Lattyak BV, Maas CS, Sykes JM. Dorsal Onlay Cartilage AutograftsComparing Resorption in a Rabbit Model. Arch Facial Plast Surg. 2003;5(3):240-243. doi:10.1001/archfaci.5.3.240
From the Division of Facial Plastic Surgery, Department of Otolaryngology–Head and Neck Surgery, University of California, San Francisco (Drs Lattyak and Maas); and Department of Otolaryngology–Head and Neck Surgery, University of California, Davis, Sacramento (Dr Sykes). Dr Lattyak is currently in private practice in Paget, Bermuda.
Objective To compare the resorption characteristics of dorsal onlay cartilage autografts from the septum, auricle, and rib.
Methods Fourteen New Zealand white rabbits underwent harvesting of equal-sized septal, auricular, and costal cartilage grafts free of perichondrium. All autografts were implanted subcutaneously on the nasal dorsum and then removed after 3 months. Graft mass, chondrocyte density, and histologic features from hematoxylin-eosin–stained sections were compared before and 3 months after implantation.
Results At 3 months after implantation, septal cartilage grafts averaged 30.8% resorption by weight, followed by auricular (23.1%) and costal (7.6%) cartilage. All 3 groups demonstrated similar changes in chondrocyte density and minor calcification at 3 months. There was no evidence of necrosis or inflammatory changes in any of the specimens.
Conclusions Although the septum is often the preferred source of autogenous cartilage for nasal reconstruction, short-term resorption of septal cartilage appears to be higher for dorsal onlay grafts. The low resorption of costal cartilage may be due in part to its compact shape compared with septal and auricular cartilage. It remains to be seen whether these differences in resorption persist in the long term.
THE ROLE OF GRAFTS and implants in aesthetic and reconstructive surgery of the face has evolved considerably during the past several decades. The search for the ideal implant material has led to an exponential growth in research and development of different alloplasts, with silicone, expanded polytef, and polyamide mesh being some of the most widely used. The advantages of a suitable alloplast are clear: unlimited supply, custom contouring, maintenance of shape, permanence, and lack of donor site morbidity. Although alloplasts have gained widespread acceptance in several areas of the body, their use in the nose has been associated with an unacceptably high rate of extrusion, infection, and rejection.1-3 Certain unique anatomic features of the nose account for this: it occupies a prominent position in the center of the face, and its lower portion is mobile. Consequently, nasal implants are subject to repeated trauma and constant functional stress, causing persistent tissue irritation and impeding stability.4-5 While recent studies demonstrating long-term stability of expanded polytef (Gore-Tex soft-tissue patch; W. L. Gore & Associates, Flagstaff, Ariz) in the nasal dorsum appear promising,6 cartilage and bone grafts remain the material of choice for dorsal augmentation by most surgeons.
Several types of cartilage and bone dorsal onlay grafts have been used, including autografts, homografts, and xenografts. The cartilage autograft is preferred for most purposes, as it offers several distinct advantages: it is perfectly biocompatible; the tissue is readily available with minimal donor site morbidity; it can be easily fashioned into any desired shape; it maintains its structural resilience, yet may be gently morselized, if needed, to break its memory; unlike bone, it does not require direct contact with bone or cartilage to survive; long-term survival is well documented; extraction, if necessary, poses minimal damage to surrounding tissues; and there is no risk of disease transmission.7
Many animal and human studies have addressed the question of cartilage graft resorption, and few have succeeded in quantifying and comparing the absorption rates of autologous cartilage from different donor sites. In addition, most animal studies have not specifically addressed the nasal dorsum as the recipient site for cartilage implants. In our clinical experience, dorsal onlay cartilage grafts undergo much more resorption than do tip grafts such as lower lateral cartilage onlays, intercrural struts, and shield grafts.
The goal of this prospective study was to quantify the level of resorption of dorsal onlay cartilage autografts taken from the septum, auricle, and rib. The adult rabbit was chosen for our animal model because (1) its cartilage is similar to human cartilage8-10 and (2) it is an established recipient site model for nasal dorsal implants.6
Fourteen adult male New Zealand white rabbits weighing 4.0 to 4.4 kg each were used. All procedures were performed under animal care guidelines of the Committee on Animal Care, Institutional Review Board. Each rabbit underwent implantation of a cartilage autograft from the auricle, septum, and rib, onto the nasal dorsum. Anesthesia consisted of ketamine hydrochloride, 40 mg/kg intramuscularly, and xylazine hydrochloride, 5 mg/kg intramuscularly for induction, followed by inhaled 1% to 2% isoflurane to effect for the duration of the procedure. The ventral surfaces of the ear, the nose, and the lower chest wall were shaved and prepared with povidone-iodine in a sterile manner.
