The comparative analysis of MRI data in the early period after lumbar microdiscectomies with epidural injection of polyacrylamide hydrogel

1 Department of Miniinvasive and Laser Neurosurgery, Romodanov Neurosurgery Institute, Kyiv, Ukraine 2 Department of CNS Trauma, Romodanov Neurosurgery Institute, Kyiv, Ukraine 3 Department of Neurosurgery, Shupyk National Healthcare University of Ukraine, Kyiv, Ukraine 4 NeuroTrauma Department, Romodanov Neurosurgery Institute, Kyiv, Ukraine 5 Radioneurosurgery Department, Romodanov Neurosurgery Institute, Kyiv, Ukraine 6 Department of Adjuvant Treatment for CNS Tumors, Romodanov Neurosurgery Institute, Kyiv, Ukraine


Introduction
"Failed back surgery syndrome (FBSS)", according to various estimates, is observed in 10-40% of patients after spinal surgery [1,2]. The development of this syndrome after microdiscectomy is associated with structural and biomechanical postoperative changes of the spine (progression of intervertebral disc degeneration (IVD) and spondylarthrosis, the development of instability of the operated spinal motion segment, etc.), as well as with iatrogenic factors, in particular with inadequate intervertebral disc degeneration hernia removal [3]. A significant proportion (about 20-36%) of FBSS cases and about a quarter of reoperations for it are caused by postoperative reactive compression cicatricial adhesive process around the dura mater and neurovascular formations of the spinal canal [4][5][6]. Epidural fibrosis as the main or concomitant pathology is practically an obligatory finding in repeated operations on the spine [6][7][8].
Repeated surgical interventions for epidural fibrosis are associated with a high risk of complications (rupture of the dural membrane, nerve root injury, epidural bleeding) and are effective only in 30-35% of patients, and in 15-20% -lead to worsening of symptoms [7,9,10]. Reoperations do not prevent recurrence of the cicatricial adhesive process and cause thickening of the dural membrane [7,11].
The pathogenesis of epidural fibrosis is a matter of debate. The formation of scar tissue is a physiological response to any surgical injury, but the intensity and http://theunj.org duration of this process depends on many factors. An important mechanism of excessive scarring is the transformation of postoperative hematoma into fibrous tissue [8,12]. A directly proportional relationship between the volume of hematoma and the severity of aseptic productive epidural inflammation has been established [13]. Chronic epidural inflammation can be maintained due to autoimmune response to the appearance of IVD elements in the spinal canal, which has antigenic properties as an extra barrier tissue [14,15]. Epidural penetration of fibroblasts from the connective tissue of paravertebral muscles along the surgical trajectory leads to an increase in collagen synthesis in this area [14]. The destruction of epidural fat and damage to the ligamentum flavum affect the aggressiveness of the epidural scarring process [16]. Some technical features of the operation contributing to its overall traumatic effects are also important [7]: the use of electrocoagulation for the purpose of intracanal hemostasis, excessive traction of the dural sac and curettage of the IVD cavity.
Considering the multifactorial pathogenesis of epidural fibrosis, various approaches have been proposed for its prevention and treatment: both conservative (non-invasive and invasive) and surgical [4,7,[17][18][19][20][21][22][23][24][25][26], but the study of their effectiveness indicates the need for further search for ways to solve this problems. One of the most promising current approaches consists in the use of barrier biological materials (autologous tissues and polymers), which separate the exposed dura mater from the surrounding tissues, minimizing the frictional effect [7,22]. However, the role of modern barrier biomaterials is not limited to a purely mechanical effects as the additional leverage to restrain the development of adhesive epidural process. Thus, polymer materials are used as carriers for controlled release of chemical compounds for targeted delivery of substances that slow down the development of epidural fibrosis (steroids, nonsteroidal anti-inflammatory, immunomodulatory, anti-cancer drugs, statins, inhibitors of transforming growth factor -β (TGF-β) and other profibrotic cytokines) [7,21,[27][28][29][30]. The use of a biopolymer scaffold with implanted autologous stem cells of adipose origin improves the regeneration of epidural adipose tissue [31], while autologous subcutaneous fat grafts demonstrate ambiguous results in reducing epidural adhesion [7,32,33].
There are many clinically or experimentally tested autotransplants as well as biopolymers -natural (chitosan, fibrin gel, hyaluronate, amniotic membrane, gelatin sponge, etc.) and artificial (polytetrafluoroethylene, nano-hydrox yapatite from multi -(amino acid) copolymer, poly lactic-co-glycolic acid, polyethylene glycol etc.) including those with a controlled resorption time [7,9,10,17,22,30,[34][35][36][37]. However, the search for an optimal anti-adhesive barrier biomaterial continues, since none of them has demonstrated a consistently high efficiency and absolute safety [5,22]. The main disadvantages of natural biopolymers are their uneven distribution and insufficient lifetime in the site of their application. Most synthetic polymeric materials are characterized by a reactive inflammatory reaction during their integration into the tissues of the human body [30,38]. Research is continued on biopolymers and methods of their use to prevent postoperative epidural adhesion [22].
Objective: to conduct a comparative analysis of magnetic resonance imaging data obtained in the early period after repeated lumbar microdiscectomies in patients with/without intraoperative epidural injection of polyacrylamide hydrogel "Nubiplant.

