Abstract
Forty-two neurologically intact adults in whom non-operative treatment of grade-I or grade-II isthmic spondylolisthesis of the most caudad lumbar segment had failed were entered into a prospective study of the results of operative treatment. Twenty patients who smoked were managed with a posterolateral arthrodesis with instrumentation (transpedicular fixation), and twenty-two patients who did not smoke were managed with a posterolateral arthrodesis without instrumentation. Of the patients who were managed with instrumentation, eight were randomized to treatment with a decompressive laminectomy and twelve, to treatment without it; in the group that was managed without instrumentation, the distribution was ten and twelve patients, respectively.The patients were followed clinically for a mean of 4.5 years (range, 3.5 to six years). Of the eighteen patients who had been managed with decompression, four had a pseudarthrosis and six had an unsatisfactory result compared with none and one of the twenty-four who had been managed without decompression (p = 0.02 and p = 0.01, respectively). In the group of twenty patients (smokers) who had been managed with instrumentation, none of the twelve managed without decompression had a pseudarthrosis compared with one of the eight managed with decompression (p = 0.2). In the group of twenty-two patients (non-smokers) who had been managed without instrumentation, none of the twelve managed without decompression had a pseudarthrosis compared with three of the ten managed with decompression (p = 0.04). In the group managed with instrumentation, two of the eight who had had decompression had an unsatisfactory result compared with none of the twelve who had not had decompression. In the group managed without instrumentation, four of the ten who had had decompression had an unsatisfactory result compared with one of the twelve who had not had decompression.The addition of decompression to arthrodesis, performed with or without instrumentation, for the treatment of low-grade isthmic spondylolisthesis in patients who do not have a serious neurological deficit does not appear to improve the result and may significantly increase the rates of pseudarthrosis and unsatisfactory results.
Low or moderate-grade isthmic spondylolisthesis in symptomatic adults usually can be treated successfully with non-operative means15. However, the type of operation that should be performed when patients have not responded to non-operative treatment remains controversial.
Some authors have recommended decompression and arthrodesis, and others have recommended arthrodesis alone1,3,6,9-15. Similarly, arthrodesis with instrumentation has been advocated by some, whereas arthrodesis without instrumentation has been recommended by others1,3,6,9-11,13. The disc (usually between the fourth and fifth lumbar vertebrae) just cephalad to the level of the arthrodesis often degenerates during the third and fourth decades of life7,15, and some authors have advocated inclusion of the fourth and fifth lumbar levels in arthrodeses performed in adults10,15. Others have suggested that the integrity of the more cephalad disc should be evaluated with magnetic resonance imaging or discography to determine if it is at risk for early failure or if it is a primary generator of pain7,15. Smoking seems to be a major risk factor for failure of lumbar arthrodesis in general and for those performed for spondylolisthesis in particular; this risk appears to be mitigated by the application of rigid transpedicular fixation of the segment to be included in the arthrodesis10,18.
The literature on this subject is confusing because clinical reviews often have involved mixed groups, such as children, adolescents, and adults; spondylolistheses with different etiologies and of different degrees of severity; patients managed with and those managed without instrumentation; smokers and non-smokers; single and multilevel arthrodeses; posterolateral and interbody techniques; and so on.
The current study involved a relatively simple, single group: patients who had had a primary posterolateral arthrodesis at the fifth lumbar and first sacral vertebral levels, with a normal disc cephalad, at the fourth and fifth lumbar levels. The results when a decompressive laminectomy had been carried out were compared with the outcomes when decompression had not been carried out. Whether or not transpedicular instrumentation was used was determined by the patient's smoking history: all patients who smoked were managed with instrumentation, and all who did not were managed without it.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Funds were received in total or partial support of the research or clinical study presented in this article. The funding source was the Division of Orthopaedic Surgery, Stanford University School of Medicine.
†Division of Orthopaedic Surgery, Room R-171, Stanford University School of Medicine, 300 Pasteur Drive, Stanford, California 94305-9805. E-mail address: carragee@leland.stanford.edu.
