Clear information regarding microarchitecture of the GT might influence the choice of anchoring techniques and implant positioning. Although the success of RC surgery is majorly determined by the stability of the suture tendon interface and the footprint restoration, evidence grows in the very recent past that the security of anchorage has been overestimated [12–15]. Therefore, poor bone quality may compromise the success of RC repair by promotion of suture anchor loosening and thereby impaired tendon healing [6, 9, 14].
Therefore, the trabecular architecture of the GT of osteoporotic humeral heads using HR-pQCT was analyzed in the present study. In accordance with previous work, this study also revealed higher values of bone volume to total volume ratio (BV/TV) in the posteromedial GT-region in comparison to the anterolateral portion . In addition, further crucial bone quality parameters such as trabecular number (Trab N), trabecular thickness (Trab Th) as well as connectivity density (Conn Dens) were assessed in the presented work resulting in a distinct pattern regarding the regional distribution within the GT. In comparison between both genders, the posterior quadrant of the GT, directly adjacent to the articular surface, appeared to be of even better bone quality in males. Regarding the comparison between the dominant and non-dominant shoulder, no significant differences for any bone quality parameter resulted. Moreover, a distinct inverse correlation of BV/TV and age was found with a more pronounced character in the female specimens.
In the present study, a HR-pQCT imaging system from Scanco Medical, Switzerland, was used with an isotropic voxel size of 82 μm providing a direct analysis of BV/TV, Trab N, Trab Th and Conn Dens. HR-pQCT is a non-destructive technique allowing for measurements of important micro-architectural parameters of bone morphology. Several authors clearly demonstrated the biomechanical consequence of the assessed parameters [16–19]. In this context, Schiuma et al. investigated the primary stability of proximal humeral nails (MultiLoc PHN) in correlation to bone properties assessed by HR-pQCT. The authors found a significant correlation between axial displacement and bone properties at the primary proximal screws .
There are several studies in the literature trying to characterize the bone architecture of the GT [5, 7, 20, 21]. However, most of these studies suffer from significant technical drawbacks such as Barber et al., who were trying to correlate biomechanical data for anchor pullout with BMD of the GT, assessed by dual-energy X-ray absorptiometry (DXA) . Although they found higher pullout forces in the posterior GT part, they failed to show any significant differences between the anterior and posterior GT areas because of the DXA technique used and the consecutively impossible measurement of microarchitecture that only evaluated bone mineral density . On the other hand, Tingart et al. performed pQCT-scans with a pixel size of 0.59 mm and a slice thickness of 2.5 mm, thereby acquiring scans of inferior resolution .
One of the relevant characteristics of bone quality is its microarchitecture, which is influenced by bone turnover and mineralization. Lill et al. reported a correlation of increasing age and decreasing BMD . In contrast to our results, they found the highest BMD in the medial and posterior regions of the cranial horizontal slice in women. They also stated that a BMD decrease for women older than 70 years is constant over all regions of the proximal humerus. In this context, our data are more in line with Barvencik et al. reporting an accentuation of bone loss in the lateral humeral head . In our study, BV/TV was higher in the medial compared to the lateral region, independent from age and sex. This is in accordance with the work of Tingart et al., however they used a pQCT with inferior resolution and thus their results are not really comparable . Khosla et al. stated that the microstructural basis for the trabecular volume differs between the sexes whereas the BV/TV declined similarly in men and women with advancing age . Regarding the number of trabeculae, the authors reported a decrease of Trab N with a corresponding increase of Trab Sep in terms of a trabecular loss in women, whereas the primary mechanism for the decrease of BV/TV is found in trabecular thinning in men. These facts have a significant impact on the age-related bone strength, since based on a finite element, modeling a decrease in Trab N has a 2- to 5-fold greater impact on bone strength in comparison with Trab Th reduction resulting in a similar decrease in bone volume . Our results are in line with these findings, as for our rather old (mean age 72 yrs) specimens a significantly higher BV/TV and Trab N was found for the males, whereas no significant difference was found for Trab Th between both genders. Although the trabecular microarchitecture of the humeral head was assessed non-invasively