Indian Journal of Medical Sciences Medknow Publications on behalf of Indian Journal of Medical Sciences Trust ISSN: 0019-5359 EISSN: 1998-3654 Vol. 61, Num. 6, 2007, pp. 347-355

Indian Journal of Medical Sciences, Vol. 61, No. 6, June, 2007, pp. 347-355

Original Contributions

Immunomodulatory role of Tinospora cordifolia as an adjuvant in surgical treatment of diabetic foot ulcers: A prospective randomized controlled study

Purandare Harshad, Supe Avinash

Department of Surgery, Seth G. S. Medical College and K. E. M. Hospital, Mumbai
Correspondence Address:Sangeeta Towers, 6th Floor, Chembur, Mumbai - 400 071, avisupe@vsnl.com

Code Number: ms07056

Abstract

Keywords: Diabetic foot ulcers, immunomodulation

Diabetic ulcers of the lower extremity are a major concern, preceding most of the amputations in general surgery. More than 7 million people in India suffer from diabetic foot ulcers, with a recurrence rate of 53%. The expenditure on diabetic foot may amount to 25-30% of a patient′s annual income. ^[1],[2]

Diabetes mellitus is a secondary immunodeficiency disorder. ^[3] There is impaired chemotactic, phagocytic, lytic and intracellular killing capacity of the neutrophils, macrophages and monocytes; defective cellular immunity and T-cell-dependent IgG/IgM response and impaired complement activity. ^[4] Non-healing diabetic foot ulcers have deficient synthesis growth factors such as PDGF, βFGF, TGF-β, GM-CSF, as these factors are trapped in pericapillary fibrin cuffs or destroyed by metalloprotein enzymes.

Recombinant REGRANEX (Becaplermin; rhPDGF-BB), approved by the US FDA (1997), is found to impair wound contraction, epithelialization and wound strength in some cases. ^[5] Though a 40% increase in healing of noninfected neuropathic foot ulcers was observed, a 30% recurrence was noted at 6 months. ^[6] Treatment of such immunodeficiency by immunomodulators or growth factors is technologically expensive. The average cost of US$10/day is substantial in the Indian setting. Based on the Ayurvedic concept of use of plant extracts to improve body′s natural resistance, we used immunomodulation to overcome shortcomings of specific targeted immunopotentiation. Macrophages play a key role in phagocytosis, production of cytokines and growth factors and potentiation of neutrophil-platelet functions. They were the target of our therapy. The aqueous extract of the creeper Tinospora cordifolia (family Menispermaceae), purified and biostandardized, is well tolerated and induces leukocytosis with neutrophilia and improves PMN-macrophage phagocytic, lytic and intracellular killing activity by the enhancement of GM-CSF activity. ^[7] Clinical studies show reduced complications in intra-abdominal sepsis, enteric fistulas and obstructive jaundice with early resolution in pulmonary tuberculosis, cirrhosis, Chronic Suppurative Otitis Media (CSOM), burns sepsis, as also lymphatic and hemopoetic malignancies, chemotherapy-induced neutropenia and fungal infections. ^[8] GM-CSF increases the levels and activity of platelet- and neutrophil-derived PDGF and acts in a positive feedback loop, a vital step in wound healing. The global stimulatory role of GM-CSF and actions of T. cordifolia to stimulate GM-CSF have led to this study, which aims to evaluate the efficacy of T. cordifolia as an adjuvant in the treatment of diabetic foot ulcers. We also aim to study its influence on determinants of wound healing, bacterial eradication and phagocytic function of PMN cells.

Materials and Methods

A double-blind prospective randomized placebo-controlled study lasting for over 18 months (August 1999 to February 2001) was carried out at a reputed public hospital after authorization from the hospital′s Ethics Review Committee. Patients admitted to the surgical wards were screened as per the inclusion-exclusion criteria, and 50 patients entered the study.

Inclusion criteria

1. Adult patients of either sex with diabetic foot ulcer not less than 4 cm in diameter
2. Ulcers of Wagner′s system′s Grade I or Grade II; and digital, ray or forefoot amputations with non-healing ulcer ^[9]

1. Patients with ulcer of any other etiology
2. Patients with local or systemic disease or on therapy that may interfere with wound healing
3. Ulcer with radiographic evidence of osteomyelitis or involving joint spaces or bones (Wagner′s Grade III and IV) ^[9]

On inclusion into the trial, a detailed history was taken with emphasis on the duration of ulcer and diabetes, its treatment, previous ulcerations, amputations and other associated illnesses. A detailed clinical examination was performed. The ulcer included in the trial was classified by wound morphology, severity and location with a total wound severity score according to a published grading system by Pecoraro and Reiberg ^[10] [Table - 1] 1 week after the initial debridement. The mean target ulcer area and wound perimeter were calculated from wound contour tracing over a sterile transparent plastic film. Ulcer depth was measured by means of a blunt-tip calibrated probe from the level of wound surface. Swabs were taken for Gram staining and culture. Wound photographs were taken at a standard focal length. Clinical assessment of neuropathy and peripheral vascular disease with flow-pressure assessment using an arterial color Doppler was done. Baseline investigations including a complete hemogram, differential count, ESR, fasting and postprandial blood sugar, liver and renal function tests were also performed.

