Treatment of Periprosthetic Fractures

Leesa M. Galatz, MD

Alexander J. Vervaecke, MD

Brad Parsons, MD

INTRODUCTION

Periprosthetic shoulder fractures are potentially devastating complications and can present substantial challenges for the treating orthopedic surgeon. Currently considered relatively rare, the increasing use of shoulder arthroplasty in general will ensure a similar rise in the absolute number of complications, making them more common.¹ The periprosthetic fractures may be anatomically located at the humerus or scapula (glenoid, acromion, or coracoid), and a clear distinction is made between an intra- and postoperative occurrence. The scope of this chapter will mainly be focused on periprosthetic humerus fractures and to a lesser extent on periprosthetic glenoid fractures. While the incidences vary widely in the literature, intraoperative humerus fractures are most frequently reported and occur in 1.2% to 6.5% of primary shoulder arthroplasties. Postoperative humerus fractures most often originate from traumatic events, and an incidence between 0.2% and 12.9% is reported.² The majority of modifiable risk factors for fractures are directly related to technical errors during surgery such as poor exposure, endosteal notching, and cortical perforation and can therefore potentially be avoided. While multiple classifications systems are present, they are mostly based upon small unvalidated series that somewhat limits their transferability to relevant treatment algorithms. Higher nonunion rates are reported for periprosthetic fractures in comparison to native fractures, and most individuals undergoing shoulder arthroplasty tend to belong to an older and more osteopenic population. Considering the already limited bone stock of the proximal humerus and glenoid vault, these factors further increase treatment difficulties. Intraoperative fractures should be stabilized at the time of the index procedure. Postoperative fractures may require revision surgery but can often be treated nonoperatively. Two important factors have to be assessed when determining optimal treatment: (1) implant stability and (2) fracture stability. Generally, with a stable implant and a stable fracture, nonoperative treatment can be utilized. Unstable fractures at the tip or below the stem will often need to be reduced and fixated as typical diaphyseal humerus fractures. Unstable implants in combination with periprosthetic fractures warrant implant revision and fracture fixation.

INTRAOPERATIVE INCIDENCE

Fractures are the most frequent intraoperative complications during primary and revision shoulder arthroplasties.³,⁴ A single-center retrospective analysis of 2588 total shoulder arthroplasties (TSAs) and 1431 hemiarthroplasties (HAs) performed over 32 years reported intraoperative fractures in 47 (40 humerus, 5 glenoid, 2 unspecified) and 15 (8 humerus, 7 glenoid) cases, respectively.⁵ Furthermore, an analysis of the National Joint Registry of England, Wales, Northern Ireland and the Isle of Man (NJR) between 2011 and 2015 reported 315 intraoperative complications in 12,559 primary shoulder arthroplasties, of which the majority were humeral fractures (110 humerus fractures, 87 glenoid fractures).³ While these results are consistent with other published reports, the incidence is most likely underestimated due to the inherent limitation of database studies that rely on surgeons reporting their complications. Higher incidences are reported in revision arthroplasty procedures. Athwal et al retrospectively analyzed all primary and revision shoulder arthroplasties performed in their center between 1980 and 2002 and found that fractures were more than twice as likely to occur in revision surgery (14 of 422, 3.3%) compared to those in primary arthroplasty (31 of 2666, 1.2%).⁶ Additionally, Wagner et al evaluated the medical records of 230 revision procedures performed between 2005 and 2012 and found an incidence of 15.7%.⁷ Utilizing the same aforementioned NJR database, Ingoe et al further confirmed this increased risk as fractures occurred in 50 revision cases.⁴ Summarized, the intraoperative incidence varies between 1.2% and 15.7% for both primary and revision shoulder arthroplasty (TABLE 40.1).

POSTOPERATIVE INCIDENCE

The incidence of postoperative periprosthetic humeral fractures as reported in the literature is quite variable. Boyd et al identified seven cases with postoperative periprosthetic humeral fractures between 1974 and 1988 resulting in an incidence of 1.6%.⁹ Similar results
were found by Wright et al (1.8%) and Worland et al (2.38%).¹⁰,¹¹ In the study by Singh et al, 43 (36 humerus, 5 glenoid, 2 unspecified) postoperative humeral fractures occurred in 4019 primary shoulder arthroplasties (0.9%).⁵ Chin et al reported on only 1 postoperative fracture in 431 patients (0.2%) with a mean follow-up of 4.2 years, whereas Atoun et al reported four fractures in a series of 31 short-stemmed reverse total shoulder arthroplasty (RTSA) implants (12.9%).⁸,¹² In summary, the postoperative incidence varies between 0.2% and 12.9% (TABLE 40.2).