using HR-pQCT, the presented results in female specimens are consistent with the work of Parfitt et al., who analyzed transiliac biopsies of pre- and postmenopausal women using histomorphometry . Similarly to our data, they showed that postmenopausal women had significantly reduced BV/TV along with reduced Trab N and correspondingly increased Trab Sep. These facts formed the basis for the possible hypothesis that a menopause-associated increase in osteoclastic activity leads to a loss of all trabeculae with a corresponding conversion of the continuous trabecular network characteristic for young women into the discontinuous network of older women . Since we reported similar changes in bone quality at the humeral head using high-resolution pQCT of rather old postmenopausal females, an effect of menopause on this process cannot be excluded but seems rather probable. In contrast, there are micro-CT studies with discordant findings highlighting the need for further studies to analyze the changes of the trabecular structure in women across the menopausal transition . Regarding the male specimens, these findings are consistent with those published by Aaron et al. on cadaveric transiliac bone biopsies describing a parallel decrease of BV/TV with advancing age in men and women and similar to the presented data a significant decrease of Trab N in the course of a lifetime in women, but not in men . Moreover, the presented data is also consistent with the findings on Trab Th as described by Khosla et al. showing a more pronounced decrease of Trab Th at the wrist in men compared to women in the course of a lifetime .
Bone quality is an important factor not only for fracture management but also for the preoperative planning of RC repair, especially in the elderly population. There are currently no concepts to quantify the local degree of osteoporosis of the humeral head before surgical treatment. However, in their study Diederichs et al. reported indications for the existence of alternative BMD measurement sites in case of proximal humerus fractures using QCT, such as the contralateral side helping the physician to choose the adequate surgical procedure for treatment . In their study, a high correlation of BMD values between the right and left proximal humerus resulted. Our results are in line with theirs since the comparison of bone quality parameters of the right and left humeral head assessed with HR-pQCT revealed no significant differences, neither for gender nor for any bone quality parameter, which might be due to similar biomechanical loading conditions.
With increasing age, the overall bone mass decreases, leading to a reduction in trabecular thickness and trabecular connectivity . This reduction in bone mass and bone quality is predominantly seen in women older than 50 years and older in comparison to men . These changes seem to be most prominent in the proximal regions of the long tubular skeletal bones such as the proximal femur, the tibial head and also in the humeral head, as shown in the study by Hepp et al. . However, the results of the histomorphometric analysis of these other mentioned skeletal regions cannot be transferred directly to the situation of the humeral head. The knowledge about the total mineral density of the humeral head mostly derived from DXA is not sufficient to differentiate between regions with adequately stable bone versus bone with minor mechanical strength. This information on bony mechanical strength which is considered to be a function of age and gender is essential to achieve minimally invasive but maximally stable osteosynthetic results in case of fractures as well as stable suture anchor fixation in case of RC tear.
Amling et al. and Mosekilde et al. reported a correlation between increasing age and diminished BV/TV in females for spinal and proximal femoral specimens [30, 32]. The presented analysis for BV/TV and trabecular network of the humeral head showed a relationship between age and gender as well as bone quality and quantity. In the female specimens, a significantly inverse correlation between age and BV/TV was found, whereas such an inverse correlation was not as pronounced in male specimens. This inverse correlation between the genders can be explained by a reduction in trabecular number which can be derived from the significantly inverse correlation of trabecular number with advancing age in females. Another reason might be the reduction of trabecular node number with a significantly negative correlation in females as described by Hepp et al. which is partly a contrast to our results, since we did not find any inverse correlation between Conn Dens and age for both genders .
One potential drawback of our study was the enrollment of shoulder specimens with a macroscopically intact RC only. Of course, this situation does not reflect the actual clinical situation and future examinations need to focus on this problem which will be part of ongoing trials of our study group.