The etiology of adolescent idiopathic scoliosis (AIS), which affects 1%–3% of adolescents, remains unknown [1]. The incidence of Lenke type 1 scoliosis accounts for 51% of AIS [2]. Selected thoracic fusions (STFs) are performed on these patients to achieve coronal and sagittal balance and to preserve more lumbar motion segments. Postoperative distal adding-on is one of the complications after STF in Lenke 1 AIS.

Postoperative distal adding-on is defined as a progressive increase in the number of vertebrae included distally in the primary curvature, combined with a deviation of <5 mm from the center sacral vertical line (CSVL) to the lower instrumented vertebrae (LIV) or angulation of the first disk caudal to the LIV of more than 5° [3]. The occurrence of postoperative distal adding-on is closely related to decreased postoperative satisfaction, back pain, increased cost, and reoperation [4,5].

In 2017, the concept of fusion mass shift (FMS) was proposed, with an intraoperative aim of balanced fusion mass with FMS of <20 mm to avoid postoperative distal adding-on [3]. Therefore, we retrospectively reviewed the Lenke 1 AIS cases from the last decade to validate the influence of FMS on the occurrence of postoperative distal adding-on and to investigate the risk factors for postoperative distal adding-on in Lenke 1 AIS.

This is a retrospective study of patients from a single institution. Permission to conduct this retrospective study was obtained from the ethics committee of Shandong Provincial Hospital Affiliated to Shandong First Medical University (IRB approval no., NSFC-2020-528), and written informed consent was obtained from all subjects who participated in this study.

The study included 78 patients with Lenke 1 AIS who underwent selective thoracic fusion surgery between 2008 and 2019. Of them, 60 patients were included in the study, and 18 were excluded (Fig. 1). They all underwent surgical treatment by the same surgeon in a university-affiliated hospital. The fusion levels for selective thoracic fusion in Lenke 1 AIS were assessed using the fulcrum bending radiograph, a method described by Luk et al. [6].

The inclusion criteria of this study are as follows: (1) patients aged between 10 and 18 years, (2) patients with AIS who have main thoracic (MT) scoliosis, (3) the surgical procedure is selective thoracic fusion, (4) fixation with pedicle screw system, and (5) patients with a minimum of 2 years of follow-up.

The exclusion criteria of this study are as follows: (1) patients with other types of AIS, (2) patients with other types of scoliosis, (3) patients with a prior history of spinal surgery, (4) patients with a history of spinal infection, and (5) patients with postoperative fusion mass Cobb (FMC) angle of more than 25°.

Coronal spinal alignment parameters on standing anteroposterior (AP) plain radiographs were analyzed preoperatively, postoperatively, and at the final follow-up according to the established positioning protocol. The coronal spinal alignment parameters included the proximal thoracic (PT) curve, MT curve, thoracolumbar/lumbar (TL/L) curve, clavicle angle (CA), thoracic trunk shift (TTS), T1 tilt, LIV rotation, FMC angle (Cobb angle between the superior endplate of the upper instrumented vertebrae and the inferior endplate of LIV), FMS (distance from the center of the superior endplate of upper instrumented vertebrae to a perpendicular line of the inferior endplate of LIV erected from the center of the LIV), FMSA (angle between a line from the center of the superior endplate of upper instrumented vertebrae to the center of the inferior endplate of LIV and a perpendicular line of the inferior endplate of LIV) (Fig. 2). FMC, FMS, and FMSA were measured on the postoperative films. In CA and T1 tilt, the left higher was positive, and the right higher was negative. Furthermore, age, sex, triradiated cartilage, and Risser sign were recorded before surgery. The Scoliosis Research Society 22-item patient questionnaire (SRS-22) was used to evaluate clinical improvement preoperatively, postoperatively, and at the final follow-up [7].

The distal adding-on phenomenon was assessed by comparing postoperative standing AP radiographs with final follow-up standing AP radiographs. The postoperative FMS was divided into two groups: the balanced group (FMS ≤20 mm) and the unbalanced group (FMS >20 mm).

