Abstract
Background
There is a lack of comparative studies examining patient-reported outcomes (PROs) after cervical spine surgery between rheumatoid arthritis (RA) and non-RA patients. The objective of this study was to compare postoperative outcomes after cervical spine surgery between RA and non-RA patients.
Methods
This retrospective cohort study utilized prospectively registered data from patients older than 18 years who underwent primary cervical spine surgery from April 2017 to March 2022 at 13 institutions in the Greater Tokyo metropolitan area. Demographics, perioperative complications, surgical details, and PROs were recorded. Propensity-score overlap weighting was used to adjust for baseline and surgical differences between RA and non-RA groups, and group differences were reported with 95% confidence intervals.
Results
In this cohort of 2,729 participants, 2.5% (67 individuals) were diagnosed with RA. These RA patients were predominantly older women with lower body mass index and more frequently required upper cervical or fusion surgery. The average duration of RA was 21 years, with treatment involving steroids in 56.7%, methotrexate in 41.8%, and biologics in 19.4%. After overlap weighting, RA patients had poorer preoperative and postoperative PROs than non-RA patients. However, RA patients showed improvements in several symptoms and quality-of-life measures. The overlap-weighted improvement in EQ-5D was 0.14 in RA patients and 0.13 in non-RA patients [difference, 0.01; 95% confidence interval (CI): −0.06 to 0.09]. The Neck Disability Index (NDI) showed only limited improvement in RA patients, whereas postoperative satisfaction was not clearly lower in RA patients.
Conclusions
Patients with RA had poorer PROs before and after cervical spine surgery than non-RA patients. Nevertheless, selected RA patients showed improvement in several symptoms and quality-of-life measures, and postoperative satisfaction was not clearly lower. Disability measured by the NDI may remain less responsive, and the limited number of RA patients and wide confidence intervals preclude any conclusion of equivalence with non-RA patients.
Keywords: Atlanto-axial dislocation, myelopathy, radiculopathy, upper cervical spine, patient-reported outcomes (PROs)
Highlight box.
Key findings
• Rheumatoid arthritis (RA) accounted for 2.5% of patients undergoing cervical spine surgery and was associated with older age, female sex, lower body mass index, and more frequent upper cervical or fusion surgery.
• After propensity-score overlap weighting, RA patients had poorer preoperative and postoperative patient-reported outcomes than non-RA patients, but improvements were observed in several symptoms and quality-of-life measures.
• Postoperative satisfaction was not clearly lower in RA patients, although the small number of RA patients and the 95% confidence intervals do not support a conclusion of equivalence between RA and non-RA patients.
What is known and what is new?
• RA has traditionally been associated with higher perioperative risk and poorer overall health status in spine surgery.
• Comparative data on patient-reported outcomes after cervical spine surgery, adjusted for the marked baseline differences between RA and non-RA patients, have been limited.
• This multicenter weighted analysis suggests that improvement in several symptoms and quality-of-life measures is still achievable in selected RA patients, although disability measured by the Neck Disability Index may remain less responsive and equivalence with non-RA patients cannot be concluded.
What is the implication, and what should change now?
• Cervical spine surgery should not be discounted in RA patients solely because baseline quality of life is poor.
• Counseling should emphasize that selected RA patients may improve despite poorer absolute patient-reported outcomes and that postoperative satisfaction is not clearly lower, while avoiding an overstatement that outcomes are equivalent to those of non-RA patients.
Introduction
Rheumatoid arthritis (RA) is a chronic, systemic autoimmune disorder primarily affecting synovial joints, leading to significant morbidity. A critical challenge in RA’s clinical management is cervical spine involvement, impacting a considerable patient population. A recent systematic review indicated that cervical spine involvement in RA patients ranges from 16% to 70% (1). Reports indicating a higher frequency of cervical spine involvement in RA often include cases with radiographic subaxial instability (2), while the actual incidence of cases presenting neurological symptoms varies, with reported rates ranging from 7% to 56% (3). Moreover, progressive joint damage and ligamentous laxity can lead to subluxation and instability (4). The cervical spine’s pathophysiology in RA involves a complex interplay of inflammatory, immunological, and mechanical factors (5). Prolonged inflammation contributes to synovial joint degeneration, resulting in ligamentous laxity and subsequent spinal instability. This instability is most pronounced in the atlantoaxial (C1–C2) region but can extend to the subaxial spine (C3–C7) (6). Clinical manifestations range from asymptomatic radiological abnormalities to severe neurological deficits, such as myelopathy, highlighting the importance of early diagnosis and intervention (7).
