The Effect of Cigarette Smoking and Vitamin D Status on Fusion Rates after Posterior Lumbar Interbody Fusion Surgery: Does Sound Radiological Fusion Correlates with Better Clinical Outcomes?

Background Data: Lumbar interbody fusion surgeries are the gold standard surgical option in degenerative disc disorders and spondylolisthesis. Achieving solid arthrodesis at the operated segment remains the main goal of surgery. Yet, the correlation of sound radiological fusion with better clinical outcomes is not well established. In recent years, spinal surgeons had much greater awareness of the influence of vitamin D deficiency and cigarette smoking on spinal fusion. Purpose: To assess the influence of vitamin D deficiency and cigarette smoking on the rate of spinal arthrodesis after posterior lumbar interbody fusion (PLIF) in degenerative disc disorders and whether successful fusion correlates with clinical outcomes. Study Design: A prospective cohort study. Patients and Methods: The study was conducted on 67 patients (with a total of 92 levels) who underwent operation for PLIF with pedicle screw fixation. Twenty-six patients had degenerative spondylolisthesis, 20 had herniated discs, and 21 patients showed degenerative disc changes with segmental canal stenosis. In 29 patients, a local bone graft was used as soft PLIF. In the remaining 38 patients, a PEEK interbody cage was utilized. Patients were categorized according to both their serum vitamin D status (39 low; 28 normal) and cigarette smoking habits (18 smokers; 49 non-smokers). The final assessment was done 6 months postoperatively. Clinical outcomes were measured by the Oswestry Disability Index (ODI) and Visual Analogue Scale (VAS) for low back pain. Radiological fusion status was assessed on plain radiographs and graded according to Bridwell’s interbody fusion grading system. The relationship between vitamin D status and cigarette smoking and radiological outcomes was also evaluated. Results: Twenty-two males and 45 females were included in this study. The mean age was 55.8±±3.78 years. L4-L5 was the most commonly operated level, followed by L5-S1 and L3-4 (51.1%, 36.95%,


INTRODUCTION
It is estimated that about 77% of the population will present with low back pain at least once during their lifetime. 22 Intervertebral disc degeneration can be held accountable in most of these cases. A variety of conditions that arise secondary to the process of disc degeneration, collectively called degenerative disc disorders (DDD), can contribute to the patient's complaint. These include the development of lumbar canal stenosis, degenerative spondylolisthesis, facets arthropathies, and the so-called "black disc". 1,47 Although most patients can be initially managed conservatively, surgical intervention is still indicated in those in which conservative measures fail or those who develop severe or progressive symptoms. Interbody fusion surgeries remain the most popular surgical option for most of these cases, and their use has dramatically increased worldwide over the last two decades. The rationale behind fusion procedures is to achieve solid arthrodesis at the affected segment, thus eliminating any segmental instability. However, contradicting evidence exists regarding the relationship between sound radiological fusion and better clinical outcomes. 13 In addition, studies assessing the factors that may influence the success of arthrodesis are still lacking or inconclusive. Therefore, in our study, we aimed to assess the influence of some factors, theoretically thought to affect the normal bone metabolism, such as vitamin D status of the individual, and cigarette smoking on spinal fusion after PLIF surgery. In addition, we aimed to demonstrate whether there is a correlation between successful radiological fusion and better clinical outcomes.

