Spinal trigeminal nucleus lesions in trigeminal zoster-associated neuralgia: A retrospective cohort study on diagnostic value and prognostic factors

Abstract

Background

Trigeminal zoster-associated neuralgia (TZAN) is a refractory intractable craniofacial neuropathic pain with a high risk of progressing to postherpetic neuralgia. Abnormal T2-weighted magnetic resonance imaging (MRI) hyperintensity in the spinal trigeminal nucleus (STN) has been reported in TZAN patients, but its diagnostic value, influencing factors and association with prognosis remain insufficiently explored. This study aimed to investigate the correlation between STN lesions and TZAN and identify factors affecting STN signal characteristics and TZAN therapeutic efficacy.

Methods

This retrospective cohort study included 105 TZAN patients, 105 non-TZAN normal controls (NC) and 287 classical trigeminal neuralgia (CTN) patients who underwent cranial MRI at Shandong Provincial Hospital Affiliated to Shandong First Medical University between September 2018 and March 2024. Propensity score matching (1:1, caliper = 0.1) was used to balance baseline differences between the TZAN and CTN groups. STN lesions were evaluated in a blinded manner by two radiologists, and clinical data (pain Numeric Rating Scale scores, pregabalin dosage) were collected via medical records and 6-month telephone follow-up. Multivariate logistic regression analyzed factors associated with STN hyperintensity and TZAN efficacy (pain relief ≥ 80% defined as “excellent response”).

Results

STN lesion detection rate was significantly higher in TZAN than in NC (62.9% vs. 1.0%, p < 0.001) and matched CTN (68.9% vs. 0%, p < 0.001), with 62.9% sensitivity, 99.0% specificity and 81.0% accuracy for TZAN diagnosis. Age (odds ratio (OR) = 1.06, 95% confidence interval (CI) = 1.006–1.116, p = 0.030) and subacute phase (vs. acute: OR = 32.01; vs. chronic: OR = 46.40, both p < 0.001) independently predicted STN hyperintensity. STN-normal patients had better short-term efficacy (discharge Numeric Rating Scale = 2 (1, 3) vs. 3 (2, 4), p = 0.042; excellent response rate: 38.5% vs. 19.7%, p = 0.036) but no long-term difference (3 and 6 months post-discharge, p > 0.05). Chronic phase predicted poor prognosis (vs. acute: OR = 6.55, p = 0.005; vs. subacute: OR = 5.39, p = 0.017).

Conclusions

STN lesions are highly specific for TZAN and may serve as an auxiliary diagnostic indicator. STN hyperintensity is most prominent in elderly TZAN patients in the subacute phase, potentially acting as a subacute-phase imaging marker. Early intervention in subacute TZAN may be critical for improving prognosis.

Keywords

diagnostic marker magnetic resonance imaging (MRI)postherpetic neuralgia (PHN)spinal trigeminal nucleus (STN)trigeminal zoster-associated neuralgia (TZAN)

Introduction

Herpes zoster (HZ) is caused by the reactivation of varicella-zoster virus (VZV) in sensory ganglia, and zoster-associated neuralgia (ZAN) affects 10–30% of patients with HZ.¹ Trigeminal zoster-associated neuralgia (TZAN), a subtype of ZAN involving the trigeminal nerve, accounts for 58% of head and facial HZ cases, with a higher risk of progressing to postherpetic neuralgia (PHN) (35–50%) than trunk or limb ZAN.² TZAN is characterized by severe burning or stabbing pain, allodynia, and sleep or anxiety disorders, imposing substantial physical and socioeconomic burdens.²

The spinal trigeminal nucleus (STN) is a key relay station for orofacial nociception, integrating pain signals via its caudal subnucleus (also known as the “medullary dorsal horn”).³ Recent case reports have described T2-weighted imaging (T2WI) hyperintensity in the STN of TZAN patients, which is attributed to inflammation caused by the anterograde spread of VZV.^4–6 Niemeyer et al.³ further showed that VZV triggers the release of proinflammatory cytokines (interleukin (IL)-6 and tumor necrosis factor (TNF)-α) in the trigeminal ganglion and STN, exacerbating neural damage and pain. However, most existing studies are case-based, and systematic data on the prevalence of STN lesions, their influencing factors and their relevance to TZAN prognosis remain lacking.

This retrospective cohort study analyzed the clinical and imaging data of TZAN patients to: (i) compare the rates of STN lesions across the TZAN, normal control (NC) and classical trigeminal neuralgia (CTN) groups; (ii) identify factors affecting STN signals; and (iii) explore the association between STN lesions and efficacy of treatment in TZAN. The findings aim to provide evidence for TZAN diagnosis and early intervention.

Methods

Participants

This retrospective cohort study was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement. This retrospective study was approved by the Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong First Medical University (Approval No.: SWYX: 2025-076) and adhered to the Declaration of Helsinki. Written informed consent was waived given the retrospective nature of the study, while verbal informed consent was obtained and documented during telephone follow-up. All data were anonymized to protect patient privacy.

