REVISION SURGERY FOR PATIENTS WITH PSEUDOARTHROSIS AND ADJACENT SEGMENT DISEASE IN DEGENERATIVE DYSTROPHIC DISEASES OF LUMBOSACRAL SPINE Abakirov M.Dzh., Nurmukhametov R.M., Mamyrbaev S.T., Al-Bavarid O.

Peoples' Friendship University of Russia,

Central Clinical Hospital of the Russian Academy of Sciences,

City Clinical Hospital No. 17 of Moscow City Health Department, Moscow, Russia

Currently, spondylodesis presents the most wide-spread technique for treatment of degenerative dystrophic diseases of the spine [1, 2]. Formation of full-featured bone block is the main task and the final aim of spondylodesis. Absence of the bone block, inconsistency or disorder of integrity of fixing system present one of the main negative results of this intervention, and it often requires for revision surgery. The incidence of formation of full-featured bone block for posterolateral spondylodesis is about 65 % without use of tools, and 90 % for all other techniques [3]. However, despite of successful spondylodesis with instrumental fixation, there is a risk of poor results of surgical treatment in long term period.

One of the important indications for revision surgery after spondylodesis of the spine is pseudoarthrosis [4]. The risk factors of pseudoarthrosis are systemic causes: osteoporosis, diabetes mellitus, rheumatoid arthritis, smoking, anti-inflammatory therapy [5]. The rates of lumbar spine pseudoarthrosis vary from 5 to 35 % [3]. There are some convincing biomechanical and clinical findings that spinal spondylodesis creates significant compensatory increase in movement in the adjacent segment as result of high rigid fixation at the operated level [6]. Development of degeneration or disease of the adjacent segment is considered as a potentially long term complication of spinal spondylodesis [7-8]. The incidence of adjacent degeneration varies significantly and achieves 100 %. It is explained by inevitability of progression of the involutional process. The incidence of clinically significant pathology does not exceed 27.5 %, and the incidence of revision interventions − 15 % [9].

According to the data from Xiao-PengXia et.al., the appearance of degeneration and disease of the adjacent segment after spondylodesis varies within 5-77 % and 0-27 % with combined incidence of 26.6 % and 8.5 % correspondingly [8]. There are a lot of risk factors of diseases of the adjacent segment such as age, gender, body mass index (BMI), surgical techniques and change in sagittal balance [10]. According to some authors, the amount of patients who require for revision surgery for the adjacent segment is 5.6-13.2 % after 7 years [11-12].

The results of revision surgery for long term complications such as pseudoarthrosis and disease of the adjacent segment after decompressive stabilizing interventions have not been studied properly, but are important for estimation of risks and advantages of revision surgery with specific indications. Therefore, this study is oriented to estimation and comparison of clinical and radiological results of revision surgery for pseudoarthrosis and disease of the adjacent segment after primary decompressive stabilizing surgery.

Objective − to conduct a comparative analysis of results of revision surgery in pseudoarthrosis and adjacent segment disease after decompressive stabilizing interventions.

MATERIALS AND METHODS

Totally, 44 patients with pseudoarthrosis and disease of the adjacent segment with revision surgery in 2017-2010 were included. The patients were distributed into two groups, depending on pathology. The group 1 included 20 patients with pseudoarthrosis who were operated with ALIF in combination with transpedicular fixation. The patients' age was 36-68 (М ± SD = 55.6 : 9.9). There were 9 (45 %) men and 11 women (55 %). Pain estimated with VAS before surgery: М ± SD = 7.3 : 0.8 points. Oswestry index М ± SD = 55.4 : 10.0. The group 2 consisted of 24 patients at the age from 38 to 79 (М ± SD = 58.9 : 12.1) with diagnosis adjacent segment disease, who received decompressive stabilizing intervention with TLIF. In this group, the men : women ratio was 10 (41.7 %) : 14 (58.3 %). Pain according to VAS − М ± SD = 6.7 : 0.9. Oswestry index М ± SD = 48.4 : 10.4 (%) (the table 1).

Table 1

Presurgical demographic and clinical comparison of data

Characteristics	Number of patients (%)		p*
	Group I (n = 20)	Group II (n = 24)
Gender:			p > 0.05
male	9 (45.0 %)	10 (41.7 %)
female	11 (55.0 %)	14 (58.3 %)
Age, M ± SD; range (years)	55.6 ± 9.9; 36.0-68.0	58.9 ± 12.1; 38.0-79.0	p > 0.05
Body mass index М ± SD	31.1 ± 4.9	49.3 ± 10.2	p = 0.05
Disk degeneration stage according to Pfirrmann (III:IV)	12 (60 %) : 4 (20 %)	18 (75 %) : 6 (25 %)
ODI before surgery М ± SD	55.4 ± 10.0	48,4 ± 10,4	p = 0.05
VAS before surgery М ± SD	7.3 ± 0.8	6,7 ±0,9	p > 0.05

Note: group I − patients with diagnosis of pseudoarthrosis; group II − patients with diagnosis of adjacent segment disease;* − level of significance of differences according to Mann-Whitney's test.

