ARA 290, a Nonerythropoietic Peptide Engineered from Erythropoietin, Improves Metabolic Control and Neuropathic Symptoms in Patients with Type 2 Diabetes
- 36 Downloads
Although erythropoietin ameliorates experimental type 2 diabetes with neuropathy, serious side effects limit its potential clinical use. ARA 290, a nonhematopoietic peptide designed from the structure of erythropoietin, interacts selectively with the innate repair receptor that mediates tissue protection. ARA 290 has shown efficacy in preclinical and clinical studies of metabolic control and neuropathy. To evaluate the potential activity of ARA 290 in type 2 diabetes and painful neuropathy, subjects were enrolled in this phase 2 study. ARA 290 (4 mg) or placebo were self-administered subcutaneously daily for 28 d and the subjects followed for an additional month without further treatment. No potential safety issues were identified. Subjects receiving ARA 290 exhibited an improvement in hemoglobin A1c (Hb A1c) and lipid profiles throughout the 56 d observation period. Neuropathic symptoms as assessed by the PainDetect questionnaire improved significantly in the ARA 290 group. Mean corneal nerve fiber density (CNFD) was reduced significantly compared with normal controls and subjects with a mean CNFD >1 standard deviation from normal showed a significant increase in CNFD compared with no change in the placebo group. These observations suggest that ARA 290 may benefit both metabolic control and neuropathy in subjects with type 2 diabetes and deserves continued clinical evaluation.
The current worldwide epidemic of type 2 diabetes constitutes one of the major unmet challenges of public health. The chronic diabetic state is characterized by hyperglycemia- and lipid-induced inflammatory processes that damage tissues in a potentially self-amplifying manner (1). While current therapy of diabetes has targeted glucose control, treatment of diabetic complications, for example, neuropathy, aside from striving for good metabolic control has been only symptomatic. An integrated therapeutic approach to diabetes would target both the underlying cause of glucose intolerance as well as activate repair systems to prevent or reverse end organ damage, that is, a disease modifying therapeutic.
Over the past 15 years, preclinical studies have identified the presence of an endogenous protective system that is activated by inflammation, metabolic stress and tissue injury (2). The receptor for this response, innate repair receptor (IRR) is a member of the type I cytokine receptor family and is a complex consisting of β common receptor (CD131) and erythropoietin (EPO) receptor subunits (3). The principal mediator of this homeostatic response is hypoglycosylated EPO produced in situ by many cells as a stress response. This locally produced EPO has been shown in diverse model systems to antagonize the production and effects of proinflammatory molecules, for example, tumor necrosis factor (TNF) (4), as well as activate healing processes. However, proinflammatory agents also attenuate the local production of EPO. In many disease states, therefore, the balance between tissue damage and healing is tipped toward injury due to inadequate EPO production.
Administration of recombinant human EPO (rhEPO) has been shown to be effective in the reduction of inflammation and activation of the healing process in a wide variety of preclinical models as well as in patients. With respect to diabetes, substantial recent work has shown that rhEPO can improve the diabetic state in animal models (5). However, in addition to interacting with the innate repair receptor in a paracrine/autocrine manner, EPO also functions as a circulating hormone via its interaction with the EPO receptor homodimer that mediates hematopoiesis. Therefore, use of rhEPO for treatment of tissue injury is limited by the potential side effects of increased erythrocyte mass and endothelial cell activation, which, in combination with activated platelet production, predisposes to thrombosis (6,7). To circumvent this problem, we have developed engineered proteins and peptides that interact only with the innate repair receptor (8,9).
One promising candidate is ARA 290, which is an 11-amino acid peptide modeled from the three dimensional structure of helix B of the EPO molecule that interacts with the IRR (8). ARA 290 has been evaluated extensively in a wide spectrum of preclinical models, including diet-induced insulin resistance (10), diabetic retinopathy (11), diabetic autonomic neuropathy (12), myocardial infarction (13), chronic heart failure (14), burns (15), traumatic brain injury (16,17) and shock-induced multi-organ failure (18), among others. The results of these studies show that ARA 290 prevents tissue injury, reduces inflammation and activates healing. However, in contrast to rhEPO that has shown side effects in clinical trials, to date ARA 290 has been shown to be safe when formally evaluated in preclinical animal toxicology, normal human volunteers and patients with sarcoidosis (19, 20, 21).
