Translate this page into:
Ream-and-run technique offers equivalent clinical outcomes as anatomical total shoulder arthroplasty but with a high rate of complications: A systematic review and meta-analysis
, Robert W. Jordanb, Tanujan Thangarajahc, Simon MacLeand, Jarret Woodmasse, Peter D'Alessandrof,g, Shahbaz S. Malikh
⁎Corresponding author: Omar E. S. Mostafa. omar.mostafa1@nhs.net
-
Received: ,
Accepted: ,
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
Abstract
Abstract
This review aims to explore if the Ream and Run (RnR) technique is associated with better outcomes and lower complications than the traditional Anatomical Total Shoulder Replacement (aTSA) for osteoarthritis.
A systematic search of the literature was conducted using Medline, Embase and Cochrane in accordance with the PRISMA guidelines on March 2, 2024. Only comparative studies of adult patients with glenohumeral osteoarthritis (OA) comparing RnR and aTSA were included. Basic demographics, patient-reported outcome measures (PROMs) and complications were extracted and analysed. Quality assessment was performed using the Newcastle-Ottawa Score (NOS) tool and meta-analysis of outcomes reported by two or more studies was performed using Cochrane RevMan Web.
A total of 1548 patients were pooled from eight studies [RnR 738 vs aTSA 810]. Mean age in the RnR group ranged from 52.8 to 60.3 years with 93.6 % being male, compared with age range of 53–67.5 years in the aTSA group with 56 % being male. Five patients in the RnR group were Walch grade C or D, compared with three patients in the aTSA. No difference was observed between the two groups in post-operative SST score [P < 0.04], post-operative ASES score [P = 0.57] or degree of post-operative forward flexion [P < 0.41]. There was a statistically significant improvement in post-operative degree of external rotation, favouring RnR [MD -8.35, 95 % CI -14.69 to −2.01, P < 0.01] but without a significant clinical importance. The overall rate of complications in the RnR group was 15.4 % and 5.3 % in the aTSA group. The commonest reported complication in RnR group was chronic pain and stiffness (3.9 %) and soft tissue failure in the aTSA group (2.7 %). Overall rate of return-to-theatre was 7 % in RnR and 2.7 % in aTSA group.
Both aTSA and RnR offer improvement in shoulder PROMS. However, the overall re-operation rate and complications appeared high in RnR group. The choice of technique should be tailored to the patient's pre-operative baseline, activity level and desired goals.
Keywords
Shoulder osteoarthritis
Ream n run
Shoulder arthroplasty
Glenoid wear
Bone preservation
1 Introduction
Anatomical Total Shoulder Arthroplasty (aTSA) for shoulder arthritis in younger patients has shown good outcomes in terms of shoulder function and range of motion.1,2 However, challenge still remains about the longevity of aTSA in younger population.2,3 In the United Kingdom, the National Joint Registry (NJR) estimated a five-year revision rate of TSA up to 11 % in patients aged under 55 years, compared with figures between 3 % and 7 % in over 55 years.4 Several large, single-centre cohort studies reported the five-year survivorship of aTSA estimated to be up to 96 %5 with a risk of revision varying between 2 % and 23 % across the literature.6–8 At 10-years, the glenoid component survival has been reported to average between 62 % and 96 % for the general population, and 87 % for patients aged 55 and below with radiological glenoid loosening being reported in up to 40 % of the younger patients undergoing aTSA.9,10 Across stemmed and stemless aTSA, glenoid loosening is reported in up to 43 % of cases analysed,11 and both implant types are reported to have similar complication profiles and revision rates in the short term.12 One recent retrospective study examining mid-term outcomes of stemless aTSA reported no incidence of glenoid loosening in a population of 60 patients.13 However, there remains a paucity of literature on the rate of glenoid loosening within stemless aTSA. Other common complications reported in stemless aTSA included dislocation of prosthesis14 but no differences in rates of periprosthetic fractures or revision surgery were observed between stemless or stemmed aTSA.15
The ‘Ream-and-Run’ (RnR) technique has been advocated as an alternative option to overcome the long-term issues with glenoid wear and subsequent bone osteolysis.3 Originally described by Matsen III and Lippitt,16,17 this technique involves appropriate glenoid reaming and maximal bone preservation in the humeral component followed by a strict rehabilitation plan. Their technique demonstrated promising outcomes in a case series of 544 patients; a slightly higher average Simple Shoulder Test (SST) score with RnR compared with aTSA (10 vs 9.5, respectively).17,18 However, to our knowledge, no previous meta-analysis has been performed to compare the outcomes of both techniques. The primary aim of this systematic review and meta-analysis is to report on patient reported outcomes, range of motion and rates of revision in patients undergoing RnR versus aTSA.
2 Methods
This systematic review and meta-analysis were reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement19 using Medline, Cochrane, Embase and Google Scholar databases. The review was registered on the INPLASY database (ID: INPLASY202470094). The search and selection of studies was performed by 2 authors independently on March 2, 2024. Additional studies were identified by screening bibliographies of studies meeting inclusion criteria. Any disagreements were resolved by a third senior author.
2.1 Inclusion criteria
Comparative randomised controlled trials, case-control studies and cohort studies that reported on RnR and aTSA patients were included. Only studies with adult patients (age >18 years) and who reported either functional scores, revision rates, complications or radiographic outcomes were included. Case reports, case series, conference abstracts, editorials and reviews on this topic were excluded.