Patients who underwent cranial magnetic resonance imaging (MRI) at Shandong Provincial Hospital Affiliated to Shandong First Medical University between 1 September 2018 and 31 March 2024, were screened for inclusion. The inclusion criteria for the TZAN group were: a definite history of head/facial HZ in trigeminal dermatomes (confirmed by dermatologists); moderate-to-severe pain (Numeric Rating Scale (NRS) ≥ 4) at inpatient admission; age ≥ 18 years; and complete clinical data and 6-month follow-up data. All TZAN subjects were treated with analgesic medications combined with interventional therapies, including nerve block and neuromodulation. For the CTN group, inclusion criteria were adherence to the International Classification of Headache Disorders, 3rd edition, criteria,⁷ moderate-to-severe pain (NRS ≥ 4) at inpatient admission, age ≥ 18 years, and complete clinical data and 6-month follow-up data. The NC group comprised non-TZAN patients without any head/facial pain or relevant disorders, who had complete clinical and imaging data; all NC subjects were recruited from the health check-up population and non-pain clinic inpatients of our hospital during the same study period.

Baseline characteristics

Demographic factors (age, gender, body mass index), smoking/alcohol history and comorbidities (hypertension, diabetes, coronary heart disease and cerebral infarction) were systematically recorded as potential confounders. Additionally, disease course (acute phase (AHN, < 30 days), subacute phase (SHN, 30–90 days), chronic phase (PHN, > 90 days)) and affected side were recorded for the TZAN and CTN groups. For both groups, cranial MRI examinations were completed within 2 days after inpatient admission, and so the timing of MRI scanning was deemed consistent with the disease course staging by default.

MRI was performed using a 3.0T MAGNETOM Prisma scanner (Siemens, Erlangen, Germany) with the following parameters: repetition time (TR) = 4000 ms, echo time (TE) = 100 ms, matrix = 320 × 156, slice thickness = 5 mm and 3-mm thin-slice reconstruction for the STN region (the medulla oblongata, ranging from the pyramidal decussation to the upper olive level) (Figure 1). Two radiologists with ≥ 5 years of neuroradiology experience independently reviewed T2WI images in a blinded manner (unaware of patients’ clinical diagnoses). Disagreements were resolved by a senior radiologist with ≥ 10 years of experience. Examples of typical imaging findings are shown in Figure 2.

Figure 1.

Horizontal section of the medulla oblongata (at the level of the pyramidal decussation).

Figure 2.

Brain magnetic resonance imaging (MRI) examination image (T2-weighted imaging). (a) MRI shows abnormally high signal intensity in the left spinal trigeminal nucleus (STN) of the medullary plane at decussation of lemniscus height (arrow). (b) MRI shows abnormally high signal intensity in the right STN of the medullary plane at decussation of pyramids height (arrow). (c) MRI shows abnormally weak high signal intensity in the left STN of the medullary plane at the height of the upper olive (arrow). (d) MRI shows abnormally high signal intensity with a large area in the left STN of the medullary plane at the height of the middle olive (arrow).

To minimize the impact of potential confounders, propensity score matching (PSM) (1:1 nearest-neighbor matching, caliper = 0.1, no replacement) was used to balance baseline differences between the TZAN and CTN groups. Matching variables included age, gender, disease course, smoking history, diabetes and coronary heart disease.

(A) MRI shows abnormally high signal intensity in the left STN of the medullary plane at decussation of lemniscus height (arrow); (B) MRI shows abnormally high signal intensity in the right STN of the medullary plane at decussation of pyramids height (arrow); (C) MRI shows abnormally weak high signal intensity in the left STN of the medullary plane at the height of the upper olive (arrow); and (D) MRI shows abnormally high signal intensity with a large area in the left STN of the medullary plane at the height of the middle olive (arrow).

Outcome measures and follow-up

To investigate the association between STN lesions and the efficacy of therapy in TZAN, the following outcome parameters were assessed at baseline, discharge, 3 months post-discharge and 6 months post-discharge: pain intensity, pain relief rate, sleep quality, medication use and complications. Pain intensity was evaluated using the validated NRS, a 0–10 scoring system with ‘0’ indicating no pain and ‘10’ representing the most severe pain, while the pain relief rate was calculated using: [(Baseline NRS score – Follow-up NRS score)/Baseline NRS score] × 100%. A simplified sleep quality assessment was employed to evaluate patients’ sleep status, categorized as excellent (difficulty falling asleep, defined as inability to fall asleep within 30 min, ≤ 1 episode per week, normal sleep duration), moderate (difficulty falling asleep > 1 episode per week and ≤ 3 episodes per week with slightly shortened sleep duration) or poor (difficulty falling asleep > 3 episodes per week with short sleep duration). Complications assessed in this study included pain recurrence, pruritus, numbness and other related symptoms.

Statistical analysis

All data compilation and statistical analysis were performed using Excel (Microsoft Corp., Redmond, WA, USA), SPSS, version 27.0 (IBM Corp., Armonk, NY, USA), R, version 4.4.2 (R Foundation, Vienna, Austria) and Prism, version 10.1.2 (GraphPad Software Inc., San Diego, CA, USA). All results were analyzed using descriptive statistical methods. The Shapiro–Wilk test was used to test the normality of continuous data. Data conforming to a normal distribution were expressed as the mean ± standard deviation; data not conforming to a normal distribution were expressed as median (first quartile, third quartile) (M (Q1, Q3)). Categorical data were described using case count (percentage) (n (%)). For inter-group comparisons, an independent samples t-test was used for variables with a normal distribution and a Mann–Whitney U-test or a chi-squared test was used for variables without a normal distribution. Multivariate binary logistic regression analysis was conducted to evaluate factors influencing STN signals and TZAN efficacy, with odds ratio (OR) and 95% confidence interval (CI) calculated. P < 0.05 was considered statistically significant.

