U S E OF A B O N E B L O C K G R A F T F R O M T H E I L I A C C R E S T W I T H R I G I D F I X A T I O N TO C O R R E C T D I A P H Y S E A L D E F E C T S

OF T H E R A D I U S A N D U L N A

C. H BARBIERI, N. MAZZER, C. A. ARANDA and M. M. de O. PINTO

From the Hand Surgery and Microsurgery Service, Department of Orthopaedics and Trauma Surgery, Ribeirto Preto School of Medicine University Hospital, Ribeir~o Preto, Brazil

We report our experience with the use of a bone block graft from the lilac crest to correct diaphyseal defects of the forearm bones. The technique was used in 12 patients (ten men and two women, average age 29 years) for defects resulting mainly from either closed or compound fractures, which later developed infection and bone tissue loss. The average dimensions of the graft required to correct the defect was 3.5 x 1.8 cm. The graft application was combined with rigid fixation with an AO 3.5 mm D C P plate, permitting early active motion. The graft incorporated without any additional grafting procedure in ten cases within 17.2 weeks on average. The most frequent complication was infection (four cases), controlled by means of d~bridement, cleansing and antibiotics. A comparative analysis of the immediate and final radiographs of the graft showed an average 30% loss of bone mass despite integration. We conclude that the technique of bone block grafting to correct diaphyseal defects of the radius or ulna is relatively easy to carry out and has a high success rate. Journal of Hand Surgery (British and European Volume, 1997) 22B: 3." 395-401

Surgical correction of wide diaphyseal segmentary bone defects and non-unions associated with extensive bone loss is still a challenging problem, despite progress in the techniques of osteosynthesis and bone grafting. This is particularly true for the forearm, where massive loss of bone, apart from other tissues, may be produced by trauma and infection, so that function and working capacity are severely impaired.

Many conventional grafting techniques have been proposed to treat segmentary defects in general. Boyd (1943) devised the "dual graft" technique, which consists of two blades of cortical bone removed from the anteromedial aspect of the tibia and fixed with screws to either side of the bone to be treated, the medullary surfaces facing inwards. The gap between them is filled with cancellous bone graft. According to his own evalu- ation, the results obtained with this technique in the treatment of pseudarthroses of various causes in many different bones, including the radius and ulna, were excellent.

Gibson and Loadman (1948) presented their experi- ence with the use of the "onlay" cortical graft, also removed from the anteromedial aspect of the tibia, in the treatment of post-traumatic non-unions of many bones, including the radius and ulna. The graft was fixed on to the bone with screws and its main advantage was mechanical strength, not obtained with the "inlay" cancellous bone graft.

Scuderi (1948) discussed three types of bone graft that he had used in the treatment of 94 cases of bone defects caused by war wounds and stated that the "onlay" cortical graft fashioned as a bone plate and fixed to the bone with screws had better mechanical properties, enhancing healing; the cancellous graft should only be used to fill defects where mechanical stability is not a problem.

Spira (1954) reported a bone block graft from the iliac crest used with intramedullary pin fixation for defects of both radius and ulna, reporting good results. The graft is harvested so as to have three cortices, which provide it with adequate mechanical strength. Shortly thereafter Nicoll (1956) presented his experience with a similar technique, which differed from Spira's technique in that the graft, despite being a bone block from the iliac crest, had no cortices when applied to small defects; in case of wider defects he would retain just one cortex to improve mechanical stability. Fixation was carried out either with intramedullary pins or with plates and screws.

Dabezies and colleagues (1971) used the technique of the tri-cortical bone block graft from the iliac crest and osteosynthesis with a compression plate in 11 patients, four of whom had a diaphyseal defect in the radius, six in the ulna and one in both forearm bones. The defect width ranged from 1.5 to 7.5cm, but healing was achieved in ten patients within a period of 9 to 13 weeks. Non-union at the distal junction of the graft with the recipient bone occurred in a single case due to infection. The authors emphasized the quick recovery of forearm mobility due to the stability and the graft healing within a relatively short period of time.

Weber and Cech (1976) suggested the use of a bone block graft from the iliac crest ("pressure-proof blocks") to correct diaphyseal defects of many different bones, including the radius and ulna. They considered that the "cancellous bone is used not only as living substance but also as mechanical building material for restoring the normal anatomy". However, they recommended the technique "for defects of not more than 3 cm, combined with a stable osteosynthesis (plate or fixateur externe)".

