Does the killer turn angle in biceps tenodesis lead to a guillotine effect? A biomechanical study

Abstract

Purpose

The purpose of this study was to investigate whether a killer turn angle in biceps tenodesis creates a guillotine-like effect under physiological loading conditions and to evaluate its influence on tendon elongation and tunnel widening.

Methods

30 fresh ovine humeri underwent biceps tenodesis using a bicortical endobutton technique. Specimens were divided into three groups according to the bone–tendon angle at the tunnel entrance (45°, 60°, and 90°). Each construct was subjected to 500 cyclic loading cycles between 0 and 150 N. Tendon elongation and tunnel widening were measured and compared among groups.

Results

No tendon rupture or construct failure was observed in any group. Tendon elongation and residual creep were significantly greater in the 90° group compared with the 45° and 60° groups (p = 0.001), whereas no significant difference was found between the 45° and 60° angles (p = 0.247). Tunnel widening did not differ significantly among groups, although a trend toward increased enlargement with higher angles was noted.

Conclusions

Sharp bone–tendon angulation increased tendon elongation during cyclic loading, suggesting the presence of a guillotine-like mechanical effect in biceps tenodesis. Reducing the killer turn angle to 60° or less appeared sufficient to mitigate this effect. Although no acute failure was observed, these findings highlight the potential relevance of the guillotine effect in shoulder surgery and underscore the importance of considering the killer turn angle, particularly in tendon transfer procedures with higher reported failure rates.

Level of evidence

Level V, biomechanical study.

Introduction

The Killer Turn phenomenon was first described in the literature following posterior cruciate ligament reconstruction, where the tendon graft exits the tibial tunnel at a sharp angle, leading to elongation over time and resulting in residual laxity [1, 5, 10, 12, 19]. Although the killer turn phenomenon is not the sole contributor to tendon graft laxity, other factors such as tunnel enlargement or graft displacement are generally not considered the primary causes of residual laxity in procedures involving tendon graft fixation [7, 9].

In shoulder surgery, a potential killer turn phenomenon that may lead to failure of techniques such as tenodesis or tendon transfer is referred to as the “guillotine effect” [17, 21]. This phenomenon has been mentioned in some studies as a potential culprit, particularly among the causes of failure following biceps tenodesis or latissimus dorsi transfer [6]. Tendon fixation in shoulder surgery is most commonly performed using suture anchors, while techniques involving bone tunnel creation are also employed in biceps tenodesis, utilizing either endobutton or bioabsorbable interference screw fixation methods [4, 14, 20]. Tendon graft laxity in anterior and posterior cruciate ligament reconstructions of the knee has been extensively investigated in the literature, but it remains unclear whether a similar phenomenon occurs in tendon fixation procedures of the shoulder, such as biceps tenodesis. Tendon elongation following biceps tenodesis would not theoretically result in joint laxity, however the sharp angle formed at the margin of the humeral tunnel created for tenodesis may, over time, lead to elongation and thinning of the long head of the biceps tendon, which could theoretically predispose it to eventual rupture.

Today, biceps tenodesis is commonly performed as onlay methods in which the tendon attaches on humeral cortex using an anchor or a unicortical endobutton, which eliminates the possibility of the killer turn phenomenon. However, some shoulder surgeons perform techniques involving bone tunnel creation combined with either bicortical endobuttons or bioabsorbable interference screws, due to the biomechanical superiority of these techniques in high demand patients [2, 14]. The aim of the study was to investigate whether the ‘killer turn’ phenomenon exists under physiological loading of the biceps tenodesis techniques that involve a sharp bone–tendon angle.