MATERIALS & METHODS
We prospectively enrolled 67 patients with symptomatic degenerative disc disorders, admitted to our institute in the period between May 2017 and May 2018. Patients suffering from one or two levels of degenerative spondylolisthesis, degenerative disc disease, or degenerative stenosis and having progressive neurological symptoms, persistent incapacitating pain interfering with normal daily activities and not improving despite at least 3 months of nonoperative treatment were recruited in this study. Patients with high-grade spondylolisthesis, osteoporosis and those with other pathologies (e.g., infection, fractures) were excluded. Twenty-six patients (38.8%) had degenerative spondylolisthesis, 20 (29.9%) had degenerated herniated discs, and the remaining 21 patients (31.3%) showed degenerative segmental stenosis. All patients underwent operation for posterior lumbar interbody fusion (PLIF) with pedicle screw fixation and were scheduled for a follow-up visit, 6 months postoperatively for final assessment. 38 patients had a PEEK cage inserted between the vertebral bodies, with a total of 52 operated levels (25 with single level, 12 with double levels, and 1 with triple levels). The remaining 29 patients had only local bone graft inserted in the interbody space with a total of 40 operated levels (18 patients with single level and 11 with double level). The degrees of pain and disability were assessed by the Oswestry Disability Index (ODI: 0%-100%) 18,20 and a Visual Analogue Scale for low back pain (VAS: 0-10). Both measures were obtained preoperatively as a baseline and once at the final follow-up, 6 months postoperatively. Preoperative radiological evaluation included MRI of the lumbosacral spine, as well as anteroposterior, lateral neutral, and dynamic plain radiographs. The final postoperative radiological assessment of fusion was done on anteroposterior and lateral plain radiographs. Fusion status was categorized according to Bridwell's classification of interbody fusion. 7 Bridwell's grades I and II were considered as having sound fusion, while grades III and IV were considered incomplete or no fusion. Moreover, patients were categorized according to their vitamin D status (39 patients with low serum vitamin D versus 28 with normal levels), represented by the serum level of 25(OH)D obtained at final follow-up. The cut-off value for serum 25(OH)D as set by the American Endocrine Society was 30 ng/mL. 26 Levels below that figure were considered low (deficient or insufficient), while levels above that figure were considered to be normal. The demographic data of both groups was comparable (Table 6). The mean age in both groups was similar. 42.9% of patients with normal levels had degenerative spondylolisthesis, 25% had disc herniation, and 32.1% had segmental stenosis, compared to 35.9%, 33,3%, and 30.8%, respectively, in the low-level group. 53.6% of patients with normal vitamin D levels had a PEEK cage inserted versus 59% in those with low levels. Patients were also categorized according to their cigarette smoking habits into smokers, who smoked at least 10 cigarettes (1/2 a pack) per day, for at least 6 months, and non-smokers (18 smokers versus 49 non-smokers). Again, the patients' demographic data was comparable in both groups (Table 6). 44.4% of the smoker group had degenerative spondylolisthesis, 27.8% had disc herniation, and 27.8% had segmental canal stenosis, compared to 36.7%, 30.6%, and 32.7%, resp., in the non-smoker group. 50% of the smoker group and 59.2% of the non-smoker group had a PEEK cage inserted. The potential influences of both smoking and vitamin D status on the clinical and radiological outcomes were also assessed at final follow-up using the Chi-Square test of independence.

