Translate this page into:
CT versus MRI planning for reverse geometry total shoulder arthroplasty
∗Corresponding author: Shane P. Russell. abscessmicrobiology@gmail.com
-
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
Preoperative planning for Reverse Total Shoulder Arthroplasty (RTSA) using CT or MRI is well described.1-7 We aimed to compare pre-operative CT versus MRI measurement accuracy for predicting intra-operative glenoid implant sizing.
All patients with a preoperative CT or MRI undergoing RTSA at our tertiary referral center from October 2017 to February 2020 were included. Data was collected from theatre and implant registers. Glenosphere Width (GW) and Baseplate Central Screw Length (BCSL) were independently predicted from pre-operative CT or MRI imaging by 2 blinded senior authors. A sub-group analysis was also performed between trauma and non-trauma CT cases. SPSS v26 was used for statistical comparison between predicted and actual implants.
71 data sets from 69 patients were included for analysis: 31 CT predictions and 40 MRI predictions. 61.3% of CT measured GW predictions were accurate compared to 82.5% of MRI predictions (p = 0.045). BCSL predictions were 77.4% and 70% accurate for CT and MRI respectively, without significant difference. There was no significant difference in sub-group analysis for trauma vs elective CT accuracy of BCSL or GW measurements.
MRI imaging may be superior to CT for predicting GW and no less accurate than CT for predicting BCSL in the elective setting. No difference in CT measurement accuracy was seen between trauma and elective settings. While simultaneously clearly defining shoulder soft tissue anatomy, MRI may also be the preferred modality for bony measurements during pre-operative planning for elective RTSA.
Keywords
Pre-operative imaging
Reverse total shoulder arthroplasty
Pre-operative predictions
Proximal humeral fracture
Shoulder arthroplasty
1 Introduction
Reverse Total Shoulder Arthroplasty (RTSA) has seen a surge in utilization in recent years and since 2014 surpasses Anatomic Total Shoulder Arthroplasty (ATSA) use in the United States.1–3 Improvements in functional outcomes have been shown for both elective indications (such as rotator cuff arthropathy, glenohumeral arthritis with severe glenoid bone loss, chronic shoulder dislocations, and rheumatoid arthritis) and traumatic indications (such as certain proximal humerus fracture types and massive rotator cuff tears).4 RTSA is increasingly the revision of choice for failed ATSA or failed hemiarthroplasties.5–10
Due to the limited bone stock in the shoulder, particularly in cases with glenoid erosion, pre-operative 3-dimensional radiological planning is a useful adjunct. While much weight has been given to the importance of implant positioning and 3D assisted navigations, to our knowledge no methods have been described for comparison between pre-operative implant sizing modalities.7,10–13
Complications such as scapular notching, glenoid loosening or failure, and prosthetic instability have been well described. Studies have also shown the importance of pre-operative soft tissue and bony orientation in aiding accurate intra-operative component positioning.4,11,14–16 Determining the superior imaging modality to assess hard and soft tissues should help improve procedure efficiency and potentially improve both patient function and implant longevity.
The aim of this study was to compare the accuracy of pre-operative shoulder CT and MRI imaging for measuring the relevant bony shoulder anatomy and therefore predicting implant dimensions for intra-op use.
2 Methods
Using a retrospective cohort design, all patient who underwent RTSA at our Major Trauma Center and elective orthopaedic tertiary referral center between October 2017 and February 2020 were identified from theatre registries. For both trauma and elective indications, the Univers Revers™ Shoulder System (Arthrex GmbH) via a standard deltopectoral approach was used in all cases. Direct intra-operative measurements were used by the senior author (DM) in all cases when choosing implant sizes.
We excluded patients who had neither a pre-operative MRI nor CT, cases where an alternative implant system was utilized and cases where the bony anatomy was deemed radiologically immeasurable by the senior authors, such as severely comminuted or displaced proximal humerus fractures.
Patient demographics and implant details were harvested from patient medical records. All patients were then assigned to a CT or an MRI sub-cohort.
Two senior authors were assigned to blinded CT and MRI pre-operative imaging. Multiple orthogonal views were used to determine the optimal implant location before predictive measurements were made. Predicted Glenosphere Width (GW) and Baseplate Central Screw Length (BCSL) were recorded. Predicted GW was recorded as the widest humeral head diameter on coronal view, demonstrated in Fig. 1 and Fig. 2. Predicted BCSL values began with the perpendicular distance from the center of the glenoid fossa (confirmed on orthogonal views) to the medial border of the glenoid cavity, as depicted in Figs. 3 and 4. Due to a 10 mm baseplate recession for the central screw, 10 mm was subtracted from the measurement for the final BCSL measurement (see Appendix 1 for derivation of the BCSL measurement) (see Fig. 5).
