Autogenous iliac crest has been the gold-standard bone graft in spinal fusion surgery. Despite its successful use for spinal fusion, the harvest of autogenous iliac crest bone graft has been hampered by the donor site morbidity that frequently results. The complication rate following autogenous iliac crest bone graft harvest has been reported to be from 2.8% to 39%.1-11
Several studies have found that a significant number of patients (15-39%) experience long-term pain at 2 years postoperatively.2,4-9,11 Significantly, the majority of studies are retrospective, and it is often noted that the follow-up with regard to pain assessment may have been incomplete and inadequate.1,2,4,5,9-12 One study noted that the patient’s report of pain on survey was not noted in the patient chart,12 and another found that the surgeon’s perception of the patient’s pain was far less than that which the patient reported to an independent observer.7 Still, most studies have relied on chart review to determine if patients had pain.
The current study reports the results of data collected as part of a prospective multicenter study that evaluated iliac crest harvest site pain in terms of both intensity and frequency together with the assessment of the patient’s perception of the graft site appearance. This review of prospective data was intended to clarify the actual incidence and persistence of postoperative response to the iliac crest graft harvest site in patients undergoing anterior iliac crest bone graft harvest for anterior lumbar interbody fusion (ALIF).
MATERIALS AND METHODS
A total of 208 patients had single-level ALIF as treatment of symptomatic degenerative disc disease. Patients had at least 6 months of nonoperative treatment prior to surgical intervention, including physical therapy, medications, braces, chiropractic care, bed rest, spinal injections, or exercise programs. They were extracted from four prospective multicenter randomized controlled clinical trials. The goal of these studies was to evaluate the efficacy of bone morphogenic protein (BMP) versus autologous bone graft for ALIF using threaded cylindrical cages. The four studies are described below:
- Food and Drug Administration (FDA)-approved InFUSE Bone graft (Medtronic Sofamor Danek, Memphis, TN)/bone dowel pivotal Investigational Device Exemptions (IDE) study; 55 patients with InFUSE Bone graft/bone dowel, 30 patients with bone dowel/autologous bone graft; 13 investigational sites.
- FDA-approved InFUSE Bone graft/INTERFIX (Medtronic Sofamor Danek) device pilot IDE study; 25 patients with InFUSE Bone graft/INTERFIX device, 20 patients with INTERFIX/autologous bone graft; 7 investigational sites.
- FDA-approved InFUSE Bone graft/LT-CAGE (Medtronic Sofamor Danek) device pivotal IDE study; 143 patients with InFUSE Bone graft/ LT-CAGE device, 136 patients with LT-CAGE/ autologous bone graft; 16 investigational sites.
- FDA-approved InFUSE Bone graft/bone dowel pilot IDE study; 24 patients with InFUSE Bone graft/bone dowel, 22 patients with bone dowel/autologous bone graft; 5 investigational sites.
All patients included in the current study were enrolled in the control group of the above clinical trials. They received two interbody implants (LT-CAGE, INTERFIX, or threaded bone dowels) at one low lumbar level (L5-S1 or L4-L5) and underwent iliac crest graft harvesting. The patients in the experimental study group received the implants with only the InFUSE bone graft. Patients enrolled in the study had symptomatic degenerative disc disease diagnosed by symptoms of intractable leg and/or back pain with positive diagnostic imaging findings.
Patients were followed at hospital discharge, 6 weeks, 3 months, 6 months, 12 months, and 24 months. The patents included 110 women and 98 men, ranging in age from 19 to 81 years, with a mean age of 42.3 years.
