Evaluation of the correlation of dorsal root ganglia and spinal nerves with clinical symptoms in patients with postherpetic neuralgia using magnetic resonance neurography

Study design and population

This study was designed with a statistical power of 80.14% using GPOWER software3.1 with the following parameters: two-tailed; α = 0.05; and n = 35. The study was conducted on 60 individuals. Thirty patients with PHN in the thoracic segments admitted to our hospital between January 2021 and February 2022, and 30 age-, sex-, height-, and weight-matched healthy controls were recruited. The patient inclusion criteria were age >18 years and pain lasting for at least 1 month after lesion crusting. The exclusion criteria were as follows: presence of severe, systemic, metabolic, or neurological diseases that could be correlated with peripheral polyneuropathies, namely multiple myeloma, diabetes mellitus, or thyroid disease; history of psychiatric diseases, other chronic pain, or substance abuse; failure or inability to complete MRN scans; and history of thoracic surgery and pain.

The primary objective was to evaluate differences in DRG volume between the lesioned side of patients and the same side of healthy controls. The secondary outcomes were to examine differences in DRG volume between the lesioned and contralateral sides, differences in the largest diameter (D_max) of spinal nerves between the lesioned side of patients and the same side of healthy controls, differences in D_max of the spinal nerves between the lesioned and contralateral sides, and correlations between clinical symptoms and differences in DRG (Fig. 1A).

Since DRG adjacent to the lesion segments may also be affected (Kramer et al., 2019), the volume and M-value of the DRG, as well as the D_max and M-values of the spinal nerves of the first segment below or above the lesioned segments, were also evaluated.

Assessment of pain, allodynia, itching, numbness, and mapping

During the patient’s first visit, a numeric rating scale (NRS; 0 = no pain, 10 = worst pain imaginable) was used to evaluate the average daily pain of patients over the past 48 h. Four areas were mapped on the skin: (1) areas of skin lesions; (2) the most painful area (MPA); (3) areas of dynamic mechanic allodynia (DMA); and (4) areas of sensory disturbance (itching and numbness) (Fig. 1B).

Skin lesion areas were tested with linen tape and estimated using a parallelogram. MPA areas were delineated according to the patients’ descriptions. The distribution of DMA was measured by light stroking with a sterile swab from the normal skin towards the most painful area in steps of 1 cm at intervals of 1 s (Petersen & Rowbotham, 2010; Reda et al., 2013). The areas of itching and numbness were also tested by gently stroking the skin with a sterile swab in steps of 1 cm at intervals of 1 s or according to the patients’ descriptions. The intensity of itching was rated on a 0–10 NRS (0 = no itching, 10 = most itching imaginable). All markings were captured as photographs after obtaining patients’ consent.

Locating lesion dermatomes

The lesion dermatomes of all patients were located according to the schematic of cutaneous nerve distribution in the trunk (Netter, 2003) by two pain specialists with ≥10 years of experience (Xueqin Cao and Xianwei Zhang).

Imaging protocol

All patients underwent MRI scans before the start of treatment. We used a 3.0 Tesla MRI scanner (Magnetom Skyra, Siemens). All participants laid supine on the scanner bed with their thorax immobilized in a tight-fitting thoracic coil, and underwent MRN as follows: Three-dimensional T2-weighted sampling perfection with application-optimized contrasts using different flip-angle evolution short-tau inversion-recovery sequence of the spine was used for DRG and spinal nerve imaging: coronal plane, repetition time/echo time 3,000/178 ms, inversion time 220 ms, the field of view 305 × 240 mm², matrix size 320 × 320, slice thickness 1.0 mm, slice number 60; no gap, voxel size 1.0 × 1.0 × 1.0 mm³, 50% phase over-sampling, and acquisition time 8 min 35 s; imaging the thoracic spine included all bilateral DRG and spinal nerves from the third upper to the third lower segment centered on the lesioned DRG and spinal nerves corresponding to the damaged dermatomes. All bilateral DRG and spinal nerves of the same segments were scanned in the healthy controls.

Image analysis

The DRG volumes were assessed using 3D-slicer version 5.0.3 (http://www.slicer.org/) and defined as the V_DRG. Signal intensity in T2-weighted imaging (M-value) was measured to indirectly reflect inflammation of the DRG and spinal nerves. The Synapse 3D software package (version 5.0.3) was used to measure the D_max of the spinal nerves and the M-values of the DRG and spinal nerves. Measurements of the M-value of the DRG were performed in the selected slice with the largest cross-sectional area. D_max and M-values of the spinal nerves were measured in the selected slice with the largest diameter (Fig. 2). The imaging data were measured and then re-measured after 2 weeks by a radiologist (Donglin Wen) to assess intra-observer reliability. Another radiologist (Gang Wu) independently performed the measurements to assess inter-observer reliability. Both radiologists were blinded to the clinical data. Intra- and inter-observer reliability were evaluated through correlation coefficients (ICCs). ICCs > 0.75 indicated good agreement.

Considering the influence of height and weight on the morphology of the DRG and spinal nerves, the maximum rate of V_DRG difference (ΔV_DRG = V_DRG (lesioned side)–V_DRG (contralateral side)/V_DRG (contralateral side)) was used to evaluate the correlation between V_DRG differences and clinical symptoms. The signal intensity on T2-weighted imaging of each patient’s normal disc was measured as a reference, and the relative M-values of the DRG and spinal nerves (M_DRG/M_disc and M_nerve/M_disc, respectively) were used for assessment.

