JBJS | Osteocutaneous Radial Forearm Free Flaps : The Necessity of Internal Fixation of the Donor-Site Defect to Prevent Pathological Fracture*

The Journal of Bone & Joint Surgery, Volume 82, Issue 5

Articles | May 01, 2000

Osteocutaneous Radial Forearm Free Flaps : The Necessity of Internal Fixation of the Donor-Site Defect to Prevent Pathological Fracture*

Kevin W. Bowers, M.D.†; Joseph L. Edmonds, M.D.†; Douglas A. Girod, M.D.†; Gopal Jayaraman, Ph.D.‡; Chee Pang Chua, M.S.‡; E. Bruce Toby, M.D.†

View Disclosures and Other Information

Investigation performed at the Section of Orthopedic Surgery, Kansas University Medical Center, Kansas City, Kansas
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Section of Orthopedic Surgery (K. W. B. and E. B. T.) and Department of Otolaryngolic Surgery (J. L. E. and D. A. G.), Kansas University Medical Center, 3901 Rainbow Boulevard, Kansas City, Kansas 66160.
‡Department of Mechanical Engineering, Michigan Technological University, Houghton, Michigan 49931.

The Journal of Bone & Joint Surgery. 2000; 82:694-694

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Abstract

Background: Osteocutaneous radial forearm free flaps have fallen from favor due to pathological fractures of the radius. The purposes of this study were to propose a means to decrease the rate of pathological fracture by prophylactic fixation of the donor-site defect and to evaluate this technique biomechanically.

Methods: Two groups of ten matched pairs of fresh-frozen cadaveric radii were harvested. In Group 1, an eight-centimeter length of radius comprising 50 percent of the cross-sectional area of the bone was removed to simulate an osteocutaneous radial forearm donor-site defect. This defect was created in one member of each pair, with the other bone in the pair left intact. In Group 2, both members of the ten matched pairs of radii had identical defects created as previously described. However, one radius in each pair had a twelve-hole, 3.5-millimeter dynamic compression plate placed across the segmental defect. In each group, five matched pairs were tested to failure in torsion and five matched pairs were tested to failure in four-point bending.

Results: In Group 1, the intact radius was a mean of 5.7 times stronger in torsion and 4.2 times stronger in four-point bending than the radius with the segmental resection. In Group 2, the radius that was ostectomized and fixed with a plate was a mean of 4.0 times stronger in torsion and 2.7 times stronger in four-point bending than the ostectomized radius.

Conclusions: Removal of an eight-centimeter segment from the radius dramatically decreased both torsion and bending strength. Application of a plate over the defect in the radius significantly restored the strength of the radius (p = 0.01).

Clinical Relevance: The segmental defect created in the radius when an osteocutaneous radial forearm free flap is harvested weakens the donor bone an unacceptable amount, resulting in a high risk of pathological fracture. We believe that prophylactic internal fixation of the donor-site defect with a plate restores strength to such a level that pathological fracture may be prevented, thus increasing the utility of the osteocutaneous radial forearm free flap.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

The radial forearm flap, or so-called Chinese flap, was first developed by Dr. Yang Guofan, Dr. Chen Baoqui, and Dr. Gao Yuzhi of the Shenyang Military General Hospital in 1978^23,24. Chinese investigators later reported on a large clinical series in which the flap was used very successfully in the reconstruction of head and neck defects²³. In 1983, Soutar et al. espoused the benefits of the radial forearm free flap, with or without bone, in the reconstruction of intraoral defects²⁴. They thought that this fasciocutaneous flap was ideally suited to intraoral reconstruction due to the pliability, thickness, and relatively hairless quality of the skin and the availability of vascularized bone for mandibular reconstruction. Furthermore, the reliability of the vascular pedicle and the cutaneous nerve innervation are well documented strengths of this technique^6,24.

Unfortunately, concerns over donor-site morbidity have caused enthusiasm for the osteocutaneous radial forearm free flap to wane. Problems with donor-site healing and appearance have been minimized by the use of various techniques, such as muscle closure over the tendon of the flexor carpi radialis or primary closure of the defect with the use of a local flap^2,10,17. However, morbidity related to the ostectomy of the radius remains a problem. The prevalence of pathological fracture has ranged from 0 to 66 percent, with an overall rate of fracture of 25 percent (twenty-eight fractures associated with 114 osteocutaneous flaps) when the results of all of these series are combined^{2,3,15,22,25,26}. Although most fractures in these series healed after treatment with immobilization, some had delayed union or nonunion, requiring additional treatment with external fixation and bone-grafting. Furthermore, some of these series demonstrated substantial impairment in pronation and supination and a 50 percent decrease in grasp and pinch strength in association with the pathological fracture of the donor site^2,3,15.

Several investigators have suggested ways to minimize the fracture rate, including the use of so-called boat-shaped, or beveled, osteotomies in preference to right-angled bone cuts; limits on bone resection to 40 percent or less of the cross-sectional area of the radius^25,26,32; and various regimens of immobilization in a cast or brace^2,24. However, in a study of donor-site morbidity in sheep tibiae that were used as a model of the ostectomized radius, Meland et al. found that even with these precautions the overall strength of the ostectomized bone in torsion was decreased by more than 70 percent¹⁸. This finding led them to recommend abandonment of the use of osteocutaneous radial forearm free flaps as a source of vascularized bone.

In the present study, we reexamined the effect of an ostectomy on the biomechanical properties of the bone with use of matched pairs of fresh-frozen cadaveric radii. We also examined the effect of prophylactic internal fixation of the ostectomized radius on biomechanical strength. The experimental hypothesis of this study was that plate fixation of ostectomized radii would increase the strength in torsion and bending to at least twice that of bones with ostectomy alone so that the number of pathological fractures would be minimized. We hypothesized this doubling of strength on the basis of our observation that the biomechanical strength of a bone with an ostectomy involving about 50 percent of the cortex was less than 50 percent of that of an intact bone. This finding agrees with those of studies in the literature that have demonstrated reductions of bone strength of much greater than 50 percent in ostectomized bone^18,27. In addition, clinical studies have documented that long bones with defects involving 50 percent or more of the cortical surface are at a greater risk (50 percent greater or more) for pathological fracture^8,9,16,21. We thought that full restoration of strength is not necessary; instead, we hypothesized that a substantial improvement (such as a doubling) of strength is sufficient to prevent most pathological fractures.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-A:: Illustration of the volar view of the ostectomized radius, showing the ostectomy defect between the insertions of the pronator teres and the brachioradialis muscles.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-B:: Illustration of the dorsal view of the ostectomized radius fixed with a plate. Note that the plate is well contoured to the dorsal-ulnar surface of the bone.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2:: Illustration of the torsion-testing apparatus. The ends of the radius were mounted in polymethylmethacrylate. The proximal end was fixed, and the distal end was twisted by the testing apparatus.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3:: Illustration of the four-point-bending testing apparatus. Note that the load points are applied to the dorsal surface and the support points are on the volar surface. LVDT = linear variable differential transducer.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-A:: Figs. 4-A and 4-B: Graphs showing the torque-versus-angle-of-twist curves. One inch-pound = 0.113 newton-meter.
Fig. 4-A: Specimen 3 from Group 1.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-B:: Specimen 11 from Group 2.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-A:: Figs. 5-A and 5-B: Graphs showing the moment-versus-center-displacement curves.
Fig. 5-A: Specimen 7 from Group 1.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-B:: Specimen 18 from Group 2.

