Glenoid Bone Loss in Primary Total Shoulder Arthroplasty: Evaluation and Management

Review Article Glenoid Bone Loss in Primary Total Shoulder Arthroplasty: Evaluation and Management

Abstract Benjamin W. Sears, MD Glenohumeral osteoarthritis is the most common reason for Peter S. Johnston, MD shoulder replacement. Total shoulder arthroplasty provides reliable pain relief and restoration of function, with implant survivorship Matthew L. Ramsey, MD reported at 85% at 15 years. Glenoid component wear and aseptic Gerald R. Williams, MD loosening are among the most common reasons for revision. Glenoid wear characteristics have been correlated with, among other things, the degree of anatomic glenoid version correction. Anatomic glenoid reconstruction is particularly challenging in the presence of glenoid bone deficiency. Walch classified glenoid morphology into five types: type A, centered, without posterior subluxation but with minor erosion (A1) or major erosion (A2); type B, posteriorly subluxated (B1) or posteriorly subluxated with posterior glenoid erosion (B2); and type C, excessive glenoid retroversion. The type A glenoid represents only 59% of patients; thus, the need to address glenoid deformity is common. Methods of correction include asymmetric reaming of the anterior glenoid, bone grafting of the posterior glenoid, and implanting a specialized glenoid component with posterior augmentation. In many cases of type C or hypoplastic glenoid, the humerus is concentrically reduced in the deficient glenoid and glenoid deformity may not need to be corrected. Severely hypoplastic glenoid may require the use of bone-sparing glenoid components or reverse total shoulder arthroplasty.

he rate of total shoulder arthro- most commonly used prostheses,3 a Tplasty (TSA) continues to in- meta-analysis of long-term studies crease in the United Sates, with ap- demonstrated the revision burden to proximately 20,000-25,000 surgeries be approximately 7%.4 Glenoid performed annually and doubling in component failure (ie, aseptic loosen- frequency approximately every 7 ing) is the most common complica- From Thomas Jefferson University 1 Medical Center, The Rothman years. Osteoarthritis (OA) remains tion following shoulder replacement Institute, Philadelphia, PA. the most common indication for and accounts for most unsatisfactory shoulder replacement, and in 95% of results.2 Factors that contribute to J Am Acad Orthop Surg 2012;20: 604-613 these patients, shoulder arthroplasty component loosening include altered is associated with pain relief, im- joint reaction forces, component http://dx.doi.org/10.5435/ JAAOS-20-09-604 proved function, and patient satis- malposition, and insufficient bony faction.2 Although implant survivor- support of the implanted glenoid Copyright 2012 by the American 2-4 Academy of Orthopaedic Surgeons. ship is estimated to be >85% at prosthesis. minimum follow-up of 15 years for Glenoid component malposition-