Clinical and neurological examination: presence of chronic pain syndrome in the back (with VAS > 4 and/or ODI > 30 % within at least 3 last months); neurogenic intermittent claudication; presence of root pain syndrome and sensitivity disorders (any variants) in absence of effect of conservative therapy; movement disorders (muscle strength − 3 points and less).

Complex of presurgical radial diagnosis included: plain and functional X-ray imaging, multi-spiral computer tomography (MSCT), magnetic resonance imaging (MRI). Multi-spiral computer tomography (MSCT) was conducted for all patients for clarification of stability of a metal construct, presence of zones of resorption of bone tissue in the contact site implant-bone, for determination of a criterion of spondylodesis and compressing action on nervous structures of the spinal channel. Lenke points were used for estimation of the bone block with use of X-ray image [13]. Lenke A and B were considered as spondylodesis of the spinal motion segment, and Lenke C and D − as absence of spondylodesis. For pseudoarthrosis diagnosis, presence of the bone block was estimated with MSCT according to recommendations by Choudhri et al. [14]. The bone block was considered as appropriate in presence of at least one continuous bone bridge between vertebral bodies. In the opposite case, (through the interbody implant and around it), the bone block was considered as inconsistent. Simultaneously, the condition of the fixing system was estimated: the fixing system was considered as inconsistent in presence of a bilateral fracture of longitudinal rods at the singe level or in presence of a fracture of both screws in a single vertebra at least and/or in presence of osteolysis around both screws in a single vertebra at least. MRI was conducted for all patients for estimation of degenerative and dystrophic changes in the intervertebral disk according to the classification by Pfirrmann (the table 2, Fig. 1) [15] and reactive changes in the subchondral part of vertebral bodies according to the classification by M. Modic [16]. Spondyloarthrosis of facet joints was estimated according to the improved classification by D. Weishaupt [17]. Diagnosis of recurrent spinal channel stenosis was based on the clinical picture with predominance of root pain syndrome and/or neurogenic intermittent claudication and estimation of radial diagnosis data. Central stenosis was estimated with the classification by Schizaset al., which is based on estimation of distribution of spinal fluid, cauda equina roots and epidural cellular tissue on T2-weighted MR images in axial plane [18]. Lateral stenosis was estimated on the basis of the classification by W.S. Bartynski et al.: degree 1 − deformation of the lateral pocket in comparison with contralateral side; degree 2 − deformation of a root with presence of liquor stripe; degree 3 − rough compression with obliteration of liquor spaces [19]. The classification by S. Lee et al. [20] was used for diagnosis of foraminal stenosis: degree 1 − obliteration of epidural fat tissue on both sides from the root in the intervertebral foramina in the sagittal scans in T1-weighted image; degree 2 − obliteration of epidural fat layer on all sides from the root, but without morphological changes; degree 3 − obliteration with morphological changes.

Table 2

Classification of stages of intervertebral disk degeneration according to Pfirrmann

Stage	Intervertebral disk structure	Signal intensity	Clearness of separation of fibrous ring and nucleus pulposus	Height of intervertebral disk
I	Homogenous	Hyperintensive (white)	Clear border	Normal
II	Non-homogenous	Hyperintensive (white)	Clear border	Normal
III	Non-homogenous	Intermediate (grey)	Non-clear	Normal or slightly decreased
IV	Non-homogenous	Intermediate (grey)	No border	Normal or moderately decreased
V	Non-homogenous	Hypointensive (black)	No border	Disk space collapse

Figure 1

Lumbar spine MRI (T2 mode) [Pfirrmann et al., 2001].

Figure 1 Lumbar spine MRI (T2 mode) [Pfirrmann et al., 2001].

Results of revision operations were analyzed after one year from surgery. Pain intensity was assessed with VAS, which was considered as significant for less than two points. Functional activity and estimation of life quality was assessed with Russian Version of the Oswestry Disability Index 2.1a [21]. For ODI of 0-20 %, vital activity disorders were considered as minimal, for 20-40 % − as moderate, 40-60 % − intense,60-80 % − disabling, 80-100 % − extremely intense or exaggerated. MacNab was used for estimation of personal satisfaction with revision surgery.

Description of statistical methods. The study materials were analyzed with use of methods of parametric and non-parametric analysis. Accumulation, correction, systematization of basic data, and visualization of results were carried out with Microsoft Office Excel 2016. The statistical analysis was conducted with IBM SPSS Statistics 20 (IBM Corporation). For comparison of two and more groups, dispersion analysis (Kruskal-Wallis test) with subsequent paired comparisons was used (Dann's test). Mann-Whitney's test was used for comparison of two groups. Presence of a relationship between studied values was analyzed with contingency tables (Fisher's exact test). Wilcoxon's test was used for analysis of recurrent changes (over time). The differences were statistically significant with p < 0.05.

