Analysis of Clinical and Radiological Predictive Factors for Moderate and Severe Pulmonary Impairment in 102 Adolescent Idiopathic Scoliosis (AIS) Patients With Major Cobb Angle ≥45°: A Study of an Asian Population

Abstract

Study Design

Retrospective cohort study.

Objective

To investigate the independent predictive factors for moderate-severe pulmonary impairment (MSPI) among adolescent idiopathic scoliosis (AIS) patients.

Methods

The preoperative pulmonary function tests (PFTs) of 102 AIS patients (major Cobb angle ≥45°) operated between 2015 and 2020 were retrospectively reviewed. Patients were divided into 2 groups based on the predicted forced vital capacity (FVC) and forced expiratory volume in one second (FEV₁), respectively. Group 1 had normal/ mild pulmonary impairment (FVC or FEV₁ >65% predicted) whereas Group 2 had MSPI (FVC or FEV₁ ≤65% predicted). Multivariate logistic regression analysis was performed to determine the predictive factors for MSPI.

Results

41.2% (N = 42) and 52.0% (N = 53) of the total patients were categorised into Group 2 (MSPI) based on predicted FVC and FEV₁, respectively. In general, Group 2 had more patients with a major main thoracic (MT) curve, larger MT curve with lower MT flexibility, a larger MT apical vertebra translation (MT-AVT), and a smaller thoracolumbar-lumbar (TL/L) AVT than Group 1 (P < .05). When analyzing the MT Cobb angle for every 10° increment, patients with MT Cobb angle ≥70° had MSPI (N = 31.4%). Body mass index (BMI) (FVC: aOR .8; FEV₁: aOR .9) and MT Cobb angle (for every 10° increment, FVC: aOR 1.7; FEV₁: aOR 1.8) were the significant independent predictive factors for MSPI.

Conclusion

MSPI was evident in patients with MT Cobb angle ≥70°, with MT Cobb angle and BMI being the significant independent predictive factors.

Keywords

adolescent idiopathic scoliosis pulmonary function tests moderate-severe pulmonary impairment

Introduction

Scoliosis is a three-dimensional (3D) deformity of the spine, and thoracic/rib cage deformity is one of the main components involved.^1,2 The impact of scoliosis on thoracic deformity was thought to be complex. Notably, scoliosis impedes ribs movement thus compromising the respiratory muscles and displacing the organs in the thoracic cavity. Eventually, this reduces the chest wall compliance directly and the lung compliance indirectly.³

Studies had been relating the spine deformity in adolescent idiopathic scoliosis (AIS) patients with its adverse effects on pulmonary functions.⁴ Machino et al⁵ concluded that AIS patients with restrictive lung disorder (RLD) had significantly smaller rib-cage volume than patients without RLD (3.94 L vs. 4.49 L, respectively). Several authors also demonstrated that pulmonary impairment was present among AIS patients with larger main thoracic (MT) Cobb angle, greater apical vertebral rotation angle, or hypokyphosis.^4,6,7 Moderate to severe pulmonary impairment had been reported as clinically relevant pulmonary impairment.^8-11 Yaszay et al⁷ reviewed preoperative pulmonary function tests (PFTs) and 3D reconstructed imaging of 163 AIS patients, and deduced that patients with moderate to severe pulmonary impairment (≤65% predicted PFTs values) had 3D coronal Cobb >80°, thoracic lordosis >20°, and absolute apical rotation >25°. Nevertheless, majority of the earlier studies did not include clinical factors ie, age, gender, or body mass index (BMI), in their analysis of predictive factors for pulmonary impairment.

Despite extensive literature review, only scarce data reported on both clinical and radiological predictive factors for poor pulmonary function among AIS patients were found. To date, controversies on the negative impact of scoliosis on pulmonary function still persist.⁴ Therefore, the purpose of this study was to investigate the clinical and radiological predictive factors for moderate-severe pulmonary impairment (MSPI) among AIS patients. The reported data would benefit surgeons in risk counselling, preoperative planning, and preoperative optimization to avoid detrimental respiratory complications.

Methodology

This was a retrospective study conducted in a single academic institution involving AIS patients who underwent posterior spinal fusion (PSF) between the years 2015 and 2020. This research was approved by our institution ethical board (MREC ID NO.: 2020108-9138). AIS patients who were of Asian ethnicity, aged between 10 and 18 years old, with a major Cobb angle of ≥45°, and inclusive of all Lenke curve types were recruited. Patients who had non-idiopathic scoliosis, anterior surgery, revision surgery, or patients with pre-existing cardiopulmonary diseases were excluded.

