6.3. Reduction of morbidity

6.3.1. Safety of trephine harvesting of the anterior iliac crest

Although previous reports have looked at the technique of trephining bone to avoid open procedures (Schwartz & Leake 1979, Altman & Blenkisopp 1994, Habal 1995) the bone harvest was not well quantified and the sites were not thoroughly evaluated for perforations. The use of a hand-held driven trephine with an unspecified diameter, that can penetrate the iliac crest up to 8 cm, was suggested for harvesting up to 6 ml of bone from the anterior iliac crest (McGurk et al. 1993). These authors suggest that care be taken at depths greater than 80 mm, a depth that is impossible with the system evaluated here, as demonstrated with the 20 intentional perforations performed on the cadavers. No peritoneal involvement was found on exploration of the intentionally perforated sites. In the 50 cadaveric sites where bone cores were harvested in a more routine fashion, the perforation rate was 4.4% of cores and occurred in the most atrophic of cadavers with core lengths of greater than 30 mm. The perforations appeared as breaches of the cortex without periosteal disruption or with fraying of the innermost aspect of the iliacus muscle. The perforations never encroached upon the peritoneum or peritoneal contents.

The instrument used in this study derives its simplicity from the cutting core drill which engages the bone as the drill is turned on. The trephine advances on its own, with minimal force or pressure and practically using only its own weight. As the cutting core advances, it is bounded by the cortex on the medial and lateral sides of the ilium. The perforations seen in the cadavers occurred at depths greater than 30 mm which corresponded to areas where the ilium is narrow and has minimal cancellous marrow. The use of the hand held guide limits the penetration of the core cutter to 38 mm and thereby reduces the likelihood of perforations laterally or medially. Despite purposeful perforation, the peritoneum was untouched.

The cadaveric study quantifies the amount of bone that can be easily harvested using the power driven trephine evaluated in this study. Adequate amounts of bone can be harvested for use in many cranio-maxillofacial procedures. Harvesting 5 to 7 cores per iliac crest site can produce 2.3–3.2 ml of bone, which can adequately augment the maxillary sinus floor to allow for dental implant placement, treat alveolar clefts, and fill bony defects. Clinically, almost all the cores are greater than 30 mm in length and it is easy to obtain 5 to 6 cores per hip. One can further increase the yield of harvested bone by changing the core direction even through the same opening.

The cadaveric bone in this study is a good estimation of the quantity of bone obtainable. The cadavers were under 30 days old but from an elderly population, which may have been slightly atrophic with some degree of degeneration even with formalin preservation. A previous attempt at this study resulted in grossly underestimated values of bone volumes due to the poor quality of bone taken from 8 month old cadavers.

6.3.2. Morbidity with trephine harvesting

In study II, a total of 333 corticocancellous bone cores were harvested from 86 iliac crests. These cores appeared as compact cylinders of bone that could be easily morcellized for packing or moulded into desired shapes. Generally four to six cores ranging in length from 32 to 38 mm could be easily obtained from a single 1 cm incision. Typically, the first 3 to 4 mm of the core was composed of cortical bone, the remainder was cancellous bone. The amount of bone obtained was related to the number of cores taken. The use of a curette introduced through the stab incision and placed through the osseous holes created by the trephine served to increase the yield of bone substantially. Using the cores alone, the mean volume obtained per site ranged from 3 to 21 ml of compacted bone.

The results of this study show a reduction in morbidity with the use of a power driven trephine to obtain anterior iliac crest cancellous cores and correspond well with previous reports (Duncan et al. 1980, Minns & Sher 1983, Caddy & Reid 1985, Williams & Ford 1986, McGurk & Barker 1993, Altman & Blenkisopp 1994, Billmire & Rotatori 1994). In a retrospective study of complications following 14,810 iliac crest biopsies from 14 centres, trephines were reported to leave minimal scars, decrease morbidity and produce less dysesthesia than open procedures (Duncan et al. 1980). Complications including local haematomas, lateral cutaneaous nerve neuropathies and pain in excess of 7 days duration, occurred in 0.63% of cases. If the vertical approach was used the rate of complications decreased to 0.36%.