Between 1989 and 1992, during the evaluation of 3589 new patients who had a spinal disorder at Stanford University Medical Center, Stanford, and the Veterans Affairs Medical Center, Palo Alto, I identified and examined 544 patients who had a primary diagnosis of lumbar spondylolisthesis. Operative treatment was recommended for sixty-three of these 544 patients, and forty-nine consecutive patients met the criteria for inclusion in the study. The criteria for inclusion were a grade-I or II7 isthmic spondylolisthesis at the most caudad mobile lumbar segment (usually the fifth lumbar vertebra), an age of more than eighteen years, persistent pain in the back and the lower extremities despite non-operative treatment for at least six months, normal findings on a discogram or a T2-weighted magnetic resonance image of the immediately cephalad intervertebral disc, no previous operation on the spine or coexisting spinal deformity, no signs or symptoms of cauda equina, and no more than trace weakness on testing of motor function (that is, motor function of at least 5-) of the fifth lumbar and first sacral muscle groups. Of the forty-nine patients who met the criteria, forty-six agreed to participate.
The patients were randomized to treatment with posterolateral arthrodesis only or to treatment with posterolateral arthrodesis and decompression. All smokers were assigned to management with instrumentation (with the Texas Scottish Rite Hospital transpedicular screw-and-rod system [Danek, Memphis, Tennessee]), and all non-smokers were assigned to management with the arthrodesis only. Both groups had placement of autogenous bone graft from the iliac crest, and allogenic bone graft (morselized cancellous-bone chips) was used as well in four patients who had decompression and in three who had an arthrodesis only.
The random assignment to treatment either with decompression and arthrodesis or with arthrodesis alone was achieved with use of sealed envelopes containing random numbers that had been generated by a computer program. Odd and even numbers determined the use of decompression or the lack of it, respectively. The envelopes were batched in groups of ten for the smoker and non-smoker groups to help to equalize the distribution.
Approval for the study was obtained, in 1989, from the institutional review board and the Administrative Panel of Human Subjects in Medical Research, according to regulations of the Department of Health and Human Services, at both the Veterans Administration Medical Center and the Stanford University School of Medicine; the approval was granted for a five-year trial, and it was renewed annually through 1992. In 1992, the protocol was discontinued because the results seemed to be poorer for the group that had been managed with decompression (laminectomy). Consent was obtained from all patients entering the study, according to guidelines of the institutional review board.
Decompression involved removal of the free lamina and decompression of the fifth lumbar nerve root and thecal sac. This involved removal of the fibrocartilage, fragmented pseudarthrosis bone fragments, and connective tissue about the nerve root to the point where the nerve exited the neural foramen. In the group that was managed with instrumentation, no attempt was made to reduce the degree of the slip or the angle of the spondylolisthesis other than by employing the prone position used routinely for the operative approach. All patients were managed with a Boston-type wraparound brace postoperatively; the brace was worn full-time for at least six weeks, with patients instructed to wean themselves from it during the next six weeks.
Preoperatively, in addition to recording a history and performing a physical examination, standardized testing was carried out. This included an assessment of pain in the back and the lower extremities with use of a 10-point visual-analog scale that ranged from 0 ("no pain") to 10 points ("pain so severe it cannot be tolerated for more than a few minutes"). The level of activity also was evaluated with use of a 10-point visual-analog scale ranging from 0 ("all activities are limited") to 10 points ("full activity without limitations"). The durations of acute and chronic symptoms, intake of pain medication, work status, and concurrent medical problems were recorded.
Of the forty-six patients who initially had been enrolled in the study, two who had been assigned to management with instrumentation were excluded when bilateral transpedicular instrumentation of the fifth lumbar and first sacral levels could not be performed. Another patient, who had been managed with decompression, died in an automobile accident at six weeks. A fourth patient, who also had been managed with decompression, was lost to follow-up shortly after the operation because he was incarcerated in a federal prison out of state. Thus, the rate of follow-up was 98 per cent (forty-two of the forty-three patients who had survived and had not been dropped from the study). Of the forty-two patients, eighteen (seven women and eleven men) were managed with decompression and twenty-four (nine women and fifteen men) were managed without decompression (Table I). Three patients (7 per cent) had litigation pending because of a motor-vehicle accident, and six (14 per cent) had a Workers' Compensation claim. Twenty patients (48 per cent) were smokers, and twenty-two (52 per cent) were not. The durations of acute and chronic low-back symptoms were similar in all four groups, as were the ratings for pain in the back and the lower extremities, intake of medication, and level of activity.
