Effect of tuberosity repair on functional outcome of reverse shoulder arthroplasty in proximal humerus fractures

Abstract

Purpose

Unsatisfactory results of hemiarthroplasty in Neer's 3- and 4-part proximal humerus fractures in elderly, have led to the shift towards reverse shoulder arthroplasty (RSA). The objective of our study was to repair the tuberosities that are generally overlooked during RSA and observe its impact on the functional outcome and shoulder scores.

Methods

We include elderly patients with acutely displaced or dislocated 3- or 4-part proximal humerus fractures from July 2013 to November 2019 who were treated with RSA along with tuberosity repair by non-absorbable sutures and bone grafting harvested from the humeral head. Open injuries and cases with neuro-muscular involvement of the deltoid muscle were excluded. According to the tuberosity healing on radiographs of the shoulder at 9th postoperative month, the patients were divided into 2 groups, as the group with successful tuberosity repair and the other with failed tuberosity repair. Statistical analysis of the functional outcome and shoulder scores between the 2 groups were done by independent t-test for normally distributed parameters and Mann-Whitney test for the parameters, where data was not normally distributed.

Results

Of 41 patients, tuberosity healing was achieved in 28 (68.3%) and failed in 13 (31.7%) cases. Lysis of the tuberosity occurred in 5 patients, tuberosity displacement in 2, and nonunion in 2. Mean age was 70.4 years (range 65 – 79 years) and mean follow-up was 58.7 months (range 18 – 93 months). There were no major complications. Group with successful tuberosity repair showed improvement in mean active range of movements, like anterior elevation (165.1° ± 4.9° vs. 144.6° ± 9.4°, p < 0.000), lateral elevation (158.9° ± 7.2° vs. 138.4° ± 9.6°, p < 0.000), external rotation (30.5° ± 6.9° vs. 35.0° ± 6.3°, p = 0.367), internal rotation (33.7° ± 7.5° vs. 32.6° ± 6.9°, p = 0.671) and in mean shoulder scores including Constant score (70.7 ± 4.1 vs. 55.5 ± 5.7, p < 0.000), American shoulder and elbow surgeons score (90.3 ± 2.4 vs. 69.0 ± 5.7, p < 0.000), disability of arm shoulder and hand score (22.1 ± 2.3 vs. 37.6 ± 2.6, p < 0.000).

Conclusion

Successful repair and tuberosity healing around the RSA prosthesis is associated with statistically significant improvement in postoperative range of motion, strength and shoulder scores. Standardized repair technique and interposition of cancellous bone grafts, harvested from the humeral head can improve the rate of tuberosity healing.

Introduction

Proximal humeral fractures (PHF) are the second most common upper extremity fracture and the third most common non-vertebral osteoporotic fracture, which also account for 10% of all fractures in the age group of 65 years and above. Displaced 3- and 4-part PHF in elderly patients have been treated with hemiarthroplasty (HA) and fixation of tuberosity. However, the functional outcome with HA has largely remained unsatisfactory in various reports.^1,2 Failure of tuberosity healing, pre-existing osteoporosis, lack of compliance with rehabilitation regime lead to a poor functional outcome.³ Fracture comminution, split humeral head, associated dislocation, osteoporosis, osteonecrosis of the humeral head and post-traumatic arthritis present numerous challenges in this fracture cohort.⁴ Many previous studies have shown that the clinical outcomes after HA for PHF are disappointing, if the tuberosity fail to heal.^1,2 Similarly, complications like instability, infection and loosening have been reported in greater number when the tuberosity fails to heal.^3,5, 6, 7, 8, 9 On the contrary, compared to HA, reverse shoulder arthroplasty (RSA) offers an easier rehabilitation, and a fairly better range of movement (ROM) and subjective shoulder scores even in the absence of tuberosity healing, which is why it is favored by many orthopaedic surgeons.8, 9, 10, 11 However, with time, it was also observed that in revision surgeries for failed HA with lysis of tuberosity, the ROM of the shoulder was not very satisfactory even when shifted to RSA.^10,11 Currently, RSA is continued to be favored by many orthopaedic surgeons as a better alternative to HA, because it is believed that functional results following RSA are not solely dependent upon anatomic tuberosity healing.12, 13, 14 Moreover the tuberosity healing is also difficult to obtain in elderly patients due to their limited potential for bone union.