Historically, spinal surgeries in RA patients have been linked to less favorable outcomes. For instance, Takenaka et al. observed a higher incidence of postoperative spinal surgical site infections (SSIs) in RA cases (8). Similarly, Streufert and colleagues reported an increased rate of postoperative SSIs and subsequent surgeries in RA patients (9). Kato et al. reported that RA patients undergoing lumbar spine surgery had poorer postoperative physical function and EQ-5D scores overall, although outcomes were more comparable in steroid-free patients after propensity-score adjustment (10). Based on the findings presented in these studies, it is anticipated that the incidence of postoperative complications, including SSIs, is comparatively elevated in spinal surgical interventions among patients diagnosed with RA. However, there is a lack of comparative studies examining surgical outcomes between RA and non-RA patients, considering demographic and procedural variables. Therefore, our study aimed to fill this knowledge gap by analyzing cervical spine surgery data from multiple institutions, comparing patient-reported outcomes (PROs) between RA and non-RA cohorts. Recognizing potential disparities in demographic and clinical characteristics between the groups, we utilized propensity score methods to ensure a fair comparison. The objective of this study is to compare postoperative outcomes after cervical spine surgery between RA and non-RA patients. We present this article in accordance with the STROBE reporting checklist (available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0149/rc).
Methods
Study design and patient population
This retrospective cohort study utilized prospectively registered data. The study was approved by the institutional review board of the Clinical Research Support Center at the authors’ institution. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from all participants in the registry. All participating institutions were informed of and agreed to this study. We included patients older than 18 years who underwent primary cervical spine surgery from April 2017 to March 2022 at one of 13 participating institutions in the Greater Tokyo metropolitan area. Patients with tumor, infection, trauma, or a history of previous cervical spine surgery were excluded. Demographic information included age, sex, body mass index (BMI), American Society of Anesthesiologists physical status (ASA-PS) classification, diabetes mellitus, RA, smoking status, steroid use, and the number of affected cervical levels.
Variables and outcomes
Surgical details recorded included surgical procedure, upper cervical involvement, occipitocervical fusion, decompression range, fusion range, operation duration, and estimated blood loss. When numeric cranial and caudal cervical levels were available in the registry, the numbers of decompressed and fused levels were calculated. PROs were assessed preoperatively using questionnaires that measured neck, arm, and low back pain and hand and sole numbness on a Numerical Rating Scale (NRS), together with the 12-item Short Form physical component summary (SF-12 PCS) and mental component summary (SF-12 MCS), the EuroQOL 5-dimension (EQ-5D), and the Neck Disability Index (NDI). Complications within 30 days after surgery were documented and categorized into neurological damage, SSI, hematoma, implant-related complications, organ injury, respiratory issues, urinary tract infection, cardiovascular or cerebrovascular events, in-hospital mortality, and others. Patients were encouraged to attend a 1-year postoperative follow-up and complete the same PRO questionnaires.
Statistical analysis
All analyses were conducted using IBM SPSS Statistics version 25 (IBM Corp., Armonk, NY, USA) and Python statistical packages. Paired t-tests, Wilcoxon signed-rank tests, or Mann-Whitney U-tests were used for continuous variables as appropriate, and McNemar’s test or the Chi-squared test was used for categorical variables. A P value <0.05 was considered statistically significant. Because this was a retrospective analysis of a prospectively maintained registry, no formal a priori sample-size calculation was performed; all eligible patients during the study period were included.
Propensity scores for RA status were calculated using logistic regression with prespecified clinically relevant variables selected to reduce overfitting while accounting for major baseline and procedural differences: age, sex, BMI, ASA-PS classification, posterior fusion, anterior or combined fusion, and upper cervical involvement. Propensity-score overlap weighting was then applied as the primary adjusted analysis. Overlap weights were defined as 1 minus the propensity score for RA patients and as the propensity score for non-RA patients. Covariate balance before and after weighting was assessed using standardized mean differences (SMDs), with absolute SMDs <0.1 considered acceptable. Weighted group differences were reported with 95% confidence intervals (CIs).
For pain, numbness, and NDI, improvement was defined as the preoperative score minus the 1-year postoperative score. For EQ-5D and SF-12, improvement was defined as the 1-year postoperative score minus the preoperative score. Postoperative PRO analyses were performed using paired complete cases with available preoperative and 1-year questionnaires for each outcome. No imputation was performed for missing postoperative questionnaire data. As a sensitivity analysis, stabilized inverse probability weighting with weight truncation at the 1st and 99th percentiles was used for the change-score analyses.