Pulmonary Function Tests (PFTs)

Preoperative PFTs were performed in 107 AIS patients. We excluded 5 patients with incomplete PFTs data. Plethysmography and PFTs were used to measure total lung capacity (TLC), forced vital capacity (FVC), and forced expiratory volume in one second (FEV₁). Each test was repeated 3 times and the single best effort was recorded. Age, gender, and height-matched standards published by the American Thoracic Society were used to generate a percentage of the predicted values for each pulmonary parameter (predicted FVC or predicted FEV₁). In this study, pulmonary function was identified as normal when FVC or FEV₁ were >80% of the predicted values. Mild pulmonary impairment was defined as FVC or FEV₁ ≤80% but >65% predicted, whereas moderate pulmonary impairment was FVC or FEV₁ ≤65% but ≥50% predicted. Severe pulmonary impairment was defined as FVC or FEV₁ <50% predicted.^8-12

The definitions of PFTs parameters are as follows⁶:

1. Forced vital capacity (FVC): maximum volume of air expelled as rapidly and completely as possible by a maximum effort after a maximum inspiration.

2. Forced expiratory volume in one second (FEV₁): The amount of air expired in one second during a rapid and maximal expiration from full inspiration.

3. Total lung capacity (TLC): summation of FVC and reserved volume (RV).

Data Collection

Clinical data collected were age, gender, menarche age, and BMI. Lenke curve types, Risser sign, MT and thoracolumbar/lumbar (TL/L) Cobb angle, side bending (SB) Cobb angle, flexibility, thoracic kyphosis (TK), location of apical vertebra, number of vertebrae involved, coronal deformity angular ratio (C-DAR), MT apical vertebra translation (MT-AVT), and TL/L apical vertebra translation (TL/L-AVT) were documented.

Below are the definitions of the radiological parameters:

1. Side bending (SB) flexibility:

= \frac{P r e o p e r a t i v e C o b b a n g l e - S i d e B e n d i n g C o b b a n g l e}{P r e o p e r a t i v e C o b b a n g l e} \times 100

2. Coronal deformity angular ratio (C-DAR): unitless number used to evaluate the severity of the coronal deformity.

= \frac{M a j o r c u r v e C o b b a n g l e}{N u m b e r o f v e r t e b r a l l e v e l s i n v o l v e d}

3. Main thoracic apical vertebra translation (MT-AVT): the horizontal distance from the midpoint of the MT curve apical vertebra (or intervertebral disc) to C7 plumb line.

4. Thoracolumbar-lumbar apical vertebra translation (TL/L-AVT): the horizontal distance from the midpoint of the TL/L curve apical vertebra (or intervertebral disc) to central sacral vertical line (CSVL).

5. Thoracic kyphosis (TK): Cobb angle between the superior end plate of T5 vertebra and the inferior end plate of T12 vertebra, measured in the sagittal plane.

Statistical Analysis

Data analysis was performed using the SPSS version 26.0 (IBM SPSS Statistics for Windows, IBM Corp., Armonk, NY, USA). Descriptive data analysis was conducted for all clinical, radiological, and PFTs data. Independent samples t-test and chi-square test were performed for continuous variables and categorical variables, respectively. For statistical analysis, all patients in this cohort were classified into 2 groups based on the predicted FVC or FEV₁, respectively. Group 1 patients had normal/ mild pulmonary impairment (FVC or FEV₁ >65% predicted) whereas Group 2 patients had moderate-severe pulmonary impairment (MSPI) (FVC or FEV₁ ≤65% predicted).

Univariate logistic regression analysis followed by multivariate logistic regression analysis was performed to identify the independent predictive factors for MSPI (FVC or FEV₁ ≤65% predicted). Factors with P < .25 identified in the univariate logistic regression analysis were subsequently included in the multivariate logistic regression analysis. Independent predictive factors were determined by P < .05. Receiver-Operating-Characteristics (ROC) curve was drawn to predict MSPI for various MT Cobb angles. Youden Index was calculated for various points on ROC curve, whereby the optimal cut-off value was defined as the point with the highest Youden Index.^13,14

