A seventy-year-old man was seen with a history of two previous operations on the lumbar spine. A laminectomy had been performed in 1964 because of pain in the lower extremities. The patient had had intermittent low-back pain since that procedure, but it had been effectively treated with manipulation and rest. In 1995, clinical symptoms of neurogenic claudication developed. These symptoms included progressive difficulty with walking secondary to pain in the low back and the buttocks as well as in both lower extremities. Sitting or leaning forward provided nearly complete relief of the symptoms. Because of the severity of the symptoms, the patient elected to proceed with operative treatment. As part of the preparation for the operation, a history was obtained and a physical examination was performed by an internal medicine specialist. The medical history included hypertension, a remote history of seizure disorder, and a thirty-year history of smoking a pipe. A radiograph of the chest showed no acute pulmonary process or mass. A laminectomy of the fourth and fifth lumbar vertebrae as well as a bilateral hemifacetectomy was performed. The laminectomy included removal of the cephalad one-half of the fifth lumbar lamina and the caudad three-fourths of the fourth lumbar lamina bilaterally. The patient had complete resolution of the preoperative symptoms.
Four months postoperatively, pain developed in the left buttock, the lateral aspect of the left thigh, and the left calf. This region corresponds with the distribution of the left fifth lumbar nerve. The patient perceived this pain to be distinctly different than the pain that he had had before the operation on the spine. The pain was aggravated by standing or walking, but, unlike before, it was not relieved by sitting or bending forward. The patient reported no sensory changes or weakness in the lower extremities. Bowel and bladder function were normal. Physical examination revealed a positive straight-leg-raising test at 70 degrees on the left. Sensation and strength in the lower extremities were normal. The deep tendon reflexes at the knee and ankle were symmetrical. The patient was referred for a lumbar epidural injection of steroid for both diagnostic and therapeutic purposes. The working diagnosis at the time of the injection was a radiculopathy of the left fifth lumbar nerve. The injection of steroid provided complete relief of the pain in the left lower extremity for two days; however, the pain returned on the third day. A magnetic resonance imaging scan demonstrated a large non-sequestered disc that extruded posterolaterally at the level between the fourth and fifth lumbar vertebrae, causing severe compression of the fifth lumbar nerve root (Fig. 1). There was no evidence of a signal change within the marrow of the lumbar vertebrae that would have indicated the presence of a tumor or an infection.
Clinically, the patient was thought to have a radiculopathy of the left fifth lumbar nerve because of the distribution of the pain, the positive straight-leg-raising test, the transient but complete relief of pain after the injection of steroid, and the appearance of an extruded disc at the level between the fourth and fifth lumbar vertebrae with compression of the fifth lumbar nerve root. Once again, the patient decided to proceed with operative treatment. As part of the preparation for the operation, an evaluation was performed by the same internal medicine specialist who had examined the patient previously. The internist noted two changes compared with the earlier findings. First, a transient ischemic attack, which was believed to have been cerebral in nature, had occurred two months previously. A computerized tomographic scan of the head was interpreted as normal. Second, the patient was found to have microscopic hematuria; a repeat urinalysis was therefore scheduled for the immediate postoperative period. The preoperative guidelines at our institution include the recommendation that a radiograph of the chest be made every two years for a patient who is more than sixty years old, who is currently a non-smoker, and who does not have a history of cardiovascular or pulmonary disease5. Because there had been no change in baseline respiratory function and no new related symptoms, and because a chest radiograph had been made eleven months earlier, radiography of the chest was not repeated.
A discectomy was done at the level between the fourth and fifth lumbar vertebrae, and a posterolateral arthrodesis was performed at the same level with use of autologous bone from the iliac crest. The aim of the bone-grafting was to obtain fusion and thereby reduce the risk of a recurrent disc herniation at this level. The operation was uneventful, and removal of the large extruded disc resulted in satisfactory decompression of the left fifth lumbar nerve root. The operative specimen appeared to consist of normal nuclear material on gross examination, and the loose aggregate measured 2.0 by 1.5 by 0.3 centimeter. The specimen was sent to the pathology department.