2.2 Data extraction
A data sheet was designed to extract comparative data between RnR and aTSA regarding demographic data and assess outcomes in terms of.1.Patient Reported Outcome Measure (PROM)2.Range of Motion (ROM)3.Radiographic outcomes4.Complications5.Revision Surgery
2.3 Evaluation of bias
The risk of bias was assessed using the validated Newcastle-Ottawa Scale (NOS) scoring tool.20 This tool considers the quality of non-randomised studies based on selection, comparability and ascertainment of the exposure or outcome of interest. Summary of the risk of bias is demonstrated in Table 1.
| Author | Year | Selection | Comparability | Outcome | Total score | Total Quality | |||||
| Representativeness of the Exposed Cohort | Selection of the Non-Exposed Cohort | Ascertainment of Exposure | Demonstration That Outcome of Interest Was Not Present at Start of Study | Comparability of Cohorts on the Basis of the Design or Analysis | Assessment of Outcome | Was Follow-Up Long Enough for Outcomes to Occur | Adequacy of Follow Up of Cohorts | ||||
| Sharareh et al. 22 | 2023 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
| Levins et al. 23 | 2023 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
| Schiffman et al. 24 | 2020 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
| Matsen et al. 18 | 2019 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
| Gowd et al. 26 | 2019 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
| Garcia et al. 27 | 2019 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
| Virk et al. 28 | 2018 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
| Clinton et al. 29 | 2007 | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | ∗ | 8 | Good |
2.4 Statistical analysis
Statistical heterogeneity was quantified using the I2 statistic; a value of 0 % represent no heterogeneity, 25 % represent low heterogeneity, 50 % represent moderate heterogeneity and >75 % represent high heterogeneity. Due to heterogeneity in some outcome reporting, only pooling and meta-analysis of patient-reported outcome measures were performed. In outcomes that were precluded from meta-analysis, descriptive statistics, including standard deviation (SD) or range, were included as reported in the original studies. Summary forest plots were generated as means of representing outcomes reported by two or more studies. Continuous variables were reported as mean (standard deviation) or as mean difference with 95 % confidence interval (CI) where appropriate. Categorical data was reported in the form of percentages or frequency. Meta-analysis was performed using Cochrane Review Manager (RevMan) [Computer Programme], version 7.2 (Cochrane Collaboration, 2024)21 based on a random-effect model. A P value of <0.05 was used where statistical significance is reported.
2.5 Sensitivity analysis
Sensitivity analysis was performed using Cochrane Review Manager (RevMan) software to investigate bias and assumptions made for meta-analysed clinical outcomes amongst studies with possible repetition of included populations and studies with high rates of losses to follow-up. This will be performed by removing studies from the meta-analysis. Sensitivity analysis is reported for each outcome reported in three or more studies.
3 Results
The literature search identified 188 studies whose titles and abstracts were screened (Fig. 1). After removal of duplicate studies and application of eligibility criteria, eight studies22–29 met the inclusion criteria; eight studies were retrospective comparative studies (Level III). No randomised trials were identified. Summary of the studies and demographic data are provided in Table 2.
| Study | LOE | Population | Intervention | Comparator | Mean f/up | Outcomes | Results (RnR vs aTSA) |
| Sharareh et al., 2023 22Combined N = 63 in final analysis | III | Mean age63.6 ± 8.5M:F 45:18Walch CxA1 – 4 vs 1A2 – 8 vs 7B1 – 7 vs 9B2 – 13 vs 7D – 0 vs 1 | RnR 34Mean age 60.3 ± 7.2M:F 33:1Implant:DePuy Synthes | TSA 29Mean age 67.5 ± 8.3M:F 12:17Implant:DePuy Synthes | 10.3 years (Range 10–11 years)Lost to follow-up: 64/127 (50 %) from originally enrolled cohort | VASSSTMCIDSCBRe-operation | SSTPreop: 5.4 ± 2.5 vs 3.8 ± 2.6Postop: 10.3 ± 2.1 vs 8.9 ± 2.6VASPreop: 6.5 ± 1.6 vs 6.6 ± 2.2Postop: 0.9 ± 1.3 vs 1.2 ± 2.3Reoperation N = 5 vs N = 0 |