All patients underwent the ALIF procedure through an open approach, via a retroperitoneal or transperitoneal exposure. Patients were placed in the supine position on the operating room table. A vertical or transverse incision was made over the lumbosacral spine. A separate incision was made for the bone graft harvest over the iliac crest. For retroperitoneal exposure, the peritoneal sac was mobilized through blunt dissection off the abdominal wall and retracted past midline. The bifurcation of the great vessels was identified. The soft tissue elements overlying the L4-L5 or L5-S1 disc space were bluntly mobilized. Care was taken to avoid the use of electrocautery during the dissection. The great vessels were mobilized, exposing the anterior surface and lateral borders of the disc space. For the transperitoneal approach, the abdominal cavity was entered through a midline incision. The small bowel was bluntly retracted from the surgical field. The great vessels and the sacral promontory were identified. At the L5-S1 disc space level, the soft tissues were divided in the midline between the bifurcations of the iliac vessels. At the L4-L5 disc space level, the soft tissues were reflected off the left lateral portion of the inferior vena cava and retracted laterally.
Most of the iliac crest graft harvests were performed on the anterior iliac crest evenly dispersed between left and right sides. Cancellous bone was taken as described by Ebraheim.13 The bone graft donor site was standardized in this prospective study by protocol and meetings with all 13 surgeons of this multicenter project. A separate incision was made over the anterior iliac crest beginning 2 cm posterior to the anterior superior iliac spine and carried posteriorly. A window was made in the cortex of the cephalad aspect of the iliac crest, and a curette harvested the cancellous bone between the iliac tables. No significant defect in the outer iliac crest was created, and no backfill material was implanted. Drains were not routinely used. Among the 208 patients, only three grafts (1.4%) were taken posteriorly and six grafts (2.9%) were tricortical. The patients evaluated graft site appearance at each follow-up interval.
The pain evaluation score is composed of two standard Visual Analog Scale (VAS) scores reporting both pain intensity and pain frequency. The scores are recorded on a scale of 0 to 10, with “0” as “no pain” and “10” as “bad as it could be” for pain intensity and “0” as “none of the time” and “10” as “all of the time” for pain frequency. The total pain score combined the pain intensity and the pain frequency scores. This score consisted of the sum of the Visual Analog Pain Scale and the frequency of the pain. The highest pain score is thus 20 (the highest intensity pain occurring all of the time).
The patient donor site pain was measured with both pain intensity and frequency scores at each follow-up period.
The evolution of the pain during the total follow-up period was statistically analyzed using the Fisher exact test and χ2 test with level of significance at P < 0.05.
The pain evaluation was independent of the surgical investigator’s evaluation, as the patients completed the follow-up forms independently and without assistance.
A total of 208 patients had an autogenous iliac crest bone graft harvested for placement in ALIF threaded cylindrical cages. Prospectively acquired data were available on 202 patients at discharge from the hospital. Only 2 patients (1%) of 202 had no bone graft donor site pain. Improvement occurred at the 6-week postoperative interval with 34 (17%) having no bone graft donor site pain in 199 completed questionnaires. By 3 months, 85 (43%) of 199 patients had no pain. Remarkably, 41% had some iliac crest bone graft donor site pain at 6 months in 192 prospectively completed data sheets. At the 1-year postoperative interval, 33% of 168 patients still had pain. At the final 2-year follow-up visit, 141 of the original 208 study patients were available for review. A consistent 31% had some degree of bone graft donor site pain (Fig. 1).
The appearance of the graft site was good at hospital discharge in 49% of patients, fair in 40%, and poor in 11%. At the 6-week postoperative interval, the graft site bothered patients very little or not at all (good) in 69%, some (fair) in 27%, and very much (poor) in 5% of patients. At 3 months, the prospective data rated the graft site as good in 75%, fair in 24%, and poor in 2%. At the 6-month interval, patients rated their graft site as good in 82%, fair in 15%, and poor in 3%. This rating remained consistent at the 1-year postoperative interval, with 82% of patients good and 19% fair or poor. At the final 2-year follow-up visit, 141 of the original 208 study patients were available for review. Consistently, 84% rated the appearance of the graft site as good, and 16% felt it was fair or poor (Fig. 2).
