Diabetes-Impaired Wound Healing Predicted by Urinary Nitrate Assay - A Preliminary, Retrospective Study
Joseph V. Boykin, Jr., MD, Lisa G. Shawler, RN, CCRC, Vicki L. Sommer, RN, Mary C. Crossland, RN, CWCN, Columbia Retreat Hospital Wound Healing Center, Richmond, VA; John E. Kalns, PhD, Davis Hyperbaric Laboratory, Brook Air Force Base, San Antonio, TX.
Wounds 11(3):62-69, 1999. © 1999 Health Management Publications, Inc
Abstract and Introduction
Diabetic wound healing is occasionally impaired and associated with chronic foot ulceration and lower extremity amputation (LEA). Topically applied platelet-derived growth factor (rhPDGF-BB) increases healing in less than 50 percent of these cases suggesting that other factors are involved. In a preliminary, retrospective study of diabetic ulcer patients receiving rhPDGF-BB (becaplermin), we show that urinary concentrations of the nitric oxide (NO) metabolite nitrate are significantly (p<0.05) reduced in patients with poor or absent ulcer healing as compared to patients with healed ulcers or non-diabetic controls. This suggests that significantly decreased endogenous NO production predicts diabetic wound outcomes in patients treated with becaplermin. These preliminary findings also appear to support further investigations of the clinical use of urinary nitrate levels as a tool to guide therapy during treatment of diabetic ulcers.
Diabetes is diagnosed in three percent of the United States population and affects an estimated 15 million people. Within this diabetic population are individuals with chronic, non-healing lower extremity ulcerations (LEU) that are associated with significant morbidity and treatment costs. Chronic, non-healing LEU precedes about 85 percent of the lower extremity amputations (LEA) that are performed on over 50,000 diabetic patients annually. While only six percent of diabetic hospitalizations are associated with LEU, the total average government reimbursement for diabetic lower extremity complications, not including amputation and rehabilitation costs, exceeded $1.5 billion in 1992. While the majority of diabetic patients exhibit slower but otherwise normal wound healing, those presenting with chronic LEU often demonstrate decreased wound inflammation, recurrent wound infections, decreased cutaneous vascular perfusion, and poor wound collagen deposition and scar maturation. Platelet-derived growth factor (PDGF) deficiency has been identified as a local factor of the chronic diabetic ulcer that contributes to impaired healing. Topical application of becaplermin restores wound PDGF levels to those associated with normal healing; however, this treatment is effective in less than 50 percent of cases suggesting that other factors are involved.
The emerging importance of NO in inflammation, tissue repair, and microvascular homeostasis,[4-6] and the observation of reduced NO production in diabetic wounds[7,8] suggest a central role for NO in the pathogenesis of chronic, non-healing LEU. Specifically, we hypothesize that NO production is reduced in the non-healing diabetic wound and that topical becaplermin therapy is effective only when NO production deficiency is corrected. We further hypothesize that below a critical or threshold level of endogenous NO production, LEU repair may not be achieved. Under this hypothesis, diabetic patients with chronic, non-healing LEU responsive to becaplermin therapy should have substantially increased NO production compared to those that fail. We examined this hypothesis by measuring plasma and urinary NO metabolites -- nitrate and nitrite -- in diabetic patients treated for LEU with becaplermin..
Study Design and Methods
Patient Selection and Outcome Evaluation
For our preliminary clinical study, ten healthy diabetic patients presenting with a history of one or more diabetic foot ulcers were chosen. All patients had previously received topical ulcer treatment with becaplermin gel under close clinical observation. Half of this group (n=5) experienced complete healing (healed diabetic patients/HD) of the ulcer by week 20 of observation. The remaining half (n=5) of this group had not experienced complete healing (unhealed diabetic patients/UHD) of the ulcer by week 20 of observation. Following the completion of becaplermin treatment, the ten diabetic subjects (HD and UHD) and ten healthy, non-diabetic controls (C) were enrolled for urine and plasma nitrate/nitrite analysis. Prior to this analysis all subjects were screened with a medical history, physical examination, and baseline hematology and serum and urine chemistry in order to eliminate subjects with active malignant disease, rheumatic or collagen vascular disease, inflammatory bowel disease, alcohol/drug abuse, cellulitis, osteomyelitis, and those requiring revascularization surgery. Patients with diabetic nephropathy (serum creatinine > 2mg/dl), severe heart failure, or any other serious illness were excluded from the study. For compliancy reasons, subjects determined to exhibit poor diabetic control were disqualified. Additionally, anyone receiving radiation therapy, systemic corticosteroids, and immunosuppressive or chemotherapeutic agents was disqualified. Informed consent was obtained from all study participants. The ten healthy, non-diabetic, control patients (C) were observed with the following clinical parameters [average value (±SE)]: age (years) - 39.4(±4.35); male/female (M/F) ratio - 5/5; weight (kg) - 84.2(±7.6); BMI (kg/m2) - 29.5 (±2.0); and serum creatinine (mg/dl) - 0.98 (±0.04). The diabetic ulcer study groups, HD and UHD, were matched similarly for age, M/F ratio, type of diabetes, initial ulcer size, weight, BMI, HgA1c, serum creatinine and creatinine clearance (Table 1). All physiological clinical parameters were within normal limits for the groups evaluated, and no significant (p<0.05) differences were observed between the diabetic ulcer study group values (unpaired t-test). Both study groups, HD and UHD, consisted of diabetic neuropathic patients with a history of foot ulceration but no peripheral vascular disease (PVD). None of the study patients were observed with claudication and/or abnormal noninvasive vascular tests. All ulcers were surgically debrided prior to becaplermin application.