Postoperatively, the patient had complete resolution of the pain in the left lower extremity. The pathology report indicated that there were extensive degenerative changes, foci of neovascularity that were consistent with disc prolapse, and multiple fragments of malignant tumor within the operative specimen (Fig. 2). The histological findings were suggestive of a high-grade metastatic carcinoma. The tumor cells were concentrated along the peripheral margin of the herniated nucleus pulposus and appeared to be encapsulating the nuclear material that was within the epidural space. A new radiograph of the chest revealed a left mediastinal lesion. Computerized tomography demonstrated a 1.6-centimeter pulmonary mass in the left lower lobe, a 3.0-centimeter right renal mass, and a 1.0-centimeter right adrenal mass. A bone scan showed increased uptake in the region of the fourth and fifth lumbar vertebrae, consistent with the recent operation. There were no changes within the vertebral bodies that would have indicated involvement with an infection or a neoplastic process. The hospital stay was prolonged because of the workup for metastatic disease, and the patient was discharged to home on the fifth postoperative day.
The patient subsequently had a biopsy of the renal mass, which was found to be a renal-cell carcinoma, and a biopsy of the pulmonary mass, which was found to be a poorly differentiated adenocarcinoma. The cells of the adenocarcinoma were histologically similar to the tumor cells that had been found at the site of the herniated nucleus pulposus. Therefore, the patient had two distinctly different tumors. The first was a renal-cell carcinoma, which had metastasized to the right adrenal gland. The second was an adenocarcinoma of the lung, which had metastasized to the epidural space of the lumbar spine. The renal-cell carcinoma was treated with a regimen of cimetidine. The patient elected not to have chemotherapy for the adenocarcinoma. Follow-up computerized tomographic scans of the chest, pelvis, and abdomen, made two and five months postoperatively, showed no change in the size of the pulmonary, renal, or adrenal mass compared with the findings on the previous computerized tomographic scans. There was no other evidence of metastatic disease.
Eight months after the most recent operation on the lumbar spine, the patient needed an emergency repair of the ascending aorta because of an intramural hematoma. There were no complications, and the patient was discharged to home on the eighth postoperative day. Two months later, a malignant left pleural effusion developed and the patient was hospitalized. A thoracocentesis was performed, and approximately 2.5 liters of bloody exudate was drained. Two weeks later, the patient had a recurrent pleural effusion. A repeat thoracocentesis was scheduled, but the patient died before it could be done. No autopsy was performed.
The need for routine pathological examination of intervertebral disc specimens has been questioned by several authors. Boutin and Hogshead as well as Daftari et al. noted that the prevalence of unexpected findings on routine histological evaluation is suspected to be very low. Furthermore, assuming no change in patient outcomes, Daftari et al. estimated that $24,000,000 could be saved annually in the United States if intervertebral disc specimens were not submitted for routine histological examination. Both sets of authors concluded that routine pathological examination of such specimens is not essential for ensuring the quality of care and that the need for pathological examination can be determined by the clinician on the basis of the clinical presentation, preoperative studies, and operative findings1,2.
Although this may be true for many patients, it is not always the case. Our patient, who had a radiculopathy of the left fifth lumbar nerve and preoperative studies that were not suggestive of a malignant lesion, was found, on histological evaluation of the operative specimen, to have an adenocarcinoma that had metastasized from the lung to the site of the herniated intervertebral disc. The proposed explanation for the metastasis to this unusual location is that the increased blood flow in the area of the acute disc herniation increased the likelihood of a metastatic focus at this site. Although the probability that a previously unsuspected carcinoma will be identified during a routine pathological examination is low, the potential effect on the treatment and the outcome can be important. Daftari et al., in their analysis of cervical disc herniations, used what they termed the risk quotient and the chance ratio in an attempt to better define the probability that a pathological finding would be missed if specimens were not routinely submitted. With use of 95 per cent confidence intervals for exact proportions, those authors demonstrated that the probability of a missed diagnosis in a sample of 500 consecutive specimens would be as high as 0.006, or six of 1000. An attempt to reduce this ratio and narrow the confidence limits even further would require a sample size of more than 500 specimens. Those authors noted that the risk quotient for a true occurrence of an occult malignant lesion in the cervical spine cannot be readily calculated or estimated on the basis of epidemiological studies2. No such chance ratio has been determined for lumbar disc herniations, to our knowledge.
The concept that routine evaluation of the specimen is not necessary after a lumbar discectomy because the pathological findings would not have a clinically important effect on the quality of care is correct in most instances. The real debate centers on what risk quotient should be considered acceptable in the effort to avoid unnecessary medical services and their associated costs.
NOTE: The authors thank Dr. Donald Schreiber for his cooperation and assistance in preparing this manuscript.