Results

This cohort of 102 AIS patients had a mean age of 16.1 ± 4.3 years. Mean MT Cobb angle was 61.1 ± 19.7°. Mean thoracic kyphosis (T5 – T12) in this study was 21.6 ± 14.1°. Lenke type 1 curve (N = 44.1%) demonstrated predominance in this study, followed by Lenke type 2 curve (N = 20.6%) and Lenke type 5 curve (N = 20.6%) (Table 1). There were 41.2% (N = 42) and 52.0% (N = 53) of the total patients categorised into Group 2 (MSPI) based on the predicted FVC and predicted FEV₁, respectively (Tables 1 and 2). In general, Group 2 patients had significantly lower BMI, more patients with a major MT curve, larger MT curve with lower MT flexibility, larger MT apical vertebra translation (MT-AVT), and smaller thoracolumbar-lumbar (TL/L) AVT compared to Group 1 patients (P < .05). Based on the predicted FVC, Group 2 patients had significantly lower BMI of 17.5 ± 3.8 kg/m² as compared to Group 1 patients (19.3 ± 3.4 kg/m²) (P = .014) (Table 1).

Table 1.

Preoperative Clinical and Radiological Data of 102 AIS Patients in Relation to the Predicted FVC (%).

Parameters	Overall (N = 102, 100.0%)	Group 1 FVC >65% Predicted (N = 60, 58.8%)	Group 2 FVC ≤65% Predicted (N = 42, 41.2%)	P value
Clinical data
Gender, N (%)				.831
Male	13 (12.7)	8 (13.3)	5 (11.9)
Female	89 (87.3)	52 (86.7)	37 (88.1)
Age (years)	16.1 ± 4.3	15.9 ± 4.2	16.2 ± 4.6	.747
Menarche age (years)	12.4 ± 1.3	12.2 ± 1.4	12.6 ± 1.2	.239
BMI (kg/m²)	18.6 ± 3.6	19.3 ± 3.4	17.5 ± 3.8	.014*
Radiological data
Risser sign, N (%)				.946
0	6 (5.9)	3 (5.1)	3 (7.1)
1	3 (5.9)	0 (.0)	3 (7.1)
2	11 (10.8)	6 (10.2)	5 (11.9)
3	11 (10.8)	8 (13.6)	3 (7.1)
4	42 (41.2)	26 (44.1)	16 (38.1)
5	29 (28.4)	16 (27.1)	12 (28.6)
Lenke curve type, N (%)				.151
1	45 (44.1)	25 (41.7)	20 (47.6)
2	21 (20.6)	9 (15.0)	12 (28.6)
3	5 (4.9)	2 (3.3)	3 (7.1)
4	1 (1.0)	0 (.0)	1 (2.4)
5	21 (20.6)	16 (26.7)	5 (11.9)
6	9 (8.8)	8 (13.3)	1 (2.4)
Curve type
Lenke Type 1 - 4, N (%)	72 (70.6)	36 (60.0)	36 (85.7)	.007*
MT Cobb angle (⁰)	61.1 ± 19.7	54.7 ± 17.5	70.1 ± 19.4	<.001*
MT side bending Cobb angle (⁰)	29.7 ± 17.8	24.0 ± 14.8	37.8 ± 18.8	<.001*
MT flexibility	54.9 ± 19.4	59.2 ± 19.7	48.9 ± 17.5	.011*
TL/L Cobb angle (⁰)	46.2 ± 17.7	47.0 ± 19.7	45.1 ± 14.5	.588
TL/L side-bending Cobb angle (⁰)	7.9 ± 16.0	8.6 ± 17.0	7.0 ± 14.6	.617
TL/L flexibility	91.0 ± 30.4	90.8 ± 32.2	91.2 ± 27.9	.948
Thoracic kyphosis (T5-T12) (⁰)	21.6 ± 14.1	21.1 ± 12.3	22.4 ± 16.4	.636
Location of apex vertebra, N (%)				.698
T1 – T4	3 (2.9)	2 (3.3)	1 (2.4)
T5 – T8	88 (86.3)	52 (86.7)	36 (85.7)
T9 – T12	8 (7.8)	3 (5.0)	5 (11.9)
L1 – L3	3 (2.9)	3 (5.0)	0 (.0)
No. of vertebral level involved	6.4 ± 1.4	6.1 ± 1.3	6.8 ± 1.3	.007*
C-DAR	11.0 ± 3.0	10.9 ± 3.1	11.2 ± 3.0	.620
MT-AVT	45.4 ± 24.2	37.9 ± 22.3	56.2 ± 22.9	<.001*
TL/L-AVT	23.4 ± 24.7	28.1 ± 26.1	16.7 ± 21.0	.024*

FVC, forced vital capacity; BMI, body mass index; MT, main thoracic; TL/L, thoracolumbar/lumbar; C-DAR, coronal deformity angular ratio; MT-AVT, main thoracic-apical vertebra translation; TL/L-AVT, thoracolumbar/lumbar-apical vertebra translation.

