Anterior cervical decompression and fusion (ACDF) is a commonly performed procedure with a good outcome when fusion is achieved. When performing ACDF, the surgeon must weigh pros and cons of graft choices to aid in fusion. The greatest benefit of autograft over allograft is increased fusion rate,1,2 especially for patients undergoing multilevel procedures and for patients who smoke.3 Failure of solid fusion results in significant morbidity.4 The major disadvantage of autograft is donor site morbidity. Much literature exists on acute morbidity.5–11 Fewer studies have evaluated graft site morbidity greater than 1 year.
One of the authors (J.S.T.) has been performing ACDF with anterior iliac crest harvest for more than 20 years. Originally, the author used a tricortical technique and noticed unacceptable morbidity. To reduce morbidity while maintaining the benefit of increased fusion rate with autograft, the author switched to a pilot hole burr technique for cancellous extraction. The objective of this study was to evaluate long-term anterior iliac crest graft site morbidity in ACDF using a pilot hole burr technique.
Materials and Methods
Institutional review board approval was obtained. Patients undergoing ACDF performed by a single surgeon (J.S.T.) at 2 hospitals were identified using Current Procedural Terminology codes 22845 and 22846. Patients who underwent ACDF with anterior iliac crest harvest using a pilot hole burr technique from April 1, 2012, to January 1, 2017; were older than 21 years; and had 6-month to 1-year follow-up were included.
Primary chronic morbidity outcomes were determined using a phone questionnaire. Questions focused on presence or absence of graft site pain, numeric pain rating scale score, medication to control pain, and impact of pain on activity. Contact was attempted on 3 occasions—2 weekdays and 1 weekend day. Acute morbidity was determined through chart review of operative reports, progress notes, phone encounters, and pain diagrams at 1 month, 3 months, 6 months, and 1 year.
The distal extent of the incision was marked 2 fingerbreadths proximal to the anterior superior iliac spine along the iliac crest to avoid injury to the lateral femoral cutaneous nerve. The mark was continued proximally in line with the iliac crest for approximately 3 cm (Figure 1). The skin incision was made with a No. 10 blade scalpel. Electrocautery was used to dissect down to the level of the abdominal fascia. Weitlaner retractors were positioned to mobilize the external oblique musculature. The deep abdominal fascia was peeled with subperiosteal dissection off the iliac crest to expose the underlying cortical bone (Figure 2). A high-speed, size 4, acorn burr was used to create a distal pilot hole at least 2 cm proximal to the anterior superior iliac spine. A proximal pilot hole approximately 2 cm proximal to the distal pilot hole was created. The 2 pilot holes were connected, resulting in a smooth-edged defect in the iliac crest (Figure 3). A curette was used to extract underlying cancellous bone. For 1-, 2-, and 3-level fusions, approximately 1 cm3, 2 cm3, and 3 cm3 of bone were harvested, respectively.
Incision drawn for anterior iliac crest harvest using a pilot hole burr technique. The anterior superior iliac spine is marked on the left side of the image. The incision was made 2 fingerbreadths proximal to the anterior superior iliac spine to avoid injury to the lateral femoral cutaneous nerve and extended 3 cm proximal along the iliac crest.
Iliac crest exposed with the initial pilot hole in the proximal (right) aspect of the wound.
After the 2 pilot holes were connected with the acorn burr, underlying cancellous bone was identified and harvested with a curette. The quantity of harvested bone was dependent on the number of cervical levels fused.
The iliac crest was copiously irrigated with sterile saline. Surgifoam (Ethicon, Sarasota, Florida) and thrombin and then bone wax were placed in the cortical defect (Figure 4). The overlying fascia was closed with 1-0 Vicryl sutures (Ethicon). Ten milliliters of 0.5% bupivacaine was injected into the graft site. The subcutaneous tissue and skin were closed with 2-0 and 3-0 Vicryl sutures, respectively.
Defect in the iliac crest filled with bone wax.
The operative approach to the cervical spine was through a standard left-sided collar incision. The Smith–Robinson technique was used for ACDF. An inter-vertebral polyetheretherketone cage was filled with bone harvested from the iliac crest and placed in the intervertebral disk. An anterior cervical plate with screws was placed in standard fashion. Remaining cortical bone was placed anterior to the intervertebral cage just deep to the plate.
Data were entered into Excel (Microsoft, Redmond, Washington) and then imported into SPSS version 24.0 software (IBM, Armonk, New York). The primary outcome was pain assessed via a 0 to 10 ordinal scale. An exploratory analysis was performed to identify characteristics associated with chronic pain. Kendall's tau was used to compare the levels fused between dichotomous chronic pain groups. Fisher's exact test was used for comparisons involving sex and surgical complication. An independent samples Student's t test was used to compare mean ages. The percentages of patients experiencing crest site pain were presented via column bar graph with 95% confidence intervals at 1 month, 3 months, 6 months, 1 year, and the time of survey completion. For the time of survey completion category, the authors included those patients with current pain, as this was thought to most closely replicate pain recorded with pain diagrams. Categorical descriptive summaries used frequencies and percentages to further describe the study cohort and any morbidity associated with surgery. All testing was 2-sided, with P<.05 considered statistically significant.
