Journal Logo

Scientific Articles

Prospective, Blinded, Randomized Controlled Trial of Stemless Versus Stemmed Humeral Components in Anatomic Total Shoulder Arthroplasty

Results at Short-Term Follow-up

Wiater, J. Michael MD1,a; Levy, Jonathan C. MD2; Wright, Stephen A. MD3; Brockmeier, Stephen F. MD4; Duquin, Thomas R. MD5; Wright, Jonathan O. MD1; Codd, Timothy P. MD6

Author Information
The Journal of Bone and Joint Surgery: November 18, 2020 - Volume 102 - Issue 22 - p 1974-1984
doi: 10.2106/JBJS.19.01478
  • Open
  • Supplementary Content
  • Disclosures
  • Data Availability
  • Infographic
  • Video Summary


As designs for anatomic total shoulder arthroplasty (aTSA) implants have evolved, the need for long humeral stems has been questioned as complications involving the humeral component are often related to the stem1,2. The idea of stemless implants was introduced in the 1980s by Levy and Copeland with the development of humeral resurfacing implants3,4. With those implants, the entire humeral head was not resected, resulting in the preservation of humeral bone stock at the cost of making glenoid exposure more challenging. Following their success, many “stemless” implants have been designed with canal-sparing stems that only engage the humeral metaphysis but still allow for standard humeral anatomic neck cuts, facilitating exposure of the glenoid.

Stemless humeral implants have several potential benefits over stemmed implants, including simplified recreation of native anatomy, particularly in cases of metadiaphyseal deformity5-7; decreased risk of intraoperative periprosthetic humeral shaft fractures resulting from stem impaction1,2; decreased operative time and blood loss8,9; and, importantly, preservation of humeral bone stock, with less bone removal at the time of the initial procedure, less bone loss over time (due to decreased stress-shielding), and easier explantation if revision is needed10-12.

Short- and intermediate-term studies of stemless implants used in Europe since 2004 have shown good success with minimal complications5-7,9,11,13-23. However, to our knowledge, there have been no large, multicenter, randomized controlled trials (RCTs) comparing stemless and stemmed implants. Therefore, we performed such a trial with the objective of evaluating both the safety and effectiveness of a stemless implant compared with a stemmed implant, hypothesizing that the stemless implant would be noninferior to the stemmed implant at the time of the 2-year follow-up.

Materials and Methods

Trial Design

The present study was a prospective, multicenter, single-blinded RCT in which patients were allocated evenly into treatment cohorts at 12 sites across the United States. Sites obtained institutional review board approval prior to study commencement, and no substantial methodological changes were made after study commencement. Three sites enrolled ≥25 shoulders, 7 enrolled 10 to 24 shoulders, and 2 enrolled <10 shoulders.


Each site screened consecutive patients undergoing aTSA for eligibility. The inclusion and exclusion criteria are shown in Table I. Patients with osteoporosis, osteomalacia, or other disorders possibly compromising fixation of a stemless implant in the metaphysis were excluded.

TABLE I - Complete Inclusion and Exclusion Criteria
Inclusion Criteria Exclusion Criteria
  • Patients with non-inflammatory degenerative joint disease, including osteoarthritis.

  • Patients in whom the device will be used for the correction of a functional deformity, specifically, deformities that prevent congruent articulation of the glenohumeral joint. Examples include but are not limited to humeral head structural deformity, osteophyte formation restricting range of motion, and so on.

  • Patients with pain and/or loss of function in the shoulder for whom other treatment modalities have been unsuccessful. Examples of such treatment include but are not limited to activity modification, physiotherapy, and anti-inflammatory or other types of medication.

  • Patients requiring unilateral or staged bilateral shoulder arthroplasty.

  • Patients who are anatomically and structurally suited to receive the implants. During the preoperative templating, it must be confirmed that the humeral neck is of sufficient diameter to implant at least the smallest stemless humeral component and that the humeral neck cortex is intact.

  • Patients who are 21-90 years of age at the time of surgery.

  • Patients who are skeletally mature.

  • Patients with an ASES score of ≤40.

  • Patients who are willing and able to return for scheduled follow-up evaluations.

  • Patients who have completed a valid, institutional review board-approved informed-consent form.

  • Patients for whom the surgeon has confirmed intraoperatively that there is no cyst measuring >1 cm and that there is not >1 cyst at the implantation site

  • Patients who agree to be blinded to treatment until evaluations are completed at the ≥22-month end point.

  • Patients diagnosed with osteonecrosis or posttraumatic arthritis of the humeral head.

  • Patients presenting with shoulder joint infection, sepsis, osteomyelitis, or distant foci of infections that may spread to the implant site.

