Background. The inflammatory index can be useful for neurosurgeons to understand and grade pain in degenerated intervertebral disc (DIVD).
Objectives. The study focused on the value of the platelet-to-lymphocyte ratio (PLR), neutrophil-to-lymphocyte ratio (NLR) and the inflammatory multiple indices (MIs), and aimed to compare its efficiency with the preoperative and postoperative pain scale and scoring algorithms.
Materials and methods. A total of 88 DIVD patients were included in this retrospective clinical cohort study. Visual Analogue Scale Back (VASB) and Visual Analogue Scale Leg (VASL), Oswestry Disability Index (ODI), Roland–Morris Disability Questionnaire (RMDQ), and walking distance (WD) were used to assess pain. The multiple index (MI) was calculated as MI-1 = PLR × C-reactive protein (CRP) and MI-2 = NLR × CRP.
Results. Comparing the MI with ODI, no correlation was found in preoperative values, while a positive correlation (MI-1: r = 0.398, p < 0.001; MI-2: r = 0.285; p = 0.007) was found between the postoperative measurements. A significant correlation was found for VASB and both MI-1 (preoperative: r = 0.373, p = 0.001; postoperative: r = 0.232, p = 0.041) and MI-2 (preoperative: r = 0.388, p < 0.001; postoperative: r = 0.206, p = 0.044). The MI-1 index showed 71.4% sensitivity and 73.3% specificity, while the MI-2 index exhibited 78.6% sensitivity and 68.9% specificity.
Conclusions. MI-1 and MI-2 showed a positive correlation with pre- and post-operative VASB score and had strong potential to predict postoperative pain in DIVD. They are easy-to-use, noninvasive and low-cost indices; therefore, our results are promising for routine application.
Key words: pain scoring, degenerated intervertebral disc, inflammation index
Low back pain is accepted as a derivative of multi-origin somatic pain with its physical, psychogenic and social aspects together, and the degenerated intervertebral disc (DIVD) problem emerges as the most important source of this pain in the lumbar spine.1 Stretching and degeneration of the facet joints, especially in the posterior location of the vertebral body, is often pointed out as an important factor contributing to pain.2 The only conventional surgical treatment available to prevent further deterioration is based on the surgical fusion of the 2 vertebral bodies, namely spinal fusion, spondylodesis or arthrodesis, although its clinical results vary.3 Currently, the biggest surgical difficulty is diagnosing the problematic disc region according to the source and degree of pain.4 Analyzing the inflammation that occurs in the painful area and evaluating the correlation with pain will be illustrative in the detection and grading of pain, and the choice of surgical approach.5
The pathophysiology of the anatomical deterioration of the disc is a set of symptoms that develop as a result of inflammatory developments and remodeling facet joint processes.6 In addition to prostaglandins, numerous cytokines such as tumor necrosis factor (TNF), interleukin 1 (IL-1) and IL-6 worsen this picture by causing osteoarthritic changes.4, 7 These cytokines and matrix-degrading enzymes disrupt chondrocyte metabolism and thus lead to cartilage transformation. In a vicious circle, changes in cartilage cause intense pathological remodeling in the subchondral bone, stimulating a recurrent inflammatory process.8, 9 It may be easier to follow these processes clinically in order to investigate their relevance to pain using hematological (platelet) indices, indirectly influenced by pain, rather than expensive and difficult-to-measure markers such as TNF and IL.5 The inflammatory process frequently causes changes in numerous hematological parameters, such as peripheral blood cell counts and levels of C-reactive protein (CRP). In comparison, peripheral blood cell counts are easy to measure, inexpensive and widely available in routine clinical practice.10, 11
Hematology-related inflammatory indices, including platelet-to-lymphocyte ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR), are used as single hematological parameters in routine laboratory examinations for inflammatory disease and are investigated for potential new roles.12, 13 However, their combination with CRP has not been performed for inflammation-related disease, which had mostly an inflammatory process similar to DIVD.14
These inflammatory indices can be useful for neurosurgeons to understand and grade pain in DIVD, and consequently, to plan the surgical approaches. The study focused on the value of multiple indices (MIs) combined with PLR, NLR and CRP for DIVD as compared to preoperative and postoperative pain scale and scoring algorithms.