604 Journal of the American Academy of Orthopaedic Surgeons Benjamin W. Sears, MD, et al ing increases stress forces across the ondary to inadequate bone support morphology. Type A glenoid mor- component and is a common reason and incomplete component seating.7,8 phology, which was seen in 59% of for failure. In their analysis of 80 In addition, the use of asymmetric patients, consisted of a centered hu- successful and 53 failed TSAs, reaming to manage glenoid bone loss meral head in the socket, with aver- Moskal et al5 found no instances of results in reduced amount of bone age glenoid retroversion of 11.5°. associated glenoid malversion among remaining for fixation, which puts These patients were subclassified by the successful replacements. How- the patient at risk of glenoid vault degree of glenoid erosion, whether ever, 46% of the failed TSAs were penetration. Unrecognized cortical minor (A1) or major (A2). Type B, found to have glenoid component penetration can result in cement ex- the next most common glenoid mor- malversion. Shapiro et al6 demon- trusion. In addition, unless the fixa- phology, was associated with poste- strated that placement of the glenoid tion device (ie, peg, screw, flute, keel) rior subluxation of the humeral head in 15° of retroversion significantly gains purchase of the perforated and asymmetric joint force distribu- decreased the glenohumeral contact vault, the initial fixation is likely to tion. Type B1 morphology is charac- area (P < 0.0002), increased contact be compromised. Uncorrected bone terized by a narrowed posterior joint pressures (P < 0.0002), and de- loss may also result in insufficient space and lack of posterior erosion. creased inferior and posterior gleno- seating at the periphery of the gle- Type B2 is characterized by posterior humeral forces (P < 0.02 and P < noid component, creating the poten- erosion and an associated biconcave 0.03, respectively), in cadaver speci- tial for eccentric loading of the im- appearance (Figure 2). Type C gle- mens. This raises concern for long- plant and increased stress levels at noid is hypoplastic, with retrover- term component viability. Farron the implant-bone interface. sion >25°. et al7 used finite element analysis to OA alters glenohumeral biome- Types B2 and C demonstrate the evaluate torque at the cement-bone chanics and modifies the anatomic most advanced changes in glenoid interface with increasing degrees of features of the glenoid. Over time, morphology and are the most diffi- glenoid retroversion. They reported these changes result in increased pe- cult to manage. To effectively treat an increase in micromotion at the ripheral glenoid contact stresses and these patients, the surgeon must have cement-bone interface of >700% and posterior glenoid bony erosion, lead- a thorough understanding of the de- stress increase of 326% with compo- ing to increasing glenoid retroversion gree of anatomic change, including nent retroversion >10°. The authors and posterior decentralization of the retroversion, degree of bone loss, concluded that glenoid retroversion humeral head from the socket. These and amount of humeral head dis- >10° should be corrected and that, if findings are common in glenohu- placement. Advanced diagnostic im- version correction is not possible, the meral OA and are seen in up to 40% aging, notably three-dimensional surgeon should consider not per- of patients with advanced disease.9 (3D) reconstruction CT, is an impor- forming glenoid resurfacing. Al- These anatomic changes add to the tant tool to evaluate glenoid bone though controversy remains regard- technical difficulty of correct glenoid stock and version.10,11 Surgical man- ing the effect of glenoid malversion component implantation and affect agement of glenoid retroversion and on implant longevity, the reported the longevity of the implant. bone loss depends on severity. The basic science and clinical literature to In 1999, Walch et al9 examined most common method to correct ver- date support restoration of normal 113 axillary CT scans of arthritic sion involves eccentric reaming of glenoid version as a means of im- glenoids and developed a classifica- the anterior glenoid, or high side, to proving component durability.2-7 tion system based on wear patterns a more neutral version. However, Glenoid bone loss has been shown and version (Figure 1). This classifi- with significantly increased version to contribute to glenoid component cation has become the most com- or bone loss, this method can result malpositioning and loosening sec- monly used to describe glenoid in removal of a substantial amount

Dr. Ramsey or an immediate family member has received royalties from, serves as a paid consultant to or is an employee of, and has received research or institutional support from, Integra (Ascension), and serves as a board member, owner, officer, or committee member of the American Academy of Orthopaedic Surgeons, the Philadelphia Orthopaedic Society, the Rothman Institute, and the Rothman Specialty Hospital. Dr. Williams or an immediate family member has received royalties from, is a member of a speakers’ bureau or has made paid presentations on behalf of, and serves as a paid consultant to or is an employee of, DePuy; has received research or institutional support from Tornier; has stock or stock options held in InVivo Therapeutics; and serves as a board member, owner, officer, or committee member of the American Shoulder and Elbow Surgeons and the Pennsylvania Orthopaedic Society. Neither of the following authors nor any immediate family member has received anything of value from or has stock or stock options held in a commercial company or institution related directly or indirectly to the subject of this article: Dr. Sears and Dr. Johnston.

September 2012, Vol 20, No 9 605 Glenoid Bone Loss in Primary Total Shoulder Arthroplasty: Evaluation and Management

Figure 1

Illustration of the Walch classiﬁcation of glenoid morphology, which is based on wear patterns and version. A, Type A1, centered humeral head with minor glenoid erosion. B, Type A2, centered humeral head with major glenoid erosion. C, Type B1, posterior subluxation with no erosion. D, Type B2, posterior erosion with a biconcave glenoid. E, Type C, severe retroversion.

of bone. In these cases, either bone ble or glenoid augmentation is re- Figure 2 grafting or use of an augmented quired. component may be required. Recog- Standard radiographs are necessary nition and management of altered for preoperative planning. AP, trans- glenoid morphology and diminished scapular lateral, and axillary lateral bone stock are important for success- views should be obtained. The AP ful shoulder arthroplasty. radiograph, which is often performed in internal and external rota- Preoperative Evaluation tion, is used to assess bone quality, note the presence of osteophytes, and Preoperative planning is essential to determine the diameter of the hu- manage bone deficiencies and defor- meral canal. The axillary view pro- mity correction in TSA. Although vides information on posterior plain radiographs continue to be the glenoid wear and subluxation. How- standard diagnostic modality, multi- ever, the axillary view has been CT scan of a type B2 glenoid. Note axial 3D imaging has enhanced the shown to overestimate retroversion the biconcave glenoid surface and interpretation of bony architectural in up to 86% of patients.12 approximately 50% posterior changes associated with shoulder Early descriptions used CT scans subluxation of the humeral head. OA. In patients with posterior ero- to accurately characterize the anat- sion, CT provides the most definitive omy and version of both nonarthritic noid version using the transverse axis evaluation of glenoid version, vault and arthritic glenoids.12-14 Friedman of the scapula, defined as a line anatomy, and severity of bone et al13 were the first to use CT to drawn from the tip of the medial loss.10-13 This assessment is helpful in characterize the association between border of the scapula to the mid- determining whether the use of a glenoid retroversion and OA. They point of the glenoid fossa. Randelli standard glenoid prosthesis is feasi- introduced a method to calculate gle- and Gambrioli14 also proposed a