  • Patients with cuff tear arthropathy.

  • Patients who have previously undergone a hemiarthroplasty, total shoulder arthroplasty, or reverse total shoulder arthroplasty in the affected shoulder.

  • Patients presenting with malunion or nonunion of the tuberosities of the proximal part of the humerus.

  • Patients with osteoporosis; osteomalacia; rheumatoid arthritis; metabolic disorders of bone, muscle, or connective tissue; gross deformity or any other condition of the proximal part of the humerus (defined as severe destruction or deformity of the proximal part of the humerus that precludes placement of the device) that in the investigator’s medical judgment could compromise implant fixation or bone-healing.

  • Patients with rapid bone destruction, marked bone loss, or bone resorption apparent on radiographs.

  • Patients with neurological or other disorders that would affect the stability of the shoulder prosthesis (e.g., Charcot joint, uncontrolled seizures, etc.).

  • Patients diagnosed with any condition (e.g., mental illness) that may limit the ability to complete the consent form or that would affect the capability or willingness to return to the clinic for assessments and/or follow directions.

  • Patients with bone cancer, either primary or secondary, affecting the shoulder.

  • Patients presenting with symptoms of chronic steroid use, defined as use of oral steroids for a chronic condition for 12 months prior to and including the date of surgery (inhaled and topical steroid use is allowed).

  • Patients with a life expectancy of <3 years.

  • Patients with severe shoulder instability.

  • Patients with subscapularis incompetence.

  • Patients with active medicolegal activity regarding the index shoulder.

  • Patients known to be pregnant or planning to get pregnant, prisoners, and/or alcohol or drug abusers.

  • Patients known to be involved in Workers’ Compensation litigation regarding index shoulder.

  • Patients with known metal allergy.

  • Patients refusing to sign the institutional review board-approved consent form.

  • Patients who are found intraoperatively to require a specific treatment and are unable to be randomized.

  • Patients who are found at the time of intraoperative examination to have a single cyst measuring >1 cm in size or multiple cysts at the implantation site.


The stemmed implants were Comprehensive Mini (83-mm) humeral stems (Fig. 1-A), which previously have demonstrated good results at short-term follow-up24-26, and the stemless implants were Comprehensive Nano humeral stems (Figs. 1-B and 1-C). Comprehensive Versa-Dial humeral heads and Comprehensive glenoid components with Regenerex posts were used in all patients (Figs. 1-B and 1-C). All implants were manufactured by Zimmer Biomet.

Figs. 1-A, 1-B, and 1-C Images of the implants used in the present study.

Fig. 1-A:
Photograph showing the Zimmer Biomet Comprehensive Mini humeral stem. (Reproduced with permission from Zimmer Biomet.)
Fig. 1-B:
Photographs showing the Zimmer Biomet Comprehensive Nano humeral stem and Versa-Dial humeral head (left) and the Comprehensive glenoid component with Regenerex post (right). (Reproduced with permission from Zimmer Biomet.)
Fig. 1-C:
Representative anteroposterior postoperative radiograph of the stemless implant, made 28 months postoperatively.

Surgical procedures were performed according to the manufacturer’s surgical techniques by surgeons with extensive experience in shoulder arthroplasty. Briefly, a deltopectoral approach was used, with subscapularis tenotomy or lesser tuberosity osteotomy and subsequent repair according to each surgeon’s preference. The humeral neck was cut in either 30° of retroversion or the patient’s native retroversion according to surgeon preference. In the stemmed cohort, the humeral canal was reamed and broached in standard fashion. In the stemless cohort, a guide pin was placed in the center of the humeral neck cut followed by a central surface reamer and broach. In both cohorts, the broach was left in place and a humeral cut protector was inserted. The glenoid was then exposed and prepared. Following trialing, the final glenoid and humeral components were implanted. A hybrid glenoid component was used in all cases, with peripheral polymethylmethacrylate-cemented pegs and a central bone-ingrowth, uncemented post. The humeral stems were uncemented in both cohorts. The subscapularis was repaired, the biceps tendon was tenodesed, and the incision was closed in a standard fashion.

Subjects were required to participate in a postoperative rehabilitation protocol as prescribed by their surgeon. A recommended rehabilitation protocol from Brigham and Women’s Hospital was provided to all sites27.


