Impact of incidental thoracic spinal stenosis on outcomes following lumbar fusion: propensity score-matched cohort study Authors: Jeremy Tze En Lim, Ashton Kai Shun Tan, Shi Ting Chiu, Christian Hwee Yee Heng, Yee Gen Lim, Marcus Zhixing Ling, Reuben Chee Cheong Soh, Lei Jiang Published: 2026 ABSTRACT AbstractBackgroundIncidental thoracic spinal stenosis (TSS) is increasingly identified on whole-spine imaging, but its clinical significance in patients undergoing lumbar fusion remains unclear. This study aimed to evaluate whether incidental radiologic TSS is associated with postoperative outcomes following lumbar fusion.MethodsA retrospective propensity score-matched cohort study was conducted using a prospectively maintained registry [2010–2023] at a single tertiary centre. Patients with incidental radiologic TSS were matched 1:4 to controls without TSS based on age, sex, body mass index, and number of fused levels. Outcomes included Visual Analogue Scale (VAS), Oswestry Disability Index (ODI), North American Spine Society (NASS), and 36-Item Short Form Health Survey (SF-36) at baseline, 6 months, and 2 years.ResultsBoth groups demonstrated meaningful improvements in pain, disability, and quality-of-life outcomes at 6 months and 2 years. No significant between-group differences were observed in pain, disability, or quality-of-life outcomes at 6 months or 2 years (all P>0.05). During follow-up, 9/27 patients (33.3%) with TSS required thoracic surgery, whereas none in the control group did.ConclusionsIn this propensity score-matched cohort, no statistically significant differences in lumbar fusion outcomes were detected between patients with and without incidental radiologic TSS at 6 months or 2 years. However, the subsequent thoracic surgery rate of 33.3% suggests that incidental TSS should not be considered clinically benign and may warrant ongoing surveillance for neurological progression.Keywords: Thoracic spine, spinal stenosis, lumbar fusion, tandem spinal stenosis, treatment outcome Highlight box Key findings • No statistically significant differences in lumbar-specific pain, disability, or quality-of-life patient-reported outcome measures (PROMs) were detected at 6 months or 2 years. • One-third of patients demonstrate progression requiring thoracic surgery. What is known and what is new? • Tandem spinal stenosis is known to affect outcomes when symptomatic • This study found no statistically significant differences in lumbar-specific PROMs between patients with and without incidentally detected radiologic thoracic spinal stenosis following lumbar fusion. What is the implication, and what should change now? • These findings do not support prophylactic thoracic intervention based on imaging findings alone. Selective thoracic imaging and close neurological surveillance remain appropriate when clinically indicated. • Lumbar pathology should remain the primary driver of surgical decision-making in the absence of clinically dominant thoracic disease. Introduction Thoracic spinal stenosis (TSS) is considerably less common than cervical or lumbar stenosis, accounting for fewer than 2% of surgically treated cases of degenerative spinal stenosis (1). This lower prevalence has been attributed to the relative biomechanical stability of the thoracic spine, where the rib cage limits segmental motion and reduces cumulative mechanical stress compared with more mobile regions (2). Nevertheless, TSS can result in progressive spinal cord compression and subtle neurologic manifestations, including imbalance, reduced proprioception, and gait disturbance, which often lack clear dermatomal or myotomal patterns (3,4). In patients with concurrent lumbar pathology, these non-specific symptoms are frequently attributed to lumbar degeneration, and thoracic imaging is therefore not routinely performed. As a result, TSS may remain under-recognised, despite incidental thoracic lesions being reported in a substantial proportion of patients undergoing spine imaging for other indications, even in the absence of overt thoracic symptoms (5). This diagnostic overlap may contribute to delayed recognition of thoracic disease in patients with multilevel degenerative pathology. While the concept of tandem spinal stenosis has been well described in the context of concurrent cervical and lumbar disease and is known to influence surgical decision-making (6,7), the clinical significance of incidental TSS in patients undergoing lumbar fusion remains poorly defined. Existing studies have largely focused on symptomatic thoracic disease or specific aetiologies such as ossification of the ligamentum flavum (8,9), with limited data addressing the impact of incidentally detected, asymptomatic TSS on postoperative recovery following lumbar fusion. Whether radiologic TSS identified incidentally