- Email: [email protected]
Shoulder arthroplasty and its effect on strain in the subscapularis muscle
Clinical Biomechanics 30 (2015) 373–376
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Shoulder arthroplasty and its effect on strain in the subscapularis muscle Thomas Wright ⁎, Thomas Easley, Jessica Bennett, Aimee Struk, Bryan Conrad Department of Orthopaedics and Rehabilitation, University of Florida College of Medicine, 3450 Hull Road, Gainesville, FL 32611, USA
a r t i c l e
i n f o
Article history: Received 5 March 2014 Accepted 16 February 2015 Keywords: Shoulder arthroplasty Subscapularis muscle Strain Prosthetic humeral head External rotation Abduction
a b s t r a c t Background: Increasing the thickness of the prosthetic humeral head on subscapularis strain in patients undergoing total shoulder arthroplasty has not been elucidated. The optimal postoperative rehabilitation for total shoulder arthroplasty that does not place excessive strain on the subscapularis is not known. We hypothesize that the use of expanded non-anatomic humeral heads during shoulder replacement will cause increased tension in the repaired subscapularis. We identified a recommended passive range of motion program without invoking an increase in tension in the repaired subscapularis, and determined the impact of the thickness of the humeral head on subscapularis strain. Methods: Eight fresh-frozen, forequarter cadaver specimens were obtained. An extended deltopectoral incision was performed and passive range-of-motion exercises with the following motions were evaluated: external rotation, abduction, flexion, and scaption. An optical motion analysis system measured strain in the subscapularis. The same protocol was repeated after performing a subscapularis osteotomy and after placement of an anatomic hemiarthroplasty of three different thicknesses. Findings: For abduction and forward flexion, we observed a trend of decreasing strain of the subscapularis, as the laxity is removed with increasing humeral head component thickness. With the short humeral head, strain was similar to native joint with passive scaption and flexion but not with external rotation or abduction. Interpretation: The passive range of motion that minimizes tension on the subscapularis is forward flexion and scaption. Therefore, passive forward flexion or scaption does not need to be limited, but external rotation should have passive limits and abduction should be avoided. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Total shoulder arthroplasty (TSA) is performed with a surgical approach through the subscapularis. After surgery, rehabilitation of the shoulder is imperative to optimize the functional result. However, the subscapularis must be respected in the process of this rehabilitation or the repair will fail, resulting in a poorly functioning shoulder and in some cases shoulder instability. Using ultrasound after subscapularis tenotomy and repair in the course of performing a TSA, failure rates of the repair have been reported between 13 and 47% (Armstrong et al., 2006; Jackson et al., 2010). Weakness, anterior instability, glenoid loosening, and decreased function have all been described after a failed subscapularis repair (Miller and Kathuria, 2006; Miller et al., 2003, 2005; Moeckel et al., 1993; Utz et al., 2011). In an effort to decrease subscapularis failure rate in the setting of a TSA, many surgeons have begun performing a lesser tuberosity osteotomy as it has been shown ⁎ Corresponding author at: Department of Orthopaedics and Rehabilitation, PO Box 112727, Gainesville, FL 32611-2727, USA. E-mail address: [email protected]fl.edu (T. Wright).
http://dx.doi.org/10.1016/j.clinbiomech.2015.02.010 0268-0033/© 2015 Elsevier Ltd. All rights reserved.
to have a decreased failure rate when compared to direct tendon repair (Qureshi et al., 2008). Rehabilitation protocols are commonly designed to protect the subscapularis repair; however, there is little science supporting their specific recommendations (Wilcox et al., 2005). The purpose of a shoulder replacement is to provide pain relief and increase function in a patient with a pathologic glenohumeral joint. The current trend is to replicate the patient's pre-existing normal shoulder anatomy as closely as possible with the shoulder replacement. However, in an attempt to stabilize the TSA many systems provide the surgeon with options of humeral heads that are much thicker than the normal humeral head. It is likely that the use of tall (thicker) or expanded non-anatomic humeral heads will cause an increase in the tension experienced in the repaired subscapularis. TSA instability is better managed with correction of glenoid version and embrication of the rotator cuff interval rather than by overstuffing the joint with a thick prosthetic head. The primary purpose of this study was to identify the optimal manner to perform a passive range of motion (PRoM) program without invoking a large increase in tension in the repaired subscapularis. The secondary purpose was to determine the impact of varying the
374
T. Wright et al. / Clinical Biomechanics 30 (2015) 373–376
