First Author (Year) | Study Type | Quality Assessment | No. of Patients | Treatment for ASD | Patient Age, y, Mean ± SD | Minimum Follow-up (mo) | Technique/Factor Investigated | Results | Association With PJK Incidence |
Buell (2021)23 | Retrospective cohort | Good | 560 | Long sacropelvic fusion | 63 ± 9 | 24 | UT vs LT UIV | No difference in reoperation rates for PJK between UT group (9.8%) and LT group (8.6%) (P = 0.81) | None |
Burks (2019)24 | Retrospective case series | Fair | 36 | Hybrid MIS-open surgical fusion, in which ≥2 most rostral levels were instrumented percutaneously | 65 ± 11 | 12 | Muscle-sparing technique at the proximal end | PJK rate, 22% (n = 8); similar to reported rates; no reoperations for PJK/PJF | No control for comparison; similar to reported literature rates |
Cazzulino (2021)25 | Retrospective cohort | Fair | 39 | Fusion using a soft-landing technique | 61 ± 10 | None specified; mean follow-up 26 months | TPH | Radiographic PJK in 16/39 patients at last follow-up; 4 patients met criteria for PJF with revision; 3 cases of compression fracture at the UIV or UIV+1 | No control for comparison; similar to reported literature rates |
Cho (2013)26 | Retrospective cohort | Good | 51 | Posterior fusion; assigned to cohorts based on UIV location in relation to UEV and HV | 68 ± 6 in the adjacent segment disease group, 63 ± 6 in the control | 24 | Selection of UIV | PJK in 5 patients, 2 requiring fusion extensions/all had UIV below UEV; junctional kyphotic angles were not different between any groups | Lower incidence of PJK with higher UIV |
Daniels (2019)27 | Retrospective cohort | Good | 303 | Posterior instrumentation | 63 ± 9 | 24 | UT vs LT UIV | Lower PJK rate in UT compared with LT fusions (OR, 0.49; 95% CI, 0.24–0.99); no difference in PJF (OR, 0.54; 95% CI, 0.24–1.2) | Lower incidence of PJK with higher UIV |
Ha (2013)28 | Retrospective cohort | Good | 89 | Various, treated 2007 to 2009 | 64 ± 7 (LT); 64 ± 11 (UT) | 24 | UT vs LT UIV | PJK in 29 patients (23 in LT, 6 in UT, P = 0.61); 8 revision surgeries for PJK in LT group and 2 in UT group (P = 0.68); higher incidence of compression fracture in LT group (16/23, 70%), higher incidence of subluxation in UT group (3/6, 50%) (P = 0.014) | None |
Hassanzadeh (2013)21 | Retrospective cohort | Good | 47 | Long (≥5 levels) spinal fusion, 2004 to 2009 | 46 (TPH); 51 (PS) | 24 | TPH | PJK in 8/27 patients in PS group compared with none in the TPH group (P = 0.02) between immediate postoperative and final follow-up; 2/8 underwent revision surgery; mean PJA was 6.4° ± 10° in the TPH group and 22° ± 14° in PS group (P < 0.001) | Lower incidence of PJK with TPH |
Kaufmann (2022)29 | Prospective cohort | Good | 76 | Posterior instrumentation and fusion of ≥3 levels; 2009 to 2017 at 1 center | 64 ± 9 (MLSS); 55 ± 20 (control) | 12 | MLSS vs standard PS | PJF in 10% of MLSS group, 31% of control (P = 0.02); less kyphosis in MLSS group (5.2° ± 6.3° compared with control 1.3° ± 5.3°, P = 0.01) | Lower incidence of PJK with MLSS |
Kim (2014)9 | Retrospective cohort | Good | 198 | Long ( >5 levels) fusion, from a multicenter database | 61 (UT); 62 (LT) | 24 | UT vs LT UIV | No difference in PJK angle at 1- and 2-year follow-up (UT 14° vs LT 14° at 1 year; 17° vs 19° at 2 years); 5 patients underwent a revision for PJK (3 in UT group and 2 in LT group) (P = 0.45) | None |
Kim (2007)30 | Retrospective cohort | Good | 125 | Instrumentation and fusion | 52 ± 11 (T9-10); 57 ± 12 (T11-12); 62 ± 12 (L1-2) | 24 | UIV location (T9-10, T11-12, or L1-2) | PJK in 51% in T0-10 group, 55% in T11-12 group, 36 % in L1-2 (P = 0.2) at final follow-up; 1 revision for PJK in T11-12 group and 3 in L1-2 group (P = 0.27); final change in PJA was not different between all groups (P = 0.46) | None |