Materials and methods
Study participants 84 patients with recurrent IVD hernia were involved in a single-center prospective cohort study. Surgical treatment and analysis of its results were carried out in the clinic of minimally invasive and laser spinal neurosurgery with the X-ray operating room of the Romodanov Neurosurgery Institute, Ukraine.
All patients gave informed and voluntary written consent to participate in the study.
The study was approved by the Committee on Ethics and Bioethics of the Romodanov Neurosurgery Institute, Ukraine (Meeting Minutes No.3 of June 6, 2016).
Inclusion criteria: -age of patients from 18 to 76 years; -repeated surgery to remove IVD hernia and decompression of the spinal root; -early postoperative period; -functional status of the patient according to the Karnofsky scale is not less than 70 points.

Characteristics of groups
The main group (MG) included 30 (35.7%) patients who at the final stage of the operation were epidurally injected with polyacrylamide hydrogel (HG) "Nubiplant", the control group (CG) included 54 (64.3%) patients who underwent reintervention without HG injection.
The age of patients in MG -from 24 to 63 years, in CG -from 20 to 75 years ( Table 1).

Study design
Patients of MG at the final stage of surgery to remove IVD hernia and decompression of the spinal root for the prevention of cicatricial-adhesive epidural fibrosis were epidurally injected with HG "Nubiplant" layer-bylayer in the amount corresponding to the volume of musculoligamentous-bone formations and adipose tissue (2.0-2.5 mm 3 ) removed during the surgical approach. First, the anterior epidural space was filled, then the root and intralaminar space were covered. Thorough hemostasis was achieved.
Patients of CG underwent repeated surgery to remove IVD hernia and decompression of the spinal root, but without the injection of HG. MRI examination was performed on the 3-15th day after surgery on an "Intera 1.5 Tl" tomograph (Philips, Netherlands) with a magnetic field induction of 1.5 T. MRI examination protocol: sagittal and axial T1w_TSE and T2w_TSE standard turbo spin echo, sagittal with suppression of MR signal from adipose tissue (STIR_ TSE), coronal T2w_TSE. Postoperative MRI data of patients of the main and control groups were compared.

Statistical analysis
Databases were formed in Excel tables. Statistical calculations were performed using specialized software Statistica 64 ver.10.0.1011.0 StatSoft Inc. For the purpose of statistical data processing, the input frequency tables were converted into contingency tables which show the frequency distribution of variables in groups taking into account the presence of low frequencies for a sign of pronounced epidural edema in the postoperative area (POA).
The null hypothesis regarding the frequency of epidural edema and hemorrhage manifestations in the groups was tested with the α-level of statistical significance of 5% by Fisher's exact test, that allows accounting for low frequencies.