Results

As illustrated in Figure 3, 524 patients with CTN and 219 patients with TZAN were admitted to the Department of Pain Medicine. Among the 524 CTN patients, 237 were excluded due to missing MRI data. Ultimately, 497 patients were enrolled in this study, comprising 105 patients in the NC group, 287 in the CTN group and 105 in the TZAN group. Additionally, the 105 TZAN patients were further subclassified into two groups based on cranial MRI findings: 66 with STN hyperintensity (STN-HS group) and 39 with normal STN signals (STN-NS group), with the symptomatic side of head and facial pain being completely consistent with the side of STN hyperintensity in all patients. Patients were also categorized into the excellent response group (n = 55) and the non-excellent response group (n = 50) according to whether pain relief was ≥ 80% at the 6-month follow-up.

Figure 3.

Patient flowchart. CTN, classical trigeminal neuralgia; MRI, magnetic resonance imaging; NC, normal control; NRS, Numeric Rating Scale; STN, spinal trigeminal nucleus; TZAN, trigeminal zoster-associated neuralgia.

Comparison between TZAN group and NC group

Although demographic characteristics (age, gender and body mass index), lifestyle factors (smoking and alcohol consumption history) and comorbidities were systematically compared between the TZAN and NC groups, no statistically significant differences were identified across any of these variables (all p > 0.05) (Table 1). Notably, STN lesions were detected in 62.9% (66/105) of TZAN patients, whereas this finding was extremely rare in the NC group, occurring in only 1.0% (1/105) of cases, comprising a difference that reached statistical significance (χ² = 92.605, p < 0.001) (Table 2). When STN hyperintensity on T2WI was evaluated as a diagnostic marker for TZAN, it exhibited a sensitivity of 62.9%, a specificity of 99.0% and an overall accuracy of 81.0%.

Table 1.

Baseline characteristics of the trigeminal zoster-associated neuralgia (TZAN) group and normal control (NC) group.

	TZAN group（n = 105）	NC group （n = 105）	t/Z/χ²	p
Age (years)	66.00 (59.00, 71.00）	62.00 (56.00, 70.00)	−1.441	0.150
Gender (n, %)			0.020	0.888
Male	63 (60.0)	64 (61.0)
Female	42 (40.0)	41 (39.0)
Body mass index (kg/m²)	24.37 ± 3.75	23.98 ± 4.16	0.704	0.482
History of tobacco (n, %)	14 (13.3)	14 (13.3)	0.000	1.000
History of alcohol (n, %)	12 (11.4)	15 (14.3)	0.383	0.536
Comorbidities (n, %)
Hypertension (n, %)	39 (37.1)	38 (36.2)	0.021	0.886
Diabetes (n, %)	31 (29.5)	29 (27.6)	0.093	0.760
Coronary heart disease (n, %)	14 (13.3)	16 (15.2)	0.156	0.693
Cerebral infarction (n, %)	9 (8.6)	10 (9.5)	0.058	0.810

Table 2.

Comparison of spinal trigeminal nucleus (STN) lesion between the trigeminal zoster-associated neuralgia (TZAN) group and normal control (NC) group (n (%)).

STN lesion	TZAN group (n = 105)	NC group (n = 105)	χ²	p
Positive	66 (62.9)	1 (1.0)	92.605	< 0.001
Negative	39 (37.1)	104 (99.0)	92.605	< 0.001

Comparison between TZAN group and CTN group

Prior to PSM, the TZAN and CTN groups differed significantly with respect to several baseline characteristics, including gender distribution, disease course, smoking history and the prevalence of diabetes mellitus and coronary heart disease (all p < 0.05). Following 1:1 PSM (caliper = 0.1), these baseline discrepancies were effectively balanced, with no statistically significant differences observed between the two groups (all p > 0.05) (Table 3). Post-PSM analysis revealed that STN lesions remained highly prevalent in the TZAN group (68.9%, 31/45) but were completely absent in the matched CTN group (0%, 0/45), a distinction that was statistically robust (χ² = 47.288, p < 0.001) (Table 4).

Table 3.

Baseline characteristics of the trigeminal zoster-associated neuralgia (TZAN) group and classical trigeminal neuralgia (CTN) group before and after propensity score matching (PSM).