Data were collected preoperatively, intraoperatively, and at 6 weeks, 3 months, 1 year, and 2 years postoperatively. Intraoperative data, including length of surgery, glenoid wear pattern, subscapularis management, and rotator cuff status, were recorded. The primary outcome measures used for noninferiority analysis were the American Shoulder and Elbow Surgeons (ASES) score at the 2-year follow-up time point, device-related complications (unanticipated adverse device events [UADEs], fracture, dislocation/subluxation, glenoid perforation, dissociation of the device, or revision/removal of any component), and radiographic signs of loosening. Secondary outcome measures were ASES scores at all other time points and Single Assessment Numeric Evaluation (SANE) scores, Constant scores, and range-of-motion measurements at all time points (except for Constant scores at 6 weeks because of ongoing healing and SANE scores preoperatively). Patients were given paper forms at each visit to complete questionnaires that were used to calculate outcome scores. Trained research staff measured range of motion with a large goniometer and measured abduction strength at 90° (for the Constant score) with a standard force gauge provided to each site. On rare occasions, if research staff were unavailable, data were collected by the surgeon. Constant scores were normalized by age and sex as described by Katolik et al.28. Intraoperative complications were recorded, and patients were assessed at each follow-up for any additional shoulder-related complications.

Radiographic Analysis

Radiographs, including glenohumeral true anteroposterior internal and external rotation views and axillary lateral views, were made at each follow-up. Radiographic technique was standardized across all sites by providing a radiographic evaluation protocol with specific instructions for making images, including equipment set-up and subject positioning, and radiographs were repeated with technical adjustments as needed to ensure similar prosthesis profiles on all radiographs. Radiographic evaluation was performed by an independent, board-certified radiographic reviewer (Medical Metrics). Radiographs were assessed for fractures, dislocation/subluxation, component complications, and signs of component loosening. Loosening was defined as ≥5 mm of subsidence of the humeral component, ≥5 mm migration of the humeral and/or glenoid component, >2 mm of progressive radiolucency around the humeral component in ≥2 contiguous zones, or >2 mm of progressive radiolucency around the glenoid component in all zones.

Sample Size

Noninferiority sample size calculations, performed with use of nQuery Advisor 7.0 software (Statsols) with a 1:1 ratio of investigational to control subjects to obtain 90% power while accounting for attrition, resulted in a sample size of 264 subjects.

Randomization and Blinding

The randomization plan was produced by the sponsor with use of SAS 9.2 (SAS Institute) with balanced, blocked randomization (with a 1:1 ratio of investigational to control subjects) on a per-shoulder basis. Randomization was blocked by site, and each site received separate randomization plans with use of a predetermined block size undisclosed to the sites.

All efforts were made to keep patients blinded to their treatment. Treatment assignment was revealed to the surgeon after surgery had commenced, while the patient was under anesthesia. The patients were not permitted to view the postoperative radiographs until all patients in the study had reached 2 years of follow-up (with the exception of 6 patients who underwent revision).

Statistical Methods

Primary end points were compared between cohorts for noninferiority with use of both a t test and the lower bound of a 1-sided 95% confidence interval (CI) for the difference between groups (investigational minus control) at ≥22 months. For the ASES score, a 9.5-point margin of noninferiority was used as studies have reported that the minimum clinically important difference (MCID) of the ASES score ranges from 9 to 21 points29,30 and because, for many patient-reported outcome measures, the MCID is approximately half the standard deviation of change31,32, which has been reported to be 19 points for ASES scores33. For radiographic signs of loosening and device-related complications, a 10% margin of noninferiority was used.

Secondary outcomes and demographic data were compared for differences with use of standard statistical tests appropriate for the data: the t test for range-of-motion measurements, the Wilcoxon rank-sum test for outcome measures, and the Fisher exact test for demographic and operative data.

Regulatory Oversight, Registration, and Safety

This study was registered with (registration number NCT01936259). The details of the study are not published on the website because the investigational device exemption (IDE) study had not yet been closed by the U.S. Food & Drug Administration (FDA), although the implant has since been cleared for use by the FDA (510[k] number K182516). The present study was carried out and documented in accordance with U.S. federal regulations. Enrolled patients signed informed consent prior to data collection. Sites reported adverse events and protocol deviations to their institutional review board as necessary.


Recruitment began in August 2013 and concluded in October 2016. Three hundred and twenty-six patients were assessed for eligibility, and 56 were excluded. Two hundred and seventy patients agreed to participate and were randomized to treatment with either a stemless (n = 133) or stemmed (n = 137) humeral component. One hundred and sixteen patients in the stemless cohort and 123 in the stemmed cohort reached the 2-year follow-up and were included in the analysis (representing follow-up rates of 87% and 90%, respectively) (Fig. 2). The trial was ended November 2018 when all enrolled subjects reached ≥22 months of follow-up.

Fig. 2:
Consolidated Standards of Reporting Trials (CONSORT) flow diagram outlining patient recruitment, allocation, and follow-up.