Materials and methods

Study design and specimen preparation

This controlled biomechanical study used 30 fresh ovine humeri obtained from 1-year-old male sheep and the specimens were randomly divided into 3 groups of 10 shoulders each. All specimens were collected and tested on the same day to prevent dehydration or biomechanical degradation. Soft tissues surrounding the humerus were carefully removed while preserving the entire biceps muscle–tendon complex. Specimens were visually inspected for cortical defects or deformities and kept moist with saline-soaked gauze throughout preparation and testing. The study utilized humeral specimens obtained from animals slaughtered for commercial purposes; therefore, no animals were sacrificed specifically for this study. The study protocol was approved by the Izmir University of Economics Animal Experiments Local Ethics Committee, Izmir, Turkey (Approval No: E-97429853-050.04-88781).

To examine the influence of killer turn angle on early tendon behavior following biceps tenodesis, specimens were randomly assigned into three groups corresponding to killer turn angles of 90°, 60°, and 45° (Fig. 1). As no standardized bone–tendon killer-turn angle has been defined in the literature for biceps tenodesis, these angles were selected to represent clinically plausible and experimentally distinguishable tunnel-entry configurations. More acute angles, such as 30°, were excluded because they were considered less representative of routine surgical practice and more likely to create a non-physiological tunnel geometry.

Fig. 1.

Schematic representation of humeral tunnel types created. The first group had a tunnel entry oriented perpendicular at 90°. In the other groups, the tunnel entrance on the load-bearing side of the biceps tendon was contoured using 60° and 45° angle-adjustable guides, resulting in killer turn angles of 60° and 45°, respectively

In each specimen, the humeral tunnel was created at the anterior cortex of the bicipital groove, replicating the anatomical entry point used clinically. A bicortical tunnel was created using a 6.5-mm drill, positioned perpendicular to the humeral shaft with the aid of a goniometer. The entrance of the cortical bone tunnel on the load-bearing cortical wall was then contoured at a specific angle using an angle-adjustable guide. Tenodesis was performed using a standard cortical endobutton fixation (11 mm) and fiber sutures (Rejoin Medical Corp., Hangzhou, China). The proximal tendon stump was trimmed to a uniform length of 50 mm and prepared with a locking Krackow-type stitch using five passes through the tendon. The native ovine biceps tendon was then passed through the tunnel and secured on the far cortex (Fig. 1).

The cortical diameter of the humerus at the tenodesis level was measured using a precision caliper to confirm comparable geometric characteristics across all specimens (Fig. 2). The proximal tendon stump was trimmed to obtain a uniform length of 50 mm across specimens. The specimens had to be loaded in a proximal direction rather than distally because the relatively short length of the biceps musculotendinous junction prevented true perpendicular distal loading in biomechanical testing device, as the humeral shaft obstructed this trajectory. However, this configuration was considered unlikely to substantially affect comparisons between groups, as it created a symmetrical loading condition across all specimens. All procedures were carried out by the same experienced orthopedic surgeon to minimize technical variability.

Fig. 2.

A Final construct appearance following biceps tenodesis (B) Measurement of tunnel diameter following biomechanical testing and removal of the biceps tendon

Biomechanical testing setup

Mechanical testing was conducted using a universal testing system (Shimadzu AGS-X series - Shimadzu, Kyoto, Japan). Each specimen was mounted vertically, with the distal humerus rigidly secured in the lower clamp and the proximal musculotendinous junction fixed in the upper clamp. Gauze was wrapped around the proximal muscular part to prevent slippage (Fig. 3). Careful alignment ensured that the applied load corresponded to the perpendicular line of pull of the biceps tendon. All tests were performed at room temperature (at 22 °C).

Fig. 3.

Image of the biomechanical testing setup

Cyclic loading protocol

Each construct underwent 500 consecutive loading–unloading cycles (between 0 and 150 N). Tensile loading was applied at a constant rate of 50 mm/min until reaching a peak load of 150 N. A peak load of 150 N was selected as it approximates the upper range of physiological forces transmitted through the long head of the biceps tendon during active elbow flexion and forearm supination. This loading magnitude has been widely used in previous biomechanical studies on biceps tenodesis to simulate early postoperative and high-demand functional conditions without exceeding physiological limits [2, 13, 16]. Specimens were continuously monitored for endobutton migration, tendon failure, or cortical damage.