*Statistically significant.

Table 2.

Preoperative Clinical and Radiological Data of 102 AIS Patients in Relation to the Predicted FEV₁ (%).

Parameters	Overall (N = 102, 100.0%)	Group 1 FEV1 >65% Predicted (N = 49, 48.0%)	Group 2 FEV1 ≤65% Predicted (N = 53, 52.0%)	P Value
Clinical data
Gender, N (%)				.654
Male	13 (12.7)	7 (14.3)	6 (11.3)
Female	89 (87.3)	42 (85.7)	47 (88.7)
Age (years)	16.1 ± 4.3	16.6 ± 4.4	15.5 ± 4.3	.211
Menarche age (years)	12.4 ± 1.3	12.3 ± 1.4	12.5 ± 1.1	.573
BMI (kg/m²)	18.6 ± 3.6	19.3 ± 3.3	17.9 ± 3.8	.069
Radiological data
Risser sign, N (%)				.306
0	6 (5.9)	1 (2.0)	5 (9.4)
1	3 (2.9)	0 (.0)	3 (5.7)
2	11 (10.8)	3 (6.1)	8 (15.1)
3	11 (10.8)	6 (12.2)	5 (9.4)
4	42 (41.2)	21 (42.9)	21 (39.6)
5	29 (28.4)	18 (36.7)	11 (20.8)
Lenke curve type, N (%)				.117
1	45 (44.1)	20 (40.8)	25 (47.2)
2	21 (20.6)	6 (12.2)	15 (28.3)
3	5 (4.9)	2 (4.1)	3 (5.7)
4	1 (1.0)	0 (.0)	1 (1.9)
5	21 (20.6)	15 (30.6)	6 (11.3)
6	9 (8.8)	6 (12.2)	3 (5.7)
Curve type
Lenke Type 1 - 4, N (%)	72 (70.6)	28 (57.1)	44 (83.0)	.005*
MT Cobb angle (°)	61.1 ± 19.7	52.1 ± 16.8	69.3 ± 18.7	<.001*
MT side bending Cobb angle (°)	29.7 ± 17.8	21.7 ± 14.0	37.0 ± 17.9	<.001*
MT flexibility	54.9 ± 19.4	61.4 ± 19.3	49.0 ± 17.7	.002*
TL/L Cobb angle (°)	46.2 ± 17.7	48.0 ± 20.1	44.6 ± 15.1	.329
TL/L side-bending Cobb angle (°)	7.9 ± 16.0	9.3 ± 17.4	6.7 ± 14.6	.399
TL/L flexibility	91.0 ± 30.4	89.8 ± 31.7	92.1 ± 29.3	.706
Thoracic kyphosis (T5-T12) (°)	21.6 ± 14.1	21.2 ± 12.4	22.0 ± 15.6	.785
Location of apex vertebra, N (%)				.595
T1 – T4	3 (2.9)	2 (4.1)	1 (1.9)
T5 – T8	88 (86.3)	42 (85.7)	46 (86.8)
T9 – T12	8 (7.8)	2 (4.1)	6 (11.3)
L1 – L3	3 (2.9)	3 (6.1)	0 (.0)
No. of vertebral level involved	6.4 ± 1.4	6.0 ± 1.3	6.7 ± 1.4	.012*
C-DAR	11.0 ± 3.0	10.7 ± 3.0	11.3 ± 3.1	.328
MT-AVT	45.4 ± 24.2	35.5 ± 21.8	54.6 ± 22.8	<.001*
TL/L-AVT	23.4 ± 24.7	30.5 ± 25.7	16.9 ± 21.9	.007*

FEV₁, forced expiratory volume in one second; BMI, body mass index; MT, main thoracic; TL/L, thoracolumbar/lumbar; C-DAR, coronal deformity angular ratio; MT-AVT, main thoracic-apical vertebra translation; TL/L-AVT, thoracolumbar/lumbar-apical vertebra translation.

*Statistically significant.

PFTs

The results of PFTs stratified by the magnitude of MT Cobb angle was summarised in Table 3. The mean absolute FVC and its percentage of predicted values in this cohort were 2.31 ± .62 L and 66.4 ± 16.1%, respectively. The mean absolute FEV₁ and predicted FEV₁ were 2.02 ± .57 L and 62.9 ± 16.6%, respectively. The predicted values (mean percentage) for FVC were ≤65% (MSPI) when MT Cobb angle exceeded 70° whereas the predicted values (mean percentage) for FEV₁ were ≤65% (MSPI) when MT Cobb angle exceeded 60° (Table 3). Meanwhile, patients with MT Cobb angle ≥90° (severe scoliosis) had a mean predicted FVC of 53.7 ± 15.0% and a mean predicted FEV₁ of 49.4 ± 13.3%. The mean TLC was 3.29 ± .87 L and the mean FEV₁/FVC ratio of this entire cohort demonstrated a restrictive pattern (FEV₁/FVC >.7).¹⁵

Table 3.

