The effect of hole size on failure loads of the distal humerus fenestration in Outerbridge-Kashiwagi arthroplasty: A biomechanical study

Abstract

Objective

The aim of this study was to biomechanically assess the effect of humeral-fenestration size in the Outerbridge-Kashiwagi arthroplasty on the ultimate failure load of the distal humerus in a synthetic bone model.

Methods

We biomechanically tested the influence of different humeral-fenestration sizes on the failure load of the distal humerus in Outerbridge-Kashiwagi arthroplasty. A total of 50 synthetic humerus models were divided into 5 groups based on the fenestration size: 10 mm, 12 mm, 15 mm, 18 mm, and 20 mm. All the samples were randomly assigned to receive either axial or anteroposterior (AP) loading and then loaded to failure at a rate of 2 mm/min on a material testing machine. The data regarding ultimate failure loads under the axial and AP loading were analyzed.

Results

Under the AP loading, the mean ultimate failure loads of the 18 mm and 20 mm groups were lower than those of the other groups. Under the axial loading, the mean ultimate failure load of the 10 mm group was significantly greater than that of the 15 mm, 18 mm, and 20 mm groups. Additionally, the ultimate failure load of the 20 mm group was significantly lower than that of the 12 mm, 15 mm, and 18 mm groups.

Conclusion

The distal humeral fenestrations with a size greater than 18 mm may offer poor biomechanical properties in the Outerbridge-Kashiwagi ulnohumeral arthroplasty.

The Outerbridge-Kashiwagi procedure, also known as the ulnohumeral arthroplasty technique, is an effective method to relieve pain and improve the motility of patients with elbow osteoarthritis (1–4). The procedure includes the removal of osteophytes from the olecranon fossa and olecranon tip, fenestration through the olecranon fossa, and removal of loose bodies in the anterior compartment of the elbow joint. Considering the higher rate of prosthetic loosening after long-term use, the Outerbridge-Kashiwagi procedure is attractive for relatively younger patients. Several studies have demonstrated the effectiveness of this procedure either through the mini-open posterior approach or arthroscopic approach (1, 2, 5–10).

It should be noted that supracondylar humeral fractures after ulnohumeral arthroplasty have been previously reported (1, 11). Allen et al. reported a supracondylar humeral fracture after the Outerbridge-Kashiwagi procedure in a young patient that occurred while playing sports, and Hong et al. reported a middle-aged patient who suffered a fracture after a fall (1, 11). Regarding this issue, some biomechanical studies have been conducted (12–14). Degreef et al. found that the failure load of distal humeri decreases by 41% after fenestration (12). Hong et al. found no difference in the failure load of distal humeri between the fenestrations of 12 mm and 15 mm (13), whereas Morrissey et al. suggested a humeral fenestration hole less than 25 mm (14).

To our knowledge, there has also been no clear suggestions about the fenestration-hole size for ulnohumeral arthroplasty. Minami et al. reported fenestration holes ranging from 12 mm to 15 mm in their surgeries (15); Degreef et al. found a mean value of 15 mm from post-operative radiographic measurements (12); Wijeratna et al. used a 12-mm bone graft drill harvester to perform the fenestration of the olecranon fossa (16), whereas Gallo et al. used a 15-mm neurosurgical dowel in their surgeries (17); and Kroonen et al. reported that the fenestration hole would usually be dilated to 10–20 mm after small drilling (8).

There were some limitations in the previous studies, such as the use of cadaver bones with poorer bone quality (12, 13) and an inadequate range of fenestration hole sizes (13, 14). Hence, the purpose of this study was to evaluate the ultimate failure loads of distal humeri with olecranon fenestrations of 10 mm, 12 mm, 15 mm, 18 mm, and 20 mm in ulnohumeral arthroplasty in a synthetic bone biomechanical model. We hypothesized that distal humeri fenestrated with 20-mm holes can exhibit significantly lower ultimate failure loads.

Materials and Methods

Specimens

A total of 50 left humerus fourth-generation synthetic bones (Item No: 3404; Sawbones, Pacific Research Laboratories, Vashon, WA, USA) were used for the biomechanical model. The shaft of the humerus was transected at mid-level with a bone saw, and the distal part of the humerus was used. Previous studies have indicated that the fourth-generation composite humerus can be a reliable substitute for cadaver specimens in biomechanical studies as it not only reproduces the biomechanical properties of the human bone but can also reduce the inter-specimen variability that occurs in the cadaveric bone (18, 19). The proximal end of the distal humerus was potted in a plastic pipe filled with industrial concrete; all specimens were fixed at distal humerus metaphysis.

The specimens were randomly assigned to different groups after being fixed in pipes. After identifying the center of olecranon fossa in each specimen, a fenestration hole was created through it with a motorized drill using drill bits with diameters of 10 mm, 12 mm, 15 mm, 18 mm, and 20 mm (Figure 1).

Figure 1. a, b.

