Assessment of the anatomic neck as an accurate landmark for humeral head resurfacing implant height placement.
In theory, by avoiding a humeral head resection, the proximal humeral anatomy is preserved, allowing a more accurate recreation and placement of a HHR. Previous authors have described using the intact humeral anatomic neck to maintain the patient's own inclination and version of the humeral head replacement. (6) While techniques have described placing a HHR parallel with the anatomic neck, (6,7) we are not aware of any description of using the anatomic neck as a reference point for the depth of reaming and the placement of the HHR. This is potentially relevant as one of the more common complications when performing HHR is overstuffing the joint, which can commonly be attributed to inadequate reaming, incomplete implant seating, or to implant design. (3,5)
Iannotti and coworkers validated the use of a sphere on preoperative 3D CT scans and a perfect circle model on postoperative coronal images to asses humeral head anatomy. (8) They used the perfect circle model in a follow-up clinical study showing that surgeons did a better job at recreating the center of rotation (CoR) when using a stemmed arthroplasty compared to when they used HHR. Interestingly, they noted that 89.3% of the time in HHR, there was improper humeral reaming. (9)
The purpose of this study was to use a new resurfacing design along with a perfect sphere model in cadaver CT scans to assess the anatomic neck of the proximal humerus as an accurate landmark for humeral head resurfacing placement, in particular looking at the deviation from the center of rotation and humeral head thickness.
Materials and Methods
Sixty-six cadaveric shoulder CT scans, 34 males (77.0 [+ or -] 8.5 years; BMI = 23.8 [+ or -] 6.7) and 32 females (76.7 [+ or -] 10.5 years; BMI = 21.3 [+ or -] 5.8) were reconstructed using Mimics (Materialise NV, Leuven, Belgium) to create 3D models of the humerus. CTs were taken with 0.5 mm slice thickness. Each 3D reconstructed CT model was analyzed in Rapidform (3D Systems) to create a best fit sphere over the humeral head articular surface for measurement of the anatomic radius of curvature, with the center of the sphere establishing the anatomic CoR, as shown in Figure 1. The anatomic neck plane was defined by selecting points circumferentially around the humeral anatomic neck, as shown in Figure 2, and creating a best fit plane that passed through all points. The anatomic neck angle vector was created for each humerus model by creating a vector normal to the anatomic neck plane, originating from the anatomic CoR. Humeral head resurfacing implant sizes (Equinoxe[R], Exactech, Inc.) were selected for each humeral model to most closely match the anatomic radius of curvature (Fig 3). Each resurfacing implant was virtually assembled onto the humeral head 3D reconstructions using Unigraphics (Siemens, Inc.) by constraining the implant coaxially to the anatomic neck angle vector, with a 2 mm gap distance between the anatomic neck plane and the base of the resurfacing implant. The 2 mm gap was selected based on the Equinoxe[R] system design to seat the HHR implant between 1 mm to 2 mm from the cortical shelf retained by the reamer. If the head is reamed to the depth of the anatomic neck, the result is that the implant will be placed at a 2 mm gap distance from the anatomic neck.
To quantify the deviation from the anatomic CoR, a true AP plane was established by defining a plane passing though the neck angle vector, parallel to the IM axis. The offset distance between the anatomic CoR and the implant CoR was measured with positive x-axis in the medial direction and positive y-axis pointing superior. To evaluate the condition in which the humeral head could be over-reamed, the deviation between the implant humeral head thickness (HHT) from the anatomic head thickness was quantified. The implant HHT was quantified by measuring the distance from the anatomic neck plane to the apex of the resurfacing implant. The anatomic HHT was measured from the anatomic neck plane to the apex of the articular surface. A Student's two-tailed, unpaired t-test was used to identify CoR, and HHT differences between implant sizes where p < 0.05 denoted a significant difference.
The mean implant CoR offset from the anatomic for all specimens was determined to be 1.03 mm [+ or -] 0.75 mm with the plotted x- and y- offset values for each implant (Fig. 4). The mean CoR offset for female specimens was 0.84 mm [+ or -] 0.68 mm, while mean CoR offset in the male group was observed to be higher than females at 1.17 mm [+ or -] 0.74 mm. The average CoR offsets for all specimens were -0.66 mm [+ or -] 0.60 mm in the x-axis and -0.62 mm [+ or -] 0.60 mm in the y-axis, which indicate an average shift in the implant CoR laterally and inferiorly from the anatomic CoR. As described in Table 1, average total CoR offset was higher for the two largest implant sizes; however, these findings were not statistically significant.
The average implant HHT deviation from the anatomic was determined to be -0.36 mm [+ or -] 0.84 mm with the average HHT offsets for each implant size (Table 1). The mean HHT offset for female specimens was determined to be -0.21 mm [+ or -] 0.76 mm while the mean HHT offset for males was -0.49 mm [+ or -] 0.90 mm. The HHT offset for each analysis is shown in Figure 5, which shows some samples with over-reaming conditions lower than 2 mm below the anatomic HHT. There were no statistically significant differences observed between HHT offsets for each implant size used.
