Examination of concomitant glenohumeral pathologies in patients treated arthroscopically for calcific tendinitis of the shoulder and implications for routine diagnostic joint exploration

We conducted a descriptive retrospective cohort study of patients presenting with symptomatic chronic CTS undergoing arthroscopic removal of CD between 2008 and 2011 that failed at least six months of conservative treatment. Patients had been referred by a general practitioner, rheumatologists, or orthopedic surgeon for treatment. Most common clinical findings were persistent shoulder pain, functional disability and the presence of symptomatic rotator cuff calcifications. Further inclusion criteria were the presence of radiographically and sonographically determined calcifications within the rotator cuff and the presence of clinically symptomatic CTS of the shoulder. Patients with previous ipsilateral shoulder surgeries or shoulder comorbidities were excluded. Calcifications were confirmed preoperatively by conventional radiographs (anterior-posterior) and bilateral ultrasound examination.

Preoperative ultrasound examination (1 to 2 days before surgery) of the rotator cuff was performed in a standardized manner by a board certified orthopedic surgeon who was experienced with musculoskeletal sonography. Calcific depositions were evaluated according to the quadrant technique within preoperative outpatient consultation [20]. The number of calcifications and dorsal attenuation were recorded accordingly. Additionally, ultrasound was used to assess the integrity of the rotator cuff and biceps tendon.

Radiographic parameters (based on standard anterior-posterior (a.-p.) and outlet radiographic views) were measured at two time points: preoperatively (1 to 2 days before surgery) and postoperatively (within 2 to 3 days after surgery). The number of calcific lesions and localization of calcifications (medial/lateral) were assessed using standardized a.-p. films. Acromion morphology was classified according to Bigliani et al. on outlet radiographs [21]. Postoperative radiographs were performed to determine the amount of CD removal.

All surgeries were performed by the same board-certified orthopedic surgeon experienced in shoulder arthroscopies (PO). Our surgical procedure of shoulder arthroscopy in patients with CTS has previously been described in detail [6]. Briefly, under general anesthesia patients were positioned laterally in a decubitus position with slight extension to the arm. Preoperatively, the exact localization of CD was determined by utilization of ultrasonographical mapping according to the quadrant technique [20]. As the patient’s arm was resting in the neutral position, the intraoperative anatomy was configured in the same exact fashion as performed for preoperative localization of the rotator cuff calcifications, which facilitated in situ localization of CD significantly. The calcifications were identified and marked with respect to the affected quadrants.

A stab skin incision was made followed by careful advancement of the scope into the glenohumeral joint through a posterior portal. The diagnostic arthroscopy was carried out in a conventional manner to rule out relevant intraarticular concomitant pathologies. Then, the subacromial space was identified and exposed using the same approach. Additionally, a lateral approach was used to insert a blunt trocar and subsequently gain access to the bursa. As the bursa was frequently hypertrophic and/or hypervascularized due to the chronic proinflammatory stimulus of the irritating intraarticular calcifications (Fig. 1), we performed a partial bursectomy by utilization of the arthroscopic shaver. The respective CD was then identified via needling and the center of the calcification was localized and incised with a needle (Fig. 1b). Calcific remnants in the needle’s tip indicated correct localization. Hereafter, a blunt hook probe was carefully advanced in order to gently release and eliminate (squeeze-and-stir-technique) the calcifications (Fig. 1c, d). Additionally, the superficial membrane was elevated in order to flush the joint followed by intermittent squeezing of the CD with the probe. This procedure was repeated until the calcific deposit was removed in a blunt fashion (Fig. 1e). Finally, a thorough irrigation of the “calcific cave” was performed by utilization of a syringe to ensure proper elimination of small calcific particles. Finally, the continuity of the affected rotator cuff tendon was confirmed precisely (risk for underestimation of rotator cuff tears following CD removal) before the wound was closed. No tendon repair was performed after CD removal.

Intraarticular concomitant pathologies were assessed within glenohumeral arthroscopy. The configuration of glenohumeral ligaments were evaluated according to Morgan et al. [22]. Glenohumeral cartilage injuries were assessed according to the International Cartilage Repair Society (ICRS, Table 1) cartilage injury classification (Fig. 2) [23]. Consistency and characteristics of calcifications were evaluated by macroscopic assessment and/or manipulation via blunt probe. Superior labral tears from anterior to posterior (SLAP lesions) were classified according to Snyder et al. (Fig. 3) [24]. Tears of the proximal biceps tendon were graded according to the thickness of injury (superficial, 25%, 50%, 75%, and full thickness rupture (100%). Rotator cuff lesions were classified according to Ellman et al. for partial tears and Bateman et al. for full thickness tears including indication of the affected tendon [25, 26].