View Larger

TABLE I: Mean Torsion Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1 (intact vs. ostectomized)	5.7 ± 2.6*	4.2 ± 2.2*	1.9 ± 1.1*
P value	0.016	0.053	0.168
Power	0.75	0.50	0.71
Group 2 (ostectomized vs. plate)	4.0 ± 1.0*	3.2 ± 1.0*	3.7 ± 3.2*
P value	0.002	0.01	0.06
Power	0.99	0.93	0.57

Add to My JBJS

TABLE I: Mean Torsion Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1 (intact vs. ostectomized)	5.7 ± 2.6*	4.2 ± 2.2*	1.9 ± 1.1*
P value	0.016	0.053	0.168
Power	0.75	0.50	0.71
Group 2 (ostectomized vs. plate)	4.0 ± 1.0*	3.2 ± 1.0*	3.7 ± 3.2*
P value	0.002	0.01	0.06
Power	0.99	0.93	0.57

View Larger

TABLE II: Individual Torsion Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/degree)	Toughness(Nmm)	Angle of Twist at Failure(degrees)	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1
1	L	Intact	4632	225	3142	60.5	9.11	5.00	3.37
	R	Cut	508	45	932	92.0
2	R	Intact	22,371	263	3783	21.5	3.92	1.45	0.74
	L	Cut	5706	181	5110	65.0
3	L	Intact	30,280	1116	6489	23.0	4.96	4.77	1.04
	R	Cut	6101	234	6255	75.5
4	R	Intact	10,056	321	5928	59.0	7.57	7.13	2.61
	L	Cut	1328	45	2269	114.5
5	L	Intact	23,162	872	12,718	50.5	2.75	2.68	1.89
	R	Cut	8417	325	6727	66.5	5.66 ± 2.62*	4.20 ± 2.19*	l.93 ± 1.09*
Group 2
11	R	Plate	11,976	389	5920	44.25	3.79	4.32	1.44
	L	Cut	3164	90	4103	107.0
12	L	Plate	15,705	299	14,281	70.0	2.99	1.89	3.75
	R	Cut	5254	158	3808	56.5
13	R	Plate	13,841	388	6384	41.5	4.45	3.92	1.46
	L	Cut	3107	99	4361	99.25
14	L	Plate	7118	154	2994	35.0	3.15	2.44	2.60
	R	Cut	2260	63	1152	39.0
15	L	Plate	8191	217	7016	62.0	5.37	3.44	9.14
	R	Cut	1525	63	768	40.0	3.95 ± 0.98*	3.20 ± 1.01*	3.68 ± 3.20*

Add to My JBJS

TABLE II: Individual Torsion Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/degree)	Toughness(Nmm)	Angle of Twist at Failure(degrees)	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1
1	L	Intact	4632	225	3142	60.5	9.11	5.00	3.37
	R	Cut	508	45	932	92.0
2	R	Intact	22,371	263	3783	21.5	3.92	1.45	0.74
	L	Cut	5706	181	5110	65.0
3	L	Intact	30,280	1116	6489	23.0	4.96	4.77	1.04
	R	Cut	6101	234	6255	75.5
4	R	Intact	10,056	321	5928	59.0	7.57	7.13	2.61
	L	Cut	1328	45	2269	114.5
5	L	Intact	23,162	872	12,718	50.5	2.75	2.68	1.89
	R	Cut	8417	325	6727	66.5	5.66 ± 2.62*	4.20 ± 2.19*	l.93 ± 1.09*
Group 2
11	R	Plate	11,976	389	5920	44.25	3.79	4.32	1.44
	L	Cut	3164	90	4103	107.0
12	L	Plate	15,705	299	14,281	70.0	2.99	1.89	3.75
	R	Cut	5254	158	3808	56.5
13	R	Plate	13,841	388	6384	41.5	4.45	3.92	1.46
	L	Cut	3107	99	4361	99.25
14	L	Plate	7118	154	2994	35.0	3.15	2.44	2.60
	R	Cut	2260	63	1152	39.0
15	L	Plate	8191	217	7016	62.0	5.37	3.44	9.14
	R	Cut	1525	63	768	40.0	3.95 ± 0.98*	3.20 ± 1.01*	3.68 ± 3.20*

View Larger

TABLE III: Mean Four-Point-Bending Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1 (intact vs. ostectomized)	4.2 ± 1.8*	5.2 ± 1.5*	3.7 ± 2.1*	0.82 ± 0.20*
P value	0.013	0.001	0.025	0.15
Power	0.81	1.00	0.72	0.99
Group 2 (ostectomized vs. plate)	2.7 ± 1.0*	2.0 ± 1.0*	4.1 ± 1.7*	1.4 ± 0.42*
P value	0.027	0.057	0.058	0.16
Power	0.83	0.87	0.44	0.92

Add to My JBJS

TABLE III: Mean Four-Point-Bending Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1 (intact vs. ostectomized)	4.2 ± 1.8*	5.2 ± 1.5*	3.7 ± 2.1*	0.82 ± 0.20*
P value	0.013	0.001	0.025	0.15
Power	0.81	1.00	0.72	0.99
Group 2 (ostectomized vs. plate)	2.7 ± 1.0*	2.0 ± 1.0*	4.1 ± 1.7*	1.4 ± 0.42*
P value	0.027	0.057	0.058	0.16
Power	0.83	0.87	0.44	0.92

