Intravenous Lidocaine Response as a Predictor for Oral Oxcarbazepine Efficacy in Neuropathic Pain Syndrome: A Prospective Cohort Study

Introduction

Neuropathic pain syndrome (NPS) is a group of conditions that have lesions or diseases of the somatosensory nervous system, primarily peripheral or central in origin. The treatment of NPS requires multiple drug combinations and multimodal treatment, yet only about 40% to 60% of patients achieve satisfactory pain relief [1–3]. Neuropathic pain has an additional negative impact on patient quality of life, regardless of the etiology of the neuropathy [4]. A consensus treatment protocol for NPS remains elusive [5–7]. Sodium-channel blockers are the second-line pharmacological treatment after tricyclic antidepressants, calcium channel α2δ1 ligands, and serotonin and norepinephrine reuptake inhibitors. The therapeutic benefits of sodium-channel blockers, such as lidocaine, can be achieved through topical, oral, and intravenous (i.v.) administration [8–10]. Lidocaine is an i.v. broad-spectrum sodium-channel blocker that is a promising treatment for pain following propofol injection [11], as well as a potential adjuvant for postoperative pain, particularly in abdominal surgery [12]. Carbamazepine, an oral sodium-channel blocker, has potentially serious adverse events (AEs) including Stevens-Johnson syndrome, with an increased incidence in Asian populations [13]. Compared with carbamazepine, oxcarbazepine (OXC) has similar efficacy, but with fewer AEs and drug interactions and better tolerance [14,15].

When administered intravenously, lidocaine has been found to decrease pain scores within 15 min and for up to 3 days after an infusion [10]. If i.v. lidocaine improves the patient’s response to oral sodium-channel blockers, it could become a valuable tool for the treatment of NPS [16]. Two studies have reported a moderate correlation between the response to i.v. lidocaine administration and the efficacy of oral mexiletine in neuropathic pain treatment [17,18]. A study by Sanchez et al also reported a good association between i.v. lidocaine and oral carbamazepine for the treatment of tinnitus [19].

A mechanism-based personalized treatment approach was studied by Demant et al, who tested the hypothesis that the response to sodium-channel blockers can be predicted by sensory profiles [20]. The study showed that OXC is more efficacious in patients with the irritable than the non-irritable nociceptor phenotype. However, the sensory profiles require quantitative sensory testing equipment, skillful personnel and, cognitively, good cooperation of the patient to perform the investigation. Furthermore an interim analysis by Schipper et al reported a low predictive value of lidocaine infusion for OXC efficacy [21]. Since this personalized medicine approach was tested in a non-Asian setting, it is important to explore the hypothesis also in an Asian population. Therefore, our primary aim was to determine whether the response to a single i.v. lidocaine infusion can predict the efficacy of oral OXC in terms of predictive value for treating chronic NPS. The secondary objective was to evaluate the efficacy of oral OXC in terms of pain intensity reduction and improvements in psychological states and quality of life via the questionnaires. Additionally, we aimed to identify AEs associated with both treatments, including abnormal or unpleasant symptoms and deviations in laboratory investigation values.

Material and Methods

PATIENTS AND STUDY DESIGN:

Following approval from the Siriraj Institutional Review Board (Si 134/2014), we undertook a prospective cohort study. Participants were recruited from the Pain Clinic of the Department of Anesthesiology at Siriraj Hospital, Mahidol University, Bangkok, Thailand. In this study, patient anonymity and confidentiality were strictly maintained. The questionnaire was designed to exclude any personal identifiers. Participants were informed about the confidentiality measures through an informed consent process. Data were collected using paper forms, ensuring no identifying information was recorded. Responses were stored in an encrypted database accessible only to the research team. The results were analyzed and reported in aggregate form to prevent the identification of individual participants.

Eligibility criteria included adults aged 18 to 65 years with a diagnosis of NPS who were classified as having probable or definite neuropathic pain according to the grading system for neuropathic pain [22], confirmed by a certified pain physician, and with an average numeric rating scale (NRS) score of at least 4 out of 10. Exclusion criteria were individuals with a history of lidocaine or OXC allergy; poorly controlled cardiovascular, hepatic, or renal disease; 12-lead electrocardiography conduction abnormalities; severe psychiatric disorders; substance abuse; cancer-related neuropathic pain; pregnancy and breastfeeding; and those unable to commit to 6 weeks of follow-up. Patients experiencing intolerable or serious AEs from lidocaine or OXC, non-compliance with the drug protocol, or failure to complete follow-up were excluded from the analysis.

