Background and Diagnostic Alternatives

JAMES ZUCHERMAN, MD
JEROME SCHOFFERMAN, MD
St. Mary's Spine Center,
San Francisco, California

More than 200,000 patients undergo lumbar spine surgery each year.[12] The patient and surgeon expect a successful outcome, defined as elimination or significant reduction of pain and disability, and markedly improved function. It is hoped that the patients will return to work, discontinue medications and resume a normal place in family and society. Unfortunately, 20-40% of patients will fail to gain the desired outcome. In fact, 1-10% of patients will be worse after initial surgery.[12] The physician who must treat the patient who has the failed back surgery syndrome (FBSS) faces an even more formidable task than the initial physician. He or she must consider the many possible causes for FBSS, which include structural and mechanical problems with the spine itself, related or unrelated to the initial surgery, poor body mechanics and deconditioning, alternative diagnoses which were the actual cause of the initial back pain or a contributing cause which arose after the surgery, psychological variables, and medication problems. An overview of each of these areas is essential to help the physician plan diagnostic evaluation and short- and long-term therapy.

ALTERNATIVE DIAGNOSIS

Only a small number of patients who fail to benefit from lumbar spine surgery will be found to have disorders unrelated to the initial indication for surgery. However, major medical problems can present as acute, chronic, or relapsing back and/or leg pain. These may have coexisted with the degenerative disease of the spine for which the surgery was done, or may have arisen later. Only when the surgery is unsuccessful does a second look provide the correct diagnosis.

NEOPLASMS

Neoplastic disease is the most feared cause of pain in the low back. It may be metastatic or primary. Neoplasm as a cause of low back pain is more likely to occur after age 50.[3] In a retrospective series of 259 consecutive patients over the age of 50 with low back pain presenting to a single practice, 13 had metastatic carcinoma and four, primary carcinoma of the spine.[3] Cancers of the lung, breast, and prostate make up 55% of carcinomas metastatic to the spine.[10] Multiple myeloma, chondrosarcoma, other metastatic malignancies, and a host of rare benign or malignant tumors can present as back pain.

The history of a patient with cancer presenting as low back pain is nonspecific. Of course, the presence of a known primary neoplasm must always arouse suspicion. Symptoms of systemic disease must be sought in all patients with back pain and may be the most useful clue. A history of pain that is worse at night and is only minimally exacerbated by activity is said to be a useful clue to the presence of malignancy, but we have found this to be unreliable.[4] Likewise, the examination is not specific unless a primary carcinoma is discovered elsewhere on examination. Signs of systemic disease may raise suspicion. The presence of neurologic signs in areas neurologically unrelated to the lumbar spine has helped us diagnose neoplasms in several patients.

Radiographic evaluation may be more fruitful. A complete lumbar spine series may reveal osteolytic or osteoblastic changes. Diffuse or localized osteoporosis may be a sign of multiple myeloma or other cancer, especially when the osteopenia is out of proportion to the patient's age and risk factors. Bone scan is indicated if metastatic cancer is suspected and is most helpful if changes are found in areas other than the previous operative site. Routine L3-to-S1 CT scans may be definitive when positive but too limited an area is visualized to be used to rule out systemic or regional cancer. Myelography is still important if other studies are unrevealing and suspicion remains. Magnetic resonance imaging is still investigational. Laboratory evaluation, particularly the hemoglobin, serum alkaline phosphatase, serum calcium, and sedimentation rate, may provide useful information. In fact, in Fernbach's series all patients with cancer presenting as low back pain had an abnormality in at least one of the latter three tests.[3]

