Dual-neuromodulation strategy in pain management of herpes zoster ophthalmicus: retrospective cohort study and literature review

Abstract

Background

Effective pain control of herpes zoster ophthalmicus (HZO) is not only essential to attenuate the clinical symptoms but to reduce the risk of postherpetic neuralgia development. Recently, neuromodulation therapy has been one promising option for neuropathic pain and increasingly applied in management of zoster-related pain. One key factor of neuromodulation treatment is the therapeutic site for the impaired nerves. In this study we aim to investigate one novel dual-neuromodulation strategy, targeting the level of the peripheral branch and trigeminal ganglion, in the pain management of HZO.

Methods

Dual neuromodulation strategy combining short-term peripheral nerve stimulation (PNS) with pulsed radiofrequency (PRF) of trigeminal ganglion was compared with single PNS treatment for HZO-related pain. Clinical recordings of patients were retrospectively reviewed. The primary outcome was the pain severity, assessed by the visual analogue scale (VAS) before and after neuromodulation therapy.

Results

PNS achieved significant relief of pain with or without PRF treatment before discharge, which provided enduring therapeutic effect up to 12-month follow-up. The mean reduction of VAS was 6.7 ± 1.4 in dual modulation therapy (n = 13) at last follow-up and 5.4 ± 1.5 in PNS subgroup (n = 20), respectively. Moreover, dual modulation strategy provided better control of pain compared with PNS therapy alone at each time point.

Conclusion

It is feasible and effective to combine the PNS and PRF in pain management of HZO. This novel dual modulation strategy of trigeminal pathway may provide additional therapeutic effects of pain symptoms in HZO population.

KEY MESSAGES

• Dual neuromodulation strategy for pain management of herpes zoster ophthalmicus is proposed, with regard to stimulation site (peripheral and trigeminal ganglion) and apparatus (electrical nerve stimulation and pulsed radiofrequency).

• Superior clinical outcome was associated with novel neuromodulation therapy with dual therapeutic targets, when compared with peripheral nerve stimulation in treatment of herpes zoster ophthalmicus.

• We conducted literature review to compare distinct pattern of neuromodulation (peripheral nerve stimulation and radiofrequency) in treatment of trigeminal neuropathic pain caused by herpes zoster.

Keywords:

• Neuromodulation
• pulsed radiofrequency
• peripheral nerve stimulation
• electrical nerve stimulation
• herpes zoster ophthalmicus
• pain
• postherpetic neuralgia
• treatment

Introduction

Herpes zoster ophthalmicus (HZO) is caused by the reactivation of latent varicella-zoster virus, typically resulting in painful vesicular rash in the distribution of the ophthalmic division of trigeminal nerve [1]. The estimation of HZO prevalence has accounted for about 10%–20% of the total herpetic infection [2]. Risk factor for herpetic infection includes ageing, immunocompromised status and comorbidities [3]. In addition, recent evidence has indicated a potential causal relationship between HZO and COVID-19 infection or vaccination [4–7]. It has been estimated that total incidence of cutaneous lesion secondary to COVID‐19 ranged between 4% and 20.4%, less risk has been reported in COVID-19 cases caused by vesicular lesions [7, 8].

One featured manifestation of HZO is the severe pain affected the ophthalmic dermatome, characterized by stabbing radiating and localized pain during acute phase of vesicular eruption [3]. Moreover, about half of the patients may develop postherpetic neuralgia (PHN) with ophthalmic involvement [2]. PHN is characterized by moderate to severe facial pain which can last for more than 3 months after initial skin lesion [9]. The neuropathic feature of PHN is characterized with continuous or spontaneous burning pain, paroxysmal electric shock-like pain, mechanical allodynia and hyperalgesia [9]. Consequently, the quality of life is significantly impaired in HZO population, and the ongoing requirement of pain management heavens the burden of public health system [10].

The principle of HZO treatment is to apply the anti-viral therapy at early stage of infection; however, it remains controversial that oral acyclovir can reduce the risk of PHN [11]. In addition to anti-viral agents, it is necessary to perform pain therapy in most cases with analgesic medication, including tricyclic antidepressants, antiseizure drugs, opioids and topical analgesics [12]. Interventional procedures may be considered when pain relief is insufficient with oral medicine.

