Fusion or Not - a Study on Low Back Pain

Chronic low back pain is one of most common reasons for reduced health related quality of life in the world. Fusion surgery is a controversial treatment for low back pain, but still widely performed. The hypothesis is that instability cause back pain and that fusion reduce instability and pain. Despite research for years this hypothesis remains obscure.

Four randomized controlled trials comparing fusion surgery with non-surgical treatment have been performed, of which three did not show a clear benefit of surgery compared to structured non-surgical treatment at short term. None of the studies had an observational group that did not receive any specific treatment. One study showed that surgery was more efficient than a referral to physiotherapy at short term, but the advantage of surgery deteriorated with longer follow-up.

The above studies have led to lack of support for fusion for low back pain in spine research societies and in some nations. However, fusion for low back pain is common world-wide. In Sweden, fusion in low back pain patients constitutes about 6% of all spine surgeries, and is performed on more than 600 patients annually.

Giving the limits of uncertainty in a study comparing a very invasive treatment such as surgery with a non-surgical treatment, which in some of the studies also was similar to the offered pre-study treatment, better study designs are desired. All the above randomized controlled trials were affected by cross-over (around one quarter of the patients over time). Crossover is only possible from non-surgical treatment to surgical treatment, making actual treatment comparisons difficult.

All patients included in studies comparing surgical and non-surgical treatment are considered to be surgical candidates, and expectations for surgery are higher than for non-operative treatment. Some of the patient-experienced effect of the fusion may be explained by the expectations for and the rituals linked to surgery. This is labelled the placebo effect. Previous double-blind trials evaluating the efficacy of commonly conducted surgeries for shoulder and knee pain found that the improvement in patients undergoing the surgical procedure was no better than in patients allocated to sham, or placebo, surgery.

The placebo response includes both the natural course and regression to the mean and possibly other causes in addition to surgery and has been shown to be an important factor for improvement. This has particularly been shown in trials where surgery is aimed at treating chronic pain. In fact, in meta-analyses, surgery was not always better than surgery without the active intervention in chronic pain conditions. Since invasive interventions are frequently associated with larger placebo-effects (compared to non-invasive treatments), there is a large risk that surgery trials give biased results unless they include a control group that undergo surgery without the active intervention.

As a result, the efficacy of surgical interventions to treat chronic pain is overestimated and widely debated, and the ethical justification is questioned, since surgical procedures involve risks for patients. For example, one quarter of patients treated with fusion for low back pain is reoperated within 5 years. Therefore, it is important to assess to what degree the positive effects are due to specific efficacy of the surgical intervention or to non-specific factors such as the placebo effect.

The only way to evaluate the causal effect of surgery is to perform a randomized controlled trial that includes a placebo, a sham, surgery arm. Based on our recent experience, performing a surgical trial with a placebo-arm is possible in Sweden, Finland, Norway and the UK.

A fusion, as mentioned previously, is expected to decrease symptoms in patients with low back pain and degeneration of one or more discs due to "stabilization" of the painful segment. There are several concerns in the rationale for doing fusion surgery for low back pain. The diagnosis is uncertain and is a combination of clinical symptoms and MRI findings of degeneration. However, findings of disc degeneration are common in an asymptomatic population.

Data from other studies indicate that low back pain is reduced by surgery even if it does not involve a fusion. Spinal surgery due to lumbar spinal stenosis, with or without fusion, decreases back pain of 2 or more on the NRS 0-10 scale. Not all fusions heal, and there is no clear indication that individuals with an unhealed fusion have a worse patient reported outcome than those with a healed fusion. Flexible but restraining posterior spinal constructs that do not lead to a completely stable segment lead to a similar outcome as posterior spine fusion. Fusion in addition to decompression is not effective in patients with a possible spinal instability, degenerative spondylolisthesis.

In fact, we do not know why spine surgery without fusion leads to improvement of back pain- is it the natural course of the disease, or are there other factors in the surgical procedure associated with the improvement?

Lumbar spine fusion is standard of care today. There are no indications that a certain type of fusion technique is superior to another according to patient reported outcome.

Any result of a high quality double blind randomized clinical trial including a non-fusion surgical arm will be of importance because it can test the possible causal effect of the fusion procedure. An outcome showing no difference between the groups will be of importance to the general low back pain patient population and have socioeconomic implications. Orthopedic surgeons will have a better reason or scientific explanation not to expose patients for a procedure that incurs significant risk and may have no benefit. In addition, financial resources can be saved if the procedure is shown to be ineffective. A result showing fusion for low back pain to reduce pain in greater extent than non-fusion surgery will support the current practice of fusion.