INFECTIONS

Infection is reported to be a cause of back or leg pain in 1-3% of patients with FBSS.[8] Infections can occur early or late in the postoperative course. They may involve the soft tissues overlying the spine, the bone, the disc space, or the epidural space, or may cause frank meningitis. Organisms may be introduced at the time of surgery but also may ascend from the urinary tract in patients with urinary tract infections or indwelling catheters. Infection may be blood-borne after introduction from a skin or intravenous site. Hematoma formation or continued wound drainage may pre-dispose to infection. A wide variety of organisms may be responsible. Staphylococcus aureus is the most common, but Staphylococcus epidermidis, streptococci, anaerobes, and gram-negative organisms are also seen. Isolated "diphtheroids" should not be dismissed as contaminants, as we have recently documented several infections due to these organisms. Patients in whom infections develop early in the postoperative period usually do not present difficult diagnostic problems. They often remain febrile beyond the first few postoperative days and their pain increases rather than decreases. Frequently, however, there is an initial decrease in pain before this secondary increase. There may be expanding erythema around the incision or purulence from the wound site, but in our experience absence of superficial evidence of infection is quite common. A hematoma may be present and, if aspirated, should be cultured. Blood cultures should be obtained at that time. Leukocytosis is not very helpful unless is has been demonstrated that the white blood cell count had returned to normal after surgery.

Other laboratory data are not useful, although studies in progress evaluating the role of the erythrocyte sedimentation rate may prove fruitful. Radiographic tests are not very helpful in identifying the early postoperative infection, nor are technetium bone scans or CT scans. Gallium scanning and labeled white blood cell scanning are of questionable value. There is no substitute for recovering an organism, and if the suspicion of infection arises, aggressive measures should be taken to obtain the organism. If infection of the meninges is suspected, cerebrospinal fluid must be obtained from a site removed from the area of incision.

Infections occurring weeks to months postoperatively may be more difficult to diagnose. We have reviewed our experience with FBSS patients who have had infections as the cause of the failures. Several had been operated on at our institution, but most were referred to us. Patients generally had improved after surgery but then experienced increased pain with obvious precipitants. They often reported low-grade fevers and/or night sweats, but this was not consistent. A useful clue has been the failure of these patients to regain their overall general health, with continuing poor appetite, weight loss, and general malaise. Leukocytosis and differential white blood cell counts have been variable and are useful only when quite abnormal. Plain xrays, technetium bone scans and CT scans are obtained but generally do not help. The Westergren sedimentation rate is usually elevated, but again, this is nonspecific and a normal rate does not rule out infection. Gallium scan has been positive in a few patients.

VISCERAL CAUSES

In our experience it has been most unusual for pain in the patient with FBSS to be the result of visceral disease. Upper back pain can rarely be a pattern for angina pectoris or may be a symptom of acute aortic dissection. Neither of these is likely to present a diagnostic problem. Pancreatic carcinoma or chronic pancreatitis can cause chronic low back pain in the area T10-L2. Of course, a large pelvic tumor can compress neural structures and cause back pain or radicular pain.

Renal sources of pain may present more difficulty. Renal infection may present as back pain near the costovertebral angle. It is usually unilateral. Fever and chills may be present. Renal carcinoma can cause a dull aching low thoracic or upper lumbar pain distribution.

Chronic prostatitis can be a source of dull low back pain or sacral pain, often associated with perineal aching. Carcinoma of the prostate is usually painless if confined to the gland but may cause pain when there is local invasion of surrounding structures. Rectal examination must be part of any thorough back examination and remains the best screening test for carcinoma of the prostate. Gynecologic sources of low back pain include endometriosis, infection, and traction caused by a large uterus.