| Levins et al., 2023 23Combined N = 78 in final analysis | III | Mean age58.6 ± 7.3 yearsM:F 78:0Walch CxA – 11 vs 7B – 26 vs 30C – 2 vs 2 | RnR 39Mean age 58.2 ± 8.8M:F 39:0Implant: N/A | TSA 39Mean age 58.9 ± 5.5M:F 39:0Implant: N/A | 4.4 ± 2.3 yearsLost to follow-up: 10/57 (17.5 %) from originally enrolled cohort | SSTASESDASHROMHRQoL (SF-6D, EQ-5D)SatisfactionVASRe-operation | SSTPreop: 4.7 ± 2.7 vs 4.1 ± 2.2Postop: 10.3 ± 2.2 vs 10.9 ± 1.9DASHPreop: 36.5 ± 14.5 vs 39.0 ± 13.4Postop: 11.1 ± 14.8 vs 7.6 ± 9.1ASESPreop: 34.8 ± 15.2 vs 33.3 ± 14.1Postop: 85.0 ± 18.9 vs 89.9 ± 12.8VASPreop: 7.0 ± 1.9 vs 7.1 ± 1.9Postop: 1.5 ± 2.1 vs 1.1 ± 1.8SatisfactionN = 35 vs N = 36Reoperation N = 3 vs N = 2 |
| Schiffman et al., 2020 24Combined N = 640 in final analysis | III | Mean ageM:F 438:202 | RnR 295Mean age 58 ± 10M:F 272:23 | TSA 345Mean age 66 ± 11M:F 166:179 | 2.9 ± 2 vs 3 ± 2Lost to follow-up: 89/640 (13.9 %) from originally enrolled cohort | SSTSANERe-operation | SSTPreop: 4.9 ± 2.5 vs 2.8 ± 2.3Postop: 10.1 ± 2.6 vs 9.4 ± 2.7Re-operation N = 26 vs N = 9 |
| Matsen et al., 2019 18Combined N = 544 in final analysis | III | Mean ageM:F 374:170Walch CxA1 – 6 vs 7A2 – 93 vs 158B1 – 40 vs 35B2 – 120 vs 77C – 2 vs 0 | RnR 263Mean age 58 ± 9M:F 242:21Implant: DePuy Synthes | TSA 281Mean age 67 ± 10M:F 132:149Implant: DePuy Synthes | N/ALost to follow-up: 11 % RnR group vs 12 % TSA group from originally enrolled cohort (raw figures not reported) | SSTRe-operation%MPI | SSTPreop: 4.9 ± 2.4 vs 2.9 ± 2.3 P < 0.01Postop: 10.0 ± 2.6 vs. 9.5 ± 2.7 P = 0.021Re-operation N = 34 N = 13 |
| Gowd et al., 2019 26Combined N = 53 in final analysis | III | M:F 48:5 | RnR 25Mean age52.8 ± 7.7M:F 23:2Implant: N/A | TSA 28Mean age53.3 ± 9.2M:F 25:3Implant: N/A | 69.1 ± 24.8 monthsLost to follow-up: 0 in RnR group vs 11/64 (17.2 %) in TSA group from originally enrolled cohort | Return to workTime to returnASESSatisfactionMCSPCSMCIDSCB | SatisfactionGood/Excellent 94.4 % vs 100 % P = 0.285MCS54.3 ± 8.4 vs 56.2 ± 7.8 P = 0.401PCS48.8 ± 9.8 vs 47.0 ± 10.0 P = 0.500Return to work100 % vs 89.3 % P = 0.091Time to return (months)2.5 ± 4.8 vs 2.0 ± 2.6 P = 0.653 |
| Garcia et al., 2019 27Combined N = 56 in final analysis | III | Mean age53.0 ± 8.5 yearsM:F 51:5 | RnR 26Mean age53.1 ± 7.7M:F 24:2Implant: N/A | TSA 30Mean age53.6 ± 9.0M:F 27:3Implant: N/A | 69.1 ± 24.8 months | Failures and re-operationSportsSatisfactionPROMsRadiographic Parameters | ASESPreop: 48.0 ± 13.9 vs 42.2 ± 14.4 P = 0.093Delta ASES: 36.8 ± 23.8 40.9 ± 23.9 P = 0.531Return to sport94.4 % vs 86.4 %, P = 0.395Return to high demand92.3 % vs 81.3 %, P = 0.390Return to same demand83.3 % vs 72.7 %, P = 0.424Reoperation N = 3 vs N = 2 |
| Virk et al., 2018 28Combined N = 44 in final analysis | III | Mean ageM:F 40:4Walch CxA1 – 11 vs NA2 – 4 vs NB1 – 3 vs NB2 – 4 vs NC – 1 vs N | RnR 21Mean age54 ± 7.1M:F 19:2Implant: N/A | TSA 23Mean age53 ± 7.2M:F 21:2Implant: N/A | RnR vs TSA37 ± 12 vs 46 ± 15 | ReoperationPROMsROMRadiographic assessment | SSTPreop: 6.2 ± 0.6 vs 5.6 ± 0.5Postop: 9.9 ± 3 vs 9.9 ± 0.5, P = 0.64ASESPreop: 45 ± 4 vs 46 ± 4Postop: 85 ± 4 vs 85 ± 3, P = 0.9VASPreop: 5.4 ± 0.5 vs 4.7 ± 0.5Postop: 1.4 ± 0.4 vs 0.9 ± 0.3, P = 0.36 |
| Clinton et al., 2007 29Combined N = 70 in final analysis | III | Mean age56.0M:F 32:3 | RnR 35Mean age56.0M:F 32:3Implant: DePuy | TSA 35Mean age56.0M:F 32:3Implant: DePuy | 2.6 years vs 2.7 yearsLost to follow-up: 2/35 (5.7 %) in RnR group vs N in TSA group from originally enrolled cohort | SST 6-monthly | SSTInitial: 4.5 vs 4.0 P = 0.366months: 6.67 vs 8.12 P = 0.2312months: 7.77 vs 9.64 P < 0.0136 months 9.45 ± 3.3 vs 9.96 ± 2.84 P=NS |
A total of 1548 patients were included: 738 patients in the RnR group and 810 patients in the aTSA group. Within the RnR group, 93.6 % of patients (n = 684) were male compared with 56 % of patients in the aTSA group (n = 454). The mean age in RnR group ranged from 52.8 to 60.3 years, compared with 53–67.5 years in the aTSA group. Mean follow up ranged between two and eleven years. Three of eight studies described the implant type, which was DePuy Synthes (Indiana, United States). The pre-operative glenoid classification was reported in four studies; in the RnR group, 137 patients were Grade A, 213 patients were Grade B and five patients were Grade C or above. In the aTSA group, 180 patients were Grade A, 158 patients were Grade B and three patients were Grade C or above. Only one study reported the exact proportion of stemless or short-stemmed versus standard humeral implant, described in Table 2.