Blood samples were also collected for the assessment by polymorphonuclear (PMN) function test: % phagocytosis evaluation by an in vitro Lehrer and Cline technique of Candida albicans phagocytosis using trypan blue dye exclusion method. ^[11] Ten ml of the patient′s heparinized blood diluted with normal saline (1:1) on Ficoll Hypaque gradient, centrifuged at 200 rpm for 40 min, and the PMN-RBC pellet were separated and mixed with 1 ml Dextran-70 and 1 ml heterologous plasma. The RBC sediment and PMN cells were counted and adjusted to 1 x 10 ⁶ cells/ml with 90% viability. The test organism was C. albicans obtained by overnight cultures in nutrient broth adjusted to 1 x 10 ⁷ spores/ml by spectrophotometric absorbance. In the assay system, 2 x 10 ⁶ PMN were suspended in Minimum Essential Medium (MEM) (Himedia, containing 100 µg/ml each of penicillin and streptomycin) and supplemented with 20% AB serum to which 1 x 10 ⁶ (0.1 ml) spores of C. albicans were added, and the suspension was incubated at 37°C in 5% CO ₂ atmosphere for 1 h. The tube was then centrifuged at 2,000 rpm for 10 min, and cytosmears were prepared, fixed and stained with Giemsa. One hundred PMN cells were scanned and the cells with ingested organisms were counted to obtain percentage phagocytosis.

On the basis of a predesigned randomization schedule, each patient was administered the drug supplied in prepacked numbered bottles for a period of 1 month. The individuals concerned with conducting the trial were blinded to the randomization. Medical therapy was optimized and all the patients were administered the drug (study drug or placebo) along with conventional therapy for the treatment of the ulcer as well as diabetes. All slough, pus pockets, clefts, necrotic tissue and infected granulation tissue in the ulcer were removed by a sharp surgical debridement. Regular debridements were done as per further wound assessment during the trial. The total number of debridements needed was assessed as a parameter of response. Standard regimen of daily wound care included gentle cleansing with half strength 1.5% hydrogen peroxide solution with ample amount of saline. Superficially infected wounds were dressed with topical antibiotics, followed by a gauze dressing. Empirical broad-spectrum systemic antibiotic coverage was initially achieved with oral antibiotic therapy using Amoxycillin-Clavulanate. Specific antibiotic coverage was reconsidered after culture results were obtained. Ambulation was minimized and protective footwear advised. Patients were started on insulin and the dose adjusted. The patients were assessed on a weekly basis till complete healing (either satisfactory or complete) was achieved. The wound severity score was determined as mentioned earlier. Wound swabs were sent for culture at 2 and 4 weeks.

A 4-week period was considered to be adequate for clinical and parametric shifts in the wound to be obvious for assessment of the effect on the progress of the wound, as also for avoiding the attrition of the sample size due to the failure of the patient to attend the OPD. ^[12] At the end of 4 weeks, a final wound severity score, the final ulcer area and perimeter were determined; and blood samples were taken at the end of the trial to measure total and differential count, liver and renal function and PMN function. For optimum results, final PMN function assessment was done at 3 months. ^[13]

Evaluation

The rate of ulcer healing was calculated by two related methods ^[14] :

• Rate of change of wound area: Leugers coefficient of healing (centimeters squared per day) was obtained as the slope of the graph of linear regression of wound area epitheliazed versus time (duration of treatment). A positive rate contributed to early healing.
• Rate of linear advancement of wound margin: Change in wound perimeter (millimeters per day), calculated by Gilman′s formula

d = ∆A / p, where

′∆A′ = A ₁ - A ₂ , in which A ₁ is the initial and A ₂ is the final wound surface area. ′p′ = p _{1 +} p ₂ , in which p ₁ is the initial and p ₂ is the final wound perimeter. The rate of closure (mm/day) is then obtained as a linear regression.

The difference in the initial and final wound severity score was calculated. An ulcer was considered to be healing when it demonstrated an overall progress in net tissue repair process as seen by a fall in the severity score or by a positive rate of healing or both. An ulcer was non-healing if there was no healing progress or if net wound expansion occurred.

