CT scans, although also not employed as part of the routine evaluation of the patient with isolated lumbar complaints, provide useful information in the preoperative assessment of the patient with recalcitrant DDD. The excellent osseous imaging capacity of CT is most useful in illustrating the presence of facet joint degeneration (Fig. 20.4) and fractures of the pars interarticularis (Fig. 20.5), which can compromise the use of anterior-based motion-sparing procedures. Furthermore, CT scanning can provide an indirect assessment of bone density as well as provide detailed measurements of the bony anatomy for implant sizing (vertebral body and pedicular anatomy). Quantitative CT may serve as an alternative to bone scintigraphy in the assessment of bone density, although there is a larger dose of radiation imparted to the patient.

MRI provides great imaging detail on both normal anatomic and abnormal pathologic entities that can affect decision-making (Fig. 20.6). The identification of diskal pathology (herniation, annular tears, dehydration), degenerative changes (vertebral endplate edema, facet joint degenerative pathology), or other compressive pathologies (ligamentum flavum infolding, facet hypertrophy) has all been greatly enhanced by the routine use of MRI. As detailed by Boden et al. [3], however, abnormal findings on MRI (such as disk degeneration, annular fissures, small protrusions, and facet arthritis) can be commonly identified in asymptomatic patients. This phenomenon requires that a measured evaluation of MR imaging occurs in light of the patient’s history, exam, age, and suspected cause of symptoms. The primary advantages of MRI in the evaluation of the lumbar patient lie in the evaluation of compressive pathology, the staging of disk degeneration, and the evaluation of facet joint degenerative changes. No validated staging system exists, however, for degenerative disk disease using MR imaging.

Bone scanning using either dual-energy X-ray absorptiometry (DEXA), dual photon absorptiometry (DPA), or quantitative CT should be performed in all patients suspected of osteopenia or osteoporosis, particularly postmenopausal women, men over age 50, and chronic smokers. DEXA scan has become the standard modality used for this purpose due to its relative low cost and lower dose of imparted radiation versus the alternatives.

As discussed in an earlier chapter, the usefulness of diskography is primarily as an adjunct to the clinical and diagnostic imaging evaluation. The current limited recommendations for diskography include the definition of the symptomatic level in the patient with multilevel pathology and in distinguishing between spinal and extra-spinal causes of pain. Although numerous studies have questioned the positive predictive value of diskography [12–14], it remains the sole modality to permit identification of an intradiskal pain generator, and it can allow for the determination of the symptomatic level in a patient with multilevel pathology. It is important to note that the adequate pressurization of the disk requires annular competency, and an incompetent annulus may result in false-negative results using diskography. More recent data has demonstrated that even using modern administration techniques, diskography has deleterious consequences to the nuclear cells, the annulus, and the disk itself [15]. And although the mechanical insult of the needle introduction and pressurization alone have been shown to contribute to the insult, both radiocontrast and local anesthetics have been shown to have markedly adverse consequences for the diskal environment [16, 17], While the highest yield for diskography may be in the symptomatic back pain patient who has undergone all methods of conservative treatment without relief and has evidence of isolated degenerative disk disease in the absence of other pathology, the risks of the procedure, its limited diagnostic yield, and its known side effects have largely relegated it to the historical realm.

Lumbar facet blocks (both with and without steroids), medial branch blocks, and radiofrequency ablation are techniques employed for both the diagnosis and treatment of disorders attributable to the zygapophyseal joints resulting in low back pain. Furthermore, these techniques have been expanded to attempt pain relief of low back pain attributable to the sacroiliac joints as well. Historical data have shown that facet blocks (both intra-articular and medial branch) provide adequate accuracy, reproducibility and safety in the diagnosis, and management of facet-mediated pain [18]. Interestingly, the addition of steroids may not add to the efficacy of this modality [19]. Based on a Cochrane-style review of the literature, diagnostic anesthetic blocks were given a level-1 recommendation, with level-II values associated with therapeutic medial branch blocks and radiofrequency ablation [20].

A successful diagnostic or therapeutic injection does not appear to accurately predict a successful surgical outcome, however [21]. In fact, one historical study appears to indicate that pain relief with temporary external fixation seems to offer better a better prediction of surgical success when compared with other modalities [22]. This is historical data, surely, but the conclusion is that despite their efficacy in diagnosis and treatment, pain management modalities cannot currently be said to predict surgical success.

SCS has emerged as a valuable technique for the management of chronic back and leg pain and unresolved symptoms after thoracolumbar spine surgery, the termed “failed back surgery syndrome” [23, 24]. Patients suffering from complex regional pain syndrome (CRPS) also may benefit from this technology [25]. The data to support its use for the patient with predominantly low back pain as a result of DDD, however, remains lacking. Leg pain as the predominant symptom has been the main if not the only predictor of positive outcomes following SCS to this point [26].

Clinical data that can be considered concerning surgery for DDD consists of randomized controlled trials (RCT) comparing arthrodesis with conservative care and RCTs comparing various procedures with arthrodesis as the control. Although there are many case series of emerging techniques and devices for diskogenic pain, there is not an existing evidence basis to support posterior interspinous devices or nucleus replacement for the treatment of diskogenic back pain.