Left side humerus composite bone that was used (a). The center of olecranon fossa was identified, and a fenestration hole was created with a metal drill (b)

Testing set-up

All specimens were mounted on the material testing machines. A material testing machine with a loading scale of 0–5 kilonewtons (kN) (AG-5kNX; Shimadzu, Tokyo, Japan) was used for the anterior-posterior (AP) loading (Figure 2a), whereas the other material testing machine with a loading scale of 0–100 kN (AGS-100kNX; Shimadzu, Tokyo, Japan) was used for the axial loading (Figure 2b). Two force directions were applied: a 5° flexion angle for the axial load application and a 75° flexion angle for the AP load application; these force directions were selected according to previous biomechanical studies (13, 20). We randomly selected the prepared samples to receive either the axial loading or AP loading. All the specimens were loaded to failure at a constant rate of 2 mm/min (12, 13). The specimen failure was defined as the occurrence of a sharp decrease in the load-displacement curve after the peak load. The ultimate failure load and the mode of failure of each specimen were recorded.

Figure 2. a, b.

Setup for biomechanical testing. A load cell with a 5 kN capacity (a) was used for the anterior-posterior loading and another load cell with a 100 kN capacity was used for the axial loading (b)

Statistical analysis

The sample size was calculated based on the ultimate failure load in the AP loading in a pilot study, in which there were a total of 15 specimens that were randomly assigned to five groups (10 mm, 12 mm, 15 mm, 18 mm, and 20 mm groups). An α of 0.05 and a power (1 - β) of 0.95 were given for this a priori power analysis model, the calculated effect size of which was 1.37. At least four specimens in each group were calculated with the use of G*Power, version 3.1.9.2 (available at http://www.gpower.hhu.de; Heinrich Heine-University of Dusseldorf, Dusseldorf, Germany). As a result, a sample size of 5 specimens for each group was determined.

The statistical analysis was conducted with Statistical Package for Social Sciences software for Windows version 20.0 (IBM SPSS Corp., Armonk, NY, USA). The descriptive statistics, including means and standard deviations, were obtained for each group. The Kruskal-Wallis test was used to compare the ultimate failure loads among the five groups. A post hoc analysis using the Mann-Whitney U test was conducted with a Bonferroni correction, resulting in the significance level being set at p≤0.01. A Chi-square test was used to compare the failure modes among the different groups. Statistical significance was defined as p≤0.05.

Results

AP loading

The data for the ultimate failure load in the AP loading are summarized in Table 1. There was a significant difference (p=0.001) in the ultimate failure load among the groups with different fenestration sizes. The post hoc analysis showed that the failure loads of the 18 mm and 20 mm groups were significantly smaller than the other three groups (Figure 3). All specimens in the 10 mm and 12 mm groups failed with the supracondylar shaft fractures, whereas the edge of the fenestration hole was the failure site for all specimens in the 15 mm, 18 mm, and 20 mm groups (Table 2).

Table 1.

The ultimate failure load of distal humeri in different fenestration groups with axial and anterior-posterior loading

Loading type	Fenestration size
	10 mm	12 mm	15 mm	18 mm	20 mm	p
AP loading	2404±160a,b	2245±79c,d	2331±305e,f	1943±83a,c,e	1837±64b,d,f	0.001*
Axial loading	7168±575g–i	5931±898j	5878±350g,k	4636±554h,l	3477±319i–l	0.001*

Ultimate failure load N (mean±SD)

AP: anterior-posterior; SD: standard deviation

^*Significant difference among groups with the Kruskal-Wallis test (p≤0.05)

^a–lSignificant differences between the two groups with the Mann-Whitney U test with a Bonferroni correction (p≤0.01)

Figure 3.

Mean ultimate failure load (N) at the anterior-posterior loading in the groups with different fenestration sizes *p≤0.01

Table 2.

Failure sites on the specimens after loading to failure

		Failure site		p
		Edge of fenestration hole	Supracondylar shaft
Axial loading	10 mm group	5	0	<0.001*
	12 mm group	5	0
	15 mm group	0	5
	18 mm group	0	5
	20 mm group	0	5
Anterior-posterior loading	10 mm group	5	0	1.00
	12 mm group	5	0
	15 mm group	5	0
	18 mm group	5	0
	20 mm group	5	0

^*Significant difference among groups with a Chi-square test (p≤0.05)

Axial loading

The data for the ultimate failure load in the axial loading are summarized in Table 1. The statistical analysis showed a significant difference (p=0.001) in the ultimate failure load among the groups with different fenestration sizes. The post hoc analysis indicated that the failure load of the 10 mm group was significantly greater than that of the 15 mm, 18 mm, and 20 mm groups (p=0.008, p=0.008, and p=0.008, respectively), and the failure load of the 20 mm group was also significantly smaller than that of the 12 mm, 15 mm, and 18 mm groups (p=0.008, p=0.008, and p=0.008, respectively) (Figure 4). All the specimens in the axial loading test failed at the edge of the fenestration hole (Table 2).