View Larger

TABLE IV: Individual Four-Point-Bending Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/cGy)	Toughness(Nmm)	Center Displacement at Failure(mm)	Rotation at Failure(cGy)	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1
6	L	Intact	18,992	1.73 x 10⁵	3090	3.48	0.11	2.85	5.97	1.42	0.48
	R	Cut	6670	2.90 x 10⁴	2183	7.27	0.23
7	R	Intact	22,687	1.62 x 10⁵	4736	4.19	0.14	3.68	4.21	3.17	0.84
	L	Cut	6164	3.85 x 10⁴	1493	4.98	0.16
8	L	Intact	9963	1.01 x 10⁵	1396	3.07	0.099	3.93	4.79	3.45	0.85
	R	Cut	2533	2.11 x 10⁴	405	3.62	0.12
9	R	Intact	36,034	1.18 x 10⁵	5717	9.48	0.305	7.31	7.42	7.06	0.99
	L	Cut	4930	1.59 x 10⁴	810	9.55	0.31
10	L	Intact	39,006	1.44 x 10⁵	5572	8.41	0.27	3.32	3.56	3.45	0.92
	R	Cut	11,741	4.05 x 10⁴	1613	9.15	0.29
								4.22 ± 1.78*	5.19 ± 1.53*	3.71 ± 2.06*	0.82 ± 0.20
Group 2
16	L	Plate	15,107	1.51 x 10⁵	2298	3.21	0.10	2.82	1.69	4.69	1.74
	R	Cut	5350	8.92 Â¥ 10⁴	490	1.85	0.06
17	R	Plate	12,677	6.34 Â¥ 10⁴	5028	6.23	0.20	1.00	0.85	1.99	1.15
	L	Cut	12,706	7.47 Â¥ 10⁴	2532	5.40	0.17
18	L	Plate	14,148	1.09 Â¥ 10⁵	2936	3.98	0.13	3.42	2.90	4.02	1.17
	R	Cut	4137	3.76 Â¥ 10⁴	731	3.39	0.11
19	R	Plate	26,127	7.68 Â¥ 10⁴	13,457	10.67	0.34	2.84	1.42	6.61	1.98
	L	Cut	9189	5.41 Â¥ 10⁴	2034	5.38	0.17
20	L	Plate	24,881	9.95 Â¥ 10⁴	8709	7.66	0.25	3.39	3.25	3.39	1.01
	R	Cut	7334	3.06 Â¥ 10⁴	2570	7.56	0.24
								2.69 ± 0.99	2.02 ± 1.7	4.14 ± 1.70	1.41 ± 0.42

Add to My JBJS

TABLE IV: Individual Four-Point-Bending Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/cGy)	Toughness(Nmm)	Center Displacement at Failure(mm)	Rotation at Failure(cGy)	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1
6	L	Intact	18,992	1.73 x 10⁵	3090	3.48	0.11	2.85	5.97	1.42	0.48
	R	Cut	6670	2.90 x 10⁴	2183	7.27	0.23
7	R	Intact	22,687	1.62 x 10⁵	4736	4.19	0.14	3.68	4.21	3.17	0.84
	L	Cut	6164	3.85 x 10⁴	1493	4.98	0.16
8	L	Intact	9963	1.01 x 10⁵	1396	3.07	0.099	3.93	4.79	3.45	0.85
	R	Cut	2533	2.11 x 10⁴	405	3.62	0.12
9	R	Intact	36,034	1.18 x 10⁵	5717	9.48	0.305	7.31	7.42	7.06	0.99
	L	Cut	4930	1.59 x 10⁴	810	9.55	0.31
10	L	Intact	39,006	1.44 x 10⁵	5572	8.41	0.27	3.32	3.56	3.45	0.92
	R	Cut	11,741	4.05 x 10⁴	1613	9.15	0.29
								4.22 ± 1.78*	5.19 ± 1.53*	3.71 ± 2.06*	0.82 ± 0.20
Group 2
16	L	Plate	15,107	1.51 x 10⁵	2298	3.21	0.10	2.82	1.69	4.69	1.74
	R	Cut	5350	8.92 Â¥ 10⁴	490	1.85	0.06
17	R	Plate	12,677	6.34 Â¥ 10⁴	5028	6.23	0.20	1.00	0.85	1.99	1.15
	L	Cut	12,706	7.47 Â¥ 10⁴	2532	5.40	0.17
18	L	Plate	14,148	1.09 Â¥ 10⁵	2936	3.98	0.13	3.42	2.90	4.02	1.17
	R	Cut	4137	3.76 Â¥ 10⁴	731	3.39	0.11
19	R	Plate	26,127	7.68 Â¥ 10⁴	13,457	10.67	0.34	2.84	1.42	6.61	1.98
	L	Cut	9189	5.41 Â¥ 10⁴	2034	5.38	0.17
20	L	Plate	24,881	9.95 Â¥ 10⁴	8709	7.66	0.25	3.39	3.25	3.39	1.01
	R	Cut	7334	3.06 Â¥ 10⁴	2570	7.56	0.24
								2.69 ± 0.99	2.02 ± 1.7	4.14 ± 1.70	1.41 ± 0.42

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

Twenty matched pairs of fresh-frozen cadaveric upper extremities were obtained. The extremities were allowed to thaw overnight, and the radii were harvested, with care taken to remove all soft tissue but to preserve as much of the periosteum as possible. All specimens were visually examined to exclude signs of previous fracture or bone defect. No external osseous abnormalities were seen, but it was not known whether any generalized bone disorders were present.

In Group 1, ten matched pairs of radii were randomly designated to have one bone in each pair remain intact and the other receive a segmental defect to simulate an osteocutaneous radial forearm donor-site defect. The defect was eight centimeters in length and comprised 50 percent of the cross-sectional area of the bone. The proximal extent of the ostectomy began at the most distal point of insertion of the pronator teres. The bone section was removed with an oscillating saw in a right-angled manner, with great care taken not to cut past the appropriate depth and thus notch the remaining portion of the radius (Fig. 1-A). Notching was prevented by making the longitudinal cut first and then placing metal ribbons into the proximal and distal aspects of the cut before the transverse cuts were performed. The transverse cuts therefore were stopped by the metal ribbons, and cutting past the appropriate depth was avoided. Eight centimeters was selected as the length of resection because it represents a length of bone that can be easily resected between the pronator teres and brachioradialis insertions, where the bone is vascularized by the radial artery. It also is a common length of bone graft that is used clinically.

Group 2 consisted of ten matched pairs of radii in which a defect that was identical to that described for Group 1 was created in both bones of each pair. However, a twelve-hole, 3.5-millimeter stainless-steel dynamic compression plate (Synthes USA, Paoli, Pennsylvania) was applied to the dorsal-ulnar surface of one ostectomized radius in each pair to span the defect (Fig. 1-B). The dorsal-ulnar surface was chosen because a plate located in this position has excellent soft-tissue coverage and does not involve the volar surface of the forearm, where the fasciocutaneous flap is elevated and which is often covered with a skin graft in patients. The plate was positioned so that at least two bicortical screws were distal to the ostectomy defect and at least two bicortical screws were proximal to it. The screw-holes were all filled with either a unicortical or a bicortical 3.5-millimeter cortical bone screw (Synthes USA). The four or five screws placed in the region of the ostectomy were unicortical screws. In both Group 1 and Group 2, five matched pairs of radii were tested to failure in torsion and five matched pairs were tested to failure in four-point bending.

A pair of custom-made end fixtures for torsion tests were designed and machined to clamp both the distal and the proximal end of the radius (Fig. 2). The end fixtures consisted of an aluminum cup with a pedestal and a solid hexagonal cylinder screwed onto the pedestal. To mount the radius, the distal end was first placed at the center of the cup with the use of four screws. While the specimen was held vertically in the cup, polymethylmethacrylate mixed with its monomer was poured into the cup until it was full. After the polymethylmethacrylate had hardened, the same procedure was repeated to fix the proximal end of the radius. All specimens were covered with moist gauze to prevent desiccation during preparation. For torsion-testing, a torsion-testing machine (model MCMT-23054; Tinius Olsen, Willow Grove, Pennsylvania) equipped with a speed controller (model 2M171E; Dayton DC, Niles, Illinois) was used along with a data acquisition system. On the torsion-testing machine, the cup at the proximal end of the radius was rigidly clamped while the distal end was forced to twist. Before each test, the specimen was twisted 5 degrees and then released back to the initial position to observe any slip between the specimen and the fixture system.