Follow-up evaluations were performed at two, six, twelve, and twenty-four weeks and then at one and two years. The most recent follow-up evaluation was done at a minimum of three years (mean, 4.5 years) after the operation. This consisted of a standardized interview, conducted either at the clinic or on the telephone. The patients were asked specifically about their current intake of medication for back or radicular pain, their work status, and their satisfaction with the result of the index operation; pain in the back and the lower extremities as well as level of activity were assessed with use of 10-point visual-analog scales in the same manner as preoperatively. The patients who were interviewed by telephone had marked the visual-analog scales before the conversation, and these were available to the interviewer. The patients who were interviewed at the clinic filled out the same forms in the waiting areas. In addition to the scores for pain in the back and the lower extremities, an over-all score was calculated by averaging the scores for pain, activity, use of medication, and satisfaction, as described previously2. Satisfactory (good and excellent) and unsatisfactory (fair and poor) results were determined and analyzed as described by Deyo et al.
At the time of the most recent follow-up visit, anteroposterior and lateral radiographs of the lumbar spine, with the patient standing, and Ferguson radiographs of the lumbosacral junction were made. Radiographs with the spine in flexion and extension had been made for all patients at twelve months postoperatively. In the patients who had been managed without instrumentation, fusion was considered to have occurred if there was a trabeculated mass between the transverse processes of the cephalad and caudad segments and less than 2 degrees of motion as seen on the radiographs of the spine in flexion and extension18. In the patients who had been managed with instrumentation, fusion was assessed in a similar manner, except that, in addition, if there was a radiolucent line about a pedicle screw (six patients) or if a screw had fractured (one patient), the radiographs with the spine in flexion and extension were repeated. In some patients (as will be described later), a pseudarthrosis was confirmed at the time of operative exploration of the fusion mass.
Data were prospectively entered into the statistical computer program StatView (Abacus, Mountain View, California). Continuous data were analyzed with use of the Student t test, and categorical data were analyzed with use of the chi-square test.
The forty-two patients were followed for a mean of 4.5 years (range, 3.5 to six years). At the most recent follow-up evaluation, four of the eighteen patients who had been managed with decompression had had a pseudarthrosis; three of these patients had had an arthrodesis without instrumentation, and one, an arthrodesis with instrumentation. All of the twenty-four patients who had been managed with an arthrodesis without decompression had a successful fusion (p = 0.02).
Six of the eighteen patients managed with decompression and one of the twenty-four managed without it had an unsatisfactory result (p = 0.01) (Table II). In the group that had had an arthrodesis with instrumentation, there was no significant difference, with the numbers available, between the outcome when a decompression had been done and the outcome when it had not been done. However, there was a trend toward better scores in all categories, particularly pain in the lower extremities and use of medication, for the patients managed without decompression. There was one poor result, in a patient who had had a solid fusion after an arthrodesis with instrumentation. This patient had rapid degeneration and collapse of the disc at the fourth and fifth lumbar levels three years postoperatively.
Of the patients who had had an arthrodesis without instrumentation, those who had been managed with decompression had poorer mean scores in all categories. This trend was most significant with regard to pain in the lower extremities (mean, 4.4 points compared with 1.9 points in the patients who had not had decompression; p = 0.06). No patient who had had an arthrodesis without either instrumentation or decompression had a poor result.
Three of the four patients who had a pseudarthrosis subsequently had an attempted revision with use of anterior interbody arthrodesis and posterior instrumentation; one of these revisions failed. The fourth patient declined additional operative treatment. All four patients who had a pseudarthrosis had an unsatisfactory clinical result (two poor and two fair results; p < 0.0001), despite a successful salvage arthrodesis in two patients who had had decompression but had not had instrumentation.