Outcome measurement

The primary outcome was tendon elongation, measured in millimeters with a resolution of 0.01 mm. Elongation was defined as the displacement observed at the moment the construct reached 150 N in each cycle, allowing quantification of cycle-to-cycle creep and early mechanical slackening of the tenodesis construct. Tunnel widening was also measured in millimeters and compared between the groups.

Statistical analysis

A prior sample size estimation using PASS 15 (NCSS, LLC, Kaysville, USA) for a one-way ANOVA with three independent groups (α = 0.05, effect size f = 0.33, power = 0.80) indicated that a minimum of 30 specimens (n = 10 per group) would be sufficient to detect statistically significant differences among groups. Data were analyzed using IBM SPSS version 23 (IBM Corp., Armonk, NY., USA). Group comparisons were performed using one-way ANOVA for three-group analyses, followed by Tukey’s HSD post-hoc test for pairwise comparisons. Normality was assessed with the Shapiro–Wilk test, and homogeneity of variances was evaluated using Levene’s test. A significance level of p < 0.05 was considered statistically significant for all analyses.

Results

Biomechanical results

Biomechanical testing was performed on 30 ovine humeri to evaluate the effect of drilling angle (45°, 60° and 90°) on tendon elongation behavior during biceps tenodesis. Descriptive measurements for all tendon elongation variables are summarized in Table 1 and statistical comparisons of residual creep are reported in Table 2.

Table 1.

Tendon elongation parameters across killer turn angles (Mean ± SD)

Parameter	45°	60°	90°
Initial minimum length (mm)	9.71 ± 2.21	12.16 ± 3.16	14.79 ± 6.69
Initial maximum length (mm)	11.54 ± 2.31	13.12 ± 2.34	17.20 ± 6.73
Final length (mm)	15.02 ± 2.92	16.11 ± 1.85	22.75 ± 7.71
Residual creep (mm)	5.31 ± 1.58	3.95 ± 1.62	7.96 ± 2.28

SD: standard deviation

Table 2.

Residual creep (mm) across angles

Test	Group	p-value
Welch’s ANOVA	45° vs. 60° vs. 90°	0.001
Tukey Post-hoc	60° vs. 45°	0.247
	60° vs. 90°	< 0.001
	45° vs. 90°	0.010
Normality (Shapiro–Wilk)	W = 0.963	0.374
Levene’s test	0.164	0.849

Bold values indicate statistically significant results (p < 0.05)

SD: standard deviation; ANOVA: analysis of variance

No tendon failure occurred in any of the specimens. Initial tendon measurements (Initial Maximum and Initial Minimum Length) showed numerical differences among groups; however, these baseline values did not demonstrate a statistical difference, indicating comparable tendon lengths prior to cyclic loading.

After 500 loading cycles, Final Length differed substantially between groups. The 90° group exhibited the greatest post-cycle elongation (22.75 ± 7.71 mm), clearly exceeding the values observed in the 60° (16.11 ± 1.85 mm) and 45° (15.02 ± 2.92 mm) groups.

Residual creep (Permanent deformation)

Residual creep, defined as the irreversible elongation after unloading, demonstrated the strongest angle-dependent effect. Values increased from 60° (3.95 ± 1.62 mm) to 45° (5.31 ± 1.58 mm) and reached their highest level at 90° (7.96 ± 2.28 mm) (Table 1). Welch’s ANOVA confirmed a significant effect of killer turn angle on residual creep (p = 0.001). Tukey post-hoc comparisons showed that the 90° group differed significantly from both 60° (p < 0.001) and 45° (p = 0.010), while 60° and 45° remained statistically similar (p = 0.247) (Table 2).

The residual creep plot demonstrated minimal overlap between the 90° group and the two lower angle groups, visually reinforcing the statistical separation and highlighting the biomechanical disadvantage of higher angles of killer turn (Fig. 4).

Fig. 4.