MEASURED OUTCOMES:

During the initial visit, demographic data, including biological sex, age, body mass index (BMI), underlying health conditions, NPS diagnosis, pain location, daily medication usage, baseline pain intensity using NRS and the Thai version of the Douleur Neuropathique 4 (DN4) questionnaire, were collected by the well-trained assistant researcher. The DN4 is a screening tool for possibly diagnosing neuropathic pain [23]. Participants also completed all the comprehensive Thai version of Depression, Anxiety & Stress Scales (DASS-21) and EuroQol-Five Dimensions-Five Levels (EQ-5D-5L) quality of life questionnaires for baseline assessments. The outcome measures are detailed below.

PAIN INTENSITY: Pain intensity was assessed using the NRS with a rating of 0 to 10 points, which has shown the most sensitivity and stability for the individual pain measure [24]. In terms of lidocaine efficacy, the NRS was used to evaluate current pain 15 min before and immediately after a single i.v. lidocaine infusion. The NRS responses range from 0 (indicating “no pain”) to 10 (indicating the “worst pain imaginable”). The percentage of pain reduction after i.v. lidocaine infusion was calculated as [(pre-infusion NRS (baseline NSR)-post-infusion NRS)/pre-infusion NRS]×100. In a similar manner of evaluating the efficacy of OXC treatment, the percentage of pain reduction after 6 weeks of oral OXC administration was calculated as [(baseline NRS-week 6 NRS)/baseline NRS]×100.

PSYCHOLOGICAL FUNCTION: Psychological function was assessed using the Thai version of the DASS-21 [25,26], which evaluates depression, anxiety, and stress through separate 7-item scales. Responses to the items are summed to create domain-specific scores, each with a maximum of 21 points.

QUALITY OF LIFE: The Thai version of the EQ-5D-5L questionnaire was used to assess the overall quality of life [27,28]. The EQ-5D-5L utility score ranges from 0 (representing death) to 1 (representing full health), with respondents able to rate their health as being “worse than death” by providing a score of less than 0. For instance, in Thai populations, scores have been reported to range from −0.42 to 0.94 [28].

ADVERSE EVENTS: Adverse events associated with i.v. lidocaine infusion included dizziness, tinnitus, perioral numbness, tongue paresthesia, and urticarial rash. Cardiac arrhythmia presence or absence was monitored by nurses during the infusion. AEs of oral OXC were also monitored and included daytime somnolence, dizziness, nausea, vomiting, rash, cytopenia, hyponatremia, abnormal renal function, and abnormal liver enzymes [29]. Blood tests, including complete blood count, sodium, creatinine, and liver enzyme levels, were conducted at baseline and 4 and 6 weeks after OXC treatment.

INTERVENTION:

Before starting the i.v. lidocaine infusion, our nurses monitored the patient by applying a blood pressure manometer, 3-lead electrocardiogram, and pulse oximetry. A single administration of i.v. lidocaine 3 mg/kg was given over 1 h [10,30,31]. The average intensity of pain score on the NRS at baseline and immediately after infusion, including any AEs, were recorded. Subsequently, 150 mg of OXC was administered orally at bedtime; the dosage was increased by 150 mg every 3 days until at least a 50% pain reduction or the maximum tolerable dose or the maximum dose of 1800 mg/day was achieved [32]. Tramadol was also provided orally as a rescue drug, with a maximum dose of 400 mg/day. Patients were further interviewed for NRS, tramadol requirements, and AEs at weeks 1, 3, and 5 by phone and at weeks 2, 4, and 6 by visiting the pain clinic. During the 6-week study, other anti-neuropathic pain medications were not permitted. The DASS-21 and EQ-5D-5L questionnaires were assessed at the baseline and repeated after week 6.

SAMPLE SIZE CALCULATION AND STATISTICAL ANALYSIS:

All statistical analyses were performed using the Statistical Package for Social Sciences version 18.0 (IBM Corp, Armonk, NY, USA). Based on the published data of Schipper et al [21], the sample size of 46 patients is required in this study, which was calculated from a 95% confidence interval (CI) of predictive value at 62.5% with an allowable error of 15% and an additional 10% dropout, as using the following formula:

The objective was to test whether a single administration of i.v. lidocaine infusion could predict the analgesic efficacy of oral oxcarbazepine in patients with NPS. The response criteria were defined as follows:

The qualitative data, including biological sex, diagnosis of NPS, and AEs, were described by number and percentage. The quantitative data regarding age, weight, BMI, NRS, and dose of tramadol (mg) per day were described using mean±standard deviation (SD), or median with minimal-maximal value. The Wilcoxon signed-rank test was used to compare the improvement of pain intensity (NRS), DASS-21, and EQ-5D-5L scores after 6 weeks of oral OXC treatment. P<0.05 was considered statistically significant.