There were nine additional complications. Two of the complications (in two [10 per cent] of the twenty patients) were associated with instrumentation. One of these patients had irritation of the fifth lumbar nerve root after instrumentation without decompression, and an operation was necessary to reposition the screw at that level. Laminotomy and exploration of the nerve root was performed at the time of the revision. The other patient had disassembly of the screw-rod coupler at two weeks postoperatively, and a revision was done. Two patients had a urinary-tract infection, which was treated with antibiotics. One patient had a dural tear. One patient had a superficial wound infection, which resolved with local care and antibiotics. One patient, who had diabetes, had acute cholecystitis on the third postoperative day. One patient had meralgia paresthetica, necessitating discontinuation of the bracing; this problem had not resolved completely at the time of the latest follow-up. One patient who had had decompression had a transient palsy of the first sacral nerve; subjective strength returned completely during the next two months, but the ankle reflex remained absent.
A screw was found to have broken, at the five-year follow-up examination, in a patient who had an apparently solid fusion and an excellent clinical result; this was not classified as a complication.
The best method of operative treatment for adults who have grade-I or grade-II spondylolisthesis is not known. The treatment variables included the use of decompression, the type and location of the arthrodesis, the use of instrumentation, the type of graft, and the number of levels included in the arthrodesis. In the current study, I tried to focus on the utility of decompression and instrumentation, performed during an isolated single-level posterolateral arthrodesis, in cohorts of adult patients who smoked and did not smoke.
Wiltse and Hutchinson as well as Gobler and Wiltse recommended lateral intertransverse-process arthrodesis alone for adults who have back pain without sciatica, and decompression and arthrodesis for young adults who have sciatica and low-back pain. Other authors have distinguished between irritation of the nerve root without demonstrable neurological loss and serious neurological injury, with decompression recommended only for the latter9. However, Peek et al. reported good results after arthrodesis alone in adults who had a high-grade slip and severe sciatica. Similarly, Cheng et al. noted good results after anterior interbody arthrodesis without decompression in adults who had a low-grade slip and sciatica; the patients routinely returned to strenuous labor. Garreau de Loubresse et al. reported the results for forty-eight patients who had had an arthrodesis for the treatment of isthmic spondylolisthesis. Of the twenty-five patients who had had an arthrodesis alone, 8 per cent had pseudarthrosis compared with only 30 per cent of the twenty-three who had had arthrodesis and laminectomy. Relief of radicular symptoms was also better in the group treated without laminectomy.
There have been several trials comparing arthrodesis with instrumentation and that without instrumentation. Zdeblick, in a randomized, prospective study of patients who had isthmic spondylolisthesis, compared the results of rigid instrumentation, those of semirigid instrumentation, and those of arthrodesis without instrumentation (control group). He found a 100 per cent rate of fusion (thirty-seven of thirty-seven) in association with rigid instrumentation, an 89 per cent rate (thirty-one of thirty-five) in association with semirigid instrumentation, and an 81 per cent rate (forty-two of fifty-two) in association with no instrumentation. He also noted a strong correlation between smoking and successful fusion after lumbar arthrodesis in general. After management without instrumentation, 53 per cent (twenty-four) of the forty-five smokers compared with 70 per cent (fifty-five) of the seventy-nine non-smokers had fusion. After management with rigid instrumentation, 87 per cent of the smokers compared with 100 per cent of the non-smokers had fusion. McGuire and Amundson prospectively randomized adult men and women in the military who had isthmic spondylolisthesis to treatment with arthrodesis with or without instrumentation, along with a Gill laminectomy. There were six single-level arthrodeses in the group that was managed with instrumentation and ten in the group that was managed without instrumentation; of these, one and three, respectively, did not result in fusion. Despite the trend toward more non-unions in the group managed without instrumentation, the authors did not believe that instrumentation was clearly beneficial10. It is noteworthy that the number of patients who smoked in each group was not given and the group that was managed with instrumentation was biased to failure as substantially more multilevel arthrodeses were performed. In an uncontrolled, prospective trial involving seventeen adults who had had an arthrodesis with transpedicular fixation from the fifth lumbar to the first sacral vertebra for the treatment of isthmic spondylolisthesis, Ricciardi et al. reported two pseudarthroses and very good clinical outcomes relative to other studies.
