Ethics statement

The study was approved by the Institutional Review Board of Tokushima University Hospital (approval number 3642) and conducted in accordance with the principles expressed in the Declaration of Helsinki. Because of the retrospective nature of this study, the requirement for written informed consent was waived by the IRB. Instead, informed consent was obtained via the opt-out method, where information regarding the study was disclosed on the hospital’s official website to provide patients with the opportunity to decline participation. All data were pseudonymized before analysis to ensure patient confidentiality. For research purposes, the medical records were accessed between 1 Dec 2025 and 15 Feb 2026. During and after data collection, the authors had no access to information that could identify individual participants.

Study population and variables

We retrospectively identified all patients who underwent TF-FESS under LA at our institution between January 2024 and March 2025. Patients were excluded if they underwent surgeries involving two or more intervertebral levels or if the surgery was indicated for spinal infection.

Baseline demographic and surgical data for these patients were collected from the electronic medical records. The patient data collected included age, sex, actual body weight, relevant comorbidities (diabetes mellitus, hypertension, chronic obstructive pulmonary disease, and congestive heart failure), and functional status (evaluated as a categorical variable). Patient frailty was assessed using the mFI-5 and a corresponding binary frailty indicator [11,12].

Operative variables included the surgical diagnosis, operative level, operative time, and total dose of 1% lidocaine. To ensure the robustness of our safety analysis, specific complications were defined as follows:

Local anesthetic systemic toxicity

Clinically significant LAST was assessed based on the occurrence of classic central nervous system symptoms (e.g., metallic taste, circumoral numbness, tinnitus, agitation, or seizures) or cardiovascular instability (e.g., bradycardia, arrhythmias, or hypotension) during or immediately after the procedure.

Postoperative delirium

Postoperative delirium was defined as any documented episode of acute confusion, disorientation, or altered consciousness recorded in the electronic medical records by nursing or medical staff during the hospital stay. We note that standardized prospective tools, such as the Confusion Assessment Method (CAM), were not utilized in this retrospective review.

Nerve root injury

Nerve root injury was defined strictly as new-onset postoperative motor weakness or sensory deficits that persisted beyond the immediate recovery period. Transient intraoperative leg pain or electric shock-like sensations during the procedure were considered essential components of real-time neurological feedback and were not classified as complications. Intraoperative tolerability measures included the need for intravenous rescue analgesics/sedatives and interruption/abandonment of the procedure as a result of pain.

Anesthetic and sedation protocol

The LA procedure was performed by the primary surgeon before the surgical incision. Following administration of 1% lidocaine into the subcutaneous tissue and fascia, the anesthetic agent was sequentially injected into the facet joint capsule, the inferior vertebral endplate, and the annulus fibrosus of the intervertebral disc. Injections were performed slowly to minimize the risk of rapid systemic absorption.

Vital signs, including peripheral oxygen saturation, electrocardiography, and non-invasive blood pressure, were monitored continuously throughout the procedure. Surgical nursing staff performed real-time clinical assessments for signs of LAST, and intravenous lipid emulsion was readily available in the operating room as a standard safety precaution.

A standardized conscious sedation protocol was used to ensure patient comfort while maintaining the capacity for neurological feedback. This consisted of intravenous administration of pentazocine (15 mg) and hydroxyzine pamoate (25 mg), which was divided into two doses and administered at 15-min intervals. This regimen allowed for an optimal balance between analgesia and preservation of real-time patient communication during the procedure.

Definition of frailty groups

For subgroup analyses, patients were categorized based on their mFI-5 scores. The cohort was divided into a non-frail group (mFI-5 < 2) and a frail group (mFI-5 ≥ 2) for evaluation of the impact of physiological reserve on surgical and anesthetic outcomes.

Safety threshold for local anesthesia

The individual safety limit for the non-frail group was defined as 7 mg/kg of actual body weight [3]. In contrast, the safety threshold for the frail group was set at a more conservative 5 mg/kg to account for their narrowed therapeutic window [3]. This lower threshold was established because frail patients often show increased drug absorption rates, reduced clearance, sarcopenia-related loss of drug reservoirs, and heightened pharmacological sensitivity [3,10]. The proportion of cases exceeding the respective dose limits was summarized for both the overall cohort and within each frailty group.