Residual creep (permanent deformation) values across drilling angles

The 90° killer turn angle consistently produced the greatest degree of tendon elongation, across all elongation-related parameters, representing the tunnel entry angle that resulted in the greatest tendon elongation during cyclic loading. In contrast, 60° and 45° killer turn angles showed nearly identical and biomechanically favorable behavior, characterized by lower elongation magnitudes and reduced permanent deformation.

Tunnel enlargement results

Killer turn angle and tunnel enlargement were also related. The 90° group showed the most significant enlargement, with an average increase of 10.4% (± 0.38 mm) in tunnel diameter. These findings suggested that the killer turn was associated with increased stress at the tendon exit, increased micromotion, and progressive bone resorption around the entry point (Table 3).

Table 3.

Tunnel enlargement at different killer turn angles

Killer turn angle (°)	Initial tunnel diameter (mm)	Mean final tunnel diameter (mm)	Enlargement ratio (%)	SD of tunnel enlargement (mm)
45°	6.50	6.95	6.97	± 0.27
60°	6.50	6.98	7.45	± 0.28
90°	6.50	7.18	10.4	± 0.38

The 45° group showed a moderate enlargement of 6.97% (± 0.27 mm); the 60° group showed a comparable, slightly higher value of 7.45% (± 0.28 mm). These findings suggested that reducing the killer turn angle from 90° to roughly 60 − 45° caused decrease in strain. However, more flattening of the angle (toward 45°) did not cause an additional biomechanical benefit.

Discussion

Although the concept of the killer turn angle has been extensively discussed in the posterior cruciate ligament surgery literature, its relevance in shoulder surgery remains incompletely understood [11]. In the shoulder, this phenomenon is more often described as the ‘guillotine effect’; however, whether such abrupt bone–tendon angulation translates into a meaningful risk of mechanical failure is still unknown. Accordingly, the present study sought to evaluate the integrity of biceps tenodesis construct under physiological loading conditions to determine whether a guillotine-type mechanical failure occurs.

Several authors have suggested that the pronounced angulation inherent to procedures such as latissimus dorsi transfer and biceps tenodesis may predispose the graft or tendon to increased stress concentration and subsequent failure, although definitive biomechanical and clinical evidence remains limited [3, 17, 21]. In addition, rupture of the biceps tendon has been reported as a complication of the biceps tenodesis technique in various studies [8, 18]. A key finding of the present study was that no tendon rupture or catastrophic mechanical failure occurred in any biceps tenodesis group, regardless of the killer turn angle. However, when the results of this study were carefully examined, it was considered that, although biceps tenodesis did not fail, configurations involving 90-degree bone–tendon angles, such as those encountered in certain tendon transfer procedures, may still harbor potential risks. We considered that the 2–6% risk of rupture reported after biceps tenodesis in the literature may, at least in part, be associated with a guillotine-like mechanical effect; however, this experimental model does not allow a direct causal relationship to be established. Notably, the majority of studies reporting such failures originated from centers employing bone tunnel–based techniques [8, 15].

In the study, following the application of 500 loading cycles, tendon elongation at a 90º angle was observed to be significantly higher than that at 45º and 60º angles. A mild decrease of the killer turn angle of the bone tunnel appeared to be sufficient to reduce tendon damage. Reducing the angle from 60 to 45 degrees did not result in a further decrease in tendon elongation. The lack of additional biomechanical benefit when reducing the killer turn angle from 60° to 45° indicates a plateau phenomenon, whereby the bone–tendon interface stabilizes once a critical angulation threshold is crossed. This plateau effect supports the notion that beyond a certain point, further flattening of the bone–tendon angle does not meaningfully reduce micromotion or improve tunnel integrity. Further reduction of the angle increases the length of tendon–bone contact, potentially augmenting frictional forces and time-dependent viscoelastic creep without conferring additional protection against the killer turn phenomenon.