confers any meaningful difference in pain relief, functional recovery, or health-related quality of life after lumbar surgery therefore remains unclear. Accordingly, the objectives of this study were twofold. First, we sought to identify the presence of incidental thoracic spinal stenosis in a large institutional cohort of patients undergoing lumbar fusion for degenerative disease, reflecting real-world imaging practices in which thoracic pathology may be detected incidentally rather than through routine screening. Second, we evaluated whether incidental radiologic TSS was associated with differences in postoperative clinical outcomes following lumbar fusion. Using a propensity score-matched cohort, we compared patient-reported pain, functional, and quality-of-life outcomes between patients with and without TSS, and additionally examined the subsequent need for thoracic surgical intervention during follow-up. This cohort differs from classical tandem spinal stenosis, as thoracic pathology was identified incidentally during evaluation for lumbar degenerative disease and was not the primary indication for surgical treatment at the time of index lumbar fusion. We present this article in accordance with the STROBE reporting checklist (available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0111/rc). Methods Study design and population This was a retrospective propensity score-matched cohort study conducted at a single tertiary spine centre. Adult patients (≥18 years) who underwent primary lumbar fusion for degenerative spinal pathology between January 2010 and January 2023 were identified from the institutional surgical registry. Within this registry, patients who had undergone spinal surgery involving the thoracic or cervical spine in addition to lumbar fusion were identified as a population at higher risk of concomitant thoracic spinal pathology. All 624 such patients were systematically screened for thoracic spinal stenosis using available scout views and/or dedicated thoracic magnetic resonance imaging (MRI), reviewed by fellowship-trained spine surgeons. The study cohort selection process is summarised in Figure 1. This study therefore evaluated the prognostic impact of incidental TSS within this screened cohort rather than aiming to estimate its population prevalence. Flow diagram illustrating registry-based cohort selection. Of 3,766 patients who underwent lumbar fusion, 624 had undergone concurrent or additional cervical or thoracic spinal surgery and were systematically screened for thoracic spinal stenosis using scout views and dedicated thoracic MRI. Of these, 27 were identified with incidental radiologic TSS and 597 demonstrated no TSS. Following 1:4 propensity score matching, the final analytic cohort comprised 27 TSS patients and 108 matched controls. BMI, body mass index; MRI, magnetic resonance imaging; NASS, North American Spine Society; ODI, Oswestry Disability Index; PROMs, patient-reported outcome measures; SF-36, 36-Item Short Form Health Survey; SGH, Singapore General Hospital; TSS, thoracic spinal stenosis; VAS, Visual Analogue Scale. Patients with available preoperative thoracic MRI and complete baseline and follow-up patient-reported outcome measures (PROMs) were included. Patients with incomplete PROM data at any timepoint were excluded prior to cohort formation; all patients retained in the final matched cohort had complete PROM datasets at baseline, 6 months, and 2 years. Thoracic MRI was reviewed to classify patients with incidental TSS and to identify controls without radiologic TSS. Controls were selected exclusively from patients who had undergone thoracic MRI and demonstrated no radiologic evidence of TSS, thereby ensuring that undetected TSS in the control group could not account for the observed between-group comparisons. As thoracic MRI was not routinely performed in all lumbar fusion patients, this study should be interpreted as an analysis of a clinically selected imaging cohort rather than a population-based prevalence study. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the SingHealth Centralised Institutional Review Board (CIRB) (No. 2023/2214), and individual consent for this retrospective analysis was waived. Patients were excluded if they had prior spinal surgery, spinal trauma, tumour, infection, or incomplete PROM data. Of 3,766 patients identified in the registry, 624 had undergone spinal surgery involving the thoracic or cervical spine in addition to lumbar fusion. All 624 were systematically screened for thoracic spinal stenosis using scout views and dedicated thoracic MRI reviewed by fellowship-trained spine surgeons, of whom 27 were identified with significant radiologic TSS and 597 demonstrated no radiologic evidence of TSS. All patients were required to have complete PROM data at baseline, 