Study II assessed the morbidity rate associated with the technique of using a power driven trephine in a minimally invasive fashion from the anterior iliac crest. Intra-operatively, the complication rate was 0.3% and involved a single broken instrument. There were no instances of medial cortex perforation or unusual bleeding. Post-operatively, the complication rate was 3.6%. Short and long-term complications including pain, paresthesia, post-operative gait disturbance, and patient satisfaction revealed that only one patient described having pain for more than 3 days. That same patient was the only one to describe having a gait disturbance in excess of 3 days. With respect to patient satisfaction 98.8% of the patients stated that they would undergo the procedure a second time if necessary. There were no reports of pain or gait disturbance that persisted longer than 3 weeks post-operatively.

In comparison to the above results, a study of open anterior iliac crest harvesting in 100 patients was associated with a 2% incidence of major thromboembolic events and a 2% incidence of permanent sensory disturbance in the distribution of the lateral femoral cutaneous nerve (van der Wal et al. 1986). As far as the patients’ expectations after surgery were concerned, at the anterior ilium it was worse than expected for 18% of patients, better for 81% and 1% had no opinion (van der Wal et al. 1986).

Study II showed that harvesting CC’s from the anterior iliac crest is a procedure that can be carried out predictably in an outpatient or ambulatory surgery environment. In the 41 cases that were booked as outpatient procedures, all 41 patients were discharged immediately after recovering from general anaesthesia, as planned. However, the ability to use this technique on an outpatient basis is almost completely dependent on the extent of the reconstructive procedure being performed, as well as the overall health of the patient. The procedure can be regarded as a cost saving measure as it can avoid the admission and hospital stay for the patient.

The results of study III demonstrated significantly less morbidity following procurement of CC grafts from the anterior ilium using a power-driven trephine than following the open harvest of a CCBG. Use of the trephine resulted in fewer days to first unassisted ambulation, shorter overall length of hospital stay, and significantly less donor site pain than following harvest of CCBG. The reduced morbidity associated with the use of the power-driven trephine was likely a result of the minimization of muscle reflection and exposure of the anterior ilium.

The concave-convex anatomy of the anterior ilium limits bone harvest to the superior 3–4 cm between the ASIS and the iliac tubercle. In the middle one-third, the medial and lateral cortices come together with little or no intervening marrow. Because of this osseous anatomy, all traditional approaches to the ilium involve the reflection of regional muscle groups. A lateral approach involves reflection of tensor fascia lata, gluteus medius, and gluteus minimus, while a medial approach involves the iliacus muscle. Tensor fascia lata acts to lift, flex and stabilize the thigh during walking, and any injury from reflection, retraction or inaccurate re-approximation can lead to marked post-operative pain and gait disturbance. The iliacus, a postural muscle, inserts into the lateral side of the psoas major muscle tendon. Trauma or haematoma formation can result in psoas muscle inflammation and may also contribute to post-operative pain and gait disturbance.

The reflection and retraction of the iliacus muscle used in the anterior medial approach to the ilium may play a significant role in patient morbidity. In addition, bleeding from the exposed cancellous marrow can lead to haematoma formation. While these effects can be minimized with clean subperiosteal reflection of the medial tissue, careful control of marrow bleeding, and precise muscle re-approximation, they cannot be abolished.

Bone procurement using a trephine can proceed without muscle reflection. The instrument described in this report is designed specifically to minimize dissection. The serrated edge of the trocar is used to engage the fascia and periosteum, and allows the procedure to be carried out through a stab incision. The pre-cutter drill is used to score the fascia, periosteum and crestal bone. The core cutter engages the recess in the cortical crestal bone made by the pre-cutter drill and advances through the cancellous marrow with minimal operator assistance. A series of clutches reverses the core from the ilium, and a built-in internal plunger is then used to eject the core from the trephine.