Subjects from all groups were brought into the hospital environment after having fasted for a ten-hour period of time. Fasting urine and plasma samples were obtained from each subject upon admission (Day 1) to provide an indication of the subject's baseline nitrate and nitrite levels. Additionally, routine labwork consisting of chemistry panel, CBC, and urinalysis was obtained from all subjects upon hospital admission. Subjects were confined to the hospital setting for 24 hours, during which time activity level, dietary intake, and other environmental factors were controlled. All subjects were restricted to bedrest with bathroom privileges, consumed the same diet during the 24-hour hospitalization (2,581Kcals; 124.2g protein; 5,779mg arginine), and were required to refrain from smoking and alcohol consumption. Vegetables, which usually have a higher nitrate content from fertilizers, and nitrate- and nitrite-preserved foods were eliminated from the study diet. Concomitant baseline medications were administered, and blood glucose monitoring was performed by the diabetic subjects per their usual home routine. Medication and dietary intake as well as urinary output were recorded and evaluated by the research team during the 24-hour confinement period. At 9:00 p.m. on the day of confinement, all subjects were required to begin another ten-hour fasting period. Prior to discharge from the hospital setting, the subjects again provided fasting plasma and urine samples (Day 2). Vital signs were monitored daily during confinement, and all subjects were evaluated for adverse events prior to discharge. Plasma and urine samples were stored at -20°C prior to analysis. All laboratory personnel were blinded as to the identity of specimens prepared for assay.
Urine and plasma samples were fluorometrically assayed for nitrite and nitrate levels using a commercial kit (Cayman Chemical, Ann Arbor, MI) according to the manufacturer's instructions. The method used is that described by Gilliam, et al. Blood was collected in a glass tube and centrifuged, and the plasma was collected and frozen until assay. The samples were thawed, vortexed, and filtered with a 10kDa size exclusion filter (Millipore, Bedford, MA). For the determination of nitrate, nitrate reductase and NADP were added and allowed to incubate at room temperature for two hours. Following incubation 2,3,diaminonapthalene followed by NaOH was added, and the fluorescence was determined with a fluorimeter where excitation is 365nm and emission 405nm. Nitrite concentration was determined using the same method with the exception that nitrate reduction steps were omitted. Urine was processed in a similar fashion except that filtration was omitted. All samples were assayed in triplicate. Concentration in patient samples (µm/l) was determined by comparison to standard nitrate and nitrite solutions. One-way ANOVA with Tukey-Kramer post test was performed using GraphPad InStat version 4.10 for Windows 98.10 P-values <0.05 were considered significant).
On Day 1, fasting urine nitrate levels (µm/l ±SE) for groups C and HD (55.88±4.49 and 54.14±3.32, respectively) were not significantly different (Figure 1). However, group UHD fasting urine nitrate levels (30.35±3.61) were significantly lower than groups C (p<0.001) or HD (p<0.01). Day 2 fasting urine nitrate levels for groups C and HD were lower (42.60±1.92 and 45.57±5.10, respectively) but again not significantly different. Similarly, group UHD fasting urine nitrate levels (22.74±3.13) were lower than Day 1 values and were again significantly lower than either group C (p<0.05) or HD (p<0.05) (Table 2). Day 1 fasting plasma nitrate levels (µm/l) for group C (4.80±0.85) and group UHD (4.05±0.37) were not significantly different (Figure 2). Group HD (11.71±2.08), however, was higher than UHD but only significantly higher than C (p<0.05). Day 2 fasting plasma nitrate levels were slightly lower for groups C (2.92±0.37) and UHD (3.16±0.61) but as before these were not significantly different. However, group HD (11.94±4.46) was significantly higher than either group C (p<0.01) or UHD (p<0.05). Urine and plasma nitrate levels were approximately 100 times greater than nitrite, and nitrite was occasionally undetectable by this methodology. For these reasons, urine and plasma nitrite levels are not reported.