Materials and methods
This study was conducted in adherence to the Declaration of Helsinki and with an approval of the ethics committee of Siirt Training and Research Hospital, Siirt, Turkey (approval No. 2021/06/15, 12782). Eighty-eight patients diagnosed with DIVD from 2018 to 2021 were enrolled into this retrospective clinical cohort study. Informed consent was obtained from all patients included into the study.
Patient selection criteria
According to our inclusion criteria, the DIVD patients enrolled in the study should have had the following characteristics: having experienced chronic low back/leg pain for at least 3 months, not responding to regular physical therapy, and receiving drugs with anti-inflammatory activity for at least 1 month. Patients with comorbidities other than DIVD were not included in the study. The NLR and PLR are useful markers for assessing inflammatory response and disease activity in autoimmune diseases, including systemic lupus erythematosus (SLE). In recent years, markers of inflammation (e.g., NLR) have also been studied in cardiovascular disease (CVD). They could influence many factors, such as malignancies and viral infections, and thus individuals with ongoing infectious or inflammatory diseases receiving treatment for a different disease or systemic disease, or having a different clinical condition diagnosed before or during surgery were excluded.
We used Carragee classification, which divides disc herniations into 4 types as follows: 1) fragment-fissure herniations (with minimal annular defect disc herniations and the presence of an extruded or sequestered fragment); 2) fragment-effect herniations (presence of extruded or sequestered fragment with extensive annular tear over 6 mm); 3) fragment-contained herniations (intact annulus but with 1 or more fragments below the annulus; such fragments are removed by oblique incision to the annulus); 4) no fragment-contained herniations (characterized by the intact annulus and no free fragments under the annulus). Since there were few cases in the type 1 and we performed a different surgical procedure than to types 2, 3 and 4, Carragee type 1 was not taken into account to ensure standardization of the study.
The surgical intervention included a single-level unilateral partial hemilaminectomy and foraminotomy to remove the herniated disc and to decompress the affected nerve root. A skin incision was made in the middle of the back over the appropriate vertebrae. The length of the incision varied depending on how many laminectomies were performed. The back muscles were split in half and moved to either side, exposing the lamina of each vertebra. After the lamina and ligamentum flavum were removed, the protective covering of the spinal cord (dura mater) was made visible. The protective sac of the spinal cord and nerve root was gently pulled back to remove the bony spurs and thickened ligaments. Facet joints, located directly above the nerve roots, were undercut to make more room for the nerve roots.
Clinical assessment and scaling
The magnetic resonance imaging (MRI) findings were recorded to clarify the diagnosis of DIVD and follow-up results of patients with similar clinical conditions, including the type and location of the herniated disc. To evaluate the clinical changes of DIVD that occurred following surgery, pain, disability and walking distance (WD) were measured and the pre-and postoperative periods were compared by means of demographic data and indices. Visual Analogue Scale Back (VASB) and Visual Analogue Scale Leg (VASL) were used to evaluate the pain conditions at preoperative and postoperative period.15 Oswestry Disability Index (ODI) was used to measure the back and leg pain in terms of pain reduction and functional improvement following the surgery.16 The ODI responses were requested from all DIVD participants of the study, taking into account the severity of their complaints in the last 10 days, and the results were expressed using the 0–50-point system. The Roland–Morris Disability Questionnaire (RMDQ) was used to evaluate the self-rated physical disability caused by low back pain.17 In addition to those, we observed the WD for all the patients at preoperative and postoperative period.18
Laboratory data analysis
Study data obtained from the participants included the demographics (such as age, gender, etc.), clinical records, and pre- and postoperative routine blood examination results. Venous samples were taken from all patients at the day before the surgery and the day following the surgery. Blood samples were taken during non-medication hours. The MIs were calculated using the following equations:
MI-1 = PLR × CRP
MI-2 = NLR × CRP
and compared with pain scales.
All data analyses were performed using the IBM SPSS v. 25 software (IBM Corp., Armonk, USA), and statistical significance was considered if p < 0.05 was obtained. The Shapiro–Wilk test was used to evaluate the distribution of the index presented in Table 1 (PLR, NLR, MI-1, and MI-2) and the surgery data presented in Table 2 (ODI, RMDQ, VASL, VASB, WD). All the parameters indicated a normal distribution. Hence, we compared patients’ preoperative and postoperative outcomes and index using a paired t-test. In the Spearman’s correlation analysis, we evaluated the relationship among all the data. The receiver operating characteristic (ROC) curve was drawn to estimate the diagnostic value between the preoperative and postoperative pain values of the participants and to determine the cutoff value for analysis (Figure 1). A receiver operating characteristic (ROC) curve analysis was performed to assess the diagnostic value of the inflammatory indices.