Results and discussion
Magnetic resonance imaging is a common method of objectifying postoperative changes in patients operated on for IVD hernias, both in the early postoperative period and in the long-term [39][40][41]. The main advantages of MRI include high resolution, which provides contrast in soft tissues imaging, the possibility of 3D imaging and the absence of radiation exposure. In patients with IVD pathology, high contrast (between foraminal or epidural fatty component and IVD or osteophyte, between the cerebrospinal fluid and the spinal cord or spinal root) in combination with multiplanar imaging allows an accurate assessment of IVD hernias and spinal stenosis, in particular during postoperative follow-up.
Despite the wide range of surgical approaches to the treatment of IVD pathology, clinical issues for MRI diagnostics in the postoperative period can be classified into several main categories, directly or indirectly related to the main causes of FBSS development [1,5,[42][43][44]: -IVD hernia (any de novo, relapse at the level of surgery, ineffective herniectomy); -postoperative cicatricial-adhesive epidural fibrosis; -cerebrospinal fluid cysts; -facet degenerative changes; -spinal stenosis (foraminal or central); -infectious complications (from soft extravertebral tissues, epiduritis, arachnoiditis, osteomyelitis, spondylodiscitis); -mechanical instability of the spinal motion segment, spondylolisthesis; -inadequacy of the surgery level; -hematoma, epidural edema, pseudomeningocele; -varicose epidural veins; -pathology unrecognized before the operation or acquired after surgery (traumatic, tumors of the spine and intracanal structures, etc.).
Outlined above categorization is generalized and does not take into account the period of postoperative follow-up. Therefore, it is clear that in a particular observation, MRI data will correlate with the duration of postoperative period and reflect the specificity of the follow-up period (early, intermediate or remote).
MRI diagnostics has some clinical and technical limitations (overweight of the patient, claustrophobia, tremor, the presence of a pacemaker and metal implants (except those made of titanium and tantalum, which cause fewer artifacts than stainless steel)) [44].
There is an ongoing discussion on the advisability of using paramagnetic intravenous contrast with gadolinium compounds in MRI assessment of the results of IVD hernia removal. In our study, intravenous paramagnetic contrast was not used, since MRI was performed in the early postoperative period. Considerations about the inconsistency of the use of contrast in the early postoperative period are based on numerous publications, which indicate a high frequency of accumulation of paramagnetic by various anatomical structures at the level of discectomy and the absence of connection between such findings and clinical symptoms. It was found that during the early and intermediate postoperative period such contrast can be attributed to the "normal variant". It should also be taken into account that, in addition to therapeutic toxicity (nephro-and hepatotoxicity) and the ability of gadolinium preparations to cause allergic reactions, MRI is one of the most expensive diagnostic procedures, especially when using intravenous paramagnetic contrast.
In the first 6 weeks after surgery for IVD hernias 67% of patients show the accumulation of paramagnetic in the space of the operated disc [45], which during this observation period is not attributed to pathological MRI findings. The "normal variant" in the first 6 weeks also includes accumulation of paramagnetic by the annulus fibrosus and spinal roots at the level of the operation. In 20-62% of asymptomatic patients 3-6 weeks after surgery, the accumulation of paramagnetic in spinal roots is observed due to local disturbance of the blood-root barrier [41]. Later than the 6th month, the accumulation of paramagnetic in the spinal root is considered pathological [45]. S.D. Boden et al. demonstrated that after a successful discectomy, accumulation of paramagnetic by facet joints at the surgical level was recorded in 88% of cases, accumulation of paraspinal muscles -in 100% of cases. As the duration of observation increased, the contrast intensity decreased [46].
Considering the above, it is inappropriate to use a paramagnetic during the MRI assessment of the state of POA after discectomy in the early postoperative period in the absence of special clinical requests that require contrast enhancement (for example, if an infectious or neoplastic process is suspected).
Assessment of MRI data obtained in the early postoperative period for IVD pathology is difficult and can lead to improper conclusions. The reasons for this are an increase in the volume of soft tissues of the anterior epidural space and the absence of a clear contour of annulus fibrosus due to edema, tissue integrity disruption and the presence of hemorrhage in POA [40,47,48]. In 80% of cases in the early postoperative period in POA there is epidural edema http://theunj.org with a mass effect, which can be incorrectly assessed as a recurrence of IVD hernia [41]. In addition to the misinterpretation of radiological data as hernia recurrence, the corresponding changes in POA during this period can be misinterpreted as the development of cicatricial-adhesive epidural fibrosis, characteristic of the late postoperative period [41,46].
The main features of the MRI assessment of the state of the POA were set up, taking into account the peculiarities of early postoperative period of lumbar microdiscectomies and the presence of HG in patients of the main group: 1. Epidural edema and increased soft tissue volume (can affect hernia recurrence).
3. The degree of deformation of the dural sac. 4. Shape, clear contours, localization of the spinal root, increase in its volume, compression, dislocation.
5. The size of the HG zone, the homogeneity of the signal from the HG and the degree of coverage of the spinal root (<90°, 90-180°, >180°).
Criteria for the correctness of HG injection according to MRI data in the early postoperative period have been developed: 1. 6. Sectoral coverage of the root by epidural HG >180°.
The degree of spinal root coverage and the duration of this effect depend on both the physical characteristics of HG (adhesion, density, fluidity, homogeneity, reverse deformation, etc.) and on injection technique, in particular careful pre-mobilization of the root by radiculolysis during IVD hernia removal. According to our intraoperative and neuroimaging comparisons, full coverage is achieved precisely when the root is displaced, which allows filling the ventral zones of the epidural space. The proposed criteria for the correctness of the HG location are simple, but help to comprehensively assess the specifics of changes after surgery. This makes it possible to unify approaches to radiological assessment of the corresponding category of patients.
The data on the degree of spinal root coverage with hydrogel in the main group are given in Table. 2.
The presence and severity of certain parameters of the location of HG "Nubiplant" in the studied patients were also assessed ( Table 3).
Tomograms of patients obtained in the early postoperative period af ter repeated removal of IVD hernias are shown in Fig. 1 (while injecting HG "Nubiplant") and Fig. 2 (without the injection of HG).
Considering that epidural edema and hemorrhage in the early postoperative period are associated with the development of cicatricial adhesive epidural fibrosis in the later follow-up period [41,46], a comparative analysis  D of the distribution of these signs in the study groups was carried out to assess the impact of HG administration on the possibility of creating preconditions minimizing the development of cicatricial-adhesive epidural fibrosis ( Table 4). Hemorrhage in POA did not cause a mass effect with the corresponding clinical symptoms and did not require repeated surgical intervention in none of the observations in MG and CG.
The null hypothesis of no difference in the frequency of MRI signs of epidural edema and hemorrhage between MG and CG was rejected (p = 0.0444 and p = 0.0288, respectively, Table 5), that is, the groups were statistically significantly different in MRI signs. The p-value for testing signs of epidural edema is close to the level of α-value.
Thus, the use of HG "Nubiplant" for repeated microdiscectomies is accompanied by a statistically significant decrease in the frequency of MRI signs of epidural edema and hemorrhage in the POA in the early postoperative period, taking into account the pathogenetic mechanisms of cicatricial adhesive epidural fibrosis. http://theunj.org