thickness of the humeral head on subscapularis strain using the same PRoM protocol. 2. Methods 2.1. Specimen preparation Eight fresh-frozen, forequarter cadaver specimens (four females and four males) were obtained following International Review Board approval. Each specimen was mounted on a surgical holding device that clamped to the medial scapula eliminating scapular motion. Using an extended deltopectoral approach, the distal subscapularis and insertion site were well visualized. A six-camera motion analysis system (Eagle, Motion Analysis Corp., Santa Rosa, CA, USA) was used to measure strain in the subscapularis. Four 2 mm reflective beads were carefully sutured to the subscapularis in a 1 cm2 square-shaped arrangement to allow tissue strains to be measured. The superior two markers were sutured to the upper subscapularis and the inferior two markers were sutured to the lower subscapularis (Fig. 1). The primary outcome measure for this study was tissue strain of the upper and lower subscapularis. Strain was measured as the peak change in displacement from the resting tissue length, during PRoM movements (initial length − final length = strain in mm). PRoM of the glenohumeral joint was measured using an electromagnetic tracking device (Liberty, Polhemus Inc., Colchester, VT, USA). The Polhemus device allowed the three-dimensional angulation to be tracked and recorded. Two Polhemus tracking sensors were attached: one to the mid-diaphysis of the humerus and one to the spine of the scapula. Before collecting data, the sensors were aligned to the anatomical axes of the specimen. A custom software program written in LabView (National Instruments, Austin, TX, USA) was used to collect data from the Polhemus system and calculate joint angles. 2.2. Experimental procedure Both strain and range of motion data were collected simultaneously as the arms were moved through a PRoM in external rotation, abduction, flexion, and scaption for five cycles of each motion. To ensure that the
tissues were at a steady state, data from the fifth cycle were used for analysis. The range of motion and strain data were synchronized post hoc by using a linear regression procedure to optimize the temporal alignment of the signals. The PRoM protocol was initially performed on each specimen in the intact condition to measure “normal” strain. The starting position for each set of conditions was always with the arm by the side and internally rotated like what would occur with the hand on the abdomen in the resting state. The exact same PRoM sequence was then repeated after placement of a short humeral head with repair of the osteotomy site and again after placement of a tall (thicker) and expanded head (even thicker). The subscapularis osteotomy was repaired back to the exact same spot after each set of conditions. No destructive loading of the subscapularis was performed. All implant components were part of the Equinoxe shoulder arthroplasty system (Exactech Inc., Gainesville, Florida, USA). Eight specimens were tested for all conditions, but two specimens were too small to be tested with the expanded humeral head arthroplasty. After exposure of the subscapularis and placement of the reflective markers, baseline strain was obtained using the above motions. The lesser tuberosity was then osteotomized in a consistent fashion, and a humeral head cut was made on the anatomic neck. The humeral canal was reamed and broached. Prior to placing the humeral stem, three drill holes were placed just lateral to the lesser footprint and three just medial to the footprint. One number 2 FiberWire (Arthrex, Naples, Florida, USA) suture was placed through the lateral hole then around the inserted humeral stem and out the medial hole. This was repeated for all three sets of holes, placing a total of three sutures. Once repaired, each suture acted as a cerclage around the lesser tubercle and stem exactly as is normally performed in surgery. This construct returned the lesser tubercle to its native footprint. Using the system replicator plate and a humeral head sized directly off the removed native head, the arthroplasty head was placed to exactly cover the native humeral head cut. The subscapularis was then repaired in the manner described. One figure of eight number two FiberWire was used in the rotator interval adjacent to the two tuberosities. The same passive motions were then performed with strain noted as in the intact state. The exact same surgical and testing procedures were repeated with the tall (4 mm thicker than short) and then expanded humeral heads (4 mm thicker than tall and 8 mm thicker than short) in each cadaver specimen. 3. Results A general trend of decrease in tissue strain on the subscapularis was observed in abduction and external rotation following implantation of the shoulder arthroplasty components (Figs. 2–5). The apparent paradox of this is due to the tendon being under increased tension (all the laxity has been removed). At point zero there was less tendon laxity,
Fig. 1. Schematic drawing of the experimental setup with markers placed on the subscapularis.
Fig. 2. Tissue strain for upper and lower subscapularis during abduction with different implant sizes.
T. Wright et al. / Clinical Biomechanics 30 (2015) 373–376
Fig. 3. Tissue strain for upper and lower subscapularis during external rotation with different implant sizes.
which resulted in less movement between the markers (decreased strain). For abduction and forward flexion, we observed a trend of decreasing laxity with increasing humeral head component thickness. For external rotation, all components displayed a similar reduction in tissue strain. For the short head, passive flexion and scaption showed strains very close to the normal state, which may suggest that these movements are optimal for postoperative rehabilitation. A summary of peak strains is in Table 1. Similar strain patterns were observed for both the superior and inferior portions of the subscapularis. This occurs because the insertion site is very broad superior to inferior and strain is distributed in all directions. 4. Discussion Most approaches to the shoulder, for the purpose of performing a shoulder arthroplasty, violate the subscapularis muscle tendon unit that is subsequently repaired. This may occur through the tendon, at the bone tendon interface, or through the lesser tuberosity. Failure of the subscapularis repair is a well-described complication of shoulder arthroplasty, which results in decreased function and instability (Jackson et al., 2010; Terrier et al., 2013; Utz et al., 2011). Failure of the upper subscapularis results in altered kinematics that is exacerbated by load, causing the humeral head to translate in an anterior superior direction (Su et al., 2009). Miller et al. (Miller et al., 2003) showed that, despite meticulous attention paid to subscapularis repair, two out of three patients would still have abnormal subscapularis function.