To determine the torsional stiffness, each test specimen was twisted 5.0, 7.5, 10.0, 12.5, and 15.0 degrees and the corresponding torques (T0) were recorded. The torque after each angle of twist was released, and the specimen was returned to its initial position. A nonphysiologically slow rate of 1 degree per second was used. A graph between torque and angle of twist was drawn, and the slope of the straight line was calculated as the torsional stiffness. All of the plots between torque and angle of twist up to 15 degrees exhibited a linear relationship between torque and angle of twist. Specimens were tested to failure to determine failure torque and angle of twist at failure. The specimens were visually examined to determine the location of the resultant spiral fracture. Photographs were made after failure, with special reference to the fracture patterns.

Five matched pairs from Group 1 and five matched pairs from Group 2 were loaded to failure in torsion, and their stiffness, torque at failure, maximum angle of twist, and torsional toughness were measured. Torsional toughness, or the energy absorbed up to failure under torsion-loading, was measured as the area under the torque-versus-angle-of-twist curve (Fig. 4-A and Fig. 4-B).

For the four-point-bending tests, load and support points were marked on the radius. To ensure that load and support points were identical for each of the specimens, the radii of a pair were aligned parallel to each other during the marking process. The load points were chosen 2.5 centimeters on either side of the center of the cut, whereas the support points were 5.0 centimeters on either side of the center of the cut. These load and support points were chosen so that the four-point-bending system applied constant bending moment across the central region of the ostectomized section of the radius. With this system, the failure tests provided the failure bending moment of the central region of the ostectomized site of the radius for direct comparison with the corresponding failure bending moments at the same sites of the intact radii and the radii fixed with a plate. Special steel clamps at the ends of the radius held the bone and prevented it from rolling while being subjected to bending forces (Fig. 3. The computer-controlled tester (remanufactured model Instron TT-D; MTS Systems, Minneapolis, Minnesota) with a capacity of 90,000 newtons was used to carry out the four-point-bending tests along with a data acquisition system. The loading was displacement-controlled at a speed of one millimeter per second. In Groups 1 and 2, the cut side of the bone was directed downward during bending. The intact radii and the cut radii were placed in the same orientation for testing.

The radii were loaded to failure with use of four-point bending, and the failure bending moment, bending toughness, and deflections both at the center and at load points were measured. The specimens were visually examined to determine the site of the fracture. The bending stiffness was calculated with use of Mohr's area-moment method¹². Equating the deflection at the load point relative to the tangent at the center to the first moment of area of the M/EI diagram¹² about the load point, the flexural rigidity EI was calculated. With substitution of this result in the calculation of the area of the M/EI diagram between the center and the load point, the bending rotation at the load point was calculated and used to determine the bending stiffness.

The bending toughness was calculated with use of the so-called area-moment method and the area under the curve in the bending-moment-versus-center-displacement graph for each radius (Fig. 5-A and Fig. 5-B).

With a sample size of five for each group, a total of twenty matched pairs of fresh-frozen cadaveric radii were utilized as a means of standardizing the data and allowing twelve direct comparisons. The paired Student t test was used to analyze the data. Additional post hoc power calculations with use of the observed variance estimates and a clinically meaningful difference in strength (ostectomized radii fixed with a plate having twice the strength of radii with ostectomy alone and intact radii having twice the strength of ostectomized radii) were carried out for all twelve comparisons.

Results

Introduction | Materials and Methods | Results | Discussion | References

In the five matched pairs in which one bone was ostectomized and the other was left intact (Group 1), the mean failure torque of the ostectomized radius was 18 percent (range, 11 to 36 percent) of that of the intact radius. This difference amounted to a mean decrease in strength of 82 percent (p = 0.016). The mean torsional stiffness of the ostectomized radius was 24 percent of that of the intact radius, for a mean decrease in torsional stiffness of 76 percent (p = 0.053, P = 0.50). The torsional toughness of the ostectomized radius ranged from 0.30 to 1.35 times that of the intact radius in the five matched pairs. The mean torsional toughness of the ostectomized radius was 53 percent of that of the intact radius (p = 0.168, P = 0.71) (Table I and Table II).

In the five matched pairs in which one bone had an ostectomy only and the other had an ostectomy and fixation with a plate (Group 2), the mean failure torque of the radius that had fixation with a plate was 3.95 times (range, 2.99 to 5.37 times) that of the radius that had an ostectomy only (p = 0.002). Application of a plate over the ostectomy defect resulted in a significant (p = 0.01) increase in stiffness as reflected by the slope of the torque-versus-angle-of-twist curves for the ostectomized specimens. The mean stiffness of the radii with the ostectomy and the plate was 3.2 times that of the radius with the ostectomy only. With the numbers available, the torsional toughness and the maximum angle of twist were not significantly different between the radii that had the ostectomy alone and those that had the ostectomy and fixation with a plate, although the difference in the torsional toughness approached significance (p = 0.06, P = 0.57). The radii that had the ostectomy and fixation with a plate were 3.7 times tougher than the radii that had the ostectomy only (Table I).

In Group 1, the failure bending moment of the ostectomized bone ranged from 14 to 35 percent of that of the intact radius. Thus, the ostectomized bone was a mean of 76 percent weaker than the intact bone (p = 0.013). The difference between the bending stiffness of the ostectomized radii and that of the intact radii was a mean of 19 percent (range, 13 to 28 percent) (p = 0.001). Similarly, the mean bending toughness for the ostectomized radii was 27 percent (range, 14 to 71 percent) of that of the intact radii (p = 0.025, Table III and Table IV).

In Group 2, the failure moments of the radii that had the ostectomy alone ranged from 29 to 100 percent of that of the radii that had the ostectomy and fixation with a plate. The radii that had the ostectomy alone were a mean of 63 percent weaker than the radii that had the ostectomy and fixation with a plate (p = 0.027). The bending stiffness of the radii that had the ostectomy and fixation with a plate ranged from 0.84 to 3.25 times (mean, 2.0 times) that of the radii that had the ostectomy alone (p = 0.057, P = 0.87). The difference in bending toughness only approached significance in Group 2 (p = 0.058, P = 0.44). The bending toughness of the radii that had fixation with a plate was a mean of 4.1 times (range, 2.0 to 6.6 times) that of the radii that had the ostectomy only (Table IV). With the numbers available, no significant difference could be detected in the mean center displacement in Group 1 or Group 2 (Table III).