During the 57-month study period, 140 patients undergoing ACDF with anterior iliac crest autograft using the pilot hole burr technique met inclusion criteria (Table 1).
Baseline Characteristics of Patients Undergoing Anterior Iliac Crest Autograft Harvest
A total of 106 patients (76%) completed the phone survey (Table 2). Mean follow-up was 38.9 months. At the time of the survey, 2 patients had died of unrelated causes. One patient refused to participate. Eleven patients (10.4%) experienced any type of graft site pain. Six patients (5.7%) reported current pain. Two patients (1.9%) reported current and constant graft site pain. Nine patients (8.5%) reported intermittent pain. The average numeric pain rating scale score for those reporting pain was 3.8 of 10. The average graft site numeric pain rating scale score for all survey participants was 0.25 of 10. No patients were taking narcotics for pain. Six of 11 patients took acetaminophen, ibuprofen, or naproxen. The remaining 5 patients took no medication for graft site pain. Two patients (1.9%) reported functional impairment, difficulty with ambulation, and difficulty with activities of daily living secondary to graft site pain. Multivariate analysis suggested no statistical significance of number of levels fused, age, sex, or acute graft site complications for chronic graft site pain (Table 3).
Chronic Morbidity of Patients With Anterior Iliac Crest Autograft Harvest
Multivariate Analysis of Chronic Donor Site Pain
Acute morbidity was evaluated (Table 4). No intraoperative complications were encountered. Nine patients (6.4%) experienced superficial infection that resolved without further complication or surgical intervention. Three patients (2.1%) experienced hematoma. All hematomas resolved without further complication or surgical intervention. One patient (0.7%) experienced meralgia paresthetica. This patient underwent local anesthetic and corticosteroid injection of the lateral femoral cutaneous nerve with 50% resolution of symptoms.
Acute Morbidity of Patients Undergoing Anterior Iliac Crest Autograft Harvest
Pain diagrams were evaluated from office visits at 1 month, 3 months, 6 months, and 1 year. Follow-up rates at these time periods were 96%, 88%, 81%, and 81%, respectively. Graft site pain was reported by 51 (38.1%), 12 (9.1%), 3 (2.4%), and 6 (5.3%) patients at 1 month, 3 months, 6 months, and 1 year, respectively (Figure 5). There was significant reduction in graft site pain between 1 month and 3 months. There was a trend but no significant reduction in graft site pain from 3 months to 6 months. There were no significant changes in incidences of graft site pain from 6 months to 1 year, or from 1 year to phone survey follow-up. Few patients experienced graft site pain for more than 3 months.
Reporting of pain at 1 month, 3 months, 6 months, and 1 year through pain diagrams. Pain at follow-up (average, 38.9 months) by survey reported as the percentage of patients with current pain at the time of questionnaire. Error bars represent 95% confidence intervals.
To the authors' knowledge, this is the first study to evaluate morbidity of anterior iliac crest harvest using a pilot hole burr technique. It is also one of the few studies to evaluate morbidity of anterior iliac crest harvest in spine surgery greater than 1 year. The results of this study indicate low long-term morbidity and low functional impact of morbidity using a pilot hole burr technique for anterior iliac crest harvest.
In this study, 11 patients (10.4%) had chronic graft site pain. Two patients (1.9%) experienced constant graft site pain, while 9 patients (8.5%) experienced intermittent pain. Average numeric pain rating scale score of all patients at the time of survey completion was 0.25 of 10. Anterior iliac crest graft site pain at greater than 1 year using traditional techniques has been reported, with an incidence of 20% to 48%.10–14 Summers and Eisenstein12 performed a retrospective evaluation at an average 5-year follow-up from anterior lumbar fusion. They reported a chronic acceptable pain rate of 17% and a chronic significant pain rate of 31%. A high pain rate in that study may have been influenced by close vicinity of a lumbar fusion to the harvest graft site. For a more direct comparison, Silber et al10 retrospectively evaluated graft site morbidity of 134 patients undergoing ACDF. A chronic pain rate of 26.1% with an average visual analog scale score of 3.8 of 10 was identified at 4-year follow-up.
In terms of functional assessment, graft site pain has been shown to diminish quality of life and impact activities of daily living.5,10,15 Silber et al10 reported graft site–induced impairment with ambulation (12.7%), recreation (11.9%), work (9.7%), sexual activity (7.5%), and household chores (6.7%). In the current study using the pilot hole burr technique, the impact of graft site pain on function was reported by 2 (1.9%) patients. These patients expressed diminished abilities with ambulation, household chores, recreation, sexual activity, and activities of daily living. No patient in this study was taking narcotic medication secondary to graft site pain. Of the 11 patients experiencing chronic pain, 3 patients were taking acetaminophen, 1 patient was taking acetaminophen and naproxen, 2 patients were taking ibuprofen, and 5 patients were taking no medications for graft site pain.