Statistical analysis

Continuous variables are summarized as the mean ± standard deviation and categorical variables (including the proportion of cases exceeding the dose limit) as the number (percentage). Between-group comparisons (non-frail vs. frail) were performed using the Welch two-sample t-test for continuous variables and the chi-squared test or Fisher’s exact test or Fisher’s exact test for categorical variables as appropriate. Linear regression analysis was performed to visualize the relationship between actual body weight and the total lidocaine dose administered. The statistical analyses were performed using Python software (libraries including pandas, NumPy, and SciPy). All statistical tests were two-sided, and a p-value of < 0.05 was considered statistically significant. The minimal underlying dataset is provided as S1 data.

Reporting guidelines

The study was conducted and reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines for cohort studies.

Study population and baseline characteristics

A total of 150 patients underwent TF-FESS during the study period. After 42 exclusions, (surgeries involving two or more intervertebral levels, n = 35; spinal infection, n = 7), the study cohort included 108 patients, with 80 (74.1%) categorized into the non-frail group (mFI-5 < 2) and 28 (25.9%) into the frail group (mFI-5 ≥ 2) (Fig 1). The mean age of the overall cohort was 61.9 ± 18.0 years, and 64 patients (59.3%) were male. The frail group was significantly older than the non-frail group (74.5 years ± 10.6 vs. 57.5 ± 18.0 years, p < 0.001). There was no significant between-group difference in sex distribution (p = 0.685) or mean body weight (67.7 ± 13.1 kg vs. 66.8 ± 12.3 kg, p = 0.739).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow Diagram of Patient Selection.

In total, 108 of the 150 patients screened between January 2024 and March 2025 met the study inclusion criteria. Participants were categorized into a non-frail group (mFI-5 < 2, n = 80) and a frail group (mFI-5 ≥ 2, n = 28) based on the mFI-5 score. Abbreviations: CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; HT, hypertension; mFI-5, modified Frailty Index-5.

https://doi.org/10.1371/journal.pone.0349672.g001

Consistent with the frailty classification, the prevalence of each of the comorbidities assessed was significantly higher in the frail group than in the non-frail group (diabetes mellitus, 60.7% vs. 6.2%, p < 0.001; hypertension, 78.6% vs. 21.2%, p < 0.001; chronic obstructive pulmonary disease, 14.3% vs. 1.2%, p = 0.016; chronic heart failure, 53.6% vs. 0.0%, p < 0.001; and dependent functional status (10.7% vs. 0.0%, p = 0.016). Furthermore, the estimated safety limit for LA was significantly lower in the frail group (334 ± 62 mg vs. 474 ± 92 mg, p < 0.001). The baseline demographic and clinical characteristics are summarized in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Patient demographics and baseline characteristics.

https://doi.org/10.1371/journal.pone.0349672.t001

Data are presented as the number (percentage) unless otherwise indicated. Abbreviations: COPD, chronic obstructive pulmonary disease; LA, local anesthesia; mFI-5, modified Frailty Index-5

Surgical characteristics and intraoperative data

The surgical diagnoses primarily included lumbar disc herniation (65.7%), followed by lumbar lateral recess stenosis (17.6%) and lumbar foraminal stenosis (16.7%). The most frequently operated levels were L5/S1 (47.2%) and L4/5 (35.2%). There was no significant difference in the distribution of surgical diagnoses (p = 0.369) or the operative level (p = 0.125) between the non-frail and frail groups.

The mean operative time for the cohort overall was 83.8 ± 20.2 min, and estimated blood loss was minimal (4.6 ± 4.2 mL). Between-group comparisons found no significant difference in operative time between the non-frail and frail groups (84.5 ± 21.4 min vs. 81.6 ± 16.5 min, p = 0.467). Although the non-frail group had slightly greater estimated blood loss (5.0 ± 4.5 mL vs. 3.4 ± 3.1 mL), this difference reached only borderline significance (p = 0.050) and was not considered clinically meaningful. The surgical data are detailed in Table 2.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Surgical characteristics and operative outcomes.

https://doi.org/10.1371/journal.pone.0349672.t002

Data are presented as the mean ± standard deviation or as the number (percentage) as appropriate. Abbreviation: LAST, local anesthesia systemic toxicity

