Patient characteristics
Between January 2013 and December 2019, 70 patients (53 men, 17 women) with a mean age of 45.5±1.7 (range, 19–73) years and BMI of 26.5±4.3 (range, 19.9–38.0) were enrolled in this study. Among this population, six patients had a history of DM, eight patients had a history of smoking, and eight reported a history of alcohol consumption.
One level of TLHD was validated in 58 patients, two levels in 10 patients, and three levels in two patients. The distributions of levels were as follows: 12 cases, T10/T11; 34 cases, T11/T12; 24 cases, T12/L1; and 14 cases, L1/2.
The distribution of each radiographic parameter and reliability of the radiographic parameters are presented in Table 1.
Table 1 The values and inter observer reliability for radiographic parameter
Parameters
|
Mean
|
Standard deviation
|
Range
|
Inter-rater ICC
|
SVA(mm)
|
27.9
|
24.8
|
-46.9-73.1
|
0.90
|
CVA(mm)
|
7.9
|
5.3
|
2.1-26.1
|
0.92
|
TK(°)
|
29.1
|
12.3
|
0.8-56.0
|
0.91
|
TLK(°)
|
21.8
|
13.1
|
0.9-57.0
|
0.88
|
LL(°)
|
47.5
|
14.9
|
7.5-71.4
|
0.84
|
T1PA(°)
|
8.2
|
7.3
|
0.3-43.3
|
0.91
|
L1PA(°)
|
6.2
|
4.9
|
0.2-19.6
|
0.89
|
PI(°)
|
45.9
|
7.5
|
30.7-63.6
|
0.86
|
SS(°)
|
34.6
|
7.9
|
14.1-53.1
|
0.92
|
PT(°)
|
11.4
|
9.4
|
0.2-38.3
|
0.90
|
PI-LL(°)
|
12.4
|
9.4
|
0.5-44.9
|
0.88
|
Subtypes of sagittal thoracic-lumbar alignment
We summarised each thoracic-lumbar alignment figure and found two subtypes differentiated by the apex of kyphotic curves (Figure 1). Type I was similar to a normal spine curvature, defined as an apex of the thoracic kyphotic curve located in the middle thoracic spine, indicating that the whole thoracolumbar region was still below the apex of the thoracic kyphosis (Figure 2). Type II was defined as an apex of the thoracic kyphotic curve located in the thoracolumbar region, indicating that the middle thoracic region was replaced by part of the thoracolumbar region as the apex of the thoracic kyphosis (Figure 3). Thirty-six patients were classified as type I, and 34 patients were classified as type II.
Details and comparison of the two subtypes
In type I patients, the apex of the thoracic kyphotic curve was at T5 in five cases, T6 in 11 cases, T7 in 12 cases, and T8 in eight cases. The apex of the lumbar lordotic curve was at L3 in one case, L3/4 in three cases, L4 in 16 cases, L4/5 in 11 cases, and L5 in five cases. The inflexion point was at T12 in 11 cases, L1 in 17 cases, and L2 in eight cases.
In type II patients, the apex of the thoracic kyphotic curve was at T10 in four cases, T10/T11 in three cases, T11 in 10 cases, T11/T12 in nine cases, T12 in two cases, T12/L1 in four cases, and L1 in two cases. The apex of the lumbar lordotic curve was at L3/4 in one case, L4 in seven cases, L4/5 in 14 cases, and L5 in 12 cases. The inflexion point was T12 in two cases, L1 in 13 cases, L2 in 16 cases and L3 in three cases.
There was a statistically significant difference in TLK (p<0.001) and LL (p=0.033) between the two subtypes, but there were no differences for other variants (Table 2).
Table 2 Comparison of parameters between subtypes
Parameters
|
Type I
|
Type II
|
p value
|
Age (yr)
|
47.4±13.7 (25-69)
|
43.7±15.4 (19-73)
|
0.242
|
BMI
|
26.1±4.4 (20.4-38.0)
|
26.9±4.2 (19.9-35.2)
|
0.381
|
DM (Yes/No)
|
2/34
|
4/30
|
0.463
|
Smoking (Yes/No)
|
3/33
|
5/29
|
0.307
|
Alcohol (Yes/No)
|
4/32
|
4/30
|
0.877
|
SVA(mm)
|
25.1±16.5 (-31.5-71.2)
|
30.9±31.7 (-46.9-73.1)
|
0.738
|
CVA(mm)
|
7.5±4.4 (2.1-21.1)
|
8.5±6.2 (2.3-26.1)
|
0.892
|
TK(°)
|
28.9±11.2 (0.8-53.9)
|
29.2±13.5 (3.4-56.0)
|
0.796
|
TLK(°)
|
14.9±7.9 (0.9-31.5)
|
29.1±13.7 (3.4-57.0)
|
<0.001**
|
LL(°)
|
51.4±10.9 (24.4-68.6)
|
43.3±17.5 (7.5-71.4)
|
0.033*
|
T1PA(°)
|
8.2±6.1 (0.3-23.6)
|
8.3±8.5 (0.4-43.3)
|
0.573
|
L1PA(°)
|
6.0±4.7 (0.3-19.6)
|
6.4±5.2 (0.2-19.1)
|
0.930
|
PI(°)
|
47.1±7.7 (30.7-63.6)
|
44.8±7.3 (32.5-60.2)
|
0.206
|
SS(°)
|
35.7±6.8 (20.7-51.1)
|
33.3±8.8 (14.1-53.1)
|
0.247
|
PT(°)
|
11.4±9.3 (0.2-30.5)
|
11.4±9.5(0.5-38.3)
|
0.507
|
PI-LL(°)
|
10.4±8.1 (0.5-36.3)
|
14.5±10.3 (0.6-44.9)
|
0.111
|
** p<0.01, * p<0.05
Association between TLDH level and regional curvatures
In type I patients, TLDH-I was recognized in 23 patients, TLDH-A in 0 patients, and TLDH-O in 13 patients. In type II patients, TLDH-I was recognised in three patients, TLDH-A in 29 patients, and TLDH-O in two patients. The Fisher’s exact test showed that TLDH-I was more frequently observed in type I patients (p<0.001), while TLDH-A was more frequently observed in type II patients (p<0.001).
Concordance with the Roussouly classification
The distribution of match/mismatch ratios of the Roussouly classification in each subtype are shown in Table 3. In type I and II patients, 58.3% and 73.5% were matched with the Roussouly classification, respectively. There was a statistically significant difference between the two subtypes (p<0.001).
Table 3 Concordance of Roussouly classfication
Roussouly Classification
|
Match/Mismatch (Type I)
|
Match/Mismatch (Type II)
|
p value
|
I
|
5/0
|
9/3
|
-
|
II
|
4/7
|
10/4
|
-
|
III
|
12/4
|
7/0
|
-
|
IV
|
0/4
|
1/0
|
-
|
Total
|
21/15
|
27/7
|
<0.001**
|
** p<0.01
Table 4 Rank sum test of sagittal modifiers (SVA, T1PA, L1PA, T1ST)
Parameters
|
JOA
|
ODI
|
VAS for back
|
VAS for leg
|
SVA
|
0.001**
|
0.165
|
0.835
|
0.066
|
T1PA
|
0.018*
|
0.001**
|
0.091
|
0.069
|
L1PA
|
0.155
|
0.033*
|
0.242
|
0.121
|
T1ST
|
0.005**
|
0.337
|
0.492
|
0.081
|
** p<0.01, * p<0.05