Fig. 5. Tissue strain for upper and lower subscapularis during scaption with different implant sizes.
It is important that we protect subscapularis healing and therefore function after shoulder arthroplasty. The optimal shoulder arthroplasty technique and rehabilitation protocol that will maximize the opportunity for subscapularis healing need to be defined. There are surprisingly few reports of strain on the subscapularis. Muraki et al. (Muraki et al., 2007) looked at subscapularis strain using cadavers with their native joint and found that the lower subscapularis generally experienced greater strain than the upper. In our study we could not clearly demonstrate the differentiation of upper and lower subscapularis strains, but this is probably due to the markers being placed in the upper and middle subscapularis rather than the far inferior lower subscapularis. The markers were not attached to the lower muscular portion of the subscapularis, which resides below the glenohumeral center of rotation, and therefore we could not demonstrate a significant difference in strain between the upper and lower subscapularis as it was not tested. Muraki et al. (Muraki et al., 2007) showed that the subscapularis had the greatest strain in external Table 1 Summary of peak strains in the inferior and superior subscapularis during different movements. Motion
Implant
Inferior medial/lateral (mean ± SD) (mm)
Superior medial/lateral (mean ± SD) (mm)
Abduction
Intact Short Tall Expanded Abduction average
4.1 ± 3 2.2 ± 1 2.7 ± 1.7 1.9 ± 1 3 ± 2.2
4.4 ± 3 3.9 ± 2.5 2.4 ± 1.5 1.3 ± 0.8 3.3 ± 2.5
External rotation
Intact
6.6 ± 6.6
7.2 ± 4.5
Short Tall Expanded External rotation average
3.9 ± 2.8 4.3 ± 3.9 3.1 ± 1.3 4.8 ± 4.6
3.5 ± 2.3 5 ± 3.5 4.1 ± 1.6 5.3 ± 3.6
6 ± 7.1
6.4 ± 3.2
Short Tall Expanded Forward flexion average
5 ± 5.3 3.4 ± 2.1 4.5 ± 3.6 4.9 ± 5.1
5.8 ± 3.5 4.1 ± 2.1 4.5 ± 2.1 5.4 ± 2.9
Intact Short Tall Expanded Scaption average
4.6 ± 2.5 4.1 ± 3 7.5 ± 8.2 5 ± 4.9 5.2 ± 4.8
5.3 ± 4.3 5.8 ± 5.3 9.2 ± 11 8.8 ± 11.9 6.9 ± 7.6
Forward flexion
Scaption
Fig. 4. Tissue strain for upper and lower subscapularis during forward flexion with different implant sizes.
375
Intact
376
T. Wright et al. / Clinical Biomechanics 30 (2015) 373–376
rotation especially with abduction and our data, although not performed in this manner, would corroborate their findings. Pure external rotation in our study was performed with the shoulder in adduction by the side. However, we noted lower strain in the subscapularis (slack removed) with abduction, as did Muraki et al.; the explanation for this is that there is obligatory external rotation with abduction. Muraki et al. did not evaluate strain after shoulder arthroplasty (Muraki et al., 2007). Although there are very few studies that have evaluated strain in the subscapularis, some investigators have reported on strains of other muscles of the rotator cuff. Kim et al. (Kim et al., 2011) used a novel ultrasound technique to measure in vivo tissue strain of the supraspinatus during scaption and found that the greatest strains occurred in the superficial region, which would correspond to the area measured in the current study. Andarawis et al. (Andarawis-Puri et al., 2010) showed that, during abduction, strain in both the supraspinatus and infraspinatus was at a minimum when the shoulder was at thirty degrees abduction. Based on our data, the subscapularis muscle has less strain after shoulder joint arthroplasty. We interpret this mechanism as follows: slack in the subscapularis has been removed because of greater size of the implant components, thus increasing preload tension on the muscle unit. This effect is more pronounced with thicker humeral head implants. This is a concern because the repaired muscle tendon unit could be under greater tension than normal, which might make it prone to failure. There are two possible explanations for this observation. One is that, in the process of repairing the osteotomy site, the site was transferred laterally or the sutures took up some of the subscapularis tendon in the process of passing them and, therefore, caused excessive tension on the subscapularis. The other explanation is that the proximal humeral reconstruction was not perfectly anatomic. The reduced strain trend was even greater with thicker humeral head implants. It appears that even the short head size humeral component trended towards decreased strain in the subscapularis compared to the intact joint. Thicker non-anatomic humeral head implants are provided to manage instability. However, instability of the shoulder should be managed by other means including correction of glenoid version and embrication of the rotator cuff interval rather than over-stuffing the joint. If the repaired muscle tendon unit is under increased tension after shoulder arthroplasty, it becomes imperative that the rehabilitation program not stress this repair until it has had sufficient time to heal. Based on our findings, limiting external rotation, abduction, and particularly the combination of abduction and external rotation is important for the rehabilitation program. With abduction and external rotation the strains decreased because the subscapularis had all its slack removed at time 0. However, passive flexion and scaption without external rotation appear to be safe with the short arthroplasty head demonstrating the same strain as the intact state for these two motions. This represents a small rehabilitation change as most protocols limit external rotation but do not push flexion and scaption. Limitations of this study were that tension could not directly be measured in the subscapularis; rather, strain is used for a surrogate. Additionally, scapular motion was constrained with all motion occurring only at the glenohumeral joint. Because we were dealing with cadaveric tissues it is possible that some stretching of soft tissues occurred in the process of multiple examinations. This was a passive test and it is probable that dynamic contraction of the subscapularis may alter strain further in vivo.