As far as the complications concerned, the major clinical and functional issues following treatment of comminuted PHF with RSA were restriction of internal or external rotation, forward flexion, overhead activities and scapular notching. Rotational movements are crucial for activities of daily living and personal hygiene. It may be safely said that reverse replacement without a successful repair of tuberosity is still better than HA without a successful repair. However, it remains a tempting concept whether we can go further in terms of the functional outcome with RSA by doing a successful tuberosity repair. Many papers have examined the overall outcome of RSA in PHF, but very few papers have investigated on this specific issue of the need of tuberosity healing and its impact on functional outcome of RSA. The present study aims to find the efficacy of meticulous anatomic tuberosity repair in RSA for displaced PHF.

Methods

An interventional study was undertaken in our institution from July 2013 to November 2019. Approval of our institutional ethical committee was obtained beforehand. Inclusion criteria were: displaced 3- or 4-part PHF (Neer's classification) less than 2-week-old duration, age ≥ 65 years, and treated with RSA along with tuberosity repair. Open injuries and cases with neuro-muscular involvement of the deltoid muscle were excluded from the study.

All the patients were operated in beach chair position, and the prosthesis (DePuy, DELTA XTEND™) was implanted through the deltopectoral approach. The tuberosities were carefully retracted, and the shoulder joint was opened through the fracture at the level of rotator interval to retrieve the humeral head and expose the glenoid. The glenoid was reamed with cannulated reamer over a guidewire with 10° inferior tilt. Glenoid base plate was impacted as low as possible and secured in place with 3–4 screws. Two 2 mm holes were drilled through the humeral shaft on either side of the bicipital groove, and then heavy non-absorbable sutures were passed through each hole. One set with 3 heavy non-absorbable sutures was passed through the bone tendon junction of infraspinatus and looped across the medial suture hole of the prosthetic, and the other set was passed through the subscapularis tendon (Fig. 1A). The greater tuberosity (GT) fragment was reduced onto the lateral aspect of humeral shaft with slight overlap for better healing. The prosthetic height was determined according to the GT reduction, of which the ideal height was obtained when the tip of the tuberosity was 5–10 mm below the summit of the prosthesis. Prosthesis was distally cemented into the diaphysis at about 30° of retroversion. The bone cement is kept below the level of the proximal metaphysis, and therefore does not expose to the tuberosities. The thin cancellous bone grafts harvested from the fractured humeral head were placed between the tuberosities and humeral shaft. Three sets of horizontal sutures were tied embracing the tuberosities around the prosthesis (Fig. 1B). Then the prepared sutures from the humeral shaft were passed through infraspinatus and subscapularis, and then were tied to create a tension band construct for preventing superior migration of the tuberosities.

Fig. 1.

Technique of tuberosity repair followed in this study. (A) A set with 3 heavy non-absorbable sutures were passed through bone tendon junction of infraspinatus and supraspinatus, and another set through subscapularis. (B) Sutures were tied embracing tuberosities around the prosthesis.

Physiotherapy and rehabilitation regime was the same for all the patients. Postoperatively, shoulder abduction orthosis was given for 2 weeks, static exercises started from day 4, and pendulum exercises and passive ROM exercises from day 7, which were supervised by a single team of physiotherapists. Passive stretching and active assisted movements began at the 4th week. Strengthening started at the 12th week postoperatively after radiological healing of the tuberosities. Patients were evaluated every 3 months for 1 year, then for 6 months. Active anterior elevation (AAE), active lateral elevation (ALE), rotational movements and functional scores i.e., Constant and Murley score,¹⁵ American shoulder and elbow surgeons (ASES) score, disability of arm shoulder and hand (DASH) score were recorded at each visit.