Pulmonary Function Tests (PFTs) Stratified by the Magnitude of Main Thoracic (MT) Curves.

MT Cobb Angle (°)	N (%)	FVC		FEV₁		FEV₁/FVC Ratio
MT Cobb Angle (°)	N (%)	L	% of Predicted FVC	L	% of Predicted FEV₁	FEV₁/FVC Ratio
<30	8 (7.8)	2.36 ± .40	69.8 ± 11.4	2.11 ± .35	66.9 ± 11.8	.90 ± .04
30 – 39	8 (7.8)	2.55 ± .43	77.9 ± 9.8	2.33 ± .38	77.5 ± 9.7	.91 ± .03
40 – 49	6 (5.9)	2.47 ± .63	71.0 ± 19.8	2.24 ± .59	69.5 ± 19.9	.91 ± .07
50 – 59	22 (21.6)	2.57 ± .49	73.0 ± 11.1	2.26 ± .42	70.0 ± 8.2	.88 ± .06
60 – 69	26 (25.5)	2.37 ± .62	66.1 ± 16.2	2.06 ± .63	61.9 ± 17.7	.86 ± .09
70 – 79	17 (16.7)	2.00 ± .65	57.9 ± 14.6	1.70 ± .54	52.5 ± 14.4	.85 ± .08
80 – 89	8 (7.8)	2.08 ± .87	59.8 ± 23.0	1.75 ± .76	56.9 ± 23.4	.85 ± .06
≥90	7 (6.9)	1.84 ± .59	53.7 ± 15.0	1.60 ± .46	49.4 ± 13.3	.87 ± .09
Overall	102	2.31 ± .62	66.4 ± 16.1	2.02 ± .57	62.9 ± 16.6	.87 ± .07

MT, main thoracic; % of Predicted, percentage of predicted values; FVC, forced vital capacity; FEV₁, forced expiratory volume in one second.

Logistic Regression Analysis

From the univariate analysis in predicting MSPI (≤65% of predicted values) using both FVC and FEV₁, we identified the following variables with P < .25: BMI, Lenke 1 to 4 curve type, MT Cobb angle, MT SB Cobb angle, MT flexibility, number of vertebral levels involved, MT-AVT, and TL/L-AVT. Further analysis using multivariate logistic regression for the predicted FVC demonstrated that BMI (aOR, .8; P = .013) and MT Cobb angle (for every 10° increment) (aOR, 1.7; P < .001) were the significant independent predictive factors for MSPI (Table 4). Similarly, multivariate analysis using predicted FEV₁ demonstrated BMI (aOR, .9; P = .045) and MT Cobb angle (for every 10° increment) (aOR, 1.8; P < .001) as the significant independent predictors for MSPI (Table 4).

Table 4.

Multivariate Logistic Regression Analysis Performed to Determine the Independent Predictive Factors for Moderate-Severe Pulmonary Impairment (MSPI) (FVC or FEV₁ ≤65% predicted).

Multivariate Analysis
	aOR	95% CI	P Value
FVC ≤65% predicted
BMI	.8	.7 - 1.0	.013
MT Cobb angle (+10°)	1.7	1.3 - 2.3	<.001
R² = .290
FEV₁ ≤65% predicted
BMI	.9	.8 - 1.0	.045
MT Cobb angle (+10°)	1.8	1.4 - 2.5	<.001
R² = .311

MT, main thoracic; FVC, forced vital capacity; FEV₁, forced expiratory volume in one second; +10°, for every 10° increment.

ROC Curve Analysis

Upon establishing that MT Cobb angle was a statistically significant independent predictor for MSPI, we proceeded to analyse the data using ROC curve analysis. Using the predicted FVC, we identified that a MT Cobb angle of 64.5° had the highest Youden Index (.481) with area under the curve (AUC) of .751 in predicting MSPI (FVC ≤65% predicted) (P < .001). This cut-off value of 64.5° for MT Cobb angle had a sensitivity of 71.4% and specificity of 76.7% (Figure 1). Similarly, ROC curve analysis using the predicted FEV₁ illustrated that a MT Cobb angle of 64.5° had the highest Youden Index (.476) with AUC of .767 (P < .001). It has a lower sensitivity of 66.0% but a relatively higher specificity of 81.6% in predicting MSPI (Figure 2).