At the end of the trial, patients were divided into two groups - Group A = drug, Group B = placebo - as per the randomization chart. The data obtained from the two groups was analyzed for comparability by validation tests like the Fisher′s F test. The various parameters mentioned above were compared between the two groups for statistical significance using the Chi-square (χ² ) test, the unpaired ′t′ test and the ANOVA (Analysis of Variance) test.

Results and Analysis

[Table - 2]

Of the 50 patients enrolled, 45 patients completed the trial - 23 in the drug study group and 22 in the placebo group. In the study group, 17 were males (73.9%) and 6 females (26.1%); while in the placebo group of 22 patients, there were 19 males (86.4%) and 3 females (13.6%). The mean age for the entire study group was 56.29 years (range: 32.4-80.6 years; study group = 56.26 years; placebo group = 56.32). The mean duration of diabetes in those patients receiving the study drug was 5.95 years (range: 0-18 years) while in those of the placebo group was 8.27 years (range: 0-22 years).

In patients receiving the study drug, the mean duration of ulcer prior to inclusion in the trial was 21.08 days (range: 12-35 days); while in those receiving the placebo, it was 30.36 days (range: 21-44 days). The Chi-squared and Fischer′s tests indicate comparable groups for inferential analysis. As many as 66% of the ulcers were involving the forefoot and midfoot, suggesting proneness to trauma.

Four patients (8.89%) had more than one ulcer, out of which only one ulcer was included in the trial as per the protocol. The total number of patients with repeated diabetic foot lesions was 10 (22.22%). Six patients (13.33%) were freshly detected to be diabetics.

Analyzing the final outcome, viz., net improvement or failure, of the 23 patients in the study drug group, 17 (73.9%) improved; while in the placebo group of 22 patients, 13 (59.1%) improved (P = 0.292). Though not statistically significant, better healing and net improvement were obtained in patients of the study drug group.

When the wounds were evaluated in terms of the two measured healing rates, viz., R ₁ - rate of change of ulcer area in cm ² /day; and R ₂ - rate of change of ulcer perimeter in mm/day, improvement was more obvious in the group receiving the drug. The mean R ₁ in the drug group was 0.149 cm ² /day (95% confidence interval: 0.06-0.36), while that in the placebo group was -0.069 cm ² /day (95% confidence interval: 0.12-0.26), indicating a net increase in the ulcer area in the placebo group in comparison with the drug group (P = 0.145). Similarly R ₂ , the mean change of perimeter, in the patient group was 0.093 mm/day (95% confidence interval: 0.09-0.1), while that in the placebo group was -0.073 mm/day (95% confidence interval: 0.06-0.08) (P = 0.0890). Thus, better results were obtained in the drug group, both the parameters indicating an expansion of the wound either in width or in depth in the placebo group, indicating non-healing.

Change in the wound depth was analyzed by applying the unpaired ′t′ test to the mean difference in the wound depth in the drug and the placebo group. It was found that d ₁ = 2.17cm, 95% confidence interval: 1.89-2.45; and d ₂ = 1.36 cm, 95% confidence interval: 1.08-1.64 respectively, a bigger positive value indicating a better improvement in the ulcer with reduction in its depth. The obtained ′P′ value of 0.096 indicates better improvement in the drug group. Thus the change in the depth of ulcers is one of the important determinants of healing.

The Wound Severity Score was measured as a qualitative parameter in wound assessment, a positive difference indicating healing. The difference in the initial and final wound score was calculated (drug score = 14.39, 95% confidence interval: 12.64-16.14; placebo score = 10.59, 95% confidence interval 8.70-12.48; P = 0.149, analyzed by ANOVA test), indicating that the drug group showed better improvement in the wound scores.

The total number of debridements needed per patient was considered as a measure of wound status, more debridements indicating a progressively worsening wound. The mean number of debridements (drug = 1.91, 95% confidence interval 1.72-2.10; placebo = 2.49, 95% confidence interval 2.17-2.81; P = 0.030) is significantly reduced in the drug group.

Finally, analyzing the change in percentage phagocytic function of PMN cells as a quantitative measure of immune function improvement, the values of 3.94% (95% confidence interval 3.12-4.76) and 2.34% (95% confidence interval 0.68-4.0) in the drug and placebo group respectively (P = 0.048) suggest a significant effect of drug on the immune function.

No difference was observed in bacterial clearance from wounds in both the groups [53% persistence of the same strain, especially in patients with polymicrobial spectrum - 34 out of 45 patients (75.55%)]; also, no difference was observed with respect to neutrophil count. Gram-negative bacteria formed 70% of the bacterial isolates.