Routine antero-posterior (AP), axial and AP in internal rotation postoperative radiographs were taken. GT was considered to be consolidated when it was visualized on AP view in neutral rotation and united with the humeral shaft.¹ Similarly lesser tuberosity union was assessed on AP view in 60° internal rotation. Inferior scapular notching (Nérot-Sirveaux system),¹⁰ heterotopic ossification, and radiolucent lines around the components were searched for in the radiographs.

Statistical analysis

Based on the tuberosity healing at 9th postoperative month, the patients were divided into 2 groups, as the group with successful tuberosity repair and the group with failed tuberosity repair. Statistical analysis of the functional outcome and shoulder scores between the 2 groups were done by independent t-test for normally distributed parameters and Mann-Whitney test for the parameters, where the data was not normally distributed. Shapiro-Wilk and Kolmogorov-Smirnov tests were used to determine the data distribution. SPSS software, version 25 was used and significance level was set at p < 0.05.

Results

Patient population

Forty-four patients, aged ≥ 65 years were included. Out of these patients, 1 died before 2-year follow-up and 2 did not turn up for further evaluation. Statistical analysis of 41 cases (41 shoulders) was achieved, who were evaluated with a minimum of 18 months of follow-up. Mean age at the time of surgery was 70.4 years (range 65 – 79 years). Mean follow-up was 58.7 months (range 18 – 93 months) with 14 women (34.1%) and 27 men (65.9%). Average injury to surgery interval was 11 days (range 3 – 19 days). Dominant side was involved in 27 shoulders (65.9%).

Tuberosity healing

Among them, 28 out of 41 cases (68.3%) had successful GT healing (Fig. 2). Five patients had either partial or total tuberosity lysis, 2 had tuberosity migration, and 2 had nonunion (Figs. 3A–C). The comparison between the 2 groups is presented in Table 1.

Fig. 2.

Healed tuberosity of antero-posterior and axial view radiograph at 5.5 years postoperatively in the same patient.

Fig. 3.

A few cases showing failure of tuberosity healing. (A) Near complete lysis of GT at 7 months postoperatively; (B) near complete lysis of GT at 1 year postoperatively and (C) partial lysis of GT at 9 months follow-up.

Table 1.

The comparison between the 2 groups.

Variables	Successful tuberosities repair (n = 28)	Failed tuberosities repair (n = 13)
Age (year)	69.3 ± 3.9 (61 – 77)	72.6 ± 3.3 (68 – 79)
BMI	24.5 ± 2.8 (18.4 – 31.8)	23.6 ± 4.1 (19.2 – 31.6)
Injury-surgery interval (day)	9.2 ± 3.4 (3 – 18)	12.8 ± 3.6 (8 – 19)
Female	8 (28.5)	6 (46.1)
Diabetes mellitus	7 (25.0)	4 (30.7)
Dislocation	4 (14.2)	3 (23.1)
3-part fractures	17 (60.7)	5 (38.4)
4-part fractures	11 (39.3)	8 (61.5)
Dominant side	21 (75.0)	6 (46.1)

Note: Data presented as mean ± SD (range) or n (%).

BMI: body mass index.

Functional results

Group with successful tuberosities repair showed improved mean active ROMs, including anterior elevation (165.1° ± 4.9° vs. 144.6° ± 9.4°, p < 0.000), lateral elevation (158.9° ± 7.2° vs. 138.4° ± 9.6°, p < 0.000), external rotation (30.5° ± 6.9° vs. 35.0° ± 6.3°, p = 0.367) and internal rotation (33.7° ± 7.5° vs. 32.6° ± 6.9°, p = 0.671) (Fig. 2, Fig. 4). Details of the results are shown in Table 2, Table 3.

Fig. 4.

Abduction and internal rotation movements at 5.5 years postoperatively in a patient with healed tuberosities of the right shoulder.

Table 2.

Cases with successful tuberosity repair.

