THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINE Volume 2, Number 3, 1996, pp. 335-343 Mary Ann Liebert, Inc.

An Aqueous Extract of the Leaves of Chromolaena odorata (Formerly Eupatorium odoratum) (Eupolin) Inhibits Hydrated Collagen Lattice Contraction by Normal

Human Dermal Fibroblasts

THANG T. PHAN, M.D./'2 MARGARET A. HUGHES, Ph.D.,^ GEORGE W. CHERRY, D,PhiL,i TRUNG T. LE, D.Sc., and HUNG M. PHAM, D.Sc.

ABSTRACT

Chromolaena odorata (formerly Eupatorium odoratum) is used as a traditional medicine in Vietnam (Nghiem, 1992), where its Vietnamese common name is "co hoi/' While it has been widely considered a weed by agriculturalists (Holm et al., 1991), the aqueous extract and the decoction from the leaves of this plant have been used throughout Vietnam for the treatment of soft tissue wounds, bum wounds, and skin infections. A number of clinical studies done by Vietnamese as well as foreign medical workers has demonstrated the efficacy of this ex­tract on the wound-healing process. In this article, the effect of the Eupolin extract on hy­drated collagen lattice contraction by human dermal fibroblasts, an in vitro model of wound contraction, is described. The significant inhibition of collagen gel contraction by Eupolin ex­tract at 50 to 200 fig/ml is demonstrated in various concentrations of collagen. When the ex­tract at 50 to 150 /ig/ml was washed out of the lattices and replaced by fresh medium with­out Eupolin, the contraction of collagen by cells was resumed. The visualization of cells in the lattices by incubation in a tetrazolium salt for 2 h showed live cells at 50 to 150 /tig/ml of extract. In contrast, all cells were killed in the higher extract doses of 300 or 400 fig/ml. These preliminary results showing the inhibitory effect of Eupolin extract on collagen contraction suggest that a clinical evaluation of its effect on wound contraction and scar quality should be made. This work illustrates that traditional remedies that are used by folk practitioners to improve healing can be examined in a scientific manner using in vitro wound-healing mod­els. It could be that the synergistic properties of components of the natural extract contribute to the positive effects demonstrated on various wound-healing mechanisms.

INTRODUCTION ous extract and the decoction from the leaves of this plant have been used since then

C hromolaena odorata was first identified in the throughout Vietnam for the treatment of soft Antilles archipelago in Central America tissue wounds, burn wounds, and skin infec-

(Nghiem, 1992). In 1930, it was found growing tions. A number of studies carried out by in-in North Vietnam (Nghiem, 1992), The aque- vestigators in Vietnam, Nigeria, and the Ivory

^Wound Healing Institute, The Churchill, Oxford, U.K. ^National Institute of Burns, Hanoi, Vietnam.

335

336 PHAN ET AL.

Chromolaena odorata (L.) R. M. King & Robinson. 1, habit; 2, flowering branch; 3, flower; 4, achene. Reproduced with permission from R. Holm et aL, The World's Worst Weeds, published 1977 (reprint edition), Krieger Publishing Company, Malibar, Florida.

Coast on clinical and experimental wounds demonstrated that the extract from Chromolaena odorata inhibited the growth of bacteria {Pseudomonas aeroginosa, Escherichia coli, and Staphylococcus aureus) (Akah, 1990; Bamba et a l , 1993; Irobi, 1992; Le, 1995; Triratana et a l , 1991). In clinical use it has been noted that the aqueous extract of Chromolaena odorata en­hances hemostasis (Akah, 1990) and stimulates granulation tissue and reepithelialization (Le,

1995). In a major clinical study on the treatment of infection in dentoalveolitis, Eupolin as an ointment was shown to be effective as a thera­peutic modality (Truong, 1989). In a previous in vitro study using cell culture, we have found that Eupolin extract enhanced both human der­mal fibroblast and endothelial cell growth as determined by a colorimetric method (Phan et al., 1996 unpublished).

Wound contraction is one of a cascade of

AN AQUEOUS EXTRACT OF THE LEAVES OF CHROMOLAENA 337

events that occur in healing and leads to scar contracture that is a critical complication of postburn trauma. One of the criteria for topical agents used for treating burns is to reduce the severity of that complication.

The aim of this present study was to investi­gate the effect of Eupolin extract on collagen gel contraction by normal human dermal fibrob­lasts, an in vitro model of wound contraction.

