Evaluation and Treatment of Tandem Spinal Stenosis

Joseph F. Baker, MCh, FRCSI


J Am Acad Orthop Surg. 2020;28(6):229-239. 

In This Article


Treatment modalities, as for most spinal conditions, can be considered nonsurgical and surgical. For stenosis of the lumbar spine, a nonsurgical approach, at least initially, is justifiable. A proportion of patients will find they become increasingly tolerant of their symptoms with time, although their walking distance may not improve. For the frail patient, nonsurgical management is also the low-risk approach. A physical therapy program may aid in maintaining cardiovascular health, muscle strength, and bone mineral density, all of which are beneficial. However, it must be acknowledged that in the setting of LSS, there is a lack of good quality evidence comparing surgical and nonsurgical treatment, and time spent counseling the patient about the pros and cons of each treatment pathway, taking into account the particulars of the individual's presentation, is needed.[23]

For stenosis of the cervical spine, judicious decision making is required. The natural history of CSM is one of stepwise progression—surgical intervention does not necessarily lead to reversal of the deficit, and so, any delay in surgery exposes the patient to potentially further loss of function. Systematic reviews have failed to find quality evidence to support nonsurgical treatment in CSM—of those managed nonoperatively, a notable number (more than 50% on one review) eventually undergo surgical intervention.[24] In patients initially managed nonoperatively, a higher rate of hospitalization for spinal cord injury is also seen.[25]

There is, therefore, little evidence on which to justify extended nonsurgical management of a patient with cervical myelopathy who is otherwise fit and well. On the other hand, the very frail or comorbid patient may decide, having being appropriately counseled that surgical intervention is highly risky for continued nonsurgical management, no matter what the consequence is.

In those who fail to adequately improve with nonsurgical measures or in those who have a clear indication to proceed to surgery, such as (progressive) neurologic deficit, surgical options should be considered and tailored to the individual. In TSS, a decision on which level to operate on first is perhaps key, and second, whether to perform decompression of all levels of stenosis in one setting or in a staged manner.

It is intuitive to operate on the cord-level stenosis first—the risk associated with ongoing compression of the cord is greater than that associated with compression below the conus. However, in selected cases, where the integrity of the spinal cord appears intact, without overt compression or signal change on either T1- or T2-weighted sequences, one may opt to address the lumbar spinal pathology first if it is clear that the claudication or lumbar radiculopathic symptoms predominate the clinical picture. In this setting, one can augment preoperative assessment by assessing for potential provocation of cervical myelopathic symptoms on clinical examination and by obtaining flexion-extension radiographs if the cervical spine excludes any instability. A suggested treatment algorithm on this basis is proposed in Figure 3. To minimize potential for further insult to the spinal cord during anesthesia and patient positioning, one can request the patient to demonstrate their comfortable range of motion of the cervical spine before anesthesia—during intubation, and patient positioning care is then taken to not exceed these limits.

Figure 3.

Proposed treatment algorithm in the setting of TSS involving both cervical and lumbar canal stenosis. CSM = cervical spondylotic myelopathy, TSS = tandem spinal stenosis

A number of studies have reported on the outcomes of surgery—these largely consist of retrospective cohort studies. Many have focused on detailing and/or comparing the clinical outcomes for either simultaneous or staged approaches to surgical management. There is no Level 1 evidence available to guide the surgeon in this clinical setting, and this remains an elusive goal given the challenge in diagnosis and lack of uniform clinical and radiographic findings.

In one of the largest cohorts, of 565 patients with LSS, 202 were demonstrated to have TSS with radiographic evidence of CSS and associated cord compression.[26] During an average of 3.2 years of follow-up, 28 patients underwent cervical spinal surgery—11 simultaneous and 17 staged. Those patients who underwent subsequent cervical surgery experienced a mean improvement in both their cervical and lumbar functional measures (C-Japanese Orthopaedic Association score [JOA] and L-JOA). Radiographic coexisting cervical stenosis that was asymptomatic did not influence outcomes for LSS surgery, but a symptomatic cervical lesion negatively affected the neurologic scoring after lumbar surgery. The same author group also reported on 172 patients with symptomatic CSM and TSS involving the lumbar spine.[27] Forty-one underwent a procedure on the lumbar spine during a 3.1-year follow-up period. Although the recovery rate and mean functional outcome scores for the TSS group were worse than those with isolated cervical disease, the additional lumbar surgery still resulted in improvements in both cervical and lumbar functional scores.

Li et al[28] reviewed 222 patients who, over a 15-year period, were prioritized to have either the cervical or the lumbar spine decompressed first. The subsequent revision surgery rate for lumbar stenosis was much lower than the revision surgery rate for cervical stenosis (23% compared with 58%). Patients undergoing cervical decompression experienced improvement in Oswestry Disability Index scores in addition to JOA and Nurick grades. However, no JOA score or Nurick grade improvement was seen after lumbar surgery alone, hence supporting the approach to address CSS first.

The common theme from these reports is of the potential for lumbar symptoms to improve with intervention for the cervical disease alone and to prioritize the cervical lesions first, especially if symptomatic. This offers a commonsense strategy for staged surgery, starting with the cord-level lesion and possibly avoiding a second procedure (Figure 4).[14,29,30] By addressing the cord-level disease first and delaying surgery for lumbar stenosis, there appears to be little downside. The converse, however, is more difficult, and one must be certain that the cord is not compromised if electing to treat the lumbar disease first and the consequences can be catastrophic.

Figure 4.

T2-weighted sagittal (A) and axial (B) images of the cervical spine and sagittal (C) and axial (D) images of the lumbar spine. Axial images of the most stenotic segments. Lateral radiograph after surgical intervention (E). A 75-year-old woman was initially referred by her primary care physician with features of both cervical myelopathy and neurogenic claudication. Given her age and comorbid conditions including insulin-dependent diabetes mellitus, a staged approach was planned. After open-door laminoplasty at C5-7, she experienced marked improvement in balance and fine motor tasks. Symptoms attributable to LSS were notably reduced, and she deferred any further surgical intervention. LSS = lumbar spinal stenosis