Figure 4.

Mean ultimate failure load (N) at the axial loading in the groups with different fenestration sizes *p≤0.01

Discussion

The Outerbridge-Kashiwagi procedure is an attractive surgical technique for young patients with elbow osteoarthritis. Although it is considered a safe technique, supracondylar humeral fractures after ulnohumeral arthroplasty have been previously reported (1, 11). The purpose of this study was to evaluate the ultimate failure loads of distal humeri with olecranon fenestrations of 10 mm, 12 mm, 15 mm, 18 mm, and 20 mm in ulnohumeral arthroplasty in a synthetic bone biomechanical model. Our results revealed that the ultimate failure loads in the AP loading were significantly smaller in the 18 mm and 20 mm groups, whereas the ultimate failure loads in the axial loading decreased in the groups with larger fenestration holes. The clinical relevance is that distal humeral fenestrations of more than 18 mm are not suggested for the Outerbridge-Kashiwagi procedure due to their poor biomechanical properties.

The Outerbridge-Kashiwagi procedure allows for the debridement of the osteophytes at the tip of the coronoid through a small posterior incision. As a result, it is believed that the larger fenestrations might improve the overall effectiveness of the surgery (14). However, whether the size of the fenestration would affect the clinical outcome is not clearly understood. Further studies on this topic may possibly be needed.

Both bone loss and stress concentration effects have been shown to contribute to decreases in the maximum loads of distal humeri after fenestration (13). The larger fenestration holes cause more bone loss, possibly leading to a greater decrease in the ultimate failure loads; meanwhile, the stress concentration effects on the edges of the fenestration holes have been found to be smaller in the groups with larger fenestrations (13). In the present study, the maximum failure loads of distal humeri after fenestration in the AP loading were similar in the 10 mm, 12 mm, and 15 mm groups; the possible reason for this is that the combined effects from the bone loss and stress concentration were similar when the distal humeri fenestration ranged from 10–15 mm.

Synthetic bones were used in this study for biomechanical testing. In some previous studies, cadaveric humeri were used to evaluate the biomechanical properties after the Outerbridge-Kashiwagi procedure; Degreef et al. reported a 41% decrease in the maximal loads of distal humeri after fenestration in a cadaveric model, whereas Hong et al. compared the ultimate failure loads of distal humeri between the 12 mm and 15 mm fenestration groups using cadaver humeri (12, 13). However, the use of cadaver bones was a limitation in these studies as these bones did not totally reflect the actual results (12, 13). In clinical practice, the Outerbridge-Kashiwagi procedures are often considered for young and active patients (5). Thus, considering their age, the use of cadaver bones is relatively inadequate. Thus, synthetic bones were used in this study to simulate young patients with good bone quality.

Although a previous biomechanical study evaluated the relationship between the size of fenestration and fracture risk, the size of the fenestration (10–31 mm) in that study did not entirely represent the clinical conditions (14). According to the literature, the fenestration holes used in clinical practice range from 10–20 mm (8, 12, 15–17). Hence, our study evaluated the ultimate failure loads of distal humeri with fenestration holes ranging from 10 mm to 20 mm.

The findings in our study are consistent with those of previous studies (13, 14). Hong et al. reported similar failure loads between the 12 mm and 15 mm fenestration groups either in the axial or AP loading in a cadaveric model (13). The findings from our synthetic bone model are similar; the ultimate failure loads in the 12 mm and 15 mm fenestration groups were similar in terms of both the axial and AP loading. Meanwhile, Morrissey et al. found that a fracture through the fenestration tends to occur in the groups with larger fenestration holes; similar to the findings of this study, the fracture through the fenestration occurred in the 18 mm and 20 mm fenestration groups in the AP loading (14).

This study had some limitations. First, the biomechanical testing in this study could not totally simulate the dynamic effects of the muscular forces on distal humeri. Second, the time zero biomechanical testing did not take the effects of bone remodeling into account. Third, as the size of the humerus differs from person to person in an actual clinical situation, it may be difficult to generalize the size based on the results of this optimal bone hole size. Finally, the biomechanical tests were performed on bones without any capsule, ligaments, and muscles; although these surrounding tissues can affect the actual fracture pattern and fracture risks, we targeted the influence on distal humeri after the Outerbridge-Kashiwagi procedure.

In conclusion, the failure loads of distal humeri became significantly smaller in the groups with larger fenestrations, especially in the 18 mm and 20 mm groups. The clinical relevance is that the distal humeral fenestrations of more than 18 mm are not suggested for the Outerbridge-Kashiwagi procedure due to their poor biomechanical properties.

HIGHLIGHTS.

• Supracondylar humeral fractures after Outerbridge-Kashiwagi procedure have been previously reported.

• The failure loads of distal humeri became significantly smaller in groups with fenestration holes larger than 18 mm.

• The clinical relevance is that distal humeral fenestrations of more than 18 mm are not suggested for the Outerbridge-Kashiwagi procedure.