The fracture patterns were noted in all radii in Groups 1 and 2. In the intact specimens, torsion produced a spiral fracture near the radial tuberosity and four-point bending produced a short oblique fracture at the midpoint of the tested segment. In the specimens that had the ostectomy only, torsion produced a short spiral fracture passing through one of the ends of the ostectomy defect. In four-point bending, a short oblique fracture was seen near the center of the applied loads. In the ostectomized bones that had been fixed with a plate, a spiral fracture (in torsion testing) or a short oblique fracture (in four-point bending) was seen passing either through screw-holes or through the ends of the ostectomized segment.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

It is well documented in the biomechanical and orthopaedic literature that when a tubular structure, such as the radius, is converted from a closed section to an open section by removal of a segment its torsional strength, stiffness, and energy-absorbing capacity are markedly reduced^7,11,13. Furthermore, Brooks et al. demonstrated that even a defect as small as a drill-hole can decrease the energy-absorbing capacity of bone prior to fracture by as much as 55 percent as a result of both the stress concentration, or stress-riser, created in the bone in the area of the hole, or defect, and the decreased amount of material remaining to share the load⁴. A three-dimensional finite element analysis of radii from which osteocutaneous vascularized bone graft had been harvested predicted a 65 percent decrease in torsional strength in the radius with the ostectomy compared with the intact radius⁵. The greater decrease in torsional strength (82 percent) that we found may be due in part to the homogeneous and isotropic assumptions made concerning the bone in the finite element analysis.

This open-section effect as it applies to the ostectomized radius has been examined in two studies. Swanson et al., in a study of embalmed cadaveric radii with a bone defect that was nine centimeters in length and one-third of the thickness of the bone, found that the strength was decreased 76 percent in four-point bending²⁷. Meland et al. used fresh-frozen sheep tibiae to model the human radius¹⁸. With bone defects of various lengths and up to 50 percent of the cross-sectional area in depth, the strength in torsion was decreased up to 85 percent.

In our study, fresh-frozen (nonpreserved) radii from human cadavera with a defect that was eight centimeters in length and 50 percent of the cross-sectional area in depth were tested in both four-point bending and torsion. We found a mean decrease in strength of 76 percent and 82 percent, respectively. These findings are comparable with those of the previously mentioned studies^18,27; however, since fresh human specimens were utilized, these numbers are likely to be the most accurate representation of the strength lost in an ostectomized radius when right-angled cuts are used. This decrease in strength in both four-point bending and torsion is much larger than a 50 percent reduction; therefore, doubling the strength with the application of a plate, as was assumed in our original hypothesis, would result in a return of only 36 percent of the strength of the intact bone in torsion and 48 percent of the strength of the intact bone in four-point bending.

Meland et al. thought that the decrease in strength and the resultant rate of fracture of the ostectomized radius were insurmountable problems and recommended the use of other sources of vascularized bone in the reconstruction of defects requiring composite grafts¹⁸. Reports in the literature have confirmed unacceptable morbidity when fractures of the donor site occur. These sequelae include decreased pronation and supination and as much as a 50 percent loss of pinch and grip strength^2,3,15. Furthermore, some of these fractures have required extensive interventions to obtain union. Inglefield and Kolhe reported a fracture at the donor site of an osteocutaneous radial forearm graft that required grafting with an osteocutaneous vascularized fibula and external fixation to heal¹⁴. While we agree that the decrease in strength and the rate of fracture of ostectomized radii result in unacceptable donor-site morbidity, we believe that this problem can be overcome by internally fixing the ostectomy defect with a plate and screws.

The specimens in the present study that had fixation with a plate (Group 2) had a 4.0-fold increase in torsional strength and a 2.7-fold increase in four-point-bending strength compared with the radii that had an ostectomy only. The mean four-point-bending strength ratio (2.7) of the specimens that had plate fixation would have been larger had it not been for the data regarding one specimen in which the strength ratio was essentially 1.0. In light of the consistent results in the other radii, we believe that this specimen had a testing error of uncertain etiology. Despite this specimen, these numbers still approach the difference in strength between the intact radius and the ostectomized radius (Group 1). Because of this strength advantage, we believe that internal fixation with a plate and screws would prevent most fractures and decrease donor-site morbidity to acceptable levels. The increase in strength associated with application of a plate was substantially greater than the doubling of strength that we believed, according to our initial hypothesis, would be required to prevent most pathological fractures.

However, use of the plates considerably increased stiffness compared with that of the bones that had the ostectomy alone. This increase in stiffness had the effect of minimizing the increase in energy absorption required to fracture the specimens that were fixed with a plate compared with those that had the ostectomy only. Thus, bones that are fixed with a plate may be less well protected against traumatic energy overload.

The increase in stiffness raises the clinical concern of stress-shielding and eventual failure of the construct. Numerous studies have implied that stress-shielding occurs with rigid plate fixation, resulting in osseous loss under the plate^1,4,29,31. The long-term effect of this is unknown. Some studies have indicated that bone loss due to stress-shielding diminishes or stops at three to six months in animals^19,28. A study of humans performed with computerized tomography showed that the stress-shielding effects in human tibiae were less than those previously reported in studies of animals, and it did not show an increase in such effects after a one-year duration³⁰. Placement of the plate also results in decreased vascularity and subsequent porosity of the bone directly under the plate²⁰. This vascular effect may be more important in terms of the temporary weakening of the bone that is seen with plate fixation, and it has led to the development of limited-contact plates.

Both stress-shielding and altered vascularity secondary to placement of the plate are of concern in the long term for patients who have prophylactic application of a plate after removal of an osteocutaneous radial forearm free flap. Plates that are less stiff, and plates that have less contact with the bone, might be advantageous in these patients. The long-term remodeling associated with plate fixation after an ostectomy, as opposed to that associated with plate fixation of a fracture, is unknown.

Since the application of a plate increases the residual strength of the radius, the surgeon is allowed to take a larger piece of bone, which may permit the reconstruction of a larger defect. In this study, we took only an eight-centimeter length of bone because in very short radii we found it difficult to obtain adequate screw fixation (that is, engagement of four cortices distal to the defect) if a longer segment of bone was harvested. However, we believe that up to a twelve-centimeter length of bone may be harvested from longer radii and still allow adequate screw fixation distally, but the strength of the construct cannot necessarily be inferred from the present study. Also, with plate fixation, 50 percent of the cross-sectional area can be harvested, resulting in a thicker piece of implantable bone. In previous studies^25,26,32, investigators have recommended that no more than 40 percent of the cross-sectional area be harvested because of the risk of fracture.

In our study, 3.5-millimeter dynamic compression plates were used to prophylactically fix the ostectomized radius, and all of the screw-holes were filled with a screw even if only unicortical purchase was obtained. We performed a preliminary study with use of formalin-fixed human cadaveric radii in which we found that filling all of the screw-holes conferred greater torsional strength than either not filling any screw-holes over the defect or filling only every other screw-hole over the defect in the ostectomized radius. On the basis of that study, we filled all of the screw-holes over the site of the ostectomy with unicortical screws. It is possible that the ostectomized radius might have been stronger in four-point bending with fewer screws or with no screws in the area of the defect, but this construct was not tested. The additional screw-holes, however, add more areas of stress concentration at which failure can occur and they further decrease the vascularity of the bone so they may be deleterious in the long term.

