Prognostic Factors in Zygapophyseal Osteoarthritis Infiltration: a Prospective Cohort Study

1. Introduction: Low back pain is one of the significant health problems; studies have shown that it can affect 80-85% of people throughout life; consisting of the second major cause of work withdrawal. In the investigation by lumbar spine imaging methods, facet osteoarthritis (FO) is one of the most prevalent diagnoses, in 59.6% for men and 66.7% for women. The process of osteoarthritis begins early in the spine; more than half of adults are under 30 years old, presenting degenerative changes in the facets, being more commonly affected by the L4-L5 level. It is estimated that the pain of facet origin corresponds to 30% of cases of low back pain. The treatment of low back pain, which presents questionable results, has a substantial economic impact on the health system and represents a growing concern for public and medical policymakers. The lack of diagnostic accuracy and the various types of treatment without evidence of appropriate effectiveness are considered the leading causes of the increasing costs of the health system. Biomechanical, neuroanatomical, and neurophysiological studies demonstrated free and encapsulated nerve endings in the lumbar facet joints and nerves containing substance P and peptides related to the calcitonin gene. Chemical or mechanical stimuli can cause pain. The pathophysiology of facet pain is not fully understood, but it is believed that facet osteoarthritis is due to repetitive mechanical stress, with subsequent inflammation and elongation of the joint capsule, generating pain.

   Several treatments are used in low back pain secondary to FO. The most used are non-hormonal anti-inflammatory drugs and physiotherapy as a conservative measure. Interventional procedures such as facet intraarticular or periarticular infiltration of corticoid for neural blockade and denervation by radiofrequency are considered minimally invasive options, indicated for cases in which conservative treatment fails. Surgical treatment of arthrodesis is also part of the therapeutic options. A systemic corticosteroid is within the available therapeutic arsenal but has undesirable side effects, and its use is restricted in a short period. In order to increase drug concentration at the symptom site, without the undesirable side effects of the systemic corticosteroid, facet intraarticular infiltration (FI) has been used as a complementary treatment. The therapeutic value of FI is not consensus, with widely divergent results in the literature. Nevertheless, it is one of the most used treatments, observing a growing demand for this procedure in the last decades. The selection of the appropriate patient, who would benefit from the treatment, is a determining factor in the outcome of the FI since, in the thesis, only the subgroup of patients with the active inflammatory process in the facet joint would have a positive result corticosteroid infiltration. Identifying this subgroup of patients is a challenge because the diagnosis of facet origin pain is often imprecise. With proper and accurate diagnosis, FI is expected to achieve better results. Magnetic resonance imaging (MRI) can identify the morphological characteristics of facet osteoarthritis, generating a classification in degrees mild, moderate, or severe. It also allows diagnosing the presence or absence of an inflammatory process in the facet joint, usually manifested by indirect signs such as bone edema, periarticular soft-tissue edema, joint effusion, and synovitis. However, morphological characteristics and inflammatory markers of MR imaging have not been tested as prognostic factors of FI efficacy. Patients' characteristics such as gender, age, schooling, family income, smoking, body mass index, professional activity may affect the expected results, but they have not been tested as prognostic factors of FI efficacy.

   1.1 Rationale: The relationship between MRI findings of facet osteoarthritis and the effect on FI of corticoid and anesthetic has not been studied. The relationship between the characteristics of age, sex, education, family income, smoking, body mass index, professional activity, and the effect on FI of steroids and anesthetics has not been studied.

   Determining the prognostic factors may impact the clinical routine of physicians and patients, guiding the treatment, and selecting those with chances of better results with the FI.

   1.2 Hypotheses: MRI findings of facet osteoarthritis and patient characteristics are prognostic factors for improving or worsening the clinical outcome after treatment with FI.