Statistical analysis was analyzed using the IBM SPSS ver. 27.0 software (IBM Corp., Armonk, NY, USA). An independent t-test was used to compare quantitative data between groups, and a chi-square test was used for qualitative data. In addition, binary logistic regression and receiver operating characteristics (ROC) curve analyses were used to identify the risk factors for postoperative distal adding-on in Lenke 1 AIS. The distribution of parameters was presented as mean and standard deviation. The significance level was set to p<0.05.

This study included 60 patients (13 males and 47 females). The mean age for surgery was 13.6 years old, with a 2-year follow-up. Twenty-three patients developed complications of distal adding-on at 2-year follow-up. Significant differences in preoperative parameters were identified between no adding-on and adding-on patients in trunk shift (13.84±14.71 mm versus 20.51±10.86 mm) and LIV rotation (0.57±0.69 grade versus 1.09±0.79 grade). In addition, significant differences in parameters were observed postoperatively between no adding-on and adding-on groups in the TL/L curve (9.97°±9.75° versus 14.9°±8.25°), LIV rotation (0.35±0.59 grade versus 1.13±0.81 grade), FMS (12.13±9.27 mm versus 27.15±14.65 mm), and FMSA (2.73°±2.1° versus 6.38°±3.48°). At the final follow-up, there were significant differences in parameters between no adding-on and adding-on groups in the TL/L curve (10.63°±8.52° versus 15.75°±10.49°) and LIV rotation (0.35±0.59 versus 1.09±0.79) (Table 1).

Age, gender, Risser sign, and preoperative and postoperative coronal spinal alignment parameters did not differ between balanced and unbalanced groups. On the other hand, patients in the unbalanced group had significantly lower mental health scores than those in the balanced group (Tables 2, 3). In addition, the TL/L curve showed significant differences between balanced and unbalanced groups at 2-year follow-up (10±8 versus 16±10), which was correlated with the incidence rate of the distal adding-on phenomenon.

Significant differences in FMC (11°±6° versus 19°±9°), FMS (11±6 mm versus 32±9 mm), and FMSA (2°±2° versus 7°±3°) were observed between balanced and unbalanced groups. Even though there were no significant differences in preoperative and postoperative LIV rotation between groups, postoperative LIV rotation was higher in the unbalanced group than preoperative LIV rotation (1.0±0.9 grade versus 0.7±0.9 grade). In addition, the unbalanced group was more likely to develop distal adding-on at 2-year follow-up (17 of 24 patients) than the balanced group (six of 36 patients) (Table 2).

Binary logistic regression analysis revealed three independent risk factors for distal adding-on incidence after surgery: FMS (odds ratio [OR], 1.115; 95% confidence interval [CI], 1.049–1.185; p<0.001), FMSA (OR, 1.590; 95% CI, 1.225–2.064; p<0.001), and postoperative LIV rotation (OR, 6.581; 95% CI, 2.280–19.000; p<0.001) (Table 4). ROC curve analysis confirmed the following risk factors: area under the ROC curve (AUC) of FMS was 0.831 (p<0.001) with an optimal cutoff value of 20°, AUC of FMSA was 0.856 (p<0.001) with an optimal cutoff value at 4.5°, and AUC of postoperative LIV rotation was 0.812 (p=0.031) with an optimal cutoff value of 1 (Fig. 3).

Postoperative distal curve adding-on is a complication of AIS undergoing selective thoracic fusion surgery. In our study, the incidence of postoperative distal adding-on was 38.3%, similar to the previous study (21%–51%) [3,8]. Previous studies reported that a low Risser grade predicted a high growth potential and was more likely to develop distal adding-on after selected thoracic fusion surgery [9,10]. The selection of LIV was another important factor influencing the incidence of postoperative distal adding-on [8,11]. According to He et al. [12], preoperative rotation of LIV was also an independent predictor of postoperative distal adding-on in Lenke 1A or 2A AIS patients.