For patients who had bone graft donor site pain, a pain score was calculated at each time. At discharge from the hospital postoperatively, the graft site pain score for the 202 patients ranged from 0 to 20 with a mean of 12.8 (SD 4.5). At the 6-week postoperative interval, the mean pain score decreased to 7.3 (SD 5.6; range 0-20). At 3 months, the mean pain score decreased to 3.8 (SD 4.6; range 0-18). The mean pain scores were 2.9, 2.4, and 1.8 at 6, 12, and 24 months, respectively. At all time intervals, P values from t tests comparing the mean with 0 were <0.001 (Fig. 3).
An analysis of total pain scores (combining intensity and frequency) of ≥10 (maximum possible 20) was performed at all time intervals. For example, a patient who had bone graft donor site pain as bad as it could be (intensity VAS 10) half of the time (frequency VAS 5) would have a total score of 15. At discharge from the hospital, 80% (161/202) had a pain score of ≥10 (range 10-20). At 6 weeks postoperatively, the percentage of patients with a pain score of ≥10 decreased to 35% (70/199; range 10-20). At 3 months, 14% (28/199) had a pain score of ≥10 (range 10-18). At 6, 12, and 24 months postoperatively, 13% (25/192), 10% (17/168), and 7% (10/141) had total pain scores of ≥10, respectively.
Only 3 patients of the total 208 had the bone graft harvested from the posterior iliac crest. All of the rest had an anterior iliac crest bone graft donor site. These three posterior harvest sites did not show any significant deviation from the mean of the anterior grafts and were thus included in the analysis. Only six of the anterior iliac crest bone grafts were harvested with a bicortical technique; the rest were unicortical. These six bicortical patients did not demonstrate significant differences in outcomes and were thus included in the analysis. The anterior unicortical bone grafts were harvested relatively equally between the left and right iliac crests and thus were not separated for the analysis.
Autogenous iliac crest bone graft has long been the gold-standard graft for spinal fusion. Donor site morbidity has been well established to be a problem with autogenous iliac crest bone graft use.1-12
Several issues addressed in the current study are significant and confirm prior study findings. Most of the morbidity associated with autogenous iliac crest bone graft involves graft site pain. It has been demonstrated that the patient’s perception of pain is different from that reported by the physician.7 Greater pain response is attained when an independent examiner interviews the patient.7 Patients who have iliac crest graft harvested for degenerative disc disease and lumbar reconstructive procedures appear to have greater pain postoperatively.3 Several studies that attempted to evaluate donor site pain demonstrated a consistent 31-39% rate of chronic pain at 2-10 years.2,4,6-8,11
Visual Analog Pain Scales are valid functional outcome tools. They have been commonly used in clinical studies since the early 1970s. VASs are more sensitive to change than are Verbal Rating Scales and so require smaller sample sizes in evaluative studies.14 VAS scores are also very reproducible, with correlations of 0.99.15 VASs provide the patients with a robust, sensitive, reproducible method of expressing pain severity. Results correlate well with other methods of measuring pain. The method is applicable to all patients regardless of language and can be used by children aged ≥5 years.16 VASs give more sensitive and precise measurements than descriptive pain scales.17
This prospective study evaluated donor site pain at each postoperative interval. Patients completed the pain intensity and frequency as well as the donor site appearance surveys unassisted by the physician. All patients were treated for degenerative disc disease with all but three grafts being harvested from the anterior iliac crest. Only six grafts were tricortical. Virtually all of the grafts in this study were harvested from the anterior iliac crest site, thus eliminating the possibility that the persistent pain was coming from a site other than the iliac crest, such as the sacroiliac joint, lumbar spine, or residual radiculopathy.
The results found in this prospective study support findings in previous retrospective studies where patient pain response data were collected independently of the operating physician. In this study, 31% of patients reported persistent donor site pain at 2 years postoperatively. Several studies have demonstrated a similar percent donor site pain response at long-term follow-up.2,4,5-8,11 In addition, 16% of patients in this study reported fair or poor appearance of their graft site at 24 months postoperatively. It should be noted, however, that although 31% of patients reported pain at 24 months, the clinical significance of this finding is not certain. It is unknown whether this pain was significant enough to impact or restrict daily activities or just cause occasional discomfort. Even so, the fact that symptoms persisted at a site that was asymptomatic prior to surgery should be of concern to the surgeon.