606 Journal of the American Academy of Orthopaedic Surgeons Benjamin W. Sears, MD, et al

Figure 3 posterior edge), and neoglenoid (pos- neutral articulation between the gle- terior erosion surface) (Figure 4). noid implant and the humeral im- These authors determined that the plant, with the aim of providing con- combination of the Friedman gruent contact between the bone and method and the intermediate line the glenoid prosthesis (Figure 5). provide the most reliable determina- Cannulated reaming systems allow tion of glenoid version in the pres- placement of a guide pin to assess ence of posterior wear. However, al- planned version correction before though CT is the most accurate reaming. These have become popular modality with which to assess gle- to help guide reaming in the correct noid version, measurements of ver- version. sion have been shown to vary by as Although eccentric reaming is com- much as 10° with only minor monly used to address minor changes in scapular rotation in the changes in glenoid version, there are coronal plane.16 no clear guidelines regarding the Recent work has demonstrated the amount of glenoid erosion that can value of 3D CT to accurately charac- safely be corrected with this method. terize glenoid morphology compared Severe retroversion requires the re- with standard two-dimensional moval of increasing amounts of ante- CT.10,11 Scalise and colleagues10,11 rior bone, thereby creating a more have published extensively on the severe glenoid deficiency. This can utility of 3D CT in preoperative lead to insufficient version correction planning and have developed soft- and placement of the glenoid compo- ware to estimate physiologic glenoid nent in residual retroversion. The im- Illustration of the landmarks used to assess glenoid version on CT. A version in patients with severe OA portance of version correction was line (ie, Friedman line [blue]) is and deformity with a glenoid vault demonstrated by Lazarus et al,17 who drawn perpendicular to the axis model. At our institution, 3D CT is reported that insufficient version cor- along the glenoid surface to deﬁne ordered for patients with radio- rection was the most common cause neutral version (yellow line). The angle from either the posterior graphic evidence of combined gleno- of poor seating of the glenoid com- (retroversion) or anterior humeral subluxation and posterior ponent in their series. Additionally, (anteversion) margins of the glenoid erosion. Three-dimensional extensive removal of glenoid bone glenoid indicates the native version. The black dashed line lies reconstruction with humeral or scap- may jeopardize component stability, across the face of the glenoid. The ular subtraction is useful to further leaving inadequate bone stock to red scapula line extends parallel to characterize bony anatomy in cases support the implant. Removal of the scapula body. of severe deformity and can provide bone also effectively medializes the a template to fabricate a reference glenohumeral joint line, thereby al- technique that uses the scapular model for preoperative planning and tering both glenohumeral biome- body line to calculate glenoid ver- intraoperative use. chanics and soft-tissue tensioning. sion. In both of these methods, a line Although humeral lateralization may drawn perpendicular to the axis be improved by increasing the size of along the glenoid surface was de- Eccentric Reaming the humeral head, the effect of in- fined as neutral version, and the an- creased humeral component size is gle from either the posterior margin The goals of glenoid resurfacing in- unknown. or the anterior margin of the glenoid clude correction of version and pro- Several studies have attempted to defined the native version (retrover- vision of adequate bone to support investigate limits to eccentric ream- sion and anteversion, respectively) the implant. The most common ing for posterior glenoid erosion. (Figure 3). More recently, Rouleau method to correct version involves Clavert et al18 eccentrically reamed et al15 characterized the glenoid sur- eccentric reaming of the anterior gle- five cadaver shoulders to create gle- face with eccentric wear using three noid to the level of the posterior sur- noid retroversion of ≥15° and im- new reference lines: the paleoglenoid face to recreate glenoid version and planted a four-pegged glenoid com- (original glenoid surface), intermedi- re-center the humeral head. This ponent in each. Subsequent CT scans ate glenoid (line from anterior and technique effectively creates a new of the shoulders demonstrated peg