There were no differences between the cohorts in terms of demographic or operative data (Table II). Of the 265 shoulders that received the allocated intervention, 176 (66%) were in male patients and 89 (34%) were in female patients. The mean age (and standard deviation) was 62.6 ± 9.3 years (62.1 ± 9.6 years in the stemmed cohort and 63.1 ± 9.0 years in the stemless cohort). The primary diagnosis was osteoarthritis for 99% of the shoulders. A lesser tuberosity osteotomy was performed in 33% of the shoulders in the stemless cohort and 31% of those in the stemmed cohort (p = 0.79), and a tenotomy was used in the remainder. The average operative time was 97.1 ± 24.9 minutes in the stemless cohort and 97.2 ± 24.9 minutes in the stemmed cohort (p = 0.99). Glenoid morphology did not differ significantly between the cohorts (p = 0.14). Blood loss data were not collected.

TABLE II - Demographic Characteristics and Operative Data
Variable Control (Stemmed) (N = 133) Investigational (Stemless) (N = 132) P Value
Demographic characteristics
 Female:male ratio (no. of shoulders) 46:87 43:89 0.80
 Age at surgery*(yr) 62.1 ± 9.6 (28-81) 63.1 ± 9.0 (42-85) 0.42
 Weight*(kg) 90.3 ± 17.4 (52-131) 91.9 ± 20.0 (48-154) 0.49
 Height*(cm) 173.2 ± 10.4 (140-196) 173.2 ± 10.9 (142-196) 0.92
 Body mass index* (kg/m 2 ) 30.1 ± 5.3 (19.5-44.6) 30.6 ± 5.8 (18.6-47.8) 0.46
 Race (no. of shoulders) 0.81
  Caucasian 125 126
  African American 6 4
  Asian 0 1
  Hispanic or Latino 2 1
 Prior contralateral shoulder arthroplasty (no. of shoulders) 19 15 0.58
Operative data
 Unilateral:bilateral (no. of shoulders) 124:9 127:5 0.41
 Total operative time*(min) 97.2 ± 24.9 (44-180) 97.1 ± 24.9 (42-163) 0.99
 State of rotator cuff (no. of shoulders) 0.64
  Degenerated/partial tear 7 (5.26%) 9 (6.82%)
  Full-thickness tear 1-3 cm 1 (0.75%) 2 (1.52%)
  Normal 125 (93.98%) 121 (91.67%)
 Subscapularis management (no. of shoulders) 0.79
  Lesser tuberosity osteotomy 41 (30.83%) 43 (32.58%)
  Subscapularis tenotomy 92 (69.17%) 89 (67.42%)
 Glenoid morphology (no. of shoulders) 0.14
  Anterior erosion 5 (3.8%) 6 (4.5%)
  Normal alignment 44 (33.1%) 51 (38.6%)
  Posterior erosion 84 (63.2%) 75 (56.8%)
*The values are given as the mean and the standard deviation, with the range in parentheses.

Primary Outcomes

The mean ASES score at 2 years was 92.5 ± 14.9 (median, 98.3; interquartile range [IQR], 92.5-100.0; n = 112) for the stemless cohort, compared with 92.2 ± 13.5 (median, 98.3; IQR, 91.7-100.0; n = 121) for the stemmed cohort; evaluation of the difference of 0.27 points and the lower bound of the 95% CI for the difference of −2.81 points with a 9.5-point margin of noninferiority showed statistical noninferiority of the stemless implant (p value for noninferiority test, <0.0001) (Fig. 3-A).

Figs. 3-A, 3-B, and 3-C Box-and-whisker plots showing secondary outcome measures for both cohorts over time. There were no significant differences between cohorts at any time point. Median values are indicated with horizontal lines, IQRs are indicated with boxes, and maximum and minimum values are indicated with I-bars.

Fig. 3-A:
ASES scores.
Fig. 3-B:
SANE scores.
Fig. 3-C:
Sex and age-adjusted Constant scores.

With regard to device-related complications, 18 procedures (9 in the stemmed cohort and 9 in the stemless cohort) were identified as failures (Table III). In the stemless cohort, the reasons for failure included revision for infection (2 shoulders), intraoperative glenoid perforation (1 shoulder), and postoperative radiographic findings identified by the independent reviewer (including 1 humeral fracture, 4 glenoid fractures, and 1 case of anterior subluxation, all of which were treated nonoperatively with satisfactory outcomes). In the stemmed cohort, the reasons for failure included revision for infection (2 shoulders), revision to reverse TSA for postoperative subscapularis tears (2 shoulders), traumatic dislocation treated with closed reduction (1 shoulder), humeral fracture noted intraoperatively during stem impaction (1 shoulder), and radiographic findings (including 2 humeral fractures and 1 case of anterior subluxation, all of which were treated nonoperatively with satisfactory outcomes). No UADEs were reported in either cohort. The success rates at 2 years were 92% and 93% for the stemless and stemmed cohorts, respectively; statistical evaluation of the difference of −0.44% and the lower limit of the 95% CI of −6.74% with a 10% margin of noninferiority showed statistical noninferiority of the stemless implant (p value for noninferiority test, 0.0063).

