Most patients with failed spine surgery have a "peripheral" mechanical
structural lesion as a source of nociception. Because of the iatrogenic
changes associated with surgical intervention, identifying the source of
symptoms is difficult. Occult lesions are expected to be more frequent in
these patients than in those not previously operated on, and may explain
the failure of surgery. A systematic approach is necessary to evaluate the
various factors in both central and peripheral nervous systems that contribute
to the patient's experience of pain. The spinal lesions most often associated
with failed surgery are discussed below.
Instability
As defined by Frymoyer and Selby, segmental instability is loss of the normal
stiffness of a motion segment, such that application of force to that motion
segment results in greater displacement than would occur in a normal structure. 10
Conceptualizing the spine as "motor segments" or "motion
segments"12 emphasizes the functional interrelationships between its components. A spinal motion segment then is comprised of the disc, facet joints, spinous processes, central and lateral canals, and ligaments. Each structure in the motion segment is interrelated in function to other structures.
For example, a degenerative narrowed disc with the resulting approximation
of the vertebral bodies on either side results in approximation of the facet
joints and subsequent increased loading and facet overriding.
Similarly, a fracture of a facet joint with resultant loss of its stabilizing effect will allow greater stresses on the intervertebral disc. Ninety per cent of patients requiring surgery have had significant episodes of symptoms prior to those that necessitated surgery. One would suspect that frequently an unstable motion segment is the cause of, or at least is associated with, disc deterioration and herniation. A segment that had deteriorated to the point of necessitating surgery could not withstand the stresses applied to it. In many cases it may be incompetent even in activities of daily living. Operating on this diseased segment without fusion weakens it further, and although neural compression may be alleviated, segmental instability is expected to be increased by removal of some of the bony and ligamentous stabilizers. This increased instability may overcome the gains of the operation and cause surgery to fail.
A typical scenario is that of the patient with radicular pain prior to laminectomy and discectomy who has relief of radicular symptoms after surgery, but develops disabling low back pain, a symptom indicative of instability. In the presence of scar tissue, with relative fixation of neural elements, we may see less tolerance to instability than preoperatively. Incipient instability can certainly be made clinically manifest by surgical exposure and further destabilization.
Levels adjacent to fusions are frequently seen to be unstable radiographically and clinically, as a result of both the multisegmental nature of degenerative disease and the increased stresses applied to the last mobile segment.
Stenosis
Stenosis may cause surgery to fail both early and late in the postoperative period. Early in the postoperative period, stenosis in the lateral canal which is unrecognized may cause persistent pain. The stenosis is usually lateral where it is not visualized from the midline and the canal size must be gauged blindly, using probes. A multiplanar CT scan is most sensitive for picking up foraminal and lateral canal narrowing, which can be missed on myelograms or CT scans without lateral reconstructions.4 When discectomy
is performed, subsequent postoperative settling occurs, resulting in further
overriding of facet joints and further subarticular narrowing of the lateral
canals. In the late postoperative period central and lateral stenosis can
occur from progressive degenerative changes, as would be expected after
the surgical violation of an already diseased segment. When fusion is performed
posteriorly, gradual overgrowth may result in late postoperative central
stenosis at the surgical site. Burton et al. found the primary factor leading
to the failed back surgery syndrome to be lateral spinal stenosis.4 This is in contrast to Frymoyer's findings, which indicated recurrent disc herniation to be the most common cause.9 Burton, et al. found lateral spinal stenosis to be the cause of failed surgery in 57% of cases, central stenosis in 7-14%, and recurrent or persistent disc herniation in 12-16%.4
Recurrent Disc Herniation
Most discectomies entail removal of only a portion of the disc material.
The remaining original disc material may exit out the annulotomy site and
result in a recurrent disc herniation. This is one of the more common correctable lesions in failed spine surgery in our experience. Frymoyer found that recurrent disc herniations were the most frequent cause of failed disc surgery in a series of 268 failed spine surgery cases.9 In their series 43% of the patients had recurrent disc herniations at the operated level and 22% had them at a different level; thus, 65% of all surgical failures appeared to be due to recurrent disc herniations. Since postoperative fibrosis tethers the aura and nerve roots, a small amount of disc material can cause disabling symptoms. This is especially true when there is a connection with the disc space and the protruding disc fragment, so that loading causes some motion of the herniation against the immobile neural elements.