Due to the laxity of the overlying skin and subcutaneous tissues, the incision can be “walked” up and down along the crest of the anterior ilium to fresh harvest sites. Careful superficial dissection should allow the operator to remain directly over the iliac crest between the medial (external oblique, iliacus) and lateral (tensor fascia lata, gluteus minimus, gluteus medius) muscle groups. Minimal bony exposure is required to harvest the CC’s themselves, and the bony septae between them, even when the ilium is curetted. Bleeding from core donor sites can be controlled with the use of gelfoam. The spread of any residual bleeding is limited by the integrity of the cortical walls of the ilium and the musculoperiosteal attachments. Without the medial or lateral subperiosteal reflection typical of the standard approach, haematoma formation is minimized.

Although not formally evaluated in this study, procurement of cancellous cores with the motorized trephine approach was always faster than harvesting CCBG’s. Minimal surgical trauma and shorter operating time resulting from the use of the motorized trephine likely contribute to the reduction in overall patient morbidity.

Trephining the anterior iliac crest for the purpose of harvesting autogenous bone is reliable, safe and predictable. It is associated with minimal intra-operative and post-operative morbidity. An advantage of this technique is that it can predictably be used on an outpatient basis eliminating the need for admission to hospital and the expenses associated with it.

6.3.3. Morbidity with coral-derived granules in the cranio-maxillofacial skeleton

In study IV, the results with CDG subperiosteal implantation in the cranio-maxillofacial skeleton were quite encouraging. With the exception of five sites of clinically evident resorption of the material, the remaining patients enjoyed a seemingly stable augmentation. Also the resorption in the five sites was not complete so that there was an improvement of the clinical situation, when compared to the premorbid condition. Even if the material required removal, as necessitated by infection, the nature of the procedure, with a small distant incision is such that the patient was not compromised aesthetically.

However, a follow-up period of 12 to 36 months may not be long enough to allow one to be conclusive with regards to the long term behaviour of xenograft derived bone in the cranio-maxillofacial skeleton. The results of study IV are in general agreement with previous reports of CDG placed into the cranio-maxillofacial skeleton (Levet & Jost 1983, Robier et al. 1987, Besins & Philipe 1988, Brasnu et al. 1988, Levet et al. 1988, Roux et al. 1988a,b), although none of these previous studies reported any long term follow-up.

As an addendum, this study initially considered only those patients operated on before June 1991. Since that time, a further 12 sites of augmentation with coral-derived granules have been performed in nine patients. Two of these nine patients developed infected sites post-operatively. The first was a 16-year-old girl who had three sites injected on the forehead, to treat sequelae of craniosynostosis surgery. All three sites developed post-surgical swelling with aseptic pus. This particular patient had an underlying case of severe acne vulgaris and required the total removal of the coral granules in order to allow resolution of the infection. The second patient had a post-traumatic asymmetry of the forehead and was treated with CDG augmentation. Two months later he was noted to have an otherwise asymptomatic fluctuant area on palpation of the forehead. A total of 1.5 ml of aseptic pus was evacuated and the infection subsided, although the CDG were not removed.

The procedure seems relatively easy to perform, but caution must be exercised. Dissection of the subperiosteal pocket must be precise to avoid deposition of granules in unintended locations. CDG must be placed in an appropriately prepared non-infected bed. The granules must be placed in a subperiosteal location with good bony contact if proper bony in-growth and replacement-resorption of this osteoconductive xenograft scaffold is to occur.

Aseptic technique is of paramount importance. The granules become implanted in a wound surrounded by blood and fibrin and present a potentially favourable environment for microbial proliferation.

It is important to differentiate between coral granular wound irritation and frank infection. In the two cases of wound irritation described in the study, one was in a subciliary wound used for a malar deposition, the other was in a scalp wound. In these cases an inflammatory reaction develops in response to the coral granules mistakenly left in a supraperiosteal location, in the superficial layers of the wound. Care must be taken to avoid leaving granules in an ectopic subcutaneous location.

In the case of frank infection, total evacuation of the coral granules may be the only recourse. The distant small incision used to introduce the granules, may pose a technical problem for complete curettage of the subperiosteal pocket.