Preoperative and postoperative comparisons
In the preoperative and postoperative pain assessment, leg pain values according to VASL decreased from 8.02 ±1.5 to 1.32 ±0.9 (p < 0.001). Similarly, according to VASB, back pain values decreased significantly following the surgery (from 4.77 ±2.26 to 2.5 ±1.12; p < 0.001). The ODI values regressed from 82.9 ±7.9 preoperatively to 14 ±10.1 postoperatively (p < 0.001). Similar to ODI, RMDQ also showed a strong statistical decrease (from 17.9 ±3.2 to 5.14 ±2.5; p < 0.001). When examined in terms of WD, a strong increase was noted as a result of the surgery. The patients who could walk 57 ±23.1 m on average before surgery reached an average WD of 430 ±114 m postoperatively (p < 0.001). Scale values before and after surgery are given in Table 2.
Correlation with demographics
In the DIVD patients, the number of comorbidities (r = 0.729, p < 0.001) and length of hospital stay (r = 0.318, p < 0.001) showed a significantly positive correlation with age. While body mass index (BMI) was not associated with the length of hospital stay, it showed a significantly correlated increase for hospitalization duration (r = 0.318, p < 0.001), similarly to correlation with age. In the analysis of the correlation of BMI and age with MI-1 and MI-2, it was observed that these 2 parameters increased with increasing age (r = 0.344 and r = 0.317, respectively; p < 0.001) but were not related to BMI and comorbidity count. There was a strong positive correlation between the length of hospitalization and MIs (r = 0.613 and r = 0.741, respectively; p < 0.001). Follow-up time showed a weak correlation with MI-1 (r = 0.232, p = 0.029) and a stronger correlation with MI-2 (r = 0.321, p = 0.002).
Correlation with pain scales
The correlation analysis was performed between the scales and the indices (MI-1 and MI-2) before and after the operation. When comparing indices with ODI, no significant correlation was found in preoperative values, while a strong positive correlation (MI-1: r = 0.398, p < 0.001; MI-2: r = 0.285, p = 0.007) was found with ODI in the postoperative period. In comparison with the RMDQ, a positive relationship was found with the MI-1 index, regarding both preoperative (r = 0.257; p = 0.015) and postoperative (r = 0.501; p < 0.001) values. As regards MI-2, positive correlations were found with both preoperative (r = 0.294; p = 0.005) and postoperative (r = 0.378; p < 0.001) RMDQ values. In terms of VASL, neither preoperative nor postoperative scores were found to be correlated with MI-2; a weak (r = 0.227; p = 0.033) positive relationship was found between MI-1 and postoperative VASL only. However, a significant correlation was found between preoperative and postoperative VASB scores and both MI-1 (preoperative: r = 0.373, p = 0.001; postoperative: r = 0.232, p = 0.041) and MI-2 (preoperative: r = 0.388, p < 0.001; postoperative: r = 0.206, p = 0.044).
The MI-1 index showed 71.4% sensitivity and 73.3% specificity when the cutoff value was set at 370, while the MI-2 index exhibited 78.6% sensitivity and 68.9% specificity when the cutoff value was set at 7.4. According to the ROC curve values, while NLR and PLR indices are not effective diagnostic tools for VASB according to ROC curve values, MI-1 and MI-2 seem to be powerful diagnostic parameters (Table 2). The ROC data of VASB including PLR, NLR, MI-1, and MI-2 are presented in Table 1.
Nowadays, when diagnostic costs have increased considerably and healthcare systems cause financial problems for governments and private institutions as a result of these costs, the need for cheap diagnostic parameters has increased more than ever. In this sense, since the investigated MIs are a strong indicator of inflammation, are very useful and have a low cost, their results and efficiency in DIVD patients shown in the present study will give them momentum to be used routinely before and after the surgery.
Although inflammation as a type of response mechanism to injury or damage to the skeletal system and related structures, inflammation plays a critical role in the development and worsening of DIVD.11 In the tissue where chronic inflammation occurs, intense inflammatory secretion formed by cells such as macrophages and platelets is observed.19 Recent studies have shown that hematological indices change as a result of the systemic inflammatory response and hematological indices are defining factors in the prognosis of various diseases and can be effectively used in follow-up.4, 13, 19, 20 Under these inflammatory conditions, the increasing neutrophil population can produce large amounts of nitric oxide, the effect of which has been shown regarding pain, resulting in worsening of the clinical condition.
