Adult Spinal Deformity What’s new in treatment? Emre Acaroglu MD Ankara Spine center Ankara, Turkey

Scheme What is the best treatment? Do complications affect results? A new score to decrease the rate of mechanical complications

European Spine Study Group Identifying The Best Treatment in Adult Spinal Deformity: A Decision Analysis Approach Emre Acaroglu1, Aysun Cetinyurek Yavuz2, Umit Ozgur Guler1, Selcen Yuksel3, Yasemin Yavuz4, Selim Ayhan1, Montse Domingo-Sabat5, Ferran Pellise6, Francisco Javier Sanchez Perez-Grueso7, Ahmet Alanay8, Ibrahim Obeid9, Frank Kleinstück10, European Spine Study Group-ESSG 1 Ankara Spine Center, Ankara, Turkey 2 Biostatistics, ClinIST-eu, Nijmegen, Netherlands 3 Biostatistics, Yildirim Beyazit University School of Medicine, Ankara, Turkey 4 Biostatistics, Ankara University School of Medicine, Ankara, Turkey 5 Fundacio Institut de Recerca Vall d’Hebron, Barcelona, Spain 6 Spine Unit, Hospital Universitari Vall d’Hebron, Barcelona, Spain 7 Spine Unit, Hospital Universitari La Paz, Madrid, Spain 8 Comprehensive Spine Center, Acibadem Maslak Hospital, Istanbul, Turkey 9 Spine Unit, Bordeaux University Hospital, Bordeaux, France 10 Spine Center, Schulthess Klinik, Zürich, Switzerland 12/7/2018 European Spine Study Group

European Spine Study Group Purpose To identify the best (optimum) treatment in ASD Using a statistical “decision analysis” model Hypotheses: Surgery may yield better overall clinical results Surgery is associated with more complications and overall burden To construct a statistical decision analysis (DA) model to identify the optimum overall treatment in ASD. 12/7/2018 European Spine Study Group

European Spine Study Group Patients and Methods International multicentre prospective database ASD Patients Overall n=968 Included n=535 (1 year f-up) non-surgical (NS): 371 surgical (S): 164 From an international multicentre database of ASD patients with a total of 968 pts at the time of analysis, 535 who had completed 1 year follow-up (371 non-surgical –NS), 164 surgical –S), constitute the population of this study. 12/7/2018 European Spine Study Group

European Spine Study Group Patients and Methods Decision Analysis Baseline Outcome probabilities improvement (↓ in ODI > 8pts) no change deterioration (↑ in ODI > 8pts) Death / complete paralysis The values of preference – utilities Combine information and assign quality adjusted life expectancy (QALE) No improvement DA was structured in two main steps of: 1) Baseline analysis (Assessing the probabilities of outcomes, Assessing the values of preference –utilities-, Combining information on probability and utility and assigning the quality adjusted life expectancy (QALE) for each treatment) and 2) Sensitivity analysis. 12/7/2018 European Spine Study Group

European Spine Study Group Results Pooled outcome probabilities of patients (n=472) Surgery presented with a higher chance of improvement (54.2%) compared to conservative treatment 12/7/2018 European Spine Study Group

Life Expectancy (default) Results Baseline and final utilities; QALE of patients Treatment (all) Baseline Utility Final Utility p (baseline vs. final) Life Expectancy (default) QALE (healthy years) Surgery 0.56 0.60 <0.0001 100 60 Non-Surgical 0.65 0.2692 65 0.0038 The non-surgical group demonstrated better final outcomes although this group had also started with higher QALE. There were improvements in overall QALE in both groups but this was significant only in the surgical group. 12/7/2018 European Spine Study Group

European Spine Study Group Conclusions No single best treatment Conservative treatment Higher QALE (up to 6%) Secondary to higher baseline QALE Surgical treatment Significantly higher increase in QALE Chances of improvement at first year significantly better than non-surgical  This study demonstrated that a single best treatment modality for ASD may not exist. Conservative treatment appears to yield higher (up to 6%) QALE when compared to surgery, probably secondary to a higher baseline QALE. On the other hand, surgery provides a significantly higher increase in QALE and chances of improvement at 1st year 12/7/2018 European Spine Study Group