	Before PSM				After PSM
	TZAN group （n = 105）	CTN group （n = 287）	t/Z/χ²	p	TZAN group （n = 45）	CTN group （n = 45）	t/Z/χ²	p
Age (years)	66.00 (59.00, 71.00)	64.00 (55.00, 70.50)	−1.365	0.172	68.00 (60.00, 71.00)	67.00 (57.00, 73.00)	−0.008	0.994
Gender (n, %)			17.753	< 0.001**			0.000	1.000
Male	63 (60.0)	104 (36.2)			25 (44.4)	25 (44.4)
Female	42 (40.0)	183 (63.8)			20 (55.6)	20 (55.6)
BMI (kg/m²)	24.37 ± 3.75	25.16 ± 3.65	−1.854	0.065	24.31 ± 3.83	24.56 ± 3.18	−0.348	0.729
Disease course (n, %)			204.250	< 0.001**			0.197	1.000
< 30 days	50 (47.6)	4 (1.4)			5 (11.1)	4 (8.9)
30–90 days	36 (34.3)	27 (9.4)			21 (46.7)	22 (48.6)
< 90 days	19 (18.1)	256 (89.2)			19 (42.2)	19 (42.2)
Side (n, %)			4.125	0.042			0.185	0.667
Left	56 (53.3)	120 (41.8)			26 (57.8)	28 (62.2)
Right	49 (46.7)	167 (58.2)			19 (42.2)	17 (37.8)
History of tobacco (n, %)	14 (13.3)	15 (5.22)	7.375	0.007**	6 (13.3)	9 (20.0)	0.720	0.396
History of alcohol (n, %)	12 (11.4)	17 (5.9)	3.401	0.065	4 (8.9)	6 (13.3)	0.450	0.502
Comorbidities (n, %)
Hypertension (n, %)	39 (37.1)	110 (38.3)	0.046	0.831	12 (26.7)	12 (26.7)	0.000	1.000
Diabetes (n, %)	31 (30.0)	26 (9.1)	25.909	< 0.001**	11 (24.4)	5 (11.1)	2.736	0.098
Coronary heart disease (n, %)	14 (13.3)	19 (6.6)	4.494	0.034*	4 (8.9)	4 (8.9)	0.000	1.000
Cerebral infarction (n, %)	9 (8.6)	13 (4.5)	2.371	0.124	3 (6.7)	0	1.379	0.240

Note: BMI, body mass index.

Table 4.

Comparison of spinal trigeminal nucleus (STN) lesions between the trigeminal zoster-associated neuralgia (TZAN) group and classical trigeminal neuralgia (CTN) group after propensity score matching (PSM) (n (%)).

STN lesion	TZAN group (n = 45)	CTN group (n = 45)	χ²	p
Positive	31 (68.9)	0	47.288	< 0.001**
Negative	14 (31.1)	45 (100.0)	47.288	< 0.001**

Note: *p < 0.01.

Comparison between STN-hs group and STN-NS group

A significant difference in disease course was observed between the STN-HS and STN-NS groups (p < 0.001): the STN-HS group had a higher proportion of SHN (51.5% vs. 5.1%) and lower proportions of AHN (36.4% vs. 66.7%) and PHN (12.1% vs. 28.2%), with no other baseline variables differing significantly (all p > 0.05) (Table 5).

Table 5.

Baseline characteristics of the spinal trigeminal nucleus (STN) hyperintensity (STN-HS) group and normal STN signals (STN-NS) group.

	STN-HS group (n = 66)	STN-NS group (n = 39)	t/Z/χ²	p
Age (years)	65.33 ± 8.64	62.26 ± 11.37	1.459	0.149
Gender (n, %)			0.333	0.564
Male	41 (62.1)	22 (56.4)
Female	25 (37.9)	17 (43.6)
BMI (kg/m²)	24.18 (21.55, 26.68)	24.73 (21.89, 26.96)	−0.849	0.396
Disease course (n, %)			23.617	< 0.001**
< 30 days	24 (36.4)	26 (66.7)
30–90 days	34 (51.5)	2 (5.1)
> 90 days	8 (12.1)	11 (28.2)
Side (n,%)			0.236	0.627
Left	34 (51.5)	22 (56.4)
Right	32 (48.5)	17 (43.6)
Related nerves (n, %)			3.392	0.663
I: ophthalmic division	20 (30.3)	17(43.6)
II: maxillary division	8 (12.1)	4 (10.3)
III: mandibular division	2 (3.0)	0
I+II	17 (25.8)	7 (17.9)
II+III	16 (24.2)	8 (20.5)
I+II+III	3 (4.5)	3 (7.7)
Allodynia (n, %)	53 (80.3)	33 (84.6)	0.308	0.579
Explosive pain (n, %)	54 (81.8)	29 (74.4)	0.824	0.364
History of tobacco (n, %)	9 (13.6)	5 (12.8)	0.014	0.905
History of alcohol (n,%)	8 (12.1)	4 (10.3)	0.000	1.000
Comorbidities (n,%)
Hypertension (n, %)	23 (34.8)	16 (41.0)	0.401	0.527
Diabetes (n, %)	17 (25.8)	14 (35.9)	1.211	0.271
Coronary heart disease (n, %)	10 (15.2)	4 (10.3)	0.508	0.476
Cerebral infarction (n, %)	7 (10.6)	2 (5.1)	0.370	0.543
Base NRS (scores)	8 (7,10)	8 (7.10)	−0.283	0.777
Base sleep quality (n, %)			−1.394	0.163
Excellent	1 (1.5)	1 (2.6)
Medium	10 (15.2)	10 (25.6)
Poor	55 (83.3)	28 (71.8)

Note: BMI, body mass index; NRS, Numerical Rating Scale.

