Indications for pars repair are: (a) patient age below 20 years; (b) isthmic defect less than 3 mm, (c) absence of significant disc degeneration, as evident on preoperative MRI scan; (d) no instability in lateral spinal flexion-extension view [19–26]. Presence of spina bifida occulta and a thin lamina are contra-indications for pars repair [19, 27]. The lamina should be at least 3 mm thick to provide adequate instrumentation purchase [20].

The theoretical advantages of pars repair are: (a) preservation of a motion segment (especially if the isthmic lesion affects levels proximal to L5), therefore minimizing the risk of adjacent segment disc degeneration; (b) no effect on spinal growth [20, 28–31]. Mild disc degeneration is not necessarily a contra-indication for pars repair as long as it is preoperatively confirmed by a local anaesthetic injection that the painful symptomatic area is the spondylolysis and not the adjacent disc [23, 30, 31]. Technically demanding positioning of the implants and need for instrumentation removal at a later stage according to the patient’s symptoms or needs (for example, if recurrent back pain develops often in athletes who have high physical demands) are the potential disadvantages of pars repair [1].

The isthmic repair represents a true reconstructive procedure of the lumbar spine. This is accomplished by debridement of the non-union, refreshing of the pars defect bony edges and autologous bone grafting. Local stabilization is achieved either though application of a brace [32, 33] or most commonly by internal fixation of the free floating posterior fragment using various techniques. These may include: (a) trans-defect screws [20, 23, 24]; (b) a figure of 8 wire fixation [34]; (c) a hook-screw construct [27, 35]; (d) a pedicle screw U-shaped rod construct [28]. The repair techniques are not recommended for gaps greater than 3 mm and the exposure of the transverse processes may lead to bleeding and a higher risk of nerve root damage. The hook-screw technique overcomes the difficulty of fixation in the presence of dysplasia of the posterior arch and avoids crossing the isthmic defect with a screw. It also provides a larger surface area for bone healing and produces a more stable construct. Various modifications of the technique have been developed since the initial report [19, 20, 30, 35, 36]. More recently, the U-shaped pedicle screw/rod technique has promoted union of the defect by applying compression forces through the spinous process of the affected segment. Advantages of this technique include stable fixation and wide surface for bone grafting which allows early patient mobilization and no need for instrumentation removal as this is not interfering with the integrity of the facets.

This is the mainstay of surgical treatment in bilateral L5 spondylolysis and associated low grade lumbosacral spondylolisthesis [1, 8, 37]. It consists of either a posterior or posterolateral fusion with the use of autologous bone graft commonly harvested from the iliac crest. A postoperative lumbar support is recommended for a period of 4–6 months.

The advantages of the in situ fusion are: (a) there is a large area for bone grafting; (b) it can address a dysplastic spondylolisthesis; (c) it can be used to treat more severe spondylolisthesis; (d) the complete fusion addresses other potentially symptomatic areas of the affected level (L5/S1 disc and lumbosacral facets); (e) it can be performed with the use of autologous bone graft alone with no need for placement or removal of spinal instrumentation; (f) it involves a lesser risk of neurological injury compared to an instrumented fusion. The restriction of movement in the fused segment is the only disadvantage of this procedure.

The most commonly used technique to achieve an in situ lumbosacral fusion is that described by Wiltse and Spencer [38]. The fusion is performed using a bilateral paraspinal sacrospinalis muscle splitting approach. The sacral ala, posterior surface of the L5 transverse processes and L5/S1 facet joints are decorticated and the area is packed with iliac crest bone autograft (ICBG). The patients are mobilized on the 2nd or 3rd postoperative day with a soft lumbar brace applied for approximately 4 months. Previous reports have combined the technique of direct pars repair with interposition and in situ fusion of the L5/S1 facet joints using autologous bone in patients who already have degenerative changes affecting the L5/S1 disc [32, 33].

There is no evidence on superiority of conservative versus operative treatment for spondylolysis and low grade spondylolisthesis in regards to clinical outcomes (Table 26.2). The long term retrospective level IV studies of the Finnish group [2, 3, 5] regarding conservative versus in-situ fusion suggest: (a) that none of the patients for whom conservative treatment was decided required operative treatment at a later stage; (b) that there is significantly higher progression of slip in the first review but not in the final follow-up in the operated group; (c) the operated group had better clinical results and less pain at final follow up; (d) that fusion operations do not significantly increase the rate of disc degeneration in the adjacent disc level above after a mean postoperative follow-up of 13.8 years.

