A cadaveric study of the three different palpation-guided techniques for glenohumeral joint injections

Abstract

Background

It is not known whether an anterior, posterior or superior approach using the Neviaser portal is more accurate for glenohumeral joint injections. The aim of this study was to evaluate the accuracy of the palpation-guided technique and compare the three different approaches.

Methods

Palpation-guided glenohumeral joint injections were performed in 48 shoulders (24 cadavers) by two operators. Each shoulder was injected by three different approaches with a different coloured latex solution. The three approaches included the anterior, posterior and superior methods. The accuracy and location of unsuccessful injections were assessed through dissection of the shoulders.

Results

Posterior injections were the most successful with an accuracy rate of 89.6%, followed by anterior injections (75%) and superior injections (54.2%). Both posterior (p = 0.0001) and anterior injections (p = 0.03) were statistically significantly more accurate than superior injections. The most common failure mode was an intratendinous or intramuscular injection, which occurred most frequently with a superior approach.

Conclusions

Use of a superior approach through the Neviaser portal for the palpation-guided technique for glenohumeral joint injections showed a statistically significant inferior accuracy when compared to both the posterior and anterior approaches and is therefore not recommended. Posterior approached injections were the most accurate.

Introduction

Intra-articular glenohumeral joint injections are a useful adjunct in the diagnostic and treatment of shoulder joint pathologies.¹ Therapeutic injections are most commonly performed in such conditions as frozen shoulder, glenohumeral osteoarthritis, articular-sided partial-thickness rotator cuff tears and long head of biceps tendon pathology. Diagnostic injections are carried out in cases when it is not completely clear whether the shoulder pain generator is in the glenohumeral joint itself or somewhere else. The referred pain could be radiating from a cervical spine pathology, subacromial bursitis or acromioclavicular joint arthritis. A correctly administered injection can improve clinical outcome and be a valuable diagnostic tool in clinical suspicious painful glenohumeral pathologies, whereas inaccurately placed injections can cause further damage to tissues such as the rotator cuff and lead to incorrect conclusions being drawn as a diagnostic injection.^2,3

The gold standard of correctly administering a glenohumeral injection is either under fluoroscopic or ultrasound guidance. However, a palpation-guided injection is still the most frequently performed method for injection, particularly for therapeutic reasons.^4,5 Several techniques have been described to approach the glenohumeral joint with injections with no consensus in the literature as to which technique is superior. Some studies have demonstrated a higher accuracy with anterior injections,⁶ whereas other studies have concluded that posterior injections are more accurate.¹

To date, no study has reported on the accuracy of a superior injection technique using the arthroscopy portal described by Neviaser.⁷ This portal has become more familiar as surgeons use it to perform arthroscopic superior capsular reconstruction in non-repairable tears of the supraspinatus tendon as it provides better access to the glenoid.⁸

The aim of this cadaveric study is to evaluate and compare the accuracy of the three different palpation-guided techniques for glenohumeral joint injection.

Methods

Forty-eight shoulders of 24 Thiel embalmed cadavers (13 men, 11 women) were used for this study. The mean age at time of death was 71 years (range, 56 to 89 years). The embalming technique by Thiel⁹ was developed over a period of 30 years. It preserves original colour, consistency and transparency of the tissues, as well as allowing an almost full range of passive movement of the joints. None of the cadavers exhibited signs of advanced arthritis, evidence of trauma, previous surgery or other pathological changes.

Three different palpation-guided techniques were used with the attempt to inject the glenohumeral joint. These were the anterior, posterior and superior approaches (Figure 1). All injections were performed in a beach chair position with the shoulder in neutral rotation and with the use of a 22-gauge needle. Each shoulder was injected using each of the three different methods, using 1 mL of fluid latex in a specific colour for each technique to distinguish the success rate and location – anterior (red), posterior (blue), superior (green). The injection order was varied, so that each technique was equally performed as first, second and third injections into the shoulders. In each technique group, the injections were performed bilaterally in all of the 24 cadavers (48 shoulders) with 24 in the left and 24 in the right shoulders.

Figure 1.

Right shoulder in beach chair position with a view from superolateral. Illustration of the three palpation-guided techniques which were used to approach and inject the glenohumeral joint: Anterior (red), superior (through the Neviaser portal; green) and posterior (blue). The anatomic landmarks are marked.

The anterior technique uses the rotator interval to reach the glenohumeral joint. The anatomical landmarks were palpated and marked at the beginning. These are the coracoid, the anterior edge of the acromion and the acromioclavicular joint. As with the standard anterior arthroscopy portal, the injection was placed 1.5 cm anterior to the acromioclavicular joint and the needle was inserted at an angle of 45° laterally and 45° caudally. The posterior injection site was 2 cm caudal and 1 cm medial to the posterolateral corner of the acromion. The needle was aimed in the direction towards the coracoid, angulated about 30° caudally.