Many diseases, including diabetes, have been associated with damage and loss of small unmyelinated or lightly myelinated fibers of the peripheral sensory and autonomic nervous system, termed small fiber neuropathy (SFN). This pathological process is characterized by neuropathic pain, sensory symptoms and/or autonomic dysfunction (22). Diabetic painful neuropathy has been shown to be mediated via an inflammatory chemokine pathway (23) and strategies to reduce this inflammation, including gene transfer of soluble TNF receptor, have been shown to alleviate pain (24). In preclinical models of neuropathy, ARA 290 has been shown to prevent and improve peripheral neuropathic pain, an action that requires the innate repair receptor (25). In these effects, ARA 290 is not acting as an analgesic agent (26), but rather reduces the underlying inflammation (27) and stimulates nerve fiber regrowth from damaged axons. Recent studies of patients with sarcoidosis and SFN have shown that ARA 290 reduces neuropathic symptoms in conjunction with an increase of small nerve fibers as assessed by changes in the density of corneal nerve fibers (19). Additionally, ARA 290 has been shown to reverse diabetes-induced autonomic nerve degeneration in a murine model (12).
The present study was initiated to determine whether ARA 290 improves metabolic control and neuropathic pain in patients with type 2 diabetes.
Materials and Methods
ARA 290 is a linear 11-amino acid peptide of molecular weight 1257 daltons (21). The material for this study was produced by Bachem (Weil am Rhein, Germany) and prepared in sterile vials allowing for daily subcutaneous self-injection (SC) of 4 mg. Placebo consisted of the vehicle (20 mmol/L sodium phosphate buffer, pH 6.5, 1% sucrose and 4% d-mannitol. The results of pharmacokinetic analyses performed in normal volunteers show that following 4 mg SC, a peak plasma level ∼3 ng/mL (∼2.4 nmol/L) was obtained with a terminal half-life of ∼20 min (21). Preclinical pharmacokinetic-pharmacodynamic data has shown that beneficial effects of ARA 290 occur when plasma concentrations exceed 1 nmol/L (2).
This double blind, placebo-controlled, investigator-initiated clinical trial is registered in the Netherlands Trial Register (NTR3858). The primary purpose was to determine the safety and efficacy of ARA 290 on metabolic control and neuropathic symptoms in subjects with type 2 diabetes. Primary endpoints were (1) collection of adverse events and laboratory parameters; (2) change in hemoglobin A1c (Hb A1c) at d 28 and 56 compared with baseline; (3) change in the scores of the Small Fiber Neuropathy Screening List (SFNSL), PainDetect, Neuropathic Pain Symptom Inventory (NPSI), and RAND-36 at d 28 and 56 compared with baseline. Secondary endpoints were change in (1) quantitative sensory testing; (2) corneal nerve fiber density (CNFD); and (3) 6 Minute Walk Test (6MWT) distance at d 28 versus baseline. Primary Inclusion criteria were (1) diagnosis of type 2 diabetes mellitus; (2) spontaneous discomfort level of 6 or greater on a numerical rating scale of the “Pain Now” (0 [none] to 10 [worst]) of the PainDetect questionnaire and SNFSL score of >22, or spontaneous discomfort level on Pain Now <5 and SFNSL >44 at screening and at first dosing visit; (3) discomfort defined as distal pain/discomfort plus one of the following: (a) paresthesia; (b) burning/painful feet worsening at night; or (c) intolerance of sheets or clothes touching the legs or feet; (4) between 18 and 70 years of age; and (5) body mass index (BMI) <40 kg/m2. Primary exclusion criteria were (1) vaccination or immunization within the month prior to screening; (2) anti-TNF therapy or other biological antiinflammatory agents administered within the 6 months prior to screening; or (3) use of erythropoiesis stimulating agents within the two months prior to screening or during the trial.
The power analysis was based on data collected from a double-blind trial of the safety and efficacy of ARA 290 in patients with sarcoidosis and symptoms of small fiber neuropathy that received 4 mg ARA 290 SC daily for 28 d (19). In this study, the mean decrease in the SFNSL score for ARA 290 group at 28 d was 9.1 points with a standard deviation (SD) of 8.5. On the basis of these data, we estimated that to determine a difference between the active and placebo group of a 9-point change in the SFNSL score at a p value of 0.05 with a power of 0.9, 40 patients (20 each arm) be required. On the basis of the prior experience in diabetic populations, we also estimated that up to 10 subjects would potentially withdraw from the study. Therefore, a total of 50 patients (25 each patient group) was required to complete this study.