MATERIALS AND METHODS

Plant extract material

An ointment formulation (Eupolin ointment) of an extract from the leaves of Chromolaena odor­ata was provided by the Vietnam National Institute of Burns, Hanoi. The Eupolin ointment prepared with a condensation ratio of 10:1 is dark brown, sweet, with a density 1.2-1.3 and pH 4.5. It contains flavonoids (salvigenin, saku-ranetin, isosakuranetin, kaemferid, betulenol, 2-5-7-3-tetra-O-methylquercatagetin, tamarixetin, chalcone 1, chalcone 2), essential oils (geyren, bomyl acetate, j8-eubeden), saponin triter-penoids, tannins, organic acids, and trace ele­ments) (Le, 1995).

For the in vitro studies the ointment was heated in water (MiUipore Q) at 80^C for 20 min and the brownish Eupolin mixture was filtered through a previously weighed Whatman filter paper No. 1 to eliminate the nonsoluble vase­line and obtain a clear brown solution. The fil­ter paper with residue was dried to constant weight and the amount of Eupolin solubilized was calculated by subtracting the weight of the residue from the initial weight of the ointment. The Eupolin solution was sterilized by filtra­tion through 0.45- and 0.2-^tm filters. The stock solution of 12 mg/ml was kept in the dark at 4°C.

Fibroblast culture

Human dermal fibroblasts (HDF) were ob­tained by outgrowth from explants from nor­mal breast skin removed during plastic surgery. Stock cultures were maintained as monolayers in DMEM ''complete" medium [i.e., supplemented with 10% fetal calf serum (PCS), 50 /ig/ml streptomycin, 50 lU/ml peni­cillin, 0.25 mg/ml fungizone (all from GIBCO)] in a humid 5% CO2, 95% air atmos­phere. Cells from passages 3 to 6 were used for experiments.

100

e <

e 1

« 200 fig. 400 f4g/m\

100 /4g/ml

Control

"^^•^^^ 10 f4g/m\

Time (Days)

FIG. 1. Effect of Eupolin extract on collagen lattice contraction by human dermal fibroblasts. Collagen concentra­tion = 1.3 mg/ml. Arrows indicate time points when medium was replaced by fresh medium with or without ex­tract. Dashed lines show the resumption of contraction when medium with extract was replaced by normal medium. Each point represents mean ± SEM of multiple lattices.

338 PHAN ET AL

Preparation of hydrated collagen lattices

Collagen lattices were prepared from type I collagen extracted from rat tail tendons in 0.1% acetic acid. Working solutions were made by diluting the collagen to 1.5-2.6 mg/ml. Cells were trypsinised, resuspended in medium, and counted using a hemocytometer. The vol­ume of suspension was adjusted to give a cell density of 3.35 X 10^ cells/ml and placed on ice at O' C while the collagen polymerising so­lution was made. The method of Montesano and Orci for preparing collagen lattices was used (Montesano et al., 1988). The cell sus­pension was uniformly distributed in the col­lagen polymerizing solution, giving a density of 6.7 X 10" cells/ml. Two milliliters of the col­lagen/cell suspension were dispensed into 32-mm-diameter petri dishes and the lattices were incubated for 20 min at 37^C in 5% COa/air at­mosphere. Then 1 ml of DMEM/10% PCS, ei­ther with or without additions of different con­centrations of Eupolin extract, was added to the collagen lattices. The gels were released from the sides and bottom of the petri dish with a spatula and this was considered as zero time. The reversibility of the effects of Eupolin extract was determined by the substitution of media containing the compound by normal

medium. The lattice diameters were assessed at various time points for several days by plac­ing plates on transparent graph paper and measuring two diameters at right angles to one another. The area was calculated from the mean of the diameters and expressed as a per­centage of initial lattice area.

Visualization of living cells in lattice

At the end of the experiment, the medium was removed from the dishes and 2 ml Hanks' bal­anced salt solution (HBSS) was dispensed to each dish. Then 200 il of the tetrazolium salt MTT[3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide] was added to permit the visualization of living cells in the lattices. The mitochondria of living cells reduce the MTT to a blue formazan product. After 2 h, the gels were observed at intervals and photomicrographs were taken with a camera attached to a Leitz Diavert inverted microscope (Parish et al, 1995).