Continuous variables are shown as mean ± standard error of the mean. For categorical variables percentages were calculated. All analyses were performed using SPSS version 20 (IBM, Armonk, NY, USA).

The study was approved by our local institutional review board and informed consent was obtained from all patients before surgery (protocol number: 598/16).

145 patients were enrolled in the study (66.2% female, Table 2). Mean age at surgery was 50.9 ± 8.9 years (range: 32–76). The major indication for arthroscopic removal of CD was symptomatic chronic CTS. The mean symptomatic time until patients decided to undergo surgical removal of CD was 3.5 years. 2.6% of patients had a previous shoulder surgery on the contralateral shoulder in the past. The most commonly treated side due to symptomatic CTS was the right shoulder in 54.5% of patients (Table 2). The dominant shoulder was affected in 33.1% (48/151).

Preoperative radiographic assessment of CTS was performed 1 to 2 days before surgery. As demonstrated in Tables 3, 72% of calcifications were localized laterally while 19.4% of calcific depots were found within the medial part of the rotator cuff. Few patients (8.3%) presented a bilateral configuration of the calcification (i.e. lateral +50% medial). Furthermore, the majority of patients (94.5%) featured a single calcific depot within the preoperative radiographic evaluation. Distribution of acromion morphology according to Bigliani et al., resulted in 31.7%, 59.3%, and 9% for a flat (type I), curved (type II) and hooked (type III) acromial shape (Table 3).

The articular cartilage of the humeral head was intact in 45.5% of patients (Table 2). While mild cartilage injury at the humeral head (grade 1 and 2) was found in 29.7% and 23.4%, respectively, severe cartilage lesions were only detected in 1.4% (grade 4) of patients (Fig. 2). Moreover, 48.3% of glenoids did not reveal any cartilage injury at all. Mild cartilage lesions on glenoid articular surfaces were diagnosed in 26.2% (grade 1) and 24.1% (grade 2) of patients (Table 2). Severe glenoid lesions (grade 3) were recorded in 2 (1.4%) patients. Typical co-pathologies that were identified during arthroscopy were partial tears of the proximal biceps tendon (2.1%), SLAP lesions (1.4%), and rotator cuff defects (0.7%; Fig. 4). SLAP lesions and partial tears of the proximal biceps tendon were incidental findings. Furthermore, the glenohumeral capsular type was evaluated according to Morgan et al. (Table 2) [22]. The vast majority of our patients were classified as capsular type I (93.1%). Capsular type II and III were recorded in 1.4% and 4.8% of patients, respectively. Sublabral foramen were observed in 6.9% of cases, whereas over 90% of patients did not present any sublabral lesions. Two of our patients (1.4%) required specific glenohumeral surgical treatment due to a SLAP lesion type III and an intraarticular rupture of calcific deposits (Figs. 4, 5, and 3). As the other SLAP lesion in our study population was mild (type I) a debridement was not indicated. Neither did the patients with SLAP lesions complain about specific symptoms preoperatively nor have the SLAP lesions been detected via ultrasound (i.e. joint effusion) before surgery. One patient (0.7%) showed a tear of the rotator cuff interval without involvement of the subscapularis or supraspinatus tendon. CD removal was confirmed postoperatively by conventional X-rays.

During arthroscopic removal of CD, rotator cuff calcifications typically presented with a tough and “tooth-paste”-like consistency (63.4%; mixed-forms were described); Table 4 and Figs. 1 and 5). We observed two cases (1.4%) of CD perforations into adjacent anatomical structures. One patient showed an extensive calcific bursitis as a result of CD rupture and perforation into the subacromial bursa. We observed CD perforation into the glenohumeral joint with both adhesive and free floating calcific components inside the joint in another patient.