3.1 Clinical outcomes
The commonest described Patient Reported Outcome Measures (PROMs) were the Simple Shoulder Test (SST) in six studies22–25,28,29 followed by American Shoulder and Elbow Surgeons (ASES) score in four studies.23,26–28 The Visual Analogue Scale (VAS) was reported in three studies.22,23,28Table 3 shows the summarised clinical outcomes of each study.
| Study | Clinical and Functional Outcomes (Pre op vs Post op) | ||||
| Ream-and-Run | Total Shoulder Replacement | Pre op vs Post op P-value between 2 groups (or post op only between 2 groups) | P value RnR | P value aTSA | |
| Sharareh et al., 2023 22N = 63 | SST 5.4 ± 2.5 vs 10.3 ± 2.1VAS 6.5 ± 1.6 vs 0.9 ± 1.3ROM not reported | SST 3.8 ± 2.6 vs 8.9 ± 2.6VAS 6.6 ± 2.2 vs 1.2 ± 2.3ROM not reported | SST .001 vs .001 VAS .001 vs .001 | SST .001 vs .001 VAS .001 vs .001 | |
| Levins et al., 2023 23N = 78 | SST 4.7 ± 2.7 vs 10.3 ± 2.2DASH 36.5 ± 14.5 vs 11.1 ± 14.8ASES 34.8 ± 15.2 vs 85 ± 18.9VAS 7.0 ± 1.9 vs 1.5 ± 2.1Forward Flexion 109 ± 24 vs 142.8 ± 12.2External Rotation 12.2 ± 22.6 vs 45.2 ± 14.9 | SST 4.1 ± 2.2 vs 10.9 ± 1.9DASH 39.0 ± 13.4 vs 7.6 ± 9.1ASES 33.3 ± 14.1 vs 89.9 ± 12.8VAS 7.1 ± 1.9 vs 1.1 ± 1.8Forward Flexion 106.8 ± 26.1 vs 144.5 ± 12.8External rotation 12.8 ± 21.8 vs 40.8 ± 12.5 | SST 0.29 vs 0.04DASH 0.43 vs 0.05ASES .67 vs .047VAS 0.9 vs 0.09FF 0.71 vs 0.69ER .98 vs .04 | ||
| Schiffman et al., 2020 24N = 640 | Pain 6.8±1.9 (2.0–10.0)SST 4.9±2.5 (0.0–11.0) vs 10.1±2.6 (0.0–12.0)SANE 41±19 (0–100) vs 80 vs 19 (0–100) | Pain 7.1±1.9 (1.0–10.0)SST 2.8±2.3 (0.0–10.0) vs 9.4±2.7 (0.0–12.0)SANE 35±21 (0–100) vs 84±17 (20–100) | Pain 0.024 vs n/aSST 0.135 vs 0.065SANE 0.168 vs 0.095 | Pain 0.808 vs n/aSST .02 vs .01SANE .058 vs < .001 | |
| Matsen et al., 2019 18N = 544 | SST 4.9 ± 2.4 (0.0–11.0) vs 10.0 ± 2.6 (0.0–12.0) | SST 2.9 ± 2.3 (0.0–10.0) vs 9.5 ± 2.7 (0.0–12.0) | SST <.001 vs .021 | ||
| Gowd et al., 2019 26N = 53 | ASES 49.1 ± 12.1Change ASES 35.9 ± 19.3Return to work 100 %Time to work 2.5 ± 4.8 | ASES 44.3 ± 14.3Change ASES 41.5 ± 22.7Return to work 89.3 %Time to work 2.0 ± 2.6 | ASES 0.12Change ASES 0.344Return to Work 0.091Time to work 0.653 | ||
| Garcia et al., 2019 27N = 56 | ASES 4.8±13.9Change ASES 36.8 ± 23.8Return to sport 94.4 %Return to high demand 92.3 %Return to same demand 83.3 % | ASES 42.2±14.4Change ASES 40.9 ± 23.9Return to sport 86.4 %Return to high demand 81.3 %Return to same demand 72.7 % | ASES 0.093 vs 0.531Return to sport 0.395Return to high demand 0.39Return to same demand 0.424 | ||
| Virk et al., 2018 28N = 44 | SST 6.2±0.6 (5–7.5) vs 9.9 ± 3 (8.5–11)ASES 45 ± 4 (37–53) vs 85 ± 4 (77–94)VAS 5.4 ± 0.5 (4.4–6.4) vs 1.4 ± 0.4 (0.6–2.3)FF 115±6 (102–127) vs 139±6 (124–151)ER 26±3 (20–33) vs 47 ± 4 (39–55) | SST 5.6±0.5 (4.4–6.7) vs 9.9 ± 0.5 (8.8–10.9)ASES 46 ± 4 (39–54) vs 85 ± 3 (78–92)VAS 4.7 ± 0.5 (3.6–5.7) vs 0.9 ± 0.3 (0.3–1.6)FF 113 ± 9 (94–133) vs 154 ± 5 (143–165)ER 39 ± 5 (29–50) vs 58 ±3 (51–65) | (Only comparison of post op values between the two groups)SST 0.64ASES 0.9VAS 0.36FF 0.06ER .02 | ||
| Clinton et al., 2007 29N = 70 | SST 4.49 ± 4.9 vs at 12 months 7.77 ± 3 | SST 3.97 ± 4.5 vs 9.64±1.63 | SST .36 vs .01 | ||
In the RnR group, mean SST score ranged from zero to 11.0 pre-operatively vs zero to 12.0 post-operatively. The mean SST score in aTSA ranged from zero to 10.0 pre-operatively vs zero to 12.0 post-operatively. The absolute post-operative SST results were pooled from six studies into the meta-analysis22–25,28,29 [Fig. 2]. No difference was observed in post-operative SST scores between RnR and aTSA [MD 0.25, 95 % CI -0.32 – 0.82, P < 0.04]. Sensitivity analysis did not result in a change in statistical significance. There was moderate to high heterogeneity, which was statistically significant [I2 = 66 %, P = 0.01].