MISCELLANEOUS CAUSES

Many other conditions can cause pain in the patient who has had prior back surgery. Degenerative arthritis due to instability and recurrent inflammation and osteophyte formation is rarely overlooked. Ankylosing spondylitis may have been present and overlooked prior to the initial surgery. The hallmarks of the disease are the presence of back pain and radiographically evident sacroileitis. Ankylosing spondylitis was first believed to be more common in men, but recent data suggest equal distributions in men and women. The onset is usually before the age of 40 and is insidious. The patient experiences pain and morning stiffness, which abate with exercise as the day progresses. There may be features of peripheral arthritis as well. Limited chest expansion, limitation of lumbar spine motion, and tenderness of the sacroiliac joint may be seen, but recent evidence has shown these physical findings to be less definitive than previously thoughts Plain x-rays of the sacroiliac joints may show sclerosis, but early in the illness CT scanning or bone scans of the sacroiliac joints may be more sensitive. The finding of bamboo spine is quite rare, and calcification of the anterior spinal ligament is a late finding and also rare. Laboratory data will show a positive HLAB-27 antigen in 90% of white patients with ankylosing spondylitis. The sedimentation rate is elevated in only 80% of patients. Rheumatoid factor is absent. Other spondyloarthropathies are less common but should be considered. Enteropathic arthropathy (Crohn's disease or ulcerative colitis), Reiter's syndrome, and psoriatic arthropathy all may cause low back pain and may have been previously overlooked. Metabolic bone disease is yet another cause of back pain and FBSS. Patients with symptomatic metabolic bone disease are similar in age to patients with degenerative disease of the lumbar spine. Once again, failure to improve after surgery should lead to the consideration of coexistent metabolic bone disease as a cause of the unresolved component of pain.

Osteoporosis is generally painless unless there are micro- or macrocompression fractures. Acute compression fractures will cause pain that lasts three to six weeks and then diminishes. However, secondary pain patterns may occur due to progressive kyphosis, mechanical back pain, or myofascial pain syndromes. Physical therapy is often helpful for these patients, and treatment of the underlying metabolic bone disease may prevent further fracture.[11]

Osteomalacia can cause diffuse pain in the areas of the lumbar spine, hips, knees, and legs. Compression fracture may occur. Diagnosis is usually made during the work-up of radiographic osteopenia. Paget's disease of bone can also cause back pain, and radiographic features are distinct. Vascular claudication can present as failed back surgery manifested as lack of alleviation of leg pain despite adequate spinal decompression. It is well appreciated that intermittent claudication can be of either vascular or neurogenic etiology. It may be less well appreciated that patients can have the two conditions concurrently.[5] Evaluation of the arterial system with noninvasive or invasive studies is warranted for patients whose claudicant symptoms persist despite adequate surgery, especially if examination discloses evidence of peripheral arterial vascular disease.

PSYCHOLOGICAL CAUSES

Currently accepted theories of pain argue convincingly that pain perception is a complex phenomenon which involves more than just the peripheral stimulus. In fact, it is the necrologic, biochemical, and emotional reaction to a particular stimulus that determines whether or not a person finds that stimulus painful. A stimulus that one individual considers a mere annoyance can in a different person evoke severe pain, use of analgesics, loss of job, and other major disruptions of life. The response to a peripheral stimulus is determined by underlying personality, primary and secondary psychosocial factors, the neurochemical milieu, and social, family, and economic factors. Ethnic influences, prior experiences with pain, and cognitive factors also contribute to the sensation of pain. Pain must be evaluated as a multidimensional experience and attention paid to each factor in order to obtain treatment success. We are totally dependent on the patient's report of pain in order to evaluate it.

The interaction of biological factors with psychological and social ones is one of the most complex and fascinating aspects of working with patients with FBSS. Even experienced clinicians are fooled by underestimating psychological changes that surface only when the patient is stressed and clearly contribute to the patient's pain. It is unusual for the reverse situation to be seen, that is, for a patient with severe psychological problems to improve psychologically after the underlying organic lesion is fixed. This interrelationship is a major reason that we prefer to hospitalize patients with FBSS for their multidisciplinary evaluation. The various physical, psychological and social stressors that occur routinely during hospitalization may help staff get much more insight into the patient's pain and the patient's behavior. Interactions with nursing staff in the evenings and nights are often quite revealing.

Most patients with FBSS have some psychological changes. In fact, denial of psychological changes by the patient leads us to believe they are playing even a larger role. Depression is the most common psychological problem seen in patients with chronic pain and FBSS.' Depression may have preceded the pain or may be a reaction to the pain, but often the temporal relationship is not clear. Vegetative signs of depression are often present. The patient may attribute them to the pain itself and deny underlying depression. Depression lowers the pain threshold and tolerance, leading to progressive pain perception which further worsens the depression, resulting in a downhill spiral.

Anxiety levels are often quite high in patients with FBSS. There may be high levels of autonomic nervous system arousal, but frequently adaptive changes have occurred. Frequently the anxiety coexists with an underlying depression.