Although so-called boat-shaped, or beveled, cuts, rather than right-angled cuts, have been recommended as a means to reduce fracture rates, the actual strength advantage provided by this variation is quite small¹². The beveled cuts have a theoretical advantage of increasing strength by diminishing the sharp corners resulting from the right-angled cuts, which serve as stress-risers where fractures may be initiated as was seen in our specimens that had plate fixation.

We perceive that decreased torsional strength is the most important parameter contributing to fracture of the radius at the donor site of an osteocutaneous radial forearm free flap. This belief is based on our clinical experience in which all of the fractures at the donor sites of osteocutaneous flaps were spiral fractures that often occurred with minimal supination or pronation force. Although our experimental data showed the strength in four-point bending to be equally affected by the donor-site defect, it is possible that clinically the intact ulna would protect a radius subjected to four-point-bending force but would confer less protection against torsional stress.

The weaknesses of this study include the fact that the specimens were from the cadavera of elderly individuals and we had no knowledge with regard to generalized bone disorders or hormonal therapy. However, since the experiments all involved matched pairs from the same cadaver and the results were expressed as ratios, we expect that changes due to these conditions were largely negated. The four-point-bending setup placed the greatest load at the center of the ostectomy site. If the maximum load had been applied elsewhere, such as at the end of the ostectomy defect, the fracture probably would have occurred at that point due to the stress-riser and the failure ratios may have differed somewhat.

In conclusion, we believe that osteocutaneous radial forearm free flaps require internal fixation of the donor-site defect to reliably prevent the sequela of fracture of the radius. Our data supports the use of prophylactic internal fixation of the donor-site defect to provide increased strength and decreased morbidity. Since donor-site morbidity has been the major disadvantage of the osteocutaneous radial forearm free flap, utilization of this technique should pave the way for more widespread use of this versatile flap in various types of microvascular reconstruction.

References

Introduction | Materials and Methods | Results | Discussion | References

Akeson, W. H.Woo, S. L.Rutherford, L.Coutts, R. D.Gonsalves, M., and Amiel, D.: The effects of rigidity of internal fixation plates on long bone remodeling. A biomechanical and quantitative histological study. Acta Orthop. Scandinavica,47: 241-249, 1976.47241 1976

Bardsley, A. F.Soutar, D. S.Elliot, D., and Batchelor, A. G.: Reducing morbidity in the radial forearm flap donor site. Plast. and Reconstr. Surg.,86: 287-294, 1990.86287 1990

Boorman, J. G.; Brown, J. A.; and Sykes, P. J.: Morbidity in the forearm flap donor arm. British J. Plast. Surg.,40: 207-212, 1987.40207 1987

Brooks, D. B.; Burstein, A. H.; and Frankel, V. H.: The biomechanics of torsional fractures. The stress concentration effects of a drill hole. J Bone Joint Surg,52-A: 507-514, April 1970.52-A507 1970

Carter, D. R.Shimaoka, E. E.Harris, W. H.Gates, E. I.Caler, W. E., and McCarthy, J. C.: Changes in long-bone structural properties during the first 8 weeks of plate implantation. J. Orthop. Res.,2: 80-89, 1984.280 1984 [PubMed]

Chicarilli, Z. N.; Ariyan, S.; and Cuono, C. B.: Free radial forearm flap versatility for the head and neck and lower extremity. J. Reconstr. Microsurg.,2: 221-228, 1986.2221 1986 [PubMed]

Clark, C. R.Morgan, C.Sonstegard, D. A., and Matthews, L. S.: The effect of biopsy-hole shape and size on bone strength. J Bone Joint Surg,59-A: 213-217, March 1977.59-A213 1977

Fidler, M.: Prophylactic internal fixation of secondary neoplastic deposits in long bone. British Med. J.,1: 341-343, 1973.1341 1973

Fidler, M.: Incidence of fracture through metastases in long bones. Acta Orthop. Scandinavica,52: 623-627, 1981.52623 1981

Fenton, O. M., and Roberts, J. O.: Improving the donor site of the radial forearm flap. British J. Plast. Surg.,38: 504-505, 1985.38504 1985

Frankel, V. H., and Burstein, A. H.: Load capacity of tubular bone. In Biomechanics and Related Bio-Engineering Topics, pp. 381-396. Edited by R. M. Kenedi. New York, Pergamon Press, 1965.

Higdon, A.; Olson, E. H.; Stiles, W. B.; Weese, J. A.; and Riley, W. F.: Flexural loading: deflections. In Mechanics of Materials, pp. 389-399. New York, John Wiley, 1995.

Hipp, J. A.Edgerton, B. C.An, K. N.;, and Hayes, W. C.: Structural consequences of transcortical holes in long bones loaded in torsion. J. Biomech.,23: 1261-1268, 1990.231261 1990

Inglefield, C. J., and Kolhe, P. S.: Fracture of the radial forearm osteocutaneous donor site. Ann. Plast. Surg.,33: 638-643, 1994.33638 1994 [PubMed]

Juretic, M.; Car, M.; and Zambelli, M.: The radial forearm free flap: our experience in solving donor site problems. J. Cranio-Maxillo-Fac. Surg.,20: 184-186, 1992.20184 1992

McBroom, R. J.; Cheal, E. J.; and Hayes, W. C.: Strength reductions from metastatic cortical defects in long bones. J. Orthop. Res.,6: 369-378, 1988.6369 1988 [PubMed]

McGregor, A. D.: The free radial forearm flap - the management of the secondary defect. British J. Plast. Surg.,40: 83-85, 1987.4083 1987

Meland, N. B.Maki, S.Chao, E. Y., and Rademaker, B.: The radial forearm flap: a biomechanical study of donor-site morbidity utilizing sheep tibia. Plast. and Reconstr. Surg.,90: 763-773, 1992.90763 1992

Moyen, B. J.-L.Lahey, P. J., Jr.Weinberg, E. H., and Harris, W. H.: Effects on intact femora of dogs of the application and removal of metal plates. A metabolic and structural study comparing stiffer and more flexible plates. J Bone Joint Surg,60-A: 940-947, Oct 1978.60-A940 1978

Perren, S. M.Cordey, J.Rahn, B. A.Gautier, E., and Schneider, E.: Early temporary porosis of bone induced by internal fixation implants. A reaction to necrosis, not to stress protection?. Clin. Orthop.,232: 139-151, 1988.232139 1988 [PubMed]

Ryan, J. R.; Rowe, D. E.; and Salciccioli, G. G.: Prophylactic internal fixation of the femur for neoplastic lesions. J Bone Joint Surg,58-A: 1071-1074, Dec 1976.58-A1071 1976

