Progressive population aging has led to an increased prevalence of lumbar degenerative diseases among older people. Posterior lumbar interbody fusion (PLIF) is widely used to treat degenerative diseases. However, several studies have demonstrated a high incidence of adjacent segment degeneration (ASD) following PLIF [1–5].

ASD can be broadly categorized into two types: radiographic and symptomatic. Over time, radiographic ASD may progress to symptomatic ASD, necessitating reoperation. Reoperations do not consistently yield favorable outcomes and can reduce patient satisfaction. Therefore, it is imperative to reduce the incidence of ASD after lumbar fusion. Previous studies have investigated the risk factors for ASD. A meta-analysis by Wang and Ding [6] identified several factors associated with ASD development, including age, body mass index (BMI), preoperative adjacent disc degeneration, long-segment fusion, fusion method, and segmental sagittal alignment. However, in the study by Zhang et al. [7], age, sex, segmental sagittal alignment, fusion method, or instrumentation were not found to increase the risk of ASD. Many factors can contribute to the risk of ASD but there is a lack of clear consensus. Several studies have focused on the sagittal plane balance of the spine, aiming to establish a correlation between sagittal plane balance and ASD. However, most of these studies primarily focused on analyzing the development of ASD within the lumbosacral and local segmental environment, and there is a paucity of studies examining the risks of ASD in relation to the overall sagittal balance of the spine [8–11].

Inflection vertebra (IV) refers to the transitional vertebra between regions of kyphosis and lordosis. Some studies have shown that the location of the IV may affect the incidental vertebral fractures [12] and the development of proximal junctional kyphosis after thoracopelvic corrective fusion for adult spinal deformities [13]. Even post-PLIF short fusion changes in the sagittal proportion of the spine may alter the geometric configuration and mechanics, especially in IV. Thus, a comprehensive exploration of ASD risk factors necessitates a study of the sagittal plane configuration of the entire spine. The objective of this study was to investigate the sagittal plane configuration of the entire spine and its association with the risk of ASD after PLIF for lumbar degenerative disease.

Of 211 individuals considered for enrollment, 99 were excluded (32, dropouts; 25, lack of follow-up whole-spine radiographs; 21, without an inflection point; 10, reoperation not due to ASD; seven, scoliosis with Cobb angle >20º; four, vertebral fracture with severe wedge-shaped vertebra). Finally, 112 patients who underwent PLIF for lumbar degenerative diseases were included in this study. The minimum follow-up period was 2 years, with an average follow-up time of 63.6 months.

Among the 112 enrolled patients, 52 (46.4%) were classified into the ASD group and 13 (25%) of these patients required reoperation due to worsening of pain and numbness and failure of conservative treatment. At the last follow-up, ASD was observed in 39 patients at the cranial segment, in 10 patients at the caudal segment, and both cranial and caudal segments in three patients. The average fusion length was 1.8 levels, with 52 patients (46.4%) having undergone one-level fusion (non-ASD:ASD, 30:22), 36 patients (32.1%) having undergone two-level fusion (non-ASD:ASD, 19:17), 18 patients (16.1%) having undergone three-level fusion (non-ASD:ASD, 8:10), and six patients (5.4%) having undergone four-level fusion (non-ASD:ASD, 3:3). Forty-one patients (36.6%) had undergone sacral fusion (non-ASD:ASD, 24:17) and 71 patients (63.4%) did not undergo sacral fusion (non-ASD:ASD, 36:35). Thirteen out of 52 patients (25%) in the ASD group underwent revision surgery and seven of these patients (7/13, 53.8%) had sacral fusion. Although we found no evidence of an increased incidence of ASD with longer segment fusions (one-level versus two-level versus three-level versus four-level, p=0.811) or sacral fusion (sacral fusion versus no sacral fusion, p=0.423), in the ASD group, patients with sacral fusion showed a greater tendency for requiring reoperation, which approached statistical significance (p=0.06).

The mean age at the surgery showed no association with the occurrence of ASD (p=0.364). The incidence of ASD was significantly higher in men than in women (p=0.041), and the ASD group had a higher mean BMI compared with the non-ASD group (p=0.036) (Table 1).

In spinal parameters, the ASD group showed a more anterior C7SVA in the preoperative phase (p=0.046) and a significantly larger TK (p=0.023), compared with the non-ASD group at the last follow-up (Table 2).

The spinal proportion was not significantly different between the ASD and non-ASD groups in terms of the TAV and TASVA, except for a more posterior TASVA (p=0.008) at the last follow-up. The location of IV was significantly more caudally and posteriorly in the ASD group, both immediately postoperation (p=0.001, p=0.045) and at the last postoperative follow-up (p=0.036, p=0.004). Additionally, the LAV was significantly more caudal in the ASD group immediately (p=0.007) after surgery, but there was no significant difference in terms of the LASVA (Table 3, Fig. 2).

Factors associated with ASD in univariate analysis were included in binary logistic regression analysis. Only immediately postoperative IV was identified as a significant risk factor for ASD (p=0.047) (Table 4). The receiver operating characteristic curve of IV to predict ASD showed an optimal cut-off of 5.2 (p=0.003). The incidence of ASD was significantly higher in the group having IV ≤5 (L1 vertebral body) than in the group having IV ≥5.5 (T12–L1 disc) (69.0% versus 38.6%, p=0.005) (Figs. 3, 4).