Statistics

Three collagen lattices were used for each set of treatments as well as for the control set, ex­cept for experiments on reversibility in which 6 lattices were prepared and the extract was washed out of pairs of lattices at two or three

o

<

100 i

80 -

60 -

40 -

20 -

0 -

I w m m ^ V\ '1

1 1 1 m m—m—m

I ^ ^ 1 — - - - ^ 1 ,

,

N / ' ^ } ^ ^ - ^ •• i.

I 1 • r » f — • 1 • 1 . 1

200, 300 ^g/ml

100 /ig/ml

50 |ig/ml

Control.

9 12

Time (Days)

15 18 21

FIG. 2. Effect of Eupolin on collagen lattice contraction by human dermal fibroblasts. Collagen concentration = 1 mg/ml. Arrows indicate time points when medium was replaced by fresh medium with or without extract. Dashed lines indicate the reversibility of gel contraction when the extract was washed out. Each point represents mean ± SEM of multiple lattices.

AN AQUEOUS EXTRACT OF THE LEAVES OF CHROMOLAENA

l oog

339

< Control.

50 pg/ml.

- » 75 ^g/ml,

1 00 ^g/ml

Time (days)

100 g

Of

(R

B Time (Days)

FIG. 3. Effect of Eupolin up to 100 iJig/m\ on collagen gel contraction by human dermal fibroblasts. Collagen con­centration — 0.75 mg/ml. Each point represents mean ± SEM of triplicate dishes. Degree of contraction depends on the extract concentrations. (A) No exchange of medium or extract doses was made during the time course of the ex­periment. (B) Reversibility of inhibition. Arrows indicate time points when medium was replaced by fresh medium with or without extract. Dashed lines show the reversibility of fibroblast inhibition in the lattices where medium with­out extract was added.

subsequent time points. For each lattice at a ments to identify significant differences at or given time point, the percentage of initial area above the 95% level. (% AQ) was calculated as %Ao == 100 X At/Ao where AQ is initial lattice area and At is lattice I?F<;TTT TQ area at time t. For the data at each time point, RESULTS one-way analysis of variance was carried out ^ using the Statview 512+ TM 1.1 package, meet of different doses of Eupolm extract Scheffe's test was used to compare the mean The effect of the extract on the contraction of percentage of initial area under different treat- collagen gels in the high collagen concentration

340 PHAN ET AL.

100

e <

e

Control.

50 /jg/ml.

75 /jg/ml.

100 ^g/ml.

- " ^ — 150 ^g/ml.

e <

tfi

^ool

8 0 -

6 0 -

40 -

2 0 -

0 -

^ys &

T 1 1 -

1 . . . . . . . , ; ^ ^

— C I J- •

I ' 1 '

1 K 71 A /{

'"^^.. ^ " ^ ^ " ^ ^ ^

1 ' r " 1 ' 1

Control

50 ^fg/ml

75 Mg/ml

1 00 Mg/ml

1 50 Mg/ml

B

6 8

Time (Days)

10 12 14

FIG. 4. Effect of Eupolin up to 150 fJig/ml on collagen lattice contraction by human dermal fibroblasts. Collagen concentration = 1 mg/ml. Each point represents mean ± SEM of multiple lattices. Degree of contraction depends on the extract concentrations. (A) No exchange of medium or extract doses was made during the time course of the ex­periment. (B) Reversibility is shown for 50 and 150 /Ag/ml. Arrows indicate time points when medium was replaced by fresh medium with or without extract. Dashed lines show the reversibility of fibroblast inhibition in the lattices where medium without extract was added.

(1.3 mg/ml) is shown in Figure L In that con­centration of collagen, gels contracted slowly. Eupolin at 100 ^tg/ml completely inhibited contraction until day 9. At that stage fresh medium with the same concentration of extract was given to half the gels while the remainder were given medium without extract. The in­hibited collagen gels began to contract but were

still significantly slower than controls (area = 242% of control on day 11, p < 0.01). The inhi­bition of contraction of collagen gels was re­versible when the medium was changed to medium without extract. At concentrations of 200 and 400 /ig/ml, the extract completely in­hibited gel contraction and the inhibition was irreversible when the drug was washed out.