TABLE 40.1 Summary of Reported Intraoperative Incidences of Periprosthetic Fractures in Both Primary and Revision Shoulder Arthroplasty

Study	Procedure	No. of Athroplasties	Incidence	Anatomical Location
Study	Procedure	No. of Athroplasties	Incidence	*Tuberosities*	*Metaphysis*	*Shaft*	*Multi*
1. Athwal et al (2009)⁶	Primary	2666	31 (1.2%)	19	6	16	3
1. Athwal et al (2009)⁶	Revision	422	14 (3.3%)	19	6	16	1
2. Singh et al (2012)⁵	Primary	4019	62 (1.5%)	Unspecified
2. Singh et al (2012)⁵	ATSA HA	2588 1431	47 (1.8%) 15 (1.0%)	Unspecified
3. Atoun et al (2014)⁸	Primary	31	2 (6.5%)	–	2	–	–
4. Wagner et al (2015)⁷	Revision	230	36 (15.7%)	30	3	3	–
5. García-Fernández et al (2015)¹⁴	Primary	163	2 (1.2%)	–	2	1	–
5. García-Fernández et al (2015)¹⁴	Revision	40	1 (2.5%)	–	2	1	–
6. Ingoe et al (2017)⁴	Revision	1455	35 (2.4%)	Unspecified
7. Cowling et al (2017)³	Primary	12,599	202 (1.6%)	Unspecified
7. Cowling et al (2017)³	ATSA HA RTSA SL	3712 1928 4590 2329	52 (1.4%) 45 (2.3%) 103 (2.2%) 2 (0.1%)	Unspecified
Incidence is given as the total number with percentage between parentheses. ATSA, anatomic total shoulder arthroplasty; HA, hemiarthroplasty; RTSA, reverse total shoulder arthroplasty; SL, stemless hemiarthroplasty.

RISK FACTORS

Intraoperative Risk Factors

Considering that patient characteristics often cannot be altered or modified, identification of the primary risk factors associated with periprosthetic fractures allows the surgeon to apply extra care and consider alternative technical approaches in those patients at high risk. Established unmodifiable risk factors for intraoperative periprosthetic humerus fractures are female gender, a preoperative diagnosis of posttraumatic arthritis or osteonecrosis, and revision arthroplasty. Although no studies were able to directly attribute a higher incidence of intraoperative periprosthetic fractures to osteopenia and osteoporosis, these conditions will certainly be a factor in the higher relative risk in female patients. Modifiable risk factors are strongly related to the choice of implant, surgical technique, and potential technical errors. The use of press-fit humeral stems showed a relative risk of 2.9× greater than the use of cemented implants (TABLE 40.3).⁶ As the endosteal surface at the level of the diaphysis is asymmetrical and elliptical in cross sections, endosteal notching can result from using cylindrical intramedullary reamers and can potentially increase the risk of fractures when inserting the broach or final press-fit implant.¹⁵,¹⁶ Cementing the humeral component allows the use of a smaller size implant, and therefore over-reaming can be avoided, thereby reducing the risk of fracture. These results, however, are based on a historical cohort. Between 1980 and 2002, press-fit humeral stems relying on diaphyseal fixation predominated. The introduction of newer stems with metaphyseal fixation and “bony ingrowth” coating potentially mitigating the increased fracture risk associated with the use of press-fit stems.¹⁷ Werthel et al found no difference in both intra- and postoperative fractures between cemented and cementless stems in 4636 primary arthroplasties performed between 1970 up to 2012.¹⁸

One study reported a lower relative fracture risk with the superior approach compared to that with the deltopectoral approach.³ The authors hypothesized that this difference was more likely attributable
to surgeon differences and the rationale of using the superior approach in less-complicated cases than to the approach itself. Higher incidences of intraoperative fractures were also found for RTSA and stemmed HAs compared to those for anatomical and resurfacing arthroplasties. This trend correlated to a relative risk of 1.4 for RTSA and 1.9 for HA compared to TSA in a multivariable analysis adjusting for age, sex, indication for surgery, and American Society of Anesthesiologists grade. Taking into consideration that the former two implant types will more often be utilized in acute or trauma settings and for patients with a larger inherent risk for fractures, it remains difficult to directly attribute the higher risk solely to the implant type when different implants would have been selected based upon the operative indications.³ While revision arthroplasty itself is a significant risk factor for intraoperative periprosthetic humeral fractures, subgroup analysis of revision cases showed that female sex, history of prior instability, and prior HA are also risk factors with an
odds ratio of 2.41 (P = 0.03), 2.65 (P = 0.02), and 2.34 (P = 0.03), respectively.⁷ Interestingly, a history of a prior cemented primary arthroplasty did not significantly increase the risk of intraoperative fractures in this study.

TABLE 40.2 Summary of Reported Postoperative Incidences of Periprosthetic Fractures in Both Primary and Revision Shoulder Arthroplasty

Study	Procedure	No of Athroplasties	Incidence	Wright and Cofield
Study	Procedure	No of Athroplasties	Incidence	A	B	C
1. Boyd et al (1992)⁹	Primary	436	7 (1.6%)	Unspecified
2. Wright et al (1995)¹⁰	Primary	499	9 (1.8%)	1	5	3
3. Worland et al (1999)¹¹	Primary	256	6 (2.3%)	2	3	1
4. Kumar et al (2004)¹³	Primary	3091	19 (0.6%)	6	6	3
5. Chin et al (2006)¹²	Primary	431	1 (0.2%)	Unspecified
6. Singh et al (2012)5	Primary	4019	43 (1.1%)	20 shaft fractures further unspecified
6. Singh et al (2012)5	ATSA HA	2588 1431	23 (0.9%) 18 (1.3%)	15 shaft fractures further unspecified
7. Atoun et al (2014)⁸	Primary RTSA	31	4 (12.9%)	4 metaphyseal fractures
8. García-Fernández et al (2015)¹⁴	Primary RTSA	203	4 (2.0%)	1	3	–
Incidence is given as the total number with percentage between parentheses. ATSA, anatomic total shoulder arthroplasty; HA, hemiarthroplasty; RTSA, reverse total shoulder arthroplasty.