Safety and outcomes of local anesthesia

In all 108 cases (100%), continuous verbal communication with the patient was successfully maintained throughout the procedure, and the surgery was completed under LA without any need for conversion to general anesthesia. The mean total dose of lidocaine administered intraoperatively was 218 ± 40 mg, with no significant difference observed between the non-frail and frail groups (219 ± 41 mg vs. 216 ± 37 mg, p = 0.722). Evaluation of the total administered volume against the prespecified weight-based safety limits (7 mg/kg for non-frail patients and 5 mg/kg for frail patients) identified only 2 patients (1.9%) in the overall cohort in whom the respective thresholds were exceeded. Subgroup analysis revealed that no patients (0.0%) in the non-frail group exceeded the limit, whereas 2 patients (7.1%) in the frail group exceeded the stricter 5 mg/kg threshold (Fig 2). This difference did not reach statistical significance (p = 0.065). Linear regression analysis demonstrated that lidocaine dosing trends were nearly identical between the non-frail and frail groups. While the regression lines remained generally below the institutional thresholds for the majority of the cohort, the frail regression line converged toward and slightly intersected the 5 mg/kg limit in the lowest body weight range (Fig 2).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Relationship Between Total Lidocaine Dosage and Actual Body Weight Relative to Frailty-Adjusted Safety Limits (n = 108).

The individual scatter plots represent the total dose of lidocaine (mg) administered during the procedure. The dashed gray line indicates the conventional safety limit for non-frail patients (7 mg/kg), while the dashed black line represents the stricter safety threshold (5 mg/kg) applied to frail patients (mFI-5 ≥ 2). The thin solid lines represent the linear regression of actual doses for each group. Gray circles denote non-frail patients (mFI-5 < 2), and black triangles denote frail patients. Two frail cases slightly exceeded the 5 mg/kg threshold; no local anesthesia systemic toxicity symptoms were observed in these cases. Abbreviation: mFI-5, modified Frailty Index-5.

https://doi.org/10.1371/journal.pone.0349672.g002

Importantly, the procedure had an excellent safety profile. Regardless of frailty status and whether or not the LA safety threshold was exceeded, there were no documented cases of clinically significant LAST symptoms, nerve root injury, postoperative delirium, or conversion to general anesthesia across the entire cohort (p > 0.999 for all complications). No intravenous rescue analgesics/sedatives were required, and no procedures were interrupted or abandoned because of intolerable pain. Given the zero-event nature of these outcomes, the upper bound of the 95% confidence interval for the event rate was approximately 3.4% for the overall cohort (0/108) and 12.3% for the frail subgroup (0/28).

Safety of local anesthesia and weight-based dosing thresholds

A primary concern during TF-FESS under LA is the risk of LAST, particularly in elderly or frail patients, in whom drug metabolism is potentially altered [3,10,14]. To address this concern, we implemented weight-based safety thresholds, namely, 7 mg/kg for the non-frail group and a more conservative 5 mg/kg for the frail group calculated based on actual body weight. While these thresholds exceed the 4.5 mg/kg limit for plain lidocaine specified in the U.S. Food and Drug Administration (FDA)-approved labeling, our clinical observations suggest they are well-tolerated in the specific context of TF-FESS [3,10,15]. In the non-frail group (n = 80), 100% of patients remained within their safety limit. In the frail group (n = 28), the 5 mg/kg threshold was complied with in 92.9% of cases; although this limit was exceeded slightly in two frail cases because of operative requirements, no clinical symptoms of toxicity were observed. This consistent safety profile across both groups could potentially be explained by the specific anatomical and vascular characteristics of the surgical site. It is hypothesized that the systemic absorption of lidocaine from the subcutaneous tissue and fascia is relatively slow. Furthermore, a substantial volume of lidocaine was administered into the facet joint capsule. The dense fibrous structure of the joint capsule is hypothesized to act as a structural barrier, potentially sequestering part of the anesthetic agent within the intra-articular space and limiting its rapid diffusion to the surrounding extra-capsular tissues, which are more highly vascularized. By restricting immediate access to the systemic circulation in this manner, the joint capsule may help mitigate a rapid increase in intravascular lidocaine concentrations, potentially resulting a more gradual systemic uptake. This, combined with the poorly vascularized nature of the vertebral endplates and intradiscal spaces might further limit rapid systemic entry, although pharmacological studies measuring serum concentrations are required to confirm this mechanism [10,15]. Our data suggest that these weight-based thresholds may serve as a useful clinical benchmark for minimizing the risk of clinically significant LAST in patients undergoing TF-FESS under LA.