Results
Baseline characteristics before weighting
The study cohort comprised 2,729 enrolled patients. RA was diagnosed in 67 individuals, accounting for 2.5% of cases (Table 1). Patients with RA were characterized by higher age, a predominance of women, lower BMI, higher ASA classification, and a greater prevalence of myelopathy. They also showed a lower frequency of ossification of the posterior longitudinal ligament and disc herniation, but a markedly higher incidence of upper cervical involvement, longer operative duration, and greater blood loss. The mean duration of RA was 21 years (range, 1–56 years), and the mean preoperative C-reactive protein level was 0.72 mg/dL (range, 0–3.1 mg/dL). Steroids were used in 56.7% of RA patients, methotrexate in 41.8%, and biologic agents in 19.4%. After overlap weighting, the variables included in the propensity-score model were well balanced between the groups (Table 1).
Table 1. Baseline characteristics before and after propensity-score overlap weighting.
| Variable | RA | RA OW | Non-RA | Non-RA OW | SMD before OW | SMD after OW |
|---|---|---|---|---|---|---|
| Age, years | 71.0 (9.8) | 71.0 (10.3) | 65.0 (13.4) | 71.0 (11.5) | 0.52 | 0.00 |
| Female sex | 53 (79.1%) | 75.8% | 797 (29.9%) | 75.8% | 0.99 | 0.00 |
| BMI, kg/m2 | 22.3 (4.0) | 22.5 (4.1) | 24.6 (5.2) | 22.5 (3.9) | −0.51 | 0.00 |
| ASA-PS classification | 2.2 (0.4) | 2.2 (0.4) | 1.9 (0.5) | 2.2 (0.5) | 0.52 | 0.00 |
| Current smoker | 4 (6.0%) | 6.7% | 385 (14.5%) | 6.1% | −0.28 | 0.02 |
| Diabetes mellitus | 13 (19.4%) | 21.0% | 561 (21.1%) | 19.2% | −0.04 | 0.05 |
| Steroid use | 38 (56.7%) | 54.7% | 54 (2.0%) | 2.5% | 1.20 | 1.16 |
| OPLL | 3 (4.5%) | 5.4% | 349 (13.1%) | 12.1% | −0.30 | −0.24 |
| Disc herniation | 2 (3.0%) | 3.6% | 350 (13.1%) | 3.2% | −0.37 | 0.02 |
| Deformity | 0 (0.0%) | 0.0% | 32 (1.2%) | 2.3% | −0.16 | −0.21 |
| Myelopathy | 60 (89.6%) | 89.3% | 2070 (77.8%) | 87.8% | 0.32 | 0.05 |
| Posterior decompression | 32 (47.8%) | 54.7% | 1781 (66.9%) | 54.7% | −0.39 | 0.00 |
| Posterior fusion | 33 (49.3%) | 41.7% | 342 (12.8%) | 41.7% | 0.79 | 0.00 |
| Anterior fusion | 2 (3.0%) | 3.6% | 528 (19.8%) | 3.4% | −0.53 | 0.01 |
| Combined anterior/posterior fusion | 0 (0.0%) | 0.0% | 11 (0.4%) | 0.2% | −0.09 | −0.06 |
| Upper cervical involvement | 26 (38.8%) | 29.4% | 87 (3.3%) | 29.4% | 0.87 | 0.00 |
| Endoscopic surgery | 5 (7.5%) | 8.8% | 308 (11.6%) | 6.2% | −0.14 | 0.10 |
| Surgical time, min | 183.1 (95.4) | 173.2 (92.0) | 149.6 (78.7) | 177.9 (92.8) | 0.38 | −0.05 |
| Estimated blood loss, g | 160.8 (214.4) | 158.4 (221.3) | 96.5 (160.4) | 129.0 (192.5) | 0.34 | 0.14 |
Overlap-weighted values are weighted mean (SD) or weighted percentages. ASA-PS, American Society of Anesthesiologists physical status; BMI, body mass index; OPLL, ossification of the posterior longitudinal ligament; OW, overlap weighting; RA, rheumatoid arthritis; SD, standard deviation; SMD, standardized mean difference.
Perioperative complications and surgical details
In the RA cohort, decompression was performed in 59 patients (88.1%), fusion in 35 patients (52.2%), upper cervical surgery in 26 patients (38.8%), and occipitocervical fusion in 9 patients (13.4%) (Table 2). Among patients with available numeric level data, the mean number of decompressed levels was 3.3 in RA patients and 3.6 in non-RA patients, whereas the mean number of fused levels was 1.3 and 0.8, respectively. Perioperative complications showed no major between-group differences except for urinary tract infection, which was more frequent in the RA cohort before adjustment (Table 3). In the overlap-weighted analysis, no complication category showed a clear adjusted increase in the RA group, although the small number of events limited precision.