Multivariate logistic regression analysis, adjusted for potential confounders, identified age (OR = 1.06, 95% CI = 1.006–1.116, p = 0.030) and disease course as independent predictors of STN hyperintensity, with the SHN phase associated with a substantially higher risk compared to both AHN (OR = 32.01, p < 0.001) and PHN (OR = 46.40, p < 0.001) (Table 6). The prevalence of STN lesions varied dynamically across disease stages: 48.0% (AHN), 94.4% (SHN), 77.8% (PHN, 3–12 months) and 10.0% (PHN, > 12 months) (Figure 4).

Figure 4.

The positive rate of spinal trigeminal nucleus (STN) lesions changes with the course of the disease course. Bubble plot showing STN lesion positive rates by disease course; bubble size represents sample size.

Table 6.

Multivariate binary logistic regression analysis of spinal trigeminal nucleus (STN) signals.

	Reference	p	OR	95% CI
Age	–	0.030*	1.06	1.006–1.116
Base sleep quality		0.168
Medium	Excellent	0.641	2.34	0.066–83.619
Poor	Excellent	0.271	7.06	0.217–229.362
Disease course		< 0.001**
SHN (30–90 days)	AHN (< 30 days)	< 0.001**	32.01	5.482–186.933
AHN (< 30 days)	PHN (> 90 days)	0.526	1.45	0.461–4.561
SHN (30–90 days)	PHN (> 90 days)	< 0.001**	46.40	6.640–324.255

Note: AHN, acute herpetic neuralgia; CI, confidence interval; OR, odds ratio; PHN, postherpetic neuralgia; SHN, subacute herpetic neuralgia.

Clinically, the STN-NS group achieved lower discharge NRS (2 (1, 3) vs. 3 (2, 4) p = 0.042) and a higher excellent response rate (38.5% vs. 19.7%, p = 0.036) than the STN-HS group, but this short-term advantage did not persist at 3- or 6-month follow-up (all p > 0.05) (Table 7 ). No significant differences in daily pregabalin dosage (admission, discharge and follow-up) or complication rates (pain recurrence, pruritus and numbness) were noted (all p > 0.05) (see Supplementary material, Tables S1 and S2) and no serious adverse events (neurovascular injury, wound infection) were reported.

Table 7.

Comparison of treatment outcome between the spinal trigeminal nucleus (STN) hyperintensity (STN-HS) group and normal STN signals (STN-NS)group (6-month follow-up).

	STN-HS group (n = 66)	STN-NS group (n = 39)	Z/χ²	p
NRS (scores)
On inpatient admission	8 (7,10)	8 (7,10)	−0.283	0.777
On inpatient discharge	3 (2,4)	2 (1,3)	−2.035	0.042*
3 months after discharge	2 (1,3)	1 (0,5)	−0.856	0.392
6 months after discharge	2 (1,3)	1 (0,4)	−0.664	0.507
Excellence rate (n, %)
On inpatient discharge	13 (19.7)	15 (38.5)	4.414	0.036*
3 months after discharge	29 (43.9)	20 (51.3)	0.531	0.466
6 months after discharge	34 (51.5)	21 (53.8)	0.053	0.817
Effective rate (n, %)
On inpatient discharge	51 (77.3)	34 (87.2)	1.560	0.212
3 months after discharge	53 (80.3)	30 (76.9)	0.169	0.681
6 months after discharge	54 (81.8)	31 (79.5)	0.086	0.769
Sleep quality (excellent/medium/poor)
On inpatient admission	1/10/55	1/10/28	−1.394	0.163
On inpatient discharge	24/36/6	17/18/4	0.555	0.579
3 months after discharge	32/31/3	22/11/6	−0.189	0.850
6 months after discharge	31/31/4	20/15/4	−0.174	0.862

Note: NRS, Numerical Rating Scale.

Table 8.

Baseline characteristics of excellent response group and non-excellent response group.

	Excellence group (n = 55)	Non-excellence group (n = 50)	t/Z/χ²	p
Age (years)	66.00 (54.00, 71.00）	67.00 (61.75, 71.25)	−1.535	0.125
Gender (n, %)			2.545	0.111
Male	37 (67.3)	26 (52.0)
Female	18 (32.7)	24 (48.0)
BMI (kg/m²)	24.34 ± 3.82	24.40 ± 3.71	−0.074	0.941
Disease course (n, %)			9.476	0.009**
< 30 days	31 (56.4)	19 (38.0)
30–90 days	20 (36.4)	16 (32.0)
> 90 days	4 (7.3)	15 (30.0)
Side (n, %)			2.062	0.151
Left	33 (60.0)	23 (46.0)
Right	22 (40.0)	27 (54.0)
Related nerves (n, %)			2.551	0.819
I: ophthalmic division	20 (36.4)	17 (34.0)
II: maxillary division	6 (10.9)	6 (12.0)
III: mandibular division	2 (3.6)	0
I+II	13 (23.6)	11 (22.0)
II+III	12 (21.8)	12 (24.0)
I+II+III	2 (3.6)	4 (8.0)
Allodynia (n, %)	48 (87.3)	38 (76.0)	2.246	0.134
Explosive pain (n, %)	45 (81.8)	38 (76.0)	0.535	0.464
High signal of STN (n, %)	34 (61.8)	32 (64.0)	0.053	0.817
History of tobacco (n, %)	8 (14.5)	6 (12.0)	0.147	0.702
History of alcohol (n, %)	8 (14.5)	4 (8.0)	1.109	0.292
Comorbidities (n, %)
Hypertension (n, %)	21 (38.2)	18 (36.0)	0.053	0.817
Diabetes (n, %)	16 (29.1)	15 (30.0)	0.010	0.919
Coronary heart disease (n, %)	9 (16.4)	5 (10.0)	0.918	0.338
Cerebral infarction (n, %)	4 (7.3)	5 (10.0)	0.022	0.881
Base NRS (scores)	8.00 (7.00, 10.00)	8.00 (7.00, 10.00)	−0.301	0.764
Base sleep quality (n, %)			−0.686	0.493
Excellent	1 (1.8)	1 (2.0)
Medium	12 (21.8)	8 (16.0)
Poor	42 (76.4)	41 (82.0)

