Surgical Versus Nonoperative Treatment: How do we Choose the Right Approach to Lumbar Disk Herniation?

KEYWORDS:

• disk herniation
• diskectomy
• nonoperative treatment
• outcomes

Choosing the appropriate treatment for radiculopathy associated with lumbar intervertebral disk herniation (IDH) is a fascinating clinical problem given that the symptoms can be quite debilitating, it can be treated safely and effectively with surgery, and it can also improve spontaneously in a relatively short period of time. The patient and provider face an interesting dilemma in that they need to choose between a relatively safe and effective surgical treatment (diskectomy) that carries with it a low but real risk of complications and nonoperative treatment that carries minimal risk but may not be as effective, or at least not effective as quickly. In order to make an informed decision, the patient and provider need to consider the likely outcomes of surgical and nonoperative treatment as well as the patient’s preferences and willingness to take on risk. If outcomes could be accurately predicted on the individual patient level, the decision would be easier. Unfortunately, the spine community still does not have a crystal ball, and decisions for individuals need to be made based on the known outcomes for ‘average’ patients in large clinical trials. We are making progress on efforts to use individual patient characteristics to guide treatment decisions, but the use of clinical prediction models is still at the research stage.

In considering the effectiveness of treatment for any condition, the natural history of the condition must be known. Fortunately, there have been many studies evaluating the natural history of IDH, with one classic study demonstrating spontaneous improvement of symptoms in 90% of IDH patients within 12 weeks of symptom onset [1]. In this series of 64 patients, only six went on to surgery, and 90% had ‘good or excellent’ outcomes with undefined nonoperative care. More modern studies have shown somewhat lower success rates with nonoperative care, but these studies generally only included patients with symptoms that had already lasted at least 6 weeks [2–4]. There is general agreement that sciatica due to IDH is frequently a self-limited condition, and that the majority of patients who experience radiculopathy will improve without any intervention within weeks to months. However, there is a substantial minority who fail to improve in the short term and need to consider other nonoperative or surgical options.

There are many nonoperative treatments for lumbar radiculopathy caused by IDH, including medication, physical therapy, injections, and alternative medicine. Unfortunately, there are no high-quality data demonstrating that any of these provide more than partial, temporary relief. In fact, it is unclear if any of these treatments significantly alter the natural history of the disease. The evidence is mixed on NSAIDs, with one meta-analysis showing a trend toward short-term benefit versus placebo [5]. There is even less data on oral corticosteroids and neuromodulators such as gabapentin. The American Pain Society Clinical Practice Guideline concludes that there is ‘fair’ evidence that epidural steroid injections provide moderate, short-term relief for sciatica [6]. The Spine Patient Outcomes Research Trial (SPORT) found that IDH patients who underwent epidural steroid injections were more likely to avoid surgery but had similar outcomes compared with those who did not have injections [7]. Physical therapy is another commonly recommended treatment for lumbar radiculopathy due to IDH, yet a systematic review concluded there was no benefit compared with no treatment [8]. Alternative treatments such as acupuncture have not been rigorously evaluated for IDH. Despite most clinical guidelines suggesting attempts at nonoperative treatment for patients with lumbar radiculopathy, there is no strong evidence indicating that any of the frequently used nonoperative therapies are effective. This lack of evidence is reflected by two recent trials, SPORT and the randomized controlled trial by Peul et al. which could not recommend a specific nonoperative treatment regimen and instead encouraged patients being treated nonoperatively to employ ‘usual care’ [2,3].

The outcomes of surgery for lumbar IDH have been better studied than the nonoperative modalities, and essentially all investigations have demonstrated that surgery leads to faster and more complete improvement of radiculopathy compared with nonoperative treatment. The SPORT, the Maine lumbar spine study, and the classic randomized controlled trial by Weber all showed that diskectomy patients had more improvement of pain and function than those treated nonoperatively at 1 year, and some of these advantages persisted at 8 and 10 years after surgery [9–11]. Using a different study design, Peul et al. compared outcomes between lumbar IDH patients with symptoms lasting for 6–12 weeks who either underwent early surgery or prolonged nonoperative treatment with surgery if they deemed it necessary [2]. While the early-surgery patients had faster improvement of their leg pain, only 44% of the prolonged nonoperative care patients eventually chose to undergo surgery, and overall outcomes were similar for the two groups at 1 year. Surgery also carries the risk of complications, though the literature indicates that microdiskectomy is a relatively safe procedure. Dural tear, the most common complication, occurred in 3% of cases in SPORT but had no effect on long-term outcomes [12]. Two percent of patients had wound infections, and one patient out of 798 had a nerve injury. No patients died within 30 days of surgery or had a death attributed to the surgery. At 8 years of follow-up, 15% of patients had undergone re-operation, with recurrent herniation at the same level being the most common reason for revision surgery (9% rate overall) [9]. The existing literature makes it clear that surgery leads to a faster and greater degree of improvement compared with nonoperative care in patients with radicular symptoms that have persisted for at least 6 weeks, but a substantial proportion of these patients can also do well with nonoperative treatment or have spontaneous improvement.

Given what we know about the condition, IDH patients and their providers are in a conundrum given that surgery will likely lead to a rapid improvement in pain and function with a low risk of complications, yet pursuing nonoperative treatment may also lead to improvement with almost no risk of complications. In an effort to better define which subgroups of patients may do markedly better with surgery or do just as well with nonoperative care, Pearson et al. did an exhaustive subgroup analysis of the SPORT IDH data to determine which characteristics were associated with the treatment effect of surgery [13]. Somewhat surprisingly, essentially all examined subgroups of patients improved significantly more with surgery than with nonoperative treatment, suggesting that a patient who met the indications for surgery (i.e., symptoms for at least 6 weeks, neurological findings on physical exam and advanced imaging showing disk herniation consistent with their symptoms) would likely improve more with surgery than with nonoperative treatment. Patients with worsening symptoms at enrollment, those without concomitant joint problems, and those who were married had the greatest treatment effects of surgery, though even single patients with stable symptoms and joint problems still improved significantly more with surgery. Putting all of this together, determining how to treat lumbar IDH in patients who meet the indications for surgery is ultimately a preference-sensitive decision [14]. Patients should go through a shared decision-making process in which they are informed that surgery will likely lead to a faster and greater degree of improvement than nonoperative treatment, but that they may also improve substantially with nonoperative treatment without the risks that go along with surgery. Patients who are able to tolerate or manage their symptoms have nothing to lose by giving prolonged nonoperative treatment a try, at least out to 6–9 months, while those who are unable to tolerate or manage their symptoms can expect a high probability of rapid symptom relief with a relatively low-risk operation. Either decision, based on the preferences of a well-informed and engaged patient, is likely to be the right one.

Financial & competing interests disclosure

AM Pearson is an associate web editor for Spine and has a position on the Spine Associate Editorial Board. JD Lurie is a consultant for the Informed Medical Decisions Foundation, FzioMed and NewVert, and owns stock options for NewVert. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Additional information

Funding

AM Pearson is an associate web editor for Spine and has a position on the Spine Associate Editorial Board. JD Lurie is a consultant for the Informed Medical Decisions Foundation, FzioMed and NewVert, and owns stock options for NewVert. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.