Local nerve block is commonly applied to achieve immediate but short-term pain relief. Recently, we demonstrated that the implantable neuromodulation therapy provided effective and enduring analgesic effect in patients with HZO [13, 14]. By placing an electrical nerve stimulator towards the supraorbital and supratrochlear nerves, continuous stimulation can be delivered to cover the painful regions. However, we found that about 22.2% of HZO patients suffered facial pain in the distribution of the second branch of trigeminal nerve (i.e. the maxillary nerve), which cannot be modulated by the peripheral ophthalmic stimulation [13]. In some severe cases, patients may suffer pain involved with all three branches of trigeminal nerve. Thus, in this study, we aim to investigate the clinical effect of one novel dual modulation strategy, targeting the peripheral nerve branch and Gasserian ganglion simultaneously.

Materials and methods

This was a retrospective medical recording review of consecutive patients with HZO orofacial pain presenting at the Department of Pain, the Third Xiangya Hospital, Central South University between January 2018 and June 2022. Surgical protocol was approved by the institutional and ethics board of the Third Xiangya Hospital of Central South University.

Participants

Patients who were diagnosed with HZO were involved in this study. The diagnosis was confirmed consistent with our previous criteria: vesicular rash and dermatomal pain associated with ophthalmic distribution [13]. Phase of HZO-related pain was categorized according to the disease duration, that acute pain indicated pain occurred within 1 month, and those with moderate to severe trigeminal neuropathic pain over 3 months were considered as PHN population. Subacute phase of HZO-related pain was defined as duration of disease ranging between one and 3 months [13]. Patients were classified into two subgroups according to the method of neuromodulation therapy. Patients who underwent the implantation of peripheral nerve stimulation (PNS) were categorized into Group 1, and those who received dual neuromodulation treatment (PNS combined with trigeminal ganglion pulsed radiofrequency [PRF]) were categorized into Group 2, respectively.

Surgery

Peripheral nerve stimulation

The surgical detail of PNS has been described previously [13]. One supine position was used with the head of patient turning slightly to the contralateral side of the surgical region. After local anaesthesia with 1% lidocaine, a 14-G Tuohy needle was introduced into the 2 cm posterolateral junction of the frontal and zygomatic portion of the orbital rim. The cannulation was then penetrated through the subcutaneous tissue and reached the plane of supra-periosteal tissue over the eyebrow in a semilunar path. The tip of cannula was slightly beyond the cranial middle line (Figure 1a). One eight-contact electrode (Model 3873; Medtronic, Minneapolis, MN, USA) was inserted into the Tuohy needle. The distal ending of electrical lead was connected to one extension multi-lead cable (Model: 355531; Medtronic), to adjust the stimulation parameter with one external neurostimulator (Model 37022; Medtronic). The stimulation electrode was programmed with a pulse width of 500 µS at a frequency of 40 Hz, to induce a sensation of paresthaesia covering the painful region. The amplitude of stimulation voltage was set between 0.5 and 3.0 mV according to the pain severity reported by the patient.

Figure 1. Surgical schematic of PNS implantation and PRF placement. (a) Anteroposterior view of fluoroscopic imaging. Red cycle indicated the foramen ovale; blue line indicated the mandibula; (b) depth of PRF cannulation was guided under the lateral view of the fluoroscopy.

Pulsed radiofrequency of Gasserian ganglion

To perform the percutaneous PRF of the trigeminal ganglion, patient was placed in one supine position with the shoulder slightly elevating for better visualization of the oval foramen. To exposure the anterior–posterior view of the oval foramen, the probe of C-arm was set about 15 degrees ipsilaterally and 30 degrees caudally to the surgical site. The entry of puncturing needle was set about 2.5 cm lateral to the corner of the mouth. Local anaesthesia of 1% lidocaine was applied during the puncturing procedure of the oval foramen. The insertion of needle was performed with the submental, oblique view until we reached the bony edge of the foramen (Figure 1b). The depth of cannulation was then confirmed under the lateral view of fluoroscopy. To target the trigeminal ganglion, we aimed to place the tip of needle 2–3 mm within the clivus line (Figure 1b). The parameter of PRF stimulation was consistent with our previous protocol [15], at a frequency of 2 Hz (20 ms pulse width) and temperature of 42 degrees. The duration of stimulation was set to 240 s and repeated for three cycles.