Note: AHN, acute herpetic neuralgia; BMI, body mass index; NRS, Numerical Rating Scale; PHN, postherpetic neuralgia; SHN, subacute herpetic neuralgia; STN, spinal trigeminal nucleus.

Comparison between excellent response group and non-excellent response group

Comparison between the excellent response and non-excellent groups at baseline revealed a statistically significant difference in disease course (Table 8). Multivariate logistic regression analysis demonstrated that disease course was a statistically significant factor associated with efficacy of treatment in TZAN (p < 0.05). Patients in the chronic phase (PHN) had a 6.55-fold higher risk of non-excellent response compared to those in the acute phase (AHN) (OR = 6.55, 95% CI = 1.774–24.148, p = 0.005) and a 5.39-fold higher risk than those in the subacute phase (SHN) (OR = 5.39, 95% CI = 1.360–21.379, p = 0.017). No significant difference in treatment efficacy was observed between the acute phase and subacute phase (p> 0.05) (Table 9).

Table 9.

Multivariate binary logistic regression analysis of efficacy of treatment in trigeminal zoster-associated neuralgia (TZAN).

	Reference	p	OR	95% CI
Age		0.368	1.02	0.976–1.069
Gender	Male	0.194	1.79	0.743–4.331
Side	Left	0.237	1.68	0.711–3.960
Allodynia	Negative	0.156	0.44	0.145–1.363
Disease course		0.017*
SHN (30–90 days)	AHN (< 30 days)	0.687	1.21	0.473–3.119
PHN (> 90 days)	AHN (< 30 days)	0.005**	6.55	1.774–24.148
PHN (> 90 days)	SHN (30–90 days)	0.017*	5.39	1.360–21.379

Note: AHN, acute herpetic neuralgia; CI, confidence interval; OR, odds ratio; PHN, postherpetic neuralgia; SHN, subacute herpetic neuralgia.

Discussion

To our knowledge, this is the first large-cohort study to demonstrate that STN lesions exhibit high specificity (99.0%) for TZAN, consistent with earlier case reports^4–6 but extending these observations to systematic, large-scale data. Importantly, the laterality of STN hyperintensity was fully concordant with the symptomatic side in all TZAN patients, and the rate of STN lesions was 62.9% in TZAN, compared to only 1.0% in NC and 0% in PSM CTN. These findings confirm that STN damage is not a non-specific manifestation of head and facial pain but a TZAN-specific radiological feature. It is hypothesized that the underlying pathway entails the transaxonal spread of VZV, where reactivation of latent VZV in the trigeminal ganglion prompts the virus to spread along axons to the STN, upregulate proinflammatory cytokines (IL-6 and TNF-α) and thereby trigger localized inflammatory responses and induce T2WI hyperintensity.^8,9 Clinically, STN hyperintensity offers valuable diagnostic utility for atypical TZAN cases (e.g. those without a characteristic rash¹⁰), which often overlap with conditions such as lateral medullary infarction (a disorder also linked to STN hyperintensity)^11,12; integrating STN imaging with VZV serology or cerebrospinal fluid (CSF) testing may thus refine differential diagnosis in ambiguous scenarios.

Although the age difference between groups was borderline significant (p = 0.149), multivariate logistic regression identified age as an independent predictor of STN hyperintensity (OR = 1.06, 95% CI = 1.006–1.116, p = 0.030). This finding suggests that age-related declines in immune function and impaired neuronal repair capacity may increase vulnerability to VZV-induced STN injury,¹³ although its clinical significance requires validation in larger cohorts. Multivariate analysis further identified the subacute phase as another independent predictor of STN hyperintensity. Notably, the detection rate of STN lesions reached 94.4% in the subacute phase (30–90 days) compared to only 48.0% in the acute phase (< 30 days). This difference may reflect the several-week window required for anterograde transaxonal spread of VZV from the trigeminal ganglion to the STN, such that STN involvement in the acute phase is often too early to be detected by conventional MRI. In the chronic phase (> 12 months), the detection rate drops to 10.0%, a phenomenon potentially explained by either resolution of VZV-induced inflammation following antiviral therapy¹⁴ or reduced T2 relaxation time due to astrocyte activation.¹⁵ These results redefine STN hyperintensity as a subacute-phase-specific marker (rather than a general “early-stage” indicator), providing actionable guidance for clinical imaging timing: MRI performed during the subacute phase maximizes the detection of STN lesions in TZAN patients.