5. Conclusions Despite meticulous surgical technique, strain did not return to the normal state after shoulder arthroplasty and a greater deviation from the normal state was noted with increasing humeral head thickness. Passive scaption and flexion appear to be relatively safe motions for postoperative rehabilitation, whereas abduction and external rotation were not as safe. It is our recommendation to use the thinnest humeral head implant that is practical and limit passive external rotation, avoid passive abduction, and not limit scaption or flexion as long as the shoulder is not externally rotated during these exercises. Based on this study we recommend against the use of expanded (thick) non-anatomic humeral head implants as there are better means for managing TSA instability. Using this protocol should limit excessive strain on the repaired subscapularis and hopefully result in a lower postoperative failure rate. Acknowledgments Funding for this study was from institutional sources. References Andarawis-Puri, N., Kuntz, A.F., Ramsey, M.L., Soslowsky, L.J., 2010. Effect of glenohumeral abduction angle on the mechanical interaction between the supraspinatus and infraspinatus tendons for the intact, partial-thickness torn, and repaired supraspinatus tendon conditions. J. Orthop. Res. 28, 846–851. Armstrong, A., Lashgari, C., Teefey, S., Menendez, J., Yamaguchi, K., Galatz, L.M., 2006. Ultrasound evaluation and clinical correlation of subscapularis repair after total shoulder arthroplasty. J. Shoulder Elb. Surg. 15, 541–548. Jackson, J.D., Cil, A., Smith, J., Steinmann, S.P., 2010. Integrity and function of the subscapularis after total shoulder arthroplasty. J. Shoulder Elb. Surg. 19, 1085–1090. http://dx.doi.org/10.1016/j.jse.2010.04.001. Kim, Y.S., Kim, J.M., Bigliani, L.U., Kim, H.J., Jung, H.W., 2011. In vivo strain analysis of the intact supraspinatus tendon by ultrasound speckles tracking imaging. J. Orthop. Res. 29, 1931–1937. Miller, D., Kathuria, V., 2006. A surgical tip for shoulder hemiarthroplasty in a patient with a deficient rotator cuff. J. Surg. Orthop. Adv. 15, 60–61. Miller, S.L., Hazrati, Y., Klepps, S., Chiang, A., Flatow, E.L., 2003. Loss of subcapularis function after shoulder replacement: a seldom recognized problem. J. Shoulder Elb. Surg. 12, 29–34. Miller, B.S., Joseph, T.A., Noonan, T.J., Horan, M.P., Hawkins, R.J., 2005. Rupture of the subscapularis tendon after shoulder arthroplasty: diagnosis, treatment, and outcome. J. Shoulder Elb. Surg. 14, 492–496. Moeckel, B.H., Altchek, D.W., Warren, R.F., Wickiewicz, T.L., Dines, D.M., 1993. Instability of the shoulder after arthroplasty. J. Bone Joint Surg. Am. 75, 492–497. Muraki, T., Aoki, M., Uchiyama, E., Takasaki, H., Murakami, G., Miyamoto, S., 2007. A cadaveric study of strain on the subscapularis muscle. Arch. Phys. Med. Rehabil. 88, 941–946. Qureshi, S., Hsiao, A., Klug, R.A., Lee, E., Braman, J., Flatow, E.L., 2008. Subscapularis function after total shoulder replacement: results with lesser tuberosity osteotomy. J. Shoulder Elb. Surg. 17, 68–72. http://dx.doi.org/10.1016/j.jse.2007.04.018. Su, W.R., Budoff, J.E., Luo, Z.P., 2009. The effect of anterosuperior rotator cuff tears on glenohumeral translation. Arthroscopy 25, 282–289. http://dx.doi.org/10.1016/j. arthro.2008.10.005. Terrier, A., Larrea, X., Malfroy Camine, V., Pioletti, D.P., Farron, A., 2013. Importance of subscapularis muscle after total shoulder arthroplasty. Clin. Biomech. 28, 146–150. Utz, C.J., Bauer, T.W., Iannotti, J.P., 2011. Glenoid component loosening due to deficient subscapularis: a case study of eccentric loading. J. Shoulder Elb. Surg. 20, e1621. http://dx.doi.org/10.1016/j.jse.2011.03.024. Wilcox, R.B., Arslanian, L.E., Millet, P.J., 2005. Rehabilitation following total shoulder arthroplasty. J. Orthop. Sports Phys. Ther. 35, 821–836. http://dx.doi.org/10.2519/ jospt.2005.35.12.821.
