Background

Surgical treatment of adult spinal deformity has progressed substantially over the past ten years. There have been significant advances in decision-making, medical management and surgical technique [1, 2]. These improvements in evaluation and treatment have broadened the applicability of adult deformity surgery and lead to more reproducible clinical benefit based upon health related quality of life (HRQOL) scores [3, 4].

Despite these positive developments, the durability of adult spinal deformity surgery remains problematic. Revision rates above 20 % have been reported, with a range of causes including wound infection, nonunion and adjacent level pathology [5–7]. While some of these complications have been amenable to changes in patient selection or surgical technique, Proximal Junctional Kyphosis (PJK) remains an unresolved challenge.

The initial description of PJK in the pediatric literature was an increased sagittal angulation, without structural failure, at the upper aspect of a fusion construct [8]. At present, the term is applied much more widely to describe any failure or loss of alignment above an instrumented segment [9, 10]. This may result from adjacent level compression fracture, spondylolisthesis or fixation failure [6–10]. In general, this has been viewed as a consequence of poor bone quality, over-aggressive deformity correction or inadequate fixation.

PJK has been the focus of intense scrutiny, with multiple studies proposing mechanical solutions including adaptations for osteoporotic bone and in particular specific sagittal alignment targets [11, 12]. Unfortunately, none of these mechanical solutions have effectively decreased the rate of PJK. The role of this study is to examine the contributions of non-mechanical factors to the incidence of postoperative sagittal imbalance and PJK after adult deformity surgery.

Methods

After receiving Institutional Review Board Approval, we reviewed a consecutive series of adult spinal deformity patients who required revision for PJK from 2013 to 2015 and examined in their medical records in detail. Standard demographic data including age, gender, smoking status, height and weight were collected. Indications for the index surgery, specifics of the index surgery including upper instrumented vertebra fixation, time to PJK diagnosis, time to PJK surgery, mode of failure. Medical records were extensively evaluated for preoperative comorbidities; specifically for preoperative neurologic disorders and other pathologies that may affect standing balance.

Results

From 2012 to 2014, 245 patients underwent surgical correction of their adult spinal deformity at our institution. A true incidence of PJK will be difficult to determine as (1) some patients presenting at our institution with PJK had their index surgery performed elsewhere and (2) some of the patients who had their index surgery at our institution could have developed PJK and had surgery elsewhere.

Twenty-nine cases of PJK requiring revision were identified (Table 1). Of these 9 (31 %) were males and 10 (34 %) were smokers. Mean age was 64.4 years. Mean BMI was 29. kg/m2. Neurologic disorders were identified in 22 (76 %) of the PJK cases reviewed in this series. Neurologic disorders included Parkinson’s disease (1), prior stroke (5), metabolic encephalopathy (2), seizure disorder (1), cervical myelopathy (7), thoracic myelopathy (1), diabetic neuropathy (5) and other neuropathy (4). Other potential comorbidities affecting standing balance included untreated cataracts (9), glaucoma (1) and polymyositis (1) (Table 2). Eight patients were documented to have frequent falls, with twelve cases having a fall right before symptoms related to the PJK were noted. Seventeen cases used an assistive device such as a cane, crutches or a walker and required a wheelchair. One patient had 5 co-morbid conditions affecting standing balance, two had 4 co-morbid conditions, four had 3 co-morbid conditions, nine had 2 co-morbid conditions and ten had only one co-morbid condition (Table 3).

Table 1 Summary of cases
Full size table
Table 2 Frequency of co-morbid conditions that can affect balance
Full size table
Table 3 Number of co-morbid conditions that can affect balance
Full size table

Discussion

Proximal Junctional Kyphosis was first identified in 1999 [8], and was initially described as a radiographic finding with limited clinical relevance [13, 14]. This sanguine assessment was short lived, as subsequent reports have documented the frequent need for revision surgery [5, 6] as well as the occurrence of catastrophic failures, termed Proximal Junctional Failure (PJF) [9, 10, 15, 16]. The reported increase in PJK was coincident with several major changes in treatment paradigm. Adult deformity surgery became more common in older patients, and more aggressive correction was undertaken using osteotomies and rigid instrumentation. Studies have highlighted these factors and examined their etiologic role in PJK and PJF [10, 17, 18].

