Effect of angiotensin converting enzyme inhibitors and angiotensin II receptor blockers on need for operative intervention for idiopathic adhesive capsulitis

Background The exact pathogenesis of idiopathic adhesive capsulitis (IAC) is not fully understood, but an inflammatory profibrotic cascade, largely mediated by transforming growth factor-beta 1 (TGF- β1) has been implicated. Angiotensin II receptor blockers (ARBs) and angiotensin-converting enzyme inhibitors (ACE-Is) both decrease the activity of TGF-β1. The aim of this study was to determine the impact of ACE-Is or ARBs use on the need for operative intervention in IAC. Methods This was a retrospective cohort study of patients from a single institutional database with IAC, divided into two cohorts, with and without ACE-I and/or ARB use as the primary exposure and a minimum 2-year follow-up. The primary outcome measured was the incidence of operative intervention including manipulation under anesthesia (MUA) and arthroscopic capsular release (ACR). Additional multivariable logistic regression analysis was performed to evaluate associations between ACE-I/ARB use and likelihood of undergoing an operative procedure. Results A total of 17,645 patients met inclusion criteria, with 5424 patients in the ACE-I/ARB cohort and 12,221 in the non-ACE-I/ARB cohort. Overall, 422 (2.4%) patients underwent surgical treatment, 378 (2.1%) ACR, and 74 (0.4%) MUA. There was no significant difference between cohorts in the frequency of surgical procedures or time to procedure since diagnosis. There were no significant differences between individual ACE-Is or ARBs, although Losartan was found to have a trend of decreased rate of intervention (31.7% vs. 36.8%, P = .209) when compared to patients not on losartan that did not reach statistical significance. Patient factors predictive of undergoing MUA/ACR were diabetes (P = .013), obesity (P < .001), and male sex (P < .001). Increasing patient age reduces the likelihood of undergoing operative intervention, with patients aged 50-70 years (P = .022) and age >70 years (P < .001) demonstrating reduced odds as compared to patients aged <30 years. Conclusion Patients with IAC have an overall low (2.4%) rate of requiring surgical intervention. While the antifibrotic mechanism of ACE inhibitors and ARBs did not significantly affect the rate of requiring surgical intervention, male gender, obesity, younger age, and diabetes, all increased the risk for operative intervention. Losartan, specifically, may have a disease modifying effect on IAC that should be investigated with larger controlled trials.

inflammatory profibrotic cascade resulting in intense synovitis and capsular thickening.Among numerous proinflammatory mediators, transforming growth factor-beta 1 (TGF-b1) is known to play a key role. 17,24,28,30,37,40,41Angiotensin II receptor blockers (ARBs) and angiotensin-converting enzyme inhibitors (ACE-Is), commonly used antihypertensive agents in the renin-angiotensin-aldosterone system, have demonstrated promising results as antifibrotic agents due to their downstream effects of blocking TGF-b1 in nephrology and also more recently in surgical specialties such as plastic surgery and scar formation. 12,14,21,35,38In orthopedics, ACE-Is and ARBs have recently gained attention for their anti-TGF-b1 effects in decreasing fibrosis and promoting cartilage healing. 20,31,33,39However, no study to our knowledge has investigated the effects of ACE-Is and ARBs on the natural history of IAC.
The purpose of this study was to compare IAC patients who were prescribed ACE-Is or ARBs with those not on those medications to identify if there were any differences in need for surgical treatment, including MUA or ACR.We hypothesized that the prediagnosis use of ACE-I/ARB medication would be associated with a decreased need for surgical treatment at final follow-up.

Study design
In this retrospective cohort study, patients diagnosed with primary IAC of the shoulder (International Classification of Diseases 10 code M75.0,International Classification Disease 9 code 726.0) between January 2010 and December 2020 were identified from a single institutional database.Inclusion criteria for this study were a diagnosis of IAC, aged 18 years or more at time of diagnosis, and minimum 2-year follow-up.Exclusion criteria consisted of patients who had prior ipsilateral shoulder surgery, those who underwent multiple shoulder procedures during the study period or those who were prescribed ACE-I or ARB after surgery for IAC as described within the single institutional database.Patients were divided into two cohorts using ACE-I and/or ARB use as the primary exposure.The primary outcome measured was rate of operative intervention in form of MUA (Current Procedural Terminology code 23700) and/ or ACR (Current Procedural Terminology 29825) following diagnosis of IAC.Covariates of interest included patient age, sex, race, smoking status, body mass index (BMI), and comorbidities (hypertension, diabetes, rheumatoid arthritis, and hypothyroidism).

