The effect of sagittal alignment, coronal balance, and segmental stability on preoperative patient-reported outcomes in patients with degenerative lumbar spondylolisthesis

After approval by the ethics committee in our hospital, a retrospective review of 247 consecutive patients who were diagnosed with DLS between January 2019 and December 2021 was performed. All patients were recruited from outpatient spinal surgery clinics and scheduled to undergo spine surgery. 3D printing molds are used for preoperative planning and surgical simulation.

The inclusion requirements included: (1) L4 DLS; (2) Meyerding grade I or II; and (3) complete radiographic data, including whole-spine anteroposterior and lateral radiographs and lateral flexion and extension X-ray of the lumbar spine.

The exclusion criteria were as follows: (1) multilevel DLS; (2) isthmic spondylolisthesis; (3) history of prior lumbar trauma, tumor, infection, cauda equina syndrome, or revision surgery; and (4) surgery requiring more than two-level instrumentation and fusion.

Demographic data were collected using electronic medical record reviews, including age, gender, height, weight, and body mass index. Overall, 101 patients were ultimately enrolled in the present study.

The ODI, Japanese Orthopedic Association (JOA) score, VAS-back pain, and VAS-leg pain were examined for all patients at the time of admission by the same experienced research assistant. The validated ODI is a self-administered questionnaire for evaluating back-specific functional disability, consisting of 10 items with scores from 0 to 5 [21]. The value of ODI is less than 20% in the asymptomatic population [22]. The JOA score is a comprehensive assessment method that comprises the rating of subjective symptoms (0–9 scores), clinical signs (0–6 scores), restriction of activities of daily living (0–14 scores), and urinary bladder function (− 6–0 scores) [23]. The JOA score is 29 in ordinary people [23]. The VAS was used to measure back and leg pain for patients based on a 100 mm line, with “painless” (0) and “most severe pain” (100 mm) at each respective end [24].

All data are presented as the mean values ± standard deviation. Normality was tested by the Shapiro‒Wilk Normality test (Additional file 1, Table S1). Univariate and multivariate linear regression analyses were conducted to identify factors relevant to PROs. Categorical variables were represented as sets of dummy variables. Variables with P < 0.05 on univariate analysis were included in the subsequent multivariate linear regression analysis. Continuous variables were compared among subgroups using one-way ANOVA and the Kruskal‒Wallis test with Bonferroni or Tamhanes T2 post hoc analysis. The Chi-square test was used to compare categorical variables among groups. Intraclass coefficients and their 95% confidence intervals for all parameters were calculated to evaluate the reliability of the intrinsic variability of radiographic measurements. Values of ICC less than 0.5, between 0.5 and 0.75, between 0.75 and 0.9, and greater than 0.90 are indicative of poor, moderate, good, and excellent reliability, respectively.

In total, 80 (79.2%) females and 21 (20.8%) males were enrolled in the present study, with a mean age of 66.34 ± 8.33 years. All baseline values of the whole cohort are summarized in Table 1. Forty-two of 101 (41.6%) patients were characterized by stable spondylolisthesis, while 59 (58.4%) patients presented unstable spondylolisthesis. Moreover, 49, 38, and 14 patients were classified as coronal balance, LCI, and GCI states, respectively. The vast majority of patients (83.2%) showed a spondylolisthesis of Meyerding grade I, whereas only 16.8% of patients exhibited SP over 25%. All measurements presented high reproducibility, as the values of ICC > 0.85. For the whole cohort, the mean values of ODI and JOA score were 57.84% and 15.96, respectively. The VAS scores of back pain and leg pain reached approximately 62 mm and 65 mm, respectively.

Univariate linear regression analysis was conducted to identify the indicators related to the preoperative PROs. ODI was associated with age, LL, SVA, SLL, and GCI. Age, LL, PT, SVA, SLL, and GCI were related to the JOA score (Table 2). VAS-back pain was correlated with LL, PT, PT/PI, SVA, SLL, unstable spondylolisthesis, LCI, and GCI. For VAS-leg pain, age, LL, SS, PT, PT/PI, SVA, SLL, LCI, and GCI were influential factors (Table 3).

The results of the multiple linear regression are presented in Table 4. Increasing age (B 0.173; 95% CI [0.053 to 0.293]; P = 0.005), higher SVA (B 0.084; 95% CI [0.055 to 0.114]; P < 0.001), and GCI (B 3.610; 95% CI [0.519 to 6.702]; P = 0.023) were independent risk factors for superior preoperative ODI. Patients with GCI (B − 1.703; 95% CI [− 2.686 to − 0.720]; P = 0.001) had lower preoperative JOA scores than those with balanced coronal alignment. In terms of preoperative VAS-back pain, unstable spondylolisthesis (B 4.582; 95% CI [2.758 to 6.406]; P < 0.001) and GCI (B 4.108; 95% CI [1.032 to 7.185]; P = 0.009) were two vital predictors. Additionally, increasing age (B 0.160; 95% CI [0.015 to 0.305]; P = 0.031), LCI (B 6.065; 95% CI [3.502 to 8.629]; P < 0.001), and GCI (B 8.477; 95% CI [4.608 to 12.346]; P < 0.001) were associated with higher preoperative VAS-leg pain.