September 2012, Vol 20, No 9 607 Glenoid Bone Loss in Primary Total Shoulder Arthroplasty: Evaluation and Management

Figure 4

Illustrations demonstrating characterization of the glenoid surface with eccentric wear using three new reference lines: Walch type B2 glenoid with posterior erosion (A), paleoglenoid (B), intermediate glenoid (C), and neoglenoid (D). The inset images indicate that the cross section in each panel is at the midportion of the glenoid. (Adapted with permission from Rouleau DM, Kidder JF, Pons-Villanueva J, Dynamidis S, Defranco M, Walch G: Glenoid version: How to measure it? Validity of different methods in two-dimensional computed tomography scans. J Shoulder Elbow Surg 2010;19[8]:1230-1237.) penetration in all specimens which, Nowak et al20 used 3D models cre- tion following surgical correction of in theory, limits the cement mantle ated from CT scans of patients with glenoid alignment in the transverse around each peg and may be a factor advanced glenohumeral OA to simu- plane and soft-tissue balancing. Ger- in glenoid component loosening. Ul- late asymmetric reaming and resur- ber et al22 reported that eccentric timately, the authors concluded that facing in varying degrees of glenoid reaming resulted in correction of retroversion ≥15° cannot be cor- retroversion. They used an in-line posterior humeral subluxation in 21 rected with eccentric reaming. In an- three-peg glenoid component as the of 23 patients (91%). No correlation other cadaver study, Gillespie et al19 resurfacing implant. Correction of was found between the ability to re- simulated posterior glenoid wear in glenoid retroversion <12° did not re- center the humeral head and the de- 5° increments, then eccentrically sult in vault penetration. However, gree of preoperative glenoid version reamed to neutral and implanted all glenoids with retroversion >18° or correction. pegged glenoid components. In four demonstrated component peg pene- Several cadaver and clinical investi- of the eight specimens, placement of tration. Ultimately, results from this gations have demonstrated the effec- a glenoid prosthesis was not possible study demonstrate that advancing tiveness and safety of computer navi- after correction of 15° retroversion degrees of retroversion beyond 12°, gation in improving measurement of secondary to inadequate bone stock and all of those over 18°, may not be version intraoperatively.23-25 Kircher and peg penetration. A 20° defor- amenable to correction by eccentric et al24 reported improved version mity was not correctable in six speci- reaming. correction using computer naviga- mens. These authors concluded that Clinical results of TSA following tion in a prospective, randomized eccentric reaming to correct poste- asymmetric reaming are limited. clinical trial. Version correction from rior wear >15° has only a 50% Habermeyer et al21 prospectively 15.4° to 3.7° was achieved with nav- chance of successful correction by ec- studied a cohort of patients undergo- igation, whereas correction without centric reaming and that instead, ing TSA to address posterior glenoid navigation resulted in improvement bone grafting should be considered wear. In all cases, the humeral head from 14.4° to only 10.9°. No to address this defect. was maintained in a recentered posi- follow-up outcome data were ob-

608 Journal of the American Academy of Orthopaedic Surgeons Benjamin W. Sears, MD, et al

Figure 5 vantages include graft dissolution, nonunion, fixation failure, and shift leading to settling. Additionally, bone grafting is technically difficult and not reproducible as a result of high variability with graft contour- ing. Multiple grafting techniques have been described, with varying results. Three retrospective reviews in the literature have reported the results of bone grafting for eccentric glenoid wear, with incidence rates ranging from 3.3% to 16% of primary shoulder arthroplasties.26-28 Neer and Morrison26 reported excellent results in 16 patients and satisfactory results (or attainment of desired limited goals) in 3 patients, and no revision surgeries. No glenoid loosening or migration had occurred at a minimum follow-up of 2 years (average, 4.4 years). Steinmann and Cofield27 reported less favorable results in their series of 28 patients who underwent bone grafting for eccentric wear; however, this study included Illustrations of eccentric reaming to create a neutral articulation between the patients with instability. Fifteen pa- bone and glenoid prosthesis. A reamer in the neutral position is used to ream tients (54%) demonstrated some de- down the high side (A) to create a perpendicular articulating surface that can accept the glenoid implant (B). Although version can be corrected, the gree of radiographic lucency, and glenoid is medialized, and the width of the remaining glenoid bone becomes three glenoids were radiographically narrower following the removal of anterior bone. loose at an average follow-up of 5.3 years (range, 2 to 11 years). Despite these findings, patient satisfaction tained; however, the use of naviga- bone grafting include uneven wear was similar to that reported by Neer tion appears to improve glenoid po- that cannot be accommodated by 26 and Morrison. In the most recent sitioning and provides immediate, small changes in glenoid or humeral series, Hill and Norris28 reported on objective measurements of glenoid component version, insufficient bone 17 shoulders that required bone version. volume to support the glenoid com- grafting secondary to anterior or Eccentric reaming to correct ver- ponent, >15° of retroversion, and posterior instability. In this challeng- sion is a common and useful surgical potential penetration of the glenoid ing patient population, there were option. However, there are limita- 26-28 vault after version corrections. five glenoid failures requiring revi- tions to the degree of version that Advantages of bone grafting in the sion as a result of instability (two pa- may be appropriately corrected. setting of posterior bone deficiencies tients), one recurrent rotator cuff include preservation of anterior gle- tear, one improper placement of gle- Posterior Bone Graft noid bone stock and maintenance of noid component, and one glenoid a more normal joint line that avoids failure. These complications oc- Bone grafting provides a biologic so- altered joint kinematics secondary to curred within 2 to 91 months of the lution in cases of substantial bone shortening of the glenoid vault. Ad- initial surgery. Nine of 17 shoulders loss that prevents adequate version ditionally, theoretically, bone graft- were stable after bone grafting correction or secure seating of a gle- ing provides a permanent restorative (53%), with a similar distribution noid component. Indications for solution to glenoid bone loss. Disad- for functional satisfaction.