Statistical analysis
A priori power analysis was performed with G*Power Version 3.1.9.7 (Heinrich Heine Universit€ at, Düsseldorf, Germany) using a 2.5% difference in procedure rates as the threshold for clinical significance.With the goal of achieving a minimum of 80% power at an alpha equal to 0.05, it was determined that 714 patients were needed per cohort.
Descriptive statistics were stratified by concomitant ACE-I/ARB use.Chi-squared or Fisher's exact test was used for analysis of categorical variables and two-sample t-test were used for analysis of continuous variables.Categorical variables were reported as frequencies (%) and continuous variables were reported as mean (standard deviation).Multivariable logistic regression analysis was performed to evaluate associations between ACE/ARB use and likelihood of undergoing MUA and/or ACR procedures in patients diagnosed with IAC.All covariates of interest were included in each model to control for confounding.The odds ratios (ORs) and 95% confidence intervals (CIs) were reported.All statistical analysis was performed in Jupyter Notebook Version 6.4.8 (Project Jupyter, New York, NY, USA) using Python programming language.For all analyses, a P value of .05 was considered to be statistically significant.

Individual ACE-I/ARB medication use
Of those taking ACE-I/ARB medication, 2487 (45.9%) were on an ACE inhibitor, while the remaining 2904 (53.5%) were on an ARB.Increasing patient age appears to reduce the likelihood of undergoing MUA and/or ACR, with patients aged 50-70 years (OR 0.44, CI 0.22-0.89,P ¼ .022)and aged >70 years (OR 0.10, CI 0.04-0.25,P < .001)demonstrating reduced odds as compared to patients aged <30 years.A comprehensive comparison list of factors and their associated ORs are available in Table III and Figure 1.

Discussion
The main finding of our study was that ACE-I or ARB use did not have a significant impact on reducing the incidence of surgical intervention (MUA and/or ACR) in patients with IAC.ACE-Is and ARBs are believed to have an antifibrotic mechanism through TGF-b1 modulation in addition to their antihypertensive effect through the renin-angiotensin-aldosterone system; 6,12,14,20,35 however, the literature within orthopedics remains in nascent stages.Additionally, we demonstrate that male sex, younger age, obesity, and diabetes were all associated with an increased likelihood of requiring surgical intervention in IAC.
Although the antifibrotic effect of ACE-Is/ARBs has been demonstrated in preclinical models, 20,29,33,39 the results in orthopedic literature have been mixed.Three recent studies have investigated the effect of ACE-Is/ARBs on the incidence of postoperative arthrofibrosis, ROM, and manipulation in total knee arthroplasty. 2,18,25None of the three studies demonstrated significant improvement in postoperative ROM, reduction in need for MUA, or decreased revision rate.The most recently published study by Arraut et al evaluated 79 patients undergoing primary total knee arthroplasty who were prescribed losartan at least three months prior to surgery, matched with a control group of patients who were not taking losartan.Postoperative ROM and change in ROM, rates of readmission, manipulation for stiffness, or all-cause revision were not different between groups. 2 In our study, the lack of difference in rates of operative intervention could be due to multiple reasons.First, the antifibrotic dosage may be different than the antihypertensive dosage for ACE-Is/ARBs.Second, although we did a preliminary power analysis, we may need a larger sample size to demonstrate a clinically significant difference.Finally, while IAC is characterized by fibrosis and contracture of the glenohumeral joint capsule with high expression of TGF-b1, high levels of matrix metalloproteinases, interleukin-1, tumor necrosis factor alpha, and cyclooxygenase (COX1 and COX2) have also been demonstrated to have abnormal expression in IAC tissues and may continue to promote a profibrotic cascade despite the anti-TGF-b1 effects of ACE-Is and ARBs. 17,28,30ultiple animal models have demonstrated losartan's ability to inhibit fibrosis, improve skeletal muscle regeneration, and decrease post-traumatic joint capsule stiffness through its antagonism of TGF-b1. 4,9,12,23While it did not reach statistical significance, our study demonstrated that patients taking losartan had a trend of decreased rate of surgical MUA and/or arthroscopic release.Losartan is an ARB and prevents direct activation of TGF-b1 as well as the angiotensin II-induced phosphorylation of Smad2 and Smad3, which translocate into the nucleus of the cell and lead to additional increased transcription of TGF-b1, procollagen, and fibronectin. 32osartan specifically has been evaluated in the orthopedic literature for its antifibrotic properties and its mechanistic differences when compared to ACE-I and other ARBs. 7,15,41,42n this study, we demonstrate that male sex, younger age, obesity, and diabetes were all associated with an increased likelihood of requiring surgical intervention in IAC.Levine et al reported that patients with more severe initial symptoms, younger age at onset, and reduction in ROM despite 4 months of therapy were more likely to require surgical intervention. 27ultiple additional studies have demonstrated an increased rate of surgical intervention in younger patients. 1,34However, it is difficult to ascertain if this is due to more severe unremitting symptoms or if younger patients are less likely to tolerate smaller decreases in ROM and/or surgeons are more aggressive in indicating younger patients for surgical intervention.Some studies have similarly shown an increased rate of surgical intervention among men, as demonstrated in our results, but this is not replicated throughout the literature. 16,34Diabetes is also a significant risk factor for developing adhesive capsulitis and increases the likelihood of requiring surgical management. 3,16,31Interestingly, BMI has not previously been associated with frozen shoulder, and a recent study investigating the effects of diabetes and BMI on frozen shoulder found no significant difference in surgical requirement or overall outcomes based on BMI more than 30. 5 Our study demonstrated that obese IAC patients were 1.62 times more likely to undergo surgery.

