The Reamer/Irrigator/Aspirator (RIA; Synthes, West Chester, Pennsylvania) was initially developed to minimize the detrimental systemic physiological effects of pressurization during reaming of long-bone fractures1. However, with increased use and experimentation by surgeons, it has also been seen to have value as an autologous bone-graft harvesting device2. Passing the aspirated irrigant through a filtration system permits the reaming debris to be captured in a sterile fashion for later use as bone graft. The literature contains numerous reports regarding reaming debris and its inherent qualities that make it a potential source of autologous bone graft3-7. These studies have repeatedly shown the viability of osteocytes and osteoblasts in the reaming debris as well as the presence of growth factors known to participate in the various phases of fracture-healing and bone formation.
Autologous bone graft from the iliac crest continues to be the gold standard to which all other substitutes are compared. Donor site morbidity and limited volume (depending on the technique and location) continue to be the primary drawbacks of this method of harvesting autologous bone graft8-13. Use of the RIA as a tool for autologous bone-graft harvesting is attractive because of its potential to decrease donor site morbidity, pain, and wound complications, and to increase the volume of graft available to the surgeon if multiple extremities can be used in one surgical procedure.
If RIA bone-grafting can possibly supplant the iliac crest as a source of autologous bone graft, it is important to determine the biological activity and cellular composition of the reaming debris and how these compare with the properties of bone graft harvested from the iliac crest. The purpose of this study was to determine the histological characteristics, cellular content, and transcriptional expression profiles of bone-graft material obtained with use of the RIA and to compare these with the properties of the autologous cancellous bone graft obtained from the iliac crest of the same patient.
Patient Selection
All studies were approved by the institutional review boards of Tampa General Hospital and the University of South Florida Medical School, Tampa, Florida. All patients gave informed consent to have bone graft harvested from both the anterior iliac crest and the long bone of their choice with the RIA. Skeletally mature patients undergoing repair and reconstruction of a diaphyseal or metaphyseal nonunion of the tibia or femur requiring autologous bone graft were considered eligible and given the opportunity to participate in the study. Patients were excluded if they had a history of infection, osteomyelitis (acute or chronic), intramedullary nailing, malignancy, or a physiological condition affecting bone metabolism.
Iliac Crest Harvesting
A 3-cm incision was made 2 cm proximal to the anterior superior iliac spine to expose the underlying external oblique fascia, which was then elevated from the iliac crest in a subperiosteal fashion. The tensor fascia and iliacus muscles were not elevated away from the bone. A “trap-door” approach was used to gain access to the medullary cavity of the iliac crest by hinging the outer cortex of the top of the crest inward on a myofascial flap connected to the cortical fragment. Cancellous graft from between the inner and outer cortices of the ilium was harvested with a curet, and the cortical trap door was then replaced and the periosteum and soft tissue were repaired14.
Intramedullary Harvesting with Use of the RIA
The medullary cavity was opened with use of an entry reamer in the standard fashion for retrograde intramedullary nailing of the femur. A long ball-tipped guidewire was then passed into the medullary cavity to the far end of the canal. The selected reamer diameter was 3 mm greater than the estimated canal diameter at the isthmus, assessed fluoroscopically with use of a radiopaque ruler on both the anteroposterior and lateral projections. The RIA collected bone graft by means of the closed, sterile suction-irrigation apparatus connected to the reamer driveshaft and tube. The aspirated reaming debris was captured in a sterile filtration basket (Biomet, Warsaw, Indiana) for later retrieval. This technique has been previously described by Kobbe et al. and Quintero et al.15,16.
The choice of bone(s) for harvesting with the reamer was dependent on the volume of bone graft desired and the clinical scenario. If possible, the bone containing the nonunion was used for RIA harvesting if the fracture had initially been treated with a plate-and-screw construct. If the fracture had been treated initially with an intramedullary device, reaming debris was obtained from another long bone (the tibia or femur, either ipsilateral or contralateral) that had not been reamed previously. Only tibiae and femora were reamed to harvest bone-graft material.
Sample Handling
At the time of harvesting, a 1-cm3 sample of each type of bone graft was divided in half and each half was placed in a separate sterile specimen vial. One of the halves was flash-frozen in liquid nitrogen, and the other was placed in neutral buffered 4% paraformaldehyde fixative. Each sample was labeled with the location of the harvesting site and a patient identifier.
Demineralized Histology
The bone fragments designated for histological assessment were fixed, decalcified, and sectioned as described previously by Gerstenfeld et al.17.
Microarray Analysis of Gene Expression
The GeneChip Human Genome U133A 2.0 (Affymetrix, Santa Clara, California) was used for these studies. All microarray analyses were performed at the Boston University Microarray Resource Facility according to the GeneChip WT (Whole Transcript) manual. Total RNA was isolated separately from the iliac crest and RIA samples with use of procedures developed for murine bones18. One microgram of RNA was labeled and used for hybridization, and three chips were used for each patient to provide replicate analyses. The array data were output to a file in CEL format and summarized with use of Affymetrix Expression Console software (version 1.1). The RMA (Robust Multi-Array Analysis) algorithm19 was used to generate gene-level data.