A similar technique was employed by Calkins and colleagues (1987) in the treatment of diaphyseal defects

395

3 9 6

of both radius and ulna resulting from gunshot wounds in five forearms. The operation was carried out around the ninth day after the accident ("delayed primary treatment") and healing was achieved in four cases, on average 6.5 months thereafter.

We have used the technique of the bone block graft from the iliac crest in 12 patients and report here our experience.

MATERIAL A N D M E T H O D S

Twelve patients (ten men and two women) with a mean age of 29 years (range: 11-71) were operated on for diaphyseal defects of the forearm bones, eight of them in the radius and four in the ulna (Table 1). The defect resulted from trauma in eleven patients and from tumour resection in one. The cases resulting from trauma included five closed fractures and six compound frac- tures of variable degree.

All five closed fractures had been submitted to open reduction and internal fixation upon arrival and later presented with complications such as infection, non- union or malunion (Tables 1 and 2).

Four of the six cases with compound fractures were G-III open fractures, with contamination and a variable loss of bone tissue. All were treated primarily by soft tissue and bone d6bridement and external fixation. All developed a variable degree of infection and were later treated by elective bone block grafting, after complete healing of the original wound. One patient (case 12) was first treated with a cortical bone graft from the tibia; this procedure failed but the bone block grafting carried out secondarily succeeded in uniting the defect.

All three patients with an infected non-union follow- ing internal fixation were treated by removal of the plate, d6bridement, cleansing and external fixation; external fixation was maintained until soft tissue healing was complete to permit secondary bone block grafting. An external fixator was used in seven patients and removed immediately before the operation for bone block grafting.

The remaining patient (case 2) was an l 1-year-old girt with Ewing's sarcoma of the ulna. She had been treated primarily by a revascularized transplant of the fibula to replace almost the entire ulna, which was resected after massive systemic chemotherapy. She devel- oped a proximal non-union, probably due to a very short proximal ulnar stump and unstable fixation; she was treated secondarily by bone block grafting, but this also failed to unit.

The time from the day of injury or first operation (tumour case) to the day of the operation for iliac bone block grafting was 11 months on average (range: 1-30 months). The defect was approached by conventional incisions, with extreme care to avoid excessive detach- ment of the muscles and remaining periosteum. The bone stumps were freshened and the medullary canal was opened. The defect was measured and a bone block

T H E J O U R N A L O F H A N D S U R G E R Y V O L . 2 2 B N o . 3 J U N E 1 9 9 7

=

@ m

@

d _= =

°~

..=

0

0

~n

0

g

o

o

. ~ .g

O O • ~ , ~ . ~

0 0 0 " ~ 0 "~

. . . . . n ~ , = ' ~ - ' - "no ~ o ~ ,

+ + + + • d ~ -d .d

+ ~ . - ~ + ~ ~ ~ . . -

~ ~~ ~° ~ ~

O O

H

E ~u

H

g

q::

II

-tn H

n~

BONE GRAFTS FOR RADIUS AND ULNA

Table 2--Distribution of the fracture cases according to type, initial treatment and result

397

Fracture type N °. In i t i a l t r ea tmen t Resu l t

E F O R I F PC Infect ion N o n - u n i o n Ma l -un ion Open 6 4 1 l 5 1 Closed 5 4 1 3 1 1

EF = external fixation; ORIF = open reduction + internal fixation; PC =plaster cast.

was obtained from the iliac crest accordingly. The length of the graft necessary to fill the defect was 3.5 cm on average (range: 2.3-5.0cm); its width was 1.8cm on average (range: 1.5-2.0 cm) and its thickness was that of the iliac bone itself (average 1.5 cm). The graft was always cut from the iliac crest with a power saw for precision and as a rule it was about 1.0 cm longer than the defect to allow for shaping.

Once shaped, the graft was fitted into the defect under pressure and fixed in place by means of a 3.5 mm DCP plate long enough to place at least three screws into each bone stump, according to the AO techniques (M011er et al, 1991 ). Dynamic compression was obtained with at least two screws, one on each end of the plate; occasionally a third compression screw was driven on one end of the plate so as to get a firmer assembly. After compression, the rest of the screws were driven in the neutral position. The graft itself was then fixed under the plate by means of one or two screws. A suction drain was left in for 24 to 36 hours. The entire procedure took 3 hours on average.

Antibiotics were started the day before the operation and maintained for 1 week even in the absence of infection; if infection occurred, they were used for as long as necessary. External immobilization by means of above elbow dorsal plaster slab was maintained for 4 to 8 weeks. Daily active assisted mobilization was instituted as early as the first week in most cases and the patients were reviewed at monthly intervals for clinical and X-ray evaluations, until the graft had taken or a new procedure became necessary.