The long term stability of coral augmentation remains an unknown. A prospective randomized trial would help to answer this question. Additionally serial computerized tomography scans with volumetric analysis may help determine the nature of the changes in the augmented volume of coral granules over time.

Coral granules seem to be well tolerated in the cranio-maxillofacial skeleton when introduced into a well vascularized, aseptic, subperiosteal pocket. Complications seem to be few in number and the result is an improvement over the pre-treatment situation in most cases. There however, is no direct evidence that coral granules actually become totally resorbed and transformed into bone in these cranio-maxillofacial sites. Future serial histological studies would be necessary to show this.

6.3.4. Morbidity with coral-derived granules in the dento-alveolar area

The results of study V show that CDG are well tolerated in the dento-alveolar region where placement of the material relies on an intra-oral approach and a wound in the oral cavity. Other studies are in agreement with this observation (Issahakian et al. 1987b, Issahakian & Ouhayoun 1988, Yukna & Yukna 1998).

However previous studies did not look at the prevention of alveolar ridge resorption in the paediatric dentition. In the deciduous paediatric dentition, loss of retained second deciduous molar, which has no succedaneous permanent replacement tooth, is associated with bone loss. This is manifested as a decrease in the width of the alveolar ridge of 25% within 3 years after extraction of the retained primary molars, and this continues to diminish a further 4% over the next 3 years (Ostler & Kokich, 1994).

This study demonstrates that CDG have been useful in the preservation of alveolar ridge dimensions following the elective removal of ankylosed primary molars lacking succedaneous premolars. This ridge preservation allowed the placement of dental implants into these sites without a bone graft in 93.5 % of sites. Conversely, grafting coral granules into defects created in the anterior maxilla following post-traumatic loss of incisors failed to preserve ridge dimensions sufficiently to permit the placement of dental implants without revisionary bone grafting. Implants were possible in these traumatized sites in 17.6 % of sites without the use of a revisional autogenous bone graft. The starkly dissimilar results between post-traumatic maxillary incisor sites and post-extraction primary molar sites might be accounted for by the distinctly different nature of the insult incurred by the tissues prior to grafting. A violent trauma to the front of the mouth may result in tooth avulsion, and also in fractures of the alveolar bone, and detachment and laceration of the mucoperiosteum. All of these injuries may lead to scarring and compromise of the local blood supply and consequently impair healing, even though the labial and buccal plates of bone were noted to be intact for all sites in the study.

In the healing of traumatized teeth a number of complications may follow that may further corrupt healing, potentially through the effects of inflammatory root resorption and replacement, ankylosis, interference with local alveolar bone growth, submergence of the ankylosed tooth and distortion of the occlusion, and surgical removal of the residual root. In contrast, the relatively atraumatic, elective removal of an ankylosed deciduous molar produces a wound with very much less collateral damage, less scarring, and fewer implications for healing and wound repair.

One other possible explanation for the different results with CDG at the two areas of the alveolus may be due to differences inherent to the two sites. The bucco-lingual width of the alveolar crest is greater in the posterior parts of the maxilla and mandible than in the anterior maxilla. In addition the rate of alveolar resorption may be higher in the anterior maxilla than in the posterior maxilla and mandible. A great deal of resorption is known to occur in the first four months following tooth-loss in the anterior maxilla (Lam 1960). This anatomical difference and possibly different rates of alveolar resorption may place the maxillary anterior alveolus at a potential disadvantage when compared to at least a theoretically more favourable posterior maxillary or mandibular alveolar site (Ostler & Kokich, 1994).

Although controlled for in this study, the wearing of a routine acrylic mucosal-borne prosthetic replacement is thought to hasten the resorptive process. The patients who had lost anterior teeth were all given prosthesis which loaded the neighbouring dentition with a bite plane, thereby removing the load from the ridge and the xenograft.

Based upon the results of this study, coral granules may be used to preserve the dimensions of the alveolar ridge following extraction of retained primary molars lacking succedaneous premolars, thereby sparing the patient the potential morbidity associated with a bone graft harvest.