Effect of Treatment Complications on Outcomes in Adult Spinal Deformity: A Decision Analysis Approach Emre Acaroglu1, Umit Ozgur Guler1, Aysun Cetinyurek Yavuz2, Selcen Yuksel3, Yasemin Yavuz4, Selim Ayhan1, Montse Domingo-Sabat5, Ferran Pellise6, Francisco Javier Sanchez Perez-Grueso7, Ahmet Alanay8, Ibrahim Obeid9, Frank Kleinstück10, European Spine Study Group-ESSG 1 Ankara Spine Center, Ankara, Turkey 2 Biostatistics, ClinIST-eu, Nijmegen, Netherlands 3 Biostatistics, Yildirim Beyazit University School of Medicine, Ankara, Turkey 4 Biostatistics, Ankara University School of Medicine, Ankara, Turkey 5 Fundacio Institut de Recerca Vall d’Hebron, Barcelona, Spain 6 Spine Unit, Hospital Universitari Vall d’Hebron, Barcelona, Spain 7 Spine Unit, Hospital Universitari La Paz, Madrid, Spain 8 Comprehensive Spine Center, Acibadem Maslak Hospital, Istanbul, Turkey 9 Spine Unit, Bordeaux University Hospital, Bordeaux, France 10 Spine Center, Schulthess Klinik, Zürich, Switzerland

Results Number of complications associated with each treatment arm Surgery n (%) Non-surgical Total None 112 (68.3) 330 (89.0) 442 (82.6) NLT 39 (23.8) 39 (10.5) 78 (14.6) LT 10 (6.1) 2 (0.5) 12 (2.2) Death 3 (1.8) 3 (0.6) Total Complications 52 (31.7) 41 (11.1) 535 All 535 patients (371 NS, 164 S) could be analysed in regard to complications. Overall, there were 78 NLT and 12 LT complications and 3 death/paralysis. Surgical treatment was significantly more prone to complications (31.7% vs. 11.1%, p<0.001) p<0.001

European Spine Study Group Results Distribution of outcomes by complications Surgery Non-Surgical n Detoriorate n (%) No Change Improve None 112 8 (9.4) 27 (31.8) 50 (58.8) 330 47 (17.5) 184 (68.7) 37 (13.8) NLT 39 2 (7.4) 13 (48.1) 12 (44.4) 6 (15.4) 26 (66.7) 2 (17.9) LT 10 1 (12.5) 4 (50) 3 (37.5) 2 1 (50) On the other hand, presence of complications did not necessarily decrease the chances of improvement, surgical patients tending to rate better in this respect (Table 1b). 02-04.September.2015 European Spine Study Group

European Spine Study Group Conclusions Surgical treatment of ASD more likely to cause complications Presence of complications Negative impact on clinical improvement Do not affect QALE at the first year  This study has demonstrated that surgical treatment of ASD is more likely to cause complications compared to non-surgical treatment. On the other hand, presence of complications although has a negative impact on the likelihood of clinical improvement, it does not affects the QALE at the first year detrimentally. 02-04.September.2015 European Spine Study Group

European Spine Study Group Impact of Resolved Early Major Complications on Two-Year Outcome Following Adult Spine Deformity Surgery Núñez S, Pellisé F, Vila A, Domingo M, Haddad S, Sánchez Pérez-Grueso F, Acaroglu E, Alanay A, Obeid I, Kleinstück F European Spine Study Group

Aim To investigate the impact of early major complications, considered to be resolved at 6 months, on PROMs and on the likeness of reaching MCID two-years after adult deformity surgery.

Baseline demographics CG (151) MCG (24)   Mean (SD) Age 48.5 (19.1) 62.3 (16.8) n (%) Female 118 (78.2%) 13 (54.2%) Previous Surgery 35 (23.2%) 12 (50%) ASA≥2 81 (53.6%) 18 (75%)

Baseline deformity parameters CG (151) MCG (24)   Mean (SD) Major Cobb 41.28 (23.4) 38.78 (21.1) SVA 21.82 (55.5) 72.69 (69.4) Global Tilt 21.02 (15.8) 31.0 (17.2) PI 52.26 (11.1) 54.25 (10.6) PT 18.89 (10.5) 23.83 (9.7) SS 33.27 (10.8) 30.42 (10.2) LL -47.56 (21.0) -35.32 (18.8)

Baseline HRQOL CG mean (SD) MCG mean (SD) ODI 37.69 (20.6)   CG mean (SD) MCG mean (SD) ODI 37.69 (20.6) 44.75 (14.1) SF-36 PCS 36.19 (9.3) 32.98 (6.3) SF-36 MCS 43.98 (11.4) 42.13 (10.9) SRS-22function 3.15 (0.91) 2.95 (0.93) SRS-22mental health 3.24 (0.8) 3.18 (0.92) SRS-22pain 2.75 (0.98) 2.65 (0.89) SRS-22satisfaction 3.0 (1.1) 3.0 (1.0) SRS-22self image 2.53 (0.73) 2.43 (0.63) SRS-22subtotal 2.91 (0.7) 2.79 (0.7)

Surgical characteristics   CG mean (SD) MCG mean (SD) Estimated Blood Loss 1395.9 (1155.8) 1880.4 (1225.7) Surgical Time 310.1 (174.2) 364.0 (148.7) CG n (%) MCG n (%) 3 Column Osteotomies 25 (16.6%) 5 (20.8%) Patients on the revision group had longer, more complex surgeries