Limitations
The retrospective nature of this study introduces the inherent possibility of selection bias and errors in the entered data.Additionally, no ROM outcomes were recorded, preoperatively or postoperatively.This is important as patients with worse presenting ROM have been shown to be more likely to require surgical intervention.Also, while only a small percentage of patients required surgical intervention, it is possible that ACE-Is/ ARBs improved ROM without significantly impacting the need for operative intervention.The dosage of ACE-I/ARB medication may also have played a role in the lack of significant findings, as all patients take just their prescribed dose of antihypertensive medication, which was not standardized among patients.However, given there are no clear optimal dosage requirements for the antifibrotic effects of these medications, we felt this lack of dosage recorded would not impact the generalizability of our results significantly.Finally, patient compliance with their ACE-I/ ARB medication was also not recorded, which while certainly a limitation, we felt this increased the external validity of the study results, as when investigating such a large cohort of more than 17,000 patients and applying the results to the general population, patients will have varying degrees of compliance in clinical practice.

Conclusion
Patients with IAC have an overall low (2.4%)rate of requiring surgical intervention.While the antifibrotic mechanism of ACE-Is and ARBs did not significantly affect the rate of requiring surgical intervention, male gender, obesity, younger age, and diabetes all increased the risk for operative intervention.Losartan, specifically, may have a disease modifying effect on IAC that should be investigated with larger controlled trials.

Disclaimers:
Funding: No outside funding or grants were received in support of the completion of this study.Conflict of Interest: The authors of this paper certify that they have no affiliations with or involvement in any organization or entity with any financial or nonfinancial interests pertinent to the subject matter discussed in this manuscript.

Figure 1
Figure 1 Odds ratios and 95% confidence intervals for factors predictive of undergoing MUA ± capsular adhesion lysis in patients with adhesive capsulitis.ACE, angiotensinconverting enzyme inhibitors; ARB, angiotensin receptor blocker; MUA, manipulation under anesthesia.
ACE-I, angiotensin-converting enzyme inhibitors; ARB, angiotensin II, receptor blockers; MUA, manipulation under anesthesia; SD, standard deviation.On average, the ACE-I/ARB cohort had been on their respective medications for an average of 28.4 ± 29.0 months prior to being diagnosed with IAC.A comprehensive list of medications is available in TableII.The most commonly prescribed medications were Losartan (1989 [36.7%]),Lisinopril (1413 [26.1%]), and Enalapril (601[11.1%]).Although there were no significant differences between individual ACE-Is and ARBs with respect to their impact on reducing the prevalence of operative intervention, but Losartan demonstrated maximum impact (31.7% vs. 36.8%,P ¼ .21)when compared to patients not on losartan.