6 months, and 2 years; patients with incomplete follow-up were excluded prior to cohort formation. From the 597 eligible controls, 108 were selected by propensity score matching, yielding a final matched cohort of 27 TSS patients and 108 controls. Definition of thoracic spinal stenosis TSS was defined as radiographic narrowing of the central thoracic spinal canal with compression of neural elements on MRI, in accordance with established definitions in the literature (2). The presence of intramedullary T2 signal change was recorded when present but was not required for diagnosis. Patients who underwent concurrent thoracic surgical intervention at the time of index lumbar fusion were excluded. As part of routine preoperative evaluation, the operating spine surgeon assessed all patients for clinical signs of myelopathy, including gait disturbance, lower limb hyperreflexia, and sensory deficits; no patient in the TSS group demonstrated overt thoracic myelopathy necessitating immediate surgical intervention. No formalised myelopathy grading instrument (e.g., mJOA or Nurick grade) was routinely applied, as detailed in “Limitations” section. All thoracic MRI scans were independently reviewed by two fellowship-trained spine surgeons who were blinded to clinical outcomes. Discrepancies were resolved by consensus. Interobserver agreement was assessed using Cohen’s kappa, yielding κ =0.68, indicating substantial agreement. Patients meeting imaging criteria were classified into the TSS group, while patients without radiologic evidence of TSS served as potential controls. Propensity score matching and statistical analysis Patients with TSS were matched in a 1:4 ratio to patients without radiologic TSS using propensity score matching based on age, sex, body mass index (BMI), and number of fused lumbar levels. Nearest-neighbour matching without replacement was performed using a caliper of 0.2 standard deviations of the logit of the propensity score (caliper =0.09), resulting in a matched cohort of 27 patients with TSS and 108 matched controls. Surgical approach (minimally invasive versus open) was not included in the matching model to preserve sample size and model stability, but was evaluated post-matching and did not differ significantly between groups. Continuous variables are reported as mean ± standard deviation. Baseline demographic and surgical characteristics were compared using independent-samples t-tests for continuous variables and chi-squared tests for categorical variables. Between-group comparisons of PROMs at baseline, 6 months, and 2 years were performed using independent-samples t-tests for approximately normally distributed variables. Ordinal or non-normally distributed outcomes, including Visual Analogue Scale (VAS) pain scores, were analysed using Mann-Whitney U tests. Change scores from baseline to 6 months and from baseline to 2 years were calculated and compared using the same statistical approach. Because propensity score matching was performed without replacement and was intended to improve baseline comparability rather than create paired observations, post-matching outcome comparisons were conducted using conventional between-group statistical methods. Effect sizes were reported to aid clinical interpretation, with Cohen’s d used for continuous outcomes and r (Z/√N) for non-parametric comparisons. Statistical significance was defined as P<0.05 using two-sided tests. Given the exploratory nature of this study and the limited sample size of the TSS cohort, no formal adjustment for multiple comparisons was performed. Accordingly, findings should be interpreted as exploratory and hypothesis-generating rather than confirmatory. All statistical analyses were conducted using SPSS Statistics (version 29.0; IBM Corp., Armonk, NY, USA). Given the retrospective design and fixed cohort size, an a priori sample size calculation was not performed. Instead, a minimum detectable difference analysis was conducted using a representative clinical outcome [Oswestry Disability Index (ODI) at 2 years]. With 27 patients in the TSS group and 108 matched controls, the study had approximately 80% power (two-sided α =0.05) to detect a between-group difference of approximately 9–10 ODI points, corresponding to a moderate effect size (Cohen’s d ≈0.6). This magnitude corresponds to the upper range of reported minimal clinically important difference (MCID) thresholds for the ODI, supporting the study’s ability to detect clinically meaningful differences if present. Outcome measures Patient-reported outcome measures (PROMs) were selected to capture pain, functional disability, and health-related quality of life following lumbar fusion and were collected preoperatively (baseline) and at 6 months and 2 years postoperatively as part of routine clinical follow-up. Pain severity