Univers Revers™ Shoulder System provides for a selection from 3 glenosphere implant diameters as well as 3 baseplate central screws lengths, summarized in Table 1. The preoperative measurements were then sorted into the nearest dimension category and this was used as the preoperative prediction. In all cases, the final prosthetic dimensions were compared to preoperative predictions. Correct and incorrect predictions were then compared for the two cohorts (CT vs MRI).
| Glenosphere Width (mm) | Baseplate Central Screw Length (mm) |
| 36 | 15 |
| 39 | 20 |
| 42 | 25 |
SPSS statistical analysis using a chi-squared test was completed to compare image modality (CT or MRI) with predictions made for GW and BCSL separately.
A sub-group analyses was also performed to examine for a significant difference between trauma and elective CT prediction accuracy.
3 Results
3 patients were excluded as neither a CT nor MRI was performed pre-operatively, whilst 7 cases were excluded due to destroyed joint anatomy precluding measurement with any reasonable accuracy. 71 data sets from 69 patients were included for analysis. 31 patients had a pre-operative CT for analysis and 40 had an MRI. In two cases both MRI and CT were used pre-operatively. In these cases, data from each imaging modality was analyzed in the respective group. The average patient age was 74.7 (Table 2).
| Age | Shoulder | Modality | |||||||
| Female | Male | <66 | 66–75 | >75 | Left | Right | CT | MRI | |
| Count | 55 | 14 | 6 | 34 | 29 | 23 | 46 | 31 | 40 |
| Average | 79.7% | 20.3% | 8.7% | 49.3% | 42.0% | 33.3% | 66.7% | 43.7% | 56.3% |
| Average Age: 74.7 | |||||||||
GW was accurately predicted in 61.3% (19 of 31) of pre-operative CT measurements and 82.5% (33 of 40) of MRI measurements (Table 3), which was statistically significant (p = 0.045).
| Modality | Total | ||||
| CT | MRI | ||||
| Glenosphere Prediction | Correct | Count | 19 | 33 | 52 |
| % within Modality | 61.3% | 82.5% | 73.2% | ||
| Incorrect | Count | 12 | 7 | 19 | |
| % within Modality | 38.7% | 17.5% | 26.8% | ||
| Total | Count | 31 | 40 | 71 | |
| % within Modality | 100.0% | 100.0% | 100.0% | ||
BCSL was accurately predicted in 77.4% (24 of 31) of pre-operative CT measurements whereas 70% (28 of 40) of MRI cases were predicted accurately (Table 4). No statistically significant difference was found between CT and MRI for predicting BSCL (p = 0.484).
| Modality | Total | ||||
| CT | MRI | ||||
| Baseplate Central Screw Length Prediction | Correct | Count | 24 | 28 | 52 |
| % within Modality | 77.4% | 70.0% | 73.2% | ||
| Incorrect | Count | 7 | 12 | 19 | |
| % within Modality | 22.6% | 30.0% | 26.8% | ||
| Total | Count | 31 | 40 | 71 | |
| % within Modality | 100.0% | 100.0% | 100.0% | ||
In addition, when a sub-group analysis was performed, no significant difference was found between trauma and elective CT measurement accuracy for either BCSL (p = 0.23) or GW (p = 0.379).
It was not possible to make a statistical comparison between trauma and elective MRI measurement accuracy as just one of 16 trauma patients had a pre-operative MRI performed.
4 Discussion/Conclusion
In analysis of GW prediction, MRI scans were found to be superior to CT. When considering the BCSL though, there was no statistically significant different found between the two. Consequently, MRI was found overall to be superior to CT for correctly predicting Univers Revers™ Shoulder System glenoid implant sizes in the preoperative setting.
This study has several limitations. MRI was not found to be superior in trauma due to an insufficient number of patients undergoing MRI prior to RTSA for traumatic indications. Cost and availability of MRI may prohibit its use in some centers, especially in the trauma setting where time to surgery is critical and trauma center resources are very often pressured. However, when the intrinsic benefits of MRI over CT such as absence of radiation exposure and vastly superior soft tissue detail are considered in combination with superior implant predictability, MRI may be the preoperative assessment tool of choice for elective RTSA.
Improved pre-operative implant predictability may also reduce intra-operative variability and therefore improve theatre efficiency, reduce operating time and ultimately reduce costs.
While our study focused on a single prosthesis system, the comparison between preoperative imaging modalities for alternative implants may also be of interest.
T our knowledge, this is the first study to describe the direct comparison of preoperative shoulder imaging for predicting intra-operative implant selection. However, we identified various studies where the use of pre-operative imaging for predicting glenoid bone stock, glenoid version or optimal glenoid implant placement was examined.7,17–21
We were also unable to identify any validated method for pre-operative glenoid measurement. King JJ et al. describe how height alone is a poor predictor.22 The glenoid prediction methods used in this study have not previously been described. A follow-up study comparing pre-operative measurements to intra-operative anatomical measurements is underway at our center with the aim of validating a preoperative measuring technique.
