvement is a key diagnostic criterion for leprosy. Future
research could explore whether combining HRUS with clinical assessment allows for
earlier and more accurate detection of nerve involvement in leprosy cases.