Clinical outcomes and follow-up

The demographic and clinical data were collected with one brief questionnaire according to the medical history by two independent researchers (Y.W. and L.Y.). To investigate the clinical outcome of neuromodulation therapy, one telephone interview was performed at 1-, 3-, 6- and 12-month follow-up. The primary clinical outcome was the pain severity assessed by the visual analogue scale (VAS), ranging from 0 (‘pain free’) to 10 (‘worst pain imaginable’).

Statistical analysis

Statistical analysis was performed by Prism 9.0 software (GraphPad, San Diego, CA, USA). Variables were presented as mean ± standard deviation. Two-way analysis of variance with repeated measures and post-hoc multiple pairwise comparison Sidak’s testing was used to assess the alteration of pain scores between two groups over time. A p-value <.05 was considered statistically significant.

Results

Demographics

A total of 33 patients (18 males and 15 females) were included in this study, with a mean age of 70.2 ± 9.0 years old. Twenty patients (11 male and 9 female) underwent PNS therapy, and 13 (7 male and 6 female) were treated with dual neuromodulation, respectively. The mean age of PNS group and PNS + PRF group were 70.25 ± 9.20 (range, 50–86) and 70.15 ± 9.00 years old (range, 50–80) separately, which had no statistically significant difference (p = .977). Most patients (76%) reported severe facial pain with VAS over 6/10 points in the ophthalmic region at admission (Group 1 vs. Group 2: 7.5 ± 0.9 vs. 7.2 ± 1.4, p = .578). The majority of participants presented with one subacute herpetic lesion, with disease duration ranging between 1 and 3 months. Only five patients were diagnosed with PHN in this study. Before neuromodulation therapy, all patients had failed to control pain with conventional treatment, including oral analgesic agents and nerve block. The clinical features of patients are shown in Table 1.

Table 1. General characteristics of patients.

Group	PNS	PNS + PRF	p Value
N	20	13
Age (years)	70.25 ± 9.20	70.15 ± 9.00	.977
Sex (male/female)	11/9	7/6	>.999
Area (n, %)
V1 branch	15 (75%)	10 (77%)	>.999
V1–2 branch	5 (25%)	3 (23%)
Duration (n, %)
<1 month	2 (10%)	4 (31%)	.301
1–3 months	15 (75%)	7 (54%)
>3 months	3 (15%)	2 (15%)
Follow-up (n, %)
3-month	0	1 (8%)
6-month	7 (35%)	6 (46%)
12-month	13 (65%)	6 (46%)
Hospitalization duration (days)	16.15 ± 5.49	15.69 ± 3.65	.767
Implantation duration (days)	12.35 ± 1.96	11.91 ± 1.93	.564
VAS
Preoperative	7.5 ± 0.9	7.2 ± 1.4	.579
Last follow-up	2.05 ± 1.12	0.54 ± 0.75	.001
Anti-virus treatment
None	3 (15%)	3 (23%)	.658
Yes	17 (85%)	10 (77%)
Preoperative pain management (n, %)
Lidocaine path	3 (15%)	4 (31%)	.423
NSAIDs	9 (45%)	3 (23%)
Gabapentin/pregabalin	20 (100%)	13 (100%)
Opioids	13 (65%)	8 (62%)
Nerve block	12 (60%)	3 (23%)

Data are presented with mean ± standard deviation. PNS: peripheral nerve stimulation; PRF: pulsed radiofrequency; NSAIDs: non-steroidal anti-inflammatory drugs; VAS: visual analogue scale.

Follow-up

All participants consent to undertake the routine telephone follow-up and accomplished the 1- and 3-month follow-up after their hospital discharge. Fourteen out of 20 patients were accomplished for the final follow-up in Group 1. One patient lost the 6-month follow-up in Group 2, and 8 out of 13 patients finished the 12-month follow-up in Group 2, respectively.