Data Acquisition, Preprocessing, and Quality Control
Microarray data quality was assessed with use of methods implemented in the Affymetrix Expression Console. Data quality was assessed on the basis of the relative log expression (RLE) values of all probe sets, the normalized unscaled standard error (NUSE) values, and the area under the receiver operating characteristic (ROC) curve comparing signal values for positive and negative control probes. Genes with a false-discovery rate of >0.05 were excluded from further analyses. A combination of statistical methods (Partek Pro software; Partek, St. Louis, Missouri) was used to evaluate statistical significance for those genes that showed differential expression between the harvest sites. The mean expression of a gene at a harvest site was calculated by averaging the signal intensity across all three replicates of all patient samples, and the relative expression of the gene was calculated as the ratio of the mean expression at the two harvest sites. Genes that showed at least a ±2.0-fold difference in expression were considered to be differentially expressed, whereas those showing reproducible expression but expressed at lower levels (less than a twofold increase) were considered to be comparably expressed.
The Ingenuity Pathway Analysis software and the Ingenuity Knowledge Base were used to identify the biological or molecular gene network that was associated with those genes that were differentially and comparably expressed between the two harvest sites. Only those groups having an enrichment score of p ≤ 0.05 were considered for this analysis.
Quantitative Reverse Transcription-Polymerase Chain Reaction (qRT-PCR) Analysis
Expression of messenger RNA (mRNA) for selected growth factors known to participate in the various phases of fracture repair and for cell markers of vascular and osseous lineages was analyzed with use of qRT-PCR. Pooled mRNAs from each patient sample were run in triplicate with use of PCR probes from Applied Biosystems (Foster City, California). All mRNA levels were normalized to that of β-actin, and each analysis was run in triplicate. Quantitation was based on the fractional cycle number at which the fluorescence passed the fixed threshold (CT value), and the ΔΔCT method was used to compare the RIA and iliac crest samples as described by Applied Biosystems.
Source of Funding
This study was funded by a research grant supplied by the Foundation for Orthopaedic Trauma (FOT). The FOT did not have a role in the design, execution, or publication of this study and is not associated with Synthes.
Autograft materials are usually well incorporated and less extensively resorbed over time than allografts because of their nonimmunogenic characteristics, the presence of viable cells, and their greater osteoinductivity and osteoconductivity20. Additionally, the use of autologous bone graft eliminates the potential of disease transmission, which can occur with cadaveric allografts21-24. Although autologous bone graft from the iliac crest continues to be the gold standard, it has the distinct disadvantages of donor site morbidity and limited volume (approximately 20 to 50 cm3). The reported rate of major complications following iliac crest bone grafting ranges as high as 25%, and the rate of minor complications ranges as high as 39%9,25. Goulet et al. reported major complications in 2.4% of patients and minor complications in 21.8%. Pain was the most frequent complaint in the first six months after surgery (in 38% of patients), with 19% of patients continuing to report pain at more than two years postoperatively12.
The morbidity associated with iliac crest bone-graft harvesting makes alternative methods of obtaining autograft attractive to orthopaedic surgeons and patients. Currently, a number of reports exist on small cohorts of patients in whom autologous bone graft obtained with use of the RIA was utilized to successfully repair nonunion defects26-28. The use of reaming debris as a source of autologous bone graft is further supported by a number of studies showing that bone material collected in this manner retains osteocyte viability as well as the presence of biochemical proteins known to stimulate bone-healing4,5,7,29.
Although the results of one other study have suggested that levels of VEGF, platelet-derived growth factor (PDGF), and BMP2 (among others) were higher in reaming debris compared with iliac crest bone graft, the sample size of that study was small, and the iliac crest samples were not obtained from the same patients as the reaming debris6. That study does, however, lend support to our findings, which directly compared bone tissues from the two sites of the same patient. Additionally, our PCR analysis demonstrated the potential osteogenic and osteoinductive properties of the reaming debris, corroborating other work performed by Frölke et al., Tydings et al., and Schmidmaier et al.3,4,6,7. Like the 2006 study by Schmidmaier et al.6, our study showed that angiogenic factors were decreased and growth factors were increased in the RIA samples. Finally, our study corroborated the findings of Tydings et al.7 and Hoegel et al.5 that viable osteocytes and osteoprogenitor cells are present in the reaming material.
It is of considerable importance to note that our analysis suggests that the RIA samples had a greater enrichment in mesenchymal stem cells than the iliac crest bone graft on the basis of both greater expression of specific human mesenchymal stem cell markers such as CD14630 and greater expression of BMP receptors31. Such findings are also consistent with those of Cox et al.32, who showed that RIA samples had a greater percentage of colony-forming fibroblastic cells. Other characteristics of our transcriptional profiling data that suggest that RIA tissues have greater regenerative characteristics compared with iliac crest bone graft include the similarity of the gene expression profiles in the RIA tissues to expression profiles in experimental models of bone-healing after marrow reaming in rats33 and during fracture-healing in mice31.
Finally, it is notable that this is one of the first transcriptional profiling studies to date that has been carried out in the appendicular skeleton of normal, healthy human subjects. One of the most intriguing features of our data in this regard is the reproducible differences between the complement of differentially expressed genes that were seen in tissue retrieved from the intramedullary space and those in bone harvested from the iliac crest.
In summary, reaming debris obtained with the RIA represents a potential alternative source of autologous bone graft. Although further studies will be needed to confirm the actual biological activity of the RIA graft material, use of reaming debris obtained with the RIA appears to be justified.