The conducted study corresponds to standards of Helsinki Declare − Ethical Principles for Medical Research with Human Subjects. The informed consent for personal data analysis was received.

RESULTS

The mean VAS and ODI was 7.3 and 55.4 before and after revision surgery in the group 1 correspondingly, and improved to 1.4 and 11.8 by the moment of the final examination (p = 0.001). The mean VAS was 6.7 and 48.4 in the group 2 correspondingly, and improved to 17.2 and 2.7 by the moment of final follow-up (p = 0.001). Therefore, the mean VAS and ODI improved in both groups. There were some statistically significant differences in VAS between the groups (p < 0.001), but ODI was significantly worse in the group 2 (p = 0.002, the table 3). The results of subjective estimation with MacNab score were mainly positive: the group 1: 54.1 % − excellent, 37.5 % − good, 8.3 % − satisfactory; in the group 2: 40 % − excellent, 30 % − good, 15 % − satisfactory, 10 % − poor. The poor results were determined by preservation or appearance of pain syndrome. The findings are presented in the figure 3. The mean body mass index was higher in the group 2 than in the group 1 (p = 0.05).

Table 3

Comparative results of ODI and VAS before and after surgical management of groups

Characteristics	Number of patients (%)						Level of intergroup significance, p*
	Group I (n = 20)			Group 2 (n = 24)			Level of intergroup significance, p*
Gender	before surgery	after surgery	p**	before surgery	after surgery	p**	p > 0.05
male	9 (45.0 %)			10 (41.7 %)
female	11 (55.0 %)			14 (58.3 %)
Age, M ± SD; range (years)	55.6 ± 9.9; 36.0–68.0			58.9 ± 12.1; 38.0-79.0			p < 0.01
ODI, М ± SD	55.4 ± 10.0	11.8 ± 4.0	p = 0.001	48.4 ± 10.4	17.2 ± 5.2	< 0.001	p = 0.002
VAS, M ± SD	7.3 ± 0.8	1.4 ± 0.4	p = 0.001	6.7 ± 0.9	2.7 ± 1.4	< 0.001	p < 0.001

Note: group I − patients with diagnosis of pseudoarthrosis; group II − patients with diagnosis of adjacent segment disease; * − level of significance of differences according to Mann-Whitney's test; ** − level of significance of differences according to Wilcoxon's test.

According to Lenke estimations with simple X-ray images after revision surgery, the group 1 included 15 cases classified as Lenke A, 4 cases with Lenke B, and one case as Lenke C. The group 2 included 16 cases cassified as Lenke A, 4 cases as Lenke B, 2 cases as Lenke C, and 2 cases as Lenke D (the table 4). The bone block was estimated with simple X-ray images and MSCT. 4 cases of malunion (Lenke C or D) were found in the group 2, and 1 cases (Lenke C) in the group 1.

Table 4

Comparison of Lenke estimates for two groups

Lenke types	Types of abnormality
	Group I (n = 20)	Group II (n = 24)	p-value
А	15 (75 %)	16 (66.6 %)	p = 0.317
В	4 (20%)	4 (16.6 %)	p = 1.000
С	1 (5%)	2 (8.3 %)	p = 1.000
D		2 (8.3 %)	p = 0.333

Note: no reliable differences; p − level of significance of differences according to Mann-Whitney.

The table 5 shows the results of intrasurgical and early postsurgical complications. One of four cases of malunion in the group 1 and two of six cases in the group 2 required for additional secondary revision surgery since conservative therapy was inefficient. Superficial infection of the surgical wound was found in 2 cases in the group 2. Sanitation, opened management of the wound, and antibacterial therapy (according to results of bacteriological studies) were carried out. For injuries to external iliac vein, the integrity of the injured part was restored with application of vascular sutures in 2 (50 %) cases. In the postsurgical period, early activation was performed, and anticoagulant therapy was carried out. There were not any events of venous insufficiency. For sudden injuries to dura matter, defects were covered with muscular flap with use of Evicel surgical glue (Biopharmaceuticals Ltd., Israel) (one case in each group). Moreover, one case of foot paresis was found in each group. These cases disappeared without treatment within six months. The total incidence of complications was higher in the group 2 (6 patients, 25 %) than in the group 1 (4 patients, 20 %, p = 0.01).

