Adult spinal deformity covers a broad range of pathologies and can be responsible for disability and altered quality of life [
ASD is commonly described with the SRS-Schwab classification [
The key parameter to evaluate postoperative alignment is the PI-LL mismatch. This parameter has been reported as a predictor of success and is correlated with health-related quality of life scores [ The L4-S1 DLL was reported to be almost constant Pelvic incidence is significantly related to the L1-L4 PLL
It seems therefore crucial to evaluate not only the global amount of LL restoration but also the PI-LL mismatch according to PLL and DLL.
This study aimed to analyze the results of ASD surgical management according to postoperative PI-LL mismatch with the hypothesis that aligned patients had a better restoration of the DLL.
After institutional review board approval (IRB 00009118), we conducted a retrospective analysis of a prospectively maintained database of ASD patients operated between 2015 and 2018. Prior to inclusion, every patient signed an informed consent.
Inclusion criteria were as follows: all adult patients managed surgically for ASD with a posterior fixation that included all the lumbar spine (upper level instrumented vertebra L1 or above, lower instrumented vertebra S1 or below), primary surgery or revision case, pre and postoperative full spine X-rays available, and one-year minimal follow-up (including full spine X-rays).
For each patient, demographic data, clinical scores (lumbar VAS and Oswestry Disability Index), and full spine X-rays (AP and lateral) were obtained preoperatively, in the immediate postoperative period, and at one-year of follow-up. Complications during the follow-up period were systematically noted.
The surgical procedure consisted of a posterior fusion and bone resection if required (Grade 2 or 3 of the Schwab classification [
On lateral radiographs, the following parameters were measured: T4-T12 thoracic kyphosis (TK), L1-S1 lumbar lordosis, L1-L4 PLL, L4-S1 DLL, SVA, pelvic parameters (PI and PT), and mismatch between pelvic incidence and lumbar lordosis (PI-LL).
Data were formulated as means and standard deviations. Comparisons were carried out using Student’s
Among the 97 patients who underwent ASD surgical correction in our institution during the inclusion period, 77 patients met inclusion criteria for the present study. Fifty-six women and 21 men with a mean age of 66.5 years old (SD = 8.8) were included. On average, 14 levels were fused during the procedure (SD = 3), and lower instrumented level included iliac screws for 43 patients and S1 screws for 34 patients.
On the whole series, the mean preoperative PI was 54.3° SD = 13 and did not change during follow-up.
With regards to regional parameters, a significant improvement of LL and TK was noted between preoperative and last follow-up evaluation (28.2° vs. 43.5°,
SVA and PI-LL mismatch were significantly reduced (81.9 mm vs. 50.9 mm,
Thirty-five patients were classified as “aligned” (PI-LL <10°) and 42 were classified as “not aligned” (PI-LL >10°). Of the 42 “not aligned” patients, 17 were revision cases, 10 requiring a grade 3 osteotomy (8 in L3 and 2 in L4). Among the 35 “aligned” patients, 19 were revision cases, 7 requiring a grade 3 osteotomy (5 in L3 and 2 in L4) (Figures
Clinical example of a patient from the “aligned group” with a low PI. Primary case, T9-S1 posterior fixation, multiples grade 2 osteotomies. Preoperative measurements (left) were PI = 41°, PT = 21°, PI-LL = 21, LL = -20°, TK = 22°, and SVA = 42 mm. One-year measurements (right) were PT = 17°, PI-LL = 0, LL = −41°, PLL = −8°, DLL = −33°, TK = 41°, and SVA = 41 mm.
Clinical example of a patient form the “not aligned” group with a high PI. Revision case, T3-S2 posterior fixation, L3 grade 3 osteotomy. Preoperative measurements (left) were PI = 59°, PT = 34°, PI-LL = 32°, LL = −27°, TK = 3°, and SVA = 93 mm. One-year measurements (right) were PT = 25°, PI-LL = 15°, LL = −44°, PLL = −29°, DLL = −15°, TK = 36°, and SVA = 91 mm.
Preoperatively, “not aligned” patients had a significantly higher pelvic incidence, higher pelvic tilt, smaller lumbar lordosis, and thoracic kyphosis, despite a nonsignificantly different sagittal vertical axis (Table
Preoperative values of radiographic parameters between “aligned” and “not aligned” groups.