TABLE III - Summary of Failures
Reason for Failure Count
Control cohort (stemmed)
 Revision for infection 2
 Revision for subscapularis tear 2
 Dislocation, traumatic 1
 Humeral fracture, intraoperative 1
 Humeral fracture, postoperative radiographic review 2
 Anterior subluxation, postoperative radiographic review 1
 Success rate 93% (114 of 123)*
Investigational cohort (stemless)
 Revision for infection 2
 Intraoperative glenoid perforation 1
 Humeral fracture, postoperative radiographic review 1
 Glenoid fracture, postoperative radiographic review 4
 Anterior subluxation, postoperative radiographic review 1
 Success rate 92% (107 of 116)*
*P = 0.0063 for noninferiority.

Last, no shoulder in either cohort had radiographic signs of loosening. No shoulder in either cohort had subsidence or migration of the humeral component. Two shoulders in the stemmed cohort showed humeral radiolucency of >2 mm, but the implants were not considered loose because the radiolucency was confined to only 1 radiographic zone; no shoulder in the stemless cohort showed humeral radiolucency of >2 mm. No shoulder in either cohort had evidence of subsidence or migration of the glenoid component. These findings demonstrated radiographic noninferiority of the stemless implant.

Secondary Outcomes

No significant differences in the ASES, SANE, or Constant scores or range-of-motion measurements were observed between the cohorts at any time point within the 2-year follow-up (p > 0.05) (Figs. 3-A through 4-B) (see Appendix).

Figs. 4-A and 4-B Line graphs showing the preoperative and postoperative mean range-of-motion measurements for both cohorts over time. There were no significant differences between cohorts at any time point in either measurement. Standard deviations are indicated with I-bars. (See Appendix for complete active and passive range-of-motion measurement data and comparisons.).

Fig. 4-A:
Active forward elevation.
Fig. 4-B:
Active external rotation with the arm at the side.

Secondary outcomes improved in both cohorts at the time of the 2-year follow-up. In the stemless and stemmed cohorts, mean active forward elevation increased from 107° ± 30° to 157° ± 21° and from 109° ± 29° to 157° ± 19°, respectively, and mean active external rotation with the arm at the side increased from 24° ± 21° to 62° ± 23° and from 29° ± 23° to 59° ± 15°, respectively. In the stemless cohort, the median normalized Constant score increased from 45.0% (IQR, 34.0-57.0%) to 104.5% (IQR, 94.0-111.0%) and the mean score increased from 46.3% ± 17.1% to 101.5% ± 16.7%. In the stemmed cohort, the median Constant score increased from 45.0% (IQR, 34.0-60.0%) to 102.0% (IQR, 92.0-111.0%) and the mean score increased from 47.2% ± 17.1% to 100.2% ± 15.0%. The median SANE score was 98.0% (IQR, 10.0%) for the stemless cohort and 95% (IQR, 10.0%) for the stemmed cohort at the time of the 2-year follow-up.


This single-blinded, multicenter RCT demonstrated that the studied stemless shoulder implant was noninferior to the studied stemmed shoulder implant at the time of short-term follow-up after aTSA for the treatment of primary osteoarthritis. All primary outcomes showed noninferiority of the stemless implant, and no secondary outcomes differed significantly between stemless and stemmed cohorts within the 2-year follow-up period. To our knowledge, the present study represents the first Level-I evidence supporting use of a stemless humeral component.

Our results compare favorably with and expand on the results of previous, smaller RCTs comparing stemless and stemmed humeral implants. Most recently, in 2017, Uschok et al. reported on a 40-patient (Level-II) RCT, which showed no significant difference between stemmed and stemless cohorts in terms of the mean adjusted Constant score and range of motion at 2 and 5 years of follow-up; the mean adjusted Constant score in the stemless cohort improved from 70.7% ± 18.8% preoperatively to 83.6% ± 19.3% and 94.9% ± 18.6% at 2 and 5 years, respectively (with no significant difference between the scores at 2 and 5 years)12. In 2013, Razmjou et al. reported on a 74-patient (Level-II) RCT comparing 3 implants (1 stemless and 2 stemmed), which showed no significant differences in outcome scores and range of motion between cohorts, except that 1 stemmed implant had decreased external rotation at 90° of abduction at a minimum of 2 years of follow-up; the mean relative Constant score in the stemless cohort was 92 ± 24 points at 2 years11. Last, in 2013, Berth and Pap reported on an 82-patient RCT, which showed no significant difference between stemmed and stemless cohorts in terms of mean Constant scores and range-of-motion measurements at a minimum of 2 years of follow-up; the mean adjusted Constant score in the stemless cohort was 73.2 ± 11.3 points at 2 years9.