Results

RESPONSE TO I.V. LIDOCAINE INFUSION AND RELATIONSHIP TO ORAL OXC EFFICACY:

From 32 analyzed patients, 11 patients (34%) had a positive response to a single i.v. lidocaine infusion, and 21 patients (66%) had a negative response. In nearly identical ratios, after 6 weeks of OXC treatment, 11 patients (34%) achieved more than 50% pain relief and 21 patients (66%) achieved less than or equal to 50% pain relief. However, only 6 patients had response to both i.v. lidocaine and oral OXC. On the other hand, there were 16 patients who had non-response to both i.v. lidocaine and oral OXC. Thus, the efficacy prediction of oral OXC by i.v. lidocaine infusion showed the negative predictive value of 76.2% at the 95% CI (61.6–86.5), while the positive predictive value was 54.5% at the 95% CI (32.0–75.4) (Table 2).

Figure 2 demonstrates the relationship of pain reduction after i.v. lidocaine infusion and after oral OXC treatment. The difference in pain reduction after i.v. lidocaine infusion and oral OXC treatment in the range of ±20% was found in 21 patients (66%). Eight patients (25%) experienced >20% more pain reduction after i.v. lidocaine infusion than after OXC treatment, while only 3 patients (9%) received >20% more pain reduction after OXC treatment than after i.v. lidocaine infusion.

EFFICACY OF OXC:

After 6 weeks of OXC treatment, 20 of 32 patients (63%) achieved pain intensity reduction of at least 30%, and 11 patients (34%) achieved more than 50% pain relief. The average NRS decreased from 6.84±1.72 to 3.81±1.97 after i.v. lidocaine infusion, and to 4.09±2.26 after 6 weeks of oral OXC treatment; P<0.001. Most patients (75%) required the same average tramadol dosage in the first week as they did in week 6, whereas 5 of 32 (16%) patients required a lower tramadol dosage.

The EQ-5D-5L demonstrated significant improvement of median scores (interquartile ranges; IQR) from 0.690 (0.503, 0.845) at baseline to 0.845 (0.575, 0.960) at week 6 (P=0.018). The DASS-21 measures 3 dimensions of negative emotional states, namely depression, anxiety, and stress/tension. The median scores for depression and anxiety were not significantly different between before and after OXC treatment, from 5.00 (3.00, 10.75) to 4.00 (1.00, 8.50); P=0.120, and from 4.00 (1.00, 7.00) to 2.00 (1.00, 4.75); P=0.072, respectively. Nevertheless, the median scores in the stress dimension were found to be statistically significantly decreased from 7.00 (5.00, 10.75) to 5.00 (2.00, 8.75); P<0.001 (Figure 3).

ADVERSE EVENTS OF I.V. LIDOCAINE INFUSION AND ORAL OXC TREATMENT:

During the i.v. lidocaine infusion, 1 patient (3%) reported dizziness, but no patients experienced AEs. Fifteen (47%) patients reported tolerable dizziness, and 19 (59%) patients experienced daytime somnolence from oral OXC. Moreover, 11 (34%) patients had nausea, and 4 (13%) patients had vomiting. Also, 5 (16%) patients developed low serum sodium levels (Table 3). After receiving sodium chloride supplements, their blood sodium level became normal. No other abnormal laboratory investigations were observed.

Discussion

The results of our study demonstrate that patients with NPS who poorly respond to i.v. lidocaine are unlikely to achieve a good analgesic effect of oral OXC. Also, in an interim analysis, Schipper et al reported a low predictive value of lidocaine infusion for OXC efficacy [21]. In our study, we found that a negative treatment response to i.v. lidocaine infusion predicted a negative response of oral OXC, with a negative predictive value of 76.2%, similar to previous studies [17,21]. Thus, i.v. lidocaine infusion can be useful to identify patients who are unlikely to benefit from oral OXC treatment, thereby avoiding the potentially serious AEs of this drug.