AN AQUEOUS EXTRACT OF THE LEAVES OF CHROMOLAENA 341

When the collagen concentration was reduced of control cells. In contrast, at the high concen-to 1 mg/ml , collagen gels contracted signifi- trations (200 to 400 /xg/ml) the fibroblasts ap-cantly faster than in 1.3 mg/ml collagen (Fig. 2) peared rounded with no cytoplasmic ap-(p < 0.001). Collagen contraction was inhibited pendages. by a range of doses from 50 to 300 /xg/ml of the The addition of MTT allowed the visualiza-extract. The inhibition of contraction by 50 tion of live cells in the lattices. At 200 / tg /ml fxg/m\ was completely reversible when the ex- cells were rounded but still alive as shown by tract was washed out, with lattices reaching the MTT reduction. However, blue formazan in the size of controls by day 14. The extract at 100 cytoplasm was only seen after 24 h incubation /xg/ml inhibited gel contraction to a much with MTT, indicating extremely slow metabo-greater degree than at 50 /xg/ml at all the dif- lism. At 300 and 400 /xg/ml, no evidence of ferent time points (p < 0.01). Reversibility of in- MTT reduction by cell mitochondria was ob-hibition of contraction of the gels was faster from served. This shows that all cells were killed by day 11 when a second "washing out" of the drug high concentrations of the extract, was performed. Contraction of collagen lattices was completely prevented by 200 or 300 /xg/ml of extract, and this was irreversible. DISCUSSION

The inhibition of collagen gel contraction by Eupolin extract from 50 to 150 /xg/ml is shown Wound healing consists of a number of corn-in Figures 3A and 4A for collagen concentrations plex and overlapping phases in which contrac-of 0.75 and 1.00 mg/ml. The inhibition of gel tion is an important component. Wound con-contraction by 75, 100, and 150 /xg/ml reached traction is a basic biological repair process significance as early as day 1 (p < 0.01) and was necessary for the closure of open wounds, almost stable at all the time intervals. Both the However, in wounds like burn injuries, wound degrees of inhibition and its reversibility de- contraction may lead to scar contracture pro-pended on the extract concentrations. Lattices ducing both functional and cosmetic deformi-that had been in 50 /xg/ml of extract quickly re- ties. Both of these can lead not only to physi-sumed contraction when normal medium was cal impairment in healing but also to substituted without the addition of extract, and psychological problems. Some of these can be further exchanges of medium to remove resid- horrendous (Rudolph et al., 1992). ual extract reduced the lattice area almost to that It is important in the treatment of burn of the control by day 10 (Figs. 3b and 4b). wounds and skin injuries to minimize scar con-

, ^ tracture and deformities. To this end, one ap-'^ ^ proach has been to use topical pharmacologi-

By light microscopy, fibroblasts associated cal agents as well as pressure garments, with lattices maintained in control medium ap- However, the latter treatment is very difficult peared elongated with a number of intercon- for patients to comply with in countries that necting cytoplasmic appendages. However, in have hot, tropical climates, lattices supplemented with Eupolin extract, fi- Previously, as mentioned in the Introduction, broblast morphology appeared variable with an extract of the leaves of Chromolaena odorata the degree of change being directly related to has been extensively used by traditional practi-the concentrations of drug used. tioners in the treatment of wounds and bums in

At 50 and 75/xg/ml of Eupolin extract the ma- Vietnam. Because of the complexity of the jority of fibroblasts appeared elongated with cy- wound healing process comprising different toplasmic processes. However, the degree of specific events such as cell proliferation, reep-elongation and the number of processes were ithelialization, angiogenesis, and contraction, less than in control cells, but, after the extract models have been used to look at the relation-was washed out, they became similar to those ship between therapeutic agents and these seen in control lattices. At 100 and 150 /xg/ml a phases of healing (Cherry et al., 1994). In our few cells had small protrusions, although the study, a simpUfied in vitro model of wound con-cell bodies appeared more rounded than those traction, the fibroblast-populated collagen lat-

342 PHAN ET AL.

tice (FPCL), initially described by Bell, Ivarsson and Merrill (Bell et aL, 1979), was used to in­vestigate the effects of an aqueous extract from Chromolaena odorata (Eupolin ointment). It was found that this agent inhibited collagen lattice contraction by normal dermal fibroblasts. The degree of inhibition depended on the concen­trations of the extract and of collagen. At the highest doses of Eupolin extract, 200 to 400 /xg/ml, the contraction of collagen gels was completely prevented (Figs. 1 and 2). No evi­dence of cell viability was shown at 300 or 400 /tg/ml. At high concentrations of collagen (1 or 1.3 mg/ml), collagen contracted slowly and 100 /xg/ml of extract completely inhibited contrac­tion up to day 7 or day 9 (Figs. 1 and 2). In ad­dition, medium supplemented with FCS can make floating collagen gels contract faster (Delvoye et a l , 1991). We also noticed this in our experiments when old medium was removed and fresh DMEM/10% FCS with or without ex­tract was added.