Therapeutic efficacy

Both groups obtained significant pain relief at discharge, with pain reduction of 4.2 ± 1.0 VAS after PNS treatment and 5.0 ± 1.4 for dual neuromodulation strategy, respectively. The VAS score between Group 1 and Group 2 was 3.1 ± 1.2 and 1.4 ± 1.4 at 1 month, 2.7 ± 1.2 and 0.8 ± 1.2 at 3 months and 2.4 ± 1.1 and 0.4 ± 0.7 at 6 months. The therapeutic effect was enduring up to 12-month follow-up, as the pain scores gradually decreased to 1.6 ± 1.3 in Group 1 and 0.5 ± 0.8 in Group 2 (Figure 2). Meanwhile, the pain severity was significantly attenuated in Group 2 at each time point of follow-up compared with the other cohort (Figure 2).

Figure 2. Pain severity assessed by the visual analogue score at each follow-up point. Patients were interviewed at baseline, discharge, 1-, 3-, 6- and 12 months after neuromodulation treatment. Two-way analysis of variance with repeated measures and Bonferroni’s post-test. ns = non-significant, *p < .05, **p < .01, ****p < .0001.

Complications

We did not observe any obvious complications (e.g. infection, haematoma, tetrode migration or nerve injury) in this study.

Literature and discussion

Significant percentages of HZO patients may suffer severe and long-lasting facial pain. Except shortening disease course and preventing ophthalmic complications, sufficient analgesia is necessary to improve the quality of life and reduce the risk of PHN. Here we demonstrated that one novel neuromodulation therapy targeting both peripheral nerve branch and Gasserian ganglion is an effective option for pain management in patients with HZO disease.

In the literature review, a series of studies focused on PRF and/or PNS neuromodulation therapy for pain management of HZO were identified (Table 2) [13, 16–35]. Despite HZO, other indication of neuromodulation usage for facial pain management includes idiopathic, postsurgical, traumatic, post-stroke, multiple sclerosis, radioactive, classic trigeminal neuralgia and other unclassified types of aetiology [16, 20–23]. Multiple phases of herpetic disease (acute, subacute and PHN) can be considered to undertake neuromodulation treatment, ranging from 2 weeks to 30 years. The majority of previous data were obtained from observational cohort studies or case series, and about 23% (5 out of 22) were randomized controlled trials [29, 30, 32, 35]. To treat HZO-related neuralgia, supraorbital and infraorbital branches are commonly targeted for stimulator implantation. Unlike classical trigeminal neuralgia, less herpetic cases present with mandibular lesions, and the implantable neuromodulation can also be applied directly to this region [18, 19, 22] or in the trigeminal ganglion [23].

Table 2. Summary of enrolled literatures..