Table 2. Surgical details.
| Surgical detail | RA | Non-RA |
|---|---|---|
| Decompression performed, n (%) | 59 (88.1) | 2,569 (96.5) |
| Fusion performed, n (%) | 35 (52.2) | 881 (33.1) |
| Upper cervical involvement, n (%) | 26 (38.8) | 87 (3.3) |
| Occipitocervical fusion, n (%) | 9 (13.4) | 20 (0.8) |
| Numeric decompression levels, mean (SD) | 3.3 (1.9); n=60 | 3.6 (1.5); n=2,483 |
| Numeric fusion levels, mean (SD) | 1.3 (1.7); n=55 | 0.8 (1.4); n=2,532 |
Numeric decompression and fusion levels were calculated only when both cranial and caudal numeric cervical levels were available in the registry. For cases without decompression or fusion, the level count was recorded as 0 when identifiable. RA, rheumatoid arthritis; SD, standard deviation.
Table 3. Perioperative complications.
| Complication | RA, n (%) | Non-RA, n (%) | OW RA (%) | OW non-RA (%) | OW difference, % (95% CI) | P value |
|---|---|---|---|---|---|---|
| Dural tear | 4 (6.0) | 92 (3.5) | 5.0 | 4.7 | 0.3 (−5.5 to 6.1) | 0.92 |
| Neurological damage | 1 (1.5) | 6 (0.2) | 1.7 | 0.2 | 1.6 (−1.9 to 5.0) | 0.37 |
| Surgical site infection | 1 (1.5) | 42 (1.6) | 1.4 | 2.3 | −1.0 (−4.4 to 2.5) | 0.59 |
| Hematoma | 0 (0.0) | 18 (0.7) | 0.0 | 0.6 | −0.6 (−1.1 to 0.0) | 0.07 |
| Reoperation | 1 (1.5) | 73 (2.7) | 1.4 | 4.9 | −3.5 (−7.6 to 0.5) | 0.09 |
| Urinary tract infection | 2 (3.0) | 10 (0.4) | 2.1 | 0.5 | 1.6 (−1.7 to 4.8) | 0.35 |
| Stroke | 0 (0.0) | 3 (0.1) | 0.0 | 0.1 | −0.1 (−0.3 to 0.0) | 0.15 |
| Respiratory complications | 1 (1.5) | 5 (0.2) | 1.8 | 1.3 | 0.5 (−3.6 to 4.5) | 0.83 |
| Serious adverse event | 1 (1.5) | 8 (0.3) | 1.8 | 1.4 | 0.3 (−3.7 to 4.4) | 0.87 |
Overlap-weighted risk differences are shown as RA minus non-RA. CI, confidence interval; OW, overlap weighting; RA, rheumatoid arthritis.
PROs
At the 1-year postoperative follow-up, PRO assessments were obtained from 1,983 respondents (72.7%), including 54 RA patients (2.7%). Patients with and without 1-year PRO data are compared in Table S1. In the overlap-weighted paired complete-case analysis, RA patients had poorer absolute EQ-5D and SF-12 PCS scores both before and after surgery (Table 4). However, RA patients showed improvement in several symptoms and quality-of-life measures. The improvement in EQ-5D was 0.14 in RA patients and 0.13 in non-RA patients [difference, 0.01; 95% confidence interval (CI): −0.06 to 0.09]. The improvement in SF-12 PCS was 4.2 and 6.2, respectively (difference, −2.0; 95% CI: −7.6 to 3.7). Neck pain and hand numbness also improved in RA patients. The NDI showed only limited improvement in RA patients (1.8 vs. 7.2; difference, −5.4; 95% CI: −14.9 to 4.2). Postoperative satisfaction was 69.1% in RA patients and 72.4% in non-RA patients (difference, −3.3%; 95% CI: −17.7% to 11.1%). Similar directions were observed in the stabilized inverse probability weighting sensitivity analysis (Table S2).
Table 4. Patient-reported outcomes before and after surgery in overlap-weighted paired complete-case analyses.
