. 2024 May 3;18:100327. doi: 10.1016/j.xnsj.2024.100327

Table 10.

Details and results of studies reporting on junctional complications.

Study	Surgical procedure Levels treated (No.)	Complications considered	Radiographic follow-up	Complication rates	Summary of results	Associated Clinical Outcomes
Barton et al. [118]	Posterior or A-P (36) fusion + osteotomy Median posterior 8 (2-17), anterior 2 (1-6)	PJF: fracture or spondylolisthesis of UIV or UIV+1	Between 24 and 60 mo	11.7% (11/94)	– Osteoporosis/osteopenia (DXA or ultrasound) an independent risk factor for PJF in 5+ level fusions (OR 10.4, p=.039)	– All PJF symptomatic and required revision (OR >19, p<.0001)
Chen et al., 2011 [60]	L4/5 PLIF	Progression of L3/4 degeneration: 1) disc height >3mm; 2) dynamic angulation >5°; 3) L3 slippage >3mm	Between 24 and 52 mo (at final follow-up)	22.01% (24/109)	– Mean lumbar T-scores: −1.23±0.23 (degeneration group) vs. −1.12±0.19 (no degeneration group), p=.08	– No significant differences in ODI or JOA based on BMD or degeneration
Duan et al. [53]	Long posterior fusion from T9-12 to sacrum	PJK^*	< 1 month and at final follow-up	53.7% (29/54)	– Patients with PJK had lower HU at the UIV (120.41 vs. 152.8, p=.011), UIV+1 (124.52 vs. 155.96, p=.02), and UIV+2 (129.28 vs. 160, p=.018) – ROC analysis: optimal HU cutoffs at the UIV, UIV+1, and UIV+2 were 104 (AUC 0.710), 113 (AUC 0.679), and 110 (AUC 0.681) – Higher rates of PJK in patients with HU < 110 (73.9% vs. 38.7%, p=.014)	N/A
Ha et al. [45]	Long posterior fusion to L5 or S1 PJF 6.1±1.1 no PJF, 6.6±1.5	Acute PJF ^b	Mean time to PJF 23.4±29.9 mo, median 8 mo (1–88)	11.5% (18/157)	– Presented differential risk profiles for PJF secondary to UIV fracture (n=5), UIV+1 fracture (n=6), UIV fixation failure (n=4), and junctional subluxation (n=3) – Mean lowest T-score: −3.3±1.1 (PJF group) vs. −1.9±1.5 (no PJF group), p<.001 – Lowest T-score was an independent risk factor for PJF (HR 0.64, p=.021)	– All patients with PJF had pain or deficits, 6 required revision.
Hiyama et al. [54]	Staged: 1) 2-4 level LLIF, 2) long posterior fusion with L5/S1 TLIF Mean 9.7±2.5	PJF: any symptomatic PJK requiring revision	1 year; mean time to revision 18.4±13.9 mo	25% (13/52)	– Mean UIV HU: 116.6±28.1 (PJF group) vs.141.8±41.8 (no PJF group), p=.049 – No significant differences in HU at the UIV+1 (p=.342) or UIV+2 (p=.787)	N/A
Hyun et al. [121]	Long posterior or AP (20) fusion with T9-L2 UIV PJK 5.6±1.4, no PJK 5.6±1.3	PJK	NR	38.6% (17/44)	– Mean T-scores: −2.5±1.2 (PJK group) vs −1.3±1.3 (no PJK group), p=.003 – Osteoporosis (T-score <-2.5) an independent risk factor for PJK (HR 2.73, p<.001)	– Lower SRS pain sub scores in PJK (p<.05), but no differences overall
Kim et al. [33]	Long posterior or AP (218) fusion	PJK: PJA > 10°	1-2 mo, 2 y, and at final follow-up	39.5% (144/364)	– Higher rates of osteoporosis in patients with PJK (20.4% vs. 9.8%, p=.02)	– Upper back pain highly predictive of PJK (OR 12.5; p<.01)
Kim et al. [125]	Long posterior or AP (32) fusion from T10-L2 to L5 or S1	PJK: angle change of >10° on dynamic x-rays	NR	32.65% (16/49)	– Mean T-scores: −2.30±0.85 (PJK group) vs. −1.01±0.67 (no PJK group), p=.027	N/A
