Introduction
Spondylolisthesis is defined as forward displacement of a vertebra over another vertebra (1). Nowadays, spondylolisthesis is the indication of lumbar fusion surgeries in 30% of cases. It can be categorized into 5 types. The degenerative type mostly affects L4-L5 level and is a more common problem in women, whereas the isthmic type more frequently involves the level of L5-S1 and is more prevalent in men. In the majority of symptomatic patients with spondylolisthesis, a mechanical low back pain that increases with activity is the primary complaint (2). 25% of patients with low grade spondylolisthesis do not respond to any type of non-surgical treatment (3). Patients in whom symptoms become unbearable and disrupt their daily function, patients undergoing a progressive course and patients with a neurological deficit are candidates of surgery (4,5). The commonly used accepted technique to reconstruct the affected segment is pedicle screw instrumentation but there is a considerable controversy about the surgical procedure of choice (4-11). PLF and PLIF are widely accepted fusion techniques (12). PLIF was firstly introduced by Cloward in 1940 (13-15). Some of the studies have represented the PLIF as the superior technique but comparable results of both techniques have been shown by other trials (12, 16-20). The aim of this study was to compare the two common techniques of fusion in terms of pain curb in patients with low grade degenerative and isthmic spondylolisthesis.

Methods and Materials/Patients 
This prospective observational study involved 102 patients with isthmic and degenerative spondylolisthesis of low grade who were admitted to Neurosurgical Department of Poursina Hospital, Guilan, Iran between 2012 and 2014. Patients were studied in two groups. Group A included 51 patients in whom the used fusion technique was PLF and group B composed of 51 patients who were operated on with PLIF. Pedicle screw fixation was the applied technique for reconstruction of the affected segment in both groups. Surgeries were carried out by a single team consisted of an associate professor of neurosurgery, an assistant professor of neurosurgery and a neurosurgical resident affiliated with Guilan University of Medical Science (GUMS). Inclusion criteria were as follows: patients with degenerative and isthmic type spondylolisthesis of grade 1 and 2 who failed to respond to conservative therapy and aged between 18 and 75. Patients who reported any prior spinal surgery for spondylolisthesis or had a history of alcohol abuse and patients with an inadequate disk space for performing PLIF were excluded. This study was approved by Ethical Committee of GUMS and all patients signed a written consent form. A questionnaire containing VAS scores was used to measure the pain in patients before exposing the patients to surgical treatment (PLF and PLIF). Moreover, thorough physical examination was done and all of the related signs and symptoms were recorded under supervision of a resident of neurosurgery. In the operation room, all patients were positioned prone. After a midline incision and complete bony exposure, subperiosteal dissection continued till transverse processes were exposed. Decompressive procedure was done thorough laminectomy, medial facetectomy and extensive foraminotomy. Then after three level pedicle screw fixation (one level above and one level below), in PLF group (A), posterolateral  fusion was done by autografting with bone chips and in PLIF group (B), after a complete discectomy, lumbar interbody fusion was done by PEEK cages. A brace was prescribed for 3 months and then was tapered off if fusion was achieved. All the patients were informed of a scheduled follow up program explained by residents and they were asked to complete the same questionnaire using VAS scores in the day after surgery, after 6 months and after 1 year. Statistical analysis was done using Repeated Measure ANOVA with post hoc tests (Bonferoni method) of SPSS (Version 21). All of the tests were two tailed and a p-value<0.05 was considered to be statistically significant.   
 

 