This is a multicenter patient and outcome assessor blinded parallel assignment two arm randomized controlled trial. The trial will use a pragmatic approach and use existing equipment and document handling at the different participating units and the existing quality registry for spine surgery, Swespine, for outcome assessments.

Randomization between the study groups will take place intraoperatively, minimizing selection bias. The operative personnel cannot be blinded but the patients and outcome assessors will be blinded for treatment allocation until the 6-month follow-up has been performed. The statistician conducting the analyses will be blinded for treatment allocation.

Intraoperatively, radiographic verification scans or computed tomography will be used for identification of operated level and any implant positioning during surgery. After surgery, no radiographs will be taken until the 6-month follow-up has been accomplished.

Patient reported outcomes will be assessed without influence of the care giver.

Patients will be assessed before surgery, during surgery and inpatient stay, at 3 months, 6 months, at 1, 2 and 5 years.

Subject enrollment and randomization Individuals submitted for evaluation of lumbar spine fusion will be informed of the study at the time of an outpatient visit at the participating clinic.

Subject eligibility will be established before treatment randomization. Subjects will be randomized intraoperatively and strictly sequentially, as subjects are eligible for randomization.

Discontinuation and withdrawal of subjects Subjects are free to discontinue their participation in the study at any time. This will not affect further treatment. Patients will be withdrawn from the study if the patient withdraws consent. Already collected study data for these patients will be kept in the study database, however new data will not be added.

Premature termination of the study The study group may decide to stop the trial or part of the trial at any time. Furthermore, the investigator should promptly inform the Swedish Ethics Review Authority and provide a detailed written explanation.

Re-screening Re-screening of a possible participant for study inclusion is allowed before surgical treatment has been performed.

Preoperative planning Preoperative planning is an important part of all surgeries that will be specifically documented before surgery. Planning of the surgery includes level of surgery and implant placement.

Surgical treatment See Arms and interventions.

The postoperative regime is according to standardized protocols for lumbar fusion surgery. Postoperative radiographs are not performed until unblinding has been performed. Patients will be informed that no radiographs are to be taken until after the 6-month follow-up.

Blinding Surgeon and surgical team are not blinded. Allocation of randomisation is given through an online platform after anaesthesia and surgical field preparation. A generic surgical note will be used in the hospital system and in Swespine so that the treatment allocation will not be revealed. The surgical note includes dimensions of planned implants in case of non-fusion surgery and actual implants in case of fusion surgery.

Patients and personnel at the ward, and personnel at the outpatient clinic are blinded for treatment until after the 6-month follow-up. Outcome assessors are blinded for treatment allocation. Treatment allocation is revealed after the 6-month follow-up.

Blinding of treatment type will be assessed by questionnaire the day after surgery and before being unblinded. Patients will be asked to guess their treatment allocation and are allowed to express uncertainty.

Randomization The subjects are randomized in a 1:1 ratio, without stratification for participating unit. Randomization will be performed through a web-based platform using an allocation sequence hidden from the healthcare personnel and provided independent from the study. Block randomization will be used. Block sizes will be unknown for the researchers. The inclusion will end when the minimum sample size has been reached in both study arms.

In the unlikely event of platform issues rendering randomization impossible, randomization will take place with use of an in advance prepared closed envelope containing treatment allocation (prepared by an independent statistician) available at the central site research coordinator, who can be reached by telephone.

Concomitant medication Patients will receive their ordinary medications and the standard pre- and postoperative treatment.

Sample size calculation

Sample size for primary objective The study is designed as a superiority trial. We hypothesize that the active intervention (fusion surgery) is more effective than the control intervention (non-fusion surgery). Two-sided tests will be used.

The sample size is based on the primary outcome, ODI (0 no disability, 100 bed-bound). The minimal important change in ODI is 10. A difference in change from baseline to follow-up of 10 points between the groups, a standard deviation of 15 points, a significance level of 5%, a power of 80% and consideration for dropout of 20%, and rounding up, gives 45 patients in each of the groups.

For the secondary outcome variable NRS back pain (0 no pain, 10 worst imaginable pain), the minimal clinical important difference is 2. Given a difference in change from baseline to follow-up between the fusion group (intervention) and the non-fusion group (control) of 2 on the NRS, a standard deviation of 3, a significance level of 5%, a power of 80% and consideration for dropout of 20%, and rounding up, gives 45 patients in each of the groups.