During these evaluations, particular attention was paid to the condition of the skin and deep soft tissues, for early detection of infection and/or sinus formation, to the stability of the bones and the plate and to the mobility of the forearm at both the elbow and wrist. The functional results were estimated according to our own criteria, which took into consideration the possible total range of motion (ROM) of the normal elbow and wrist, respectively 320 ° and 390 ° . They were rated good when they corresponded to over 80% of the normal ROM, average when between 61% and 80% of the normal, and poor when below 60%.

Apart from the overall X-ray appearance of the graft, particularly its texture, measurements were made from X-rays of the length and width of the graft. With time, the only measurement that could be made was that of the graft's width, since the junctions between graft and host bone faded by healing and remodelling.

RESULTS

Primary healing of the graft, without the need for any complementary grafting procedure, occurred in ten patients (Figs 1 and 2), but healing required on average 17.2 weeks (range: 4 40 weeks). The final width of the graft was 1.3 cm on average (range: 1.l-1.6 cm), which meant an average decrease of 0.6 cm from the first postoperative radiographs. Healing was totally unevent- ful in five patients. Infection occurred in four other patients, being deep in three and superficial in one, but was adequately controlled by cleansing, d6bridement and continuing antibiotics. It did not become chronic in any patient, but may have contributed to delayed healing in two (6 and 12).

In two cases the iliac bone block graft failed; in the first one (8), decortication, cancellous bone grafting and another plating were carried out and eventually resulted in healing. In the second, the tumour case (2), a new bone block grafting was undertaken but this also failed.

In one patient (10) radioulnar synostosis developed 4 months after the operation and was treated by resection of the bone bridge.

No complication was observed in the donor area, except for pain during the first postoperative days, which subsided spontaneously and did not interfere with walk- ing. The average range of flexion and extension at the elbow was 115 ° (range: 35°-145°), with flexion varying from 130 ° to 150 ° (average: 144 ° ) and extension from minus 95 ° (average: minus 24 ° ) to 5 ° . The average range of forearm rotation was 148 ° (range: 95°-180°), while pronation alone varied from 50 ° to 80 ° (average: 73 °) and supination from 45 ° to 100 ° (average: 75°). At the wrist, the average range of flexion and extension was 123 ° (range: 70°-180°); flexion alone was 64 ° on average (range: 30°-90 °) and extension alone was 59 ° on average (range: 10°-90°). The average range of radioulnar devi- ation at the wrist was 41 ° (range: 20°-72°); radial deviation alone was 23 ° on average (range: 20°-30 ° ) and ulnar deviation was 18 ° on average (range: 0°-52°).

The muscle strength of the operated limb was invariably decreased, so that the average hand grip power was reduced to 48% of the opposite normal hand (range: 28-74%). Thumb-index tip-to-tip pinch was reduced to 60% (range: 43-83%) and the lateral pinch to 83% (range: 65 94%).

The final functional recovery was rated good in nine patients and average in three; no patient had a poor result.

398 THE JOURNAL OF HAND SURGERY VOL. 22B No. 3 JUNE 1997

Fig 1 Closed fracture of the radius and ulna treated by primary osteosynthesis; (a) uneventful healing in the ulna, but infection, osteolysis and non-union in the radius (b) treated initially by plate removal, d6bridement and external fixation. (c) After infection was controlled and the soft tissues had healed, a 5 cm bone block graft from the iliac crest was transplanted and fixed by means of a 10-hole AO compression plate. (d) The 10-week X-ray shows integration of the graft and union at both ends.

Resumption of heavy activities took 28 weeks on average (range: 16-32 weeks).

DISCUSSION

The iliac bone block graft as used in these patients with diaphyseal defects of the radius or ulna, resulting mostly

from trauma, produced excellent results, particularly if one takes into consideration the severity of the cases and the fact that most of them had already undergone at least one previous operation, usually followed by a complication. An unfavourable bed for bone grafting certainly resulted from those events. Furthermore, in all patients the defects involved a considerable segment of

BONE GRAFTS F O R RADIUS AND ULNA 399

the bone diaphysis, usually too wide to be corrected by a more conventional procedure, and were associated with sequelae from soft tissue injuries. Therefore, a successful result in ten out of 12 patients was quite satisfactory, despite a few complications. The infection rate, for instance, was relatively high and this was also expected, if one takes into account the patients' previous histories. Fortunately they responded quite well to treat- ment by surgical cleansing and d6bridement, associated with systemic antibiotics. Although cured, infection may have been responsible for some delay in the graft healing

in two cases. Established non-union only occurred in two cases, requiring another type of treatment.

