INTRODUCTION

Fractures of the proximal humerus (FPH) are the third most frequent fractures; only surpassed by hip and distal radius fractures¹. They have an incidence of 4-6%, affecting more women than man. They mostly occur due to low-energy trauma and in the elderly people².

The type of treatment to be instituted depends on factors such as the patient's age, functional demand, dominance, comorbidities, type of fracture and its classification³.

Although most FPH are treated conservatively, it is estimated that up to 20% of them require surgical treatment. Due to the population aging, the number of FPH can increase and consequently the percentage that need a surgical intervention as well⁴.

The FDP treatments of 3 or 4 parts Neer is a challenge. Initially, hemiarthroplasty (HA) was recommended as the ideal treatment, due to the difficulty of closed reduction and the risk of avascular necrosis of the humeral head⁵. It was also recommended for fractures with a “head-split” component⁶.

However, complex FPH that affect the tuberosities are technically demanding to treat and their poor positioning is correlated with worse functional results, if the option is HA⁷.

Currently, inverse total shoulder arthroplasty (ITSA) has been shown to be effective in the treatment of 3 or 4 parts Neer’s fractures in elderly patients with rotator cuff arthropathy^8-11, with good functional results^12-14, and the latter depend less on the positioning of the tuberosities than on HA¹⁵.

The progressive increase in the number of these cases treated with arthroplasties can lead to long-term complications that require a surgeon with technical skills to resolve them. Revision surgery for an HA or ITSA also leads to less predictable functional results and higher complication rates^16-18.

Among the complications described, aseptic or septic unsealing, instability, wear of the glenoid component and incompetence/rupture of the rotator cuffs are highlighted^19-21.

The development of modular implants allows for greater ease in revision surgeries, with the possibility of preserving the humeral component and converting from an HA to an ITSA²². The time elapsed between the fracture and the surgery is considered a factor that affects the final result in HA.

This study aims to retrospectively identify, in people undergoing total shoulder arthroplasty for fractures of the proximal humerus, the type of arthroplasty used, the functional scores of the shoulder, the complications recorded, the influence of the time elapsed since the fracture and the placement of the implant in the final functional result.

METHOD

Retrospective consultation study of clinical files in which patients undergoing HA and ITSA due to FPH were included, in the period from January 1, 2014 to March 31, 2017, in an Orthopedics and Traumatology Service.

Electronic clinical files were consulted, and demographic characteristics (age and gender), type of FPH (Neer classification), time elapsed between fracture diagnosis and initial surgery for HA or ITSA, need for cementation and prosthetic modularity were recorded, need for revision, complications, follow-up time and participation in a regular rehabilitation program.

The functional results of each patient were collected from two previously applied instruments:

- Constant Shoulder Score (CSS)²³, adapted for Portugal²⁴, is a 100-point scale that is divided into four subscales - pain, activities of daily living, strength and joint range of motion.

- American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form (ASES)²⁵, Portuguese version²⁵, being divided into two sections, one clinical and one self-administered. This second section presents a 100-point scale consisting of two dimensions, a pain subscale worth 50 points, and ten daily life activities items worth the remaining 50 points.

Ethical and deontological procedures were complied with and the best interests of patients, namely their anonymity were safeguarded.

Descriptive and inferential statistical analysis of the data was performed using the IBM SPSS statistics software, version 23. The results are presented in absolute value and as a percentage when it is justified to facilitate their interpretation. Mean values ​​are shown followed by the standard deviation (mean±standard deviation). For the inferential analysis, the non-parametric Mann-Whitney U test was used to compare the groups operated early and late and the groups submitted to HA vs. ITSA. The assumed p value for this study was p ≤ 0.05.

RESULTS

During the time covered by our study, 15 shoulder arthroplasties were performed by FPH in 12 women and 3 men. The mean age of the sample was 78 years-old. (Table 1)

We recorded 6 Neer 2-part FPH cases, 2 Neer 3-part FPH cases and 7 Neer 4-part FPH cases. (Table 1)

The average time between fracture diagnosis and surgery was 29.4 days (Table 1), with 9 of the patients undergoing early surgery (up to 3 weeks after diagnosis). We registered 6 cases of late surgery (3 weeks after diagnosis).