Missed Lesions
Pathologic processes that were missed at the time of initial surgery are
numerous. Crock described postoperative persistent symptoms without temporary
relief as one of two patterns in failed spine surgery.6 This pattern is
the hallmark of missed pathologic processes. MacNab cited migration of disc
fragments into the intervertebral foremen, nerve root kinking by the pedicle,
articular process impediment on the nerve root, spinal stenosis, and extraforaminal lateral disc herniations as causes of "negative disc explorations."14 In addition, midline disc material may dissect posterior to the longitudinal ligament and be very difficult to detect at the time of surgery. Occasionally, anatomic confusion may lead to surgery on an inappropriate spinal segment; more commonly, occult symptoms from one segment are overlooked while disease at another segment that may not have been so symptomatic is surgically approached. These missed structural lesions are less likely when an extensive preoperative diagnostic work-up is performed, including multiplanar CT scanning and EMG. One disadvantage to discectomy approaches involving minimal exposure is that some of these unusual lesions will not be detected at the time of surgery. Very rarely, thoracolumbar intraspinal tumors may simulate lumbosacral herniations,
especially deceptive in the presence of lumbar discogenic disease. Occasionally, an internally disrupted disc which is most symptomatic is overlooked because a different level of herniation is grossly visible on CT scan and myelogram. We have seen this in several cases involving trauma in which lumbosacral spine pain was referred from higher lumbar discs.
Wiltse et al. have described a syndrome of far lateral stenosis seen in
young adults with isthmic spondylolisthesis and significant slippage and
in elderly persons with degenerative scoliosis.[16] This usually occurs
at the L5-S1 level as a result of impingement of the sacral alar and L5
transverse processes upon the L5 nerve root. When the problem is recognized,
it is difficult to decompress without destabilizing the segment. Wiltse
has referred to this as the "far-out syndrome."
Intraneural Fibrosis — Epidural Fibrosis
Epidural fibrosis is unavoidable to some extent and usually is not a source
of disabling pain. The factors that seem to make it symptomatic in some
patients are obscure. In many cases, it seems to be the source of a peripheral
stimulus which is greatly amplified because of a deranged pain perception
mechanism. Since this lesion is always present to some extent, it remains
somewhat of a "wastebasket" diagnosis when no other lesion can
be identified.
Similarly, intraneural fibrosis is a diagnosis that is usually made when
the EMG is abnormal and there is no lesion to account for pain other than
fibrosis about the nerve root on CT scan. Postoperative positive EMG findings
and even neurologic abnormalities are frequently present without pain. The
reason some patients with intraneural fibrosis develop disabling symptoms
and others do not is again presumably that the patients' pain perception
mechanisms differ, as do specific anatomic locations and extents of nerve
damage.13 We have seen several cases of reflex sympathetic dystrophy associated with intraneural fibrosis of exiting roots.
Arachnoiditis
Postoperative and post-Pantopaque-myelography arachnoiditis is a devastating
problem that responds relatively poorly to treatment. 1,3 It is characterized by intrathecal fibrosis, causing rootlets to adhere to each other and to the inner walls of the thecal sac. Again, some patients with anatomic lesions are not symptomatic and others are. Again, since the results from treatment are less than optimal, other explanations for pain should be ruled out before giving arachnoiditis as the functioning diagnosis. Back and/or lower extremity pain with or without necrologic abnormalities may be manifest. The diagnosis is made by myelography and sometimes is evident on CT scan.
Reflex Sympathetic Dystrophy
Reflex sympathetic dystrophy is a rare occurrence in the postoperative patent.1,2
It is associated with nerve root irritation or damage in all of the cases
that we have seen. The reflex sympathetic dystrophy seems to be a secondary
response to the somatic pain of a root lesion, causing great escalation
in its intensity. Temperature changes, sweating, and burning of the lower
extremities suggest this diagnosis. The diagnosis can be confirmed by monitoring skin temperature before and after a sympathetic block. The patient should have relief of sympathetic symptoms at the same time skin temperature elevation indicates chemical sympathectomy.
Soft-tissue Dysfunction
A poorly defined, but apparently common, cause of persistent postoperative
disability and symptoms is what we refer to as soft-tissue dysfunction.
Patients who have splinted their lumbar motion prior to surgery continue
to do so after surgical healing has occurred. They develop pain in the buttock
and lumbosacral region, with associated areas of tenderness. The pain is
worsened by holding static positions such as sitting, standing, or lying
for a prolonged period. The range of motion is limited by back tightness
and discomfort. Symptoms are relieved by changing positions. Most patients
respond to a gradual stretching program. Presumably, scar tissue and shortened
spinal ligaments are being stretched when static positions allow continuous
unidirectal forces to stretch these tightened structures. The associated
discomfort alarms the patient and causes him to splint and immobilize his
pain and continue the pathologic process. Other symptom complexes presumably
resulting from soft-tissue abnormalities are often placed under this category,
but are difficult to classify and treat. Many of these appear to be associated
with a great degree of central nervous system augmentation.