There is only level IV evidence assessing the clinical and radiological outcomes of the various pars repair techniques (Table 26.3). It seems that the rod-pedicle screw construct achieves the highest mean fusion rate (93.08 %). According to the same analysis the best clinical outcome has been achieved by the hook-screw Morsher-type repair (90.87 %); however this is a subjective result and the various authors used different criteria for evaluating their outcomes. The rod-pedicle screw technique produces 85 % good to excellent clinical results. No clear recommendation can be suggested on the basis of this level of evidence. These are complex and technically challenging procedures which require surgical expertise. It seems that every group of surgeons evolve their preferred method of pars repair with eventually improved clinical results [46] (Table 26.4).

Lumbosacral fusion is considered the surgical treatment of choice and is applied for the entire spectrum of grades of symptomatic spondylolisthesis. The studies which are mostly relevant to spondylolysis and mild spondylolisthesis are by Lenke et al. [48] and Helenius et al. [37] (Table 26.5). The fusion rate achieved after surgery varies from 50 % to 81.48 %; despite the relatively low fusion rates both studies report a high percent of good to excellent clinical outcomes.

Two author groups [32, 49, 50] compared the effectiveness of isthmic repair versus in situ fusion in the treatment of spondylolysis and low grade spondylolisthesis (Table 26.6). The conclusions of these studies regarding spondylolisthesis affecting L5/S1 are that: (a) both pars repair and fusion showed no significantly different radiological or clinical outcomes in the short-term follow-up; (b) the in situ fusion achieved superior results in the long term; (c) the theoretical advantage of motion-segment preservation in pars repair cannot be proven in the clinical setting; (d) the isthmic repair does not prevent disc degeneration; (e) fusion should be performed in patients with preoperative MRI proven disc degeneration; (f) the repair has better results in younger patients; (g) the clinical and functional scores (VAS: Visual Analogue Scale, ODI: Oswestry Disability Index, SRS: Scoliosis Research Society) are significantly better in the fusion when compared to the direct repair group.

Many authors, therefore, suggest that the repair of spondylolisthesis at L5 produces less favourable results than lumbosacral in situ fusion [1]. In contrast, repair of the pars defect is indicated for L1 through to L4 spondylolytic defects, spondylolytic defects of multiple vertebral levels, and low-grade but reducible spondylolisthesis at levels cephalad to L5 with an intact intervertebral disc at the level of the displacement [8]. This is possibly due to: (a) anatomical reasons; (b) the fact that many lytic defects at L5 are the end stage of a developmentally weakened and elongated pars; (c) fusion for an isthmic defect cephalad to L5 would create a much stiffer lumbar spine.

In the current literature, there is no solid evidence to lead to strong recommendation for the treatment of spondylolysis and/or mild spondylolisthesis (<15 % vertebral displacement). The existing studies are an amalgamation of different concepts, treatment of different levels of slips, inclusion of different age groups and application of same techniques regardless of the affected level.

The mainstay of surgical treatment in spondylolysis/mild spondylolisthesis affecting L5 is in situ lumbosacral fusion. However, adolescent patients with increased healing potential (as evident on SPECT), symptomatic (injection proven) spondylolysis (<3 mm) or early-reducible spondylolisthesis without degenerative changes affecting the discs/facets, dysplastic bony changes or spina bifida may be considered for direct pars repair. The threshold towards isthmic repair if the pathology involves the L1-L4 levels is lower.

Surgical treatment should be offered after at least 6 months of unsuccessful conservative management and should be accompanied: a) by brace immobilization of the lumbosacral junction for 4–6 months based on the surgical method and postoperative progress and b) by activity modification and limitations.

Among the available level IV studies the mean fusion rate achieved when a posterolateral fusion is performed for spondylolisthesis is 86.93 % [37, 49, 58]. The fusion rate increases above 92 % in long term follow-up series with large experience on this technique [49]. A posterior fusion is less effective reaching mean fusion rates of 82.67 % (Table 26.7). In the same line, the studies of the Finnish group (level III & IV) show that the SRS-24, ODI and VAS scores were better when the fusion was achieved through the posterolateral compared to the posterior approach [37, 49, 59]. However, the functional/clinical results are difficult to interpret, as different methods of assessment have been used among the available studies. There is no correlation between patient outcome (ODI) and abnormal lumbar MRI findings [52]. Also the development of a non-union following index surgery does not affect the final clinical outcome [59]. The presence of abnormal preoperative neurology may justify additional root decompression. Even though the difference has not reached significance, there is level III evidence that the pain in exertion and at rest is lesser with fusion alone than with fusion and decompression [58].

Posterolateral fusion in situ remains the mainstay of surgical treatment for children and adolescents with a grade I–II L5/S1 spondylolisthesis and produces satisfactory, long-lasting results [1, 4, 48, 51, 60]. The recommended technique involves a bilateral muscle-splitting approach through a midline skin incision, enabling fusion with the use of autologous bone harvested from the iliac crest across the transverse processes of L5 and the sacral ala [60]. The presence of neurological deficits associated with radiological evidence of neural compromise requires nerve root decompression which can be performed either through a posterolateral or a posterior approach to the spine.