10 ITSA were performed, with the particularity of 3 of them as salvage treatment of failure of conservative treatment (two cases of 2-part Neer FPH and one 3-part Neer FPH) and 2 of them as rescue treatment of initial treatment failure with osteosynthesis (both FPH 2 parts of Neer). 5 HA were recorded. (Table 2)

All procedures were performed by the same surgeon.

Except for 2 cases of ITSA, all implants were cemented and all placed prostheses were modular. (Table 2)

Regarding the functionality of the groups, for the group of patients undergoing HA, the mean score was 53.2% and 44.6, respectively for CSS and ASES; in the group of patients undergoing ITSA, the mean scores were 41.1% and 68.5, respectively for CSS and ASES (Table 3).

Results with statistical significance were not obtained in the comparative study of the mean values ​​of CSS and ASES between the groups which underwent early and late surgeries (Graphics 1 and 2).

In the comparative study of functional scores between patients undergoing HA and ITSA, there was a statistically significant value (p<0.05) for ASES of patients with ITSA (Graphic 2).

The complications that were recorded were divided into a case of acute infection of an ITSA, a case of proximal migration of the HA’s nail requiring revision (with conversion to ITSA), a case of sub-acromial impingement of an HA (Table 3) needing revision with

conversion into arthroscopic shoulder surgery for treatment of rotator cuff tear (ATRO). In revision surgeries, no complications were recorded.

Graphic 1 Relation between the time since fracture diagnosis and surgery and the CSS (Constant Shoulder Score) functional score 

Graphic 2 Relation between time since fracture diagnosis and surgery and ASES functional score 

DISCUSSION

Generally speaking, most FPH can be treated conservatively. On the other hand, FPH with surgery indication are typical of the elderly who, due to osteoporosis inherent to their age, increase the complexity of these fractures^16,26,27. In fact, a study recently published by Jung et al.²⁸ reported a rate of refraction of more than 18% in elderly people with osteoporosis. The mean age of patients treated in the present study was 78 years-old, which reflects the aging population of the geographic covered area and which is corroborated by several review articles published since 2015^29-32.

The currently increasing literature recommends ITSA as a first-line treatment in complex FPH in the elderly people with surgical criteria, when osteosynthesis is not indicated^8-11. ITSA is also increasingly recommended as a rescue treatment when primary treatments have failed^19-21, recognizing the importance of prosthetic modularity for revision surgeries²².

The complications recorded were those foreseen in any shoulder arthroplasty, however complications of vascular-nervous involvement were not recorded¹⁶.

In the comparative study of functional outcomes in patients undergoing HA versus ITSA, the best score was recorded for ITSA, but only with statistical significance for the ASES score (p<0.05). These data are in agreement with the literature, which states that ITSA allows better functional results when compared to HA^8-11. The results obtained by the ASES assessment are in line with the results found by other studies, namely 64.14 points described in the work published by Horneff et al.³³, 66 points in the work developed by Wagner et al.³⁴, 59 points in the article published in 2017 Holschen et al. and 65.3 points in another article published by the same team^35,36. Regarding the CSS assessment, the results found range from 57 to 45% in the study developed by Lignel et al.³⁷, 63% in the article published by Holschen et al., 52.9% in the publication by Giardella et al.³⁸ to 73% in the study carried out by Schliemann et al.³⁹. The present study was not able to assess the influence of the time interval from FPH diagnosis to surgery on functional outcomes. The comparative study of functional scores between early and late operated patients was not statistically significant (p>0.05). The fact that the sample in this study was relatively small and that 5 of the cases were implants performed as rescue treatment may have negatively influenced this comparative study.

CONCLUSION

The present study concludes that FPH with surgical criteria are typical of an aging population with poor bone stock.

For complex FPH and in the elderly without criteria for osteosynthesis, this study concludes that ITSA seems to be the best option as a first-line treatment in FPH with surgical criteria and it allows the patient to have better postoperative functionality, with regard to the ASES score.

It can also be concluded that prosthetic modularity is an asset in revision surgeries, making this procedure intrinsically more demanding, safer and simpler.

Therefore, despite a short experience and a small sample of cases, we can deduce that there are good practices regarding the treatment offered and that the most current international recommendations are followed.

In terms of the limitations of the present study, the small sample size, the relatively short follow-up time, the different contexts and experiences of the rehabilitation team and the type of prosthesis used, which was not homogeneous, can be indicated.

For the future, it is suggested to extend the follow-up time, create a specific intervention and rehabilitation team and improve the implant model.