Baseline patient characteristics.a
92.2 ± 3.8
91.7 ± 2.7
63.0 ± 1.5
63.8 ± 1.4
175.3 ± 1.8
171.0 ± 1.6
29.9 ± 1.0
31.4 ± 0.8
ARA 290 dose (mcg/kg)
45.0 ± 1.9
ARA 290 dose (mg/ m2)
1.9 ± 0.05
3.35 ± 0.86
3.73 ± 1.17
Hb A1c(%; mmol/mol)c
7.3 ± 0.4; 55.8 ± 4.4
6.9 ± 0.2; 51.7 ± 1.9
(5.4%, 6.6%, 12.3%)
(5.8%, 6.7%, 9.3%)
4.57 ± 0.18
4.59 ± 0.15
1.22 ± 0.09
1.35 ± 0.08
2.35 ± 0.25
2.30 ± 0.27
SFNSL total score
24.8 ± 2.5 (3, 25, 52)
21.3 ± 1.7 (7, 21, 38)
NPSI total score
17.0 ± 2.0 (0, 16.5, 33.3)
21.7 ± 2.0 (4.5, 22.8, 37.5)
PainDetect total score
18.0 ± 1.4 (6, 20, 27)
18.0 ± 1.3 (6, 20, 28)
410.3 ± 17.4
447.8 ± 19.6
6MWT predicted (m)c
656.0 ± 13.8
618.4 ± 13.6
24.8 ± 1.4
24.1 ± 1.7
Subjects self-injected active or placebo (0.5 cc total volume) subcutaneously into the anterior thigh using rotating injection sites. Compliance was assessed by a count and visual inspection of vials returned to the pharmacy and checked again via a daily patient diary. Subjects also were maintained on a variety of medications for treatment of diabetes and its complications, including a wide range of drugs to treat neuropathic pain, including opioids, tricyclic antidepressants, nonsteroidal antiinflammatories, serotonin reuptake inhibitors, and/or antiepileptics. These medications were continued throughout the trial.
Safety. Clinical chemistry, hematology and vital signs were recorded at enrollment, and at 14, 28 and 56 d. A continuous recording 12-lead electrocardiogram was obtained before and for 10 min following the first administration of the study drug. Serum for anti-ARA 290 antibodies was obtained at baseline and again at the end of dosing (28 d). Subjects were interviewed in person or by telephone every 2 wks and questioned about possible adverse events. Adverse events reported spontaneously by subjects outside the above-indicated time points also were documented.
Other endpoints. Neuropathic pain was assessed using the numerical rating scales of PainDetect (28) and the Neuropathic Pain Symptom Inventory (29), two validated and widely used patient reported outcomes. The versions used in this study were validated Dutch translations. Additionally, the SFNSL was used to assess the severity of small fiber neuropathic symptoms. The SFNSL was developed for Dutch-speaking individuals and has been validated in sarcoidosis patients that have symptoms of small fiber neuropathy (30). Finally, the Dutch version of the RAND-36 was utilized to determine health-related quality of life and scoring was carried out as described by Hays, et al. (31).
For missing data of individual questions in the questionnaires, the last data point recorded was carried forward. One placebo patient was missing the baseline PainDetect value and was eliminated from the PainDetect analysis.
Quantitative sensory testing was performed for the face, hand and the foot using a Medoc Advanced Medical Systems device (Ramat Yishai, Israel) in accordance with the German Research Network on Neuropathic Pain protocol (32), but was modified to include only six of the thirteen domains to reduce the amount of time subjects would be involved in testing. These domains were: cold detection threshold (CDT), warm detection threshold (WDT), thermal sensory limen (TSL), paradoxical heat sensation (PHS), allodynia (ALL) and vibration detection threshold (VDT). Normative values used were those of Rolke et al. (33).
Corneal nerve fiber images were obtained using a Rostock Cornea Module/Heidelberg Retina Tomograph III and established methodology (36). A minimum of six images from each eye were selected by a single experienced investigator blinded as to treatment, based on the presence of adequate nerve fiber contrast and full field of view. CNFD (number/mm2) was manually quantified by a blinded expert analyst according to our established criteria and evaluation methods (37) For determination of normal corneal nerve fiber density and length, 55 healthy individuals ranging in age from 28 to 73 years (median = 48 years) were evaluated.
Data was analyzed using JMP (version 11; SAS, Cary, SC, USA). All variables were considered to be continuous and confirmed to follow a normal distribution before parametric statistical methods used. Statistical testing was performed as delineated in the text. For corneal nerve density, some subjects fell within the normal range. In addition to a comparison of the full subject group, a post hoc analysis was carried out for the group of subjects having a CNFD >1 SD less than the normal mean.