One mechanism of neuropathic pain is the voltage-gated sodium-channel (NaV) [33]. Lidocaine relieves neuropathic pain by non-selectively inhibiting the voltage-gated sodium channel at the peripheral nociceptors and central neurons [34]. Therefore, neuropathic pain predominantly caused by NaV dysfunction would be predicted to usually respond well to lidocaine treatment. OXC, a carboxamide-derivative sodium-channel blocker, inhibits inactive-state sodium channels. However, a specific NaV inhibitory mechanism of OXC has not been elucidated [33]. We hypothesize that OXC is not a broad-spectrum sodium-channel blocker, different from lidocaine, which is a broad-spectrum one. Therefore, patients who respond to i.v. lidocaine infusion may not necessarily respond to oral OXC. Conversely, patients that do not respond to i.v. lidocaine would not respond to oral OXC either.

However, some patients partially responded to i.v. lidocaine but responded better to oral OXC. It is possible that i.v. lidocaine 3 mg/kg in our protocol was insufficient for some patients to experience significant pain relief, and a higher dose of i.v. lidocaine would have increased its efficacy. Second, some patients might have received other treatments during oral OXC titration that they did not disclose to the investigators.

About two-thirds of patients demonstrated a good correlation (±20% difference) of pain reduction between i.v. lidocaine infusion and oral OXC treatment. If a patient experienced nearly 100% pain relief after i.v. lidocaine infusion, for example, he would potentially have about 80% to 100% pain relief after oral OXC treatment. Therefore, in patients reporting more than 70% pain reduction after i.v. lidocaine infusion, it is reasonable to try prescribing and titrating oral OXC, with the expectation of more than 50% pain reduction.

For patients with NPS, other than trigeminal neuralgia and painful tonic spasm in neuromyelitis optica spectrum disorder (carbamazepine and OXC are the first-line treatment) [35,36], we suggest that prognostic 3 mg/kg of i.v. lidocaine infusion should be administered before considering oral OXC treatment. If patients report less than a 50% pain reduction after i.v. lidocaine infusion, oral OXC should not be considered. If patients report more than 70% pain reduction after i.v. lidocaine infusion, oral OXC should be considered and titrated. However, if patients report 50% to 70% pain reduction from i.v. lidocaine infusion, oral OXC treatment should be at the physician’s and patient’s discretion.

Consistent with a study by Magenta et al [9], about one-third of our OXC-tolerant patients achieved more than 50% pain relief. Nevertheless, a systematic review demonstrated that there was little evidence to support the effectiveness of OXC in painful diabetic neuropathy, radiculopathy, and neuropathies [29]. Most patients in our study had peripheral nerve injury, and only 1 patient had trigeminal neuralgia. This would be evidence of OXC’s efficacy for NPS, apart from trigeminal neuralgia. However, although the OXC treatment was able to decrease pain intensity and improve stress and quality of life as well, it had several tolerable and intolerable AEs [9]. Regardless, it is clear that the benefits of pain management should be weighed against potential AEs.

Although OXC is the keto analog of carbamazepine, it has different biotransformation, which confers a lower risk of allergic cutaneous reactions and better safety and tolerability [9]. The participants were not tested for the presence of the HLA-B*15: 02 allele before starting OXC. We found that 5 of 46 patients (10.9%) developed an allergic rash, which was higher than what we found in our previous retrospective study (2.4%) [14]. Therefore, cutaneous AEs from OXC should be discussed with patients before prescribing. In this study, the dropout rate was found to be quite high due to the AEs of OXC, including dizziness, headache, constipation, and hepatitis. Regardless, the AEs can be resolved simply by discontinuing the drug and providing supportive treatment and close follow-up.

Our study has some limitations. First, we used a relatively low dose of lidocaine (3 mg/kg). An increase in the dose of lidocaine, or multiple administrations, might have increased its positive predictive accuracy. Second, our study was limited by the small number of patients who had a positive response to treatment with lidocaine or OXC, partially because the dropout rate was 30%, which was higher than what we anticipated. It might be due to a quick increasing dose of OXC. Moreover, better management of AEs may have reduced the dropout rate. Lastly, we used the pain intensity score as the cut-off criterion for determining the response of lidocaine and OXC. This study showed moderate pain relief (less than 50% pain reduction) and a significant improvement in stress and quality of life. We note that chronic pain, particularly neuropathic pain, is difficult to measure due to its subjective and multidimensional nature. Thus, it requires a more comprehensive measurement than a univariable or single domain that is a surrogate for the reduction in pain [37].

Conclusions

This study tested a precision (or personalized) medicine-based hypothesis that the response to i.v. sodium-channel blocker (lidocaine) would predict the subsequent response to an oral sodium-channel blocker (OXC) in patients with neuropathic pain. The results indicated that negative responses to i.v. lidocaine do predict a lack of OXC response. A caveat to interpreting this result is that the AEs of OXC may have obscured an analgesic effect.