The present study compared the impact of STN lesions on efficacy of treatment in TZAN. Patients in the STN-NS group exhibited superior short-term therapeutic outcomes at discharge, including lower NRS scores (2 (1, 3) vs. 3 (2, 4), p = 0.042) and a higher excellent response rate (38.5% vs. 19.7%, p = 0.036), whereas no significant differences in long-term efficacy were observed at other follow-up time points (3 and 6 months) between the STN-NS and STN-HS groups. Consistent with this, logistic regression analysis revealed no significant correlation between STN hyperintensity and long-term therapeutic effects. Previous studies have demonstrated that microglia in the caudal subnucleus of the STN are activated during tissue injury and inflammatory responses, with their hyperreactivity playing a crucial role in the initiation and maintenance of chronic orofacial pain.^16,17 We thus hypothesize that VZV invasion of the STN induces microglial activation, which in turn activates astrocytes, enhances the excitability of nociceptive neurons, and promotes pain progression,¹⁸ ultimately leading to poorer short-term efficacy in patients with STN lesions. Systemic treatment effectively inhibits microglial activation, thereby attenuating its impact on therapeutic outcomes over time.¹⁹ Of critical importance, the chronic phase emerged as the sole independent predictor of poor prognosis (OR = 6.55), consistent with the well-established refractory nature of PHN.^20,21 This finding suggests the potential necessity of early intervention during the subacute phase, via targeted therapies (e.g. antiviral agents combined with anti-inflammatory drugs) to mitigate STN inflammation and prevent central sensitization.

This study has several limitations that warrant acknowledgment. First, its retrospective design may introduce selection bias, as only TZAN patients with moderate-to-severe pain and typical skin lesions were included; future prospective studies should therefore incorporate milder cases as well as zoster sine herpete (rash-negative) patients to enhance generalizability. Second, imaging assessments were limited to conventional MRI, with no dynamic imaging to track the evolution of STN signals over time; repeated MRI scans from the subacute to chronic phases could clarify the persistence and reversibility of STN lesions. Due to the insufficient sample size, quantitative analysis of STN hyperintensity signal intensity was not performed in this study; STN hyperintense lesions exhibit signal strength variability (e.g. faint lesions in Figure 2c), and, although the signal intensity may be associated with the severity of viral infection and inflammatory response, direct clinical evidence is currently lacking. Third, mechanistic insights are constrained by the lack of CSF VZV DNA testing and STN cytokine profiling, which would help validate the direct association between VZV and STN damage. Finally, sleep quality was evaluated using a simplified scale; integrating comprehensive tools such as the SF-36 or DN4 scales (22) in future studies would provide more robust assessments of quality of life and neuropathic pain characteristics.

Conclusions

In conclusion, STN lesions exhibit high specificity (99.0%) for TZAN and thus serve as a valuable auxiliary diagnostic indicator, particularly for atypical TZAN cases. Age and the subacute phase are key risk factors for STN hyperintensity, supporting the potential of STN hyperintensity as a subacute-phase imaging marker for TZAN. While STN lesions influence short-term therapeutic efficacy in TZAN, they do not affect long-term prognosis. Critically, the chronic phase of TZAN predicts poor treatment outcomes, highlighting the necessity of early intervention during the subacute phase to optimize patient outcomes. More prospective dynamic studies are needed to establish the prognostic role of subacute STN-HS and their role in guiding treatment to potentially decrease the incidence of trigeminal PHN.

Article highlights

STN lesions on T2WI are highly specific (99.0%) for diagnosing TZAN.

STN hyperintensity is most prevalent in elderly patients during the subacute phase.

STN lesions are associated with poorer short-term, but not long-term, treatment efficacy.

The chronic phase of TZAN is an independent predictor of poor prognosis.

Supplemental Material

sj-docx-1-cep-10.1177_03331024261435927 - Supplemental material for Spinal trigeminal nucleus lesions in trigeminal zoster-associated neuralgia: A retrospective cohort study on diagnostic value and prognostic factors

Supplemental material, sj-docx-1-cep-10.1177_03331024261435927 for Spinal trigeminal nucleus lesions in trigeminal zoster-associated neuralgia: A retrospective cohort study on diagnostic value and prognostic factors by Lina Wang, Yang Chen, Mingrui Li, Yin Dong, Jun-nan Wang and Tao Sun in Cephalalgia

Footnotes

ORCID iDs

Lina Wang

Yang Chen

Tao Sun

Ethical considerations

This retrospective cohort study was approved by the Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong First Medical University (Approval No.: SWYX: 2025-076) and conducted in accordance with the principles of the Declaration of Helsinki. Due to the retrospective nature of the study, written informed consent from patients was waived by the Ethics Committee.

Consent to participate

Verbal informed consent was obtained and documented during telephone follow-up for the 6-month outcome assessment, and all patients agreed to the use of their anonymized clinical and imaging data for research purposes.