| Outcome | Time | RA (n) | Non-RA (n) | OW RA, mean (SD) | OW non-RA, mean (SD) | OW difference (RA − non-RA), 95% CI | P value |
|---|---|---|---|---|---|---|---|
| Neck pain | Preoperative | 52 | 1,863 | 2.8 (2.9) | 3.3 (3.0) | −0.5 (−1.4 to 0.4) | 0.30 |
| 1-year postoperative | 52 | 1,863 | 1.6 (2.2) | 2.3 (2.6) | −0.6 (−1.4 to 0.1) | 0.08 | |
| Change/improvement | 52 | 1,863 | 1.2 (3.0) | 1.0 (3.2) | 0.2 (−0.8 to 1.1) | 0.73 | |
| Arm pain | Preoperative | 53 | 1,855 | 3.8 (3.6) | 3.2 (3.2) | 0.6 (−0.4 to 1.7) | 0.25 |
| 1-year postoperative | 53 | 1,855 | 2.5 (3.0) | 2.3 (2.8) | 0.2 (−0.7 to 1.1) | 0.67 | |
| Change/improvement | 53 | 1,855 | 1.3 (3.5) | 0.9 (3.5) | 0.4 (−0.6 to 1.5) | 0.44 | |
| Low back pain | Preoperative | 53 | 1,866 | 3.6 (3.2) | 3.1 (3.1) | 0.6 (−0.4 to 1.5) | 0.26 |
| 1-year postoperative | 53 | 1,866 | 3.4 (3.3) | 2.6 (2.8) | 0.8 (−0.1 to 1.8) | 0.09 | |
| Change/improvement | 53 | 1,866 | 0.2 (3.2) | 0.5 (3.1) | −0.3 (−1.2 to 0.7) | 0.57 | |
| Leg pain | Preoperative | 52 | 1,860 | 3.9 (3.2) | 2.9 (3.1) | 1.0 (0.0 to 2.0) | 0.04 |
| 1-year postoperative | 52 | 1,860 | 2.9 (2.9) | 2.5 (2.8) | 0.4 (−0.5 to 1.2) | 0.43 | |
| Change/improvement | 52 | 1,860 | 1.1 (3.6) | 0.4 (3.5) | 0.7 (−0.5 to 1.8) | 0.25 | |
| Hand numbness | Preoperative | 52 | 1,905 | 6.3 (3.5) | 5.9 (3.2) | 0.4 (−0.6 to 1.4) | 0.45 |
| 1-year postoperative | 52 | 1,905 | 3.8 (3.3) | 3.7 (3.1) | 0.0 (−1.0 to 1.0) | 0.97 | |
| Change/improvement | 52 | 1,905 | 2.5 (3.9) | 2.2 (3.4) | 0.4 (−0.8 to 1.6) | 0.53 | |
| EQ-5D | Preoperative | 53 | 1,919 | 0.45 (0.26) | 0.54 (0.21) | −0.09 (−0.17 to −0.01) | 0.02 |
| 1-year postoperative | 53 | 1,919 | 0.59 (0.17) | 0.66 (0.17) | −0.08 (−0.12 to −0.03) | 0.002 | |
| Change/improvement | 53 | 1,919 | 0.14 (0.26) | 0.13 (0.21) | 0.01 (−0.06 to 0.09) | 0.73 | |
| SF-12 PCS | Preoperative | 49 | 1,797 | 21.1 (14.9) | 27.1 (16.1) | −6.0 (−10.6 to −1.3) | 0.01 |
| 1-year postoperative | 49 | 1,797 | 25.3 (15.6) | 33.3 (15.8) | −7.9 (−12.8 to −3.1) | 0.001 | |
| Change/improvement | 49 | 1,797 | 4.2 (18.4) | 6.2 (16.2) | −2.0 (−7.6 to 3.7) | 0.49 | |
| Preoperative | 49 | 1,797 | 49.5 (11.3) | 49.5 (9.7) | 0.0 (−3.4 to 3.5) | 0.98 | |
| 1-year postoperative | 49 | 1,797 | 53.7 (11.0) | 53.1 (9.6) | 0.6 (−2.8 to 4.0) | 0.74 | |
| Change/improvement | 49 | 1,797 | 4.2 (11.3) | 3.7 (11.4) | 0.5 (−3.0 to 4.1) | 0.76 | |
| NDI | Preoperative | 33 | 1,445 | 31.3 (18.8) | 35.9 (18.8) | −4.7 (−11.9 to 2.5) | 0.21 |
| 1-year postoperative | 33 | 1,445 | 29.4 (17.1) | 28.7 (19.4) | 0.7 (−5.9 to 7.2) | 0.84 | |
| Change/improvement | 33 | 1,445 | 1.8 (24.3) | 7.2 (26.9) | −5.4 (−14.9 to 4.2) | 0.27 | |
| Satisfaction | Satisfied at 1 year | 50 | 1,944 | 69.1% | 72.4% | −3.3 (−17.7 to 11.1) | 0.65 |
For pain, numbness, and NDI, improvement was defined as preoperative minus postoperative score. For EQ-5D and SF-12, improvement was defined as postoperative minus preoperative score. The same paired complete cases were used for preoperative, postoperative, and change values within each outcome. Because values are rounded, displayed change values may not exactly equal the simple difference between displayed preoperative and postoperative means. CI, confidence interval; EQ-5D, EuroQOL 5-dimension; NDI, Neck Disability Index; OW, overlap weighting; RA, rheumatoid arthritis; SD, standard deviation; SF-12 PCS, 12-item Short Form physical component summary.