In our previous studies, 10 or 100 /xg/ml of Eupolin extract had mitogenic effects on fi­broblasts and endothelial cells in monolayer culture (Phan et al., 1996 unpublished). However, the metabolism of fibroblasts in a three-dimensional matrix in vitro is nearer to that in vivo and is very different in many re­spects from that in a monolayer (Coulomb et al., 1983; Dubertret et a l , 1988) and human fi­broblasts show retarded cell division when sus­pended in collagen, particularly in floating col­lagen gels (Nakagawa et al., 1989; Greco et al., 1992; Hughes et al., 1992).

We have shown that the phenomenon of re­versibility or otherwise of the contraction inhi­bition by Eupolin extract is correlated with the viability and morphology of the cells. At 200 /xg/ml or higher, inhibition of contraction was complete from time zero and correlated with rounded cell morphology and ultimate cell death. From 50 to 150 / tg/ml, it was possible to reverse the inhibition by washing out the Eupolin extract and visualization by MTT in­dicated that many of the cells were alive and capable of changing to the elongated form.

To our knowledge this is the first scientific investigation of an extract from Chromolaena odorata leaves on collagen lattice contraction as a model for wound contraction in vivo.

Clinically, wound contraction is very compli­cated and depends on many factors. The my­ofibroblasts present in neoplasm have a con­tractile nature (Skalli et al., 1988; Rudolph et a l , 1992; Hebda et al., 1993). Different phenotypes of fibroblasts derived from granulating wounds or hypertrophic scars increased colla­gen gel contraction (Hughes et al., 1992; Garner et a l , 1995). Type III collagen, which occurs at a higher percentage in healing wounds and scars than in normal skin, has been shown to contract faster than type I (Ehrlich, 1988; Rudolph et al., 1992; Cherry et a l , 1994). In ad­dition, the cell-cell interactions and the role of mediators (growth factors and cytokines) also have a considerable effect on wound contrac­tion (Souren et al., 1989; Chen et al., 1991; Hebda et al., 1993; Garner et a l , 1995).

It was interesting to note that contraction of three-dimensional collagen lattices in this study was inhibited by the same concentrations of Eupolin extract that had stimulated the pro­liferation of fibroblasts in the monolayer sys­tem observed in our previous studies. In our preliminary results, the findings that Eupolin extract inhibits collagen lattice contraction and that the inhibition of contraction is reversible support the suggestion that the extract might be of therapeutic significance in minimizing postburn scar contracture and deformities.

REFERENCES

Akah PA. Mechanism of hemostatic activity of Eupatorium odoratum. Int J Crude Drug Res 1990; 28(4):253-256.

Bamba D, Bessiere JM, Marion C, Pelissier Y, Fouraste I. Essential oil of Eupatorium odoratum. Planta-medica 1993;59(2):184-185.

Bell E, Ivarsson B, Merril C. Production of a tissue-like structure by contraction of collagen lattices by human fibroblasts of different proliferative potentials in vitro. Proc Natl Acad Sci USA 1979;76:1274-1278.

Chen JK, Green G, Weinberg CB. Cultured vascular en­dothelial cells secrete a protein factor(s) that promotes the contraction of collagen lattices made with fibrob­lasts. Exp Cell Res 1991;193:297-302.

Cherry GW, Hughes MA, Kingsnorth AN, Arnold FW. Wound healing. In: Morris PJ, Malt RA, eds. Oxford Textbook of Surgery, Vol 1. Oxford University Press, Oxford, 1994, pp. 3-23,

Coulomb B, Dubertret L, Bell E, Merril C, Fosse M, Breton-Gorius J, Prost C, Touraine R. Endogenous peroxidases

AN AQUEOUS EXTRACT OF THE LEAVES OF CHROMOLAENA 343

in normal human dermis: a marker of fibroblast differ­entiation. J Invest Dermatol 1983;81:75-78.