No clinically significant changes from baseline values of the hematology or clinical chemistry were observed at any of the time points sampled following ARA 290. Anti-ARA 290 titers at baseline and after 28 d were negative. The frequency distribution of AEs was similar in both treatment groups: for ARA 290, 54 “mild,” 9 “moderate” and 1 “severe” categorizations were recorded, while for placebo, 61 “mild” and 5 “moderate” were noted. Potential relationship to administration of study drug as assessed by the investigator was: 25 “possibly” and 39 “unlikely” for the ARA 290 group, and 40 “possibly” and 26 “unlikely” for the placebo group.
Four serious adverse events were observed in the ARA 290 treatment arm. Two were judged unlikely to be associated with ARA 290 administration. The remaining two were judged to be possibly related to ARA 290 administration. In one subject on daily furosemide therapy, the dose was increased after wk 2 of the dosing period and the previously borderline renal insufficiency worsened slightly (creatinine rose from 119 µmol/L to 159 µmol/L). The safety committee stopped ARA 290 administration. However, furosemide administration continued at the increased dose and renal function did not improve over the follow up period of 6 wks. Another subject, male 70 years of age, developed severe cellulitis of the lower extremity requiring hospitalization 2 wks after the last dose of ARA 290 and suffered a fatal myocardial infarction. The safety committee judged the event to be unrelated to ARA 290 treatment.
Diabetic therapeutic regimens varied in the patient population, ranging from no pharmacological treatment to intensive insulin therapy. The majority of subjects experienced no change in medication during the dosing or 1-month follow-up period.
NPSI. The NPSI at baseline indicated mild symptoms (score of 19.4 ± 1.4). Both groups improved significantly by ∼20% compared with baseline at d 28 (mean decreases of ARA 290 = −3.9 ± 1.9 and placebo = −3.7 ± 1.5; p = 0.05 and p = 0.02 respectively; paired t test). By d 56, neither group was significantly improved compared with baseline (ARA 290 = −1.8 ± 2.3 and placebo = −3.1 ± 2.2).
SFNSL. Similar to the NPSI, the baseline total score (out of a maximum of 84) of the SFNSL was in the lower end of the range: 24.8 for the ARA 290 and 21.3 for the placebo group respectively. Neither the ARA 290 nor placebo treatment group changed significantly from baseline for the SFNSL when assessed at d 28 or d 56.
6MWT. All subjects performed a 6MWT at baseline and d 28. The predicted 6-min walk distance calculated from Troosters et al. (35) was a mean of 637 m, whereas the mean 6-min walk distance for the patient cohort was only 429 m. Following dosing, the ARA 290 group increased their walk distance by a mean of 3.5% of baseline (p = 0.09), whereas the placebo group increased walk distance by 1.1 % of baseline (p = ns [not significant]).
The results of this study show that ARA 290 (4 mg SC) self-administered for 28 d by subjects with type 2 diabetes is not associated with significant ARA 290-related adverse events. Similarly, no clinically significant ARA 290-related alterations of baseline biochemical (electrolytes, calcium, liver, kidney or pancreas) or hematological (erythrocyte, thrombocyte or leukocyte) parameters were observed. In sum, the safety observations support the concept that the IRR, the target of ARA 290, is induced only by injury, inflammation or metabolic stress and is therefore not generally expressed by normal tissues, thus limiting systemic adverse events.
As predicted by preclinical studies with EPO (5) and ARA 290 (10,38), subjects receiving ARA 290 exhibited significant improvement in glucose control as shown by changes in Hb A1c concentration. It is notable that the beneficial effect of ARA 290 appears to be sustained for at least the 28-d follow-up period. Although the magnitude of the A1c change is small, it is consistent with the expectations for antidiabetic agents administered to subjects having excellent baseline control, and is similar to what has been observed in previous clinical trials evaluating antidiabetic agents (39). ARA 290 also benefited the lipid profile, with an improvement in the cholesterol-HDL ratio, in part, as a result of an increase in HDL levels. Additionally, although not performed on fasting serum samples, the triglyceride concentration also improved in parallel to A1c. The mechanism(s) of action underlying these changes is currently unclear, but results of preclinical work supports effects of ARA 290 on insulin resistance and muscle mitochondrial biogenesis (10) as well directly on insulin release (38).
The enrollment criteria for this trial required painful neuropathic symptoms of the extremities consistent with small fiber neuropathy. The results of the quantitative sensory testing performed are consistent with a mixed small and large fiber neuropathy. Furthermore, the reduction in corneal nerve fiber density (CNFD) confirms the presence of a structural small fiber neuropathy.