2. Purpose: To determine the prognostic factors of corticoid and anesthetic infiltration in treating facet osteoarthritis in the lumbar spine. To determine the association between MRI imaging findings and pain, functional disability, and patient's quality of life. To determine the association between age, sex, schooling, family income, smoking, body mass index, professional activity in pain, functional disability, and patient's quality of life.
3. Method: This prospective cohort study will include all patients with clinical suspicion of facet pain who have undergone MRI and have been referred to the intervention center of the image department of a tertiary hospital to perform FI with corticoid and anesthetic. On the day of the procedure, patients will be invited to participate in the study. Before performing the facet infiltration, patients will respond to an informed consent form (ICF) and five questionnaires: 1. data collection form; 2. the visual analogue pain scale (VAS); 3. Roland-Morris questionnaire on functional disability; 4. Oswestry questionnaire on functional disability; 5. EuroQol questionnaire (EQ-5D) on quality of life. All patients will be followed for one year, responding by telephone VAS, Roland-Morris, Oswestry, and EQ-5D questionnaires at 1, 3, 6, and 12 months after the FI to measure the effectiveness of the proposed treatment. The systemic use of corticosteroids, the performance of another infiltration or surgery, and the number of analgesics ingested per day will also be evaluated in the follow-up. Researchers will only apply the specific questionnaires of the project, not influence the indication, accomplishment, and treatment of FI complications or the clinical routine of the intervention center. The following variables will be collected to characterize the sample: age (years), sex (male or female), schooling (years of study), smoking (number of cigarettes per day), family income, professional activity (type of occupation, number of hours sitting or standing, carrying weight or not) and oral analgesic use (number of tablets taken per day). The FI procedure will be guided by computed tomography. The radiation dose used will be the minimum possible to obtain images with sufficient quality to guide the location and progression of the needle, being within limits established in the institutional protocols and already practiced in the department. After imaging confirmation of the correct intraarticular positioning of the needle, a solution containing triamcinolone 20 mg/ml and hydrochloride of naropin in a ratio of 1: 1, in the volume of 1.5 to 2.0 ml, will be injected. After the procedure, the patient will be discharged and advised to do home rest for 24 hours. The study design proposes to determine the correlation between the clinical outcome measured by the results of the VAS, Roland-Morris, Oswestry, and EQ-5D questionnaires, and the variables identified by MRI and also the correlation between the clinical outcome and the variables: age, sex, educational level, smoking, family income, body mass index (BMI), professional activity and amount of analgesics ingested per day. All patients will be followed for one year, responding by telephone VAS, Roland-Morris, Oswestry, and EQ-5D questionnaires at 1, 3, 6, and 12 months after the IF to measure the effectiveness of the proposed treatment. The systemic use of corticosteroids, the performance of another infiltration or surgery, and the number of analgesics ingested per day will also be evaluated.

3.1 Image exams and evaluator: All the lumbar spine MRI scans performed by the patients will be analyzed. Two radiologists will read MRI exams without knowledge of clinical, laboratory, or patient evaluation and questionnaire results. The variables degree of facet osteoarthritis, bone edema, periarticular edema, synovitis, effusion, osteophytes, subchondral sclerosis, subchondral cysts, sliding, sagittal or horizontal orientation, identified by MRI, will be tabulated.

3.2 Rating: MRI studies will be evaluated respecting patients' anonymity. Patients' characteristics, MRI imaging findings, and VAS, Roland-Morris, Oswestry, and EQ-5D questionnaires will be tabulated according to the following characteristics: age, sex, schooling, family income, smoking, weight, height, body mass index (BMI), occupational activity, analgesic ingestion, laterality, location, degree of facet arthritis, bone edema, periarticular edema, synovitis, effusion, osteophytes, subchondral sclerosis, subchondral cysts, orientation sagittal or horizontal, facet sliding, VAS, Roland-Morris, EQ-5D, and Oswestry.

3.3 Sample size estimation: Taking into account the primary variable of the study as the presence of inflammatory process in the facet joint, identified at MRI, with an expected prevalence of 30% in the sample and as the primary outcome the functional disability, as measured by the Roland-Morris questionnaire, when checking the sample size needed to detect a difference between the groups, with and without inflammatory process, before and after the FI procedure, it will require 117 patients in total, with a difference in the mean of the groups of 4.0 points. This calculation considered that the standard deviation of the difference of averages between the groups would be homogeneous and 7.0 points, as observed in the literature. Therefore, this study will include 147 patients considering a loss of follow-up in 20%. The sample size needed to compare the groups will be recalculated once the study has a three-month follow-up of 70 patients based on the SD and the proportion of facet inflammation observed. The calculations were made considering 80% test power and a significance level of 5%.

3.4 Statistical analysis: Patient characteristics, MRI findings and pain (VAS), disability (Roland Morris and Oswestry), quality of life (EQ-5D) scores will be described using absolute frequencies and percentages in categorical variables, or through means, standard deviations or medians and quartiles, in addition to minimum and maximum values, in quantitative variables. The MRI characteristics analyzed by two observers will be evaluated for agreement using correlation coefficients appropriate to the observed distribution of cases, such as the Kappa coefficient. To evaluate MRI findings and patient characteristics (age, sex, smoking, family income, BMI, education, occupation, and analgesic intake) in the evolution of pain (VAS), disability (Roland Morris and Oswestry), and quality of life (EuroQol) during twelve months after infiltration, will be used mixed linear models or generalized linear models, in multiple approaches, being such models appropriate for measurements obtained from the same patient at different times. The results of the models will be presented by effect estimates (change in mean pain or means of functional disability and quality of life) and 95% confidence intervals.

We plan to carry out the analysis with the statistical package R and consider the significance level 5%.