In the RnR group, the mean ASES score ranged from 34.8 to 53 pre-operatively vs 77 to 94 post-operatively. In the aTSA group, the mean ASES score ranged from 33.3 to 54 pre-operatively vs 78 to 92 post-operatively. The absolute post-operative ASES results were pooled from two studies into the meta-analysis23,28 [Fig. 3]. No difference was observed in post-operative ASES scores between RnR and aTSA [MD -1.20, 95 % CI -5.34 – 2.93, P = 0.57]. There was a low to moderate heterogeneity between studies, although without significance [I2 = 40 %, P = 0.20]. The change in ASES (Δ ASES) results was pooled from another two studies into the meta-analysis26,27 [Fig. 4]. No difference was observed in the change in ASES scores between RnR and aTSA [MD -4.93, 95 % CI -13.32 – 3.47, P = 0.25]. There was no heterogeneity between studies, although without significance [I2 = 0 %, P = 0.86].
In the RnR group, the mean VAS ranged from 1.0 to 10 pre-operatively vs 0.6 to 2.3 post-operatively. In the aTSA group, the mean VAS ranged from 1.0 to 10.0 pre-operatively vs 0.3 to 1.6 post-operatively. The post-operative VAS results were pooled from three studies into the meta-analysis22,23,28 [Fig. 5]. No difference was observed in post-operative VAS scores between RnR and aTSA [MD -2.34, 95 % CI -6.77 – 2.08, P = 0.30]. On exclusion of the study by Sharareh et al. sensitivity analysis found a significant difference between aTSA and RnR, in favour of aTSA [MD 0.49, 95 % CI 0.29–0.70, P < 0.00001]. There was significantly high heterogeneity between studies [I2 = 99 %, P < 0.00001].
3.2 Range of motion
In the RnR group, the mean FF ranged from 85° to 132° pre-operatively vs 124° to 155° post-operatively. In the aTSA group, the mean degree of FF ranged from 80 to 133 pre-operatively vs 131° to 165° degrees post-operatively. Only two studies reported on ROM which were pooled into the meta-analysis.23,28 Forward Flexion (FF) and External Rotation (ER) were the two reported outcomes. No difference was observed in degree of post-operative FF between RnR and aTSA [MD -6.81, 95 % CI -23.17 to 9.56, P < 0.41] [Fig. 6]. There was significantly high heterogeneity between studies [I2 = 96 %, P < 0.00001].
Mean external rotation in the RnR group ranged from 0° to 34° pre-operatively vs 30° to 60° degrees post-operatively. In the aTSA group, the mean ER ranged from 0° to 50° pre-operatively vs 28° to 65° post-operatively. There was a statistically significant difference in post-operative degree of ER, favouring RnR [MD -8.35, 95 % CI -14.69 to −2.01, P < 0.01] [Fig. 7]. There was high heterogeneity between studies, although without significance [I2 = 75 %, P = 0.05].
3.3 Radiographic outcomes
Only three studies reported radiographic outcomes in RnR and aTSA, but no meta-analysis was conducted due to heterogeneity in radiographic assessment methods across the included studies and lack of data to produce a pooled estimate.23,27,28 Cases of posterior humeral subluxation were reported in two studies23,28; RnR (n = 2) and aTSA (n = 1). Glenoid loosening was reported in one study23; three cases within the aTSA group compared to none in the RnR group. One study reported radiographic post-operative decentring and medialisation of the humerus27; RnR had greater degree of medialisation of the humeral head (RnR −2.4 ± 5.0, TSA -2.2 ± 5.7) but a lower percentage of decentring (RnR 3.6 % ± 2.6 %, TSA 4.3 % ± 3.3 %). The authors found no statistical significance was reported between RnR and aTSA in both outcomes (p = 0.913 and p = 0.795).
3.4 Complications
Complications were reported in seven of eight studies.22–28 The overall rate of complications in the RnR group was 15.4 % (114 of 738) and 5.3 % (43 of 810) in the aTSA group. The three commonest reported complications in RnR group were chronic pain and stiffness, 3.9 % (n = 29), humeral head problems, 2.6 % (n = 19) and culture-positive infection, 2.0 % (n = 15). Similarly, the three most reported complications in the aTSA group were soft tissue failure, 1.2 % (n = 10), chronic pain and stiffness, 1.2 % (n = 10) and loosening of glenoid prosthesis, 0.86 % (n = 7).
