[Postgraduate Medicine]

VOL 119 / NO 2 / July-August 2006 / POSTGRADUATE MEDICINE


When back pain is not benign

by Krzysztof Siemionow MD Robert F. McLain MD

Low back pain affects more than 85% of adults at some point in their life.1 It has a point prevalence of about 30% and is one of the most common complaints encountered by primary care physicians.1 Back pain may stem from pathologic changes in intervertebral disks, vertebrae, ligaments, neural structures, muscles, and fascia. The causes of pain typically relate to age-related degenerative changes or to minor trauma, and 90% of affected patients improve with simple supportive care and physical therapy.

Although serious, unrecognized diseases are uncommon causes of low back pain, they achieve clinical significance well out of proportion to their prevalence. Even experienced physicians are concerned with the possibility of overlooking serious or life-threatening pathologic conditions. Deyo and associates2 showed that malignancy, osteomyelitis, ankylosing spondylitis, and epidural abscess accounted for less than 1% of all cases of low back pain in a primary care practice. However, missing these diagnoses has serious consequences, and considerable time, effort, and anxiety are invested in ruling out these disorders. With a carefully taken history and physical examination, knowledge of a few specific signs and symptoms, and proper selection of diagnostic studies, the physician can confidently rule out more ominous underlying disease and focus the patient's attention on the proper course of rehabilitation and low back care.

Cancer of the spine

Unremitting pain often raises fears—in both patients and physicians—that “something bad” is going on, and cancer is the bad thing that most people fear. Because back pain is the presenting symptom in 90% of patients with spinal tumors, neoplasia does belong in the differential diagnosis of any case of persistent, unremitting back pain.3

The character of pain in patients with spinal cancer usually differs from that in common low back pain. Benign back pain is activity related, is relieved by rest, and may be precipitated by a recognized injury. Acute idiopathic back pain typically begins to subside after 4 to 6 weeks. Pain caused by spinal neoplasia is persistent, progressive, and not alleviated by rest. The pain seems to intensify at night and often awakens the patient from sleep. Back pain is typically focal and may be associated with beltlike thoracic pain or radicular symptoms of pain or weakness in the legs. A spinal mass can cause neurologic symptoms by directly compressing the spinal cord or cauda equina.4 Pathologic fractures resulting from vertebral destruction may be the first but, unfortunately, a late presentation of a tumor.


Signs and symptoms of systemic disease include fatigue, weight loss, abnormal bleeding, abdominal swelling, subcutaneous masses, and lymphadenopathy. Symptoms typical of common adenocarcinomas, such as hematochezia, hemoptysis, nipple discharge, atypical vaginal bleeding, and change in bowel habits, should prompt a specific diagnostic approach5 (table 1).

Clinical features associated with the likelihood of neoplastic disease include age greater than 50 years, previous history of cancer, duration of pain greater than 6 weeks, failure to improve with conservative therapy, an elevated erythrocyte sedimentation rate (ESR), and anemia.6

Physical examination

Carcinomas of the lung, breast, prostate, kidney, colon, and thyroid, along with multiple myeloma, account for 88% of all spinal tumors. These organs and systems should be thoroughly examined when cancer is sus-pected.7 Examination of the spine should identify sites of focal pain and elicit any signs of neurologic compromise or abnormal reflexes. Signs of spinal cord compression should be investigated immediately.

Diagnostic studies

The ESR, which is nonspecific but highly sensitive, is almost always elevated in cases of systemic neoplasia. A workup including chest radiography, mammography, measurement of prostate-specific antigen, and abdominal computed tomography usually reveals the underlying primary malignancy, if one exists.5

Basic laboratory studies may reveal anemia, hypercalcemia, and elevated levels of alkaline phosphatase. Serum and urine protein electrophoresis is specific for myeloma or plasmacytoma. Urinalysis may reveal hematuria, which suggests renal cell carcinoma.


Spinal tumors are poorly visualized on radiographic examina-tion8 until significant destruction has occurred. Bone scan is positive earlier and is considered a good screening test. However, magnetic resonance imaging (MRI) can screen the whole spine, which allows identification of lesions in patients with both normal radiographs and bone scans, and is the study of choice for ruling out spinal neoplasia9,10 (figure 1).