Another shortcoming of this study is that the amount of bone harvested from the iliac crest was not measured or recorded. It is therefore not possible to determine if the amount of bone graft taken affected the patient’s pain response.
Although it is intuitive to add a frequency measure to the standard Visual Analog Pain Score to determine how often the most intense pain occurs, this functional outcome tool has not been validated. Unfortunately, no validated instrument is available to provide these data.
Because this is a multicenter study, it is possible that the bone graft harvest protocol was not followed exactly by all of the surgeons.
Analyzing the outcome of one wing of the study group from another study with different aims and objectives always has a disadvantage. The current article is not comparable with a prospective outcome study with the primary goal to determine the effect of harvesting iliac crest bone graft. The primary studies were intended to compare the results of a bone graft substitute with those of autologous bone graft. One of the hypotheses of the primary studies was that the disadvantage of harvesting autologous iliac crest bone graft might be avoided when a bone graft substitute was used. It is therefore possible to have a bias toward overestimation of the disadvantages of harvesting autologous iliac crest bone graft.
Persistent donor site pain remains a problem with harvest of autogenous iliac crest bone graft for spinal fusion. This prospective study, the first such study reported for iliac crest harvest for ALIF, confirms that donor site pain remains a significant postoperative management problem when harvested in patients undergoing lumbar spine surgery for degenerative disc disease.
1. Ahlmann E, Patzakis M, Roidis N, et al. Comparison of anterior and posterior iliac crest bone grafts in terms of harvest-site morbidity and functional outcomes. J Bone Joint Surg Am
2. Banwart JC, Asher MA, Hassanein RS. Iliac crest bone graft
harvest donor site
morbidity. A statistical evaluation. Spine
3. Fernyhough JC, Schimandle JJ, Weigel MC, et al. Chronic donor site
pain complicating bone graft harvesting from the posterior iliac crest for spinal fusion. Spine
4. Frymoyer JW, Hanley E, Howe J, et al. Disc excision and spine fusion in the management of lumbar disc disease. A minimum ten-year follow-up. Spine
5. Goulet JA, Senunas LE, DeSilva GL, et al. Autogenous iliac crest bone graft
and functional assessment. Clin Orthop.
6. Hacker RJ. A randomized prospective study of an anterior cervical interbody fusion device with a minimum of 2 years of follow-up results. J Neurosurg (Spine 2).
7. Heary RF, Schlenk RP, Sacchieri TA, et al. Persistent iliac crest donor site
pain: independent outcome assessment. Neurosurgery
8. Robertson PA, Wray AC. Natural history of posterior iliac crest bone graft
donation for spinal surgery: a prospective analysis of morbidity. Spine
9. Sawin PD, Traynelis VC, Memezes AH. A comparative analysis of fusion rates and donor-site morbidity for autogeneic rib and iliac crest bone grafts in posterior cervical fusions. J Neurosurg
10. Schnee CL, Freese A, Weil RJ, et al. Analysis of harvest morbidity and radiographic outcome using autograft for anterior cervical fusion. Spine
11. Skaggs DL, Samuelson MA, Hale JM, et al. Complications
of posterior iliac crest bone grafting in spine surgery in children. Spine
12. Arrington ED, Smith WJ, Chambers HG, et al. Complications
of iliac crest bone graft
harvesting. Clin Orthop.
13. Ebraheim NA, Elgafy H, Xu R. Bone-graft harvesting from iliac and fibular donor sites: techniques and complications
. J Am Acad Orthop Surg
14. McDowell I, Newell C. Measuring Health. A Guide to Rating Scales and Questionnaires
. 2nd ed. New York: Oxford University Press; 1996: 343.
15. Scott J, Huskisson EC. Vertical or horizontal visual analogue scales. Ann Rheum Dis
16. Huskisson EC. Measurement of pain. J Rheumatol
17. Jensen MP, Karloy P, Braver S. The measurement of clinical pain intensity: a comparison of six methods. Pain