Figure 1. Fasting urine nitrate levels (µm/l) for controls (C), healed diabetic patients (HD), and unhealed diabetic patients (UHD) on Days 1 and 2 (mean ± SE). P-values as compared to C (PC) and HD (PHD) for each day.
Figure 2. Fasting plasma nitrate levels (µm/l) for controls (C), healed diabetic patients (HD), and unhealed diabetic patients (UHD) on Days 1 and 2 (mean ± SE). P-values as compared to C for each day except ‡, which compares HD and UHD for Day 2 only.
These data clearly establish a relationship between decreased (abnormal) fasting urine nitrate levels and the impaired healing of a diabetic ulcer unsuccessfully treated with becaplermin. Also, within the two diabetic groups (HD and UHD), the significantly decreased fasting urinary nitrate levels were paralleled by a significantly reduced fasting plasma nitrate level on Day 2. These findings show that diabetic ulcer patients unresponsive to becaplermin treatment have significantly reduced excretion of nitrate suggesting that endogenous NO production may be reduced.
The results indicate that the chronic, non-healing LEU diabetic population experience significantly decreased urinary nitrate excretion compared to diabetic patients with "unimpaired" wound healing or non-diabetic controls. If these results are further substantiated, we believe that a new strategy for clinical diabetic ulcer treatment based on the systemic or local (wound) enhancement of NO production may be justified. This new strategy in diabetic LEU treatment may provide an alternative approach for the management of other complications of diabetes, which may also be related to reduced levels of NO production.
NO is a hydrophobic, gaseous free radical, which is an important physiologic mediator for autonomic functions such as vasodilatation, neurotransmission, and intestinal peristalsis. It provides cellular signaling by activation of its target molecule, guanylate cyclase, which elevates intracellular concentrations of cyclic guanosine monophosphate (cGMP). Cellular signaling is performed without mediation of channels or cellular membrane receptors and is dependent upon the concentration of NO in the cellular environment. NO is generated in biologic tissues by three isoforms of nitric oxide synthase (NOS) that metabolize l-arginine and molecular oxygen to citrulline and NO. Two of the three isoforms are constitutive enzyme systems (cNOS) that are described in neuronal cells (nNOS) and in endothelial cells (eNOS). With these isoforms, increased levels of intracellular calcium activate the enzymes via calmodulin. The calcium dependent cNOS systems produce low (picomolar) quantities of NO. The third system is the inducible isoform (iNOS), which is calcium independent. Expression of iNOS is controlled by tissue-specific stimuli such as inflammatory cytokines or exogenous materials, i.e. bacterial lipopolysaccharide (LPS). Once induced, production of NO within tissue can increase as much as 1,000-fold, thereby producing an environment that is toxic to invading microorganisms. Currently, it appears that the cNOS enzymes are involved in maintaining skin homeostasis and providing regulatory function. The iNOS enzymes appear to be mainly associated with inflammatory and immune responses that are also implicated in certain skin diseases. In human skin keratinocytes, fibroblasts, and endothelial cells possess both the cNOS and iNOS isoforms. The wound macrophage and keratinocyte possess the iNOS isoform. In wound healing studies, NO synthesis has been shown to occur for prolonged periods (10-14 days) after wounding, and macrophages appear to be the major cellular source. As a mediator of tissue repair, NO has been demonstrated to promote angiogenesis and cellular migration, increase wound collagen deposition and collagen cross-linking, regulate microvascular homeostasis (vasodilatation), inhibit platelet aggregation, inhibit endothelial-leukocyte cell adhesions, modulate endothelial proliferation and apoptosis, increase the viability of random cutaneous flaps, and enhance cellular immunomodulation and bacterial cytotoxicity. Conversely, chronic diabetic ulcer patients may produce less NO with a marked deficiency in their ability to achieve normal wound repair. In general, during the wound healing process, NO-mediated "cellular signaling" and cytotoxicity provide enhancement of tissue oxygen availability, the inflammatory mediation of repair mechanisms, and wound matrix development and remodeling (Figure 3). This influence of NO on the wound healing process appears to be continuous over the time (delta-T) required for repair to be completed. Interestingly, despite documentation of simultaneous cytotoxic and protective activities within the wound environment, NO provides substantial promotion of tissue repair and maintenance of the microcirculation. The major metabolic pathway for NO is to nitrate and nitrite, which are stable metabolites within tissue, plasma, and urine. Tracer studies in humans have demonstrated that perhaps 50 percent of the total body nitrate/nitrite originates from the NO synthesis substrate, l-arginine.[19,20] Although nitrate and nitrite are not measures of biologically active NO, plasma and urine samples obtained with individuals on a controlled diet (low nitrate/low arginine) and after a suitable period of fasting allow us to use observations of variations in nitrate and nitrite values as an index of alterations in NO production..