The outcomes of the present study also compare favorably with those of previous large, prospective cohort studies of stemless prostheses. Churchill et al., in a study of a cohort of 157 patients who underwent stemless aTSA for the treatment of glenohumeral osteoarthritis, reported a mean adjusted Constant score of 104.1% ± 14.8% and a mean ASES score of 91.9 ± 11.4 at 2 years of follow-up15. Brunner et al., in a prospective multicenter study of 233 patients who underwent stemless aTSA, reported a mean relative Constant score of 78.9% (89% for patients with primary osteoarthritis) at an average of 23 months postoperatively22.

Two other findings of the current study warrant mentioning. First, while decreased operative time has been reported as a potential benefit of stemless TSA8,9, our data showed no significant difference between the 2 cohorts in terms of operative time. Possible explanations for this finding include (1) a learning curve effect in the current study and (2) the fact that the study by Berth and Pap compared cementless and cemented prostheses9, whereas both cohorts in the current study received cementless prostheses. Second, 4 glenoid fractures in the stemless cohort were noted at the time of postoperative radiographic evaluation and not during surgery. The reason for this finding is unclear as there were no differences between the stemmed and stemless procedures with regard to the humeral neck cut or the glenoid exposure and preparation. All were fractures treated conservatively with observation, and all 4 patients did well clinically.

The current study had several strengths, the most notable being study design. The study was planned and powered such that statistical noninferiority of the stemless implant could be concluded. The multicenter nature of the study was another strength. In addition, no patients could cross over between treatment arms, avoiding confounding of the results.

The present study also had limitations. First, follow-up was only 2 years. Longer-term radiographic and clinical follow-up may reveal differences that are not apparent at short-term follow-up. However, it is worth noting that studies of similar stemless implants have shown no deterioration of shoulder function at intermediate-term follow-up7,12,16,17,34. Second, surgeons could not be blinded to treatment. To address this limitation, primary end points were chosen such that surgeon bias should have played little role: ASES scores are determined on the basis of a patient questionnaire, device failure modes were independent of the surgeon, and an independent radiographic reviewer assessed the radiographs. Third, given the low rates of device-related complications and component loosening in the present study, the 2-year follow-up rates of 87% and 90% may have introduced error; if even only a few of the patients who were lost to follow-up had device-related complications or component loosening, it could have had a substantial impact on the results. Last, it should be noted that the scope of this study was intentionally limited in order to answer a specific question, that is, whether a stemless implant is noninferior to a stemmed implant in patients with adequate metaphyseal bone quality. A single stemless implant was compared with a single stemmed implant, and patients who were considered to be poor candidates for a stemless implant (i.e., those with osteoporosis, osteomalacia, or other disorders possibly compromising fixation of the implant in the metaphysis) were excluded. The findings of this trial cannot necessarily be applied to other stemless designs, nor should they be applied to patients with poor metaphyseal bone quality.

In conclusion, this trial provides Level-I evidence that the short-term safety and effectiveness of this stemless implant was noninferior to a stemmed implant for patients who met the criteria to receive a stemless implant. While further studies are warranted to assess the long-term success of this implant, these early results are promising given the potential benefits of stemless designs over stemmed humeral components.


Supporting material provided by the authors is posted with the online version of this article as a data supplement at (

Note: The authors would like to acknowledge the following surgeons who contributed patients to this study: Paul Kiritsis, MD, Midlothian, VA; Patrick Massey, MD, Shreveport, LA; Shane Barton, MD, Shreveport, LA; Thomas Throckmorton, MD, Germantown, TN; David Bailie, MD, Scottsdale, AZ; Jason Hurst, MD, New Albany, OH; Joaquin Sanchez-Sotelo, MD, Rochester, MN; and John Sperling, MD, Rochester, MN.