Major Complications 27 major complications in 24 patients, 18 reoperations + 1 cholecystectomy

Change Over Time and MCID/SCB Control Group Baseline 2 years Change MCID SCB ODI 37.68 23.1 -14.58 -12.8 -18.8 SF-36 PCS 36.44 43.8 7.36 4.9 6.2 SRS-22 subtotal 2.91 3.67 0.76 0.43 0.69 MCID for ODI, SF-36 Copay 2008 J Spine, SCB Glassman 2008 JBJS MCID for SRS-22 Crawford III 2015 Spine, SCB Crawford 2016 Spine Deformity

Change Over Time and MCID/SCB Major Complication Group Baseline 2 years Change MCID SCB ODI 44.75 32.92 -11.83 -12.8 -18.8 SF-36 PCS 33.48 38.49 5.31 4.9 6.2 SRS-22 subtotal 2.82 3.45 0.63 0.43 0.69 MCID for ODI, SF-36 Copay 2008 J Spine, SCB Glassman 2008 JBJS MCID for SRS-22 Crawford III 2015 Spine, SCB Crawford 2016 Spine Deformity

Conclusions Patients with major complications; If resolved six months after surgery, achieved a significant improvement reaching MCID values for SF-36 and SRS-22 at two years.

A score to predict/prevent mechanical complications

Background: Sagittal Plane Analysis There is a need for a new look into the ‘ideal’ sagittal plane Spinal curvatures and alignment must be viewed in light of each other Chain of correlations PI influences SS SS influences LL LL influences TK TK influences CL Pelvic incidence is a (relatively) constant morphological parameter that describes the ‘pelvic size’ for any given person PI = A signature All sagittal plane parameters Should be evaluated Proportional to PI (rather than absolute numeric) To assess disproportion compared with the calculated ideal

Methods From the ESSG database 222 patients randomly assigned to ≥4 levels posterior fusion ≥2 years follow up 222 patients (168F, 54M) were included Mechanical Complications PJK / PJF DJK Rod breakage Implant related complications Screw loosening, fracture, pull out Interbody, hook or set screw pull out 222 patients randomly assigned to derivation (n = 148, 66.7%) and validation (n = 74, 33.3%) cohorts Mean age 52.2 ± 19.3 (range 18-84) Mean follow-up 28.8 ± 8.2 (24-62) months

Global Alignment & Proportion : GAP Score New Method of Analyzing Sagittal Plane Offers individualized sagittal plane analysis Instead of population norms & mean values Uses PI-based proportional radiographic parameters Instead of absolute numerical values Denotes “normal” and “pathologic” standing sagittal alignment and shape as a single score for every magnitude of pelvic incidence. Radiographic parameters RPV: Relative Pelvic Version (Measured-Ideal SS) RLL: Relative Lumbar Lordosis (Measured-Ideal LL) LDI : Lordosis Distribution Index (L4-S1 / L1 – S1 x100) RSA : Relative Spinopelvic Alignment (Measured-Ideal GT) Age Factor

Mechanical Complication Results   Mechanical Complication - + n (%) 2 P RPV Anteversion 5 (62.5) 3 (37.5) 2(3)=39.1 0.000 Aligned 50 (78.1) 14 (21.9) Moderate Retroversion 22 (44) 28 (56) Severe Retroversion 3 (11.5) 23 (88.5) RLL Hyperlordosis 1 (16.7) 5 (83.3) 2(3)=37.4 63 (75) 21 (25) Moderate Hypolordosis 13 (36.1) 23 (63.9) Severe Hypolordosis 3 (13.6) 19 (86.4) LDI Hyperlordotic Maldistribution 3 (10.7) 25 (89.3) 2(3)=36.2 69 (71.1) 28 (28.9) Moderate Hypolordotic Maldistribution 6 (37.5) 10 (62.5) Severe Hypolordotic Maldistribution 2 (28.6) 5 (71.4) RSA Negative Malalignment 2 (50) 2(3)=43.8 44 (77.2) 13 (22.8) Moderate Positive Malalignment 30 (62.5) 18 (37.5) Severe Positive Malalignment 4 (10.3) 35 (89.7) Age <60 years 51 (66.2) 26 (33.8) 2(1)=9.6 0.002 ≥60 years 29 (40.8) 42 (59.2) Risk Factors   β regression coefficient (SE) OR (95% CI) P Statistical Weights RPV Anteversion 0.762 (0.790) 2.1 (0.5 – 10.1) 0.335 1 Aligned - Moderate Retroversion 1.514 (0.415) 4.5 (2.0 – 10.3) 0.000 2 Severe Retroversion 3.310 (0.684) 27.4 (7.2 – 104.7) 3 RLL Hyperlordosis 2.708 (1.124) 15 (1.7 – 135.8) 0.016 Moderate Hypolordosis 1.669 (0.429) 5.3 (2.3-12.3) Severe Hypolordosis 2.944 (0.670) 19 (5.1-70.7) LDI Hyperlordotic Maldistribution 3.022 (0.655) 20.5 (5.7-73.5) Moderate Hypolordotic Maldistribution 1.413 (0.563) 4.1 (1.4-12.4) 0.012 Severe Hypolordotic Maldistribution 1.818 (0.866) 6.2 (1.1-33.6) 0.036 RSA Negative Malalignment 1.219 (1.049) 3.4 (0.4 – 26.4) 0.245 Moderate Positive Malalignment 0.708 (0.434) 2.0 (0.9 – 4.7) 0.103 Severe Positive Malalignment 3.388 (0.615) 29.6 (8.8 – 98.9) Age <60 years ≥60 years 1.044 (0.341) 2.8 (1.4-5.5) 0.002