Lafage (2017)31 | Case-control | Good | 252 | Posterior fusion and instrumentation | 61 ± 10 | 24 | UT vs LT UIV | PJK in 49 % of UT UIV group vs 64 % of LT/TL group (P = 0.02); smaller UIV inclination between PJK and non-PJK groups when stratified into both UT (P = 0.005) and LT/TL (P < 0.001) groups | Lower incidence of PJK with higher UIV |
Line (2020)32 | Retrospective cohort | Good | 625 | Long (≥5 levels) spinal fusion | 62 ± NA | 12 | No proximal fixation augmentation vs various augmentation options, including TPH | PJF in 20% of no implant group, 11% of the implant group; 115 in TPH group, 7% rate of PJF (lowest of all augmentation), but 8.7% underwent surgical revision because of discrepancies in the definition for PJF used in the study (P < 0.05 for aforementioned rates of PJK) | Lower risk of PJK with TPH |
Matsumura (2018)33 | Retrospective cohort | Fair | 39 | Corrective surgery performed, 2009 to 2013 | 67 ± NA | 24 | TPH | PJK in 18% of TPH group vs 27% in PS group (P = 0.47); change in PJA greater in PS group (19°) than TPH group (5°) (P = 0.04) | Lower incidence of PJK with TPH |
O’Shaughnessy (2012)34 | Retrospective cohort | Good | 58 | Fusion including the sacrum, treated 2002 to 2006 | 55 ± 9 (UT); 56 ± 8 (LT) | 24 | UT vs LT UIV | PJK in 18% of LT group, 10% of UT group (P = 0.476); surgical PJK 2.6% in LT group, 0 in UT group (P > 0.99); no other differences in complications between groups | None |
Sandquist (2015)35 | Prospective cohort | Good | 15 | Posterior instrumentation and fusion of ≥3 levels, 2009 to 2012 (subset from a study by Kaufmann et al) | 66 ± NA | 12 | MLSS vs standard PS | Mean change in PJA was 4.0° (range, –0.92 to 9.13); no cases of PJK or PJF were recorded | Lower incidence of PJK with MLSS |
Scheer (2015)36 | Retrospective cohort | Good | 165 | PSO, from a multicenter database | 60 ± 11 (UT); 60 ± 11 (LT) | 24 | UT vs LT/TL UIV | PJK in 52% of UT group, 48% of LT/TL group (P = 0.85); 11 PJK cases requiring revision, 9/11 in TL/LT group and 2/11 in UT group (P = 0.03) | No significant association with PJK incidence, higher incidence of PJF with lower UIV |
Tsutsui (2022)37 | Retrospective cohort | Good | 53 | Fusion from pelvis to T9 or T10 | 73 ± 4 (TPH); 72 ± 4 (PS) | 12 | TPH | Higher incidence of PJK in TPH group (36%) vs PS group (8%) (P = 0.01); in TPH group, PJK caused in all cases by UIV or adjacent segment fracture with hook dislodgement | Higher incidence of PJK with TPH |
Wang (2017)38 | Retrospective cohort | Good | 242 | Posterior fusion and instrumentation of ≥4 levels, from 2004 to 2014 | 59 ± 6 (PAS); 60 ± 6 (MAS) | None; mean follow-up 25 ± 4 months | PAS vs MAS at UIV | PJK in 26/117 (22%) in MAS group, 30/125 (24) in PAS group (P = 0.73); greater change in PJA in PJK subgroup (2.9° vs 1.7°) (P = 0.03) | None |
Yoshida (2020)39 | Case-control | Good | 113 | Surgery for degenerative spinal disorders, including both ASD (n = 45) and non-ASD patients (n = 68) | 67 ± 8 (ASD); 57 ± 20 (non-ASD) | 12 | UIV to C2 plumb line distance | PJK in 10/45 ASD patients; sub-analysis of PJK vs non-PJK patients showed significantly greater distances from UIV to both C7 and C2 plumb lines on standing and sitting radiographs for PJK group; on logistic regression analysis, UIV to C2 distance was found to be significantly associated with PJK (OR 1.2; 95% CI 1.0–1.3) | UIV farther from C2 plumb line associated with a higher incidence of PJK |
Abbreviations: ASD, adult spinal deformity; CI, confidence interval;HV, horizontal vertebra; LT, lower thoracic;MAS, monoaxial screw; MIS, minimally invasive surgery; MLSS, multilevel stabilization screw; NA, not available; OR, odds ratio;PAS, polyaxial screw; PJA, proximal junctional angle; PJF, proximal junctional failure; PJK, proximal junctional kyphosis; PS, pedicle screw(s);PSO, pedicle subtraction osteotomy; TL, thoracolumbar; TPH, transverse process hooks; UEV, upper-end vertebra; UIV, upper instrumented vertebra; UT, upper thoracic.