was assessed using the VAS for back pain and leg pain. Functional disability was measured using the ODI. Health-related quality of life was evaluated using the 36-Item Short Form Health Survey (SF-36), including eight domains: physical function, role physical, bodily pain, general health, vitality, social functioning, role emotional, and mental health. A descriptive overall SF-36 summary score was reported for completeness but was not treated as a primary outcome or used to drive inferential conclusions. Symptom burden was additionally assessed using the modified North American Spine Society (NASS) symptom score. These outcome measures were selected to evaluate postoperative pain, disability, symptom burden, and health-related quality of life following lumbar fusion. Results Patient demographics and surgical characteristics Baseline demographic and surgical characteristics are presented in Table 1. There were no significant differences between groups in age, sex, BMI, number of fused levels, or surgical approach (minimally invasive versus open), indicating effective propensity score matching (all P>0.05). Table 1Baseline demographic and surgical characteristics of the propensity-matched cohortVariableOverall (n=135)TSS (n=27)No TSS (n=108)P valueAge (years)67.8±13.768.2±12.567.7±14.00.91Male69 (51.1)14 (51.9)55 (50.9)0.93BMI (kg/m2)27.8±4.727.6±4.827.8±4.70.88Number of fused levels3.6±1.13.6±1.13.6±1.1>0.99Surgical approach0.33 MIS51 (37.8)8 (29.6)43 (39.8) Open84 (62.2)19 (70.4)65 (60.2)Values are expressed as mean ± standard deviation or n (%). Propensity score matching was performed at a 1:4 ratio based on age, sex, BMI and number of fused levels. Comparisons were calculated using χ2 test or Fisher’s exact test for categorical variables, as appropriate, and independent-samples t-test for continuous variables. BMI, body mass index; MIS, minimally invasive surgery; TSS, thoracic spinal stenosis. Baseline demographic and surgical characteristics of the propensity-matched cohort Values are expressed as mean ± standard deviation or n (%). Propensity score matching was performed at a 1:4 ratio based on age, sex, BMI and number of fused levels. Comparisons were calculated using χ2 test or Fisher’s exact test for categorical variables, as appropriate, and independent-samples t-test for continuous variables. BMI, body mass index; MIS, minimally invasive surgery; TSS, thoracic spinal stenosis. Characteristics of thoracic spinal stenosis Thoracic stenosis predominantly involved the lower thoracic spine and thoracolumbar junction, with T11–12 representing the most frequently affected level. One patient demonstrated additional rostral involvement at T6–7. Etiology subclassification and formal stenosis severity grading were not systematically captured in the registry. Available imaging reports documented the presence and level of canal narrowing with neural element compression but did not uniformly characterise the underlying pathology, such as ossification of the ligamentum flavum, ossification of the posterior longitudinal ligament, or thoracic disc herniation. Intramedullary T2 cord signal change was identified in 1 of 27 patients (3.7%); therefore, no subgroup analysis was performed. All patients in the TSS cohort met the study definition of radiological thoracic canal stenosis with associated neural element compression. Subsequent thoracic surgical intervention During postoperative follow-up, 9 of 27 patients (33.3%) in the TSS cohort underwent subsequent thoracic surgical intervention for progression of stenosis or development of myelopathic symptoms, compared with none in the control cohort. All procedures consisted of posterior thoracic decompression with instrumented fusion, most commonly at the thoracolumbar junction (T9–L1), frequently requiring extension of existing lumbar constructs; vertebroplasty-augmented fixation was performed in one patient. In the majority of cases, 2-year PROM assessments were completed prior to thoracic surgery, and thoracic intervention was indicated by clinical progression of pre-existing stenosis rather than new or adjacent pathology. Given the small number of events, formal sensitivity analyses were not performed; however, between-group PROM differences remained non-significant on descriptive review when these patients were considered separately. Preoperative outcomes Preoperative patient-reported outcome measures are presented in Table 2. At baseline, there were no statistically significant differences between the TSS and control groups across pain scores (VAS back and leg), disability (ODI), SF-36 domain scores, or NASS neurogenic symptom scores (all P>0.05), demonstrating good baseline comparability following propensity score matching. Table 2Preoperative patient-reported outcome measures following propensity score