The rate of return to theatre for open revision surgery in the RnR group was 7 % (52 of 738) and 2.7 % in the aTSA group (22 of 810). Tissue and culture samples were obtained either routinely or for a high suspicion of infection. Of those returning to theatre, 13.5 % (7 of 52) were suspected to have an infection in the RnR group and. 4.5 % (1 of 22) in the aTSA group. Following histology and microbiology results, in the RnR group, 11 of 52 patients (21 %) were reported to have culture-positive results, compared with 2 of 22 patients (9 %) in the aTSA group. The organism reported in all cases was Cutibacterium.
3.5 Risk of bias assessment
The results of quality assessment are demonstrated in Table 1. Overall, the quality of non-randomised, observational studies was rated ‘good’ for their level of selection of patient cohorts, comparability in design and the follow-up process. All studies scored eight or more. Although follow-up was high within the studies with a low rate of attrition reported as demonstrated in Table 2), the non-randomised nature of those studies inherently predisposes them to a high level of selection and reporting biases.
4 Discussion
The main findings of this systematic review and meta-analysis are that no difference was observed between RnR and aTSA regarding PROMs but a significant improvement in range of motion was observed post-operatively, favouring RnR over aTSA. Based on findings from comparative literature studies, RnR has also been reported to be associated with a higher incidence of revision surgery and infection but limitations of the studies precludes drawing of firm conclusions.
Our meta-analysis demonstrated no difference in pooled effects of post-operative SST, ASES or VAS scores [P = 0.20, P = 0.71 and P = 0.08, respectively] between RnR and aTSA. No difference was observed in pre-vs-post ASES across both groups [P = 0.25]. Previous meta-analyses reported improved post-operative SST, ASES and VAS scores across RnR but did not report on statistical significance.30 Different outcomes have been observed in retrospective cohort studies included in this review; Sharareh et al.22 reported significant improvement in post-operative SST (RnR p = 0.001, aTSA p = 0.001) and VAS (RnR p = 0.001, aTSA p = 0.001) scores. However, Schiffman et al.24 reported post-operative SST statistical improvement only within the aTSA group post-operatively (aTSA p = 0.01, RnR p = 0.65). In single-arm studies, retrospective cohort analysis of RnR outcomes showed significant improvement in PROM; Somerson et al.31 reported significant improvement in SST, ASES and VAS post-operatively (P < 0.001). Harold et al.32 reported no significant difference in SST scores in the aTSA group after an 8-year follow-up period (P = 0.39). Thus, it is clear from the previous systematic reviews and this meta-analysis is that both procedures improve patient functional outcomes.
Range of motion was only reported in two studies.23,28 The improved degree of post-operative ER [MD -8.35°] achieved in RnR post-operatively was statistically significantly higher than in aTSA group [P < 0.01], but likely without any clinically meaningful difference. There was no difference observed in degree of post-operative FF between the two groups [P = 0.41]. Kelly et al.30 reported similar improvements within RnR in their meta-analysis but no comparisons with aTSA were carried out. Single-arm, non-comparative studies also reported similar outcomes in RnR; Levins et al.33 and Somerson et al.31 reported significant improvement in post-operative ER (P < 0.01 and P = 0.002) and FF (P < 0.01 and P < 0.001) respectively. Similarly, Vervaecke et al.34 reported statistically significant improvement in ER and FF in aTSA group after a 26 month follow-up period.
Radiographic outcomes were only reported in three studies.23,27,28 One study reported no statistical significance in post-operative radiographic decentring or humeral head medialisation between RnR or aTSA (p = 0.913 and p = 0.795). There were more cases of glenoid loosening (n = 3) and less posterior subluxation (n = 1) with aTSA compared with RnR (loosening n = 0 and subluxation n = 2, respectively). Kelly et al. reported five patients (n = 5) with posterior humeral head subluxation within RnR across two retrospective cohort studies with a follow-up of two years or more. The authors reported no significant difference in PROM of those with or without subluxation in RnR group (P = 0.794). These findings converse with the purpose that RnR is supposed to achieve, particularly in patients with an arthritic triad; RnR provides more stabilisation through conservative, concentric reaming whilst improving function.35 Kelly et al. did not postulate a reason for their review finding. Still, this difference in outcomes may be attributed to a higher percentage of patients with pre-operative Walch B2 glenoid morphology (46.3 % of the total pooled values) across their review, in addition to possible inaccuracy in delineating between humeral subluxation or glenoid loosening.36 To our knowledge and search, there is no literature that explores if the difference in rates of subluxation is secondary to surgical technique, which calls for robust controlled trials to delineate the superiority of either technique on rates of subluxation in patients of different glenoid morphology.
Matsen III and Lippitt described the ream-and-run technique for those more active, higher physical demand individuals with a better likelihood for engaging with physiotherapy and post-operative rehabilitation.16 They report that the technique is technically challenging, and the outcomes depend on the surgeon's experience and patient co-operation.16,17 Our review reported a greater proportion of male patients (93.6 %) and younger age range (52.8–60.3) within the RnR group, similar to that reported in previous studies; a systematic review by Kelly et al.30 reported a mean age of 56.7 years (range 22–81) and 91 % males. Levins et al.33 cohort study included 87 % of males within RnR group with an age group of 47–75 years. Similarly, Matsen III et al.25 consecutive series of 544 patients included a higher degree of male patients (RnR 92 %, aTSA 47 %) and younger average patients' age (RnR 58 years, aTSA 67 years).25 Their regression analysis demonstrated a correlation between age, gender, and outcomes. The authors report that younger patients and males had worse results with aTSA (p < 0.001) suggesting that the RnR technique could be better implemented in young males patients. The reason behind this remains poorly understood, but it is theorised that variation in glenoid anatomy across the genders may play a role in glenoid implant success.37 The lack of subgroup analysis within previous studies suggests the need for future research to investigate correlation more robustly.