Orienting the bone–tendon construct at approximately 60° appeared to reduce killer-turn–related stress in this experimental setting, without introducing additional mechanical disadvantages. These findings are consistent with previous studies in ligament reconstruction showing that acute tunnel angles intensify local stresses (killer-turn), increase micromotion, and are associated with tunnel widening and durability concerns [5, 19]. From a surgical perspective, a more oblique bone–tendon configuration may be facilitated by careful selection of the tunnel entry point, adjustment of drilling direction, and smoothing of the cortical edge at the tunnel aperture. Although the present biomechanical model does not allow direct procedural recommendations, these factors may be considered when attempting to reduce stress concentration at the tendon–bone interface in bone tunnel–based tenodesis techniques.

On the other hand, tunnel widening did not differ significantly among the groups. However, trend toward increased tunnel widening from 45º to 90º suggested that statistical significance might have been observed with a higher number of loading cycles or a larger sample size. Increased tunnel enlargement in higher angles further supported the concept that acute tendon angulation increases micromotion and localized stress at the tunnel aperture. This phenomenon closely parallels tunnel widening observed in cruciate ligament reconstruction, where killer turn–induced graft motion and stress concentration contribute to progressive bone resorption [9, 12].

From a clinical perspective, these experimental findings suggest that biceps tenodesis techniques involving sharp bone–tendon angulation should be interpreted with caution, particularly in high-demand scenarios. Considering that bioabsorbable screws are commonly used for tenodesis in patients with high functional demands and that these screws may cause initial tendon damage by compressing the biceps at the tunnel entrance, making the inferior tunnel aperture smoother, even when the procedure is performed arthroscopically, was considered a safer approach. In this study, bioabsorbable screws were not used in order to isolate and evaluate the effect of the tunnel entrance itself, independent of any potential compressive damage to tendon caused by the fixation devices.

Several limitations of this study should be acknowledged. First, an ovine model was used, which may not fully replicate the biomechanical and viscoelastic properties of the human humerus and long head of the biceps tendon. Second, the experimental design focused on early mechanical behavior under controlled cyclic loading and did not account for biological healing or long-term in vivo adaptations. Loading was applied along a single line of pull, whereas physiologic biceps forces are multidirectional and vary with joint position. In addition, the number of loading cycles represents early postoperative conditions and may not reflect long-term repetitive loading. Finally, although the fixation method was deliberately selected to minimize implant-related tendon damage and to isolate the mechanical influence of the tunnel aperture and killer turn angle, only one fixation technique and tunnel configuration were evaluated, which limits the direct generalizability of these findings to other tenodesis methods or implant constructs.

Conclusion

This study shows that the guillotine effect in shoulder surgery develops as a gradual mechanical process, marked by increased tendon elongation, residual creep, and early tunnel changes. Reducing the bone–tendon angle clearly lessened these effects, whereas sharper angles amplified them. Although biceps tenodesis rarely fails clinically, optimizing tunnel orientation may further improve mechanical reliability. Within the confines of this experimental model, bone–tendon angles of 60° or less were associated with a more favorable mechanical behavior, whereas near-perpendicular orientations around 90° were linked to increased mechanical stress. Additional studies specifically investigating the guillotine effect in the shoulder are needed to clarify its role in tendon transfer failure and other procedures where graft integrity is more critical.

Abbreviations

SD

Standard deviation

ANOVA

Analysis of variance

Author contributions

All authors contributed to the study conception and design. Material preparation, data collection, and biomechanical testing were performed by MK, BU, BD and UAK. Data analysis was performed by BD and BU. The first draft of the manuscript was written by MK, BU, AB and OB and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

The authors did not receive any funding for this study.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted using ovine humeri obtained from animals slaughtered for commercial purposes. No animals were sacrificed specifically for this study. The study protocol was approved by the Izmir University of Economics Animal Experiments Local Ethics Committee, Izmir, Turkey (Approval No: E-97429853-050.04-88781).

Consent for publication

Not applicable. This study was an experimental animal biomechanical study and did not include any human participants, identifiable personal data, or patient images.

Competing interests

The authors declare no competing interests.