In the present study, the addition of decompression to posterolateral arthrodesis without instrumentation in non-smokers conferred no advantage; in fact, it was associated with an increase in the rates of both pseudarthrosis and unsatisfactory clinical results. Similarly, the results after arthrodesis with instrumentation were better when decompression had not been done than when it had been performed. I did not randomize instrumentation; all patients who smoked were managed with it. However, instrumentation appears to have been beneficial in the groups managed with decompression; the rate of pseudarthrosis was higher in the patients who did not smoke (and thus were treated without instrumentation) and were managed with decompression than in those who smoked and were managed with decompression, whereas the results of previous studies suggested that there would be more non-unions in patients who smoked3,5,10,13,15,17,18. A clear advantage of instrumentation with arthrodesis in the treatment of low-grade isthmic spondylolisthesis has not previously been demonstrated, to my knowledge. Given the strong association between smoking and pseudarthrosis documented in the literature and the generic benefit of transpedicular instrumentation in increasing the rates of fusion, the ethics of denying instrumentation to smokers in a controlled clinical trial were questionable and influenced this study's experimental design. The need to maximize the initial rate of fusion is highlighted in this and other studies10,13 in which, despite successful revision arthrodesis, the clinical result could not be consistently improved after the initial failure.
I attempted to eliminate confounding factors by screening for degeneration of the intervertebral disc immediately cephalad to the slip. Despite this effort, one patient (in whom the fourth and fifth lumbar disc had appeared normal on preoperative magnetic resonance imaging) had symptomatic degeneration and deformity of the disc within five years after an arthrodesis with instrumentation. The possibility of increasing the rate of degeneration of adjacent discs with use of rigid instrumentation has been discussed previously13,15,17,18. Since the beginning of this study, positive findings on discography for degeneration of discs that appeared normal on magnetic resonance imaging have been reported8. Perhaps such discs should be evaluated more closely, with discography, particularly if they are in older patients, in order to avoid this complication.
Finally, the scores for pain, activity, use of medication, and satisfaction for the patients managed without decompression in the current study were very good at the time of the latest follow-up evaluation (a good or excellent result in eleven of the twelve patients managed without instrumentation and in all twelve of those managed with instrumentation). It is important to remember that I systematically chose patients who were more likely to have a successful outcome by limiting enrollment to those scheduled to have a single-level arthrodesis, those who had a normal cephalad disc, those who had a low-grade slip, and those who had no neurological deficits. Only six patients (14 per cent) had a Workers' Compensation claim—a low rate compared with those in other studies5,7,9,10. No patient had major motor loss or a cauda equina syndrome. Neurologically impaired patients may need decompression, and perhaps instrumentation should be used in all such patients regardless of whether or not they smoke.
There were several limitations to the current study. The use of instrumentation was not randomized, so the results of the comparison between the groups managed with and without instrumentation must be interpreted with caution. The numbers of patients who smoked and of men were higher than in the general population. This may be due the demographic bias at the Veterans Affairs Medical Center. Therefore, generalization of these data may not be reliable. Similarly, all of these operations involved teaching of residents and fellows, who participated in the care at all levels. To some extent, this may have biased referral patterns and selection of patients. Finally, it is worth re-emphasizing that a relatively small percentage of patients had a Workers' Compensation claim, and this may have influenced the clinical results and also may not be generalizable to practices with higher percentages of patients receiving such compensation.
In conclusion, the addition of decompression to arthrodesis with or without instrumentation for the treatment of low-grade isthmic spondylolisthesis in patients without a serious neurological deficit does not appear to improve the outcomes and may significantly increase the rates of pseudarthrosis and clinical failure.
