Proximal junctional kyphosis in pediatric spinal deformity surgery: a systematic review and critical analysis

Abstract

Purpose

Proximal junctional kyphosis (PJK) is a commonly encountered clinical and radiographic phenomenon after pediatric and adolescent spinal deformity surgery that may lead to post-operative deformity, pain, and dissatisfaction. Understanding the risk factors of PJK can be useful for pre-operative informed consent as well as to identify any potential preventative strategies.

Methods

We performed a systematic review and critical analysis following the PRISMA statement in July 2019 by searching the PubMed, Scopus, and Embase databases, including all prior published studies. We included articles with data on PJK in patients with operative pediatric and adolescent scoliosis and those that detailed risk factors and/or preventative strategies for PJK. Levels of evidence were determined based on consensus. Findings were summarized and grades of recommendation were assigned by consensus. This study was registered in the PROSPERO database; 202,457.

Results

Six hundred and thirty five studies were identified. Thirty-seven studies met criteria for inclusion into the analysis. No studies including neuromuscular scoliosis met inclusion criteria. No findings had Grade A evidence. There were 4 findings found to contribute to PJK with Grade B evidence in EOS: higher number of distractions, disruption of posterior elements, greater sagittal plane correction. There was no difference in incidence noted between etiology of the curvature. Five findings with Grade B evidence were found to contribute to PJK in AIS populations: higher pre-operative thoracic kyphosis, higher pre-operative lumbar lordosis, longer fusion constructs, greater sagittal plane correction, and posterior versus anterior fusion constructs.

Conclusion

Greater sagittal plane correction has Grade B evidence as a risk factor for PJK in both EOS and AIS populations. In EOS patients, an increased number of distractions and posterior element disruption are Grade B risk factors. In AIS patients, longer fusion constructs, higher pre-operative thoracic kyphosis and lumbar lordosis, and posterior (as opposed to anterior) constructs also contributed to PJK with Grade B evidence. These findings can guide informed consent and surgical management, and provide the foundation for future studies.

Appendices

Appendix 1

EOS findings with inconclusive evidence

Coronal plane curve magnitude

Rationale: two Level 3 manuscript reported > 40° of proximal thoracic scoliosis as a risk factor [14, 19]. One Level 3 paper found coronal curve magnitude did not affect rates of PJK [5].

Pre-operative cervical lordosis

Rationale: one Level 3 study found that patients who developed PJK had higher pre-operative cervical lordosis [13].

Pelvic incidence

Rationale: one Level 4 study reported a lower PI associated with increased PJK [12] while two Level 3 studies reported increased risk ratio of developing PJK with a higher PI [13, 19].

Malpositioned screw at UIV

Rationale: one Level 3 study reported that 6/7 patients who developed PJK had ≥ 1 malpositioned screw [17].

Longer fusion constructs

Rationale: two Level 3 studies reported a greater number of fused levels in the PJK group when compared to controls [6, 9].

Selection of UIV influences the rate of PJK

Rationale: one level 3 study reported that UIV distal to T2 (hazard ratio = 5.474; P = 0.044) were found to be independent risk factors of PJK in the multivariate analyses [19].

Selection of LIV influences the rate of PJK

Rationale: one Level 3 study described LIV at or cranial to L3 as a risk factor for PJK [14].

Gender

Rationale: one Level 3 study reported higher incidence of PJK in males [11] while a Level 4 study reported a higher incidence in females [12]. One Level 3 study showed no difference [19].

AIS findings with inconclusive evidence

High lordosis/pelvic incidence ratio

Rationale: one Level 2 study reported higher pre-op and post-op lumbar lordosis/PT ratios in patients who developed PJK [26].

Disruption of posterior elements

Rationale: one Level three study reported a greater frequency of disrupted junctional ligaments in the PJK group [24], and another study reported a single case of posterior ligamentous failure in a patient with PJK [27].

Rod material

Rationale: a Level 3 study found no difference in PJK risk between titanium and cobalt chrome rods [30]. Another Level 3 study reported that significantly higher risk with cobalt chromium (28%), compared to titanium and stainless steel (12%) (p < 0.0001) [38].

Bone graft material

Rationale: a single Level 3 study described autologous bone graft as a risk factor for PJK [31].

Use of distraction

Rationale: a single Level 3 study described distraction during deformity correction as a risk factor [31].

Body mass index

Rationale: one Level 2, one Level 3 and one Level 4 study found no significant difference in BMI between PJK and non-PJK cohorts [26, 33, 38]. A level 3 study reported a significantly increased BMI in the PJK group [4].