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Shoulder arthroplasty and its effect on strain in the subscapularis muscle Thomas Wright ⁎, Thomas Easley, Jessica Bennett, Aimee Struk, Bryan Conrad Department of Orthopaedics and Rehabilitation, University of Florida College of Medicine, 3450 Hull Road, Gainesville, FL 32611, USA
a r t i c l e
i n f o
Article history: Received 5 March 2014 Accepted 16 February 2015 Keywords: Shoulder arthroplasty Subscapularis muscle Strain Prosthetic humeral head External rotation Abduction
a b s t r a c t Background: Increasing the thickness of the prosthetic humeral head on subscapularis strain in patients undergoing total shoulder arthroplasty has not been elucidated. The optimal postoperative rehabilitation for total shoulder arthroplasty that does not place excessive strain on the subscapularis is not known. We hypothesize that the use of expanded non-anatomic humeral heads during shoulder replacement will cause increased tension in the repaired subscapularis. We identified a recommended passive range of motion program without invoking an increase in tension in the repaired subscapularis, and determined the impact of the thickness of the humeral head on subscapularis strain. Methods: Eight fresh-frozen, forequarter cadaver specimens were obtained. An extended deltopectoral incision was performed and passive range-of-motion exercises with the following motions were evaluated: external rotation, abduction, flexion, and scaption. An optical motion analysis system measured strain in the subscapularis. The same protocol was repeated after performing a subscapularis osteotomy and after placement of an anatomic hemiarthroplasty of three different thicknesses. Findings: For abduction and forward flexion, we observed a trend of decreasing strain of the subscapularis, as the laxity is removed with increasing humeral head component thickness. With the short humeral head, strain was similar to native joint with passive scaption and flexion but not with external rotation or abduction. Interpretation: The passive range of motion that minimizes tension on the subscapularis is forward flexion and scaption. Therefore, passive forward flexion or scaption does not need to be limited, but external rotation should have passive limits and abduction should be avoided. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Total shoulder arthroplasty (TSA) is performed with a surgical approach through the subscapularis. After surgery, rehabilitation of the shoulder is imperative to optimize the functional result. However, the subscapularis must be respected in the process of this rehabilitation or the repair will fail, resulting in a poorly functioning shoulder and in some cases shoulder instability. Using ultrasound after subscapularis tenotomy and repair in the course of performing a TSA, failure rates of the repair have been reported between 13 and 47% (Armstrong et al., 2006; Jackson et al., 2010). Weakness, anterior instability, glenoid loosening, and decreased function have all been described after a failed subscapularis repair (Miller and Kathuria, 2006; Miller et al., 2003, 2005; Moeckel et al., 1993; Utz et al., 2011). In an effort to decrease subscapularis failure rate in the setting of a TSA, many surgeons have begun performing a lesser tuberosity osteotomy as it has been shown ⁎ Corresponding author at: Department of Orthopaedics and Rehabilitation, PO Box 112727, Gainesville, FL 32611-2727, USA. E-mail address: [email protected]fl.edu (T. Wright).
http://dx.doi.org/10.1016/j.clinbiomech.2015.02.010 0268-0033/© 2015 Elsevier Ltd. All rights reserved.
to have a decreased failure rate when compared to direct tendon repair (Qureshi et al., 2008). Rehabilitation protocols are commonly designed to protect the subscapularis repair; however, there is little science supporting their specific recommendations (Wilcox et al., 2005). The purpose of a shoulder replacement is to provide pain relief and increase function in a patient with a pathologic glenohumeral joint. The current trend is to replicate the patient's pre-existing normal shoulder anatomy as closely as possible with the shoulder replacement. However, in an attempt to stabilize the TSA many systems provide the surgeon with options of humeral heads that are much thicker than the normal humeral head. It is likely that the use of tall (thicker) or expanded non-anatomic humeral heads will cause an increase in the tension experienced in the repaired subscapularis. TSA instability is better managed with correction of glenoid version and embrication of the rotator cuff interval rather than by overstuffing the joint with a thick prosthetic head. The primary purpose of this study was to identify the optimal manner to perform a passive range of motion (PRoM) program without invoking a large increase in tension in the repaired subscapularis. The secondary purpose was to determine the impact of varying the
374
T. Wright et al. / Clinical Biomechanics 30 (2015) 373–376
thickness of the humeral head on subscapularis strain using the same PRoM protocol. 2. Methods 2.1. Specimen preparation Eight fresh-frozen, forequarter cadaver specimens (four females and four males) were obtained following International Review Board approval. Each specimen was mounted on a surgical holding device that clamped to the medial scapula eliminating scapular motion. Using an extended deltopectoral approach, the distal subscapularis and insertion site were well visualized. A six-camera motion analysis system (Eagle, Motion Analysis Corp., Santa Rosa, CA, USA) was used to measure strain in the subscapularis. Four 2 mm reflective beads were carefully sutured to the subscapularis in a 1 cm2 square-shaped arrangement to allow tissue strains to be measured. The superior two markers were sutured to the upper subscapularis and the inferior two markers were sutured to the lower subscapularis (Fig. 1). The primary outcome measure for this study was tissue strain of the upper and lower subscapularis. Strain was measured as the peak change in displacement from the resting tissue length, during PRoM movements (initial length − final length = strain in mm). PRoM of the glenohumeral joint was measured using an electromagnetic tracking device (Liberty, Polhemus Inc., Colchester, VT, USA). The Polhemus device allowed the three-dimensional angulation to be tracked and recorded. Two Polhemus tracking sensors were attached: one to the mid-diaphysis of the humerus and one to the spine of the scapula. Before collecting data, the sensors were aligned to the anatomical axes of the specimen. A custom software program written in LabView (National Instruments, Austin, TX, USA) was used to collect data from the Polhemus system and calculate joint angles. 