Delvoye P, Wiliquet P, Leveque JL, Nusgens BV, Lapiere CM. Measurement of mechanical forces generated by skin fibroblasts embedded in a three-dimensional col­lagen gel. J Invest Dermatol 1991;97:898-902.

Dubertret L, Coulomb B. Dermo-epidermal interactions, wound healing and psoriasis. In: Ryan TJ, (ed.) Beyond Occlusion: Dermatology Proceedings (Int. Congress Symp. Series. 137. R). Royal Society of Medicine, London, 1988, p. 35.

Ehrlich PH. The modulation of contraction of fibroblast populated collagen lattices by type I, II and III colla­gen. Tissue Cell 1988, 20(l):47-50.

Garner WL, Rittenberg T, Ehrlich PH, Kamiol S, Rodriguez JL, Smith DJ, Phan SH. Hypertrophic scar fi­broblasts accelerate collagen gel contraction. Wound Rep Reg 1995;3:185-191.

Greco RM, Ehrlich PH. Differences in cell division and thymidine incorporation with rat and primate fibrob­lasts in collagen lattices. Tissue Cell 1992;24(6):843-851.

Hebda PA, Collins MA, Tharp MD. Mast cell and myofi­broblast in wound healing. Dermatol Clin 1993;11(4): 685^96.

Holm R, et al. The World's Worst Weeds: Distribution and Biology. Kreiger, Melbourne, Florida, 1991.

Hughes MA, Cherry GW, Dawber RPR, Ryan TJ. Minoxidil-induced changes in the contraction of colla­gen lattices by human skin fibroblasts. Plast Reconstr Surg 1992;89(4):722-730.

Irobi ON. Activities of Chromolaena odorata (Compositae) leaf extract against Pseudomonas aeroginosa and Streptococcus faecalis. J Ethnopharmacol 1992;37(1): 81-83.

Le TT, The use of Eupolin prepared from Eupatorium odoratum to treat soft tissue wounds. The 5th European Tissue Repair Society Meeting, Padova, Italy, 1995 (Abstract 30).

Montesano R, Orci L. Transforming growth factor /3 stim­ulates coUegen-matrix contraction by fibroblasts: Implications for wound healing. Proc Natl Acad Sci USA 1988;85:4894-4897.

Nakagawa S, Pawelek P, Grinnell F. Long-term culture of fibroblasts in contracted collagen gels: Effects on cell growth and biosynthetic activity. J Invest Dermatol 1989;93:792-798.

Nghiem DP. The therapeutic effects of the extract from the leaves of Eupatorium odoratum on the infection of soft tissue and non-healing wounds (summary of Ph.D. thesis). Code 3.01.21. Hanoi, 1992 (in Vietnamese).

Parish JL, Hughes MA, Cherry GW, Ferguson DJP. The effect of minoxidil analogues and metabolites on the contraction of collagen lattices by human skin fibrob­lasts. Br J Plast Surg 1995;48:15^160.

Rudolph R, Berg JV, Ehrlich HP. Wound contraction and scar contracture. In: Cohen IK, Diegelmann RF, Lindblad WJ, eds. Wound Healing: Biochemical and Clinical Aspects. Sanders, Philadelphia, 1992.

Skalli O, Gabbiani G. The biology of the myofibroblast re­lationship to wound contraction and fibrocontractive diseases. In: Clark RAF, Henson PM, eds. The Molecular and Cellular Biology of Wound Repair-Plenum, New York, 1988.

Souren JM, Ponec M, Wijk RV. Contraction of collagen by human fibroblasts and keratinocytes. In Vitro Cell Dev Biol 1989;25(11):1039-1045,

Triratana T, Suwannuraks R, Naengchomnong W. Effect of Eupatorium odoratum on blood coagulation. J Med Assoc Thailand 1991;74(5):283-286.

Truong MK. The therapeutic effects of the extract from the leaves of Eupatorium odoratum (Eupolin) on the in­fection in dento-alveolitis (summary of Ph.D. thesis). Code 3.01.52. Hanoi, 1989 (in Vietnamese).

Address reprint requests to: MA. Hughes

Wound Healing Institute Department of Dermatology

The Churchill Old Road

Headington, Oxford OX3 7LJ, UK