Table 5

Comparison of complications after revision surgery

Complications	Types of abnormality
	Group I (n = 20)	Group II (n = 24)	p-value
General amount of patients	4 (20 %)	6 (25 %)	p = 0.01
Pseudoarthrosis	1 (25 %)	2 (33.3 %)	p = 0.667
Neurological deficiency	1 (25 %)	1 (16.7 %)	p = 1.000
Superficial infection	-	2 (33.3 %)	p = 1.000
Dura mater injury	-	1 (16.7 %)	p = 1.000
Iliac vein injury	2 (50 %)	-	p = 1.000

Note: no reliable differences; p − intergroup level of differences according to Mann-Whitney's test.

DISCUSSION

In this study, the patients who received revision surgery for adjacent segment disease after decompressive stabilizing surgery showed worse clinical and radiological results in comparison with patients with revision surgery for pseudoarthrosis. In our study, the patients with diagnosed pseudoarthrosis received surgical interventions with ALIF in combination with spinal transpedicular fixation. An example is presented in the figure 2.

Figure 2

The patient, age of 58. Diagnosis: pseudoarthrosis at L5-S1, non-stable metal construct. VAS before surgery − 8, ODI − 45 %. Revision surgery was conducted, screws were dismounted. The revision screws were installed. Interbody fusion with the cage and autobone was conducted according to ALIF: a) resorption of bone tissue around the screw; b) absence of bone block 1 year after surgery; c) installment of the interbody cage; d) postsurgical X-ray control. The values after 1 year from surgery: VAS − 2, ODI − 15 %.

Butterman et al. conducted a study and performed a retrospective analysis of 38 patients after ALIF for pseudoarthrosis. The rate of spondylodesis was 95 % [22]. High rates of spondylodesis can be achieved with big square of bone surface in the anterior supporting column, good vascularization after preparation of end plates and compression load to implants [23]. According to some authors [24-25], ALIF for pseudoarthrosis leads to radiological improvement, as well as to significant improvement in life quality and decreasing pain syndrome. However there are some studies, which show opposite results. So, according to Seung-Pyo Suh et.al., the clinical and radiologic results were worse in patients after revision surgery for pseudoarthrosis as compared to patients after revision surgery for adjacent segment disease. It is explained by the fact that 29 % of patients after revision surgery for pseudoarthrosis received anterior and posterior spondylodesis, whereas most (71 %) patients received only posterior lumbar fusion (PLF) [26]. Isolated PLF for revision surgery for pseudoarthrosis is complicated by high rate of poor outcomes. West et al. [27] reported on low rate of spondylodesis (65 %) with high incidence of poor outcomes (47 %) in patients with pseudoarthrosis who received posterior spondylodesis with instrumental fixation. Lauerman et al. also reported on low incidence of bone block formation (49 %) in 40 patients with pseudoarthrosis who received revision surgery and posterior fusion with instrumental fixation [23].

The final rate of spondylodesis was equal in both groups. However, the incidence of recurrent surgery was significantly higher in the group 2. The clinical results from perspective of VAS, ODI and subjective satisfaction were better in the group 1. Among the risk factors, high BMI and presurgical disk degeneration at the adjacent level were reliably and independently associated with appearance of diseases of the adjacent segment and could be estimated before surgery. In our study, the mean BMI was higher in the group 2 than in the group 1 (p = 0.05).

Symmons et al. [28] studied women at the age from 45 to 64. The mean follow-up was 9 years. They demonstrated that the BMI increase is the risk factor of intervertebral disk degeneration. Liuke M. et al. [29] also showed that BMI > 25 kg/m² increases the risk of intervertebral disk degeneration in the lumbar spine. According to the above-mentioned results, we supposed that body mass control before and after surgery can decrease the incidence of adjacent segment disease. Anandjiwala et al. [30] performed a prospective study of 74 patients who received instrumental spondylodesis. The minimal follow-up was 5 years. They showed higher risk of adjacent segment disease in patients with presurgical disk degeneration at the adjacent segment. In this study, we confirm that presurgical disk degeneration at the adjacent level is the risk factor for development of adjacent segment disease. So, in the group 2, the degrees 3-4 of degeneration were noted in 18 (75 %): 6 (25 %) as compared to the group 1: 12 (60 %) : 4 (20 %) (the table 1).

Figure 3

Results of subjective estimation of revision surgery according to MacNab.

Figure 3 Results of subjective estimation of revision surgery according to MacNab.

CONCLUSION

In our study, the patients after revision surgery for pseudoarthrosis showed better clinical and radiological results than patients with adjacent segment disease. Patients with excessive body weight or obesity, with presurgical disk degeneration at the adjacent segment, should be completely informed on the risk of adjacent segment disease. ALIF presents the theoretically substantiated and acceptable variant for treatment of pseudoarthrosis and achievement of sagittal balance by means of preservation of lumbar lordosis, resulting in decreasing risk of adjacent segment disease.

Information on financing and conflict of interests

The study was conducted without sponsorship. The authors declare the absence of clear and potential conflicts of interest relating to this article.