Preoperative | Aligned | Not aligned | |
---|---|---|---|
Pelvic incidence (°) | 52 | 61 | 0.009 |
Pelvic tilt (°) | 26 | 33 | 0.001 |
L1-S1 lumbar lordosis (°) | 34 | 23 | 0.004 |
L1-L4 PLL (°) | 1.5 | 0.8 | 0.9 |
L4-S1 DLL (°) | 33.3 | 28.3 | 0.254 |
PI-LL (°) | 16 | 34 | <0.001 |
T4-T12 thoracic kyphosis (°) | 37 | 25 | 0.008 |
Sagittal vertical axis (mm) | 71 | 91 | >0.05 |
With regards to preoperative LL distribution, the L1-L4 PLL (1.5° vs. 08°,
Postoperatively, “aligned” patients showed a significantly higher improvement of LL, TK, and SVA (Table
Postoperative values of radiographic parameters between “aligned” and “not aligned” groups.
Postoperative | Aligned | Not aligned | |
---|---|---|---|
Pelvic incidence (°) | 52 | 61 | 0.009 |
Pelvic tilt (°) | 24 | 33 | <0.001 |
L1-S1 lumbar lordosis (°) | 49 | 37 | <0.001 |
L1-L4 PLL (°) | 18 | 16 | 0.39 |
L4-S1 DLL (°) | 31 | 22 | 0.003 |
PI-LL (°) | 4 | 23 | <0.001 |
T4-T12 thoracic kyphosis (°) | 48 | 40 | 0.002 |
Sagittal vertical axis (mm) | 23 | 63 | <0.001 |
With regards to postoperative LL distribution, the L1-L4 PLL was not significantly different between groups (18° vs. 16°
The PI-LL mismatch was significantly correlated with the L4-S1 DLL (rho = 0.407,
During the follow-up period, 7 postoperative infections were noted (4 in the “not aligned” group and 3 in the “aligned” group) that required surgical debridement and adapted antibiotics.
At last follow-up, 1 patient out of 35 (3%) in the aligned group had a mechanical complication (rod breakage) that required revision surgery. Six patients out of 42 (14%) of the “not aligned” group had a mechanical complication that required revision surgery (4 rod breakage and 2 proximal junctional failure).
With regards to final follow-up clinical scores, patients from the “aligned” group had a significantly lower lumbar VAS (2.8/10 vs. 4.8/10,
Surgical management of ASD remains challenging, and optimal management is still under debate. The complication rate after realignment procedures has been reported around 16.5% at 2 years of follow-up and 50% at 10 years of follow-up [
According to the literature, a large proportion (up to 50%) of patients remains undercorrected after surgery [
Restoration of pelvic tilt might therefore be a crucial point for sagittal realignment procedures, especially in patients with a high pelvic incidence. Lafage et al. [
Another important parameter for realignment procedures is the postoperative location of the LL apex. Based on Roussouly classification, Pizones et al. recently reported a decrease in mechanical complications with an adapted postoperative lumbar apex position and lumbar shape restoration [
In our study, the choice was made to stratify patients according to postoperative PI-LL mismatch. Our results revealed that postoperatively “aligned” patients had a preservation of L4-S1 DLL and a higher pelvic incidence when compared to “not aligned” patients. These results are consistent with the study of Ylgor et al. who suggested taking into account lumbar lordosis distribution [
Results of this study confirm that restoration of L4-S1 DLL is a crucial objective for ASD patients as recently reported by Lafage et al. [
This study presents several limitations such as the one-year follow-up and a limited number of patients included. Further studies will be needed to confirm these results.
Adult spinal deformity is a frequent and challenging condition for spine physicians. Based on postoperative PI-LL mismatch, the results of this study revealed that maintenance or restoration of L4-S1 DLL is crucial for postoperative alignment. Specific attention must be paid to restore optimal distal lumbar lordosis in order to decrease the rate of undercorrected patients and to improve outcomes.
The data used to support the findings of this study are available from the corresponding author upon request.
The study was approved by IRB 00009118.
Each patient signed an informed consent prior to inclusion.
B Blondel is a consultant at Medicrea International, Implanet, 3 M, Vexim-Stryker. S Fuentes is a consultant at Medicrea International, Medtronic, Stryker. P Tropiano is a consultant at France Rachis, Depuy Synthes, LDR Zimmer. S Prost, S Pesenti, A. Muñoz McCausland, and K Farah declare that they have no conflicts of interest.