Author (year)	Country	No.	Age	Sex (F/M)	Pian duration	Intervention (target)	Aetiology	Clinical outcome	Follow-up
Manning et al. (2019)	UK	7	32–66	6/1	4–14 years	PNS (supraorbital or infraorbital branches)	Idiopathic TNP, PHN, postsurgical TNP	NRS, analgesic consumptions, side effects	3, 6 and 12 months
Wan et al. (2021)	China	68	65–93	37/31	15–90 days	PNS (supraorbital branch)	HZO	VAS, SF-36, analgesic consumptions	3, 7, 14, 30, 90 and 180 days
Han et al. (2020)	China	18	50–86	7/11	1–3 months	PNS (supraorbital branch)	HZO	VAS, response rate, analgesic consumptions, PHN incidence	1, 3, 6, 12 months
Liu et al. (2021)	China	26	69	14/12	1–4 months	PNS (supraorbital, maxillary branch or ramus of mandible skin field)	Herpes zoster-related TN	VAS, dosages of pain medication, PHN rate	1, 3, 6 months
Jakobs et al. (2016)	Germany	8	28–81	7/1	NR	PNS (infraorbital, supraorbital or mandibular nerve)	PHN, classical TN	VAS, attack frequencies, oral drug intake, complications and side effects	6–29 months
Johnson et al. (2004)	USA	10	33–86	3/7	7–144 months	PNS (supraorbital or infraorbital branches)	PHN, posttraumatic neuropathic pain	Complication rate, preoperative symptom duration, degree of pain relief, preoperative and postoperative work status, postoperative changes in medication usage and overall degree of therapy satisfaction	26.6 ± 4.7 months
Texakalidis et al. (2020)	USA	15	43–71	11/4	0.75–17 years	PNS (supraorbital and infraorbital nerve)	PHN, traumatic, post-stroke, idiopathic facial pain	VAS	5.8 months
Klein et al. (2015)	Germany	8	57–87	4/4	2–17 years	PNS (infraorbital, supraorbital, occipital or mandibular nerve)	Classical TN, PHN, idiopathic facial pain, multiple sclerosis, radiative injury	VAS, analgesic usage	11.3 months (5–28)
Taub et al. (1997)	Canada	34	NR	20/14	0.5–18 years	PNS (Gasserian ganglion)	Peripheral, central, PHN and unclassifiable	Success rate	6-month interval
Jia et al. (2022)	China	85	23–82	44/41	27–66 days	PRF (Gasserian ganglion, peripheral branch)	Acute zoster-related TN	NRS, response rate	1, 3 months
Ding et al. (2019)	China	90	52–91	48/42	NR	PRF (Gasserian ganglion, peripheral branch)	PHN	VAS, SF-36, efficacy rate, analgesic usage	1, 3, 6 and 12 months
Zhang et al. (2020)	China	36	73	17/19	NR	PRF (supraorbital nerve)	HZO	NRS, spontaneous pain, allodynia, gabapentin dosage, opioids usage, PGIC, Barrow numbness score, recurrence rate	30, 90, 180, 360 days
Fan et al. (2022)	China	106	67/69/67	34/72	NR	PNS (supraorbital and supratrochlear nerve), PRF (Gasserian ganglion)	HZO	NRS, PSQI, concomitant pain medication usage, relapse rate	1, 2, 4 weeks; 3, 6, 12 and 24 months
Liu et al. (2021)	China	32	56–86	12/20	0.5–3 years	PRF (ophthalmic branch of trigeminal nerve)	PHN	VAS, sleep quality, treatment satisfaction	1, 4, 12 weeks
Wan et al. (2019)	China	93	65	52/41	2 months	PRF (Gasserian ganglion)	Acute/subacute zoster-related TN	VAS, SF-36, pregabalin dosage	3, 7, 14 days; and 1, 3, 6 months
Wan et al. (2021)	China	115	70	71/44	2 months	PRF (Gasserian ganglion)	Acute/subacute zoster-related TN	VAS, SF-36, pregabalin dosage	3, 7, 14 days; and 1, 3, 6 months
Li et al. (2021)	China	52	64/66	30/22	2 months	PRF (supraorbital nerve)	PHN	VAS, SF-36, treatment efficiency	1, 4, 12 and 24 months
Li et al. (2022)	China	106	63/62	54/52	NR	PRF (Gasserian ganglion)	PHN	VAS, QoL, rescue drug dosage	3, 7, 14 days; and 1, 3, 6 months
Ding et al.(2019)	China	84	34–72	39/45	3–20 months	PRF (stellate ganglion)	PHN	VAS, QoL, PCS, MCS, efficacy rate	1, 2, 4, 12 and 24 weeks
Zhang et al. (2022)	China	91	67/66	47/44	3–360 months	PRF (DRGs)	PHN	VAS, tactile sensation, efficacy rate	3, 6, 12 months
Zhang et al. (2022)	China	64	63/66	31/33	50 days	PRF (corresponding spinal nerve segment and trigeminal nerve branches)	Herpes zoster neuralgia	VAS, PSQI, SAS, SDS, infrared thermograms, body area rating scale, morphine consumption, PHN incidence, serum inflammatory index	1, 3, 5 days; and 1, 2 months

DRG: dorsal root ganglion; HZO: herpes zoster ophthalmicus; MCS: mental component summary; NR: none reported; NRS: numerical rating scale; PCS: physical component summary; PHN: postherpetic neuralgia; PNS: peripheral nerve stimulation; PRF: pulsed radiofrequency; PSQI: Pittsburgh Sleep Quality Index; QoL: quality of life; SAS: Self-Rating Anxiety Scale; SDS: Self-Rating Depression Scale; SF-36: Short Form-36 Questionnaire; TN: trigeminal neuralgia; TNP: trigeminal neuropathic pain; VAS: visual analogue scale.