Facet Syndrome
As mentioned above, surgery may increase segmental instability at the level
adjacent to a fusion or at the surgical level. In some cases, this is manifest
most by facet arthropathy. Partially resecting facet joints and posterior
annulus fibrosis probably results in increasing instability, stresses and
loading to the remaining facet joints. Disc settling after discectomy or
chemonucleolysis causes rapid facet subluxation and articular approximation
which can result in arthropathy. Diagnosis is suggested by degenerative
changes on CT scan and x-ray. It is confirmed by relief with facet injections.5
8 In our experience, most of these patients seem to eventually develop more
generalized symptomatic involvement of the entire motion segment.15
Internal Disc Disruption
Internal disc disruption has been most thoroughly described by Crock.7 It is typically characterized by severe pain associated with any loading of the disc in a young adult with a history of trauma. Myelograms and CT scans are normal. In these patients the annulus appears to be exquisitely painful, although there is no herniation. Diagnosis is confirmed by discograms and disc saline acceptance tests. These are typically exquisitely painful and usually reveal circumferential and radial tears radiographically.7 It is important to keep this entity in mind, because the usual diagnostic tests will appear normal. One may unwittingly operate on a disc at the wrong level if other disease is present unless the entity is expected and discograms are obtained. Crock feels that autoimmunity may play an important role in this disease process.7 In our experience and that of others, lymphadenopathy has been a common finding at the time of anterior interbody fusion for this entity.15
Some patients will improve with trunk strengthening, lumbosacral orthosis,
and time. Classic cases of total disability are unusual in our experience,
but there seems to be a spectrum of disorders that are very commonly associated
with painful non-herniated discs. We have seen several patients who at the
time of surgery had posterior lateral fusion without discectomy because
a disc was not herniated on direct inspection. The preoperative symptoms
persisted, though diminished, until the discs were removed under posterior
solid fusions. The persistence of symptoms is presumably caused by transfer
of stresses across the posterior fusion mass into the disc, allowing some
loading and tensile forces on the painful annulus fibrosis. Crock considers
anterior lumbar interbody fusion to be the operation of choice for lumbar
internal disc disruptions.
Pseudarthrosis
It is our impression that spine pseudarthrosis in and of itself is not a
source of pain. In the presence of hardware, scar tissue, stenosis, and/or
tethered neural elements, however, the associated instability may definitely
cause pain. Posterior lateral fusion pseudarthroses typically occur through
the foremen and lateral canal, which may allow nerve root irritation proportional to the amount of motion present. When there is a lot of motion in the pseudarthroses, it will be evident on flexion-extension films. Minimal motion is very difficult to detect. Reconstructive CT scan views will visualize the fusion mass and delineate some pseudarthroses. Even at surgical exploration, pseudarthroses can often be missed unless the fusion mass is denuded of overlying soft tissue and stresses are applied under direct visualization.
Metallic Implants
The recent explosion in internal fixation techniques of the spine and their
application to degenerative disease has brought a host of new reasons for
failed back surgery syndrome. Knodt rods and Harrington rods may cause neural
element irritation at the hook sites. Pseudomeningoceles resulting from
direct pressure of hooks gradually eroding the thecal sac are not uncommon.
Fracture may occur at the hook site or when pedicular screws at the level
of the pedicle compress neural elements. Stenotic syndromes may be caused
by bunching of the ligamentum flavum at the site of the upper hooks in Knodt
and Harrington rods. The loss of lumbar lordosis which is produced by distraction rods at the lumbosacral angle causes compensatory hyperlordosis of the first mobile segment above the fusion and greater posterior element stresses on that segment. This may result in facet arthropathy. Since the CT scan imaging is greatly distorted by the presence of hardware, specific anatomic diagnosis is difficult. We rely on epidural and selective nerve root injections to delineate whether the pain is structural and from what location within or outside of the surgical site it arises.
References
1. Benna B: Spinal arachnoiditis. Spine 3:40, 1978. back
2. Bernini S: Reflex sympathetic dystrophy associated with low lumbar disc herniation. Spine 6:180, 1981. back
3. Burton CV: Lumbosacral arachnoiditis. Spine 3:24, 1978. back
4. Burton CV, Kirkaldy-Willis WH, Yong-Hing K, et al: Causes of failure of surgery on the lumbar spine. Clin Orthop 157:191, 1981. back
5. Carrera GF: Lumbar facet joint injection in low back pain and sciatica, preliminary results. Radiology 137:665, 1980. back
6. Crock HV: Observations on the management of failed spine operations. J Bone Joint Surg 58B:193, 1976. back
7. Crock HV: The Practice of Spinal Surgery. New York, Springer-Verlag, 1983. back
8. Fairbank JCT, et al: Apophyseal injection of local anesthetic as a diagnostic aid in primary low back pain syndromes. Spine 6:598-605, 1981. back
9. Frymoyer J: The role of spine fusion. Spine 6:284, 1981. back
10. Frymoyer J. Selby D: Segmental instability, rationale for treatment. Spine 10:280, 1985. back
11. Hall S. et al: Lumbar spinal stenosis. Ann Intern Med 103:271-275, 1985. back
12. Junghans J. Schmorl G: The Human Spine in Health and Disease. Second edition. New York, Grune and Stratton, 1977. back
13. Loeser J: Pain due to nerve injury. Spine 10:232, 1985. back
14. MacNab I: Negative disc exploration. J Bone Joint Surg 53A:891, 1971. back
15. Selby D: Personal communication, March 1986. back
16. Wiltse L, et al: Alar transverse process impingement of the L5 spinal nerve: The far-out syndrome. Spine 9: 31-41, 1984. back |
