Degeneration and consequent tearing of the rotator cuff (RC) have been largely attributed to the tissue's physiological senescence and intrinsic factors. Intrinsic factors fall into two categories: medical conditions (comorbidity) and lifestyle factors. The comorbidities thought to be associated with RC tear include diabetes [1-3], hypercholesterolemia [4-6], uncontrolled arterial hypertension [7], metabolic syndrome, and obesity [8,9]. Additional lifestyle factors that may predispose an individual to RC tear include smoking [10-12], disordered eating (eating disorders) [8,9], and excessive alcohol consumption [13], each of which may contribute to peripheral microcirculatory disease. The microcirculation of the RC is of particular interest with respect to the etiology of RV tear [14].

The tendons of the RC are supplied by the suprascapular, thoracoacromial, scapular circumflex, and anterior and posterior humeral circumflex arteries [15-19]. The posterior humeral circumflex artery supplies the teres minor tendon [20]; however, its role in the vascularization of the supraspinatus tendon is still a matter of debate, whereby this role is accepted by Determe et al. [21] but denied by Põldoja et al. [15]. The anterior and posterior circumflex humeral arteries cross the osteotendinous junction to provide most of the blood supply to the posterosuperior cuff [20,22]. Brooks et al. [23] found no significant difference between the vascularity of the supraspinatus and infraspinatus tendons from their humeral insertions. The supraspinatus muscle receives its blood supply from the suprascapular and dorsal scapular arteries, while the infraspinatus muscle is supplied by the suprascapular and circumflex scapular arteries [17]. However, the subscapular artery has been found to supply the greater portion of the infraspinatus muscle circulation through its dorsal or circumflex scapular branch [24]. The blood supply of the teres minor muscle derives from several vessels, typically including the branch arising from the posterior humeral scapular circumflex artery [24].

In this study, we performed arthroscopic visualization of the distribution of visible vessels on the articular and bursal sides of the posterosuperior cuff to investigate whether there were any macroscopic differences between subjects with and without the above-noted conditions and lifestyle factors. Our hypothesis was that these conditions compromised the vascular supply of the RC. 

The study protocol was approved by the Institutional Review Board of Sapienza University of Rome, which waived the requirement for informed consent since a standard surgical procedure was performed and the anonymity of the patients was ensured during data collection.

This prospective study initially included 128 patients aged 50 to 72 years at Istituto Chirurgico Ortopedico Traumatologico (ICOT) who underwent shoulder arthroscopy from January 2017 to December 2022 for examination of the macroscopic vasculature of the posterosuperior RC. Twenty-one patients diagnosed with RC tears, frozen shoulder, rheumatic, autoimmune diseases, and fracture sequelae were excluded. Patients who had undergone injections (intra-articular or subacromial) of aspirin, antiplatelet agents, steroids, or any other viscosupplementary substance were excluded [25].

The final study group comprised 107 patients (59 male and 48 female). The mean age was 58.2±4.1 years. Groups I, II, and III comprised 36, 45, and 26 patients, respectively (Table 1). These patients underwent arthroscopic examination to assess glenohumeral instability without RC tear (28 patients), acromioclavicular injury (type III and V) with possible concomitant pathologies (e.g., superior labrum anterior to posterior; 32 patients), acute rupture of the biceps long head tendon with healthy cuff (29 patients), and low-grade (Samilson-Prieto 1) glenohumeral arthropathy without synovitis after undergoing chondroplasty (18 patients).

The vessels identified in the myotendinous junction after arthroscopic examination of the articular side of the posterosuperior cuff are shown in Fig. 1. Differences between the three patient groups were significant (χ²=14.21, df=4, P=0.007), with a Cramer’s V=0.258 for k=3, which indicates a moderate strength of association. These vessels were extremely thin, ran parallel to each other and perpendicular to the tendon insertion, and usually ran in pairs. When visible, the vessels did not reach the tendon insertion, which macroscopically appeared not to be vascularized (see critical area identified in Fig. 2). Some pairs of vessels formed terminal loops. The loops suggest that the vessels were interrupted, and that their inability to be visualized is not simply due to their deep position in the tissue.

The arterial vessels (pulsating) identified as originating from the osteotendinous junction are shown in Fig. 3. A significant difference between the three patient groups was found (χ²=19.576, df=4, P<0.001) with a Cramer’s V=0.302 for k=3, which indicates a moderate strength of association. When visible, these vessels were thin, immediately branched into vessels of smaller caliber, and were not observed after 3−7 mm (Fig. 4, Supplementary Video 1).

When viewed from the bursal side of the posterosuperior cuff, a dense anastomotic network (Fig. 5) was identified in 94% (34) of group I patients (clearly evident, 28 patients; poorly evident, 6 patients), 80% (36) of group II patients (clearly evident, 26 patients; poorly evident, 10 patients), and 35% (9) of group III patients (clearly evident, 2 patients; poorly evident, 7 patients) (Fig. 6). These differences were significant (χ²=37.15, df=4, P<0.001), with Cramer’s V=0.417 for k=3, which indicates a strong association. These vessels were very superficial and of small caliber. It was sufficient to place a hooked probe about 8−10 mm from the osteotendinous insertion, without exerting pressure on the underlying tissue, to interrupt the blood flow. By detaching the probe from the cuff, the flow immediately resumed, and the area containing the anastomosis, previously and temporarily avascular, was once again supplied by vessels with flow in the opposite direction (Fig. 7).