Deformity surgeons clearly recognize PJK and PJF as important challenges, but often regard these complications as mechanical problems for which there should be a straight forward mechanical solutions. As osteoporosis is commonly identified as an etiology of PJK, surgeons have pursued options to offset poor bone quality. Strategies have included prophylactic medical treatment of low bone density, strengthening proximal instrumented and adjacent vertebral levels with cement injection. Other strategies have included decreasing rod rigidity, and softening the transition to unfused levels using hooks rather than screws [11, 19, 20]. Another major focus has been on selection of fusion levels and restoration of sagittal alignment [12, 18, 21, 22]. Studies have advocated both more aggressive and less aggressive deformity correction. Maruo et al. report that restoration of normal sagittal alignment protected against PJK, and that greater than 30-degree increase in lumbar lordosis was a significant risk factor for PJK. [18] As increase in lumbar lordosis is generally the mechanism by which normal sagittal alignment is restored, these observations appear contradictory.

The findings of the present study suggest that our failure to control the rate of PJK may be related in part to the narrow focus on mechanical factors. This study demonstrates that 76 % of patients with PJK after spinal deformity correction have co-morbidities that adversely affect standing balance, regardless of alignment. These include neuromuscular disease, history of cerebral vascular accident, cervical myelopathy and neuropathy. All of these conditions may contribute to an inability to rebalance through unfused segments after deformity correction. This phenomenon is clearly recognized with substantial neurologic impairment such as patients with Parkinson’s disease [23], but has not been clearly defined in those patients with less severe neurologic impairment.

Beyond potential neurogenic causes of standing imbalance, other factors such as visual impairment, vestibular dysfunction and severe muscular deconditioning also impact balance and gait [24, 25]. Visual impairment was noted in 40 % of PJK cases and more than a single potentially relevant co-morbidity was noted in more than 66 % of cases. While these findings do not implicate neuromuscular disease as the direct cause of PJK, they certainly suggest a multi-factorial etiology.

The mechanisms by which these non-mechanical risk factors contribute to PJK are not well defined, and probably do not represent a unique common pathway. In some instances, such as patients with neuropathy or central neurologic deterioration, an impaired feedback loop may limit the ability to compensate appropriately after mechanical realignment. In essence, the patient’s brain does not properly register the “improved alignment” as determined by radiographic assessment. In other cases, lack of appropriate sensory feedback may result in accelerated proximal segment degeneration, akin to the appearance of a Charcot joint. In patients with severe deconditioning, muscular support may be inadequate regardless of mechanical alignment.

It is not completely clear how best to apply these observations in clinical practice. Our case series methodology cannot provide a relative risk assessment for any of the individual co-morbid conditions, and to-date no diagnostic test has been developed to quantify a global risk for post-operative standing imbalance or PJK. It is also unknown as to whether these risks can be modified by pre-operative interventions such as balance training, in the same way that treatment of osteoporosis is thought to reduce the risk of post-operative vertebral fracture or screw pull-out.

Weaknesses of this study include firstly the case series methodology. As some of the patients had their index procedure elsewhere, we do not have an accurate denominator to assess the incidence of PJK in the primary cohort. This series is also relatively small, so that the relative risk of the various co-morbidities cannot be effectively compared. Despite these weaknesses, this study clearly supports the role of concomitant neuromuscular disease in the development of post-op standing imbalance and PJK. The data does not provide a specific threshold at which surgery should be withheld, but certainly emphasizes the importance of including an assessment of associated neuromuscular disease in pre-operative planning and shared decision-making.

Spine surgeons have devoted a great deal of time and effort to defining optimal sagittal alignment, but sagittal balance is more than just alignment. Dubousset outlined the many interactive systems that contribute to ambulation and stated, “good alignment is preferable in order to obtain a good balance, but it is not sufficient” [26]. Understanding and avoiding PJK requires that we move beyond the one-dimensional view that finding an ideal sagittal alignment, softening the transition at the proximal aspect of the instrumented segment, or improving the adjacent bone strength will solve the problem of PJK. Thinking about PJK more broadly is a step in the right direction.

Conclusions

PJK is an important contributing factor to the substantial and unsustainable rate of revision surgery following adult deformity correction. Multiple efforts to avoid PJK via alterations in surgical technique have been largely unsuccessful. This study suggests that non-mechanical neuromuscular co-morbidities play an important role in post-operative sagittal imbalance and PJK. Recognizing the multi-factorial etiology of PJK may lead to more successful strategies to avoid PJK and improve surgical outcomes.