The superior injection site was at the soft spot at the site of the arthroscopic Neviaser portal, which is 1 cm medial to the medial edge of the acromion. This spot is central to the sagittal border, which is the lateral clavicle anterior and the spine of the scapula posterior. The needle was directed caudally and about 15° laterally.

The injections were performed by a shoulder fellowship trained orthopaedic surgeon and an orthopaedic surgical resident, of which each person performed 72 injections. In all of the injections, only one attempt was performed. The aim of the injection was to get through the resistance of the joint capsule and perform a glenohumeral injection. It was not intended primarily to feel a contact of a bony structure, such as the humeral head; however, if contact was made the needle was pulled back slightly to allow an injection without resistance.

After all injections were performed, the shoulders were dissected by an independent anatomy professor. In the cases with a clear connection between the glenohumeral joint and the subacromial space, as in the situation of a full-thickness rotator cuff tear the cadaver was excluded and both shoulders of another cadaver were injected instead. It was evaluated and noted, if the injection was glenohumeral or not (Figure 2). If it was extraarticular, the location of failure was noted and distinguished between intracapsular, intratendinous, intramuscular and subacromial.

Figure 2.

Posterosuperior view after dissection of a left cadaveric shoulder with removal of the infraspinatus tendon. The green coloured latex (injected by a superior approach through the Neviaser portal) is located intratendinous at the musculotendinous junction of the supraspinatus muscle. The blue (posterior approach) and red (anterior approach) coloured latex are accurately located in the glenohumeral joint.

SSP: supraspinatus; ISP: infraspinatus; HH: humeral head.

Statistical analysis was performed using IBM SPSS® statistics software (version 22.0, Armonk, NY). A χ² test was used to assess correlations between the success rate of the three injection techniques. A p-value < 0.05 was deemed to be statistically significant.

Results

The overall accuracy of all three injection techniques was 72.9% with a successful glenohumeral joint injection in 105 of 144 shoulders. Posterior injections were the most successful with an accuracy rate of 89.6% (43 out of 48). This was statistically significant compared to superior injections, which had an accuracy rate of only 54.2% (26/48; p = 0.0001). Anterior injections had an accuracy rate of 75% (36/48) and were significantly more accurate than superior injections (p = 0.03). There was no statistically significant difference between the success rate of the anterior and posterior techniques (p = 0.06). The most common mode of failure was an intratendinous or intramuscular injection, which occurred in 39% of shoulder injections (15 of 39 failures). This was followed by subacromial with 33% (13/39) and intracapsular injection with 28% (11/39). Detailed results are listed in Table 1.

Table 1.

Results of palpation-guided glenohumeral joint injection with three different approaches: anterior, superior (through the Neviaser portal) and posterior.

Technique	Anterior	Posterior	Superior	Overall
Accuracy rate	75% (36/48)	89.6% (43/48)	54.2% (26/48)	72.9% (105/144)
Non-successful injections	12	5	22	39
Subacromial	3	1	9	33.3% (13/39)
Intracapsular	8	1	2	28.2% (11/39)
Intramusculous/-tendinous	1	3	11	38.5% (15/39)

Six cadavers were identified to have large cuff tears with a clear connection between the glenohumeral joint and the subacromial space and were replaced with six further cadavers. There was no statistically significant difference between the accuracy rate of the two operators (Table 2) who performed the injections, neither within the groups nor overall (p = 0.09). However, the more experienced surgeon reached a higher accuracy rate (79.2% vs. 66.7%).

Discussion

The most important finding of the study was that a palpation-guided injection of the glenohumeral joint was most accurate with a posterior approach. Accuracy of both the posterior and anterior approach was statistically significantly higher in comparison to a superior approach. The superior technique through the Neviaser portal was the least successful with an accuracy rate of only 54%. Intratendinous or intramuscular placed injection happened most commonly with the superior technique. These findings are especially important as an injection of glucocorticoids into a tendon has the potential of further tissue damage and should be clearly avoided in the authors’ opinion.²

In the current literature, accuracy rates of image-guided injections to the glenohumeral joint range from 72% to 100% under fluoroscopic guidance and from 63% to 100% under ultrasound guidance in awake patients.¹⁰ Unguided or palpation-guided techniques had accuracy rates from 10% to 100% in clinical studies and 73% to 100% in cadavers. To date, it is not clear which technique have more accurate results. In a systematic review, Daley et al. concluded that posterior injections of the glenohumeral joint are statistically significantly more accurate than anterior injection techniques.¹ However, other comparative studies as by Tobola et al. had higher accuracy rates with an anterior technique but not statistically significant.⁶ Interestingly, in this study more experienced providers did not have a higher accuracy with their injections. We found in our study, the accuracy rate of a fellowship trained shoulder surgeon was higher compared to an orthopaedic resident (79% vs. 67%), although these differences were not statistically significant. We further found in contrast to Tobola et al. that the highest accuracy of injecting the glenohumeral joint was by the palpation-guided technique through a posterior portal. However, these differences (87% vs. 75%) were not statistically significant.