September 2012, Vol 20, No 9 609 Glenoid Bone Loss in Primary Total Shoulder Arthroplasty: Evaluation and Management

31 Figure 6 cacy. Neer et al initially presented an augmented glenoid component consisting of a sloped posterior buildup. However, description of this device was limited, and we are un- aware of any results on its use. Rice et al30 reported on their results using a Cofield 2 keeled all-polyethylene augmented glenoid (Smith & Nephew, Memphis, TN), which was implanted in 14 shoulders with OA and associated posterior glenoid deficiency. This augmented component provided version correction of approximately 4°. Despite satisfactory midterm pain relief and functional outcome, use of this component was A, Preoperative axillary radiograph of a type B2 glenoid demonstrating discontinued as a result of persistent approximately 7 mm of posterior bone erosion. B, Postoperative axillary instability. radiograph following implantation of the Global Steptech APG augmented Other augmented glenoid compo- glenoid component (DePuy, Warsaw, IN) and 7 mm of posterior buildup. nents have been recently introduced, such as the Global Steptech APG At our institution, bone loss >10 to temporarily fixed with small Stein- (DePuy, Warsaw, IN) and the Poste- 15 mm as measured on axial CT mann pins, which are replaced with rior Augment Glenoid (Exactech, scans is an indication for bone graft- countersunk 3.5-mm screws posi- Gainesville, FL). These offer an alter- ing. Ideally, the native osteotomized tioned to avoid glenoid implant pegs. native to asymmetric reaming and humeral head can be used as a struc- The graft contour is then further re- bone grafting (Figure 6). The long- tural graft. There are no current pub- fined with a burr, and the holes are term clinical performance of these lished investigations regarding this drilled for glenoid implantation. augmented components has not been technique; however, we have found Trial glenoid implantation is essen- evaluated. utilization of the native humeral tial to ensure secure seating and fit of head to be very useful as it is typi- the glenoid prosthesis to the bone. cally structurally sound and essen- Dysplastic Glenoid tially precontured to match the glenoid defect. The center of the glenoid Augmented Glenoid Walch et al9 defined type C glenoid can be identified with the help of Component morphology as glenoid retroversion specialized cannulated sizer disks. of >25° regardless of the degree of Using these disks, a centralized New prosthetic solutions to glenoid erosion. This is most commonly as- guide-wire is drilled perpendicular to bone loss have been driven by limita- sociated with congenital or dysplas- the neutral glenoid axis, which al- tions in other options, such as dem- tic development. Normal glenoid lows for version correction through onstrated lack of durability and early fossa development occurs from prox- limited reaming of the high side of glenoid failure in patients with imal and distal ossification centers.32 the glenoid. The deficient side is then eccentric wear.29,30 Although aug- Incomplete ossification of one or lightly prepared with a burr to ex- mented components with a fixed both of these centers may cause un- pose a bleeding surface to receive the polyethylene buildup theoretically derdevelopment of the posterior and bone graft. The graft is contoured to may ensure more reliable restoration inferior cartilaginous glenoid and fit the defect and placed with the ar- of the joint line, the mechanical neck of the scapula.33 Although gle- ticular side (ie, subchondral bone) properties of polyethylene in this set- noid dysplasia is thought to be rare, facing the glenoid. The curvatures ting are unknown. reports indicate that it may be more typically closely match because the Although the concept of glenoid common than previously specu- native humeral head is accountable component augmentation is not new, lated.9,34 Edelson34 examined 1,150 for creating the defect. The graft is there is a paucity of data on its effi- scapular bone specimens from sev-