GAP Score The GAP score, calculated by adding the scores for relative pelvic version, relative lumbar lordosis, lordosis distribution index, relative spinopelvic alignment, and the age factor, ranged from 0 to 13 points. A GAP score of 0 to 2 was categorized as indicating a proportioned spinopelvic state; 3 to 6, as moderately disproportioned; and ≥7, as severely disproportioned.

GAP SCORE (Global Alignment and Proportion) PARAMETERS SCORING CATEGORIES Relative Pelvic Version ( RPV = Measured - Ideal Sacral Slope * ) RPV Subgroups Score The parameters column includes a scale-based view of the subgroups of each parameter using the cutoff points. The scoring column includes the statistical weights of the parameter subgroups. The categories column includes the categorization of the GAP score. Severe Retroversion Moderate Retroversion Aligned Anteversion < -15 : -15 – -7.1 : -7 – 5 : > 5 : Severe Retroversion.................................. Moderate Retroversion............................ Aligned...................................................... Anteversion............................................... 3 2 1 -15o -7o +5o * Ideal Sacral Slope = PI x 0.59 + 9 Total Score : 0 – 2 Relative Lumbar Lordosis ( RLL = Measured - Ideal Lumbar Lordosis ✢ ) RLL Subgroups Proportioned Severe Hypolordosis Moderate Hypolordosis Aligned Hyperlordosis < -25 : -25 – -14.1 : -14 – 11 : > 11 : Severe Hypolordosis.................................. Moderate Hypolordosis............................ Aligned...................................................... Hyperlordosis............................................. 3 2 -25o -14o +11o ✢ Ideal Lumbar Lordosis = PI x 0.62 + 29 Total Score : 3 - 6 Lordosis Distribution Index ( LDI = L4-S1 Lordosis / L1- S1 Lordosis x100 ) LDI Subgroups Moderately Disproportioned Severe Hypolordotic Moderate Hypolordotic Aligned Hyperlordotic < 40 % : 40 – 49 % : 50 – 80 % : > 80 % : Severe Hypolordotic Maldistribution........ Moderate Hypolordotic Maldistribution.. Aligned....................................................... Hyperlordotic Maldistribution................... 2 1 3 40% 50% 80% Total Score  7 Relative Spinopelvic Alignment ( RSA = Measured - Ideal Global Tilt † ) RSA Subgroups Severely Disproportioned Severe Positive Moderate Positive Aligned Negative > 18 : 18 – 10.1 : 10 – -7 : < -7 : Severe Positive Malalignment……………. Moderate Positive Malalignment……….. Aligned……………...................................... Negative Malalignment............................ 3 1 +18o + 10o -7o † Ideal Global Tilt = PI x 0.48 – 15 Age Factor Age Subgroups Elderly Adult < 60 years : ≥ 60 years : Adult…………………………………...... Elderly Adult...…….............................. 1 60

Mechanical Complication & Revision Rates GAP Score in predicting mechanical complications GAP Score GAP Categories

Conclusion GAP score is a new PI-based proportional method of analyzing the individualized sagittal plane is an all-inclusive single score that offers a ‘one-size fits all’ solution for every size of pelvic incidence Preoperative planning & setting surgical goals in the sagittal plane on the basis of the individualized proportional indices reflected by the GAP score may decrease the rate of mechanical complications.

Summary What’s new in ASD treatment? Surgery has been shown to be a better treatment modality over non-surgery Within indications Surgery has been shown to be more prone to complications That affect outcomes Yet provide significant HRQoL improvements GAP score appears to be highly predictive of mechanical failures A quantification of the ‘proportional’ sagittal alignment May potentially be used as a planning tool