Kuo et al. [59]	Thoracolumbar fusion	PJK and PJF requiring revision	NR	29.3% (34/116): PJK 24.1% (28), PJF 8.6% (10)	– Mean VBQ scores: 3.13±0.46 (PJF group) vs. 2.46±0.49 (no PJF group), p<.001 – VBQ score was the only independent risk factor for PJF (OR 1.74, p<.001) – ROC analysis: VBQ of 2.85 best predicted PJF (AUC 0.943) – VBQ score strongly correlated with PJA measurements (r = 0.786) – PJF developed in 26/29 (89.6%) with VBQ > 2.85 vs. 3/116 (2.5%) with VBQ < 2.85	NA
Kurra et al. [47]	Long fusion to pelvis Mean 10.7 (5-17)	PJK	NR	35.8% (33/92): PJK 23.9% (22), VCF excluding PJK (11)	– Mean UIV-1 HU: 131±40 (VCF group), 158±55 (PJK group), 159±45 (no PJK group) – Mean UIV+1 HU: 127±28 (VCF group), 152±50 (PJK group), 162±54 (no PJK group) – Mean UIV+1 HU: 126±33 (VCF group), 162±51 (PJK group), 171±50 (no PJK group) – No significant HU differences associated with PJK in the absence of VCF	N/A
Mikula et al. [55]	Long instrumented fusion from T10-L2 to pelvis	PJF: PJK requiring revision	Mean time to PJK 22±18 mo and PJF 19±18 mo	PJK/PJF 31.33% (47/150)	– Mean UIV/UIV+1 HU: 120 (PJK/PJF group) vs. 149 (no PJK group), p<.001 – Mean FN T-score: -1.5±1.0 (PJK/PJF group) vs. 1.0±1.0 (no PJK group), p<.05 – UIV/UIV+1 HU was the only independent risk factor for PJK (UOR 0.94, p=.031) – ROC analysis: optimal HU cutoff of 122 at UIV/UIV+1 for predicting PJK (AUC 0.89) – PJK rates for HU < 110, 110-160, and >160 were 63%, 27%, and 12% (p<.001)	N/A
Mikula et al. [56]	Long instrumented fusion from T1-T6 to pelvis	PJF: PJK requiring revision	Mean time to PJK 22 mo, PJF 14 mo	PJK/PJF 33% (27/81): PJK 26% (21), PJF 19% (15)	– Mean UIV/UIV+1 HU: 148±43 (PJK/PJF group) vs. 192±47 (no PJK group), p=.001 – Mean L3/4 HU: 91±26 (PJK/PJF group) vs. 146±49 (no PJK group), p<.05 – Mean FN T-score: -1.7±0.85 (PJK/PJF group) vs. -1.2±0.84 (no PJK group), p<.05 – UIV/UIV+1 HU was the only independent risk factor for PJK (UOR 0.96, p=.005) – ROC analysis: optimal HU cutoff of 159 at UIV/UIV+1 for predicting PJK (AUC 0.77)	N/A
Park et al., [52]	Long posterior (24) or AP (39) fusion from T11-L1 to sacrum	PJF: PJA >20°, UIV or UIV+1 fracture, UIV fixation failure, myelopathy, or need for proximal extension	Mean time to PJF 9.3±14.1 mo (1.2–55)	36.5% (23/63)	– Higher rates of osteoporosis (DXA) in patients with PJF (43.5% vs. 20%, p=.046) – Osteoporosis an independent risk factor for PJF (OR 4.459, p=.033)	– Worse ODI and SRS-22 in PJF at last follow-up – 6 (26.1%) revisions, 3 recommended but refused
Wang H et al., 2016 [57]	Long posterior fusion from T9-L3 to L4-S1	PJK or spontaneous adjacent VCF	NR	17.3% (17/98)	– Mean T-scores: −1.4±0.8 (PJK group) vs. −0.7±0.3 (no PJK group), p<.001 – Osteoporosis (T-score < -2.5) an independent risk factor for PJK (OR 3.27, p<.001)	N/A
Wang et al. [61]	TLIF (98) or PLIF (139) 1 (176), 2 (59)	Symptomatic adjacent segment degeneration	NR	6.3% (15/237)	– Mean T-scores: -1±0.2 (degeneration group) vs. -1.2±-0.3 (no degeneration group), p=.413	N/A
Wang et al. [138]	Long instrumented fusion Median levels: PJF 5 (4-8), control 7 (4-12)	PJF: UIV or UIV+1 fracture, screw loosening or pullout at UIV	Median time to PJF 10 mo (2-45); 86.95% occurred within 2 y	22.1% (23/104)	– Mean L1 HU: 80±22.2 (PJF group) vs. 111±29.9 (no PJF group), p<.001) – ROC analysis: L1 HU cutoff of 89.25 best predicted PJF (AUC 0.799) – L1 HU ≤89.25 an independent risk factor for PJF (HR 8.98, p<.001) – Higher rates of PJF in patients with HU ≤ 89.25 (52.9% vs. 7.1%, p<.001)	N/A