The ream-and-run technique can pose challenges to the operating surgeon, and therefore, surgical expertise can dictate success or complication rate.16 This review observed a greater overall incidence of complications in the RnR group compared with aTSA; 15.4 % vs 5.3 %, respectively. The RnR rate is similar to that reported by Kelly et al. (15.2 % in RnR), but they did not make any comparison to TSA30 unlike this systematic review and meta-analysis. The commonest reason for complication in RnR was pain (acute or chronic) and stiffness, whilst in the aTSA group it is failure of the soft tissue and requirement for further corrective surgery. Non-comparative, single-arm studies also reported similar findings to our review; Somerson et al.31 reported three (n = 3, 15 %) patients within an RnR-only study who had revision surgery to aTSA secondary to pain or instability. Levins et al.33 reported six (n = 6, 12.2 %) patients who had to undergo revision in an RnR-only study, of which one (n = 1, 2 %) had an acute infection that required revision, and the rest had revision due to ongoing pain. Furthermore, although the suspected rates of infection in both groups was reported to be low (RnR 13.5 %, aTSA 4.5 %), the incidence of culture and tissue bacterium-positive samples was higher in the RnR group compared with aTSA; 21 % vs 9 %, respectively. Levins et al.33 reported four (n = 4, 66.6 %) patients with a positive culture of Cutibacterium who were asymptomatic initially. The relatively younger age demographic in our cohort is a recognised risk factor for such micro-organisms38 but the heterogeneity amongst methodologies and patient demographics in other studies precludes this review from a conclusion regarding the higher incidence of tissue and culture microorganism positivity in the RnR group compared with aTSA. This is an important consideration in further studies, and sub-group analysis should be performed in future reviews to evaluate for impact of such characteristics on patients short and long-term outcomes including prosthesis infections.
5 Limitations
The heterogeneity in reporting outcomes and methodology precludes this review from performing meta-analysis for radiographic outcomes and complications. This heterogeneity was reported to be significant within certain PROMs, such as SST, VAS, and degree of FF [I2 = 66 %, 99 % and 96 %, respectively]. Additionally, some patient characteristics were not explored in our review that could have impacted the significance of these findings, such as smoking status, presence of medical co-morbidities and medication use.16,17However, we included the pre-operative glenoid classification in the results to reflect the greater proportion of worse glenoid pathology within the RnR group, as this would affect our interpretation of complications. Additionally, the effect that high proportion of male patients with a younger mean age range in our cohort may not be translated to other subgroups, such as female gender; the outcomes of RnR on such a group of patients remains unclear. The methodological heterogeneity of included studies has precluded us from performing a meaningful subgroup analysis of the two interventions by age, gender or glenoid classification. Furthermore, all the studies included were of level III the lack of robust, randomised controlled trials inadvertently causes selection bias. Nonetheless, the NOS tool indicated a good level of individual study design. Future research should use robust methodology to exclude the confounding factors highlighted in this limitation section and focus on the differences in outcome of the two surgical techniques.
6 Conclusion
Accounting for methodological limitations, both aTSA and RnR offer improvement in shoulder PROMS. However, the overall re-operation and complication rate appeared high in RnR group compared to aTSA. Therefore, the choice of surgical technique should be individualised to the patient's pre-operative activity and baseline, including their desired goals from surgery. This is certainly more applicable in younger, more active cohorts of patients. Further randomised control trials are needed to assess superiority in clinical, functional, and long-term outcomes of one technique over another.
Informed consent
Not Applicable.
Gaurdian consent
No patient consent/guardian consent is required as this is a systematic review of published data.
Ethical approval
Not Applicable.
Guarantor
SSM.
Contributorship
OESM and SSM conceptualised the idea and performed a literature search. OESM and RWJ performed the literature review and data extraction. OESM, TT, SM and HW contributed to data analysis and review. OESM, RWJ, JW and PA completed the original draft writing. Supervision and validation of work provided by JW, PA and SSM. Final draft and approval contributed to by all authors.
Ethical approval
No ethical approval was required as this is a systematic review of published work.
Funding statement
No funding was applicable/acquired.
References
- Outcomes of total shoulder arthroplasty in patients younger than 65 years: a systematic review. J Shoulder Elb Surg. 2017;26:1298-1306.
- [Google Scholar]
- Glenoid bone loss in primary total shoulder arthroplasty: evaluation and management. JAAOS - J Am Acad Orthop Surg. 2012;20:604.
- [Google Scholar]
- Revision rates and associated risk factors after shoulder arthroplasty. J Clin Med. 2022;11:7256.
- [Google Scholar]
- Analysis on revision rates of shoulder arthroplasty at a single referral center in Canada. Semin Arthroplasty JSES. 2023;33:535-541.