Figure 1.

Receiver-operating-characteristic (ROC) curve for the MT Cobb angle cut-off value of 64.5° in predicting moderate-severe pulmonary impairment (MSPI) (FVC ≤65% predicted) in a single curve. Star indicated 71.4% sensitivity, 76.7% specificity, youden index of .481, and AUR of .751.

Figure 2.

Receiver-operating-characteristic (ROC) curve for the MT Cobb angle cut-off value of 64.5° in predicting moderate-severe pulmonary impairment (MSPI) (FEV₁ ≤65% predicted) in a single curve. Star indicated 66.0% sensitivity, 81.6% specificity, youden index of .476, and AUR of .767.

Discussion

Clinical studies had explored the connection between spinal deformities and pulmonary function, particularly in examining the link between AIS and pulmonary impairment.^4,8,16,17 Clarification on the widely held presumption that only severe scoliosis is associated with pulmonary impairment remains inconclusive, despite numerous prior studies. In the study of the natural history of scoliosis, Weinstein et al¹⁸ reported that having a Cobb angle of ≥50° at skeletal maturity was a significant predictor for reduced pulmonary function. Early study by Szeinberg et al¹⁹ reported that pulmonary impairment was even found among idiopathic scoliosis patients with Cobb angle <60°. A large series study involving fifteen centers by Johnston et al⁸ later reported that clinically relevant pulmonary impairment was present among AIS patients with mean MT curves of more than 70°. The authors concluded that PFTs correlated with the severity of MT curve, sagittal plane hypokyphosis (T5-T12 < 10°), and to a lesser extent in proximal thoracic (PT) curve (PT Cobb angle >30°).⁸

Likewise, our study revealed that moderate pulmonary impairment (FVC or FEV₁ ≤65% but ≥50% predicted) was present among AIS patients with MT Cobb angle of ≥70° and severe pulmonary impairment (FVC or FEV₁ <50% predicted) could be observed notably among patients with severe scoliosis (MT Cobb angle ≥90°). In the recent systematic review and meta-analysis by Kan et al,⁴ the team concluded that larger PT and MT Cobb angles, hypokyphosis, greater lumbar lordotic angles, longer thoracic curve, larger rib hump, increased apical vertebral rotation angles, and smaller rib cages, were the variables associated with poorer pulmonary functions. In a long term follow-up study of more than 27 years after surgery, 27.7% of the AIS patients in the study had restrictive pulmonary function whereby large rib humps and large vertebral rotation were the significant factors identified for poor predicted FVC.²⁰ Nevertheless, in our study cohort, the only significant radiological predictive factor identified for MSPI was the MT Cobb angle.

Newton et al¹¹ reported that AIS patients with hypokyphosis (≤10°) were more likely to have MSPI. Dreimann et al¹² concluded that both spectrums of thoracic kyphosis with combination of thoracic curve magnitude have significant impact on predicted FVC. Patients with lordoscoliosis as well as kyphoscoliosis have an increased risk for reduced predicted FVC. Meanwhile, in our current study, the mean thoracic kyphosis (T5-T12) for Group 2 patients (MSPI) were 22.4 ± 16.4° and 22.0 ± 15.6° for the predicted FVC and predicted FEV₁, respectively. However, no statistical significance was identified when they were compared to Group 1 patients. Our data analysis also revealed that Group 2 patients (MSPI) had a significantly more severe MT curve with lower MT flexibility, larger MT-AVT, and smaller TL/L-AVT. Similarly, Wang et al⁶ conducted a correlation analysis between PFTs and radiological parameters of the main right thoracic curve in AIS patients and deduced that the predicted FVC had significant negative correlations with MT Cobb angle, MT curve-rib hump, MT curve apical vertebral body-to-rib ratio (MT-AVB-R), and MT-AVT.

By analysing the clinical parameters, we identified that BMI was the other significant independent predictive factor for MSPI (FVC: aOR .8; FEV₁: aOR .9) (P < .05). Besides denoting nutritional status, BMI was also reported in prior research findings to have a correlation with poor pulmonary function.^21-23 In a prospective study by Xu et al,²³ a positive correlation was found between BMI and pulmonary function whereby low BMI (BMI <17.7 kg/m²) was significantly associated with pulmonary impairment. Our logistic regression analysis affirmed the results reported in the aforementioned studies regarding the relationship between BMI and pulmonary impairment. Based on the predicted FVC, Group 2 patients (MSPI) in our study had significantly lower BMI of 17.5 ± 3.8 kg/m² (P = .014).