Smith, A. A.Bowen, C. V.Rabczak, T., and Boyd, J. B.: Donor site deficit of the osteocutaneous radial forearm flap. Ann. Plast. Surg.,32: 372-376, 1994.32372 1994 [PubMed]

Song, R.Gao, Y.Song, Y.Yu, Y., and Song, Y.: The forearm flap. Clin. Plast. Surg.,21-26, 1982.21 1982

Soutar, D. S.Scheker, N. S.Tanner, N. S., and McGregor, I. A.: The radial forearm flap: a versatile method for intra-oral reconstruction. British J. Plast. Surg.,36: 1-8, 1983.361 1983

Soutar, D. S., and McGregor, I. A.: The radial forearm flap in intraoral reconstruction: the experience of 60 consecutive cases. Plast. and Reconstr. Surg.,78: 8, 1986.788 1986

Swanson, E.; Boyd, J. B.; and Manktelow, R. T.: The radial forearm flap: reconstructive applications and donor-site defects in 35 consecutive patients. Plast. and Reconstr. Surg.,85: 258-266, 1990.85258 1990

Swanson, E.; Boyd, J. B.; and Mulholland, R. S.: The radial forearm flap: a biomechanical study of the osteotomized radius.. Plast. and Reconstr. Surg.,85: 267-272, 1990.85267 1990

Terjesen, T., and Benum, P.: The stress-protecting effect of metal plates on the intact rabbit tibia. Acta Orthop. Scandinavica,54: 810-818, 1983.54810 1983

Terjesen, T., and Apalset, K.: The influence of different degrees of stiffness of fixation plates on experimental bone healing. J. Orthop. Res.,6: 293-299, 1988.6293 1988 [PubMed]

Terjesen, T.; Nordby, A.; and Arnulf, V.: The extent of stress-protection after plate osteosynthesis in the human tibia. Clin. Orthop.,207: 108-112, 1986.207108 1986 [PubMed]

Uhthoff, H. K., and Finnegan, M.: The effects of metal plates on post-traumatic remodelling and bone mass. J Bone Joint Surg,65-B(1): 66-71, 1983.65-B(1)66 1983

Urken, M. L.: The restoration or preservation of sensation in the oral cavity following ablative surgery. Arch. Otolaryngol.,121: 607-612, 1995.121607 1995

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-A:: Illustration of the volar view of the ostectomized radius, showing the ostectomy defect between the insertions of the pronator teres and the brachioradialis muscles.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-B:: Illustration of the dorsal view of the ostectomized radius fixed with a plate. Note that the plate is well contoured to the dorsal-ulnar surface of the bone.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-A:: Figs. 4-A and 4-B: Graphs showing the torque-versus-angle-of-twist curves. One inch-pound = 0.113 newton-meter.
Fig. 4-A: Specimen 3 from Group 1.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-B:: Specimen 11 from Group 2.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-A:: Figs. 5-A and 5-B: Graphs showing the moment-versus-center-displacement curves.
Fig. 5-A: Specimen 7 from Group 1.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-B:: Specimen 18 from Group 2.

View Larger

TABLE I: Mean Torsion Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1 (intact vs. ostectomized)	5.7 ± 2.6*	4.2 ± 2.2*	1.9 ± 1.1*
P value	0.016	0.053	0.168
Power	0.75	0.50	0.71
Group 2 (ostectomized vs. plate)	4.0 ± 1.0*	3.2 ± 1.0*	3.7 ± 3.2*
P value	0.002	0.01	0.06
Power	0.99	0.93	0.57

Add to My JBJS

TABLE I: Mean Torsion Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1 (intact vs. ostectomized)	5.7 ± 2.6*	4.2 ± 2.2*	1.9 ± 1.1*
P value	0.016	0.053	0.168
Power	0.75	0.50	0.71
Group 2 (ostectomized vs. plate)	4.0 ± 1.0*	3.2 ± 1.0*	3.7 ± 3.2*
P value	0.002	0.01	0.06
Power	0.99	0.93	0.57

View Larger

TABLE II: Individual Torsion Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/degree)	Toughness(Nmm)	Angle of Twist at Failure(degrees)	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1
1	L	Intact	4632	225	3142	60.5	9.11	5.00	3.37
	R	Cut	508	45	932	92.0
2	R	Intact	22,371	263	3783	21.5	3.92	1.45	0.74
	L	Cut	5706	181	5110	65.0
3	L	Intact	30,280	1116	6489	23.0	4.96	4.77	1.04
	R	Cut	6101	234	6255	75.5
4	R	Intact	10,056	321	5928	59.0	7.57	7.13	2.61
	L	Cut	1328	45	2269	114.5
5	L	Intact	23,162	872	12,718	50.5	2.75	2.68	1.89
	R	Cut	8417	325	6727	66.5	5.66 ± 2.62*	4.20 ± 2.19*	l.93 ± 1.09*
Group 2
11	R	Plate	11,976	389	5920	44.25	3.79	4.32	1.44
	L	Cut	3164	90	4103	107.0
12	L	Plate	15,705	299	14,281	70.0	2.99	1.89	3.75
	R	Cut	5254	158	3808	56.5
13	R	Plate	13,841	388	6384	41.5	4.45	3.92	1.46
	L	Cut	3107	99	4361	99.25
14	L	Plate	7118	154	2994	35.0	3.15	2.44	2.60
	R	Cut	2260	63	1152	39.0
15	L	Plate	8191	217	7016	62.0	5.37	3.44	9.14
	R	Cut	1525	63	768	40.0	3.95 ± 0.98*	3.20 ± 1.01*	3.68 ± 3.20*

Add to My JBJS

TABLE II: Individual Torsion Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/degree)	Toughness(Nmm)	Angle of Twist at Failure(degrees)	Strength Ratio	Stiffness Ratio	Toughness Ratio
Group 1
1	L	Intact	4632	225	3142	60.5	9.11	5.00	3.37
	R	Cut	508	45	932	92.0
2	R	Intact	22,371	263	3783	21.5	3.92	1.45	0.74
	L	Cut	5706	181	5110	65.0
3	L	Intact	30,280	1116	6489	23.0	4.96	4.77	1.04
	R	Cut	6101	234	6255	75.5
4	R	Intact	10,056	321	5928	59.0	7.57	7.13	2.61
	L	Cut	1328	45	2269	114.5
5	L	Intact	23,162	872	12,718	50.5	2.75	2.68	1.89
	R	Cut	8417	325	6727	66.5	5.66 ± 2.62*	4.20 ± 2.19*	l.93 ± 1.09*
Group 2
11	R	Plate	11,976	389	5920	44.25	3.79	4.32	1.44
	L	Cut	3164	90	4103	107.0
12	L	Plate	15,705	299	14,281	70.0	2.99	1.89	3.75
	R	Cut	5254	158	3808	56.5
13	R	Plate	13,841	388	6384	41.5	4.45	3.92	1.46
	L	Cut	3107	99	4361	99.25
14	L	Plate	7118	154	2994	35.0	3.15	2.44	2.60
	R	Cut	2260	63	1152	39.0
15	L	Plate	8191	217	7016	62.0	5.37	3.44	9.14
	R	Cut	1525	63	768	40.0	3.95 ± 0.98*	3.20 ± 1.01*	3.68 ± 3.20*