Surgical Technique
Patients underwent operation through the standard posterior approach with a midline posterior incision and subperiosteal dissection of the paraspinal muscles to expose the affected segment. This was followed by adequate decompression of the spinal canal with removal of the spinous process, interspinous ligament, lamina, hypertrophic ligamentum flavum, and medial parts of the facet joints. Foraminotomy and nerve root decompression were routinely done. Pedicle screws were inserted under fluoroscopy guidance. The intervertebral disc was then entered after securing the thecal sac and nerve root medially and inferiorly using nerve retractors. Radical discectomy with end-plate preparation was done using a combination of standard disc rongeurs, disc shavers, and curettes. In all patients, rods of desired length were contoured to the appropriate lordotic curve and were applied over the pedicle screws. An appropriate degree of distraction was applied across the disc space and a local bone graft obtained from the removed posterior arch elements (i.e., spinous process, lamina, and facets), and a PEEK cage, when used, was placed and impacted into the disc space for interbody fusion. After that, the previously applied distraction was released so that the graft is well impacted in the disc space. A drainage catheter was then inserted after proper hemostasis. Fascial, subcutaneous, and skin layers were then closed in a layer-by-layer fashion.
Our series included 45 females and 22 males, with a mean age of 55.81±3.78 years (range, 46-65). L4-L5 level was by far the most commonly operated level, followed by L5-S1 and L3-4 levels (51.1%, 36.95%, and 11.95%, resp.). Descriptive data of the studied patients is summarized in Table 1. There were significant improvements in the clinical outcome measures at final follow-up at 6 months. The mean ODI improved from 54.84±8.07 preoperatively to 22.93±3.99 at final follow-up (P<0.001), while mean VAS scores improved from 7.55±1.02 preoperatively to 2.27±0.89 at the final postoperative follow-up (P<0.001). 58 of our patients (86.6%) were considered to have satisfactory radiological fusion (Bridwell's grades 1 and 2), while only 9 patients (13.4%) had incomplete or no fusion (Bridwell's grades 3 and 4) at final postoperative assessment. In our study, we found that both smoking and vitamin D status significantly influenced the success of radiological fusion. Cigarette smoking was also found to have a significant negative effect on the fusion process, with 33.33% of the smokers (N=6) showing incomplete or no fusion, compared to a nonfusion rate of only 6.12% (N=3) among non-smokers (P=0.004). Similarly, 20.51% of the patients (N=8) with low serum vitamin D had incomplete radiological fusion on final followup, compared to only 3.57% (N=1) of those who had normal levels. The difference in radiological fusion rates between the 2 groups was found to be statistically significant (P=0.045). However, neither vitamin D status nor cigarette smoking habits had a statistically significant effect on clinical outcomes measures (i.e., ODI and VAS). Moreover, we could not find a significant correlation between sound radiological fusion and better clinical outcomes. Patients who achieved sound radiological fusion at final follow-up (N=58) had mean ODI and VAS of 25.59±3.89 and 2.18±0.87, respectively, while those with unsatisfactory radiological fusion (N=9) had mean ODI and VAS of 25.11±4.14 and 3.17±0.75, respectively. No statistically significant relationship could be found between radiological fusion and either measure at final follow-up (P=0.077 and 0.157 for ODI and VAS, resp.).

DISCUSSION
Lumbar degenerative disc disease is a common cause of low back pain and sciatica in adults. 2, 15,16,36 The process of degeneration starts with dehydration of the nucleus pulposus together with decreased tensile strength of the annulus fibrosus, leading eventually to decreased disc height, facets hypertrophy and subluxation, and segmental instability of the affected segment. 37 In our institute, posterior lumbar interbody fusion (PLIF) procedure remains the most commonly used surgical option for treatment of degenerative disc disorders and spondylolisthesis. They have the advantage of providing immediate restoration of disc height, decompression of the nerve roots, and anterior column reinforcement. 11, 34 We most commonly use the polyetheretherketone (PEEK) cage to support the bone graft. However, local bone grafts are sometimes used solely depending on the surgical preference of the operating surgeon, as well as the economic status of the patient. However, our study has some limitations: first, different surgeons operated on the patients and thus some variations in surgical skills and techniques used could exist. Second, the postoperative followup period was 6 months, which is a relatively a short follow-up period compared to similar previous studies. Third, the assessment of fusion was not a straightforward task. Blumenthal and Gill 6 stated that surgical re-exploration is the only reliable method to assess fusion but this was obviously impractical to be used in our study. Although computed tomography (CT) appears to be more accurate to assess fusion status than plain X-rays, 9 we could not also include it in our study due to the high costs involved. Consequently, we used plain lateral radiographs for assessment of radiological outcomes. In our study, patients seemed to have an improvement in their back pain despite using an artificial cage or only local bone graft. However, the cage group showed better functional outcomes as reflected by ODI scores. Various studies assessing the outcomes of interbody fusion procedures reported significant improvement in clinical outcomes measures at final follow-up. Hashimoto et al. 23 assessed the outcomes after single-level PLIF surgery using carbon cages and reported on average 83% improvement rate after a mean follow-up period of 31 months. Hioki et al. 25 assessed the outcome of double-level PLIF in patients with degenerative disc disorders and similarly reported significant improvement in functional status at final follow-up. In Sears et al. 41 study assessing outcomes of PLIF surgery in degenerative spondylolisthesis, mean VAS improved from 5.3±2.2 to 2.2±2.1 after a mean follow-up period of 21.4 months, and 91% of his patients reported good or excellent results. Kim et al. 29 assessed the outcomes of PLIF surgery using a single interbody cage in degenerative lumbar disorders. The patients reported an improvement in mean ODI score from 68.7 preoperatively to 37.9 and in VAS from 6.5 to 1.8 at final postoperative follow-up (average 31.1 months). In addition, Yu et al. 51 reported that using an artificial cage resulted in better postoperative ODI scores compared to a bone graft alone. Meanwhile, various other studies failed to show a significant advantage in using an artificial cage on clinical outcomes, compared to bone graft alone. 5,31,32