Data Sharing:

A data-sharing statement is provided with the online version of the article (


1. Athwal GS, Sperling JW, Rispoli DM, Cofield RH. Periprosthetic humeral fractures during shoulder arthroplasty. J Bone Joint Surg Am. 2009 Mar 1;91(3):594-603.
2. Singh JA, Sperling J, Schleck C, Harmsen W, Cofield R. Periprosthetic fractures associated with primary total shoulder arthroplasty and primary humeral head replacement: a thirty-three-year study. J Bone Joint Surg Am. 2012 Oct 3;94(19):1777-85.
3. Levy O, Copeland SA. Cementless surface replacement arthroplasty of the shoulder. 5- to 10-year results with the Copeland Mark-2 prosthesis. J Bone Joint Surg Br. 2001 Mar;83(2):213-21.
4. Levy O, Copeland SA. Cementless surface replacement arthroplasty (Copeland CSRA) for osteoarthritis of the shoulder. J Shoulder Elbow Surg. 2004 May-Jun;13(3):266-71.
5. Huguet D, DeClercq G, Rio B, Teissier J, Zipoli B; TESS Group. Results of a new stemless shoulder prosthesis: radiologic proof of maintained fixation and stability after a minimum of three years’ follow-up. J Shoulder Elbow Surg. 2010 Sep;19(6):847-52. Epub 2010 Mar 19.
6. Collin P, Matsukawa T, Boileau P, Brunner U, Walch G. Is the humeral stem useful in anatomic total shoulder arthroplasty? Int Orthop. 2017 May;41(5):1035-9. Epub 2017 Jan 3.
7. Hawi N, Magosch P, Tauber M, Lichtenberg S, Habermeyer P. Nine-year outcome after anatomic stemless shoulder prosthesis: clinical and radiologic results. J Shoulder Elbow Surg. 2017 Sep;26(9):1609-15. Epub 2017 Apr 11.
8. Athwal GS. Spare the canal: stemless shoulder arthroplasty is finally here: commentary on an article by R. Sean Churchill, MD, et al.: “Clinical and radiographic outcomes of the Simpliciti canal-sparing shoulder arthroplasty system. A prospective two-year multicenter study”.” J Bone Joint Surg Am. 2016 Apr 6;98(7):e28.
9. Berth A, Pap G. Stemless shoulder prosthesis versus conventional anatomic shoulder prosthesis in patients with osteoarthritis: a comparison of the functional outcome after a minimum of two years follow-up. J Orthop Traumatol. 2013 Mar;14(1):31-7. Epub 2012 Nov 9.
10. Denard PJ, Raiss P, Gobezie R, Edwards TB, Lederman E. Stress shielding of the humerus in press-fit anatomic shoulder arthroplasty: review and recommendations for evaluation. J Shoulder Elbow Surg. 2018 Jun;27(6):1139-47. Epub 2018 Feb 5.
11. Razmjou H, Holtby R, Christakis M, Axelrod T, Richards R. Impact of prosthetic design on clinical and radiologic outcomes of total shoulder arthroplasty: a prospective study. J Shoulder Elbow Surg. 2013 Feb;22(2):206-14. Epub 2012 Jul 21.
12. Uschok S, Magosch P, Moe M, Lichtenberg S, Habermeyer P. Is the stemless humeral head replacement clinically and radiographically a secure equivalent to standard stem humeral head replacement in the long-term follow-up? A prospective randomized trial. J Shoulder Elbow Surg. 2017 Feb;26(2):225-32. Epub 2016 Nov 15.
13. Bell SN, Coghlan JA. Short stem shoulder replacement. Int J Shoulder Surg. 2014 Jul;8(3):72-5.
14. Teissier P, Teissier J, Kouyoumdjian P, Asencio G. The TESS reverse shoulder arthroplasty without a stem in the treatment of cuff-deficient shoulder conditions: clinical and radiographic results. J Shoulder Elbow Surg. 2015 Jan;24(1):45-51. Epub 2014 Jul 11.
15. Churchill RS, Chuinard C, Wiater JM, Friedman R, Freehill M, Jacobson S, Spencer E Jr, Holloway GB, Wittstein J, Lassiter T, Smith M, Blaine T, Nicholson GP. Clinical and radiographic outcomes of the Simpliciti canal-sparing shoulder arthroplasty system: a prospective two-year multicenter study. J Bone Joint Surg Am. 2016 Apr 6;98(7):552-60.
16. Beck S, Beck V, Wegner A, Dudda M, Patsalis T, Jäger M. Long-term survivorship of stemless anatomical shoulder replacement. Int Orthop. 2018 Jun;42(6):1327-30. Epub 2018 Jan 24.