A study by Smeets et al. demonstrated that small geographic distances between Dutch and Flemish shoulder specialists have an impact on the preferred technique for glenohumeral joint injections.⁵ Whereas 69% of Flemish shoulder surgeons use a posterior technique, 77% of the neighbouring country the Netherlands prefer an anterior approach. Regional differences in training and injection technique can therefore influence the accuracy that can be achieved. Patel et al. found in a cadaver study an overall success rate of 83% with a posterior technique, reaching 93% with ultrasound guidance and 73% without guidance.¹¹ Kraeutler et al. performed an anterior needle placement for a glenohumeral injection and evaluated the rate of successful intraarticular needle tip placement with an arthroscopy whilst viewing from the posterior portal.¹² The needle could be visualized in all cases (100%) and fluid could be expressed through it in 93% of cases. Powell et al. had an 89% accuracy rate with palpation-guided glenohumeral joint injections with gadolinium performed through an anterior approach, where the success was confirmed by magnetic resonance arthrography.¹³ Shao et al. described a 96% accuracy rate in patients with idiopathic adhesive capsulitis.¹⁴ In this study, the success of the injection was confirmed by the intraarticular contrast seen on the first fluoroscopic picture. Brockmeyer concluded in an editorial commentary that recent results of unguided injections are comparably accurate and could be therefore considered in the hands of experienced shoulder arthroscopists.⁴ On the other hand, radiologically guided injections are highly accurate, reproducible but they are more costly and time consuming requiring special imaging devices.

In the authors’ opinion, further comparative studies are necessary to verify the effectiveness of unguided injections as this has not been proven scientifically to date. High accuracy rates of up to 96% are achievable but also as low as 46%, which are not acceptable in the authors’ opinion in performing both therapeutic and especially diagnostic glenohumeral injections. In this cadaveric study, the highest accuracy could be reached with a posterior approach (90%). In our practice, fluoroscopic-guided injections remain the gold standard for diagnostic and therapeutic injections. However, palpation-guided injections with a posterior approach and ultrasound-guided injections with an anterior approach remain an option for repetitive therapeutic glenohumeral injections, especially in setting of a private practice without fluoroscopic availability.

The main limitation of this study is the fact that it was performed in cadavers. However, Thiel embalmed cadavers have been proven to be suitable for joint injection studies with several studies publishing this.^15–17 Regarding the different palpation-guided injection techniques used in this study, one has to acknowledge there is no clear standardization in the literature of the approaches used, in particular the anterior approach. Several techniques have been described in the literature;^6,12–14 however, we used a clearly defined and standardized anterior approach similar to that described by Kraeutler et al.¹² and Powell et al.¹³ as a highly accurate technique. A strength of the study is that the three different techniques were directly compared to each other with 48 shoulders per technique used. The coloured latex solution allowed comparison of the success rate of the techniques in the same shoulder. This limits a possible bias due to anatomic differences between the shoulder joints. A further strength of the study is that injections were independently performed by two different experienced operators.

Conclusions

To conclude, a superior approach through the Neviaser portal for palpation-guided glenohumeral joint injections showed a statistically significant lower accuracy rate when compared to both posterior and anterior approaches and is therefore not recommended. A posterior approach through the standard posterior arthroscopic portal was the most accurate in this cadaveric study with a correct intraarticular injection in 90% of cases.

Table 2.

Results of palpation-guided glenohumeral joint injection of two operators, of who both injected 72 shoulders.

	Success	No success	Success rate
Operator 1	48	24	66.70%
Anterior	16	8	66.67%
Posterior	22	2	91.67%
Superior	12	12	50.00%
Operator 2	57	15	79.20%
Anterior	20	4	83.33%
Posterior	23	1	95.83%
Superior	14	10	58.33%

Note: Operator 1 is a shoulder fellowship-trained orthopaedic surgeon, operator 2 is an orthopaedic surgery resident. The bold values in the table signifies p = 0.09 – overall between operator 1 and 2.

Ethical Review and Patient Consent

Not applicable. The bodies were donated to the Institute of Macroscopical and Clinical Anatomy of the Medical University of Graz according to the strict donation rules of the Institute and according to the Austrian Law.