Statistical analysis In order to test differences in continuous variables measured on at least interval scale between the groups analysis of covariance will be used, adjusting for the baseline data. The non-parametric Mann-Whitney test will be used to test ordinal data. In order to evaluate hypotheses of variables in contingency tables, the Chi-square test will be used or, in the case of small expected frequencies, Fisher's Exact Test. Regression analysis will be used to evaluate the dependency between variables and the Pearson correlation coefficient will be used in order to test independence between variables. A linear mixed model with a subject-specific random intercept and with the outcome variable at baseline, time, intervention groups and the interaction between time and intervention groups as fixed effects will be used for longitudinal data. In addition to that descriptive statistics will be used to characterize the data; 95% confidence intervals, means, medians and interquartile ranges, and number (proportions) for categorical data.

Data from the different groups will be compared based on the 'intention to treat' principle. An intention to treat (ITT) analysis means that all patients, regardless of treatment change, loss to follow-up or drop-out, remain in the analysis of the group to which they were randomized.

Statistical expertise, blinded to treatment allocation, will perform the statistical analysis.

Interim analysis The study plan intends complete inclusion before analysis; therefore, an interim analysis is not planned.

Safety committee An independent safety committee will monitor the study with the primary aim to detect adverse events.

Data storage and management All data is recorded, handled, and stored in a way that allows its accurate reporting, interpretation and verification. All source data including informed consents, the completed study database, original protocol with amendments and the final report will be stored at each of the participating units/clinics. Copies will be kept at the Uppsala University for a minimum period of 10 years after termination of the trial.

At the conclusion of the study, the occurrence of any protocol deviations will be determined. After these actions have been completed and the database has been declared to be complete and accurate, it will be locked and available for data analysis.

The research coordinator will have regular contacts with the research team and participating centers to confirm that the investigational team is adhering to the protocol. The investigators should ensure that all persons assisting with the trial are adequately informed and trained about the protocol and their trial related duties.

Audits and inspections Authorized representatives of the Uppsala University and Uppsala University Hospital may perform audits. The investigators must ensure that all study documents are accessible for auditing and inspection. The purpose of an audit or inspection is to systematically and independently examine all study-related activities and documents, to determine whether these activities were conducted, and data were recorded, analyzed and accurately reported according to the protocol, and any applicable regulatory requirements.

Ethics The study will be performed in accordance with the protocol, with the latest version of the Declaration of Helsinki, and applicable regulatory requirements. The Swedish Ethical Review Authority has reviewed and approved the study protocol. The Principal Investigator is responsible for informing the Ethical Review Authority of any amendment to the protocol, in accordance with local requirements.

Informed consent The investigators will ensure that the subject is given written information about the nature, purpose and possible risks and benefits of the study. Subjects must also be notified that they are free to discontinue from the study at any time. The subject should be given the opportunity to ask questions and allowed time to consider the information provided.

The subject's signed and dated informed consent must be obtained before conducting any procedure specifically for the study.

The original, signed Informed Consent Form (ICF) must be stored at the study site.

Subject data protection The Informed Consent Form incorporate wording that complies with relevant data protection and privacy legislation, and about the collection of data for the purposes of the study.

The Informed Consent Form explains that study data will be collected from questionnaires, hospital files, images and health databases/registries and will be stored in a computer database, maintaining confidentiality in accordance with national data legislation.

Insurances The study subjects are covered by the Swedish Patient Injury Act.

Protocol modifications Modifications to the signed protocol are only possible through approved protocol amendments. Details of non-substantial amendments are to be clearly noted in the amended protocol, and in the clinicaltrials database.

In case of a substantial protocol amendment (e.g. change of; main purpose of the trial, primary/secondary variable, measurement of primary variable), the Swedish Ethical Review Authority must be informed and should be asked for its opinion/approval prior to implementation of amended protocol, as to whether a full re-evaluation of the ethical aspects of the study is necessary by the committee. This should be fully documented.

The Investigator must not implement any deviation from, or change to the protocol, without discussion with, and agreement by the study group and prior review and documented approval/favorable opinion of the amendment from the Swedish Ethical Review Authority, except where it is necessary to eliminate an immediate hazard to study subjects, or where the change(s) involves only logistical or administrative aspects of the study (e.g. change of telephone numbers).

Reports and publications After completion of the study, the results will be analyzed, and a clinical study report will be prepared. Upon study completion and finalization of the study report the results of this trial will be submitted for publication and posted in a publicly accessible database of clinical trial results.

Definition of "End of study" End of study is defined as the last follow-up of the last subject.