2.2. Experimental procedure Both strain and range of motion data were collected simultaneously as the arms were moved through a PRoM in external rotation, abduction, flexion, and scaption for five cycles of each motion. To ensure that the
tissues were at a steady state, data from the fifth cycle were used for analysis. The range of motion and strain data were synchronized post hoc by using a linear regression procedure to optimize the temporal alignment of the signals. The PRoM protocol was initially performed on each specimen in the intact condition to measure “normal” strain. The starting position for each set of conditions was always with the arm by the side and internally rotated like what would occur with the hand on the abdomen in the resting state. The exact same PRoM sequence was then repeated after placement of a short humeral head with repair of the osteotomy site and again after placement of a tall (thicker) and expanded head (even thicker). The subscapularis osteotomy was repaired back to the exact same spot after each set of conditions. No destructive loading of the subscapularis was performed. All implant components were part of the Equinoxe shoulder arthroplasty system (Exactech Inc., Gainesville, Florida, USA). Eight specimens were tested for all conditions, but two specimens were too small to be tested with the expanded humeral head arthroplasty. After exposure of the subscapularis and placement of the reflective markers, baseline strain was obtained using the above motions. The lesser tuberosity was then osteotomized in a consistent fashion, and a humeral head cut was made on the anatomic neck. The humeral canal was reamed and broached. Prior to placing the humeral stem, three drill holes were placed just lateral to the lesser footprint and three just medial to the footprint. One number 2 FiberWire (Arthrex, Naples, Florida, USA) suture was placed through the lateral hole then around the inserted humeral stem and out the medial hole. This was repeated for all three sets of holes, placing a total of three sutures. Once repaired, each suture acted as a cerclage around the lesser tubercle and stem exactly as is normally performed in surgery. This construct returned the lesser tubercle to its native footprint. Using the system replicator plate and a humeral head sized directly off the removed native head, the arthroplasty head was placed to exactly cover the native humeral head cut. The subscapularis was then repaired in the manner described. One figure of eight number two FiberWire was used in the rotator interval adjacent to the two tuberosities. The same passive motions were then performed with strain noted as in the intact state. The exact same surgical and testing procedures were repeated with the tall (4 mm thicker than short) and then expanded humeral heads (4 mm thicker than tall and 8 mm thicker than short) in each cadaver specimen. 3. Results A general trend of decrease in tissue strain on the subscapularis was observed in abduction and external rotation following implantation of the shoulder arthroplasty components (Figs. 2–5). The apparent paradox of this is due to the tendon being under increased tension (all the laxity has been removed). At point zero there was less tendon laxity,
Fig. 1. Schematic drawing of the experimental setup with markers placed on the subscapularis.
Fig. 2. Tissue strain for upper and lower subscapularis during abduction with different implant sizes.
T. Wright et al. / Clinical Biomechanics 30 (2015) 373–376
Fig. 3. Tissue strain for upper and lower subscapularis during external rotation with different implant sizes.
which resulted in less movement between the markers (decreased strain). For abduction and forward flexion, we observed a trend of decreasing laxity with increasing humeral head component thickness. For external rotation, all components displayed a similar reduction in tissue strain. For the short head, passive flexion and scaption showed strains very close to the normal state, which may suggest that these movements are optimal for postoperative rehabilitation. A summary of peak strains is in Table 1. Similar strain patterns were observed for both the superior and inferior portions of the subscapularis. This occurs because the insertion site is very broad superior to inferior and strain is distributed in all directions. 4. Discussion Most approaches to the shoulder, for the purpose of performing a shoulder arthroplasty, violate the subscapularis muscle tendon unit that is subsequently repaired. This may occur through the tendon, at the bone tendon interface, or through the lesser tuberosity. Failure of the subscapularis repair is a well-described complication of shoulder arthroplasty, which results in decreased function and instability (Jackson et al., 2010; Terrier et al., 2013; Utz et al., 2011). Failure of the upper subscapularis results in altered kinematics that is exacerbated by load, causing the humeral head to translate in an anterior superior direction (Su et al., 2009). Miller et al. (Miller et al., 2003) showed that, despite meticulous attention paid to subscapularis repair, two out of three patients would still have abnormal subscapularis function.