Case No.	AAE	ALE	AER	AIR	CSP	CSA	CSM	CSS	CS	ASES	DASH	Age (year)
1	170	170	25	35	15	6	34	14	75	93.3	19.2	71
2	170	150	45	40	15	2	30	14	62	88.3	20.8	74
3	165	155	30	25	15	6	38	11	72	85.0	26.0	65
4	155	160	45	45	15	2	30	8	64	88.3	20.4	67
5	170	170	40	35	15	6	34	14	68	89.9	22.0	69
6	160	155	40	35	15	6	32	11	68	88.3	25.0	77
7	170	160	30	30	15	2	34	8	74	89.9	26.0	65
8	170	165	30	25	15	6	30	8	72	91.6	22.0	68
9	165	160	25	30	15	6	32	11	68	91.6	24.0	76
10	165	155	30	45	15	6	34	14	70	88.3	23.0	72
11	170	160	30	50	15	6	34	14	77	93.3	19.0	69
12	165	160	25	40	15	6	34	11	74	89.9	20.0	67
13	170	160	30	30	15	6	32	11	71	93.3	22.0	70
14	165	150	20	35	15	2	28	14	69	91.6	25.0	71
15	160	140	30	40	15	6	30	11	74	89.9	22.5	65
16	170	160	40	25	15	6	36	8	76	91.6	21.5	77
17	165	160	30	30	15	6	32	14	77	86.6	23.0	71
18	170	155	20	20	15	6	32	8	71	94.9	20.0	67
19	170	160	30	35	15	6	34	11	74	93.3	19.2	65
20	165	145	20	40	15	6	30	11	72	91.6	19.2	73
21	160	160	25	30	15	6	30	8	62	91.6	20.8	69
22	155	165	30	35	15	6	30	11	68	86.6	26.0	74
23	165	155	25	30	15	2	32	14	70	88.9	24.0	66
24	165	165	40	40	15	6	34	11	71	88.3	22.5	67
25	155	170	35	35	15	6	28	14	69	93.3	25.0	65
26	165	155	30	40	15	6	32	8	68	87.9	19.2	68
27	170	160	25	20	15	6	34	11	77	91.6	20.0	71
28	160	170	30	25	15	2	32	14	68	92.3	23.0	68
mean	165.1	158.9	30.5	33.7	15	5.1	32.2	11.3	70.7	90.3	22.1	69.3

AAE: active anterior elevation, ALE: active lateral elevation, AER: active external rotation, AIR: active internal rotation, CSP: Constant score pain, CSA: Constant score activity, CSM: Constant score mobility, CSS: Constant score strength, CS: Constant score total, ASES: American shoulder and elbow surgeons score, DASH: disabilities of the arm, shoulder and hand score.

Table 3.

Cases with failed tuberosity repair.

Case No.	AAE	ALE	AER	AIR	CSP	CSA	CSM	CSS	CS	ASES	DASH	Age (year)
1	135	130	40	25	10	2	26	8	52	68.3	35.0	75
2	130	135	25	30	10	2	24	5	49	56.6	40.0	79
3	135	130	30	25	15	2	24	8	55	60.0	38.0	68
4	155	150	25	45	15	2	28	5	54	66.6	36.0	69
5	155	120	25	40	10	2	28	5	48	69.9	35.0	73
6	135	135	25	30	15	2	26	5	54	71.6	42.0	74
7	145	135	30	40	10	2	26	5	52	71.3	38.0	75
8	150	145	35	25	10	2	28	8	68	68.6	35.0	68
9	135	140	40	30	15	2	26	5	64	72.4	35.0	73
10	145	155	40	35	10	2	24	5	54	67.7	40.0	76
11	155	150	35	35	15	2	28	8	62	76.8	42.0	71
12	150	135	35	40	10	2	26	8	54	77.6	35.0	74
13	155	140	40	25	10	2	26	5	56	69.7	38.0	70
mean	144.6	138.4	35	32.6	11.9	2	26.1	6.1	55.5	69.0	37.6	72.6

AAE: active anterior elevation, ALE: active lateral elevation, AER: active external rotation, AIR: active internal rotation, CSP: Constant score pain, CSA: Constant score activity, CSM: Constant score mobility, CSS: Constant score strength, CS: Constant score total, ASES: American Shoulder and Elbow Surgeons score, DASH: disabilities of the arm, shoulder and hand score.