Yagi et al., [140]	Anterior (14), posterior (82) or AP (61) fusion Mean 10.7 (6-15)	PJK	Final follow-up (mean 4.3 y); 75% occurred within 2 y	20% (32/157)	– Mean FN BMD: 0.691±0.194 (PJK group) vs. 0.787±0.182 (no PJK group), p=.16 – Low BMD associated with 22.9% increased risk of PJK (p=.055)	– No difference in SRS or ODI overall, but worse in symptomatic (n=6) PJK – 4 underwent revision
Yagi et al. [141]	Anterior (4), posterior (35), or AP (37) fusion PJK 10.8±3.9, no PJK 11.2±3.6	PJK	2-3 mo, 2 and 5 y, and at final follow-up; 76% occurred within 3 mo, none after 5 y	22.4% (17/76)	– Mean FN T-scores: -1.32±0.34 (PJK group) vs. -1.08±0.32 (no PJK group), p=.011 – Low BMD associated with 30.9% increased risk of PJK (p=.04).	– No significant differences in SRS or ODI in patients with PJK – 4 symptomatic, 2 underwent revision.
Yagi et al., 142]	Long thoracolumbar fusion S-group 10.2±2.3, M-group 9.8±2.4	PJF: PJA increase ≥20° with deterioration of 1+ SRS-Schwab sagittal modifier grade, or any PJK requiring revision	Within 2 y	25% (29/113) PJK, 19% (22) PJF	– Mean T-scores: −1.5±0.5 (S-group, propensity matched) vs. −0.6±0.6 (M-group), p<.001 – Higher incidence of PJF in S-group (T-score < -1.5) both before (40% vs. 4%; OR 14.3, p<.01) and after propensity-score matching (33% vs. 8%; OR 6.4, p<.01).	– 3 (2.8%) underwent revision
Yao et al. [49]	Long posterior fusion Bony PJK 9.7±4.3, non-bony PJK 11.9±4.2, no PJK 10.75±3.9	PJK	6 wk, 6 mo, 1 year; 65% and 87% occurred within 6 wk and 6 mo, respectively	36.5% (23/63)	– Mean UIV/UIV+1 HU: 141.7±32.4 (non-bony PJK group) vs. 168.7±66.8 (no PJK group), p=.622 – UIV/UIV+1 HU moderately correlated with PJA measurements (r = −0.475, p<.01)	– 2 required revision for progressive malalignment and intolerable pain
Ye et al. [62]	TLIF 1 (988), 2 (270)	Symptomatic adjacent level disease requiring revision	Mean time to presentation 68.3±25.1 mo (20–123)	6.5% (65/1258)	– Incidence of DXA-diagnosed osteoporosis: 30.7% (symptomatic degeneration group) vs. 17.5% (no degeneration group), p=.069	– All symptomatic requiring revision, 2 required a second revision.
Yuan et al. [51]	Long posterior fusion with T9-L2 UIV PJK 6.47±2.10, No PJK 5.87±1.27	PJK	Within 6 wk and at final follow-up	20.24% (17/84)	– Mean T-scores: −2.36±0.79 (PJK group) vs. −1.61±0.91 (no PJK group), p=.01 – Osteoporosis (T-score < -2.5) an independent risk factor for PJK (OR 14.12, p=.028)	– Worse VAS low back (p=.015) and SRS-22 (p=.008) scores in PJK at final follow-up
Zhang et al. [58]	Posterior thoracolumbar fusion PJK 4.3±1.7, No PJK 3.8 ± 1.3	PJK	1, 3, 6, 12, 24, and 36 mo	32.4% (108/333)	– Mean UIV HU: 107.07±30.62 (PJK group, propensity matched) vs. 123.28±35.59 (no PJK group), p=.002 – ROC analysis: optimal cutoff for predicting PJK of 120.87 (AUC 0.646)	N/A

Study acronyms are explained in the first footnote to Table 3. Abbreviations: OR, odds ratio; UOR, unit odds ratio; VAS, visual analog scale; ODI, Oswestry disability index; SRS-22 score, Scoliosis Research Society; JOA, Japanese Orthopaedic Association score; AUC, area under curve; AP, anterior-posterior combined approach.

^⁎

PJK defined as proximal junctional angle (PJA), measured as the sagittal Cobb between the inferior endplate of the UIV and superior endplate of UIV+2, that is both >10° and at least 10° greater than the preoperative measurement [50].