Fig. 5. Tissue strain for upper and lower subscapularis during scaption with different implant sizes.
It is important that we protect subscapularis healing and therefore function after shoulder arthroplasty. The optimal shoulder arthroplasty technique and rehabilitation protocol that will maximize the opportunity for subscapularis healing need to be defined. There are surprisingly few reports of strain on the subscapularis. Muraki et al. (Muraki et al., 2007) looked at subscapularis strain using cadavers with their native joint and found that the lower subscapularis generally experienced greater strain than the upper. In our study we could not clearly demonstrate the differentiation of upper and lower subscapularis strains, but this is probably due to the markers being placed in the upper and middle subscapularis rather than the far inferior lower subscapularis. The markers were not attached to the lower muscular portion of the subscapularis, which resides below the glenohumeral center of rotation, and therefore we could not demonstrate a significant difference in strain between the upper and lower subscapularis as it was not tested. Muraki et al. (Muraki et al., 2007) showed that the subscapularis had the greatest strain in external Table 1 Summary of peak strains in the inferior and superior subscapularis during different movements. Motion
Implant
Inferior medial/lateral (mean ± SD) (mm)
Superior medial/lateral (mean ± SD) (mm)
Abduction
Intact Short Tall Expanded Abduction average
4.1 ± 3 2.2 ± 1 2.7 ± 1.7 1.9 ± 1 3 ± 2.2
4.4 ± 3 3.9 ± 2.5 2.4 ± 1.5 1.3 ± 0.8 3.3 ± 2.5
External rotation
Intact
6.6 ± 6.6
7.2 ± 4.5
Short Tall Expanded External rotation average
3.9 ± 2.8 4.3 ± 3.9 3.1 ± 1.3 4.8 ± 4.6
3.5 ± 2.3 5 ± 3.5 4.1 ± 1.6 5.3 ± 3.6
6 ± 7.1
6.4 ± 3.2
Short Tall Expanded Forward flexion average
5 ± 5.3 3.4 ± 2.1 4.5 ± 3.6 4.9 ± 5.1
5.8 ± 3.5 4.1 ± 2.1 4.5 ± 2.1 5.4 ± 2.9
Intact Short Tall Expanded Scaption average
4.6 ± 2.5 4.1 ± 3 7.5 ± 8.2 5 ± 4.9 5.2 ± 4.8
5.3 ± 4.3 5.8 ± 5.3 9.2 ± 11 8.8 ± 11.9 6.9 ± 7.6
Forward flexion
Scaption
Fig. 4. Tissue strain for upper and lower subscapularis during forward flexion with different implant sizes.
375
Intact
376
T. Wright et al. / Clinical Biomechanics 30 (2015) 373–376
rotation especially with abduction and our data, although not performed in this manner, would corroborate their findings. Pure external rotation in our study was performed with the shoulder in adduction by the side. However, we noted lower strain in the subscapularis (slack removed) with abduction, as did Muraki et al.; the explanation for this is that there is obligatory external rotation with abduction. Muraki et al. did not evaluate strain after shoulder arthroplasty (Muraki et al., 2007). Although there are very few studies that have evaluated strain in the subscapularis, some investigators have reported on strains of other muscles of the rotator cuff. Kim et al. (Kim et al., 2011) used a novel ultrasound technique to measure in vivo tissue strain of the supraspinatus during scaption and found that the greatest strains occurred in the superficial region, which would correspond to the area measured in the current study. Andarawis et al. (Andarawis-Puri et al., 2010) showed that, during abduction, strain in both the supraspinatus and infraspinatus was at a minimum when the shoulder was at thirty degrees abduction. Based on our data, the subscapularis muscle has less strain after shoulder joint arthroplasty. We interpret this mechanism as follows: slack in the subscapularis has been removed because of greater size of the implant components, thus increasing preload tension on the muscle unit. This effect is more pronounced with thicker humeral head implants. This is a concern because the repaired muscle tendon unit could be under greater tension than normal, which might make it prone to failure. There are two possible explanations for this observation. One is that, in the process of repairing the osteotomy site, the site was transferred laterally or the sutures took up some of the subscapularis tendon in the process of passing them and, therefore, caused excessive tension on the subscapularis. The other explanation is that the proximal humeral reconstruction was not perfectly anatomic. The reduced strain trend was even greater with thicker humeral head implants. It appears that even the short head size humeral component trended towards decreased strain in the subscapularis compared to the intact joint. Thicker non-anatomic humeral head implants are provided to manage instability. However, instability of the shoulder should be managed by other means including correction of glenoid version and embrication of the rotator cuff interval rather than over-stuffing the joint. If the repaired muscle tendon unit is under increased tension after shoulder arthroplasty, it becomes imperative that the rehabilitation program not stress this repair until it has had sufficient time to heal. Based on our findings, limiting external rotation, abduction, and particularly the combination of abduction and external rotation is important for the rehabilitation program. With abduction and external rotation the strains decreased because the subscapularis had all its slack removed at time 0. However, passive flexion and scaption without external rotation appear to be safe with the short arthroplasty head demonstrating the same strain as the intact state for these two motions. This represents a small rehabilitation change as most protocols limit external rotation but do not push flexion and scaption. Limitations of this study were that tension could not directly be measured in the subscapularis; rather, strain is used for a surrogate. Additionally, scapular motion was constrained with all motion occurring only at the glenohumeral joint. Because we were dealing with cadaveric tissues it is possible that some stretching of soft tissues occurred in the process of multiple examinations. This was a passive test and it is probable that dynamic contraction of the subscapularis may alter strain further in vivo.