It was proven that ethnicity difference does have a role in affecting patients’ pulmonary function. Generally, smaller reduction (∼11%) in FVC and FEV₁ were observed in Southeast Asian patients as compared to patients of Caucasian ethnicity.²⁴ Therefore, the results in our study would be relatively more pertinent to an Asian population. The variability of accuracy in pulmonary function testing could be one of the main limitations of this study.^11,25 There were theoretical concerns regarding the usage of normative data based on height and gender to derive predicted values for PFTs in the evaluation of a condition affecting the height of the test subjects, for instance, AIS population in this study.²⁵ The results of the current PFTs in this study were not corrected for the height difference due to scoliosis. This could potentially lead to the underestimation of the reduced pulmonary function. Nevertheless, the authors made an attempt to reduce the impact of other confounding variables by excluding patients with known cardio-pulmonary diseases. Despite PFT being an objective measure of pulmonary function, another notable drawback in this study was the lack of patient reported outcome measurements (PROMs) to reflect the clinical relevance in terms of respiratory symptoms. Future studies could incorporate PROMs into the analysis.

This study will be crucial for deformity surgeons to be vigilant in identifying patients at high risk for MPSI (especially those do not comply with PFT) before proceeding with surgical procedures that could further compromise the pulmonary function, such as thoracotomy, thoracoplasty, and rib resections with 3-column osteotomies. We also found that larger MT curves were associated with moderate to severe pulmonary impairment with a 1.7 to 1.8 times increased likelihood of having MSPI for every 10° increment in Cobb angle. Therefore, corrective surgeries should be performed as early as possible among AIS patients with worsening thoracic curvatures to prevent deterioration in pulmonary function.

Conclusion

MSPI was evident in AIS patients with a MT Cobb angle of ≥70°. MT Cobb angle and BMI were significant independent predictive factors for MSPI in AIS patients with a major Cobb angle of ≥45°. The ROC curve analysis demonstrated that a MT Cobb angle of 64.5° is the cut-off value for predicting MSPI in AIS patients, with a sensitivity of 71.4% and specificity of 76.7% for predicted FVC, and sensitivity of 66.0% and specificity of 81.6% for predicted FEV₁.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Chee Kidd Chiu

Chris Yin Wei Chan

Mun Keong Kwan

References

Delorme

Violas

Dansereau

de Guise

Aubin

Labelle

. Preoperative and early postoperative three-dimensional changes of the rib cage after posterior instrumentation in adolescent idiopathic scoliosis. Eur Spine J. 2001;10:101-107. doi:10.1007/s005860000196

Smith

Sciubba

Samdani

. Scoliosis: a straightforward approach to diagnosis and management. J Am Acad Physician Assistants. 2008;21:40-45. doi:10.1097/01720610-200811000-00009

Koumbourlis

. Scoliosis and the respiratory system. Paediatr Respir Rev. 2006;7:152-160. doi:10.1016/j.prrv.2006.04.009

Kan

MMP

Negrini

Di Felice

, et al. Is impaired lung function related to spinal deformities in patients with adolescent idiopathic scoliosis? A systematic review and meta-analysis—SOSORT 2019 award paper. Eur Spine J. 2023;32:118-139. doi:10.1007/s00586-022-07371-z

Machino

Kawakami

Ohara

Saito

Tauchi

Imagama

. Three-dimensional reconstruction image by biplanar stereoradiography reflects pulmonary functional states inadolescent idiopathic scoliosis. J Clin Neurosci. 2021;88:178-184. doi:10.1016/j.jocn.2021.03.043

Wang

Yang

Wang

, et al. Correlation analysis between the pulmonary function test and the radiological parameters of the main right thoracic curve in adolescent idiopathic scoliosis. J Orthop Surg Res. 2019;14:443. doi:10.1186/s13018-019-1451-z

Yaszay

Bastrom

Bartley

Parent

Newton

. The effects of the three-dimensional deformity of adolescent idiopathic scoliosis on pulmonary function. Eur Spine J. 2017;26:1658-1664. doi:10.1007/s00586-016-4694-y

Johnston

Richards

Sucato

, et al. Correlation of preoperative deformity magnitude and pulmonary function tests in adolescent idiopathic scoliosis. Spine. 2011;36:1096-1102. doi:10.1097/BRS.0b013e3181f8c931

Kim

Lenke

Bridwell

Cheh

Whorton

Sides

. Prospective pulmonary function comparison following posterior segmental spinal instrumentation and fusion of adolescent idiopathic scoliosis: is there a relationship between major thoracic curve correction and pulmonary function test improvement? Spine. 2007;32:2685-2693. doi:10.1097/BRS.0b013e31815a7b17

10.