Spinal infections can be either acute or chronic. Acute infections are most often pyogenic, whereas chronic infections may result from pyogenic, fungal, or granulomatous disease. Vertebral osteomyelitis represents 2% to 7% of all cases of osteomyelitis and is an uncommon but significant cause of back pain.11 Half of affected patients are more than 50 years old, and two thirds are male. The most frequent cause is urinary tract infection, but any source of infection (eg, dental abscess, pneumonia) can seed the spine. Immunocompromised and diabetic patients are at particular risk.12


The presenting symptom is focal back pain that is worsened by weight-bearing and activity. Patients often experience exquisite pain that is relieved only when they lie down. Radicular signs are seen in 60% of patients, and spinal cord signs are found in 29%.13 Fever, chills, headache, and systemic illness are variably present. Chronic infection is often associated with weight loss and fatigue, and fevers and night sweats are common.

Physical examination

Pain is usually well localized and can be reproduced by palpation or percussion over the involved level. Severe pain may be elicited simply by having the patient sit up or change positions. If the vertebra has collapsed, focal kyphosis may be detected. Neurologic signs should be carefully sought.

Diagnostic studies

The ESR (the most sensitive test) is increased in 90% of patients with infection and may be the only abnormal laboratory finding. The C-reactive protein level is also usually elevated, but 40% of patients with infection have a normal white blood cell count. Other laboratory values are typically normal. A tuberculin skin test should be performed, with appropriate controls, in at-risk patients (ie, emigrants from areas of endemic disease, immunocompromised patients, non-Hispanic blacks, and patients with known exposure) (figure 2). Blood cultures should be obtained in any patient who has high fever, chills, or rigors.


Radiographic findings characteristic of osteomyelitis are not apparent for at least 4 to 8 weeks.14 Narrowing of the disk space is the earliest and most consistent radiographic finding but is nonspecific.13 MRI is as accurate and sensitive as nuclear medicine scanning (sensitivity 96%, specificity 93%, accuracy 94%),15 and it is capable of differentiating degenerative and neoplastic disease from vertebral osteomyelitis.16 Epidural abscess occurs in 10% of spinal infections, yet the condition is misdiagnosed on initial evaluation in 50% of affected patients.17,18 Patients initially report local spine pain, followed by radicular pain, weakness and, finally, paralysis. A history of trauma preceding the infection has been reported in 12% to 30% of patients.18 Radiographic findings are often equivocal, and MRI is the study of choice.19 The sensitivity of MRI is further increased by gadolinium enhancement.15


Fractures of the spine are not always a result of trauma and may be asymptomatic. They can be a result of osteoporosis, metabolic disorders, malignancy, or infection. Osteoporotic vertebral fractures occur in one third of American women older than 65 years, which makes them the most commonly encountered fractures in primary care.20 Fractures in normal bone are almost always related to a traumatic event.21


The presenting symptom may be back pain or neurologic deficit associated with a history of trauma. In osteoporotic patients, the trauma may be minimal—a sneeze, a fall from a chair, or a fall in the home. Osteoporotic compression fractures are highly associated with localized vertebral pain, age greater than 65 years, female sex, and European descent. Any patient with back pain who is receiving cortico-steroid therapy should be considered to have a compression fracture until proven otherwise.2 Compression fractures are rarely associated with neurologic deficits. New opportunities for intervention make timely evaluation clinically relevant.22 Any suspicion of underlying infection or malignancy should be investigated.

Physical examination

Localized pain over the involved vertebral level is usually present. The patient may report inability to flex or extend the involved segment because of pain or muscle spasm. Absence of the midline crease may be noted, or a kyphotic deformity may be pres-ent. A thorough neurologic workup is needed.

Diagnostic studies

Routine laboratory studies and thyroid function tests should be obtained. A 24-hour urine specimen can be assayed for collagen breakdown products and levels of calcium, phosphate, and creatinine. Serum and urine protein electrophoresis should be performed if there is a question of pathologic fracture due to myeloma.

Anteroposterior and lateral radiographs are the initial studies of choice.21 If a fracture is diagnosed (figure 3) or the examination is equivocal, then MRI is appropriate. MRI can also help resolve any questions about spinal cord involvement or fracture chronicity. Bone scan is reserved for patients in whom metastatic disease is suspected.

Visceral disease mimicking back pain

A variety of disorders of the abdominal viscera can produce low back pain, mimicking lumbar or thoracolumbar spinal disease. Some of these disorders are imminently life-threatening and should not be overlooked, even though they are rare.


Back pain caused by visceral sources is most often unrelated to activity and may be described as colicky or throbbing. Pain that is associated with eating, is related to the menstrual cycle, is colicky or cramping, or is reproduced by abdominal examination is unlikely to originate from spinal tissues. Fever, nausea, constipation, or diarrhea also may be present, depending on the cause.