Shoulder scores

There was statistically significant improvement across all mean shoulder scores in the group of tuberosity repair in comparison to the group of tuberosity not healing i.e., Constant score (70.7 ± 4.1 vs. 55.5 ± 5.7, p < 0.000), ASES score (90.3 ± 2.4 vs. 69.0 ± 5.7, p < 0.000), DASH score (22.1 ± 2.3 vs. 37.6 ± 2.6, p < 0.000) (Table 4). The difference in functional outcome and shoulder scores between the 2 groups was statistically significant (Table 4).

Table 4.

Comparison of functional outcome between the 2 groups.

Variables	Successful GT repair (n = 28)	Failed GT repair (n = 13)	p value
Active shoulder ROM
AAE	165.1 ± 4.9 (155 – 170)	144.6 ± 9.4 (130 – 155)	< 0.000
ALE	158.9 ± 7.2 (140 – 170)	138.4 ± 9.6 (120 – 155)	< 0.000
AER	30.5 ± 6.9 (20 – 50)	35.0 ± 6.3 (25 – 40)	0.367
AIR	33.7 ± 7.5 (20 – 50)	32.6 ± 6.9 (25 – 45)	0.671
Constant score
CSP	15.0 ± 0.0 (15 – 15)	11.9 ± 2.5 (10 – 15)	0.001
CSM	32.2 ± 2.3 (28 – 38)	26.1 ± 1.5 (24 – 28)	< 0.000
CSS	11.3 ± 2.3 (8 – 14)	6.1 ± 1.5 (5 – 8)	< 0.000
CSA	5.1 ± 1.6 (2 – 6)	2.0 ± 0.0 (2 – 2)	< 0.000
CS	70.7 ± 4.1 (62 – 77)	55.5 ± 5.7 (48 – 68)	< 0.000
ASES	90.3 ± 2.4 (85.0 – 94.9)	69.0 ± 5.7 (56.6 – 77.6)	< 0.000
DASH	22.1 ± 2.3 (19 – 26)	37.6 ± 2.6 (35 – 42)	< 0.000

Note: Data present as mean ± SD (range).

GT: greater tuberosity, ROM: range of motion, AAE: active anterior elevation, ALE: active lateral elevation, AER: active external rotation, AIR: active internal rotation, CSP: Constant score pain, CSA: Constant score activity, CSM: Constant score mobility, CSS: Constant score strength, CS: Constant score total, ASES: American shoulder and elbow surgeons score, DASH: disabilities of the arm, shoulder and hand score.

Complications

One patient developed infection within the 2nd postoperative week which responded to irrigation and debridement, for whom antibiotics were changed as per the culture sensitivity report. Eventually there was healing of tuberosities in this patient. Inferior glenoid notching of Sirveaux grade 2 was found in 6 (15%) cases, 3 cases in either group. No neurovascular complications or dislocations were seen.

Discussion

Biomechanics and functional outcome of HA prosthesis is dependent upon anatomic tuberosity healing.^1,8 Repair and healing of tuberosities are difficult in complex PHF in elderly, and even new implant design does not overcome this challenge.⁵ Zyto et al.¹⁶ found the results to be disappointing with median Constant score of 46°, and 70° of mean flexion and abduction each. Demirhan et al.¹⁷ showed that failed tuberosity repair and non-anatomic healing of the tuberosities badly influence on the clinical outcome following HA. Anjum et al.¹⁸ reported a median Constant score of 47.5 at a mean follow-up of 33 months with disappointing ROMs in the elderly.

