Role of Clavicle Chest Cage Angle Difference in Predicting Postoperative Shoulder Balance in Lenke 5C Adolescent Idiopathic Scoliosis Patients after Selective Posterior Fusion

Abstract

Objective

To evaluate the role of preoperative clavicle chest cage angle difference (CCAD) on postoperative radiographic shoulder imbalance, patient's satisfaction and surgeon's fulfillment in Lenke 5 adolescent idiopathic scoliosis (AIS). CCAD, as a novel radiographic parameter, has proven to be a reliable predictor for postoperative shoulder imbalance in Lenke 1 AIS patients. However, the value of CCAD in predicting shoulder balance has never been evaluated in Lenke 5 AIS patients.

Methods

A total of 42 Lenke 5C AIS patients aged from 10 to 18 years old with a minimum 2‐year follow‐up were enrolled for evaluation. All patients underwent selective posterior spinal instrumentation and fusion using the all segmental pedicle screw technique by the same surgical team. The fusion levels were determined according to the Lenke criteria. Shoulder height difference (SHD) and CCAD were measured on anteroposterior (AP) standing radiographs. The patients’ satisfaction and the surgeons’ fulfillment were evaluated using a questionnaire. A receiver operative characteristic curve analysis was performed to explore the threshold values of preoperative CCAD in the prediction of the final follow‐up radiographic shoulder imbalance, patients’ satisfaction and surgeons’ fulfillment.

Results

The average preoperative Cobb angle of the main curve was 46.8° ± 4.8°, and the average immediate postoperative Cobb angle was 13.3° ± 2.6°, representing an average surgical correction rate of 75.6% ± 8.5%. The average follow‐up time was 29.2 months. At the last follow‐up, the value of preoperative CCAD was significantly higher in patients with unbalanced shoulders (SHD ≥ 10 mm). At the final follow‐up, 66.7% (28/42) of the patients were satisfied with their appearance, while 33.3% (14/42) of the patients were not satisfied with their appearance. At the final follow‐up, 61.9% (26/42) of the surgeons were fulfilled with their operation, while 38.1% (16/42) of the surgeons were not. For patients’ satisfaction and surgeons’ fulfillment, the preoperative CCAD was significantly greater in patients with unsatisfied outcomes.

Discussion

Clavicle chest cage angle difference could be a reliable predictor for evaluating postoperative shoulder imbalance in AIS patients undergoing selective posterior fusion for Lenke 5C curves. A greater preoperative CCAD was significantly correlated with a postoperative radiographic imbalance of shoulders and dissatisfaction, which will guide spine surgeons in their preoperative planning and in the surgical management of AIS to reduce postoperative shoulder imbalance.

Introduction

The main goal of the surgical treatment for adolescent idiopathic scoliosis (AIS) is to obtain a 3‐dimensional correction and a stable spinal arthrodesis1, 2, 3. One of the most important measurements of successful scoliosis surgery is achieving good shoulder balance, which influences patients’ appearance and satisfaction postoperatively4, 5. Previous studies have identified several risk factors for postoperative shoulder imbalance, including stiffness of proximal thoracic curve, positive T₁ tilt, selection of fusion level, preoperative shoulder imbalance and positive clavicle angle6, 7.

Several studies related to the evaluation of shoulder balance in AIS patients have been on the basis of radiographic parameters, which poorly correlate with patients’ satisfaction and surgeons’ fulfillment. Cosmetic shoulder balance, a major concern of patients and their parents8, 9, often draws little attention from surgeons, who care more about the correction rate on the postoperative X‐ray film. Qiu et al. report that none of the existing radiographic parameters can precisely reflect the cosmetic appearance of the shoulders10. Meanwhile, a well‐accepted index to predict shoulder balance after correction surgery for AIS is lacking, although variable radiographic factors have been identified as risk factors for postoperative shoulder imbalance5, 11, 12.

Over the past 3 years, many surgeons have focused on the proximal thoracic curvature corrections in specific types of scoliosis (e.g. Lenke type 1 and 2) with proven good results; however, many other types of scoliosis that should be treated exist. Hong et al. found that significant shoulder imbalance could also be found in thoracolumbar/lumbar scoliosis, which has previously been neglected13. Sanders et al. report that the predictors for successful selective fusion in Lenke 5C AIS included skeletal maturity, a thoracolumbar/thoracic Cobb ratio less than 1.25, and a preoperative thoracic best bend of less than 20°14. For the postoperative shoulder balance, Yaszay et al. found that the preoperative lumbar Cobb angle was a significant predictor, yet the reason remains unclear. In these cases, the selective or non‐selective fusion was not clearly classified for analysis15.