For the auricle autografts, the technique described by Brown et al11 was used. A 3-cm incision was made on the dorsal aspect of the base of the ear. A flap of ventral skin and perichondrium was raised at the base of the ear, and a 15 × 10-mm rectangle of auricular cartilage was taken. The donor site was closed primarily with interrupted 4-0 chromic catgut in a single layer.
For the septum, the approach described by Gubisch et al9 was used. This involved folding back the alar ridge of the nose to obtain extramucosal exposure of the septum. After hemitransfixion incisions, bilateral mucoperichondrial flaps were elevated, and a 15 × 10-mm rectangle of cartilage was removed.
The technique described by Coutts et al12 was used for the rib autografts. An oblique incision was made along the inferior costal margin. The medial cartilaginous portion of the ribs was exposed, and a 15-mm length of the inferior-most cartilaginous rib was excised by means of meticulous dissection. The chest was closed in 3 layers with 4-0 polyglactin 910 for muscle and dermis closure and 4-0 chromic catgut for skin.
Each donor cartilage specimen was cleaned of surrounding fibrinous attachments and soft tissue, washed with isotonic sodium chloride solution, and blotted dry. A No. 11 blade was used to carefully cut the specimens into 2 pieces, the first 10 mm long and the second 5 mm long. The smaller portion of each graft was placed in 10% buffered formalin and set aside for histologic analysis. The larger portion was weighed under sterile fashion on an analytic balance and set aside for implantation.
To prepare the recipient site, as previously described by Maas et al,6 a 1.0-cm anterior incision through skin and subcutaneous tissues was made on the dorsum of the nose. A subcutaneous pocket extending 4 cm in a cranial direction was formed directly over the bony dorsum superficial to the periosteum. The 3 cartilage grafts were placed in linear fashion in the pocket, each separated by 0.5 cm. The incisions were closed in 1 layer with 4-0 chromic catgut.
Postoperatively each rabbit was examined daily for signs of wound infection, seroma, hematoma, wound dehiscence, flap necrosis, or pneumothorax, until all wounds were healed. Subsequently, the animals were examined twice weekly.
At 3 months, the rabbits underwent anesthesia in a manner similar to that described already and then were killed with pentobarbital sodium, 200 mg/kg intravenously. A midline incision was made along the dorsum. The cartilage implants were freed from surrounding fibrous attachments. After removal, each implant was gently washed, dried, and weighed before sectioning for histologic analysis.
Preimplant and postimplant specimens were prepared for serial section and staining by hematoxylin-eosin and toluidine blue O. The auricular and septal grafts were oriented to produce 10-mm-long sections, the rib grafts to produce cross-sectional slices. For each graft, 3 random sections were chosen for analysis, distributed evenly among the entire length of the specimen. Each section was examined for evidence of necrosis, calcification, and presence of inflammatory changes. To determine the chondrocyte count, 3 sections were chosen from each graft, distributed evenly over its entire length. Chondrocyte counts were taken from a single 2 × 2-mm high-power field in the center of each section, and then averaged.
For weight analysis, each autograft was carefully dried and weighed immediately before implantation and again on removal at 3 months.
Preimplant and postimplant weight measurements and chondrocyte counts were compared by standard χ2 analysis. Because there were 3 different autograft types, we chose the nonparametric paired Wilcoxon test to compare resorption between the 3 groups. This involved isolating 2 of the 3 groups and comparing resorption between them independent of the third group. Consequently, multiple paired comparisons were produced: rib-septum, rib-auricle, and auricle-septum.
Table 1 shows the weight change for each group of autografts. Septal cartilage autografts averaged a weight loss of 30.8%, followed by auricular cartilage (23.1%) and costal cartilage (7.6%). A pairwise comparison of resorption between the 3 autograft types is shown in Table 2, with the use of the 2-tailed paired Wilcoxon test. With a Bonferroni correction of 3 for the multiple comparisons, the cutoff for statistical significance was P<.05/3 or .0167. The mean difference between septum and rib weight changes was −23.2% (range, +1.8% to −71.9%), which was statistically significant (P<.01). The difference between auricle and rib resorption was marginally statistically significant (mean, −15.5%; range, +59.6% to −64.9%; P<.02). The difference between septum and auricle resorption did not reach statistical significance (mean, −7.7%; range, +50.7 to −75.5%; P>.05).
Preimplant and postimplant chondrocyte density (Table 3) differed only marginally in all cartilage types and did not reach statistical significance according to a 2-tailed paired Wilcoxon test with Bonferroni correction. Histologic comparison of hematoxylin-eosin–stained slides showed minor levels of calcification in about one third of the postimplant specimens, regardless of cartilage type. There was no evidence of necrosis or inflammatory changes.