Such unpredictable results achieved with HA in elderly patients, have prompted the use of RSA in them for complex PHF.^2,3,5,7,10,19, 20, 21, 22 A consensus regarding the advantage of tuberosity fixation and predictability of tuberosities healing in elderly patients is still lacking. Low rates of tuberosity healing as less as 40% have been reported with RSA for PHF in elderly.^5,20,21 Design biomechanics of RSA (inferiorization and medialization of the center of rotation, and deltoid becoming prime mover) can overcome the cuff deficiency,^12,13 which made a few authors recommend that repair was not necessary in RSA for displaced PHF.7, 8, 9^,14 However, the stabilizing effect of cuff muscles and the force couple relationship between cuff and deltoid, which enables subscpularis, teres minor, and infraspinatous to provide a fulcrum to the deltoid and supraspinatous for optimal functioning, has been demonstrated by Inman et al.²² in their classic article on shoulder motion, and further emphasized in multiple other electrophysiologic studies. In the present study, we have attempted to repair the tuberosities and analyzed its effect on the functional outcome of RSA.

The results of our study showed that tuberosity healing can be achieved in a more predictable manner. We obtained a higher rate (68.3%) of tuberosity healing by following a standard meticulous repair technique and interposition of cancellous bone graft, harvested from the fractured humeral head. We filled the cement below the level of the proximal metadiaphysis, which did not expose to tuberosities. This might have prevented cement-induced thermal necrosis of the inner surface of the tuberosities. Apart from promoting bone healing, bone grafts can also act as a barrier to prevent this thermal necrosis.³ Levy and Badman³ reported 86% union rate in their series of 7 patients, where they had used bone graft in a horse shoe fashion around proximal humeral side. Garofalo et al.²⁰ used the similar technique in patients with a mean age of 76 years, and the healing rate was 75% at an average follow-up of 2-year. Boileau et al.¹⁴ found a healing rate of 84% using similar technique of repair and bone grafting. As far as comparison between the 2 groups is concerned, we observed that older age, female gender, greater body mass index, delayed surgery (greater injury to surgery interval), displaced 4-part fractures (vs. 3-part fractures) and underlying diabetes mellitus to be more prevalent in the group with non-healing of the tuberosities (Table 3). The tuberosity healing rate was higher, when injury involved the dominant side shoulder.

In our study, the anterior elevation, lateral elevation, strength, and shoulder scores showed statistically significant improvement in the patients with successful repair and healing of the tuberosities in comparison to those with failure of tuberosity repair (Table 4). Mean value of active external and internal rotation was also better in the patients with successful repair of tuberosities, though the difference was not statistically significant. Failure of repair, tuberosity migration, resorption, or nonunion were associated with lower ROM and poor shoulder scores in all the cases with failure of tuberosity repair. All 13 patients with failed tuberosity healing complained of difficulties in overhead actions and some essential activities which require active rotational movements. In the report of Gallinet et al.,⁸ 15 patients underwent RSA with excision of cuff and tuberosities. The main issue in their series was an almost systematic postoperative lack of active external rotation (AER) in adduction, with 12 patients having 0° of AER. The single patient who has been reattached the tuberosities, had 150° of AAE and ALE each, 80° of AER and 60° of AIR. Boileau et al.¹⁴ treated all 38 patients with tuberosity reconstruction and concluded that tuberosities healing in RSA for PHF improved active mobility in forward flexion, external rotation, shoulder scores as well as patient satisfaction. In case of tuberosity nonunion, patients had poor ROM and difficulties in many activities of daily living.

