Abstract

Background Total shoulder arthroplasty (TSA) is considered a successful curative procedure for many stiff painful shoulder disorders. However, it may be associated with many complications. Glenoid loosening is thought to be the most common complication of anatomic total shoulder arthroplasty (ATSA); its underlying causes could be mechanical (abnormal loading), septic (infections) or aseptic (autoimmune reaction). This study discusses the mechanical glenoid component loosening after ATSA. II. Hypotheses (Hi, Hii & Hiii) (Hi) The recorded mean peak pressure values of the ATSA components are expected to vary greatly according to the motion type; (Hii) the recorded mean peak pressure values are expected to vary between the initial and final phases of each tested specimen; and (Hiii) the occurrence of glenoid component loosening and its degree of extension are expected to be related to the changes of the obtained mean peak pressure values. III. Objectives This study’s aim was to conduct a comprehensive experimental biomechanical evaluation of the stability of ATSA components under phasic cyclic loading, as follows: (i) testing of the degree of artificial glenoid component stability under repetitive phasic cyclic loading; (ii) testing of the relation between the criteria of the applied cyclic loading according to our testing plan and the occurrence of glenoid component loosening; (iii) measurement and assessment of the values, patterns and magnitudes of the contact pressure between the joint components under cyclic loading; (iv) comparison between the obtained mean peak contact pressure values under cyclic loading in the initial and final phases to detect any relations and/or differences; (v) correlation of the measured pressure values during testing with the QCT findings with respect to glenoid component loosening. IV. Materials A series of six fresh-frozen complete cadaveric shoulder joint specimens (bones and soft tissues) was used in this study. The specimens were implanted with ATSA components and tested successively by mounting them on the shoulder simulator. To measure the values mentioned above, we used a TekScan system with a group of two-headed pressure sensor foils, QCT, shoulder pointer and a digitalized 3Dimaging Zebris system with US, in addition to the routinely used surgical and lab instruments in such experiments. V. Methodology The specimens were scanned prior to experimentation to evaluate their articular surfaces morphology. Then the specimens were implanted with ATSA components and a pressure sensor was inserted within the joint cavity of each specimen and situated on the glenoid component surface. The six specimens were successively mounted on the shoulder simulator and each was tested through three phases of cyclic loading in the three directions of motion. The 1st and 3rd short phases took place for each specimen with insertion of a pressure sensor within the joint cavity, while the 2nd long phase took place without sensor insertion. After the completion of all experiments, the specimens were again scanned with QCT to evaluate the position of the implanted glenoids and any presence of radiolucency and/or loosening. VI. Findings (Observations & Examinations) Two specimens were severely unstable during testing, even with the application of lower loads, particularly during abduction/adduction motion cycles. Provisional and/or subsequent controlling physical examinations revealed either a malposition of the glenoid component or a suspected abnormal glenoid morphology. The other four specimens were completely stable during testing in all motion directions with the application of different loading forces and stabilizing weights. Four specimens were radiologically determined to have a massive glenoid component loosening after the completion of testing. VII. Results The recorded mean peak pressure values varied greatly between the testing phases, testing cycles and motion directions. The highest mean peak pressure values were recorded during AA testing episodes, followed by FE testing episodes. The lowest mean peak pressure values were recorded during IE testing episodes. However, high mean peak pressure values were also recorded during IE testing episodes, but with a low frequency. In seven testing episodes, the recorded mean peak pressure values were higher by 16.7 % in all directions of motion in the final testing phase than those recorded in the initial phase of all testing episodes (42 testing episodes). According to the computed t-test values between the initial and final phases per motion direction/per specimen, null hypothesis (Hypothesis (Hii)) was accepted in the whole AA & FE testing cycles with a percentage of 100%, while it was rejected in only one relation of IE testing cycles with a percentage of 5% and accepted in six relations of IE testing cycles with a percentage of 95%. In total, null hypothesis (Hypothesis (Hii)) was rejected in only one relation of the testing cycles, with a percentage of 5%, and accepted in twenty relations of the testing cycles with a percentage of 95%. According to the calculated t-test values between all initial and final phases for each specimen, null hypothesis (Hypothesis (Hii)) was rejected in two experiments with a percentage of 28.6% and accepted in five experiments with a percentage of 71.4%. Four specimens (three keeled and one pegged) were found to be loose, representing 66.7% of all specimens; one of them was unstable during the testing, representing 25% of the loose specimens and 16.7% of all specimens. VIII. Conclusion The recorded mean peak pressure values and load quantities of the tested shoulder joint varied greatly between motion phases, motion cycles and motion types. The resulting contact pressures across the shoulder joint during its action varied greatly according to the acting force, motion type, muscles status and pathologies within the joint and were directly proportional to the motion type, being higher during AA and FE motion cycles than during IE motion cycles. Also, they were directly proportional to the contact surface area and to the degree of compression between joint articulating surfaces during motion. The greatest degree of variability of SD and mean peak pressure values was seen during FE testing cycles. Shoulder joint instability after ATSA could result from component malposition and/or the articular surface morphological abnormalities. Both glenoid loosening and joint instability could incite the occurrence of the other and could worsen its course in a devastating vicious circle. We concluded that glenoid component loosening could be related to joint stability, loads and the mode of load application in relation to the application duration, and to some extent to the component type, which was apparently evident in our study. The first and third hypotheses were approved, while the second hypothesis was statistically rejected (according to the computed t-test values), which may require a further evaluation in future studies.