As a novel preoperative predictor of postoperative shoulder imbalance, the clavicle chest cage angle difference (CCAD) was proposed, which was demonstrated to be reliable and easy to perform in thoracic AIS patients15, 16. However, the curve patterns of the patients varied (Lenke types 1, 2, 3, and 4 AIS) and the cases included those with or without fusion of the pelvic tilt (PT) curve. In our latest study, we found that in AIS patients undergoing posterior selective fusion for a right thoracic Lenke 1 curve, a greater preoperative CCAD (>5.58°, in particular) significantly correlates with a postoperative radiographic imbalance of their shoulders, and it is also associated with the patients’ satisfaction and surgeons’ fulfillment postoperatively16. The potential importance of CCAD in predicting the postoperative shoulder imbalance in Lenke 5C patients, however, has never been reported.

In light of these points, the present study evaluated the role of CCAD in predicting the postoperative shoulder imbalance in Lenke 5C patients after selective posterior fusion to guide spine surgeons in their surgical decision‐making preoperatively (Figs. 1, 2).

Figure 1.

A 13‐year‐old girl with Lenke 5C AIS had selective posterior fusion. (A) Preoperative anteroposterior (AP) view: main lumbar curve (48°), clavicle chest cage angle difference (CCAD) (−6°). (B) Final follow‐up AP view: main lumbar curve 10°, shoulder height difference (SHD) −5.3 mm, and CCAD −4°. LTCCA, left clavicle chest angle; RTCCA, right clavicle chest angle.

Figure 2.

A 12‐year‐old girl with Lenke 5C AIS had selective posterior fusion. (A) Preoperative anteroposterior (AP) view: main lumbar curve (46°), clavicle chest cage angle difference (CCAD) (−17°). (B) Final follow‐up AP view: main lumbar curve 8°, shoulder height difference (SHD) −15.4 mm, and CCAD −14°. LTCCA, left clavicle chest angle; RTCCA, right clavicle chest angle.

Materials and Methods

Approval for conducting this research has been obtained from the Institutional Review Board of the authors’ institute. A single center, retrospectively‐collected database of AIS patients was used in this study.

Inclusion and Exclusion Criteria

The inclusion criteria for patients were as follows: (i) those aged from 10 to 18 years at the time of the surgery; (ii) diagnosed with Lenke 5C AIS; (iii) treated with posterior selective spinal fusion; and (iv) complete radiographic data with a minimum 2‐year follow‐up. The following patients were excluded: (i) previous spine or shoulder injury; and (ii) having revision surgeries. A total of 42 patients were enrolled in our study. The radiographic shoulder height difference (SHD), CCAD, magnitude and flexibility of major coronal curve, age, gender, patients’ satisfaction, and surgeons’ fulfillment were recorded.

Surgical Procedure

All patients underwent selective posterior spinal instrumentation and fusion using the all segmental pedicle screw technique by the same surgical team. The fusion levels were determined according to the Lenke criteria17. The proximal fusion level was at T₁₀ in 15 patients, T₁₁ in 11 and T₁₂ in 16 patients. The distal fusion level was at L₃ in 19 patients and L₄ in 23 patients.

Radiographic Measurements

All patients had a full‐spine anteroposterior (AP) radiograph in a relaxed standing position. Patients were asked to stand facing straight ahead (not with chin down) with their knees fully extended, their feet apart to shoulder width, and their arms straight at the side. This posture was maintained in all the subjects when they had their AP radiographs before surgery, immediately after surgery and at follow‐ups. The SHD was measured as the height difference of the soft tissue shadows directly superior to the acromioclavicular joints (positive is defined as left shoulder up/right shoulder down)18. Because SHD <10 mm was considered the upper limit of balanced shoulder in previous studies, all patients were divided into two groups according to the final follow‐up: Group 1, with unbalanced shoulders (SHD ≥10 mm), and Group 2, with balanced shoulders (SHD <10 mm)18, 19. The CCAD was measured on the AP X‐ray films by using the following criteria:

1. The center chest cage line (CCL) was drawn from the centroid of T₁ to the centroid of T₁₂. A line was drawn perpendicular to the first line.

2. Another line was drawn from the middle of the proximal end of the clavicle to the middle of the distal end of the clavicle in both the right and left clavicles. The angles between the second line and the above two lines are the right clavicle chest angle (RTCCA) and left clavicle chest angle (LTCCA), respectively.