Most commonly, treatment of symptomatic CTS can be managed non-operatively (90% of patients) by physiotherapy, analgesics, and injections [27]. Surgical treatment is reserved for patients in which conservative therapy has failed, such as prolonged periods of functional disability, severe pain, and when calcific deposits do not resolve spontaneously. To date, most surgeons perform an arthroscopic removal of CD with/without subacromial decompression and glenohumeral exploration [7]. Recently, our group demonstrated that good to excellent results can be achieved in 90% of patients with blunt arthroscopic removal of calcific lesions, without performing a subacromial decompression. Therefore, we pursue a minimal-invasive strategy which rather tolerates minor remnant calcifications in favor of rotator cuff integrity [6]. In the present study, the vast majority of our patients did not present any intraarticular pathologies requiring surgical treatment. If intraarticular pathologies were detected, most likely these lesions did not cause any symptoms. Our results confirm the conclusion of Sirveaux et al., who compared clinical and radiographic outcome in patients with CTS treated either by CD removal alone via an arthroscopic bursal approach or CD removal combined with a standardized glenohumeral exploration [18]. The authors did not identify a single glenohumeral injury that required surgical treatment. Moreover, duration of postoperative pain and latency of return-to-work were significantly shorter in patients who received CD removal solely compared to patients that underwent CD removal including an additional glenohumeral exploration. After 6-month follow-up, functional outcome and radiographic CD disappearance did not reveal significant differences. Therefore, we agree on the authors’ statement, that a glenohumeral exploration can be avoided as the additional approach might include an increased risk for complications as well as it presumably leads rather to prolonged rehabilitation than clinical benefit. Furthermore, prolonged duration of surgery, increased postoperative pain and subsequently prolonged length of stay potentially cause higher costs [28, 29]. A glenohumeral exploration may be justified in specific scenarios: I.e. if a relatively large and consistent CD removal has been performed and an intra-articular assessment of the rotator cuff’s integrity is pursued. A thorough evaluation of the rotator cuff (especially of deep layers) may be facilitated by means of arthroscopic visualization. Nevertheless, for the majority of patients with CTS, an additional glenohumeral exploration seems to be an unnecessary and an expandable risk for complications as it contains an additional approach to the joint.

The amount of CD removal is another matter of debate: several authors favor a complete removal of CD while other authors presented good outcome with incomplete excision of CD [6, 7, 10, 30‐32]. A total removal of CD may potentially result in a full thickness rotator cuff tear. Evidence is lacking whether an immediate reconstruction or no repair at all leads to superior outcome [6, 9, 31, 33, 34]. Last but not least, the size and consistency of CD lesions are highly important for the removal of rotator cuff calcifications. In the present study, the vast majority of CD lesions was either tough and tooth paste-like or presented as snowy powder, indicating that an easy removal is feasible without the application of intense forces.

In summary, instead of a standardized glenohumeral exploration within arthroscopic removal of CD lesions, we suggest a patient-specific treatment algorithm, that is individualized on the patient’s presenting complains in order to optimize the risk-benefit-ratio.

CTS is predominantly an extraarticular phenomenon rather than a glenohumeral joint disease [1, 34]. Consequently, we expect an equivalent prevalence of glenohumeral pathologies in patients with CTS compared to healthy subjects. Certainly, an arthroscopic glenohumeral exploration provides the opportunity to detect intraarticular co-pathologies. However, our and previous data demonstrate that intraarticular co-pathologies in patients with CTS were hardly detected and moreover, if an intraarticular pathology was present, this almost never caused any procedural alteration [18]. Therefore, it is reasonable to question whether an additional glenohumeral arthroscopy can be justified in patients with CTS considering the increased risk for complications such as infection or shoulder stiffness against the lack of a true benefit through arthroscopy [18]?

The overall risk for complications in shoulder arthroscopy ranges between 4.8% and 10.6% [35‐37]. Typical complications in patients having CTS are prolonged pain, secondary adhesive capsulitis, rotator cuff tears, ossifying tendinitis, and osteolysis of the greater tuberosity [38]. The incidence of a frozen shoulder after shoulder arthroscopy is 2% to 5% in the general population [39, 40]. Postoperative shoulder stiffness after rotator cuff repair ranges between 4.9% and 32.7% [41‐43]. Shoulder stiffness is not well tolerated by patients with CTS; it does not resolve easily and may require long-term rehabilitation [44]. Reasons for shoulder stiffness are supposed to be the manipulation of capsule and/or residual calcium debris [6, 34]. Moreover, postoperative shoulder stiffness may be related to rotator cuff tear morphology, postoperative immobilization, glenohumeral adhesion, capsular contracture, or underlying predisposing patient comorbidities [42, 44, 45]. Reduction of approach-related comorbidity may potentially reduce complication rates and offers the opportunity to perform this procedure as an outpatient surgery that allows for immediate rehabilitation, which subsequently reduces the risk for postoperative stiffness [18]. Our current rate of postoperative arthrofibrosis is relatively low compared to previous studies – however, if the risk for postoperative arthrofibrosis can be further reduced without performing a glenohumeral approach while at the same time achieving equivalent outcome, one may assume that it might be reasonable to abstain from a glenohumeral exploration in routine fashion when treating patients with CTS arthroscopically [46]. However, this needs to be further confirmed in prospective comparative investigations.