- [Google Scholar]
- Serious adverse events and lifetime risk of reoperation after elective shoulder replacement: population based cohort study using hospital episode statistics for England. Br Med J. 2019;364
- [Google Scholar]
- Revision surgery following total shoulder arthroplasty: analysis of 2588 shoulders over three decades (1976 to 2008) J Bone Joint Surg Br. 2011;93-B:1513-1517.
- [Google Scholar]
- Anatomical total shoulder arthroplasty used for glenohumeral osteoarthritis has higher survival rates than hemiarthroplasty: a Nordic registry-based study. Osteoarthr Cartil. 2018;26:659-665.
- [Google Scholar]
- Mid- to long-term follow-up of total shoulder arthroplasty using a keeled glenoid in young adults with primary glenohumeral arthritis. J Shoulder Elb Surg. 2013;22:894-900.
- [Google Scholar]
- Management of glenohumeral arthritis in the young patient – a systematic review. J Arthrosc Jt Surg. 2019;6:9-20.
- [Google Scholar]
- Stemless anatomic total shoulder arthroplasty: surgical technique and review of the literature. JSES Rev Rep Tech. 2022;2:442-450.
- [Google Scholar]
- Anatomic total shoulder arthroplasty with stemless humeral component, nonspherical head, and inlay glenoid: clinical outcomes at mean 4.4-year follow-up. Semin Arthroplasty JSES. 2024;34:354-363.
- [Google Scholar]
- Stemless reverse total shoulder arthroplasty: a systematic review. J Shoulder Elb Surg. 2022;31:1083-1095.
- [Google Scholar]
- Outcomes between stemmed and stemless total shoulder arthroplasty: a systematic review and meta-analysis of randomized controlled trials. JAAOS Glob Res Rev. 2022;6
- [Google Scholar]
- The ream and run: not for every patient, every surgeon or every problem. Int Orthop. 2015;39:255-261.
- [Google Scholar]
- Shoulder hemiarthroplasty with nonprosthetic glenoid arthroplasty: the ream-and-run procedure. JBJS Rev. 2021;9
- [Google Scholar]
- The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71.
- [Google Scholar]
- Minimum 10-year follow-up of anatomic total shoulder arthroplasty and ream-and-run arthroplasty for primary glenohumeral osteoarthritis. J Shoulder Elb Surg. 2023;S1058274623007073
- [Google Scholar]
- Comparison of humeral-head replacement with glenoid-reaming arthroplasty (ream and run) versus anatomic total shoulder arthroplasty: a matched-cohort study. J Bone Jt Surg. 2023;105:509-517.
- [Google Scholar]
- Impact of previous non-arthroplasty surgery on clinical outcomes after primary anatomic shoulder arthroplasty. J Shoulder Elb Surg. 2020;29:2056-2064.
- [Google Scholar]
- Ream and run and total shoulder: patient and shoulder characteristics in five hundred forty-four concurrent cases. Int Orthop. 2019;43:2105-2115.
- [Google Scholar]
- Comparative analysis of work-related outcomes in hemiarthroplasty with concentric glenoid reaming and total shoulder arthroplasty. J Shoulder Elb Surg. 2019;28:244-251.
- [Google Scholar]
- Return to sport following hemiarthroplasty with concentric reaming versus total shoulder arthroplasty: a matched pair analysis. Orthopedics. 2019;42:276-284.
- [Google Scholar]
- Nonprosthetic glenoid arthroplasty with humeral hemiarthroplasty and total shoulder arthroplasty yield similar self-assessed outcomes in the management of comparable patients with glenohumeral arthritis. J Shoulder Elb Surg. 2007;16:534-538.
- [Google Scholar]
- Outcomes of hemiarthroplasty and concentric glenoid reaming: a systematic review and meta-analysis of the ream and run technique. Semin Arthroplasty JSES. 2022;32:571-580.
- [Google Scholar]
- Self-assessed and radiographic outcomes of humeral head replacement with nonprosthetic glenoid arthroplasty. J Shoulder Elb Surg. 2015;24:1041-1048.
- [Google Scholar]
- Total shoulder arthroplasty in patients with a B2 glenoid addressed with corrective reaming: mean 8-year follow-up. J Shoulder Elb Surg. 2023;32:S8-S16.
- [Google Scholar]
- Early outcome of humeral head replacement with glenoid reaming arthroplasty (Ream and Run) for treatment of advanced glenohumeral osteoarthritis. J Shoulder Elb Surg. 2022;31:1846-1858.
- [Google Scholar]
- Clinical and radiographic results of eccentric glenoid reaming in anatomical total shoulder arthroplasty using 3D planning and PSI. Semin Arthroplasty JSES. 2021;31:587-595.
- [Google Scholar]
- Can the ream and run procedure improve glenohumeral relationships and function for shoulders with the arthritic triad? Clin Orthop. 2015;473:2088-2096.
- [Google Scholar]
- Clinical and radiographic outcomes of the ream-and-run procedure for primary glenohumeral arthritis. J Bone Joint Surg Am. 2017;99:1291-1304.
- [Google Scholar]
- Gender differences in glenoid anatomy: an anatomic study. Surg Radiol Anat SRA. 2009;31:183-189.
- [Google Scholar]
- Factors predictive of Cutibacterium periprosthetic shoulder infections: a retrospective study of 342 prosthetic revisions. J Shoulder Elb Surg. 2020;29:1177-1187.
- [Google Scholar]