To better understand the characteristics of patients with different coronal alignments, we compared the demographic, radiographic, and clinical parameters among groups (Table 5). Demographic parameters, PI, and the constituent ratio of unstable spondylolisthesis were approximate in all three groups. The GCI group exhibited the highest SVA (P = 0.003) with the lowest LL and SLL (P = 0.001, 0.002, respectively). In addition, PT and PT/PI increased, while SS decreased significantly (P < 0.001) from the balance group to the GCI group, although there was no significant difference between the LCI and GCI groups based on post hoc tests. Furthermore, outcomes of clinical assessment were distinctly worse in the GCI group than in the other groups, including the highest ODI (P = 0.005), VAS-back pain (P < 0.001), and VAS-leg pain (P < 0.001), as well as the lowest JOA score (P = 0.006). Representative cases of the three different groups are illustrated in Fig. 4.

Spinal sagittal alignment is an essential consideration for the comprehensive assessment of the physiopathology, classification, and treatment options of DLS, but there remains no consensus regarding the relevance between sagittal alignment and clinical symptoms [10, 11]. Multiple sagittal indicators, including LL, SLL, PT, PT/PI, and SVA, were observed to be associated with preoperative PROs in the present study (Tables 2, 3). Similar to the findings of our research, Gille et al. presented that lower LL and higher SVA were associated with worse ODI and short form-12 questionnaire scores in a retrospective study containing 166 patients suffering from DLS [12]. Moreover, SVA was the only sagittal parameter that was an independent influencing factor of preoperative PRO based on our results (Table 4), which highlights the vital implication of global sagittal balance in the evaluation and therapy of DLS. As suggested by previous studies, the sagittal imbalance subtype had the most severe symptoms, and the correction of sagittal deformity to reach a favorable global sagittal alignment was the primary surgical target of such patients [11, 12]. However, in another prospective cohort study of 320 patients diagnosed with DLS, Karim et al. proposed that sagittal malalignment, such as SVA > 50 mm or PI-LL > 10°, did not appear to be correlated with baseline numeric rating scale (NRS) back pain, NRS-leg pain, or ODI. This result might be due to the fact that up to 77% of patients in their cohort had a major complaint of neurogenic claudication with no persistent pain. Thus, more comprehensive studies with larger cohorts are still needed to further elucidate how sagittal alignment affects PROs, although the correlation between higher SVA and worse PROs has been presented in this study.

If the degeneration of the intervertebral disc and incompetence of the facet joints evolve asymmetrically, the odds of a combination of coronal malalignment will increase significantly in patients with DLS [26]. Approximately 50% (52/101) of patients showed LCI, and 13.9% (14/101) of patients were simultaneously accompanied by GCI based on our results (Table 1). In contrast, 31.7% and 19.8% of patients had LCI in the retrospective studies of Takahashi et al. and Pan et al., respectively [14, 18]. Higher age was a potential reason for this disparity, as patients with LCI in our cohort were 5–10 years older than patients in the research mentioned above. Moreover, patients with LCI or GCI presented a prominent loss of LL, significant increases in PT/PI and SVA, and worse PROs compared with patients with balanced coronal alignment (Table 5), which was consistent with previous results [14, 18]. Therefore, although LCI and GCI were found to be independent influencing factors of most PROs according to the results of multiple linear regression analysis (Table 4), it cannot be ignored that patients with coronal imbalance were usually accompanied by severe sagittal malalignment and a significant increase in energy expenditure for pelvic compensation; this may have contributed to a higher severity stratification of preoperative symptoms.

The presence of segmental instability is also common in patients with DLS. As reported by Kirkaldy-Willis, the accumulation of microscopic degenerative damage caused decreased intervertebral disc height and subluxation of the facet joints, which finally led to increased intervertebral instability [27]. Compared with the ratios ranging from 31 to 40% reported by previous studies [28‐30], the proportion of unstable spondylolisthesis reached 58.4% in our study (Table 1). These discrepancies might be related to age and racial/ethnic composition. Another issue that merits concern is that the abnormal movement of the diseased segment might contribute to back pain in patients with DLS [31]. In a prospective study of 880 patients with low back pain, Kanemura et al. found that patients with dynamic instability at the L4/5 level exhibited more severe pain than those with stable segments [32]. Moreover, ST change > 3 mm was reported to have the most substantial effect on symptoms, followed by IDA change > 10° [32]. Our observations agree with the previous conclusion that unstable spondylolisthesis is an independent risk factor for higher preoperative VAS-back pain (Table 4). Consistently, Simmonds et al. demonstrated that the degree of back pain aggravated with dynamic stability deteriorated in a qualitative systematic review, and they suggested that decompression and 360° fusion surgery were necessary for patients with unstable DLS [13].

Last but not least, we observed that increasing age was closely associated with higher ODI (Table 4). The patients with GCI were also older than the other groups, although the difference was not significant (Table 5). As previously demonstrated, DLS is the consequence of long-term intersegmental instability at the lumbar motion segment and its predominance in patients older than 50 years [26]. Indeed, the increase in the degree of spinal degeneration with age is another important factor leading to the progression of symptom severity. In summary, DLS patients with increasing age, higher SVA, unstable spondylolisthesis, LCI, or GCI were prone to more severe preoperative subjective symptoms.

There still exist some limitations in our study. First, this was a retrospective study with potential bias. All clinical parameters were evaluated by the same professional research assistant on the patient’s admission. Standardization in the radiographic parameter acquisition was attempted by conducting comprehensive training before the study and illustrating each measurement on the statistical tables. Second, data discussed in the present study were concentrated on patients who were treated by surgical procedure, and there might have been a limited reference value for patients who elected to have conservative treatment. Third, further investigation is still needed to ascertain whether postoperative radiological and clinical outcomes are affected by various preoperative spinal alignments among patients with DLS, and research in this direction is ongoing.