The results of this study demonstrate that using the anatomic neck as a landmark for depth of reaming allowed for accurate restoration of the anatomic CoR and HHT with humeral head resurfacing. While it has been described to use the anatomic neck as a guide to determine humeral height positioning, including inclination and version, (6,7) to our knowledge it has not been described to use the anatomic neck as the landmark for determining the appropriate depth of the resurfacing implant position. We believe that this could account, at least in part, for some of the reports of joint overstuffing after a resurfacing procedure (5) and may explain some of the discrepancies found in the literature pertaining to the ability of a HHR to recreate the proximal humeral geometry. The discrepancies may also be implant specific, and our results may also be related to a thinner, more anatomic design. The Equinoxe[R] resurfacing is a modular implant that utilizes a cannulated system to insert a caged peg into the humeral head allowing for bony through-growth and modular assembly to better reconstruct a patient's anatomy (Fig 3). The humeral prosthesis is only 1.5 mm thick, which is noticeably thinner than the traditional 4 mm thick Copeland HHR device. The instrument system is designed so that the HHR implant sits off the remaining cortical shelf formed from the reamer base. By reaming down to the anatomic neck, the implant is automatically placed at a set distance from the anatomic neck when fully seated.
Mechlenburg and associates concluded, after evaluating the Length of the Gleno-Humeral Offset (LGHO) on standardized radiographs, that the Copeland resurfacing increased the postoperative LGHO, and that this overstuffing of the joint led to a high revision rate (14%). (5) Mansat and colleagues, however, found that HHR did restore the humeral anatomy in their radiographic study, (3) and in a computer model of non-arthritic joints, Hammond and coworkers concluded that the HHR more closely restored the geometric center than stemmed hemiarthroplasty when using an Arthrosurface[R] implant. (4)
In their clinical study, Alolabi and associates reported that the average deviation of the CoR for HHR was 3.8 mm [+ or -] 2.1 mm. (9) When using a computer model to identify and place the humeral component, we found that our average deviation was decreased to 1.03 mm [+ or -] 0.75 mm. Obviously, some of this improvement can be explained by the ease of identifying and placing the prosthesis virtually in the computer model, as opposed to identifying the anatomic neck of a deformed humerus intraoperatively and placing a prosthesis based on this. Alolabi and associates stated in their study that finding the anatomic neck in vivo is often difficult and could explain the increased deviation. (9) Whether due to difficulty identifying the anatomic neck or failure to ream to the proper depth, they showed that 89.3% of HHR had improper reaming. (9) We believe that by using the anatomic neck as a landmark for humeral head reaming and subsequent placement of a new, thinner HHR, we are better able to recreate the proximal humeral anatomy decreasing the deviation of the CoR and HHT.
There are several limitations of this study. First, it is a computer model using non-arthritic shoulders. It is much harder to identify the anatomic neck intraoperatively on an arthritic or other deformed shoulder, and future studies will need to evaluate if this technique can be safely and reliably reproduced clinically. Secondly, the use of the anatomic neck as a guide for the reaming depth of the HHR may be design specific, and further studies need to be performed specifically comparing different prosthetic designs placed at the same anatomic location. Although the computer study shows that the HHT can be accurately reproduced, it will need to be determined, clinically, whether or not reaming the humerus to the anatomic neck removes too much subchondral bone to implant a HHR safely. Finally, more clinical studies will need to be performed to evaluate HHR outcomes and whether decreasing the deviation from the CoR improves outcomes.
This computer model demonstrated that using the anatomic neck as a landmark for the depth of reaming with HHR accurately restores the CoR and HHT of the proximal humerus, better than previously reported. Future studies should investigate if this method for HHR placement can be reproduced in situ and if these results improve clinical performance.
Conflict of Interest Statement
Ryan Simovitch, M.D., Felix Savoie, M.D., and Curtis R. Noel, M.D., are consultants for Exactech, Inc., and receive royalties on products related to this article. Emmon J. Chen, M.S., is an employee of Exactech, Inc., Gainesville, Florida.
(1.) Levy O, Copeland SA. Cementless surface replacement arthroplasty of the shoulder: 10-year results with the Copeland Mark-2 Prosthesis. J Bone Joint Br. 2001 Mar;83(2):213-21.
(2.) Mullett H, Levy O, Raj D, et al. Copeland surface replacement of the shoulder. Results of an hydroxyapatite-coated cementless implant in patients over 80 years of age. J Bone Joint Surg Br. 2007 Nov;89(11):1466-9.