Publications are showing that the measurement of neutrophils in the blood or the values of NLR can predict the outcome of diseases associated with the presence of inflammation.13 In this manner, lymphocyte and platelet cells, similar to neutrophil cells, can secrete cytokines, which can further reduce the number and function of lymphocytes, and, in consequence, weaken immunological surveillance and defense.10 In such a case, increased neutrophil-platelet and decreased lymphocyte count can be considered a response to systemic inflammation and accepted as a predictor and damage indicator.14
We acknowledged that systemic and local inflammatory conditions occurring around the skeletal system have a critical role in both development and healing of disc herniation. In a recent study, Ethemoğlu and Erkoç analyzed 126 patients with neck pain.21 They recorded the PLR, NLR, and CRP, and compared them. The cervical disc hernia showed a higher CRP and NLR when compared to discs with a healthy MR image. In their study, comparing patients with cervical disc herniation to patients with neck pain, they found that in those with disc herniation, NLR and CRP were higher. The authors made a recommendation that early preventive approach applied by physicians in patients with high NLR and CRP may play a preventive role in disc degeneration and hernia. In a study conducted by Bozkurt et al. to investigate the relationship between NLR and pain intensity, a significant positive correlation of VAS scores was found when they was used to assess pain in the preoperative and postoperative period in the patient group with lumbar disc hernia.22 Although they found negative correlation coefficients between pain levels and preoperative VAS scores, they found a positive correlation between postoperative VAS and pain intensity. Bozkurt et al. assessed the relationship between NLR and pain severity in the preoperative and postoperative period using VAS in patients with lumbar disc herniation and found a significant positive correlation. Negative correlation coefficients were found between the serum urate (SU) levels and preoperative VAS scores; in contrast, positive correlation coefficients were found between the SU levels and the postoperative VAS scores.22
In our study, in preoperative and postoperative pain assessment, leg pain values according to VASL decreased following the surgery. Similarly, according to VASB, back pain decreased significantly following the surgery. Likewise, ODI and RMDQ showed a strong decrease after the surgery. When examined in terms of WD, a strong increase was noted as a result of the surgery. When the indices were compared with ODI, no significant correlation was found in preoperative values, while a strong positive correlation was found between the indices and ODI in the postoperative period. When the MIs were compared with the RMDQ, a positive correlation was found with the MI-1 index (with both preoperative and postoperative values). Similarly, regarding MI-2, positive correlations were found with both preoperative and postoperative RMDQ values. In terms of VASL, neither preoperative nor postoperative scores were found to be correlated with MI-2 scores. A significant correlation was found between preoperative and postoperative VASB scores and both MIs.
The MI-1 and MI-2, whose diagnostic value we investigated, are useful guides for us because they are cost-effective and easily accessible. In ROC analysis, the MI-1 showed 71.4% sensitivity and 73.3% specificity, while the MI-2 exhibited 78.6% sensitivity and 68.9% specificity. With the inclusion of CRP into both indices, they have become much stronger diagnostic indices than PLR and NLR. Thus, MI-1 and MI-2 are more objective markers than PLR and NLR, and paved the way for the use of MI in DIVD prognosis.
There were also some limitations to the present study. First, as it is a retrospective cohort study, patients’ selection bias may exist and the number of participants was limited due to the hospital records. The 2nd limitation is that hemogram results may be affected by some unexpected factors such as undetectable infections, abnormal blood circulation, local inflammation, and nutrition, and we were unable to measure how it affected the index. We tried to obtain the best cutoff value using the ROC. However, the best approach to solve this problem is to perform a prospective, large, single-center study.
The assessment of the systemic inflammatory condition in DIVD may prove beneficial to predict preoperative and postoperative pain. The 2 MIs, MI-1 and MI-2, had a strong significant correlation with preoperative and postoperative VASB scores. They also allowed for a reliable prediction for the pain in the postoperative period in DIVD. These parameters could give surgeons an idea about the pre- and postoperative pain and inflammation in the patients, and thus they would provide cost-free gains in drug administration in order to alleviate pain before and after the operation, and in predicting the healing processes. In addition, the success of performed surgery measured in terms of pain can be an important issue that can be followed up indirectly. These indices are universally available, noninvasive and low-cost, and hence there is a perspective of their routine application.