610 Journal of the American Academy of Orthopaedic Surgeons Benjamin W. Sears, MD, et al eral museum collections representing Figure 7 several different ethnic populations. He found localized hypoplasia of the posteroinferior glenoid in 20% to 35% of specimens. Even so, a direct link between glenoid dysplasia and secondary development of glenohumeral OA has not been established. Edwards et al35 reported that only 3.5% of patients with primary OA undergoing TSA were found to have glenoid dysplasia. Although excessive glenoid retroversion is characteristic of the dysplastic shoulder, the humerus and surrounding soft tissues appear to adapt to the glenoid morphology, Axial CT scan (A) and sagittal three-dimensional CT reconstruction (B)ofa thereby allowing the humeral head dysplastic glenoid. Although this glenoid is in >35° of retroversion, it remains contained within the anterior and posterior conﬁnes of the glenoid surface. to remain centered in the retroverted This ﬁnding distinguishes the morphology of the dysplastic glenoid from the socket. That is, even though the cen- type B2 glenoid, in which the humeral head lies subluxated posteriorly and ter of the humeral head lies posterior the glenoid has a biconcave appearance. to the scapular plane, it is contained within the anterior and posterior confines of the glenoid surface (Fig- provement in outcome following ar- placement of reverse TSA, which is a ure 7). This important feature distin- throplasty, including increased pain more constrained device, may be guishes the hypoplastic glenoid from relief, motion, and function. In cases considered to stabilize the glenohu- the biconcave (type B2) glenoid, of severe dysplasia, the amount of meral joint. which is associated with extensive bone available for fixation may be In cases in which bone erosion or glenoid bone erosion and posterior inadequate for standard fixation op- version is too significant to allow humeral head subluxation with re- tions. In these patients, the use of an placement of a glenoid component, spect to the glenoid surface. Poste- inset bone-sparing glenoid compo- replacement of the humeral head rior glenoid erosion was minimal in nent with a single, short peg may be without glenoid resurfacing has been one series of 15 patients with OA helpful to avoid cortical penetra- shown to result in acceptable out- and dysplasia.35 Additionally, these tion.29 Complete correction of ver- comes, even in young patients.37 The patients had developed posterior sion in dysplastic glenoids can result criteria used by Edwards et al35 for soft-tissue hyperplasia that seemingly in substantial loss of internal rota- glenoid resurfacing was ≥15 mm of prevented posterior subluxation. tion as a result of excessive tension- glenoid bone depth on axial CT. During glenoid preparation, the in- ing of posterior tissues. Unlike the These investigators found that pa- vestigators reamed to the native gle- type B2 glenoid, the dysplastic gle- tients who underwent hemiarthro- noid (ie, no eccentric reaming) and noid is associated with minimal pos- plasty still had significant improve- made no attempt to correct version terior subluxation of the humerus ments in pain scores and functional or augment posterior soft tissue. At and, therefore, the posterior tissues measures. A recent investigation by our institution, the posterior capsule are not excessively lax. Thus, aggres- Bonnevialle et al37 reported marked is also preserved in such cases as a sive version correction can place ten- improvement in pain scores, func- checkrein to posterior displacement sion on the posterior soft tissues and tion, and outcome measures at a of the humeral head. limit the ability of the humerus to ro- minimum 2-year follow-up in nine Although there are limited reports tate internally. However, partial cor- patients treated with hemiarthro- on surgical treatment in symptomatic rection may be attempted to restore plasty for dysplasia. The authors patients with dysplastic glenoids, native retroversion if the humeral concluded that hemiarthroplasty is a shoulder arthroplasty has been head has eroded into the posterior reliable management option for this shown to be a viable option.35,36 Ed- glenoid. In sedentary persons and el- patient population. Glenohumeral wards et al35 reported significant im- derly persons (aged >70 years), arthritis secondary to glenoid dyspla-

