Beta-adrenergic antagonist for the healing of chronic diabetic foot ulcers: study protocol for a prospective, randomized, double-blinded, controlled and parallel-group study

Diabetic foot ulcers (DFUs) account for significant morbidity and immense biomedical burden. Aside from the management of multiorgan comorbidities in diabetic patients, DFUs alone substantially impact our health care system with both economic and psychosocial effects. A diabetic patient has a 25% lifetime risk of developing a DFU [1]. As the DFU becomes intractable, the patient’s quality of life and productivity are considerably affected [1‐3]. There is a one in six risk that patients with a DFU will have an amputation, with a 66% risk of recurrence and a 47% increase in mortality, as high as colon cancer [3‐5]. Notably, according to the US Census Bureau, there are 21.8 million veterans in the US, and nearly 25% have diabetes compared to 10.5% of the general population [6‐10]. A study involving veterans with diabetes and a foot ulcer noted that there was a 2.39 increased relative risk of death compared to those without a foot ulcer [11]. Preventative measures are taken to reduce the burden of diabetes and aggressive treatment is used for the ulcers before they advance to amputation.

Health care systems invest enormous sums of money in improving the healing of DFUs, with an estimated annual cost for DFU treatment in the US exceeding $10.9 billion [12]. Several advanced treatments for DFUs exist, but they are expensive, difficult to use in the clinic or at home, and have shown limited success in the healing of DFUs. The cost per episode of DFU treatment can easily exceed $38,000 plus associated expenses related to hospitalization, ulcer recurrences, amputations, home health care, dressings, decreased productivity and premature disability, social isolation, and depression [1‐3, 8, 13].

The standard of care (SOC) for a DFU generally consists of the debridement of necrotic tissue, application of a moist dressing, and the use of offloading devices (such as foot orthoses, total contact cast and/or with the use of crutches, wheelchair, scooter, or other assistive ambulatory devices) that protect the wound from pressure or trauma related to ambulation and other acts of daily living, and management for infection if indicated [2, 3, 13‐18]. Nevertheless, despite wound specialists adhering to the best standard wound care regimens, only 31% of DFUs heal after 20 weeks of care [18, 19]. Such an unfavorable cure rate has prompted increased research for therapeutic alternatives and novel approaches to optimize wound healing, particularly for the growing Veterans Affairs (VA) diabetic population. Therefore, preventing further complications and the high costs associated with DFU treatment is critical. Thus, we are investigating using a safe, inexpensive, well-characterized, easy to use drug that could be implemented system-wide to enhance DFU repair and may have far-reaching consequences.

In prior work our laboratory has examined the effects of beta-adrenergic antagonists (βAA) on skin and skin-derived cells. Of the several classes of adrenergic receptors, predominant expression of the β2 subtype has been identified on major cell types of the skin, including human keratinocytes, melanocytes, and dermal fibroblasts [20‐23]. Keratinocytes can also synthesize catecholamines such as epinephrine and norepinephrine [20‐24], in essence creating a self-contained, catecholamine signaling network. The functional role of this network has been elusive. Our work suggests that it contributes to the control of cell migration, and thus to skin wound healing [25‐28].

When skin is wounded, repair mechanisms are activated to restore skin integrity. The repair process involves the orchestration of interactions between cellular components, growth factors, chemokines, and extracellular matrix proteins that regulate the migration and proliferation of the keratinocytes into the wound [29‐31]. This directional migration of keratinocytes into the wound is critical for repair and reestablishment of epithelial coherence. Our laboratory work has shown that activation of βAR by stress-catecholamine agonists, such as epinephrine and norepinephrine, decreases keratinocyte migration, decreases the ability to heal an in vitro scratch wound, and impairs healing of acute wounds in animal models [23, 27, 28, 31‐34]. Importantly, these βAR agonists are found in significant concentrations in human DFU tissues [28].

The natural corollary to the finding of stress catecholamine βAR ligands within the wound environment that can impair pro-reparative functions of skin-derived cells [23‐28, 30, 31, 35] was to determine the effects of blocking their action. Importantly, and specifically relevant to our proposed clinical trial, we have shown that blockade of the βAR with antagonists improves healing in vitro and in animal models [24, 26, 28]. Work by other investigators also supports the hypothesis that βAR antagonists can improve DFU healing. Collagen synthesis (in a pulmonary injury model) has been observed to be increased by βAR antagonists [36]. More specifically, Gulcan and colleagues demonstrated that the topical application of βAR antagonists to wounds in diabetic rats improves not only the rate of healing, but wound vascularity [37]. Indeed, a patent application on the use of a βAR antagonist for the healing of DFUs has been filed by other investigators [38], albeit with no human clinical data. Our goal with this clinical trial is to generate these unequivocal data and to demonstrate the safety and efficacy of this approach to improve healing in human chronic DFUs.