It is important to apply anti-viral treatment at early stage of herpes zoster to attenuate pain and prevent the development of PHN. However, it remains a great challenge to efficiently control the HZO-related pain with conventional anti-viral agents, which is consistent with our previous report [13]. They may need to take anticonvulsants and opioid analgesics for pain relief, which commonly brings in side effects (e.g. gastrointestinal and psychiatric dysfunction, addiction and tolerance) [36]. Thus, alternative option of pain management in HZO cohort is necessarily needed.

Neuromodulation therapy has been one promising method in treatment of intractable neuropathic pain. One strategy of neuromodulation is to apply PRF approaching the dorsal root ganglion or trigeminal ganglion, which has been applied in treatment of herpetic neuralgia [29, 37]. Recently, an emerging neuromodulation of implantable PNS has been introduced to treat trigeminal herpetic pain [14]. Unlike radiofrequency treatment, the implantable device can provide persistent electrical stimulation to the affected nerves. Thus, one advantage of this novel dual neuromodulation therapy is to provide enduring therapeutic effect compared with short-term PRF stimulation. In addition to stimulation duration, the mechanism underlying the analgesic effect may also vary between PNS and PRF due to the site of stimulation. Thus, we think it necessary to investigate the clinical effect of the combination of PNS and PRF in HZO pain management.

Consistent with our previous study [13], PNS achieved significant pain relief of HZO-related pain with average 4.2 VAS reduction at discharge. Moreover, we found that dual neuromodulation treatment achieved better clinical outcome compared with PNS, as demonstrated by the significantly lower pain severity at short- and long-term follow-up (Figure 2). The estimated recurrence rate of herpetic zoster-related pain was about 37% in one cohort who underwent PRF treatment only [38]. Given the high risk of recurrent pain, we did not recommend the patient to only take PRF procedure according to the clinical routine in our centre, which is consistent with recent findings that combination of PNS and PRF or PNS alone provided superior analgesic function over single PRF treatment [27]. Specifically, pain severity and functional deficiency were both significantly lower in patients treated with PNS combined with or without PRF from 1 to 24 months after treatment, compared with PRF treatment [27].

The well-known gate control theory of pain may play an important role in the analgesic effect of neuromodulation [39]. This theory proposed that the selective stimulation of non-nociceptive afferents (Aβ fiber) can inhibit the nociceptive afferents (Aθ and C fibers) and then decrease the perception of pain [40, 41]. Mechanism underlying the analgesic effect of PRF includes inhibition of inflammatory responses and interfering ectopic firing at ganglion [42, 43]. Recently, we have demonstrated that supraspinal mechanism may partially mediate the therapeutic function of PNS [44].

There are several limitations in our study. First, this is a single-centre study with relatively small sample number. Second, the retrospective and nonrandomized nature of study design. In addition, we did not enrol one group with only PRF treatment in this study, mainly due to the insufficient therapeutic effect. Finally, we did not compare the effect of different stimulation parameters, including stimulation frequency, voltage and duration on the therapeutic effect. We aim to perform randomized, well-controlled, clinical trials with large sample size in future studies to investigate the optimal strategy of neuromodulation treatment in pain management of HZO.

Conclusions

It is feasible and effective to apply dual neuromodulation techniques at trigeminal ganglion and peripheral nerve branch. Better clinical outcomes may be achieved by this novel neuromodulation therapy in pain control for HZO patients, compared with individual PNS treatment.

Authors contributions

Conceptualization: H.Z. and D.H. Methodology: J.M., Y.W., L.Y., D.H. and H.Z. Validation: J.M. and Y.W. Formal analysis: J.M. Investigation: J.M. and H.Z. Resources: H.Z. and D.H. Data curation: J.M. Writing – original draft preparation: J.M. and H.Z. Writing – review and editing: H.Z. Visualization: H.Z. Supervision: H.Z. Project administration: H.Z. Funding acquisition: H.Z. All authors have read and agreed to the published version of the manuscript.

Disclosure statement

The authors declare no competing interests.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article.

Additional information

Funding

This research was funded by the National Natural Science Foundation of China [Grant No. 82271259 to H.Z.], Excellent Youth Foundation of Hunan Scientific Committee, Key Laboratory of Hunan Province [Grant No. 2018TP1009], and Huizhiyucai Project of the Third Xiangya Hospital, Central South University.