Many patients with FBSS tend to be overly concerned with bodily functions and sensations. They tend to elaborate symptoms involving multiple organ systems, which too frequently leads to unnecessary and costly tests and procedures. Some patients may be classified as hypochondriacal by DSM-3 criteria. Classic somatization disorder and conversion disorder are seen only rarely. We also rarely see patients who are frank malingerers.

Substance abuse problems are quite common and exist conjointly with other problems of the psychological and social realms. Patients may be dependent on narcotic analgesics or sedative hyptonics and the medications themselves become part of the pathological problem. There may be dependence, drug-seeking behavior, worsening of depression, and episodes of withdrawal which are manifested as increased pain, anxiety, or sleep disturbance. There is frequently impairment of familial, social, or occupational roles directly related to misuse of narcotic analgesics or sedatives.[1] Social factors interact with psychological and physical factors, and the borders become blurred. Secondary gains are often identified and may be of various types. Pain may be a means of gaining needed attention from family, friends, or co-workers. Pain may also serve as a way to avoid unpleasant family or job situations.

The effects of worker's compensation payments or pending litigation on the outcome of treatment in FBSS remains controversial.[2,6] Traditionally these patients have been considered to be malingering, consciously or unconsciously exaggerating their pain complaints, and to be suffering from greater psychological problems compared with other patients with chronic pain who are not involved in the worker's compensation system or litigation.[6,7] Despite clinical impressions that these patients are more difficult to help, the body of current literature does not support the feeling that they have less severe organic disease, more psychopathology, or less pain.[2,7] It is difficult to believe that an injured worker with a family who has his income cut by 50-75% and no legal possibility of a large settlement at the conclusion of the case has a meaningful financial secondary gain. In fact, many patients we have seen have lost their homes and have undergone marked deterioration of their standards of living. Potentially large personal-injury suits may be more likely to influence outcome of treatment, but the literature is not definitive. It does seem important to discuss realistic expectations for outcome, including pain and future employment possibilities, with the patient while treating FBSS.

REFERENCES

1. Aronoff GM: Psychological aspects of nonmalignant chronic pain: a new nosology.
In: Aronoff GM: Evaluation and Treatment of Chronic Pain. Baltimore, Urban and Schwarzenberg, 1985, pp 471-484.
2. Dworkin RH, Handlin DS, Richlia DM, et al: Unraveling the effects of compensation, litigation, and employment on treatment response in chronic pain. Pain 23:49-59, 1985.
3. Fernbach JC, Langer F. Gross AK: The significance of low back pain in older adults. Canad Med Assoc J 115:898-900, 1976.
4. Francis KC, Hutter VP: Neoplasms of the spine in the aged. Clin Orthop 26:54, 1963.
5. Johansson JE, Bassington TW, Ameli M: Combined vascular and neurogenic claudication. Spine 7:150-158, 1982.
6. Melzack R. Katz J. Jeans ME: The role of compensation in chronic pain: analysis using a new method of scoring the M.G.H. pain questionnaire. Pain 23:101-112, 1985.
7. Mendelson G: Compensation, pain complaints and psychological disturbance. Pain 20: 169-177, 1984.
8. Russell AS, Maksymowych W. LeClerq S: Clinical examination of the sacroiliac joints: a prospective study. Arthritis Rheum 24:1575-1577, 1981.
9. Selby DK: When to operate and what to operate upon. Orthop Clin North Am 19:577588, 1983.
10. Schoberg J. Gainor BJ: A profile of metastatic carcinoma of the spine. Spine 10:19-21, 1985.
11. Schofferman J: Osteoporosis. In: White AH, Rothman R (eds): Lumbar Spine Surgery: Techniques and Complications. St. Louis, C.V. Mosby Co., 1986.
12. Wilkinson HA: The role of improper surgery in the etiology of the failed back syndrome. In Wilkinson NA (ed): The Failed Back Syndrome. Etiology and Therapy. Philadelphia, J.B. Lippincott, 1983, pp 15-16.