Discussion
Our study’s demographic findings, particularly the higher mean age and female predominance in the RA group, align with broader epidemiological trends in RA (9). The elevated ASA classification in our RA cohort indicates a higher systemic disease burden, which were compatible with previous reports that have demonstrated an increased comorbidity among RA patients (11). This demographic pattern is crucial not only for understanding the progression of RA but also for customizing surgical interventions. Furthermore, we observed that surgeries involving the upper cervical spine were notably prevalent in RA patients, accounting for approximately 40% of cases, with fusion procedures comprising about half of these surgeries. More detailed surgical data showed that 88.1% of RA patients underwent decompression, 52.2% underwent fusion, and 13.4% underwent occipitocervical fusion. The predominance of older, lean females in the RA group, a demographic often associated with osteoporosis, necessitates heightened vigilance (12,13). This increased comorbidity complicates surgical outcomes and underscores the need for a more nuanced approach to perioperative care, which should include meticulous monitoring and management of comorbid conditions. Additionally, the use of biologic agents, as noted in our study, marks a significant advancement in RA management. For example, Kaito et al. indicated that biologics might positively influence the progression of cervical spine lesions in RA (14). Furthermore, Stein et al. noted an increase in surgeries for degenerative spine diseases unrelated to RA in well-controlled RA patients in the United States (15). A recent updated review also emphasized that cervical spine involvement remains clinically relevant in the modern treatment era, particularly once structural instability has developed (16). Consequently, it is plausible to anticipate a reduction in the frequency of complex surgeries in the upper cervical spine in our cohort in the foreseeable future.
The complication rates in our study, particularly the higher incidence of urinary tract infections in RA patients before adjustment, add to the existing literature on surgical risks in this population. Bernstein et al. reported higher incidences of urinary tract infection (UTI) and SSI in RA patients following spine fusion surgery (17). This may be attributed to the immunocompromised state often seen in these patients, exacerbated by the use of immunosuppressive therapies. In the overlap-weighted analysis, however, no complication category showed a clear adjusted increase in the RA group, although the number of events was small and the confidence intervals were wide. The high usage of steroids in our RA cohort is noteworthy, as steroids can impact surgical results as well as infection rates, although it remains unclear whether the high steroid usage in our cohort directly affected the incidence of cervical lesions (10). The notable utilization of methotrexate in our cohort of RA patients warrants careful consideration, highlighting the critical balance required between effective RA management and the minimization of surgical risks (18). Additionally, as previously mentioned, RA patients tend to have a predisposition towards osteoporosis, which becomes particularly problematic in lumbar spine surgeries or spinal deformity surgeries due to potential loosening or breakage of implants (12). While cervical spine surgeries may not be as affected by osteoporosis as thoracolumbar surgeries, the importance of paying attention to bone quality in surgeries involving implants and fixation cannot be overstated. The osteoporotic and immunosuppressive characteristics of RA patients add layers of complexity to surgical decision-making, requiring careful evaluation of bone quality and fusion strategies.
PROs in our study offer critical insights into the lived experiences of RA patients post-cervical spine surgery. Notably, RA patients reported lower scores in quality-of-life assessments, particularly in the EQ5D and the PCS of SF-12, both before and after surgery. This finding aligns with the systemic and pervasive nature of RA, where the disease’s impact extends beyond spinal pathology to affect multiple body systems and extremities. The comprehensive impairment of quality of life in RA patients is a multifaceted issue, influenced not only by the physical limitations and pain associated with spinal conditions but also by the broader systemic effects of RA, including fatigue, joint pain in other areas, and psychological distress. Nevertheless, selected RA patients showed improvement in several symptoms and quality-of-life measures after surgery, and postoperative satisfaction was not clearly lower than that of non-RA patients. The overlap-weighted analyses showed improvement in EQ-5D, SF-12 PCS, neck pain, and hand numbness in RA patients. Together, these findings support the view that improvement is still achievable in selected RA patients, although disability measured by the NDI may remain less responsive. These findings should not be interpreted as evidence of equivalence between RA and non-RA patients, because the number of RA patients was limited and the 95% confidence intervals were wide. While Suzuki et al. reported that 56% of RA patients experienced neck pain (19), our study found that, although there was no significant difference in the severity of neck pain between RA and non-RA patients preoperatively, a significant improvement in neck pain was observed in RA patients postoperatively. Specifically, 77% of RA patients who initially reported a NRS score of 4 or higher experienced an improvement of 2 or more points post-surgery (data not shown). This suggests that RA patients also experience postoperative improvement in neck pain. Conversely, minimal improvement was observed in the NDI among RA patients. One limitation of our study was the absence of radiographic evaluation, which might have provided insights into cervical spine mobility or alignment issues in RA patients, potentially explaining the lack of significant improvement in NDI scores. Additionally, the NDI includes activities like reading, driving, sleeping, and lifting (20), which could be challenging to improve in RA patients with extensive motor disabilities affecting limbs and the torso. Although quality of life and neck pain may improve following cervical spine surgery in RA patients, these findings suggest that there may be limitations in surgical treatment concerning cervical spine function.