September 2012, Vol 20, No 9 611 Glenoid Bone Loss in Primary Total Shoulder Arthroplasty: Evaluation and Management sia appears to be relatively uncom- 29, and 34 are level III studies. 12. Nyffeler RW, Jost B, Pfirrmann CW, mon, but it may be effectively man- References 3, 22, 26, 27, 30, 32, 33, Gerber C: Measurement of glenoid version: Conventional radiographs aged with TSA or hemiarthroplasty and 35-37 are level IV studies. versus computed tomography scans. with limited correction of the natural Reference 31 is level V expert J Shoulder Elbow Surg 2003;12(5):493- 496. version of the glenoid. opinion. 13. Friedman RJ, Hawthorne KB, Genez References printed in bold type indicate BM: The use of computerized those published within the past 5 years. tomography in the measurement of Summary glenoid version. J Bone Joint Surg Am 1. Day JS, Lau E, Ong KL, Williams GR, 1992;74(7):1032-1037. Ramsey ML, Kurtz SM: Prevalence and 14. Randelli M, Gambrioli PL: Advanced glenohumeral OA is asso- projections of total shoulder and elbow Glenohumeral osteometry by computed ciated with increased glenoid retro- arthroplasty in the United States to 2015. J Shoulder Elbow Surg 2010; tomography in normal and unstable version, posterior erosion, and gle- 19(8):1115-1120. shoulders. Clin Orthop Relat Res 1986; noid bone loss. Surgical resurfacing (208):151-156. 2. Norris TR, Iannotti JP: Functional of the glenoid requires correction of outcome after shoulder arthroplasty for 15. Rouleau DM, Kidder JF, Pons- these anatomic alterations to allow primary osteoarthritis: A multicenter Villanueva J, Dynamidis S, Defranco M, study. J Shoulder Elbow Surg 2002; Walch G: Glenoid version: How to placement of the implant in near- 11(2):130-135. measure it? Validity of different methods neutral version with adequate bony in two-dimensional computed 3. Cil A, Veillette CJ, Sanchez-Sotelo J, tomography scans. J Shoulder Elbow structural support. This is done in an Sperling JW, Schleck CD, Cofield RH: Surg 2010;19(8):1230-1237. effort to avoid early failure of the re- Survivorship of the humeral component 16. Bokor DJ, O’Sullivan MD, Hazan GJ: placement. Selection of surgical tech- in shoulder arthroplasty. J Shoulder Elbow Surg 2010;19(1):143-150. Variability of measurement of glenoid nique depends on the severity of de- version on computed tomography scan. 4. Bohsali KI, Wirth MA, Rockwood CA J Shoulder Elbow Surg 1999;8(6):595- formity and the degree of bone loss. Jr: Complications of total shoulder 598. Typically, minor changes in version arthroplasty. J Bone Joint Surg Am 2006; 88(10):2279-2292. 17. Lazarus MD, Jensen KL, Southworth C, and limited bone loss can be ade- Matsen FA III: The radiographic quately corrected with eccentric 5. Moskal MJ, Duckworth D, Matsen FA: evaluation of keeled and pegged glenoid Abstract: An analysis of 122 failed component insertion. J Bone Joint Surg reaming of the glenoid to neutral or shoulder arthroplasties. J Shoulder Am 2002;84(7):1174-1182. near-neutral version. However, ex- Elbow Surg 1999;8:554. 18. Clavert P, Millett PJ, Warner JJ: Glenoid tensive glenoid bone loss may require 6. Shapiro TA, McGarry MH, Gupta R, resurfacing: What are the limits to augmentation with either bone graft Lee YS, Lee TQ: Biomechanical effects of asymmetric reaming for posterior glenoid retroversion in total shoulder erosion? J Shoulder Elbow Surg 2007; or specialized glenoid implants to re- arthroplasty. J Shoulder Elbow Surg 16(6):843-848. store version and support glenoid 2007;16(3 suppl):S90-S95. 19. Gillespie R, Lyons R, Lazarus M: component implantation. Extensive 7. Farron A, Terrier A, Büchler P: Risks of Eccentric reaming in total shoulder bone loss may make glenoid resur- loosening of a prosthetic glenoid arthroplasty: A cadaveric study. implanted in retroversion. J Shoulder Orthopedics 2009;32(1):21. facing impossible. Elbow Surg 2006;15(4):521-526. Glenoid bone loss is a challenging 20. Nowak DD, Bahu MJ, Gardner TR, 8. Iannotti JP, Norris TR: Influence of et al: Simulation of surgical glenoid problem that is commonly encoun- preoperative factors on outcome of resurfacing using three-dimensional tered with the osteoarthritic glenoid. shoulder arthroplasty for glenohumeral computed tomography of the arthritic osteoarthritis. J Bone Joint Surg Am glenohumeral joint: The amount of Appropriate assessment and correc- 2003;85(2):251-258. glenoid retroversion that can be tion are critical for long-term out- corrected. J Shoulder Elbow Surg 2009; 9. Walch G, Badet R, Boulahia A, Khoury 18(5):680-688. comes following TSA. Further re- A: Morphologic study of the glenoid in search is necessary to determine the primary glenohumeral osteoarthritis. 21. Habermeyer P, Magosch P, Lichtenberg J Arthroplasty 1999;14(6):756-760. S: Recentering the humeral head for most effective approach to glenoid glenoid deficiency in total shoulder management in these difficult cases. 10. Scalise JJ, Bryan J, Polster J, Brems JJ, arthroplasty. Clin Orthop Relat Res Iannotti JP: Quantitative analysis of 2007;457:124-132. glenoid bone loss in osteoarthritis using three-dimensional computed tomography 22. Gerber C, Costouros JG, Sukthankar A, References scans. J Shoulder Elbow Surg 2008; Fucentese SF: Static posterior humeral 17(2):328-335. head subluxation and total shoulder arthroplasty. J Shoulder Elbow Surg Evidence-based Medicine: Levels of 11. Scalise JJ, Codsi MJ, Bryan J, Brems JJ, 2009;18(4):505-510. Iannotti JP: The influence of three- evidence are listed in the table of dimensional computed tomography 23. Edwards TB, Gartsman GM, O’Connor contents. In this article, references 2, images of the shoulder in preoperative DP, Sarin VK: Safety and utility of planning for total shoulder arthroplasty. computer-aided shoulder arthroplasty. 8, 9, 17, and 24 are level II studies. J Bone Joint Surg Am 2008;90(11):2438- J Shoulder Elbow Surg 2008;17(3):503- References 1, 4, 5, 12, 13, 21, 23, 28, 2445. 508.