Figure 3. Schematic representation of the possible role of nitric oxide (NO) in wound repair regulation. Wound NO-mediated "cellular signaling" and cytotoxicity appear to enhance the inflammatory mediation of repair, wound oxygen availability, and wound matrix remodeling and maturation. This influence of NO on the wound healing process appears to be continuous over the time (delta-T) required for repair to be completed.
In the diabetic wound, growth factor receptor expression is reduced during the repair process, and experimental studies suggest that a certain expression level of PDGF and its receptors is essential for normal repair. However, during normal wound repair, PDGFs stimulate the chemotaxis of polymorphonuclear (PMN) leukocytes and macrophages as well as the proliferation of smooth muscle and endothelial cells. Macrophage-dependent processes are crucial to the dynamics of PDGFs in wound repair. PDGFs appear early after wounding with an early peak of activity that has decreased by the time of wound healing. Increased wound tensile strength, granulation tissue production, epithelialization, and angiogenesis have been documented following becaplermin treatments.[25,26] The effect of becaplermin therapy is not limited to increased cellular proliferation but is also associated with increased recruitment of activated repair and inflammatory cells to the wound. In this regard, another vulnerary mechanism of becaplermin therapy for the diabetic ulcer is probably its correction of deficient wound NO production by the recruitment of activated macrophages and leukocytes laden with NO.
Unfortunately, for many diabetic ulcer patients, the severity of wound NOS and NO deficiency promotes the failure of the repair process at a more fundamental cellular level than that addressed by corrective PDGF therapy alone. Here, the absence of effective NO mediation by the wound fibroblast (WF) leads to reduced wound collagen deposition and mechanical strength. The absence of normal NO mediation in the activated macrophage and PMN lead to ineffective cellular and bacterial cytotoxicity and compromised immune modulation. The absence of the appropriate influence of NO mediation during the remodeling phase of wound repair leads to early cell apoptosis and poor matrix maturation. Decreased NO production in diabetes has also been suggested to correlate with endothelial damage and increased platelet activation and leukocyte adhesions at the endothelial interface which precede the development of diabetic angiopathy. Microvascular homeostasis becomes impaired in an environment deficient in NO, and tissue perfusion and viability are jeopardized. In this regard, NO becomes a primary regulator of the cellular processes that promote wound healing. Additionally, our clinical findings suggest that an appropriate threshold level of endogenous NO production is also a priority for the initiation of unimpaired wound repair.
Decreased NO activity has been well documented in the diabetic and steroid-treated wounds.[5,30] Current clinical treatment modalities for deficient wound NO would include dietary (or parenteral) l-arginine and hyperbaric oxygen (HBO) therapy. Targeted delivery of the iNOS gene using adenoviral vectors and topical application of NO producing polymers are also being investigated as means of correcting wound NO deficiency. In this regard, the fasting urine nitrate assay would be beneficial as a clinical screening tool and as a quantitative monitor for the implementation of NO directed therapies in selected patients.
In conclusion, in a preliminary, retrospective study, we have demonstrated that diabetes-impaired wound healing following becaplermin treatment is associated with a deficiency in nitrate metabolism. This deficiency was assayed from a fasting urine sample that correlated with fasting plasma nitrate levels in diabetic ulcer patients with physiologic renal function and impaired healing. If further substantiated, we feel that this assay will provide a valuable screening tool for diabetic ulcer management. Additionally, it may also allow us to formulate novel therapies for the treatment of diabetic-impaired wound healing that may significantly reduce the tissue and limb loss associated with this disease.
Table 1. Characteristics of study subjects HD and UHD
Data are means ± SE or n (%).
Table 2. Fasting urine and plasma nitrate values
P = P-value. C = Control group. HD = Healed diabetic patients. UHD = Unhealed diabetic patients.
Appreciation is expressed to Doris Sandy, RD, for the design of the diet used in this study, and to the nursing staff of 5-East, Columbia Retreat Hospital.
Address correspondence to: Joseph V. Boykin, Jr., MD, Columbia Retreat Hospital, Wound Healing Center, 110 N. Robinson Street, Suite 403, Richmond, VA 23220, Phone (804) 353-8100, Fax (804) 353-4658, E-mail: firstname.lastname@example.org.
All Rights Reserved @ Hi-Tech Hyperbaric Medical Centre, Malaysia.