The risk factors for ASD in previous studies can be broadly divided into two categories: characteristics of the patient and characteristics of the operation. The main risk factors among patient characteristics include age >50 years, BMI ≥25.0 kg/m², postmenopausal women, and poor bone mineral density [15–17]. The risk factors related to the operation characteristics include the surgical approach, disrupted integrity of the posterior complex, length of fusion, rigid instrumentation, sacrum fusion, and pre-existing disc degeneration [18–21]. Although there is no clear consensus regarding some risk factors for ASD [6,7], most studies have demonstrated a high incidence of ASD. Therefore, it is imperative to analyze the causes of ASD from various perspectives.

Changes in the sagittal spinal proportion may alter the geometrical configuration, affecting the mechanics. Thus, a comprehensive exploration of ASD occurrence necessitates a study of the sagittal plane configuration of the entire spine. Yamato et al. [12] and Jakinapally et al. [13] showed that the original sagittal spine geometry differs between males and females with incidental vertebral fractures; they also underscored the significant influence of the immediately postoperative IV cranial to T12 as a significant risk factor for the occurrence of proximal junctional kyphosis (PJK) following corrective fusion surgery for adult spinal deformities. However, in the present study, the IV was found to be located significantly caudally and posteriorly, and the LAV was located even more significantly caudally in the ASD group in the immediate postoperative period (Fig. 3). This result is completely different compared to that of PJK in adult patients with spinal deformity. We analyzed the differences attributed to the occurrence of PJK in patients who underwent long-segment fusion from the lumbar to the thoracic region for adult spinal deformities. In these cases, the morphology of the sagittal plane changed above the fusion segment, primarily within the thoracic region. The IV and LAV were predetermined and remained unchanged after surgery. However, in patients with ASD who underwent only lumbar fusion, the local lumbar sagittal plane was adjusted, and the whole-spine proportion may have changed. In the ASD group, the IV in the immediate postoperative period was located caudally and posteriorly, and the LAV was located caudally, indicating that LL and TK were incongruent, and the fused lumbar segments could not adjust to increased TK. This restricted the mobility of the distal segments, leading to a proximal transfer of stress and acceleration of the onset of ASD. This also explains the development of ASD in the cranial segment in 75% (39/52) of the patients. ASD is more likely to occur in the cranial direction, as corroborated by many previous studies [22,23]. After PLIF, the instrumented segment will generate nonphysiological loads, inevitably altering the geometric parameters, and resulting in changes in the proportions of the spine. However, in the same operation, we have the option of reducing stress concentrations and preventing the possible occurrence of ASD by changing the spinal proportions. Therefore, it is necessary to identify risk factors for ASD that are indications for such modifications.

Matsumoto et al. [11] found that PI–LL mismatch was significantly associated with ASD. However, there was no significant difference in the incidence of ASD based on the PI–LL mismatch in our study. Both preoperative and immediately postoperative, the ASD group and the non-ASD groups were in a state of “moderate” according to the Scoliosis Research Society (SRS)-Schwab classification [24]. The small variation in the PI–LL mismatch was the likely reason for the lack of significant difference.

The Roussouly classification categorizes spine function and spinal alignment into five types. It helps identify areas of high local stress in the spine. Type 1 and 2 are characterized by much lower LL or flatter back, leading to a greater impact on the intervertebral discs, increasing the risk of lumbar disc herniation. Type 4 has much higher and bigger LL, with greater impact on the posterior part (especially the facet joints), making it vulnerable to spondylolisthesis. Significant complaints are rarely seen in patients classified as type 3 and type 3-anterverted [25]. Bari et al. [26] reported that the ideal postoperative Roussouly classification (restored) using the Roussouly flowchart was associated with a lower incidence of mechanical complications compared to non-restored in adult patients with spinal deformity. However, only short-segment fixation of the lumbar spine is unlikely to affect the whole sagittal plane of the spine. To avoid local stress concentration above the fusion segment, it is imperative to shape a larger LL and a higher IV. Although there were significant differences between the two groups in terms of sex, BMI, preoperative SVA, IV, LAV, and immediate postoperative IV, only immediate postoperative IV was identified as a risk factor (p=0.047). The incidence of ASD was significantly higher in the group where IV was lower than L1 (69.0% versus 38.6%, p=0.005).

The IV position cannot be directly manipulated during lumbar fusion surgery, but it is important to recognize that the spine functions as a unified entity; IV positioning can be indirectly influenced by altering the LL. Several authors also have emphasized that obtaining appropriate LL and maintaining a postoperative segmental lordotic angle of at least 20º is an effective method for preventing ASD [9,27]. The results of this study suggest the use of the bending rod technique with implantation of a suitable fusion cage to restore the LAV to level ≥2.5 (the L3–L4 disc) to bring the IV location to at least that of the T12–L1 disc, thereby decreasing the junctional stress at the upper instrumented vertebra and thus the incidence of ASD.

Some limitations of this study should be acknowledged. First, the study focused on radiographic changes, and the clinical outcomes of ASD were not investigated, potentially overlooking patients with clinical symptoms but no radiographic changes. Second, only the correlation between ASD and spinal proportion was investigated; further research is required to investigate how the location of the LAV affects the IV and changes the spinal proportion to decrease ASD. Additionally, ASD may result from the combined influence of multiple factors yet to be determined.

Spinal sagittal geometric mechanics and spinal proportions can interact. These factors should be considered to develop a more appropriate preoperative plan to prevent the occurrence of ASD. Our study indicated that a low IV level is a significant risk factor for ASD. Although it is difficult to control IV levels intraoperatively, we noted a high risk of ASD in patients with an IV location lower than T12–L1.