3. CCAD was calculated as follows: CCAD = LTCCA‐RTCCA.

Clinical Assessment

The patients’ satisfaction was evaluated by asking the following question at the final follow‐up: “Are you satisfied with your appearance?” The patients had two answers to choose from: “Yes” or “No” (categorized as Group A and Group B, respectively). The surgeons’ fulfillment was assessed by asking the question at the final follow‐up: “Are you satisfied with the patient's clinical appearance?” The surgeons also had two answers to choose from: “Yes” or “No,” which were categorized as Group A′ and Group B′, respectively. These two questions were a part of the questionnaire prepared by Smyrnis et al. 4, and were the most relevant to the satisfaction of appearance in the questionnaire.

Statistical Analysis

All of the parameters were measured by three experienced spine surgeons independently and the average value was applied. Analyses were performed using the Statistical Package for the Social Sciences (SPSS, Chicago, IL). Repeated measures analysis of variance (ANOVA) was applied to compare the CCAD preoperatively, postoperatively, and at final follow‐up. Two‐tailed Student's t‐tests were used to examine intergroup comparisons. Pearson correlation analysis was performed to confirm whether there was a correlation between preoperative CCAD and the final follow‐up SHD.

Results

Radiographic Outcomes

There were 40 female and 2 male patients in this study, with a mean age of 14.4 ± 1.4 years (range, 12–18 years). The average magnitude of the main curve was 48.9° ± 4.3°, and the correction rate of main curve was an average 76.0% ± 7.2% at the final follow‐up visit. The mean follow‐up time was 26.7 ± 1.9 months (range, 24–30 months), during which the overall CCAD changed from −10.4° ± 9.1° in the preoperative setting to −12.8° ± 8.7° in the immediate postoperative setting and −11.4° ± 8.7° at the final follow‐up (P = 0.44). For the radiographic shoulder balance, preoperative SHD averaged −13.8 ± 7.1 mm for all subjects, whereas the initial postoperative SHD was −10.5 ± 9.7 mm. At the final follow‐up, SHD was −10.4 ± 9.5 mm. According to the final follow‐up SHD, 13 patients with unbalanced shoulders were categorized into Group 1 (SHD ≥10 mm) and 29 patients with balanced shoulders were categorized as Group 2 (SHD <10 mm). The two groups were found to be statistically equivalent in terms of the age, Risser sign, and magnitude of the preoperative main curves (Table 1). However, the preoperative CCAD of Group 1 were significantly greater than the corresponding values of Group 2 (−17.6° ± 2.2° vs. −9.9° ± 9.0°, P = 0.04, Table 1). The Pearson correlation coefficient between preoperative CCAD and the final follow‐up SHD was 0.641 (P < 0.001).

Table 1.

Demographic data and preoperative radiographic parameter in the two groups (mean ± standard deviation)

Groups	Cases	Age (years)	Risser sign	Cobb angle (°)	Pre‐op SHD (mm)	Pre‐op CCAD (°)	Post‐op CCAD (°)	Follow‐up CCAD (°)
Group 1 (SHD ≥ 10 mm)	13	14.4 ± 1.5	3.1 ± 0.8	49.0 ± 4.1	−21.3 ± 4.1	−17.6 ± 2.2	−18.1 ± 2.8	−16.3 ± 4.5
Group 2 (SHD < 10 mm)	29	14.5 ± 1.1	2.8 ± 0.7	48.8 ± 5.0	−10.4 ± 5.4	−9.9 ± 9.0	−10.4 ± 9.4	−9.2 ± 9.1
t‐value	–	0.34	0.87	0.08	6.46	3.02	2.87	2.65
P‐value	–	0.74	0.39	0.94	<0.001	0.004	0.007	0.01

CCAD, clavicle chest cage angle difference; SHD, shoulder height difference.

Clinical Outcomes

According to the responses of the patients, 61.9% (26/42) were satisfied with their appearance at the final follow‐up, while 38.1% (16/42) of the patients were not satisfied with their appearance (Table 2). The preoperative CCAD was significantly lower in Group A (satisfied group) than Group B (Table 3, −8.4° ± 8.2° vs. −18.6° ± 3.3°, P < 0.001).

Table 2.

Grouping strategies according to patients’ satisfaction and surgeons’ fulfillment (cases)

Surgeons’ fulfillment	Patients’ satisfaction		Total
	Group A (satisfied)	Group B (unsatisfied)
Group A′ (satisfied)	24	2	26
Group B′ (unsatisfied)	4	12	16
Total	28	14	42

Table 3.