Physical examination

Percussion over the costovertebral angle of the back typically reproduces pain of pyelonephritis or renal stone. In patients with peptic ulcer or colorectal disease, guaiac-positive stool is identified on rectal examination. Depending on the cause, the abdomen may be either tender and hyperdynamic or silent and rigid. Deep palpation may reveal guarding, rebound, or focal tenderness. Signs of an acute abdomen or palpation of a pulsatile mass should prompt an immediate surgical workup.

Diagnostic studies

In addition to the routine metabolic panel, an abdominal flat-plate radiograph may reveal evidence of free air, small-bowel obstruction, biliary disease, or aortic aneurysm. Abdominal computed tomography can further elucidate these findings, if indicated.

Findings in selected disorders

Knowledge of presentations and findings specific to various visceral diseases can aid accurate diagnosis.

Abdominal aortic aneurysm: This is the most serious vascular problem presenting as low back pain. The condition is misdiagnosed on initial presentation in 30% of affected patients.23 The pain of the aneurysm can be generated by compression of adjacent structures by the aorta or, in late stages, by dissection of the arterial wall. The pain of aortic dissection is intense and undiminished by narcotics. A palpable pulsatile abdominal mass is found in almost all cases. Lower-extremity pulses may be diminished or asymmetric. Patients with risk factors for peripheral vascular disease (eg, smoking, hypertension, diabetes) should be assessed for an abdominal aortic aneurysm anytime they present with atypical back pain.

Visceral disorders: Ulcers, especially those involving the posterior duodenal wall, may cause upper lumbar pain. A past history of ulcers is often prominent.

Renal pain is usually referred to the thoracolumbar junction and flank. Pyelonephritis, renal artery occlusion, and nephrolithiasis all may cause severe, colicky back pain. Bladder disorders may cause low back symptoms, usually concurrent with suprapubic discomfort and urinary symptoms.

Pancreatic disease produces pain in the upper lumbar region that is worse with recumbency. A past medical history of pancreatitis, jaundice, or alcoholism in patients with increased levels of amylase and lipase differentiates pancreatic pain from spinal pain. Pancreatic cancer is another potential cause of upper lumbar or thoracolumbar pain. As the tumor expands, it may irritate the peritoneal lining or directly distort the nerve endings encapsulating the pancreas.

Pain of pelvic origin caused by ovarian torsion or rupture, ectopic pregnancy, endometriosis, or fibroids may present as back pain unrelated to changes in body position or movement.

Spinal cord and cauda equina compression

The patient presenting with acute or progressive weakness or bowel or bladder dysfunction should immediately trigger greater concern and attention than any patient with back pain alone. Acute cauda equina compression is occasionally caused by a massive disk herniation but is more often the result of fracture, tumor, or epidural hematoma or abscess. The presence of a cord-level or cauda equina deficit should trigger an aggressive search for the cause.24

Patients with cauda equina syndrome typically present with urinary retention, whereas those with cord compression present with incontinence. The classic symptoms of low back pain, bilateral sciatica, saddle anesthesia, and lower-extremity weakness progressing to paraplegia are variably present, requiring a high degree of suspicion by the examining physician.25 Hyporeflexia is typically a sign of cauda equina compression. Hyperreflexia, clonus, and a positive Babinski's test all suggest spinal cord compression, necessitating an evaluation of the cervical and thoracic spine. MRI is the diagnostic study of choice. Surgical decompression is warranted on an emergent basis if a compressive cause is identified.

Rarely, low back pain is caused by an epidural hematoma. The clinical presentation of a hematoma may mimic a disk herniation.26 Epidural hematomas most often occur after spine surgery or in patients receiving anticoagulation therapy. Patients with motor deficits or urinary retention need emergent surgical decompression of the hematoma.


Spinal disorders can be complex challenges, and the ultimate treatment is sometimes highly specialized. However, the evaluation and diagnosis of the most dangerous causes of back pain are more a matter of careful medical evaluation—taking a thorough history and performing a careful physical examination—than of specialized spinal knowledge or testing. Once dangerous pathologies are ruled out, both the patient and the physician can confidently focus on more routine back care.

Dr Siemionow is a resident, department of orthopedic surgery, Cleveland Clinic Foundation. Dr McLain is professor of surgery, department of orthopedic surgery, Lerner College of Medicine, and a member of The Cleveland Clinic Spine Institute, Cleveland Clinic Foundation.