(3.) Mansat P, Coutie AS, Bonevaile N, et al. Resurfacing humeral prosthesis: do we really reconstruct the anatomy? J Shoulder Elbow Surg. 2013 May;22(5):612-9.
(4.) Hammond G, Tibbone JE, McGarry MH, et al. Biomechanical comparison of anatomic humeral head resurfacing and hemiarthroplasty in functional glenohumeral positions. J Bone Joint Surg Am. 2012 Jan 4;94(1):68-76.
(5.) Mechlenburg I, Amstrup A, Klebe T, et al. The Copeland resurfacing humeral head implant does not restore humeral head anatomy. A retrospective study. Arch Orthop Trauma Surg. 2013 May;133(5):615-9.
(6.) Thomas SR, Sforza G, Levy O, Copeland SA. Geometrical analysis of
Copeland surface replacement shoulder arthroplasty in relation to normal anatomy. J Shoulder Elbow Surg. 2005 Mar-Apr;14(2):186-92.
(7.) Jensen KL. Humeral resurfacing arthroplasty: Rationale, indications, technique, and results. Am J Orthop(Belle Mead NJ). 2007 Dec;36(12 Suppl 1):4-8.
(8.) Youderian AR, Ricchetti ET, Drews M, Iannotti JP. Determination of humeral head size in anatomic shoulder replacement for glenohumeral osteoarthritis. J Shoulder Elbow Surg. 2014 Jul;23(7):955-63.
(9.) Alolabi B, Youderian AR, Napolitano L, et al. Radiographic assessment of prosthetic humeral head size after anatomic shoulder arthroplasty. J Shoulder Elbow Surg. 2014 Nov;23(11):1740-6.
Emmon J. Chen, M.S., Ryan Simovitch, M.D., Felix Savoie, M.D., and Curtis R. Noel, M.D.
Emmon J. Chen, M.S., Exactech, Inc., Gainesville, Florida. Ryan Simovitch, M.D., The Shoulder Center at Palm Beach Orthopaedic Institute, Palm Beach Gardens, Florida. Felix Savoie, M.D., Tulane University, New Orleans, Louisiana. Curtis R. Noel, MD, Crystal Clinic Orthopaedic Center, Akron, Ohio.
Correspondence: Curtis R. Noel, M.D., email@example.com.
Caption: Figure 1 Computer best-fit sphere derived from the humeral head articular surface 3D computer models. The center of the sphere establishes the anatomic CoR. AP view shown on top, axillary view shown below.
Caption: Figure 2 Computer image demonstrating the creation of the anatomic neck plane. Multiple points are selected around the humerus along the anatomic neck, and a computer best-fit plane is placed through all points selected to establish the anatomic neck plane.
Caption: Figure 3 Equinoxe[R] Resurfacing Humeral Head (Exactech, Inc., Gainesville, FL)
Caption: Figure 4 Implant CoR deviation from the anatomic CoR in the xand y-axes of all specimens. The anatomic CoR is depicted as the plot origin (0,0). The x-axis represents offset in the medial-lateral direction with positive indicating medial. The y-axis represents offset in the superior-inferior direction with positive indicating superior. Both axes are in millimeters.
Caption: Figure 5 Implant HHT deviation from the anatomic HHT, grouped by implant sizes used for each specimen. Positive offset values indicate resurfacing implant placement above the anatomic articular surface, and negative offset indicates placement below the native articular surface.
Table 1 The Average Offset of the Implant CoR from the Anatomic CoR and the Average Deviation of the Implant HHT from the Anatomic HHT of the 3D Reconstructed Humerus Model for Each Resurfacing Implant Size Used Implant CoR Offset * HHT Average Deviation Size ([+ or -] Standard ([+ or -] Standard N Deviation) Deviation) 41 mm 0.89 mm [+ or -] 0.76 mm -0.39 mm [+ or -] 0.71 mm 7 44 mm 0.89 mm [+ or -] 0.73 mm -0.17 mm [+ or -] 1.00 mm 9 47 mm 0.88 mm [+ or -] 0.66 mm -0.31 mm [+ or -] 0.72 mm 19 50 mm 1.35 mm [+ or -] 1.06 mm -0.73 mm [+ or -] 0.61 mm 8 53 mm 1.15 mm [+ or -] 0.72 mm -0.33 mm [+ or -] 0.99 mm 23 * Absolute value.
Please note: Illustration(s) are not available due to copyright restrictions.
|Printer friendly Cite/link Email Feedback|
|Author:||Chen, Emmon J.; Simovitch, Ryan; Savoie, Felix; Noel, Curtis R.|
|Publication:||Bulletin of the NYU Hospital for Joint Diseases|
|Date:||Oct 1, 2015|
|Previous Article:||Role of subscapularis repair on muscle force requirements with reverse shoulder arthroplasty.|
|Next Article:||Optimization of cemented glenoid peg geometry: a comparison of resistance to axial distraction.|