612 Journal of the American Academy of Orthopaedic Surgeons Benjamin W. Sears, MD, et al

24. Kircher J, Wiedemann M, Magosch P, bone-grafting of the glenoid. J Bone Radiol 1998;28(1):30-37. Lichtenberg S, Habermeyer P: Improved Joint Surg Am 2001;83(6):877-883. accuracy of glenoid positioning in total 34. Edelson JG: Localized glenoid shoulder arthroplasty with intraoperative 29. Gunther SB, Lynch TL: Total shoulder hypoplasia: An anatomic variation of navigation: A prospective-randomized replacement surgery with custom glenoid possible clinical significance. Clin clinical study. J Shoulder Elbow Surg implants for severe bone deficiency. Orthop Relat Res 1995;(321):189-195. 2009;18(4):515-520. J Shoulder Elbow Surg 2012;21(5):675- 35. Edwards TB, Boulahia A, Kempf JF, 684. 25. Nguyen D, Ferreira LM, Brownhill JR, Boileau P, Némoz C, Walch G: Shoulder et al: Improved accuracy of computer 30. Rice RS, Sperling JW, Miletti J, Schleck arthroplasty in patients with assisted glenoid implantation in total C, Cofield RH: Augmented glenoid osteoarthritis and dysplastic glenoid shoulder arthroplasty: An in-vitro component for bone deficiency in morphology. J Shoulder Elbow Surg randomized controlled trial. J Shoulder shoulder arthroplasty. Clin Orthop Relat 2004;13(1):1-4. Elbow Surg 2009;18(6):907-914. Res 2008;466(3):579-583. 36. Sperling JW, Cofield RH, Steinmann SP: 26. Neer CS II, Morrison DS: Glenoid bone- 31. Neer CS II, Watson KC, Stanton FJ: Shoulder arthroplasty for osteoarthritis grafting in total shoulder arthroplasty. Recent experience in total shoulder secondary to glenoid dysplasia. J Bone J Bone Joint Surg Am 1988;70(8):1154- replacement. J Bone Joint Surg Am Joint Surg Am 2002;84(4):541-546. 1162. 1982;64(3):319-337. 37. Bonnevialle N, Mansat P, Mansat M, 27. Steinmann SP, Cofield RH: Bone grafting 32. Wirth MA, Lyons FR, Rockwood CA Jr: Bonnevialle P: Hemiarthroplasty for for glenoid deficiency in total shoulder Hypoplasia of the glenoid: A review of osteoarthritis in shoulder with dysplastic replacement. J Shoulder Elbow Surg sixteen patients. J Bone Joint Surg Am morphology. J Shoulder Elbow Surg 2000;9(5):361-367. 1993;75(8):1175-1184. 2011;20(3):378-384. 28. Hill JM, Norris TR: Long-term results of 33. Currarino G, Sheffield E, Twickler D: total shoulder arthroplasty following Congenital glenoid dysplasia. Pediatr

September 2012, Vol 20, No 9 613