Liang

Qiu

Shen

, et al. Predictive factors of postoperative pulmonary complications in scoliotic patients with moderate or severe pulmonary dysfunction. J Spinal Disord Tech. 2010;23:388-392. doi:10.1097/BSD.0b013e3181b55ff4

11.

Newton

Faro

Gollogly

Betz

Lenke

Lowe

. Results of preoperative pulmonary function testing of adolescents with idiopathic scoliosis. A study of six hundred and thirty-one patients. J Bone Joint Surg Am. 2005;87:1937-1946. doi:10.2106/jbjs.D.02209

12.

Dreimann

Hoffmann

Kossow

Hitzl

Meier

Koller

. Scoliosis and chest cage deformity measures predicting impairments in pulmonary function: a cross-sectional study of 492 patients with scoliosis to improve the early identification of patients at risk. Spine. 2014;39:2024-2033. doi:10.1097/brs.0000000000000601

13.

Fluss

Faraggi

Reiser

. Estimation of the Youden index and its associated cutoff point. Biom J. 2005;47:458-472. doi:10.1002/bimj.200410135

14.

Perkins

Schisterman

. The inconsistency of “optimal” cutpoints obtained using two criteria based on the receiver operating characteristic curve. Am J Epidemiol. 2006;163:670-675. doi:10.1093/aje/kwj063

15.

Martinez-Pitre

Sabbula

Cascella

. Restrictive lung disease. In: StatPearls. Treasure Island (FL) with Ineligible Companies. Disclosure: Bhanusivakumar Sabbula Declares No Relevant Financial Relationships with Ineligible Companies. Disclosure: Marco Cascella Declares No Relevant Financial Relationships with Ineligible Companies. St. Petersburg, FL: StatPearls Publishing LLC; 2023.

16.

Deng

Luo

Zhou

, et al. Relationship between pulmonary function and thoracic morphology in adolescent idiopathic scoliosis: a new index, the “apical vertebra deviation ratio”. As a Predictive Factor for Pulmonary Function Impairment. Spine. 2021;46:87-94. doi:10.1097/brs.0000000000003731

17.

Ledonio

Rosenstein

Johnston

Regelmann

Nuckley

Polly

. Pulmonary function tests correlated with thoracic volumes in adolescent idiopathic scoliosis. J Orthop Res. 2017;35:175-182. doi:10.1002/jor.23304

18.

Weinstein

. The natural history of adolescent idiopathic scoliosis. J Pediatr Orthop. 2019;39:S44-S46. doi:10.1097/bpo.0000000000001350

19.

Szeinberg

Canny

Rashed

Veneruso

Levison

. Forced vital capacity and maximal respiratory pressures in patients with mild and moderate scoliosis. Pediatr Pulmonol. 1988;4:8-12. doi:10.1002/ppul.1950040104

20.

Akazawa

Kuroya

Iinuma

, et al. Pulmonary function and thoracic deformities in adolescent idiopathic scoliosis 27 years or longer after spinal fusion with Harrington instrument. J Orthop Sci. 2018;23:45-50. doi:10.1016/j.jos.2017.08.016

21.

Tarrant

Queally

Moore

Kiely

. Prevalence and impact of low body mass index on outcomes in patients with adolescent idiopathic scoliosis: a systematic review. Eur J Clin Nutr. 2018;72:1463-1484. doi:10.1038/s41430-018-0095-0

22.

Tung

Uvodich

Anderson

Carpenter

Sherman

Lozano

. Do heavier patients with adolescent idiopathic scoliosis have more preserved thoracic kyphosis and pulmonary function? Spine Deform. 2018;6:704-706. doi:10.1016/j.jspd.2018.05.007

23.

Sun

Zhu

Qiao

Mao

Qiu

. Body mass index as an indicator of pulmonary dysfunction in patients with adolescent idiopathic scoliosis. J Spinal Disord Tech. 2015;28:226-231. doi:10.1097/BSD.0b013e31825d97df

24.

Stocks

Sonnappa

Lum

. Lung function testing in children: importance of race and ethnic-specific reference equations. Expet Rev Respir Med. 2014;8:527-531. doi:10.1586/17476348.2014.927317

25.

Baur

Degens

Heitmann

, et al. Lung function testing: the dilemma of predicted values in relation to the individual variability. Respiration. 1996;63:123-130. doi:10.1159/000196531