Treatment and follow-up

The patients’ treatment plans were solely formulated according to the patients’ clinical conditions. Patients were followed up via phone at 1-, 3-, and 6-months post-treatment to assess NRS ratings for pain and itching intensities.

Statistical analysis

Data analyses were performed using SPSS 22 software (SPSS Inc., Chicago, IL, USA). The normality of continuous variables was confirmed with Shapiro–Wilk tests, and the homogeneity of variance was evaluated using Levene’s test. Normally distributed data were described as $\overline{\chi }$ ± SD. The data were compared between the two groups using paired-samples or independent-samples t-tests. Non-normal data were described as the median (interquartile range), and between-group differences were examined using the Wilcoxon or Mann–Whitney U tests. Correlations among normally distributed continuous data were analyzed using Pearson’s correlation coefficient and simple linear regression. Non-normal continuous data and categorical variables were evaluated using Spearman’s correlation and the Chi-square test. Statistically significant variables were selected for multilinear regression analysis. The assumption of independence for the residuals was assessed using the Durbin–Watson test. The assumptions of normality and homogeneity of variance for residuals were assessed using histograms and residual plots, respectively. Multicollinearity was assessed using Pearson’s correlation coefficient statistics and by checking the variance inflation factor for the same dependent and independent variables. All statistical tests were two-sided, and statistical significance was set at p < 0.05.

Characteristics of patients and healthy controls

The mean age of the patients was 60.7 ± 10.18 years. Forty-nine clinical lesion dermatomes were identified in 30 patients. Age, sex, height, and weight did not differ between the patient and healthy control groups (all p > 0.050). The demographic and clinical characteristics of the two groups are summarized in Tables 1 and 2, respectively. Intra-and inter observer reliability was evaluated through correlation coefficients and ICCs were 0.91 and 0.80, respectively, indicating good stability.

Comparison of DRG and spinal nerves between the lesioned side of patients and the same side of healthy controls

The mean V_DRG and relative M-value of DRG were significantly higher in the lesioned side of patients than in the same side of healthy controls (p = 0.013, p < 0.001, respectively). The mean D_max and relative M-value of the spinal nerves were significantly higher on the lesioned side of patients than on the same side of healthy controls (p < 0.001, p < 0.001, respectively) (Fig. 3).

Comparison of DRG and spinal nerves between the lesioned and contralateral sides in patients

No difference was found in the mean V_DRG and relative M-value of DRG between the lesioned and contralateral sides (p = 0.089, p = 0.715, respectively). The mean D_max and relative M-value of the spinal nerves were significantly higher on the lesioned side than on the contralateral side (p = 0.001, p = 0.005, respectively) (Fig. 4).

Correlation of ΔV_DRG with age, sex, disease duration, and skin lesion area

Age and sex were not correlated with ΔV_DRG (p = 0.738 and p = 0.456, respectively). Disease duration and skin lesion area were significantly negatively correlated with ΔV_DRG (r = −0.435, p = 0.003; r = −0.540, p = 0.002, respectively).

Correlation of ΔV_DRG with allodynia, numbness, itching, pain severity, and itching severity

Itching and ΔV_DRG were significantly positively correlated (r = 0.443, p = 0.014). The NRS values for pain and itching severity were not significantly associated with ΔV_DRG (p = 0.929 and p = 0.228, respectively). Additionally, allodynia (p = 0.360) and numbness (p > 0.999) were not significantly correlated with ΔV_DRG.

Effect of disease duration, area of skin lesion, and itching on ΔV_DRG

Multiple linear regression analysis showed that itching and disease duration were independent risk factors for ΔV_DRG (β = 0.376, p = 0.017; β = −0.388, p = 0.013, respectively). Skin lesion area was not significant independent risk factor for ΔV_DRG (β = −0.163, p = 0.308) (Table 3).

Difference in morphology and M-value of bilateral DRG and spinal nerves in the first segment adjacent to lesion segments

No significant difference was found in DRG or spinal nerves between the two sides in the first segment adjacent to lesion segments (p > 0.05). Also, no significant difference was found in DRG or spinal nerves between the two sides in the third segment away from the lesion segments (p > 0.05).

NRS score of pain and itching during follow-up

No significant correlations were found between ΔV_DRG and itching severity at 1-, 3-, and 6-months post-treatment (p = 0.097, p = 0.808, and p = 0.295, respectively). Similar results were observed for the correlations between ΔV_DRG and pain severity at 1-, 3-, and 6-months post-treatment (p = 0.124, p = 0.790, and p = 0.295, respectively). However, patients with positive ΔV_DRG recovered better than those with negative ΔV_DRG. At 3 months post-treatment, all patients reported an NRS score of pain and itching of <5. Among the 12 patients with negative ΔV_DRG, the incidence of NRS >3 was 25% (pain) and 8.3% (itching); this result was 16.7% (pain) and 0% (itching) in the 18 patients with positive ΔV_DRG. At 6 months post-treatment, the NRS scores for pain and itching were <5. Among the 12 patients with negative ΔV_DRG, the incidence of NRS >3 was 41.7% (pain) and 16.7% (itching); this result was 0% (pain) and 0% (itching) in the 18 patients with positive ΔV_DRG.