The most common sources of cartilage autografts are the nasal septum, auricle, and rib. Septal and auricular cartilage, when available, are preferred, as rib cartilage has a tendency to warp, although internal stabilization of rib grafts with wire has been reported.13 Rib cartilage autografts also carry the added risk of significant donor site morbidity. In addition, the tendency of rib cartilage to calcify with age makes costal grafts difficult to carve.14
One of the most important issues regarding cartilage grafts in the nose is the degree of resorption and remodeling the graft may undergo. Irradiated costal cartilage homografts have been studied extensively, with resorption levels quoted in human studies ranging from 0% to 75%.7, 15-17 Donald18 investigated animal costal cartilage homografts and calculated total resorption levels varying from 40% in sheep cartilage stored in thimerosal to 89% in irradiated dog cartilage. In another study using radiographic analysis, Donald et al19 found minimal resorption (1 of 20) in irradiated cartilage homografts implanted in the sheep facial skeleton. Babin et al20 found that fresh cat costal cartilage autografts and homografts underwent minimal resorption, but that both frozen autografts and irradiated homografts resorbed considerably. Breadon et al21 reported no significant difference in resorption between crushed and uncrushed auricular cartilage autografts in the rabbit model. Moreover, they found that the presence of perichondrium had no significant effect on graft viability. In contrast, Brown et al11 found considerably higher rates of resorption in crushed than in uncrushed rabbit auricular cartilage homografts implanted in the forehead. Adlington et al22 found that different modes of preparation of mouse costal cartilage homografts (irradiation, formalin, glutaraldehyde, and alcohol) did not influence observed fibrosis and resorption.
Recently, Tjelmeland and Stal23 examined rabbit auricular and costal cartilage autograft resorption at 1 year. They found 18.5% resorption by weight of auricular cartilage implanted in the nose, compared with 14.5% implanted in a control site (occiput). There was essentially no resorption of costal cartilage. Septal cartilage was not examined.
The aim of the present study was to compare the resorption of the 3 most commonly used sources of autogenous cartilage when used as dorsal onlay grafts. The results show a statistically significantly lower rate of resorption by weight of costal cartilage than either septal or auricular cartilage at 3 months. Changes in chondrocyte density were negligible.
Since the specimens were washed with isotonic sodium chloride solution and blot dried before weighing, the difference in preimplant and postimplant weights may merely reflect the tissues' ability to retain fluid rather than a change in amount of ground substance. However, the fact that there was no significant change in chondrocyte density makes this scenario less plausible.
Another possible reason for the observed difference in weight changes is that costal cartilage has a much lower surface area per unit weight than either septal or auricular cartilage. Cartilage resorption is thought to be caused at least in part by exposure to local mediators of wound healing such as interleukin 1.24 The compact shape of rib cartilage may limit the exposure to these local factors and thus reduce resorption.
Septal cartilage showed the highest resorption of the 3 cartilage types. There are 2 possible explanations for this finding. First, septal cartilage has the highest surface area per unit weight, being flat like auricular cartilage but much thinner in the rabbit. Second, the thin, delicate nature of the rabbit septum makes it much more susceptible to traumatic injury during harvest, which may result in cartilage resorption.11 Of course, in humans, septal cartilage tends to be thicker and more firm than auricular cartilage. Thus, the relatively higher resorption rates seen in rabbit septal cartilage may not translate to humans.
Another explanation for the variable but significant resorption of all 3 types of cartilage is the recipient site. We believe that dorsal onlay cartilage grafts act much more like synthetic bulking implants and may be subject to more resorption than tip grafts, eg, intercrural struts, lower lateral cartilage onlay grafts, and shield grafts. This may be due to the increased pressure exerted on dorsal onlay grafts, as they are sandwiched between the tight skin–soft tissue envelope and a fixed bony framework.
It is clear from this study that dorsal onlay cartilage grafts show variable but significant resorption at 3 months. Septal and auricular grafts undergo a considerably higher level of resorption than rib. This is significant, since the septum is often the preferred source of autogenous cartilage in nasal reconstruction. In our clinical experience, tip grafts undergo much less resorption and remodeling than do dorsal onlay grafts. Further studies are needed to determine the extent of cartilage resorption and remodeling over the long term, particularly in onlay and nasal tip grafts.
Corresponding author: Corey S. Maas, MD, Division of Facial Plastic Surgery, University of California, San Francisco, 2400 Clay St, San Francisco, CA 94115.
Accepted for publication August 26, 2002.