In 2017, the concept of balanced fusion mass was proposed [3], which helped to reduce postoperative distal adding-on. In the present study, a distal adding-on phenomenon was observed in 16.7% (six of 36 patients) of the balanced group (FMS ≤20 mm) and 70.8% (17 of 24 patients) of the unbalanced group (FMS >20 mm). This was consistent with previous results, which showed that 12.2% (five of 41 patients) were in the balanced group, and 54.5% (six of 11 patients) were in the unbalanced group [3]. Therefore, the aim of corrective surgery should be to achieve a balanced fusion mass. This should be done intraoperatively to reduce the risk of postoperative distal adding-on. FMS, FMSA, and rotation of LIV are parameters that determine whether a balanced fusion mass is achieved.

Aside from FMS, FMS angle was found to be an independent risk factor for postoperative distal adding-on. FMS could help us preoperatively determine fusion levels through fulcrum bending radiographs [6]. We can also use FMS and FMSA during the surgery to assess whether the balanced fusion mass is achieved. However, intraoperative fluoroscopy was typically performed using a C-arm X-ray machine, and measuring the FMS without scale was challenging. FMSA, which is highly correlated with FMS, is easier to be measured in fluorography without scale. Furthermore, intraoperative FMS is not weight-bearing, and FMSA may be more sensitive than FMS in short fusion in Lenke 1 AIS patients undergoing selected thoracic fusion surgery. In our study, avoiding a residual FMSA of more than 4.5° could help reduce the incidence of the postoperative distal adding-on phenomenon.

FMS occurred before the postoperative distal adding-on phenomenon. It occurred during the operation at the completion of the fixation and remained unchanged after the operation. However, the postoperative distal adding-on phenomenon gradually developed at serial postoperative follow-up X-rays (Fig. 4). In our study, an FMS of >20 mm was one of the causes of the postoperative distal adding-on phenomenon.

FMS can be assessed intraoperatively using a long cross-metal bar perpendicular to the lower endplate of LIV. It facilitates the assessment of the fusion mass. However, because this assessment is performed in a nonweight-bearing position, the unfused segments can further compensate postoperatively when the patient assumes an erect posture. Therefore, a postoperative adding-on phenomenon will occur in the presence of a large residual FMS or FMSA.

The postoperative LIV rotation was also an important risk factor contributing to postoperative distal adding-on in Lenke 1 AIS patients. The residual rotation could cause a distal adjacent segment to be offset. The concept of neutral vertebrae was proposed by Suk et al. [13], who suggested that it should be fused to neutral vertebrae when neutral vertebrae were two vertebrae caudal to lower end vertebrae. The selection of LIV should take into account the vertebrae rotation. Correction of LIV rotation may be beneficial in reducing distal adding-on.

A previous study reported that postoperative shoulder imbalance had significantly associated with distal adding-on in Lenke 2 AIS patients [14]. However, in our study, postoperative shoulder imbalance was not an independent risk factor for distal adding-on in Lenke 1 AIS. The PT curve was the main compensatory to balance the postoperative shoulder. Although postoperative CA was not significantly associated with distal adding-on in the present study, it may compensate for postoperative shoulder imbalance [15].

In our study, no patients with postoperative distal adding-on had a second surgery. In most cases, the truncal shift with distal adding-on phenomenon could be well controlled after strengthening the back muscles, as shown in Fig. 4. She was a 14-year-old female with a 52° MT preoperatively. After surgery, the MT was well corrected but with a large FMS and FMSA. The distal adding-on phenomenon occurred at the 6-month follow-up, with a deviation of 6 mm from the CSVL to the LIV. After strengthening the lumbar and back muscles, the adding-on phenomenon is controlled and does not require revision surgery at a 4-year follow-up.

This study has some limitations. First, the sample size is small due to the single-institution research design. A larger sample of multicenter studies is being conducted. Second, only Lenke 1 AIS was included in the study. Third, this study did not analyze whether sagittal lumbar alignment parameters influence the adding-on.

Distal adding-on phenomenon occurred in 38.3% of patients. Therefore, achieving a balanced fusion mass intraoperatively was important to avoid postoperative distal adding-on, with FMS of <20 mm and FMS angle of <4.5°. Furthermore, correcting LIV rotation helped decrease the incidence of postoperative distal adding-on.