matchingOutcomeTSS (n=27)Control (n=108)P valueCohen’s dSFPF29.44±26.6234.91±26.920.38−0.19SFRPF12.04±25.4111.81±24.590.970.01SFBP24.70±23.4827.83±23.490.54−0.13SFGH67.81±20.3263.67±22.060.430.16SFVI56.30±24.0258.38±23.330.60−0.09SFSF43.06±37.8148.31±38.220.48−0.14SFRFE65.43±40.3771.23±38.330.41−0.15SFMH64.04±22.3468.15±20.400.32−0.19SFAVE46.45±21.9649.72±21.750.44−0.18VASBP6.74±2.886.26±3.170.480.15VASLP6.30±3.636.07±3.520.730.07ODI56.07±20.7749.63±21.980.150.30NASS54.94±20.5653.47±23.440.780.07Values are presented as mean ± standard deviation. Higher scores indicate better health status for all SF-36 domains; higher scores indicate worse disability for ODI and NASS. Between-group comparisons were performed following propensity score matching. NASS, North American Spine Society; ODI, Oswestry Disability Index; SF-36, 36-Item Short Form Health Survey; SFAVE, SF-36 average summary score (exploratory); SFBP, SF-36 bodily pain; SFGH, SF-36 general health; SFMH, SF-36 mental health; SFPF, SF-36 physical function; SFRFE, SF-36 role emotional; SFRPF, SF-36 role physical; SFSF, SF-36 social functioning; SFVI, SF-36 vitality; TSS, thoracic spinal stenosis; VASBP, Visual Analogue Scale—back pain; VASLP, Visual Analogue Scale—leg pain. Preoperative patient-reported outcome measures following propensity score matching Values are presented as mean ± standard deviation. Higher scores indicate better health status for all SF-36 domains; higher scores indicate worse disability for ODI and NASS. Between-group comparisons were performed following propensity score matching. NASS, North American Spine Society; ODI, Oswestry Disability Index; SF-36, 36-Item Short Form Health Survey; SFAVE, SF-36 average summary score (exploratory); SFBP, SF-36 bodily pain; SFGH, SF-36 general health; SFMH, SF-36 mental health; SFPF, SF-36 physical function; SFRFE, SF-36 role emotional; SFRPF, SF-36 role physical; SFSF, SF-36 social functioning; SFVI, SF-36 vitality; TSS, thoracic spinal stenosis; VASBP, Visual Analogue Scale—back pain; VASLP, Visual Analogue Scale—leg pain. Six-month outcomes Six-month absolute scores and changes from baseline are summarised in Table 3. At 6 months, both groups demonstrated numerical improvement across pain, disability, and functional domains compared with baseline. There were no statistically significant differences in absolute 6-month scores between groups for any SF-36 domain, ODI, VAS back or leg pain, or NASS scores (all P>0.05). Table 3Six-month absolute scores and change from baseline in patient-reported outcomesOutcomeTSSControlP valueCohen’s dSix-month absolute scores SFPF55.6±26.060.5±26.70.38−0.19 SFRPF50.0±45.546.3±42.90.700.08 SFBP58.0±29.259.1±29.90.85−0.04 SFGH61.4±21.766.2±23.30.31−0.21 SFVI60.4±20.967.3±21.70.13−0.32 SFSF67.6±39.677.7±32.30.22−0.30 SFRFE86.4±33.789.2±30.90.69−0.09 SFMH77.0±18.580.5±18.00.38−0.19 SFAVE64.5±22.668.4±21.10.43−0.18 VASBP2.0±3.41.3±2.50.370.24 VASLP1.0±2.60.8±2.20.710.09 ODI24.4±20.920.2±16.60.330.22 NASS19.1±20.716.5±21.20.560.12Change from baseline to 6 months Δ SFPF+26.16±38.08+25.59±36.930.25+0.02 Δ SFRPF+37.96±44.73+34.49±42.730.62+0.08 Δ SFBP+33.30±33.79+31.27±36.270.89+0.06 Δ SFGH−6.41±20.52+2.53±19.860.67−0.45 Δ SFVI+4.10±22.84+8.92±23.340.95−0.21 Δ SFSF+24.54±28.59+29.39±27.780.87−0.17 Δ SFRFE+20.97±45.12+17.97±43.710.89+0.07 Δ SFMH+12.96±24.29+12.35±22.780.79+0.03 Δ SFAVE+18.05±24.62+18.68±23.920.63−0.03 Δ VASBP−4.74±3.68−4.96±3.580.84+0.06 Δ VASLP−5.30±4.19−5.27±4.040.83−0.01 Δ ODI−31.67±23.69−29.43±22.500.72−0.10 Δ NASS−35.84±24.04−36.97±25.850.83+0.04Values are presented as mean ± standard deviation. Δ, change from baseline to 6 months. For ODI, NASS, and VAS outcomes, negative Δ values indicate improvement. Higher SF-36 scores indicate better health status. Between-group comparisons were performed using independent-samples t-tests or Mann-Whitney U tests, as appropriate. Effect sizes are reported as Cohen’s d. NASS, North American Spine Society; ODI, Oswestry Disability Index; SF-36, 36-Item Short Form Health Survey; SFAVE, SF-36 average summary score (exploratory); SFBP, SF-36 bodily pain; SFGH, SF-36 general health; SFMH, SF-36 mental health; SFPF, SF-36 physical function; SFRFE, SF-36 role emotional; SFRPF, SF-36 role physical; SFSF, SF-36 social functioning; SFVI, SF-36 vitality; TSS, thoracic spinal stenosis; VAS, Visual Analogue Scale; VASBP, Visual Analogue Scale—back pain; VASLP, Visual Analogue Scale—leg pain. Six-month absolute scores and change from baseline in patient-reported outcomes Values are presented as mean ± standard deviation. Δ, change from baseline to 6 months. For ODI, NASS, and VAS outcomes, negative Δ values indicate improvement. Higher SF-36 scores indicate better health status. Between-group comparisons were performed using independent-samples t-tests or Mann-Whitney