Furthermore, infection following shoulder arthroscopy is another relevant risk factor. Pauzenberger et al. observed infections following arthroscopic rotator cuff repair in 0.009% [47]. The authors identified sex (male), age (≥60 years), and length of surgery (≥90 min) to be significantly associated with postoperative infection. Other groups reported overall infection rates after shoulder arthroscopy between 0.03% and 3.4% [37, 48‐50]. Especially, joint infections due to Propionibacterium acne (P. acne) are currently controversially discussed in association with shoulder arthroscopy as conventional perioperative antibiotic- or preoperative prophylaxis do not seem to sufficiently decrease the risk for postoperative infections [47, 51]. Moreover, presurgical skin preparations do not entirely eliminate P. acne [52‐54]. Current evidence suggests P. acne being the most frequent identified organism in shoulder infections [51, 55, 56]. Both, open and arthroscopic surgery provide approach related-opportunities for P. acne to be transferred from skin to deep layers and thus potentially causing glenohumeral joint infections [57]. Seth et al. observed differences in positive skin cultures contaminated by P. acne that were assessed before skin incision (15.8%) and directly before wound closure (40.4%), which underlines the association between length of surgery and potential risk for surgical side infections [58]. In order to minimize the risk for infections by P. acne, it is suggested to reduce the size and contamination of surgical approaches [59]. Due to the fact that glenohumeral joint infections are associated with disability and significant direct and indirect socioeconomic costs, we suggest to perform a glenohumeral joint exploration only if an intraarticular injury that requires surgical treatment is highly expected.

In general, the impact of glenohumeral exploration within arthroscopic removal of calcifications in CTS remains an under-investigated issue. To date, postoperative pain and latency of return-to-work were found to be significantly shorter in patients who received CD removal solely compared to patients that received an additional glenohumeral exploration [18]. It is reasonable that any additional manipulation/invasive maneuver during shoulder surgery might affect the clinical and/or functional outcome and potentially increases the likelihood of complications since this has already been demonstrated in various other surgical procedures [13, 37, 60, 61]. Investigations comparing complications of different arthroscopic approaches have not been performed yet. Sufficient data exist on 30-day readmission rates as well as risk factors for postoperative complications following shoulder arthroscopy as described by Shields et al. [45]. The authors found shoulder arthroscopy to have a 1% thirty-day complication rate. Age ≥ 60 years, operating room (OR) time ≥ 90 min, chronic obstructive pulmonary disease, inpatient status, disseminated cancer, and nicotine abuse are risk factors for postoperative complications [60]. These results have been confirmed by other authors such as Moody et al. who demonstrated, that prolonged operative time, more invasive and/or additional surgical approaches increase the risk for complications as well as health care expenditures [29, 60].

In the present study, intraarticular pathologies that needed surgical treatment have only been observed in few patients. For the majority of patients with CTS, an additional glenohumeral exploration might be an unnecessary risk and would most likely not translate into a clinical benefit. As patients undergoing surgery in CTS already reflect a “negative selection” due to failed conservative treatment, a reasonable risk-benefit-ratio by means of minimal approach-related comorbidity and perioperative risk for complications should be pursued. As there is currently no evidence supporting that patients with CTS benefit from glenohumeral explorations during arthroscopic removal of calcifications, the additional glenohumeral approach should only be performed in case of founded suspicion of relevant concomitant intraarticular pathologies.

The retrospective study design is associated with certain limitations, such as loss of information (missing reports), heterogeneous preoperative conservative treatment, and unequal group power (i.e. gender, affected shoulder, etc.). Additionally, functional and clinical outcome data was not implemented into this study. Moreover, this study was of descriptive nature and did not include a control group. Nevertheless, as far as we are aware, there is no previous study specifically investigating intraarticular glenohumeral pathologies during CD removal in patients with calcifying tendinitis. Furthermore, a huge advantage of the present study is that all patients were treated by the same orthopedic surgeon resulting in high consistency and homogeneous evaluation of intraarticular conditions.