Comparison of CCAD (°) between different clinical outcomes (mean ± standard deviation)

Groups	Pre‐operation	Post‐operation	Follow‐up	P‐value
Patients’ satisfaction
Group A (28)	−8.4 ± 8.2	−8.9 ± 8.6	−8.3 ± 8.4	0.96
Group B (14)	−18.6 ± 3.3	−19.1 ± 3.6	−16.6 ± 6.5	0.29
P‐value	<0.001	<0.001	0.002	–
Surgeon’ fulfillment
Group A′ (26)	−9.8 ± 9.0	−10.2 ± 9.4	−8.8 ± 9.0	0.82
Group B′ (16)	−17.2 ± 3.6	−17.9 ± 3.9	−16.7 ± 9.9	0.73
P‐value	0.005	0.006	0.004	–

CCAD, clavicle chest cage angle difference.

With respect to surgeons’ fulfillment, 66.7% (28/42) of the patients were categorized as Group A′ (satisfied group), with a lower preoperative CCAD than Group B′ (Table 2, 3, −9.8° ± 9.0° vs. −17.2° ± 3.6°, P = 0.005). The opinions of patients and surgeons were consistent in 85.7% (36/42) of the patients (Table 2).

Discussion

Clavicle Chest Cage Angle Difference in Predicting Shoulder Imbalance in Thoracic Curve

As a novel preoperative factor, CCAD has proven to be effective in predicting the presence of postoperative shoulder imbalance in AIS patients undergoing posterior spinal fusion (PSF)20; however, the curve pattern of patients, the approaches of the surgery and the fusion levels have varied in previous studies. In our preliminary work, quantitative analysis of the effect of preoperative CCAD on postoperative SHD has been performed in Lenke 1 patients. For further investigation into the role of CCAD in predicting posterior shoulder imbalance, the current study was restricted to Lenke 5C AIS patients who were treated with selective posterior fusion. The slight variability in the level of upper instrumented vertebra (T₁₁ or T₁₂) reflected the homogeneity of our patients. On the basis of this homologous cohort, our results revealed that a greater preoperative CCAD is associated with an increased risk of postoperative radiographic shoulder imbalance. To evaluate the precise effect of preoperative CCAD on postoperative shoulder imbalance, another ROC curve analysis was performed, which showed that preoperative CCAD >5.5° was strongly predictive of slight shoulder imbalance (SHD ≥1 cm) at the final follow‐up.

Clavicle Chest Cage Angle Difference in Predicting Shoulder Balance in Lumbar Curve

Shoulder balance used to be referred to as radiographic shoulder balance, with less attention paid to the patients’ satisfaction. However, there is a current trend toward greater emphasis on patient‐centered outcomes21, 22, 23. Together with patients’ satisfaction, the spine surgeons’ fulfillment plays an important part in postoperative shoulder balance, which should not be neglected by us4, 24. Therefore, we evaluated both patients’ satisfaction and surgeons’ fulfillment and observed a significant greater preoperative CCAD in the subjects with unsatisfied clinical outcomes, clarifying the correlation between preoperative CCAD and patients’ satisfaction and surgeons’ fulfillment. In this study, we explored the precise value of CCAD in predicting postoperative shoulder imbalance in left lumbar Lenke 5C AIS patients with upper instrumented vertebra (UIV) at T₁₀ or T₁₁, and proved that decreased preoperative CCAD is associated with more radiographically balanced shoulders after surgery and a better outcome in terms of patients’ satisfaction and surgeons’ fulfillment. Although preoperative CCAD cannot predict the postoperative cosmetic shoulder imbalance, our results did show that preoperative CCAD is a risk factor for radiographic shoulder imbalance, as well as for dissatisfaction of patients and surgeons after scoliosis surgery. Because shoulder balance, as well as patients’ satisfaction and surgeons’ fulfillment, are complex phenomenon affected by various factors, further studies are underway to fully understand which aspects lead to dissatisfaction after scoliosis surgery.

Limitations

Several limitations of the current study exist. The current study could only analyze the correlation between preoperative CCAD and posterior cosmetic appearance, while the anterior appearance may be important for the patients themselves when looking into the mirror. Thus, the relationship between preoperative CCAD and anterior cosmetic shoulder balance needs to be further explored based on the premise that patients’ privacy is preserved. However, we analyzed the correlation between preoperative CCAD and postoperative shoulder balance in a specific curve type (Lenke 5), which eliminated potential confounding factors and allowed us to obtain a more convincing conclusion.

Conclusion

In AIS patients undergoing selective posterior fusion for Lenke 5C curves, CCAD could be a reliable predictor for evaluating postoperative shoulder imbalance. A greater preoperative CCAD was significantly correlated with a postoperative radiographic imbalance of their shoulders and dissatisfaction, which will guide spine surgeons in their preoperative planning and in the surgical management of AIS to reduce postoperative shoulder imbalance.