The landmark paper published by Fritzell et al. in 2001 was the first to compare surgical treatment with nonoperative treatment of chronic low back pain in a well-matched randomized trial. This study was notable for the improved low back pain outcomes seen in the surgical group (33 %) versus the control (9 %) at 2-year follow-up. Criticisms of this trial include the indeterminate nature of the causative entity of the low back pain seen in the control patients. Diagnosis was made on the basis of history, physical examination, and radiographs alone, and MRI and diskography were not employed. Furthermore, the patients randomized to the surgical group were treated with three different fusion options: posterolateral uninstrumented, posterolateral instrumented with pedicle screws, and circumferential (360°) fusion. Also, for the patients treated conservatively, there was considerable variability in the nonoperative care. No formal physiotherapy protocol was defined, and the use of injections and other pain modalities (acupuncture) was not standardized [27]. This paper was followed by an RCT from Norway by Brox et al. comparing posterolateral arthrodesis with cognitive intervention and exercise for chronic low back pain [28]. The authors used a more defined nonoperative treatment evaluation employing patient education, range of motion exercises, and intensive physiotherapy. At 1-year follow-up, there were no noted differences in improvement in low back pain or in disability as measured using the Oswestry Disability Index (ODI). Fairbank and colleagues performed a similar RCT comparing surgery with conservative treatment and found reductions in disability in both groups 2 years after treatment, but minimal differences using the ODI and a shuttle walking test [29]. A systematic review of these trials was performed in 2007 by Mirza et al. The authors noted that surgery, when compared with cognitive-behavioral therapy and a highly structured rehabilitation protocol for “diskogenic” back pain, resulted in only a modest improvement in back-specific disability [30]. In 2013, Mannion et al. reported long-term follow-up of three randomized trials comparing surgery and nonoperative treatment (multidisciplinary cognitive-behavioral and exercise rehabilitation) for chronic low back pain and found no differences in patient-reported outcomes at an average of 11 years [31].

Numerous RCTs have been performed to assess the efficacy of lumbar arthroplasty for the treatment of diskogenic low back pain, in each case using arthrodesis as a control. In general, the clinical results comparing disk arthroplasty with lumbar fusion have failed to demonstrate significant differences between the interventions in terms of functional outcomes, pain, use of medication, or occupational disability [32–34], with the exception of the investigational device exception (IDE) trial examining the ProDisc-L, which demonstrated marginal superiority of arthroplasty results on the VAS, ODI, and return to employment. However, although the treatment alternatives have been demonstrated to be comparable in this patient population, approximately half of the clinical trial patients, despite the stringent eligibility criteria, appear to not meet the criteria for clinical success. Furthermore, when following the criteria used for the FDA studies, it becomes apparent that the contraindications to this procedure are common when considering the surgical population in a spine practice, as was demonstrated in a review by Huang et al. [35].

When considering outcomes following any surgical intervention (using health-related quality of life, or HRQOL, instruments), it is important to gauge the magnitude and clinical significance of the score difference and not only the presence of a statistically measurable difference. Using functional scoring scales such as the ODI, for example, the degree of change in the ODI (or ODI delta) may be a more accurate predictor of the functional impact of the intervention. The “minimal clinically important difference” (MCID) has also been described in order to clarify the minimum amount of necessary change in an outcome score to suggest a tangible clinical benefit to the patient [36]. For the ODI, the MCID has been defined variously as ranging from 5.2 to 16.3 historically [37, 38]. Copay, Glassman, et al. performed a prospective study of patients gathered from the Lumbar Spine Study Group and determined the MCID for patients as assessed by the ODI to be 12.8 points for patients undergoing lumbar spinal surgery [39].

It is in this context, then, that the more recent trials for lumbar surgery, including trials examining surgery for DDD, should be viewed. Weinstein et al., in the degenerative spondylolisthesis arm of the SPORT (Spine Patient Outcomes Research Trial) trial, reported a mean improvement in the ODI of −24.2 in the surgery group at 2 years, versus −7.5 in the nonoperative group, for a treatment effect of −16.7 (in light of a MCID of 12.8). In the larger trials of lumbar arthroplasty, the ODI has served as a critical tool in the functional outcome assessment and as a statistical piece of the overall assessment of clinical success. The degree of the clinical effect, however, remains an issue of discussion. In the pivotal trial of the Charite prosthesis, Blumenthal et al. [34] noted changes in the arthroplasty group from a mean preoperative ODI score of 50.6–37.7 (−12.9), 29.9 (−20.7), 27.5 (−23.1), 26(−24.6), and 26.3(−24.3) at 6 weeks and 3, 6, 9, 12, and 24 months, respectively. In the pivotal trial of the ProDisc-L implant, Zigler et al. also used the ODI as part of functional outcome assessment the FDA IDE study [40]. The investigational group had a mean initial ODI of 63.4 (reflecting use of an alternative ODI scale, which can result in a higher disability core), 25.3 % higher than in the Charite trial. At intervals of 6 weeks, 3 months, and 6 months, the authors reported statistically significant differences from control in the ODI and a trend toward significance at 24 months. Versus the baseline values, although precise data were not published, it appears that the range of mean ODI following arthroplasty was between 34.5 and 47, with a value (34.5) provided at 24 months. From baseline ODI, this value reflects a −28.9 point change (46.1 %). The FDA criteria for outcomes were specified to include a decrease of 15 % on the ODI for “ODI success.” Although there was a significant decrease following arthroplasty (and fusion, which resulted in a mean ODI of 39.8 at 24 months, for a change of −22.9 points, or 36 %), the patients remained with some degree of disability as reflected by an ODI of 34.5 at 24 months in the investigational group. Furthermore, the differences in the initial disability between the two trials demonstrate that the administration of functional scoring scales has some inherent variability depending on the content and application of the instrument. These data would suggest that the procedure (and the control), although they meet the statistical criteria, the specified FDA criteria, and MCID criteria for success of the intervention (and the control), results in improvement with a remnant of continued functional disability (mean ODI of 26.3 in the Charite trial and 34.5 in the ProDisc trial).