The operative technique for applying the bone block graft is relatively simple, but requires some extra care in preparing the recipient site by removing all devitalized tissue. It also requires full knowledge of the current techniques of compression osteosynthesis. These make the bone-graft-bone assembly completely stable, a necessary condition to allow for complete integration of the graft. Otherwise, the graft would undergo excessive osteolysis and resorption and the healing time would be

400 THE JOURNAL OF HAND SURGERY VOL. 22B No. 3 JUNE 1997

Fig 2 (a) Patient with an atrophic non-infected non-union of the radius, following a closed fracture of both forearm bones primarily treated by osteosynthesis, (b) treated by bone d6bridement and a 5 cm bone block graft, fixed with a 10-hole AO compression plate. (c) Uneventful healing at 8 weeks.

longer; even non-union could be the result of a loosely fixed graft.

infection must be cured prior to any reconstructive procedure to avoid spread and the risk of graft resorp- tion. This was why six of the patients studied here were treated by an external fixator and one or more d6bride- ments, until infection had subsided and the soft tissues were healed, so as to allow for another procedure (Fig 1).

The harvesting of the bone block graft from the iliac crest is also an easy procedure, but it is recommended that a power saw be used for precision and to prevent further trauma to the bone and to the surrounding soft tissues. In this series no complication was noticed in the donor area, as has been reported previously (Kurz et al. 1989).

In all cases, even those with quick and complete integration, some degree of graft resorption was

BONE GRAFTS FOR RADIUS AND ULNA

observed by measuring and comparing its width on the immediate postoperative and final radiographs. Although an average 30% bone loss was estimated (Fig 2) the final graft width was about the same as that of the recipient bone and in some cases was even greater, as a result of always using a graft of slightly broader dimensions than necessary.

Despite the considerable reduction in muscle power resulting from fibrosis combined with disuse, the overall functional results were good, although not directly dependent upon the bone block graft itself. However, the use of a compression plate provided such a solid assembly that early mobilization was possible and this certainly accounted for a quicker and more complete recovery.

There are other techniques for the treatment of wide diaphyseal defects of the forearm bones, including the revascularized transplant from the fibula. Although sub- ject to a number of complications, it certainly produces good results, but only in the hands of experienced surgeons with specific training, while the bone block technique is within the reach of any orthopaedic surgeon with conventional training. We also have a considerable experience with the revascularized fibular transplant and we believe that the final results of the two techniques are comparable. Therefore, the technique of the bone

401

block graft from the iliac crest is recommended for treatment of the type of diaphyseal defects discussed here.

References Boyd H B (1943). The treatment of difficult and unusual non-unions. Journal

of Bone and Joint Surgery, 25: 535-552. Calkins M S, Burkhalter W, Reyes F (1987). Traumatic segmental bone defects

in the upper extremity. Journal of Bone and Joint Surgery, 69A: 19-27. Dabezies E J, SteWart W E, Goodman F G, Defter P A (1971). Management

of segmental defects of the radius and ulna. Journal of Trauma, 11: 778-788. Gibson A, Loadman B (1948). The bridging of bone defects. Journal of Bone

and Joint Surgery, 30A: 381-396. Kurz L T, Garfin S R, Booth R E (1989). Harvesting autogenous iliac bone

grafts. A review of complications and techniques. Spine, 14: 1324-1331. Mtiller M E, Ailg6wer M, Schneider R, Willenegger H. Manual of internal

fixation: techniques recommended by the AO-ASIF group 3rd Edn, Berlin, Springer-Verlag, 1991: 466-475,

Nicoll E A (1956). The treatment of gaps in long bones by cancellous insert grafts. Journal of Bone and Joint Surgery, 38B: 70-82.

Seuderi C (1948). Restoration of long bone defects with massive bone grafts. Journal of the American Medical Association, 137:1116-1121.

Spira E (1954). Bridging of bone defects in the forearm with iliac graft combined with intramedullary nailing. Journal of Bone and Joint Surgery, 36B: 642-646.

Weber B G, ~ech O. Pseudarthrosis. Bern, Hans Huber Publishers, 1976: 85-94.

Received: 16 July 1996 Accepted after revision: 18 September 1996 Dr C.H. Barbieri, Department of Orthopaedic and Trauma Surgery, Ribeir~o Preto School of Medicine, 14049-900 Ribeir/Io Preto (SP), Brazil.

© 1997 The British Society for Surgery of the Hand