  1. DeyoRA, Tsui-WuYJ . Descriptive epidemiology of low-back pain and its related medical care in the United States Spine 1987; 12( 3): 264-8

  2. DeyoRA, RainvilleJ, KentDL . What can the history and physical examination tell us about low back pain? JAMA 1992; 268( 6): 760-5

  3. GilbertRW, KimJH, PosnerJB . Epidural spinal cord compression from metastatic tumor: diagnosis and treatment Ann Neurol 1978; 3( 1): 40-51

  4. RosenP . Emergency medicine: concepts and clinical practice 4th ed. St Louis: Mosby 1998: 2100-2

  5. AbbruzzeseJL, AbbruzzeseMC, LenziR, et al. Analysis of a diagnostic strategy for patients with suspected tumors of unknown origin J Clin Oncol 1995; 13( 8): 2094-103

  6. DeyoRA, DiehlAK . Cancer as a cause of back pain: frequency, clinical presentation, and diagnostic strategies J Gen Intern Med 1988; 3( 3): 230-8

  7. McCarthyEF, FrassicaFJ . Pathology of bone and joint disorders with clinical and radiographic correlation Philadelphia: Saunders 1998

  8. EdelstynGA, GillespiePJ, GrebbellFS . The radiological demonstration of osseous metastases. Experimental observations Clin Radiol 1967; 18( 2): 158-62

  9. SeldenNR, QuintDJ, PatelN, et al. Emergency magnetic resonance imaging of cervical spinal cord injuries: clinical correlation and prognosis Neurosurgery 1999; 44( 4): 785-92; discussion 792-3

  10. CitrinDL, BessentRG, GreigWR . A comparison of the sensitivity and accuracy of the 99TCm-phosphate bone scan and skeletal radiograph in the diagnosis of bone metastases Clin Radiol 1977; 28( 1): 107-17

  11. WaldvogelFA, MedoffG, SwartzMN . Osteomyelitis: a review of clinical features, therapeutic considerations and unusual aspects N Engl J Med 1970; 282( 4): 198-206

  12. CarrageeEJ . Pyogenic vertebral osteomyelitis J Bone Joint Surg Am 1997; 79( 6): 874-80

  13. KapellerP, FazekasF, KrametterD, et al. Pyogenic infectious spondylitis: clinical, laboratory and MRI features Eur Neurol 1997; 38( 2): 94-8

  14. DigbyJM, KersleyJB . Pyogenic non-tuberculous spinal infection: an analysis of thirty cases J Bone Joint Surg Br 1979; 61( 1): 47-55

  15. KukerW, MullM, MayfrankL, et al. Epidural spinal infection. Variability of clinical and magnetic resonance imaging findings Spine 1997; 22( 5): 544-50; discussion 551

  16. TungGA, YimJW, MermelLA, et al. Spinal epidural abscess: correlation between MRI findings and outcome Neuroradiology 1999; 41( 12): 904-9

  17. NussbaumES, RigamontiD, StandifordH, et al. Spinal epidural abscess: a report of 40 cases and review Surg Neurol 1992; 38( 3): 225-31

  18. DannerRL, HartmanBJ . Update on spinal epidural abscess: 35 cases and review of the literature Rev Infect Dis 1987; 9( 2): 265-74

  19. AngtuacoEJ, McConnellJR, ChadduckWM, et al. MR imaging of spinal epidural sepsis AJR Am J Roentgenol 1987; 149( 6): 1249-53

  20. RiggsBL, MeltonLJ3rd . Evidence for two distinct syndromes of involutional osteoporosis Am J Med 1983; 75( 6): 899-901

  21. McLainRF, BensonDR . Thorac-olumbar spine fractures: evaluation, classification, and initial managementIn: ChapmanMW, ed. Chapman's operative orthopaedics 3rd ed. Philadelphia: Lippincott Williams & Wilkins 2001: 3713-24

  22. LiebermanIH, DudeneyS, ReinhardtMK, et al. Initial outcome and efficacy of “kyphoplast” in the treatment of painful osteoporotic vertebral compression fractures Spine 2001; 26( 14): 1631-8

  23. MarstonWA, AhlquistR, JohnsonGJr, et al. Misdiagnosis of ruptured abdominal aortic aneurysms J Vasc Surg 1992; 16( 1): 17-22

  24. SpangfortEV . The lumbar disc herniation. A computer-aided analysis of 2,504 operations Acta Orthop Scand Suppl 1972; 142: 1-95

  25. TayEC, ChachaPB . Midline prolapse of a lumbar intervertebral disc with compression of the cauda equina J Bone Joint Surg Br 1979; 61( 1): 43-6

  26. Alexiadou-RudolfC, ErnestusRI, NanassisK, et al. Acute nontraumatic spinal epidural hematomas. An important differential diagnosis in spinal emergencies Spine 1998; 23( 16): 1810-3