Conflict of interest
The authors declare no conflict of interest.

Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent
The informed voluntary written consent to participate in the study was obtained from each patient.

Funding
The research had no sponsor support.
A promising line of research is a comparative analysis of MRI data in patients with recurrent IVD hernias, who were injected with HG "Nubiplant" for lumbar microdiscectomy, and those who did not use it, in later periods of observation. In our opinion, such findings will allow us to study the predictive effect of the use of this anti-adhesive barrier biomaterial on the development of postoperative epidural fibrosis and FBSS. In addition, it is advisable to conduct a multivariate analysis of the influence of various clinical factors (in particular body mass index, degree of sarcopenia, comorbidity, degenerative-dystrophic changes of the spine, etc.) on the results of repeated microdiscectomies using HG "Nubiplant". Such studies, provided that the sample size is increased to enhance the statistical significance of obtained results, would contribute to the development of treatment algorithms with a high degree of evidence and individualization of approaches to the choice of optimal surgical approach while ensuring a high quality of life in the postoperative period. Thus, our study indicates the need for further comprehensive analysis of the results of lumbar microdiscectomies when using HG "Nubiplant" in later follow-up periods.

Conclusions
1. It was found that in the early period after repeated lumbar microdiscectomies (on the 3-15th day) intraoperative epidural injection of "Nubiplant" hydrogel is accompanied by a statistically significant decrease in the frequency of MRI signs of epidural edema and hemorrhage in the postoperative area (p = 0,0444 and p = 0,0288, respectively).
2. When assessing the correctness of epidural injection of "Nubiplant " hydrogel during lumbar microdiscectomies in the early postoperative period, it is advisable to use the following MRI criteria: no deformation of the dural sac and dislocation of the spinal root, clear contours of the spinal root, uniformity of the signal from the hydrogel on T-2 and T-1 weighed images, nonvisualization of hydrogel areas outside the postoperative area, sufficient sectoral coverage of the root with the hydrogel (optimally >180°).