Boos, N.; Marchesi, D.; Heitz, R.; and Aebi, M.: Traitement chirurgical des spondylolisthésis de l'adulte à déplacement modéré par fixation pédiculaire et fusion postéro-latérale. Rev. chir. orthop.,78: 228-235, 1992.78228
1992
[PubMed]
Carragee, E. J., and Kim, D. H.: A prospective analysis of MRI findings in patients with sciatica and lumbar disk herniation: correlation of outcomes with disk fragment and canal morphology. Orthop. Trans.,20: 74, 1996.2074
1996
Chang, P.; Seow, K. H.; and Tan, S. K.: Comparison of the results of spinal fusion for spondylolisthesis in patients who are instrumented with patients who are not. Singapore Med. J.,34: 511-514, 1993.34511
1993
[PubMed]
Cheng, C. L.; Fang, D.; Lee, P. C.; and Leong, J. C. Y.: Anterior spinal fusion for spondylolysis and isthmic spondylolisthesis. Long term results in adults. J. Bone and Joint Surg.,71-B(2): 264-267, 1989.71-B(2)264
1989
Deyo, R. A.; Cherkin, D. C.; Loeser, J. D.; Bigos, S. J.; and Ciol, M. A.: Morbidity and mortality in association with operations on the lumbar spine. The influence of age, diagnosis, and procedure. J. Bone and Joint Surg.,74-A: 536-543, April 1992.74-A536
1992
Garreau de Loubresse, C.; Bon, T.; Deburge, A.; Lassale, B.; and Benoit, M.: Posterolateral fusion for radicular pain in isthmic spondylolisthesis. Clin. Orthop.,323: 194-201, 1996.323194
1996
[PubMed]
Gobler, L., and Wiltse, L.: The Adult Spine: Principles and Practice, pp. 1655-1704. Edited by J. W. Frymoyer. New York, Raven Press, 1991.
Guyer, R. D., and Ohnmeiss, D. D.: Lumbar discography. Position statement from the North American Spine Society Diagnostic and Therapeutic Committee. Spine,20: 2048-2059, 1995.202048
1995
[PubMed]
Johnson, L. P.; Nasca, R. J.; and Dunham, W. K.: Surgical management of isthmic spondylolisthesis. Spine,13: 93-97, 1988.1393
1988
[PubMed]
McGuire, R. A., and Amundson, G. M.: The use of primary internal fixation in spondylolisthesis. Spine,18: 1662-1672, 1993.181662
1993
[PubMed]
Österman, K.; Schlenzka, D.; Poussa, M.; Seitsalo, S.; and Virta, L.: Isthmic spondylolisthesis in symptomatic and asymptomatic subjects, epidemiology, and natural history with special reference to disk abnormality and mode of treatment. Clin. Orthop.,297: 65-70, 1993.29765
1993
[PubMed]
Peek, R. D.; Wiltse, L. L.; Reynolds, J. B.; Thomas, J. C.; Guyer, D. W.; and Widell, E. H.: In situ arthrodesis without decompression for grade-III or IV isthmic spondylolisthesis in adults who have severe sciatica. J. Bone and Joint Surg.,71-A: 62-68, Jan. 1989.71-A62
1989
Ricciardi, J. E.; Pflueger, P. C.; Isaza, J. E.; and Whitecloud, T. S., III: Transpedicular fixation for the treatment of isthmic spondylolisthesis in adults. Spine,20: 1917-1922, 1995.201917
1995
[PubMed]
Soini, J.; Laine, T.; Pohjolainen, T.; Hurri, H.; and Alaranta, H.: Spondylodesis augmented by transpedicular fixation in the treatment of olisthetic and degenerative conditions of the lumbar spine. Clin. Orthop.,297: 111-116, 1993.297111
1993
[PubMed]
Stillerman, C.; Schneider, J.; and Gruen, J.: Evaluation and management of spondylolysis and spondylolisthesis. Clin. Neurosurg.,40: 384-415, 1993.40384
1993
[PubMed]
Wiltse, L. L., and Hutchinson, R. H.: Surgical treatment of spondylolisthesis. Clin. Orthop.,35: 116-135, 1964.35116
1964
[PubMed]
Yuan, H. A.; Garfin, S. R.; Dickman, C. A.; and Mardjetko, S. M.: A historical cohort study of pedicle screw fixation in thoracic, lumbar, and sacral spinal fusions. Spine,19(20S): 2279S-2296S, 1994.19(20S)2279
1994
Zdeblick, T. A.: A prospective, randomized study of lumbar fusion. Preliminary results. Spine,18: 983-991, 1993.18983
1993
[PubMed]