Results
At beginning of study, a total of 102 patients with low grade spondylolisthesis of isthmic and degenerative type were entered into our study including 51 patients in group A (PLF) and 51 patients in group B (PLIF). Of 102 patients, 15 patients dropped out due to incomplete follow-up and one patient died due to cardiovascular disease. Finally, 86 patients, 45 patients in group A (PLF) and 41 patients in group B (PLIF), were studied. The mean age in patients of group A and B was 55.1±9.0 and 52.1±8.0, respectively. This observed difference between the two groups was not statistically significant (p>0.05). Based on distribution of types of spondylolistesis, the proportion of degenerative I and II and Isthmic I in patients group A (PLF) was more than patients group B (PLIF). These observed differences in proportion between two groups were statistically significant (Table 1). The most common clinical presentation was a combination of low back pain, radiculopathy, paresthesia and positive Straight leg raising test (SLR) with a prevalence of 22.2% (10 patients) in group A and 19.5% (8 patients) in groups B. The mean values of VAS scores before surgery, the day after surgery and one year later showed no significant difference between the groups but the difference in mean values after 6 months of surgery turned to be statistically significant (Table 2). Analyzing the course of VAS scores over the study period using repeated measure ANOVA, both groups  followed a descending pattern that was statistically significant for both of them (p<0.0001) (Table 2) (Figure 1). Comparison of the overall mean of VAS scores in two groups over the whole period of study showed that the mean of VAS score in group A (48.49±4.39) was higher than group B (41.02±3.41). This observed mean difference was statistically significant (p<0.05). For each group, mean values of VAS scores before surgery represented a significant difference compared to the value of the day after surgery, 6 months later and one year after surgery. Comparisons of mean values of VAS scores at different points in times have been illustrated in Table 3. The interaction of sex and age with the courses of mean values of VAS scores during the study period has been displayed in Figures 2 to 6. Analyzing the interactions using two ways ANOVA revealed a significant effect of sex and insignificant effect of age on the course of VAS scores related to two methods of surgery (p<0.05  and p>0.05, respectively).  

Discussion
In our study, both methods of fusion resulted in a remarkable decrease of VAS scores after one year of follow-up. Compared with PLF group, PLIF group reported a less degree of pain which was of statistical significance. Measuring and comparing the pain between two groups at four points in time showed a confusing result. After six months of surgery, PLIF turned to be the superior method in curbing the pain. Although the mean value of VAS related to PLIF group was lower than the PLF group during the whole period of study, this difference was not significant early after surgery and one year later. The very similar mean scores of VAS between groups before the surgery could serve as a rationalization to justify the non-significant difference of VAS values between groups the day after surgery. The effect of fusion techniques on VAS scores in age groups was similar but considering the gender, the effect of surgical methods on VAS score in men differed from women. In Ekman P et al. ´s study (2007) on a population including 163 patients with isthmic spondylolisthesis, pain index was measured before surgery, at 1 year, and 2 years of follow-up. Similar to our study, the pain index had a significant decline from pre-operative
 

 

period till 2 years after surgery but unlike our study, no statistical difference was observed in pain at any time interval between the PLIF and PLF groups. Pain was reported not to be significantly different between sexes in this study (17). Cheng et al. conducted a prospective study on 138 patients with spondylolisthesis in 2009. They concluded that VAS score after 4th year was less than the pre-operative period in both PLIF and PLF groups, however, unlike what we concluded, no statistical difference between two groups was seen (4). In 2010, Bròdano et al. in their study on 71 patients with low grade isthmic spondylolisthesis showed that both PLIF and PLF techniques had an acceptable clinical outcome but without statistically significant differences. They assessed

 

 

 

clinical outcome and to define it, what they used was ODI, RMDQ (Roland Morris Disability questionnaire), VAS (leg score and back score), persistent low back pain and persistent sciatica (20, 21). Unlike our study, they did not assess pain separately. Three years later in a meta-analysis by Ye et al. they revealed that three of five studies used VAS to evaluate low back pain, radicular pain or leg pain reduction in PLF and PLIF treated groups (0.5-6 years follow-up). Pain relief over time significantly improved in PLF and PLIF groups (3). In a recent study on 50 patients with lumbar ishmicspondylolisthesis, Habib (2014) also used visual analogue scale to study pain. He demonstrated a significant better long-term VAS score for back pain in PLIF group. A follow-up program of 18 months was scheduled for patients (12). His findings were in consistent with the results of our study. Because the surgery of choice for spondylolisthesis is the field of conflict and pain control in these patients is an important target of management, our findings may hold a clue to the better fusion technique. The limitation of our study was that the two groups were not matched in terms of the level of spondylolisthesis which can act as a confounder. 

Conclusion
As well as the substantial reduction of pain resulted by both of techniques, marked post-operative pain relief in PLIF group was noteworthy.

Acknowledgment
The authors offer their special thanks to the Clinical Research Development Unit of Poursina Hospital, Guilan University of Medical Sciences, Rasht, Guilan, Iran.

Funding
None.

Conflicts of Interest 
The authors report no conflict of interest.


References