Discussion

Recent molecular studies indicate a defect in immune functions in diabetics. Growth factors and cytokines regulating proteolytic activity and matrix synthesis are reduced in chronic wounds. These chronic wounds have a persistent bacterial load, possibly related to local immune dysfunction and growth factor deficiency. The development of REGRANEX (recombinant PDGF-BB) evoked interest in immune therapy in diabetic non-healing ulcers. It showed beneficial effect, primarily in noninfected neuropathic ulcers. However, therapy was extremely costly (approximately US$9.70 per day), and long-term effects on wound contraction and wound strength, as well as recurrences, were not satisfactory. Though these products are still unavailable in our country, their usage is bound to be highly cost-ineffective from both their clinical benefit and economic standpoint.

The modulation and stimulation of GM-CSF and related cytokines thus improving the macrophage function were considered as a therapeutic modality in non-healing diabetic ulcers. The biostandardized aqueous extract of stem of T. cordifolia, an Ayurvedic agent, has a proven stimulatory effect on GM-CSF and is extremely economical (approximately Rs. 11 per day).

Sixty percent of the ulcers involve the distal plantar aspect of the foot, indicating a proneness to minor trauma, explained on the basis of occupational activity, higher incidence of foot deformities, neglected footwear precautions and the higher incidence of lean non-obese diabetics in India, further complicated by a very poor knowledge of foot care and improper treatment of diabetes itself. A net final outcome of better healing in the drug group correlated with a better rate of change of ulcer area (R1) and a better rate of change of ulcer perimeter (R2). Moreover, the drug group showed better improvement in the wound score. These mentioned parameters are an assessment of qualitative as well as parametric evaluation of overall wound healing. An improvement in the wound score is an indirect indicator of increased phagocytic activity as well as of enhancement of granulation tissue formation in a chronic wound. Positive values of R1 and R2 in the drug group with a corresponding improvement in wound scores suggest an overall positive outcome, though both these parameters did not reach statistical significance in the current study. The ulcer heals with wound contraction as well as the development of proliferating granulation tissue from its floor, r with reduction in its depth. This is seen as a change in the wound depth and is an important determinant of wound healing. At cellular level, the above changes are probably related to enhanced macrophage and monocyte activity, leading to a substantiation of growth factor and increase in scavenging activity.

A statistically significant lower number of debridements were needed in the drug group, which correlates with better improvement in wound depth. This indicates halting of progression of infection to deeper planes of the wound. Improved scavenging and defense mechanisms involving PMN and macrophages play a vital role in this aspect.

A statistically significant improvement in the percentage phagocytic function of PMN cells in the drug group is a quantitative and functional corroboration of the role of this drug in cellular immune potentiation.

Though the results of this study showed no significant difference between the two groups, T. cordifolia has been known to induce an increase in circulating neutrophils. Coexistent polymicrobial bacterial infection and chronic local sepsis probably account for this disparity. Vasculopathy and neuropathy, with their effects on microcirculation, further confound results and support non-healing, as seen in 66.67% of patients in the drug group and 77.33% of patients in the placebo group. These had variable degrees of neuropathy and/or vasculopathy, and non-healing was more common in patients in both the groups who had these changes. T. cordifolia, unfortunately, does not seem to have an effect on these neuropathic, autonomic and microvascular dysfunctions. Infection was another important determinant of healing in both the groups. No difference was observed in bacterial clearance from wounds in them though polymorphonuclear functions were enhanced, possibly due to the chronicity and polymicrobial nature of most infections. Almost 45% of the patients with improved score still showed bacterial persistence.

Conclusion

The overall beneficial effect of Tinospora cordifolia appears to be primarily dependent on enhancement of scavenging and desloughing action of PMN cells and macrophages. These immunomodulated cells with their improved phagocytic function create a favorable environment in an otherwise non-healing wound. This probably facilitates a series of pro-healing processes like growth factor activation, angiogenesis and granulation tissue formation - all processes inhibited and suppressed in chronic wounds. The role of such immunomodulated cells in direct or indirect growth factor stimulation is not proven in this study and is only speculative, based on the observation of better wound healing in the drug group.

As this study was conducted with an observation period of only 1 month to facilitate in-patient assessment and prevent attrition of the sample size through loss of follow-up, a number of parameters were possibly not altered due to chronic and non-healing nature of these wounds to an extent obvious enough to cause statistically significant observations. A longer duration would definitely aid a more comprehensive evaluation of the effects of the drug.

Though not proven beyond doubt, the initial results of this study are promising. Considering the cost-effectiveness of the drug and the fact that almost 70% of our patients depend on traditional or alternative medicine, further study with a larger patient group and a longer follow-up is definitely indicated. T. cordifolia is also seen to have some degree of anti-hyperglycemic effect; and if accepted, oral T. cordifolia may be the future of immunotherapy in diabetic wounds, aimed at an overall enhancement of immune function.

Acknowledgments

References

Copyright 2007 - Indian Journal of Medical Sciences

The following images related to this document are available:

Photo images