The strength of this study lies in its investigation of PROs in RA patients undergoing cervical spine surgery, a relatively underexplored area, particularly in comparison with non-RA patients. While the number of RA patients in this study may not be deemed sufficient, considering that RA surgical patients accounted for 2.5% of all cervical spine surgeries in our cohort, the research is considered to hold significant value.
However, several limitations are present. First, the limited number of RA patients restricted our ability to perform a detailed examination of outcomes based on surgical procedures. The limited RA sample also reduced the precision of the estimated group differences, and the wide 95% confidence intervals preclude any conclusion of equivalence between RA and non-RA patients. The heterogeneous nature of the patient population is a concern, indicating the need for studies in more homogenous groups. Second, although we adjusted the backgrounds using propensity scores, it is essential to note that the impact on quality of life stemming from lumbar or limb conditions fundamentally differs between RA and non-RA patients, which could have influenced the surgical outcomes. In addition, approximately one quarter of patients did not have 1-year PRO data; complete-case analysis without imputation may therefore have introduced attrition bias, despite the responder analysis shown in Table S1. Although overlap weighting improved the balance of measured variables, unmeasured confounding remains possible. Third, the absence of radiographic data, as well as comparisons of bone density, should be noted as a limitation that may become significant in the long term. Fourth, the absence of specific RA disease activity metrics, such as a disease activity score (DAS), represents a significant gap in our analysis. Although preoperative CRP levels indicated adequate control of RA, the absence of one-year postoperative data could have influenced the outcomes. Studies exploring the impact of RA severity and duration on surgical outcomes are vital. Fifth, the 1-year follow-up period, especially in cases involving instrumentation surgeries, may not be adequate to fully assess long-term outcomes. Longer follow-up durations are essential to comprehensively evaluate the sustained effects and potential complications of these surgeries. Addressing the limitations identified in this study will necessitate conducting larger-scale, longitudinal research, incorporating a broader range of cases. Furthermore, as the use of biologic agents in RA management becomes more prevalent, it is reasonable to expect that these treatments will significantly influence the progression of cervical spine pathology, thereby impacting surgical decision-making and long-term patient outcomes. These developments highlight the need for continuous updates in surgical strategies and prognostic assessments to align with emerging therapeutic options.
Conclusions
Patients with RA had poorer PROs before and after cervical spine surgery than non-RA patients. Nevertheless, selected RA patients showed improvement in several symptoms and quality-of-life measures, and postoperative satisfaction was not clearly lower. Disability measured by the NDI may remain less responsive, and the limited number of RA patients and wide confidence intervals preclude any conclusion of equivalence with non-RA patients.
Supplementary
The article’s supplementary files as
Acknowledgments
The authors would like to thank all participating institutions and members of the University of Tokyo Spine Group (UTSG) for their contributions to this study.
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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the institutional review board of The University of Tokyo Hospital (No. 10335), and written informed consent was obtained from all participants. All participating institutions were informed of and agreed to this study.
Footnotes
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0149/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-2026-0149/coif). The authors have no conflicts of interest to declare.