Simovitch et al.²³ reported that fatty infiltration of the teres minor muscle was associated with poor AER and function after RSA to treat painful cuff tear arthropathy or irreparable cuff tear with pseudoparesis. They suggested that reverse shoulder replacement alone does not lead to desired functional result, unless the deficit of AER is addressed by some different techniques. Deficiency in the posterior cuff is associated with poor external rotation in adduction. They recommended repair of the posterior cuff for a better functional outcome with respect to rotation. Studies have also found poor functional outcome where infraspinatus and teres minor muscle were either absent or deficient.³ Better functional outcome following successful tuberosity repair in our study supports this concept of attempting repair of tuberosities. If tuberosities heal, then the strength and ROM of shoulders will be even better. Erickson et al.²⁴ showed significant improvement in ASES score following RSA in rotator cuff arthropathy, which had previously undergone rotator cuff repair with higher scores than control. This further reinforces the idea that although the biomechanics of the RSA prosthesis make it less dependent on rotator cuff integrity, there is still some room for improvement in shoulder function. The tuberosity repair and the dynamic interplay between cuff muscles and deltoid can help us achieve a near normal function. Similarly, the subscapularis has an enhancement to the anterior stability of the prosthesis and internal rotation. Klein et al.⁹ in their study have also mentioned that their regimen of resecting the tuberosities has no correlate in the literature which needs to be critically investigated, and functional rotator cuff reconstruction by tuberosity repair should be performed.

Tuberosity healing may also be related to reducing complication rates after RSA. Complications up to 40% are reported in some series of RSA for PHF, where tuberosity repair was not attempted or not successful.^3,5,8, 9, 10^,19,25 Cazeneuve et al.¹⁹ in a series of 30 acute fractures treated with RSA without tuberosity repair, reported 2 patients with instability, 1 with dislocation, 2 with implant loosening, 1 with infection, and 7 with proximal humeral bone lysis. Klein et al.⁹ reported 2 early infections and 2 dislocations in a series of 20 fractures treated with RSA and excision of the tuberosities. Bufquin et al.⁵ treated 36 cases with repair of tuberosities, out of which the repair failed in 19 (53%) patients. One case developed non-traumatic anterior dislocation. Gallinet et al.⁸ reported 2 cases of infection out of 15 patients, who underwent RSA with excision of cuff and tuberosities. We did not observe these complications in our series. Inferior notching is another reported complication.^10,21 Boileau et al.¹⁴ reported notching in 68% of their patients. Seebauer et al.²⁶ observed only grade 1 or 2 inferior glenoid notching in 49 out of 80 cases (63.6%). Significantly, notching affected the Constant score in those reports. We maintained an inferior tilt of 10° in all cases as recommended in the study of Gutierrez et al.²⁷ We found inferior notching in 6 (14.6%) cases (grade 2), 3 in either group. However, our mean duration of follow-up is comparatively short to comment regarding notching. Near anatomic reconstruction of the bony and soft tissue envelope all around the prosthesis might have added to the stability and prevented these postoperative complications in our series.

Limitations of our study are the relatively smaller sample size and the lack of a control group. A randomized controlled prospective study comparing “repair” vs. “not-repair” of tuberosities could have strengthened our concluding remarks. We have also not compared the prevalence of osteoporosis and bone mineral density values between the 2 groups.

In conclusion, tuberosity repair and interposition of cancellous bone grafts, harvested from the fractured humeral head can improve the probability of a successful tuberosity healing while performing RSA for PHF in elderly. Statistically significant improvement in AAE, lateral elevation, strength and shoulder scores were obtained in our series in the patients with successful repair of tuberosities around the prosthesis. Stability of the joint obtained after a successful tuberosity healing might have been the reason behind a lower rate of postoperative complications like instability, dislocation and loosening of the components. Patients with eventual failure of tuberosity healing showed a comparatively poor functional outcome, lesser shoulder scores and difficulty in daily activities. We recommend routine fixation of tuberosities for all the PHF in elderly patients while performing RSA.

Funding

Nil.

Ethical statement

Our institutional ethical committee has approved this study.

Declaration of competing interest

The authors declare that they have no competing interest.

Author contributions

Nirmal Chandra Mohapatra conceptualised the study and designed the analysis. Udit Sourav Sahoo conceived the study, designed analysis, collected and analysed the data, and wrote the paper. Madan Mohan Sahoo Contributed in data analysis and writing the paper. All authors read and finalised the manuscript.