Our study showed that subjects with comorbidities or lifestyle habits capable of compromising the peripheral microcirculation had a less evident vascular texture of the RC on both the articular and bursal sides. This association was stronger for patients with more comorbidities or unhealthy lifestyle factors. The etiology or mechanism underlying this association remains a matter of debate. Various pathological conditions, including hyperglycemia, hyperlipidemia, hypertension [27-29], aging [27] and exposure to specific drugs [27], may impact endothelial function by disrupting the molecular mechanisms regulating nitric oxide (NO) bioavailability. NO is the major endothelium-derived second messenger responsible for the maintenance of vascular homeostasis. Reduction in NO bioavailability, resulting from reduced NO production and/or increased NO degradation by superoxide anion, marks the onset of endothelial dysfunction [30].

The aforementioned pathologies are also responsible for an increase in the plasma concentrations of reactive oxygen species (ROS), which are formed as byproducts of physiological metabolism of molecular oxygen [27,30] or during bacterial killing and host defense responses [27,30-32]. At physiological concentrations, ROS act as important second messengers that transduce intracellular signals involved in various biological processes [28,33]. However, when aberrant production of ROS exceeds the buffering capacity of the antioxidant defense systems or when antioxidant enzymes are defective, oxidative stress occurs. Several studies have shown that oxidative stress plays a pivotal role in mediating the production and secretion of cytokines [30-33], linking ROS with inflammation and endothelial activation and dysfunction. Accumulation of ROS may induce apoptosis of tenocytes through the action of mediators such as metalloproteinases, cytokines, and c-Jun N-terminal kinases, which, when phosphorylated, leads to cell death [27,33]. Over time, apoptosis causes matrix degeneration and reductions in collagen synthesis and mechanical tendon resistance, leading to tissue degeneration and disruption/tearing [27,33].

Chansky and Iannotti [19] and Rothman and Parke [6] postulated that hypovascularity plays a significant role in the genesis of RC degeneration and tearing. Rathbun and Macnab [34] undertook a microangiographic study in which injected specimens were sectioned to show a histologic correlation between the area of vascularity and the degenerative pattern observed. Those researchers found that degenerative changes first occurred in the area of hypovascularity and were always more extensive in that area.

As previously noted by Lohr and Uhthoff [22], we observed that the articular side of the tendon is at a disadvantage in terms of blood supply in comparison with the bursal side. It might be hypothesized that this deficient vascularization increases susceptibility of the articular side to degenerative changes and, consequently, to tears.

Our study also showed that the vessels on the bursal side of the RC form a dense anastomotic network and are thin in caliber. Furthermore, we found that the blood flow of these vessels can be easily interrupted by placing a hooked probe on them without exerting pressure. These findings indicated that the vascular supply of these vessels is precarious. Some authors have suggested that the vascular supply of the cuff can be changed according to the position of the arm [34]. These data led us to ask if double row (DR) complex constructs are strongly ischemic and predisposed to failure. Although the literature provides indications that the DR offers greater likelihood of a mechanical seal than simple constructions (single row [SR]) [35-41], the failure rate of DR is still high.

There is an apparent contradiction with respect to the DR technique, in that it is a biomechanically superior construct but is hypothetically a more ischemic construct compared to the SR technique. However, none of the published studies comparing the two techniques have taken into account the state of degeneration of the tissue to be repaired, which is independent of the macroscopic characteristics of the tissue and the tear size. For example, Sethi et al. [42] found that the degree of tendinopathy was not correlated with the morphological appearance of the tendon, and neither of these parameters were correlated with healing or patient outcome. The authors concluded that the degree of tendinopathy, unlike muscle atrophy, may not be predictive of outcomes, and that a tendon that appears to be of poor quality may have adequate healing capacity. Therefore, abnormal gross tendon appearance should not affect the repair effort or technique. Nor, in order to compare the two techniques, they have been considered parameters, such as osteoporosis [43], comorbidities [14], genetics [44], or direction of the anchors into the greater tuberosity [20] have been found to influence tendon healing. In the last two decades, the focus has shifted from forcing the tendon to remain in contact with the bone through the use of the biomechanically more tenacious construct to optimize healing of the RC by ensuring the tendon had the vascular supply necessary for healing.

Our study demonstrated that the superficial and macroscopically vascularization of the RC is reduced when the shoulder is in extension in patients with comorbidities or lifestyle habits that predispose them to impairment of the peripheral microcirculation. Therefore, it can be postulated that deficient vascularization makes the RC tissue more susceptible to degenerative changes and, consequently, to tears. Furthermore, because the macroscopically visible vessels were thin and their blood flow was easily interrupted upon contact with a probe with little or no pressure exerted, we suggest re-evaluation of the long-term effectiveness of securing a torn RC with strong and taut sutures that may ultimately result in local ischemia and failure of the repair. Unfortunately, the available studies comparing the outcomes of the DR vs. SR techniques for RC repair have not taken into account the initial state of degeneration of the tissue to be repaired, and the findings cannot be used to support our hypothesis.

The limitations of this study included the small sample size and the subjective grading of vascularity of the shoulder joint tissues. Furthermore, no statistical correlation was observed between vascular appearance and age. As acute trauma (acromioclavicular injury) may affect the vasculature of the RC, any patient having a hemarthrosis on preoperative magnetic resonance imaging was excluded.

The vascularization of the RC is macroscopically influenced by factors than may impair peripheral microcirculation, including comorbidities (diabetes, hypercholesterolemia, uncontrolled arterial hypertension, metabolic syndrome, or obesity) and lifestyle factors.