U tests, as appropriate. Effect sizes are reported as Cohen’s d. NASS, North American Spine Society; ODI, Oswestry Disability Index; SF-36, 36-Item Short Form Health Survey; SFAVE, SF-36 average summary score (exploratory); SFBP, SF-36 bodily pain; SFGH, SF-36 general health; SFMH, SF-36 mental health; SFPF, SF-36 physical function; SFRFE, SF-36 role emotional; SFRPF, SF-36 role physical; SFSF, SF-36 social functioning; SFVI, SF-36 vitality; TSS, thoracic spinal stenosis; VAS, Visual Analogue Scale; VASBP, Visual Analogue Scale—back pain; VASLP, Visual Analogue Scale—leg pain. Similarly, the magnitude of improvement from baseline to 6 months was comparable between groups across all outcome measures, with no significant differences in change scores (all P>0.05). Effect sizes for between-group differences were small across domains, suggesting minimal clinical divergence in early postoperative recovery between patients with and without incidental TSS. Two-year outcomes Two-year absolute scores and changes from baseline are presented in Table 4. At 2 years postoperatively, pain (VAS back and leg), disability (ODI), SF-36 domain scores, and NASS symptom scores were comparable between the TSS and control cohorts, with no statistically significant between-group differences observed (all P>0.05). Table 4Two-year absolute scores and change from baseline in patient-reported outcomesOutcomeTSSControlP valueCohen’s dTwo-year absolute scores SFPF51.67±37.9458.31±39.320.46−0.17 SFRPF36.67±44.9339.29±46.300.81−0.06 SFBP55.94±33.0259.60±35.610.67−0.11 SFGH59.94±19.1761.07±18.360.88−0.06 SFVI62.06±21.8267.14±23.550.39−0.16 SFSF75.98±27.6277.86±26.650.76−0.07 SFRFE67.16±44.7669.76±43.200.80−0.06 SFMH63.75±23.6566.43±22.350.62−0.11 SFAVE56.04±23.7860.40±24.140.44−0.18 VASBP1.59±2.871.30±2.590.710.11 VASLP0.89±2.580.80±2.290.960.04 ODI18.19±17.3215.88±14.890.530.15 NASS17.65±25.5712.87±19.680.510.23Change from baseline to 2 years Δ SFPF+22.23±39.07+23.40±36.920.46−0.03 Δ SFRPF+24.63±45.96+27.48±41.700.81−0.07 Δ SFBP+31.24±33.02+31.77±35.610.67−0.02 Δ SFGH−7.87±19.17−2.60±18.360.88−0.28 Δ SFVI+5.76±21.82+8.76±23.550.39−0.13 Δ SFSF+32.92±27.62+29.55±26.650.76+0.13 Δ SFRFE+1.73±44.76−1.47±43.200.80+0.07 Δ SFMH−0.29±23.65−1.72±22.350.62+0.06 Δ SFAVE+9.59±23.78+10.68±24.140.44−0.05 Δ VASBP−5.15±3.54−4.96±3.720.87−0.05 Δ VASLP−5.41±4.48−5.27±4.260.90−0.03 Δ ODI−37.88±23.12−33.75±21.650.98−0.19 Δ NASS−37.29±25.67−40.60±25.600.55+0.13Values are presented as mean ± standard deviation. Δ, change from baseline to two years. For ODI, NASS, and VAS outcomes, negative Δ values indicate improvement. Higher SF-36 scores indicate better health status. Between-group comparisons were performed using independent-samples t-tests or Mann-Whitney U tests, as appropriate. Effect sizes are reported as Cohen’s d. NASS, North American Spine Society; ODI, Oswestry Disability Index; SF-36, 36-Item Short Form Health Survey; SFAVE, SF-36 average summary score (exploratory); SFBP, SF-36 bodily pain; SFGH, SF-36 general health; SFMH, SF-36 mental health; SFPF, SF-36 physical function; SFRFE, SF-36 role emotional; SFRPF, SF-36 role physical; SFSF, SF-36 social functioning; SFVI, SF-36 vitality; TSS, thoracic spinal stenosis; VAS, Visual Analogue Scale; VASBP, Visual Analogue Scale—back pain; VASLP, Visual Analogue Scale—leg pain. Two-year absolute scores and change from baseline in patient-reported outcomes Values are presented as mean ± standard deviation. Δ, change from baseline to two years. For ODI, NASS, and VAS outcomes, negative Δ values indicate improvement. Higher SF-36 scores indicate better health status. Between-group comparisons were performed using independent-samples t-tests or Mann-Whitney U tests, as appropriate. Effect sizes are reported as Cohen’s d. NASS, North American Spine Society; ODI, Oswestry Disability Index; SF-36, 36-Item Short Form Health Survey; SFAVE, SF-36 average summary score (exploratory); SFBP, SF-36 bodily pain; SFGH, SF-36 general health; SFMH, SF-36 mental health; SFPF, SF-36 physical function; SFRFE, SF-36 role emotional; SFRPF, SF-36 role physical; SFSF, SF-36 social functioning; SFVI, SF-36 vitality; TSS, thoracic spinal stenosis; VAS, Visual Analogue Scale; VASBP, Visual Analogue Scale—back pain; VASLP, Visual Analogue Scale—leg pain. Improvement from baseline appeared to be maintained at two years in both groups, and change scores did not differ significantly across any outcome measure (all P>0.05). Effect sizes remained small, indicating similar long-term recovery trajectories irrespective of the presence of incidental TSS. Discussion Main findings In this propensity score-matched cohort of patients undergoing lumbar fusion, no statistically significant differences in pain, disability, or health-related quality-of-life outcomes were detected between patients