One patient in this study experienced meralgia paresthetica, a complex set of pain, numbness, tingling, and paresthesias in the anterolateral thigh. Meralgia paresthetica can occur spontaneously, but it is likely secondary to iatrogenic injury to the lateral femoral cutaneous nerve after iliac crest harvest. Meralgia paresthetica is a known complication of anterior iliac crest harvest that generally occurs at a low rate.6,16 Nonoperative management for meralgia paresthetica has a high rate of success.17,18 The patient in this study who experienced meralgia paresthetica had 50% relief with local anesthesia and corticosteroid injection and decided not to undergo further operative intervention based on level of symptoms.
In this study, there was no impact of sex, age, number of levels fused, or acute graft site complications on chronic pain. The major implications of these findings are that multilevel fusions requiring greater bone and acute complications such as stitch abscess and hematoma may not impact chronic morbidity at the graft site.
Although not the primary goal of this study, acute morbidity of anterior iliac crest harvest is significant and should be considered when deciding on autograph vs allograft. Acute morbidity for the pilot hole burr technique was consistent with that reported in the literature.
The incidence of acute pain at 1 month has been reported to be between 10% and 90%.5,6 The incidence of early graft site pain, however, tends to diminish with time.5,9 This finding was consistent with the current study. Pain at 1 month and 3 months was 38.1% and 9.1%, respectively, resulting in a statistically significant reduction in pain between these time points. Results suggest few patients experience graft site pain after 3 months, but many patients with pain at 3 months may expect to have some form of chronic discomfort. Local analgesia has been shown to reduce immediate perioperative pain and graft site pain for up to 20 weeks.19,20 In this study, 10 mL of 0.5% bupivacaine was injected into the bone graft site, which may have contributed to a reduction in short-and intermediate-term pain rates.
The rate of superficial infection with traditional techniques ranges from 1% to 7%.5–10 The pilot hole burr technique in this study resulted in a superficial infection rate of 6.4%. These resolved in all cases with oral antibiotics and local wound care. The rate of deep infection requiring operative intervention is reportedly 1% to 2%.6,10 No deep infections were encountered in this study.
The rate of hematoma for traditional techniques ranges from 0% to 3%.6–8,10 The rate of hematoma in this study was 2.1%. All hematomas were treated conservatively and resolved over time. The authors did not encounter rarer complications such as iliac crest fracture, herniation, or bowel perforation. The authors believe that in addition to using the pilot hole burr technique, small incisions, careful dissection, limited retraction of soft tissues, prophylactic antibiotics, Surgifoam and thrombin, and local anesthetic all contribute to a reduction of acute and chronic morbidity.
The importance of this and related studies lies in the decision of graft use in ACDF. Morbidity of anterior iliac crest harvest must be weighed against the risk of pseudarthrosis, which is dependent on type of graft used. Pseudarthrosis, or failure of fusion, may be the most common complication of ACDF,21 with rates ranging from 0% to 46%.1,2,22 Pseudarthrosis is frequently symptomatic4 and has the potential to diminish outcomes with recurrence of presurgical symptoms, adjacent segment disease, or progression of kyphosis. The impact of graft type on pseudarthrosis has to do with graft biomechanical properties. Autogenous bone retains the presence of live osteogenic cells. Allograft has osteoconductive and osteoinductive properties but is not osteogenic.23
Many studies have clearly shown the superiority of autograft in clinical application. Shriver et al1 performed a meta-analysis and reported a pooled pseudarthrosis rate of 4.8% with allograft vs 0.9% with autograft. Benefits of autograft appear to be greatest for smokers and for patients undergoing multilevel fusion3 but may extend to 1- and 2-level ACDFs.1,2 In summary, morbidity at the harvest site must be weighed against the reduction in pseudarthrosis using autograft. Autograft may become an option for some surgeons who would otherwise use allograft secondary to concerns over harvest site morbidity.
Limitations to this study included those inherent to retrospective reviews. As a retrospective case series, this study did not directly compare traditional or alternative harvest techniques with the pilot hole burr technique. Data from prior studies were included in an attempt to provide a comparison. The questionnaire was subject to information and selection bias. Variations in response may occur based on verbal presentation of a given questionnaire. The authors attempted to minimize variations by having a formal script for presentation and a single individual completing phone surveys. The surgeon performing the anterior iliac crest autograft using a pilot hole burr technique has been practicing for more than 20 years. At that point in one's career, it is likely that surgical technique would show reduced morbidity. To further evaluate application of the burr down technique broadly, additional high-quality randomized control trials would be beneficial.
Anterior cervical decompression and fusion using autologous graft results in increased fusion rates, especially for patients undergoing multilevel cervical fusion and for patients who smoke. Iliac crest harvest, however, has been associated with significant morbidity. The pilot hole burr technique in this study resulted in low long-term morbidity and low impact of morbidity on function. There was no significant influence of age, sex, number of levels fused, or acute graft site complications on chronic graft site pain. The pilot hole burr technique may offer an alternative to traditional methods for those wishing to use autologous graft in ACDF. Prospective studies would be useful to further evaluate the pilot hole burr technique as a potential method of autograft harvest in ACDF.