17. Gallacher S, Williams HLM, King A, Kitson J, Smith CD, Thomas WJ. Clinical and radiologic outcomes following total shoulder arthroplasty using Arthrex Eclipse stemless humeral component with minimum 2 years’ follow-up. J Shoulder Elbow Surg. 2018 Dec;27(12):2191-7. Epub 2018 Aug 6.
18. Rasmussen JV, Harjula J, Arverud ED, Hole R, Jensen SL, Brorson S, Fenstad AM, Salomonsson B, Äärimaa V. The short-term survival of total stemless shoulder arthroplasty for osteoarthritis is comparable to that of total stemmed shoulder arthroplasty: a Nordic Arthroplasty Register Association study. J Shoulder Elbow Surg. 2019 Aug;28(8):1578-86. Epub 2019 Apr 28.
19. Kadum B, Mafi N, Norberg S, Sayed-Noor AS. Results of the Total Evolutive Shoulder System (TESS): a single-centre study of 56 consecutive patients. Arch Orthop Trauma Surg. 2011 Dec;131(12):1623-9. Epub 2011 Aug 7.
20. Schoch C, Huth J, Aghajev E, Bauer G, Mauch F. Die metaphysär verankerte prothese bei posttraumatischer und primärer omarthrose. Obere Extrem. 2011;6(4):275-81.
21. Berth A, Pap G, Lohmann C. Indikationsspezifische mittelfristige ergebnisse der anatomischen kurzschaft-schulterprothese. Obere Extrem. 2012;7(2):83-90.
22. Brunner UH, Fruth M, Rückl K, Magosch P, Tauber M, Resch H, Habermeyer P. Die schaftfreie Eclipse-Prothese – indikation und mittelfristige ergebnisse. Obere Extrem. 2012;7(1):22-8.
23. Ballas R, Béguin L. Results of a stemless reverse shoulder prosthesis at more than 58 months mean without loosening. J Shoulder Elbow Surg. 2013 Sep;22(9):e1-6. Epub 2013 Feb 15.
24. Romeo AA, Thorsness RJ, Sumner SA, Gobezie R, Lederman ES, Denard PJ. Short-term clinical outcome of an anatomic short-stem humeral component in total shoulder arthroplasty. J Shoulder Elbow Surg. 2018 Jan;27(1):70-4. Epub 2017 Jul 19.
25. Giuseffi SA, Streubel P, Sperling J, Sanchez-Sotelo J. Short-stem uncemented primary reverse shoulder arthroplasty: clinical and radiological outcomes. Bone Joint J. 2014 Apr;96-B(4):526-9.
26. Jost PW, Dines JS, Griffith MH, Angel M, Altchek DW, Dines DM. Total shoulder arthroplasty utilizing mini-stem humeral components: technique and short-term results. HSS J. 2011 Oct;7(3):213-7. Epub 2011 Sep 27.
27. The Brigham and Women’s Hospital,, Inc. Department of Rehabilitation Services. Total Shoulder Arthroplasty/Hemiarthroplasty Protocol. 2007. Accessed 2013 Feb 13.
28. Katolik LI, Romeo AA, Cole BJ, Verma NN, Hayden JK, Bach BR. Normalization of the Constant score. J Shoulder Elbow Surg. 2005 May-Jun;14(3):279-85.
29. Tashjian RZ, Hung M, Keener JD, Bowen RC, McAllister J, Chen W, Ebersole G, Granger EK, Chamberlain AM. Determining the minimal clinically important difference for the American Shoulder and Elbow Surgeons score, Simple Shoulder Test, and visual analog scale (VAS) measuring pain after shoulder arthroplasty. J Shoulder Elbow Surg. 2017 Jan;26(1):144-8. Epub 2016 Aug 18.
30. Werner BC, Chang B, Nguyen JT, Dines DM, Gulotta LV. What change in American Shoulder and Elbow Surgeons score represents a clinically important change after shoulder arthroplasty? Clin Orthop Relat Res. 2016 Dec;474(12):2672-81. Epub 2016 Jul 8.
31. Murray DW, Fitzpatrick R, Rogers K, Pandit H, Beard DJ, Carr AJ, Dawson J. The use of the Oxford hip and knee scores. J Bone Joint Surg Br. 2007 Aug;89(8):1010-4.
32. Singh J, Sloan JA, Johanson NA. Challenges with health-related quality of life assessment in arthroplasty patients: problems and solutions. J Am Acad Orthop Surg. 2010 Feb;18(2):72-82.
33. Angst F, Goldhahn J, Drerup S, Aeschlimann A, Schwyzer HK, Simmen BR. Responsiveness of six outcome assessment instruments in total shoulder arthroplasty. Arthritis Rheum. 2008 Mar 15;59(3):391-8.
34. Habermeyer P, Lichtenberg S, Tauber M, Magosch P. Midterm results of stemless shoulder arthroplasty: a prospective study. J Shoulder Elbow Surg. 2015 Sep;24(9):1463-72. Epub 2015 Apr 11.

Supplemental Digital Content

Copyright © 2020 The Authors. Published by The Journal of Bone and Joint Surgery, Incorporated All rights reserved.