Here we propose a clinical translational study, moving our original laboratory research into an early clinical trial to test this hypothesis. Therefore, the main aim of this proposal is to establish timolol ((S)-1-[(1,1-dimethylethyl)amino]-3-[[4-(4-morpholiny)-1,2,5-thiadiazol-3-yl]oxy]-2-propanol (Z)-2-butenedioate(1:1) (salt)), a non-selective beta-adrenergic antagonist, as a novel therapeutic alternative in response to the challenging clinical management of DFUs. We expect this study will demonstrate that timolol, a low-cost therapy, improves the rate of wound healing, which in turn will have tremendous long-term benefits, improving morbidity, quality of life, as well as the negative psychological and social issues associated with DFUs.

This is a clinical translation study, moving the investigators’ pre-clinical laboratory research into the first randomized clinical study. Our previous work has demonstrated that stress-catecholamines such as epinephrine and norepinephrine, βAR agonists, impair wound healing as evidenced by decreased keratinocyte migration and a decrease in keratinocytes’ ability to heal wounds in vitro [23, 27, 28, 31, 33, 34]. Interestingly, not only can stress-induced elevation of systemic catecholamines impair healing [69, 70], but as we and others have shown, the keratinocytes themselves can generate epinephrine [25, 71] and significant levels of catecholamines are present within the immediate vicinity of the wound [24]. Conversely, our laboratory investigations have shown that βAAs increase keratinocyte migration speed (a surrogate for epithelial healing), the healing of scratch wounds in confluent keratinocyte cultures [31], and wound epithelialization in vivo in a number of animal wound models [24, 26, 28, 72]. Additionally, in a human skin ex vivo burn wound model, cultivation with βAA (timolol) improves wound re-epithelialization relative to the culture medium control. Further, our studies have shown that catecholamines increase neutrophil dwell time in the wound, delaying healing in an animal wound model, and treatment with βAA reverses this, restoring healing time to normal [73]. We and others have also documented and reported on cases of venous leg ulcers and other chronic wounds in humans that have improved healing after topical timolol application [38, 61, 62, 67]. These multiple laboratory studies and anectdotal clinical reports indicate that βAAs (timolol) promote wound healing and shorten time to healing; thus, a clinical trial is the next logical step to determine if this approach is truly efficacious in the clinical setting of DFUs.

We hypothesized that subjects with DFU treated with a βAA (timolol) will have more rapid complete wound closure compared to those in the SOC group. The null hypothesis is that there is no difference in the proportion of subjects with complete wound closure of DFUs between the two treatment arms.

The Cochrane Wounds Group has indicated that the healing rates and healing time commonly reported in the majority of wound studies and clinical trials area deemed appropriate primary outcomes [74]. In fact, most studies have demonstrated that the proportion of ulcers healed at 12 weeks of treatment is an appropriate outcome measure [18, 19]. Thus, for our study purposes, the primary endpoint is complete ulcer closure as defined by “skin re-epithelialization without drainage or dressing requirements” by week 12 (end of the active phase of the study).

We will also assess secondary endpoints that will focus on measurements obtained throughout the clinical trial, including change in percentage reduction in wound size between the two study arms, calculated as difference in cm² of the randomization measurement after 12 weeks of the active phase and after the follow-up phase. We will consider patient’s quality of life using the VR-36 Health Survey, the Lower Extremity Functional Scale [39], and the Charlson Comorbidity Index [40].

AEs will be assessed for safety parameters that specifically include SAEs, hospitalization, study exit, unanticipated adverse device effects, and therapy-related incidences. We will analyze and compare the data between these two treatment arms, including AEs at an incidence of 1% or greater, adverse reactions to the test therapy at an incidence of 1% or greater, incidence of immediate reactions, incidence of local and systemic reactions, incidence of osteomyelitis, incidence of wound infection, and incidence of SAEs or unanticipated adverse device effects. Potential confounding variables include Basic Metabolic Index (BMI), wound size (which we will adjust for using the subcategories < 0.4 cm², > 0.5–0.9 cm², > 1.0–1.9 cm², > 2.0–2.9 cm², > 3.0 cm²), as well chronicity of the wound by year (adjusted for using < 1 year, > 1–3 years, > 3–5 years, > 5 years, > 10 years). Safety data collected from the clinical trial will significantly contribute toward establishing the safety profile of topical βAA use in DFUs.

There may be several potential limitations to this study involving study subjects’ demographics. As this study occurs solely at veteran medical centers, there may be bias for recruiting mainly veteran male patients. Compared to the US general population (49% male, 50% female), females represent only 9% of the veteran population nationwide [9, 75]. Similarly, there may also be bias toward recruiting middle-age, elderly, and white veterans at the Veteran Health System [76], but it should be noted that DFUs are more common between the ages of 60 and 80 years [2, 4, 5, 8, 14‐16, 55, 77].

Protocol date and version: December 5, 2019, version 4.

Patient recruitment began in August 2018 at the Sacramento VA Medical Center, Mather, CA, USA. Study enrollment and recruitment are estimated to be completed within 4 years.