Assessment of Fusion
For a solid spinal arthrodesis to be achieved, the graft material used should provide good osteoinductive, osteoblastic, and osteoconductive properties. 35 Using iliac bone graft provided excellent fusion rates, but was associated with significant donor site morbidity. 43 Posterior elements local bone graft is an alternative source of graft material to be used for interbody fusion. Nevertheless, when used alone, bone grafts may result in collapsed union and failure to maintain the corrected disc height. 5,51 Therefore, the use of an artificial cage has the theoretical advantage of maintaining corrected disc height and lumbar lordosis and providing immediate anterior column load sharing, while also providing satisfactory radiological fusion rates. 51 On the other hand, artificial cages have the disadvantage of reducing the available contact area for bony fusion 3 . Thus, we routinely add local bone graft anteriorly prior to cage insertion. In our study, PLIF surgery resulted in excellent overall sound fusion rates (86.6%) at 6-month postoperative follow-up visit, regardless of whether an artificial cage was used or not. Several previous studies reported similarly excellent fusion rates. Csecsei et al. 12 reported fusion rates of 95.7% after a mean follow-up period of 27.3 months. Kai et al. 27 reported 92.9% fusion rates using local facet joint bones after a long-term follow-up of 8.5 years. Kim et al. 29 reported a 94.4% solid fusion rates at 6 months postoperatively using a single artificial cage in degenerative lumbar conditions. Zhao et al. 53 reported 92.6 % fusion rate using a single diagonal artificial cage at 1-year follow-up, The EGYPTIAN SPINE Journal which increased to 100% 2 years postoperatively. In addition, some authors reported comparable fusion rates using artificial cage or bone graft alone after variable follow-up periods. 31,32,51

Correlation between Radiological Fusion and Clinical Outcome
One of the integral goals of spinal fusion surgeries is to achieve solid arthrodesis across the operated segment. It is thus expected that patients with solid fusion would report better functional outcome than those with incomplete fusion or pseudoarthrosis. However, data from previous studies shows contradicting results on whether there is actual correlation between radiological fusion and functional outcome. In our series, there was no significant correlation between sound radiological fusion and better clinical outcomes. Similar results were obtained by Thalgott et al. 45 in his series of patients undergoing anterior lumbar interbody fusion (ALIF). Penta and Fraser 38 also failed to find such correlation between successful fusion and superior functional outcome after a long-term follow-up (10 years). Epstein 17 found no correlation between fusion and clinical outcome in her series of patients who underwent noninstrumented fusion surgeries. On the other hand, other authors reported better clinical outcomes in patients with sound radiological arthrodesis. Christensen et al. 10 published a prospective study on 148 patients who underwent posterolateral fusion (PLF) either with pedicle screws fixation alone or with anterior lumbar interbody fusion (ALIF) in addition. They found a significant positive relationship between radiological solid fusion and better functional outcome and quality of life. Kim et al. 29 randomized his patients to have either PLF alone or PLIF alone or both of them simultaneously. They reported 91% of patients with fusion had superior functional outcome compared with only 41% of those with nonunion. In another study by Wetzel and colleagues, 49 patients were evaluated for 2 years, and a positive correlation could be found between solid radiological fusion and better clinical outcomes. Zdeblick et al. 52 similarly found a positive relationship between fusion status and better clinical outcomes in a prospective study with a follow-up period of 1 year. Djurasovic et al. 14 reported a 65% improvement in ODI in patients with solid fusion compared with 32% in non-fusion patients, and this difference was found to be significant.