5. Conclusions Despite meticulous surgical technique, strain did not return to the normal state after shoulder arthroplasty and a greater deviation from the normal state was noted with increasing humeral head thickness. Passive scaption and flexion appear to be relatively safe motions for postoperative rehabilitation, whereas abduction and external rotation were not as safe. It is our recommendation to use the thinnest humeral head implant that is practical and limit passive external rotation, avoid passive abduction, and not limit scaption or flexion as long as the shoulder is not externally rotated during these exercises. Based on this study we recommend against the use of expanded (thick) non-anatomic humeral head implants as there are better means for managing TSA instability. Using this protocol should limit excessive strain on the repaired subscapularis and hopefully result in a lower postoperative failure rate. Acknowledgments Funding for this study was from institutional sources. References Andarawis-Puri, N., Kuntz, A.F., Ramsey, M.L., Soslowsky, L.J., 2010. Effect of glenohumeral abduction angle on the mechanical interaction between the supraspinatus and infraspinatus tendons for the intact, partial-thickness torn, and repaired supraspinatus tendon conditions. J. Orthop. Res. 28, 846–851. Armstrong, A., Lashgari, C., Teefey, S., Menendez, J., Yamaguchi, K., Galatz, L.M., 2006. Ultrasound evaluation and clinical correlation of subscapularis repair after total shoulder arthroplasty. J. Shoulder Elb. Surg. 15, 541–548. Jackson, J.D., Cil, A., Smith, J., Steinmann, S.P., 2010. Integrity and function of the subscapularis after total shoulder arthroplasty. J. Shoulder Elb. Surg. 19, 1085–1090. http://dx.doi.org/10.1016/j.jse.2010.04.001. Kim, Y.S., Kim, J.M., Bigliani, L.U., Kim, H.J., Jung, H.W., 2011. In vivo strain analysis of the intact supraspinatus tendon by ultrasound speckles tracking imaging. J. Orthop. Res. 29, 1931–1937. Miller, D., Kathuria, V., 2006. A surgical tip for shoulder hemiarthroplasty in a patient with a deficient rotator cuff. J. Surg. Orthop. Adv. 15, 60–61. Miller, S.L., Hazrati, Y., Klepps, S., Chiang, A., Flatow, E.L., 2003. Loss of subcapularis function after shoulder replacement: a seldom recognized problem. J. Shoulder Elb. Surg. 12, 29–34. Miller, B.S., Joseph, T.A., Noonan, T.J., Horan, M.P., Hawkins, R.J., 2005. Rupture of the subscapularis tendon after shoulder arthroplasty: diagnosis, treatment, and outcome. J. Shoulder Elb. Surg. 14, 492–496. Moeckel, B.H., Altchek, D.W., Warren, R.F., Wickiewicz, T.L., Dines, D.M., 1993. Instability of the shoulder after arthroplasty. J. Bone Joint Surg. Am. 75, 492–497. Muraki, T., Aoki, M., Uchiyama, E., Takasaki, H., Murakami, G., Miyamoto, S., 2007. A cadaveric study of strain on the subscapularis muscle. Arch. Phys. Med. Rehabil. 88, 941–946. Qureshi, S., Hsiao, A., Klug, R.A., Lee, E., Braman, J., Flatow, E.L., 2008. Subscapularis function after total shoulder replacement: results with lesser tuberosity osteotomy. J. Shoulder Elb. Surg. 17, 68–72. http://dx.doi.org/10.1016/j.jse.2007.04.018. Su, W.R., Budoff, J.E., Luo, Z.P., 2009. The effect of anterosuperior rotator cuff tears on glenohumeral translation. Arthroscopy 25, 282–289. http://dx.doi.org/10.1016/j. arthro.2008.10.005. Terrier, A., Larrea, X., Malfroy Camine, V., Pioletti, D.P., Farron, A., 2013. Importance of subscapularis muscle after total shoulder arthroplasty. Clin. Biomech. 28, 146–150. Utz, C.J., Bauer, T.W., Iannotti, J.P., 2011. Glenoid component loosening due to deficient subscapularis: a case study of eccentric loading. J. Shoulder Elb. Surg. 20, e1621. http://dx.doi.org/10.1016/j.jse.2011.03.024. Wilcox, R.B., Arslanian, L.E., Millet, P.J., 2005. Rehabilitation following total shoulder arthroplasty. J. Orthop. Sports Phys. Ther. 35, 821–836. http://dx.doi.org/10.2519/ jospt.2005.35.12.821.