However, the interpretation of the 0% incidence of LAST in this study should be cautious. The retrospective design and reliance on medical record documentation primarily reflect the absence of documented, clinically significant LAST requiring medical intervention rather than the complete absence of all LAST-related symptoms. Additionally, the conscious sedation protocol using pentazocine and hydroxyzine could have potentially masked early, mild neurological signs of toxicity, such as metallic taste or circumoral numbness, which might have skewed the reported safety data.

Regarding the frailty spectrum, it is important to note that our frail group (mFI-5 ≥ 2) predominantly consisted of patients with mild frailty (mFI-5 = 2, 82.1%). The study lacked patients with severe frailty (mFI-5 ≥ 4), which may limit the generalizability of our findings to the most vulnerable populations. However, it is noteworthy that we applied the same conservative 5 mg/kg safety threshold to these mildly frail patients as would be used for more severely frail individuals. The fact that the vast majority of these patients were successfully managed within this stricter limit, with dosing trends remaining consistent and generally below the threshold. Further underscores the safety margin of our protocol. Nevertheless, our results highlight that TF-FESS under LA is a viable and low-stress alternative for mildly frail patients who might otherwise face higher risks under general anesthesia.

Clinical safety: Prevention of postoperative delirium and nerve root injury

The complete absence of documented postoperative delirium and nerve root injury across our entire cohort of 108 patients represents a noteworthy clinical observation. Postoperative delirium is a common and distressing complication in elderly patients undergoing spinal surgery under general anesthesia [4]. As shown in Table 3, Tacconi et al. reported a 14.5% postoperative delirium rate under general anesthesia in an elderly cohort, whereas no cases occurred under spinal anesthesia [16]. Although direct comparison is limited by differences in patient age and surgical techniques among these studies, our findings suggest that TF-FESS under LA may offer similar potential advantages in a broader surgical population. The retrospective identification of delirium through medical records without standardized tools like the Confusion Assessment Method may have under-detected hypoactive cases. By mitigating the physiological stress associated with general anesthesia and maintaining the patient’s conscious state, this approach suggests a potential for minimizing the risk of postoperative cognitive decline, even in individuals with frailty.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Selected reports of perioperative complications in patients who underwent lumbar spine surgery.

https://doi.org/10.1371/journal.pone.0349672.t003

Furthermore, the 0% nerve root injury rate suggests that TF-FESS under LA can be performed with a favorable safety profile. In contrast with historical data for endoscopic surgery under general anesthesia, where injury rates reached as high as 14.6% even in younger populations (Table 3) [17,18], use of LA allowed for continuous real-time neurological feedback from the patient [7,8]. This “biological monitoring” was effective regardless of frailty status, ensuring that any proximity to neural structures was immediately communicated to the surgeon [7,8,19]. The consistency of these results across our entire cohort underscores the potential of TF-FESS under LA not only as a viable alternative for high-risk patients but also as a valuable strategy for improving procedural safety in degenerative spinal surgery.

Limitations

This study has several limitations. First, it had a retrospective, single-center design and a relatively small sample size in the frail group (n = 28), which primarily consisted of mildly frail patients. While no major complications were observed, larger multicenter prospective studies are needed to confirm the generalizability of our findings. Second, we did not measure serum lidocaine concentrations; therefore, our assessment of local anesthetic safety remains based on clinical observations. Third, the possibility of under-detection of postoperative delirium and mild LAST symptoms as a consequence of the retrospective nature of the research cannot be excluded. Given that these outcomes were identified by review of medical records rather than standardized prospective screening tools, such as the Confusion Assessment Method, subclinical or hypoactive delirium and transient systemic toxicity might have been overlooked. However, it is important to note that no clinically significant episodes (i.e., those requiring medical intervention, causing safety concerns, or impacting the clinical course) were documented throughout the study period. Moreover, mild central nervous system symptoms of LAST could have been partially masked by the effects of the conscious sedation protocol. Fourth, the long-term clinical outcomes beyond the perioperative period were not evaluated. Future research should compare the long-term functional recovery and mortality rates between TF-FESS under LA and TF-FESS under general anesthesia in the frail population.