with and without incidental radiologic TSS at either 6 months or 2 years. These findings should not be interpreted as demonstrating equivalence or absence of clinical effect, particularly given the relatively small TSS cohort and the limitations of PROM-based assessment. Radiologic TSS identified incidentally may represent coexisting degenerative pathology that was not associated with detectable differences in lumbar-specific PROMs in the present cohort. However, these findings should be interpreted within the context of the study design, sample size, and outcome measures used. Rather, incidental TSS may reflect coexisting degenerative pathology rather than a determinant of lumbar surgical outcomes. Clinically, this is relevant given that TSS is frequently overlooked in patients with multilevel degenerative disease and is often detected incidentally on extended imaging (1,5). A noteworthy observation in our cohort was that 9 of 27 patients (33.3%) with incidental TSS subsequently underwent thoracic surgical intervention during postoperative follow-up for progression of stenosis or development of myelopathic symptoms. No patients in the matched control group required thoracic surgery. These subsequent procedures uniformly involved posterior thoracic decompression with instrumented fusion, but the small number of events precluded meaningful subgroup analyses. Although our study was not designed to identify predictors of progression, this observation suggests that incidentally detected TSS may not be clinically inert in all patients and may evolve over time in a subset, but should be interpreted descriptively rather than causally given the small sample size and retrospective design. Although thoracic disease progression occurred in a subset of patients, no statistically significant deterioration in lumbar-specific PROMs was detected at the cohort level. Given the limited sample size and retrospective design, the relationship between thoracic disease progression and lumbar recovery cannot be determined from the present study. TSS remains challenging to detect clinically. Early manifestations such as gait hesitation, lower limb stiffness, imbalance, and unexplained fatigue are often subtle and non-specific (3,4,8), and unlike cervical disease, TSS seldom presents with distinct motor or sensory levels on routine neurologic examination (2,5). In this context, selective whole-spine imaging may be considered in patients with unexplained gait disturbance, widespread degenerative disease, or systemic complaints not fully accounted for by lumbar imaging alone (4). Such an approach may facilitate earlier recognition of multiregional spinal pathology and inform postoperative surveillance, rather than mandate routine thoracic imaging or pre-emptive intervention. The optimal strategy for managing incidental TSS remains unclear. Evidence from the tandem spinal stenosis literature—predominantly focused on cervical-lumbar disease—suggests that both staged and simultaneous decompression strategies yield comparable functional and neurologic outcomes, with surgical sequencing often guided by symptom predominance rather than radiographic findings alone (6,7,9,10). However, these data largely exclude thoracic-predominant pathology, and their applicability to thoracolumbar combinations remains limited. As this study did not compare routine versus selective thoracic imaging strategies, it cannot determine whether universal thoracic MRI should or should not be performed before lumbar fusion. Rather, the findings suggest that incidental TSS alone should not automatically prompt prophylactic thoracic intervention in the absence of clinically dominant thoracic pathology. Further prospective studies are needed to better define the natural history and progression of incidentally detected TSS and to clarify optimal management strategies in this population (1,7). Notably, because thoracic myelopathy differs biomechanically and neurologically from cervical disease, extrapolation from tandem cervical-lumbar stenosis literature to thoracic-lumbar pathology should be interpreted cautiously. For clinicians, these findings suggest that the presence of incidental TSS alone should not delay or alter surgical decision-making for lumbar fusion when lumbar pathology is concordant with symptoms. However, recognition of subclinical thoracic disease may inform postoperative counselling and longer-term clinical surveillance for neurologic progression. Limitations This study has several limitations. First, its retrospective design and reliance on propensity score matching limit causal inference. Although matching was performed for key variables such as age, sex, BMI, and number of fused levels, unmeasured confounders—including gait mechanics, sarcopenia, or frailty—may have influenced recovery