Author contributions

LW ,YC and ML: Responsible for data collection, statistical analysis, and drafting of the manuscript; YD and JnW: Participated in patient screening, clinical data verification, and follow-up management; TS: Conceptualization and design of the study, supervision of data analysis, revision of the manuscript for important intellectual content, study supervision, ethical approval application, and final approval of the submitted version.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the Shandong Provincial Natural Science Foundation (ZR2021QH060) and the Shandong Provincial Medical and Health Science and Technology Development Plan Project (202109010812).

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Data availability statement

All data are stored in the medical record system of Shandong Provincial Hospital Affiliated to Shandong First Medical University and will be shared in compliance with ethical and legal regulations, ensuring the protection of patient privacy.

Supplemental material

Supplemental material for this article is available online.

References

Gerada

Campbell

Kennedy

, et al. Manipulation of the innate immune response by varicella zoster virus. Front Immunol 2020; 11: 1.

Niemeyer

Harlander-Locke

Bubak

, et al. Trigeminal postherpetic neuralgia: from pathophysiology to treatment. Curr Pain Headache Rep 2024; 28: 295–306.

Kim

Chung

Xing

, et al. The trigeminal sensory system and orofacial pain. Int J Mol Sci 2024; 25: 11306.

Oliveira

Corrêa

Alves Júnior Jr

. Trigeminal neuralgia due to varicella-zoster virus reactivation. Rev Soc Bras Med Trop 2024; 57: e00919.

Himeno

Takeshima

Kubo

, et al. Trigeminal herpes zoster presenting with high-intensity signals for the spinal trigeminal nucleus and tract on diffusion-weighted image: a case report of stroke mimic. Brain Nerve 2018; 70: 161–164. Japanese..

D’Amico

Russo

Ugga

, et al.

Can pontine trigeminal T2-hyperintensity suggest herpetic etiology of trigeminal neuralgia?

Quant Imaging Med Surg 2016; 6: 490–495.

Headache Classification Committee of the International Headache Society (IHS). The International Classification of Headache Disorders. 3rd ed. Cephalalgia 2018; 38: 1–211.

Yang

Han

Wang

. Transaxonal spread of varicella-zoster virus in trigeminal zoster followed by Ramsay Hunt syndrome. Neurol India 2021; 69: 1899–1900.

Hakami

Khan

Abdulaziz

, et al. Varicella-zoster virus-related neurological complications: from infection to immunomodulatory therapies. Rev Med Virol 2024; 34: e2554.

10.

Cordano

Caverzasi

Spigno

, et al. MRI Findings in Varicella zoster trigeminal neuritis without rash. Headache 2018; 58: 764–765.

11.

Ahmed Ibrahim

Bakir

Osman Sidow

, et al. Lateral medullary syndrome: uncommon form of brainstem stroke. Ann Med Surg (Lond) 2023; 85: 542–544.

12.

Gomez

. A diagnosis that’s hard to swallow: case report of delayed onset lateral medullary syndrome presenting with only dysphagia. Int J Emerg Med 2024; 17: 44.

13.

Wang

Tao

Jiang

, et al. Risk factors for postherpetic neuralgia: a meta-analysis based on demographic, clinical features, and treatment characteristics. Front Immunol 2025; 16: 1667364.

14.

Patil

Goldust

Wollina

. Herpes zoster: a review of clinical manifestations and management. Viruses 2022; 14: 92.

15.

Alshelh

Di Pietro

Mills

, et al. Altered regional brain T2 relaxation times in individuals with chronic orofacial neuropathic pain. NeuroImage Clin 2018; 19: 167–173.

16.

Kamimura

Hossain

Takahashi

, et al. Attenuation of allodynia and microglial reactivity by inhibiting the degradation of 2-arachidonoylglycerol following injury to the trigeminal nerve in mice. Heliyon 2022; 8: e10034.

17.

Asano

Hayashi

Iwata

, et al. Microglia-astrocyte communication via C1q contributes to orofacial neuropathic pain associated with infraorbital nerve injury. Int J Mol Sci 2020; 21: 6834.

18.

Chen

Wei

Zhang

, et al. Bone morphogenetic protein 4 derived from the cerebrospinal fluid in patients with postherpetic neuralgia induces allodynia via the crosstalk between microglia and astrocyte. Brain Behav Immun 2024; 119: 836–850.

19.

Zhou

, et al. Intravenous lidocaine alleviates postherpetic neuralgia in rats via regulation of neuroinflammation of microglia and astrocytes. iScience 2021; 24: 102108.

20.

Zhang

Fan

, et al. Analysis of the risk factors and a prediction model for postherpetic trigeminal neuralgia recurrence after extracranial nonsemilunar ganglion radiofrequency thermocoagulation. Pain Physician 2023; 26: E719–E724.

21.

Song

Guan

, et al. Efficacy of Gasserian ganglion high-voltage, long-duration pulsed radiofrequency combined with block on acute/subacute zoster-related trigeminal neuralgia. Pain Res Manag 2024; 2024: 1992483.

22.

Yun

Wang

Junnan

Chen

Yang

, et al.. Comparative long-term efficacy of short-term spinal cord stimulation versus bipolar pulsed radiofrequency for refractory postherpetic neuralgia: a 24 month prospective study. European Journal of Medical Research 2025; 30: 272.