View Larger

TABLE III: Mean Four-Point-Bending Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1 (intact vs. ostectomized)	4.2 ± 1.8*	5.2 ± 1.5*	3.7 ± 2.1*	0.82 ± 0.20*
P value	0.013	0.001	0.025	0.15
Power	0.81	1.00	0.72	0.99
Group 2 (ostectomized vs. plate)	2.7 ± 1.0*	2.0 ± 1.0*	4.1 ± 1.7*	1.4 ± 0.42*
P value	0.027	0.057	0.058	0.16
Power	0.83	0.87	0.44	0.92

Add to My JBJS

TABLE III: Mean Four-Point-Bending Test Results for the Two Groups

*The values are given as the mean and the standard deviation.

	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1 (intact vs. ostectomized)	4.2 ± 1.8*	5.2 ± 1.5*	3.7 ± 2.1*	0.82 ± 0.20*
P value	0.013	0.001	0.025	0.15
Power	0.81	1.00	0.72	0.99
Group 2 (ostectomized vs. plate)	2.7 ± 1.0*	2.0 ± 1.0*	4.1 ± 1.7*	1.4 ± 0.42*
P value	0.027	0.057	0.058	0.16
Power	0.83	0.87	0.44	0.92

View Larger

TABLE IV: Individual Four-Point-Bending Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/cGy)	Toughness(Nmm)	Center Displacement at Failure(mm)	Rotation at Failure(cGy)	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1
6	L	Intact	18,992	1.73 x 10⁵	3090	3.48	0.11	2.85	5.97	1.42	0.48
	R	Cut	6670	2.90 x 10⁴	2183	7.27	0.23
7	R	Intact	22,687	1.62 x 10⁵	4736	4.19	0.14	3.68	4.21	3.17	0.84
	L	Cut	6164	3.85 x 10⁴	1493	4.98	0.16
8	L	Intact	9963	1.01 x 10⁵	1396	3.07	0.099	3.93	4.79	3.45	0.85
	R	Cut	2533	2.11 x 10⁴	405	3.62	0.12
9	R	Intact	36,034	1.18 x 10⁵	5717	9.48	0.305	7.31	7.42	7.06	0.99
	L	Cut	4930	1.59 x 10⁴	810	9.55	0.31
10	L	Intact	39,006	1.44 x 10⁵	5572	8.41	0.27	3.32	3.56	3.45	0.92
	R	Cut	11,741	4.05 x 10⁴	1613	9.15	0.29
								4.22 ± 1.78*	5.19 ± 1.53*	3.71 ± 2.06*	0.82 ± 0.20
Group 2
16	L	Plate	15,107	1.51 x 10⁵	2298	3.21	0.10	2.82	1.69	4.69	1.74
	R	Cut	5350	8.92 Â¥ 10⁴	490	1.85	0.06
17	R	Plate	12,677	6.34 Â¥ 10⁴	5028	6.23	0.20	1.00	0.85	1.99	1.15
	L	Cut	12,706	7.47 Â¥ 10⁴	2532	5.40	0.17
18	L	Plate	14,148	1.09 Â¥ 10⁵	2936	3.98	0.13	3.42	2.90	4.02	1.17
	R	Cut	4137	3.76 Â¥ 10⁴	731	3.39	0.11
19	R	Plate	26,127	7.68 Â¥ 10⁴	13,457	10.67	0.34	2.84	1.42	6.61	1.98
	L	Cut	9189	5.41 Â¥ 10⁴	2034	5.38	0.17
20	L	Plate	24,881	9.95 Â¥ 10⁴	8709	7.66	0.25	3.39	3.25	3.39	1.01
	R	Cut	7334	3.06 Â¥ 10⁴	2570	7.56	0.24
								2.69 ± 0.99	2.02 ± 1.7	4.14 ± 1.70	1.41 ± 0.42

Add to My JBJS

TABLE IV: Individual Four-Point-Bending Test Results

*The values are given as the mean and the standard deviation.

Pair	Side	Procedure	Strength(Nmm)	Stiffness(Nmm/cGy)	Toughness(Nmm)	Center Displacement at Failure(mm)	Rotation at Failure(cGy)	Strength Ratio	Stiffness Ratio	Toughness Ratio	Center Displacement Ratio
Group 1
6	L	Intact	18,992	1.73 x 10⁵	3090	3.48	0.11	2.85	5.97	1.42	0.48
	R	Cut	6670	2.90 x 10⁴	2183	7.27	0.23
7	R	Intact	22,687	1.62 x 10⁵	4736	4.19	0.14	3.68	4.21	3.17	0.84
	L	Cut	6164	3.85 x 10⁴	1493	4.98	0.16
8	L	Intact	9963	1.01 x 10⁵	1396	3.07	0.099	3.93	4.79	3.45	0.85
	R	Cut	2533	2.11 x 10⁴	405	3.62	0.12
9	R	Intact	36,034	1.18 x 10⁵	5717	9.48	0.305	7.31	7.42	7.06	0.99
	L	Cut	4930	1.59 x 10⁴	810	9.55	0.31
10	L	Intact	39,006	1.44 x 10⁵	5572	8.41	0.27	3.32	3.56	3.45	0.92
	R	Cut	11,741	4.05 x 10⁴	1613	9.15	0.29
								4.22 ± 1.78*	5.19 ± 1.53*	3.71 ± 2.06*	0.82 ± 0.20
Group 2
16	L	Plate	15,107	1.51 x 10⁵	2298	3.21	0.10	2.82	1.69	4.69	1.74
	R	Cut	5350	8.92 Â¥ 10⁴	490	1.85	0.06
17	R	Plate	12,677	6.34 Â¥ 10⁴	5028	6.23	0.20	1.00	0.85	1.99	1.15
	L	Cut	12,706	7.47 Â¥ 10⁴	2532	5.40	0.17
18	L	Plate	14,148	1.09 Â¥ 10⁵	2936	3.98	0.13	3.42	2.90	4.02	1.17
	R	Cut	4137	3.76 Â¥ 10⁴	731	3.39	0.11
19	R	Plate	26,127	7.68 Â¥ 10⁴	13,457	10.67	0.34	2.84	1.42	6.61	1.98
	L	Cut	9189	5.41 Â¥ 10⁴	2034	5.38	0.17
20	L	Plate	24,881	9.95 Â¥ 10⁴	8709	7.66	0.25	3.39	3.25	3.39	1.01
	R	Cut	7334	3.06 Â¥ 10⁴	2570	7.56	0.24
								2.69 ± 0.99	2.02 ± 1.7	4.14 ± 1.70	1.41 ± 0.42