trajectories. Formal thoracic myelopathy severity instruments such as the modified Japanese Orthopaedic Association (mJOA) score and Nurick grade were not routinely collected. As such, subclinical myelopathic signs detected on routine examination were documented clinically but were not quantified using a validated scale, which limits grading of thoracic disease severity and represents a recognised limitation of retrospective cohort studies in this setting. Consequently, subtle thoracic neurological dysfunction may have been under-recognised and thoracic disease severity could not be quantified systematically. Formal pre- and post-matching standardized mean differences were not available from the original matching workflow; covariate balance was instead assessed using conventional comparisons of matched variables, with no significant differences observed across matching covariates (Table 1). Second, the number of patients with incidental thoracic spinal stenosis was relatively small (n=27), reflecting the frequency of incidentally detected disease within the available imaging framework. While the study was powered to detect moderate differences in disability, it may have been underpowered to detect small between-group effects. Importantly, between-group differences were consistently small with correspondingly small effect sizes, suggesting that clinically meaningful differences are less likely, although the possibility of small undetected effects cannot be excluded. Third, thoracic MRI was selectively obtained rather than performed routinely in all lumbar fusion patients. Consequently, undetected TSS in patients who did not undergo dedicated thoracic imaging cannot be excluded. Although this workflow reflects real-world clinical practice, it introduces potential selection bias, as patients undergoing thoracic MRI may represent a higher-risk subgroup. Incidental thoracic pathology may also have been missed in cases of subtle disease or incomplete scout view reporting, potentially leading to under-detection of asymptomatic TSS and underestimation of its true prevalence. In addition, imaging reports did not consistently characterise stenosis severity or subtle myelopathic features, which may have limited phenotypic granularity. Fourth, the PROMs used in this study were designed primarily to assess lumbar pain, disability, and quality of life rather than thoracic myelopathy-specific symptoms. Subtle manifestations of thoracic cord dysfunction, including gait imbalance, proprioceptive impairment, or spasticity, may therefore not have been fully captured. Responder analyses based on MCID thresholds for ODI and VAS were not performed in the current study and represent an avenue for future work. Finally, follow-up was limited to 2 years. Although this timeframe captures early to mid-term postoperative recovery, it may not fully reflect longer-term neurologic progression, delayed reoperations, or late complications. Future studies incorporating systematic whole-spine imaging, larger cohorts, and extended follow-up are warranted to better define the natural history and clinical significance of incidental TSS. A formal sensitivity analysis excluding patients who underwent subsequent thoracic surgery prior to their 2-year assessment was not performed given the small number of events (n=9); future studies with larger TSS cohorts should consider time-to-event analyses and sensitivity analyses treating thoracic surgery as a competing endpoint. Conclusions In this selected imaging cohort, no statistically significant differences in pain, disability, or health-related quality-of-life outcomes were detected between patients with and without incidental radiologic TSS at 6 months or 2 years following lumbar fusion. However, the subsequent thoracic surgery rate of 33.3% suggests that incidental TSS should not be regarded as clinically benign. These findings support lumbar pathology-guided surgical decision-making, with selective thoracic imaging and ongoing neurological surveillance when clinically indicated. Supplementary The article’s supplementary files as Acknowledgments None. Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the SingHealth Centralised Institutional Review Board (CIRB) (No. 2023/2214), and individual consent for this retrospective analysis was waived. Footnotes Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0111/rc Funding: None. Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0111/coif). The authors have no conflicts of interest to declare. Data Sharing Statement Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0111/dss References Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0111/prf Supplementary Materials The article’s supplementary files as Data Availability Statement Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0111/dss