Data Sharing Statement
Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0149/dss
References
- 1. Joaquim AF, Appenzeller S. Cervical spine involvement in rheumatoid arthritis--a systematic review. Autoimmun Rev 2014;13:1195-202.. 10.1016/j.autrev.2014.08.014 [DOI] [PubMed] [Google Scholar]
- 2. Yurube T, Sumi M, Nishida K, et al. Progression of cervical spine instabilities in rheumatoid arthritis: a prospective cohort study of outpatients over 5 years. Spine (Phila Pa 1976) 2011;36:647-53.. 10.1097/BRS.0b013e3181da21c5 [DOI] [PubMed] [Google Scholar]
- 3.Dreyer SJ, Boden SD. Natural history of rheumatoid arthritis of the cervical spine. Clin Orthop Relat Res 1999;(366):98-106. [DOI] [PubMed]
- 4. Shlobin NA, Dahdaleh NS. Cervical spine manifestations of rheumatoid arthritis: a review. Neurosurg Rev 2021;44:1957-65.. 10.1007/s10143-020-01412-1 [DOI] [PubMed] [Google Scholar]
- 5. Shen FH, Samartzis D, Jenis LG, et al. Rheumatoid arthritis: evaluation and surgical management of the cervical spine. Spine J 2004;4:689-700.. 10.1016/j.spinee.2004.05.001 [DOI] [PubMed] [Google Scholar]
- 6. Gillick JL, Wainwright J, Das K. Rheumatoid Arthritis and the Cervical Spine: A Review on the Role of Surgery. Int J Rheumatol 2015;2015:252456.. 10.1155/2015/252456 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Castro S, Verstraete K, Mielants H, et al. Cervical spine involvement in rheumatoid arthritis: a clinical, neurological and radiological evaluation. Clin Exp Rheumatol 1994;12:369-74. [PubMed] [Google Scholar]
- 8. Takenaka S, Kashii M, Iwasaki M, et al. Risk factor analysis of surgery-related complications in primary cervical spine surgery for degenerative diseases using a surgeon-maintained database. Bone Joint J 2021;103-B:157-63.. 10.1302/0301-620X.103B1.BJJ-2020-1226.R1 [DOI] [PubMed] [Google Scholar]
- 9. Streufert BD, Onyedimma C, Yolcu YU, et al. Rheumatoid Arthritis in Spine Surgery: A Systematic Review and Meta-Analysis. Global Spine J 2022;12:1583-95.. 10.1177/21925682211057543 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kato S, Nakamoto H, Matsubayashi Y, et al. Postoperative outcomes after degenerative lumbar spine surgery in rheumatoid arthritis patients -a propensity score-matched analysis. BMC Musculoskelet Disord 2022;23:380.. 10.1186/s12891-022-05326-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Wolfe F, Mitchell DM, Sibley JT, et al. The mortality of rheumatoid arthritis. Arthritis Rheum 1994;37:481-94.. 10.1002/art.1780370408 [DOI] [PubMed] [Google Scholar]
- 12. Seki S, Hirano N, Matsushita I, et al. Lumbar spine surgery in patients with rheumatoid arthritis (RA): what affects the outcomes?. Spine J 2018;18:99-106.. 10.1016/j.spinee.2017.06.030 [DOI] [PubMed] [Google Scholar]
- 13. Park JS, Shim KD, Song YS, et al. Risk factor analysis of adjacent segment disease requiring surgery after short lumbar fusion: the influence of rheumatoid arthritis. Spine J 2018;18:1578-83.. 10.1016/j.spinee.2018.02.005 [DOI] [PubMed] [Google Scholar]
- 14. Kaito T, Hosono N, Ohshima S, et al. Effect of biological agents on cervical spine lesions in rheumatoid arthritis. Spine (Phila Pa 1976) 2012;37:1742-6.. 10.1097/BRS.0b013e318256b584 [DOI] [PubMed] [Google Scholar]
- 15. Stein BE, Hassanzadeh H, Jain A, et al. Changing trends in cervical spine fusions in patients with rheumatoid arthritis. Spine (Phila Pa 1976) 2014;39:1178-82.. 10.1097/BRS.0000000000000376 [DOI] [PubMed] [Google Scholar]
- 16. Song J, Mai E, Zhang J, et al. Rheumatoid Arthritis in the Cervical Spine: An Updated Review of Epidemiology, Imaging, and Surgical Indications. Spine Surg Relat Res 2026;10:331-42.. 10.22603/ssrr.2025-0423 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Bernstein DN, Kurucan E, Menga EN, et al. Comparison of adult spinal deformity patients with and without rheumatoid arthritis undergoing primary non-cervical spinal fusion surgery: a nationwide analysis of 52,818 patients. Spine J 2018;18:1861-6. Erratum in: Spine J 2020;20:2048.. 10.1016/j.spinee.2018.03.020 [DOI] [PubMed] [Google Scholar]
- 18. Loza E, Martinez-Lopez JA, Carmona L. A systematic review on the optimum management of the use of methotrexate in rheumatoid arthritis patients in the perioperative period to minimize perioperative morbidity and maintain disease control. Clin Exp Rheumatol 2009;27:856-62. [PubMed] [Google Scholar]
- 19. Suzuki A, Yamada K, Yabu A, et al. Neck pain and related factors in patients with rheumatoid arthritis. Mod Rheumatol 2023;33:503-8.. 10.1093/mr/roac049 [DOI] [PubMed] [Google Scholar]
- 20. Vernon H, Mior S. The Neck Disability Index: a study of reliability and validity. J Manipulative Physiol Ther 1991;14:409-15. Erratum in: J Manipulative Physiol Ther 1992;15:followi. [PubMed] [Google Scholar]