Vitamin D Levels
Vitamin D plays an important role in maintaining adequate serum calcium and phosphorus levels by promoting intestinal absorption. It is also important for proper bone mineralization by promoting the process of bone remodeling via action on osteoblasts and osteoclasts. 44 Recent evidence suggests that vitamin D deficiency may be associated with less favorable surgical outcomes after spinal fusion procedures and increased risk of pseudoarthrosis. 39 Epidemiological studies showed a wider prevalence of vitamin D deficiency than initially thought. One study reported about 41.6% of the general American population was found to be vitamin D deficient. 19 Another retrospective study of patients undergoing elective spinal surgeries reported that 27% of them had vitamin D deficiency and about 57% had insufficiency. 42 These data suggest that vitamin D disorders may represent a modifiable risk factor that may affect the outcomes after spinal fusion surgeries. As one would expect, vitamin D deficiency would thus affect the physiology of bone fusion after spinal surgeries. This was demonstrated in our series, where data analysis showed significantly better fusion rates in those with normal serum vitamin D levels compared to those with low serum levels. Yet again, variations in vitamin D status among patients were not reflected on their final clinical outcome measures. Such a relationship between vitamin D status and fusion rates was also demonstrated by some previous studies. 39,50 Other studies also reported an influence of vitamin D status on clinical outcome measures. Ravindra et al. 39 found that patients with vitamin D deficiency showed significantly higher rates of nonunion and longer mean time to fusion when compared with patients with normal levels. Xu et al. 50 showed that vitamin D supplementation in deficient patients resulted in better fusion rates and functional outcomes compared to the control group. Studies by Kim et al. 30 and Waikakul et al. 48 also reported worse clinical outcomes in patients with vitamin D deficiency after lumbar surgeries. On the other hand, Schofferman et al. 40 found no significant association between serum vitamin D level and rate of non-fusion after spinal fusion procedures.

Smoking
As for cigarette smoking, it was found to affect the local vasculature and thus the adequacy of blood supply to the fusion bed, as well as expression of genes coding for different cytokines that are important for proper vasculature and healing process in the fusion bed. 28,47 Based on these findings, we included smoking as a variable that may affect the success of fusion after spinal surgeries. Data analysis showed that the rate of nonfusion among smokers was significantly higher than in non-smokers. Similar findings were reported by Glassman et al. 21 in patients who underwent instrumented lumbar fusion surgery. In addition, patients who stopped smoking for more than 6 months showed an improvement in their fusion rates. Andersen et al. 4 found that smoking more than 10 cigarettes per day affected spinal fusion rates negatively. Hermann et al. 24 showed a higher rate of reoperations for pseudoarthrosis among smokers as compared to non-smokers. On the other hand, a study by Luszczyk et al. 33 compared fusion rates after anterior cervical discectomy and fusion surgery among smokers and non-smokers and found no significant difference between the 2 groups regarding radiological fusion. Bydon et al. 8 could only demonstrate a negative effect of smoking on fusion rates among patients undergoing double-level lumbar fusion but failed to show a similar effect in those undergoing singlelevel fusion surgeries.

CONCLUSION
Smoking habits and low serum vitamin D seem to have a significant negative effect on the success of radiological spinal fusion. There was no significant relationship between radiological fusion and clinical outcomes.