For our study, the indication for surgery inspired the preferred imaging modality, thereby risking a selection bias. The majority of patients who underwent RTSA for trauma underwent preoperative CT, whereas those with elective indications underwent MRI. This was foreseeable not only due to the relative ease of access to CT at short notice but also its established ability to well describe fracture configurations for operative planning. Conversely, rotator cuff pathologies were well described with pre-operative MRI in the elective setting. Further studies may have the ability to match these two cohorts for degree of anatomical preservation or determine if the indication for surgery impacts prosthesis predictions.
However, having performed a sub-group analysis to test for a difference between trauma and elective measurement accuracy, it is reassuring to note no significant difference was found between trauma and elective CT accuracy.
Intra-operatively, the glenoid surface is debrided and reamed in preparation for glenosphere implantation. The volume of glenoid bone stock resected influences BCSL. Variability in surgical technique may therefore lead to varying degrees of glenosphere medialization and BCSL shortening. For our study, this was controlled for by a single surgeon performing all procedures using a standard technique to minimize variation.
Interobserver variability was not measurable for this study. However, all observers were blinded to implant details during the measurement process. Further work to assess intra and inter-observer reliability will help validate our novel GW and BCSL measurement method.
In conclusion, we determine that pre-operative MRI is an accurate predictor of correct implant size during RTSA. We would favour a pre-operative MRI in the elective settings for RTSA.
References
- Current trends in the use of shoulder arthroplasty in the United States. Orthopedics. 2018;41(3):e416-e423.
- [Google Scholar]
- Reverse shoulder arthroplasty: the Singapore General Hospital experience and a simple method of measuring change in the center-of-rotation. J Orthop. 2015;12(2):97-101.
- [Google Scholar]
- Changing trends in the management of proximal humerus fractures in New York State. J Orthop. 2020;21:127-130.
- [Google Scholar]
- Clinical implications of scapular notching at 2 and 5-year follow-up after reverse total shoulder arthroplasty. J Orthop. 2020;21:384-389.
- [Google Scholar]
- Reverse total shoulder arthroplasty for primary glenohumeral osteoarthritis in patients with a biconcave glenoid. J Bone Joint Surg Am. 2013;95(14):1297-1304.
- [Google Scholar]
- Effects of acquired glenoid bone defects on surgical technique and clinical outcomes in reverse shoulder arthroplasty. J Bone Joint Surg Am. 2010;92(5):1144-1154.
- [Google Scholar]
- Pre-operative planning for reverse shoulder replacement: the surgical benefits and their clinical translation. Annals of Joint. 2019;4
- [Google Scholar]
- Reverse total shoulder arthroplasty for treatment of 3- and 4-Part Proximal humeral fractures: clinical and radiological analysis with minimum follow-up of 2 years. Geriatr Orthop Surg Rehabil. 2020;11
- [Google Scholar]
- Indications for reverse total shoulder arthroplasty in rotator cuff disease. Clin Orthop Relat Res. 2010;468(6):1526-1533.
- [Google Scholar]
- Preoperative planning for accurate glenoid component positioning in reverse shoulder arthroplasty. Orthop Traumatol Surg Res. 2017;103(3):407-413.
- [Google Scholar]
- Preoperative planning for accurate glenoid component positioning in reverse shoulder arthroplasty. J Orthop Traumatol: Surgery & Research. 2017;103(3):407-413.
- [Google Scholar]
- Predictors of scapular notching in patients managed with the Delta III reverse total shoulder replacement. J Bone Joint Surg Am. 2007;89(3):588-600.
- [Google Scholar]
- Biomechanical relevance of glenoid component positioning in the reverse Delta III total shoulder prosthesis. J Shoulder Elbow Surg. 2005;14(5):524-528.
- [Google Scholar]
- Scapular notching in reverse shoulder arthroplasty. J Shoulder Elbow Surg. 2008;17(6):925-935.
- [Google Scholar]
- Early dislocation after reverse total shoulder arthroplasty. J Shoulder Elbow Surg. 2014;23(5):737-744.
- [Google Scholar]
- The effect of metaglene lateralization on joint mobility of reverse shoulder arthroplasty: a cadaveric biomechanical study. J Orthop. 2021;24:9-14.
- [Google Scholar]
- Accuracy of the subchondral smile and surface referencing techniques in reverse shoulder arthroplasty. Orthopedics. 2016;39(4):e615-e620.
- [Google Scholar]
- Three-dimensional preoperative planning software and a novel information transfer technology improve glenoid component positioning. J Bone Joint Surg Am. 2014;96(9):e71.
- [Google Scholar]
- Reverse shoulder arthroplasty - a literature review. Open Orthop J. 2013;7:366-372.
- [Google Scholar]
- Comparison of standard two-dimensional and three-dimensional corrected glenoid version measurements. J Shoulder Elbow Surg. 2011;20(4):577-583.
- [Google Scholar]
- Radiographic geometry and clinical glenohumeral range of motion after reverse shoulder athroplasty, a retrospective cohort study. J Orthop. 2021;25:283-287.
- [Google Scholar]
- Optimal glenosphere size cannot be determined by patient height. J Shoulder Elbow Surg. 2020;29(2):258-265.
- [Google Scholar]