The sensory-related neuropathic symptoms, as assessed by PainDetect, are notable for being quite severe in these well-controlled subjects and are comparable to other trials using this instrument in diabetic subjects (40). Results of a recent study (41) show that this instrument has excellent repeatability and can be used to assess neuropathic symptoms longitudinally. In the current study, ARA 290 significantly improved the total score compared with the placebo group. Changes noted in the scores of individual questions show that symptoms indicative of small fiber function are improved.
Subgroup analysis of those subjects having reduced corneal nerve fibers, and therefore more definitive evidence of small fiber neuropathy (37), show that ARA 290 treatment is associated with an increase in corneal nerve fiber density, analogous to the improvement seen after improved metabolic control (42) and after simultaneous pancreas and kidney transplantation (43). Additionally, the observation that changes in symptoms as assessed by PainDetect are correlated with changes in the CNFD in the ARA 290 treatment group alone, further support a functional relationship between small fiber density and neuropathic symptoms. It is interesting, however, that evaluation of small fiber function as performed by QST (that is, thermal thresholds) did not show significant changes following ARA 290 administration. This is in contrast to a previous trial in sarcoidosis where changes were observed in several temperature-related evaluations (19). Further study is needed to determine the underlying reason for these differences.
Finally, quality of life, as assessed by RAND36, was similar in most dimensions to a large control population with chronic disease in spite of the excellent glycemic control. It is interesting that the changes observed in the ARA 290 group occurred primarily in the vitality (energy/fatigue) dimensions. Fatigue has been observed to be a prevalent and clinically important symptom and major cause of functional disability of subjects with sarcoidosis and SFN (44). It is likely that fatigue is an important component of the symptom complex of these diabetic subjects, which was not formally assessed in this trial.
In summary, ARA 290 shows significant potential for the treatment of diabetic small fiber neuropathy. The two major limitations of this trial are that the treatment duration was only for 28 d and the subjects were under generally excellent metabolic control. Longer duration of therapy provided to poorly controlled subjects could result in larger changes in Hb A1c as well as more substantial changes in neuropathic symptoms and objective assessment of small nerve fiber density and function. The observed excellent safety profile, along with the positive results presented herein, support further evaluation of ARA 290 as a disease modifying agent in subjects with painful diabetic neuropathy.
M Brines, AN Dunne, and A Cerami are officers of Araim Pharmaceuticals and own stock and/or stock options.
M Brines designed the trial, analyzed data and wrote the manuscript. AN Dunne designed the trial, oversaw subject recruitment and trial management and wrote the manuscript. M van Velzen conducted clinical aspects of the trial, researched data and participated in the data analyses. PL Proto conducted clinical aspects of the trial, researched data and participated in the data analyses. C-G Ostenson participated in interpretation of the data. RI Kirk contributed to preparation and management of the clinical trial. IN Petropoulos and S Javed undertook quantification of corneal nerve morphology. RA Malik oversaw the CCM quantification and interpretation. A Cerami designed the trial, participated in data analysis and wrote the paper. A Dahan was the Principal Investigator and was involved in all aspects of the trial. All authors contributed to discussion, reviewed and edited the manuscript, as well as approved its final form.
This work was supported in part by a grant from the Dutch government to the Netherlands Institute for Regenerative Medicine (NIRM, grant no. FES0908, the Swedish Research Council, ALF, and the Swedish Diabetes Association). The authors thank Ferdinand C Breedveld and Geertrui Betgen for invaluable assistance, as well as the subjects and their families for agreeing to participate in this trial.
- 34.(2002) ATS statement: guidelines for the six-minute walk test. Am. J. Respir. Crit. Care Med. 166:111–7.Google Scholar
- 36.Tavakoli M, Malik RA. (2011) Corneal confocal microscopy: a novel non-invasive technique to quantify small fibre pathology in peripheral neuropathies. J. Vis. Exp. doi: 10.3791/2194.Google Scholar
- 38.Muller C, et al. (2013) The nonhematopoietic erythropoietin analogue ARA 290 improves glucose tolerance by stimulating insulin secretion in spontaneously type 2 diabetic Goto-Kakizaki rats. Diabetologia. 56(Suppl 1):S268.Google Scholar
- 41.Baron R, et al. (2013) Test and re-test of the painDETECT-Questionnaire. Presented at: Proceedings of the 4th International Congress on Neuropathic Pain; 2013 May 23–26; Toronto, Canada.Google Scholar
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, and provide a link to the Creative Commons license. You do not have permission under this license to share adapted material derived from this article or parts of it.
The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this license, visit (https://doi.org/creativecommons.org/licenses/by-nc-nd/4.0/)