Effects on wound healing of zinc oxide in a

hydrocolloid dressing
Magnus S. Agren, phD,a, b Lennart Franzen, MD,b and Milos Chvapil, MDc
Miami, Florida; Linkoping, Sweden; and Tucson, Arizona

Background: Zincoxideincorporated ingauzeenhances healing ofchronic wounds inhumans

and experimental pig wounds.
Objective: The purposeof this study wasto examine the effects of zincoxide added to a hy-
drocolloid dressing on the healing of surgical wounds in domestic pigs.
Methods: Fortypartial-thickness wounds (2.2 X 2.2ernand 400,um deep) were treated with
differentzinc oxide concentrations, and epithelialization wasevaluated morphometrically in
a total of 320 histologic sections. Wound closure, bacterialgrowth, and inflammation were
studiedin eightfull-thickness wounds (2.5 X 4.5em).The level ofserumzincwas determined
beforeand after treatment.
Results: In partial-thickness wounds, concentrations of zinc oxide at or below 1.0% (wt/wt)
inhibited epithelialization, whereas noeffect wasobserved at zinc oxide concentrations from
2% to 6%.In full-thickness wounds, zincoxide (6%) reduced bacterial growth by about210g
unitsand increased the inflammatory response in the granulation tissue, but hadnoeffect on
healing when compared with control (hydrocolloid alone). Serum zinc levels remained un-
changed throughout the treatment period.
Conclusion: Apart frominhibiting bacterialgrowth, noadditional beneficial effects onwound
healing in nutritionally balanced pigswerefound by supplementing a hydrocolloid dressing
with zinc oxide.
(J AM ACAD DERMATOL 1993;29:221-7.)

Topical zinc oxideis a common therapy for var-
ious skin disorders because of its protective, astrin-
gent, and antiseptic properties. We have recently
shown that zinc oxide incorporated in gauze pro-
moteshealingof leg ulcers in humans and of acute
wounds in experimental pigs. I, 2
Zinc oxide is the major activecomponent of the
Unna boot. Because the Unna boot has beenfound
to be either superior' or equal4, 5 to a hydrocolloid
occlusive dressing in the treatment of leg ulcers in
humans, wedecided to studythe combination ofzinc
oxide and a hydrocolloid in an experimental model.
Sixconcentrations of zincoxideweretested, andthe
From the Department of Dermatology and Cutaneous Surgery, Uni-
versity of Miami School of Medicine, Miami"; Department of
Pathology, Facultyof HealthSciences, Linkoping''; and the Depart-
ment of Surgery,University of Arizona, Tucson,"
Accepted for publication Feb. II, 1993.
Reprint requests: MagnusS. Agren, Coloplast A/S, Holtedam I, DK-
3050 Humlebaek, Denmark.
Copyright 1993 by the American Academyof Dermatology, Inc.
0190.9622/93 $1.00 +.10 16/1/46482
effects on the healing of partial- and full-thickness
wounds were examined in domestic pigs.
MATERIAL AND METHODS
Dressings
Thehydrocolloid dressing consists ofanadhesive mass,
which is laminated to a polyurethane foam. Polyisobuty-
lene(a hydrophobic synthetic polymer) andthreetypes of
hydrophilic particles (carboxymethyl cellulose sodium,
pectin, and gelatin) compose the adhesive mass, which
adheres bothto dryand wetskin surfaces. Zincoxide was
mixed homogeneously into the adhesive mass. Theincor-
poration of zincoxide (2%, wt/wt) altered the adhesion
to intact human skin only slightly (10% increase com-
pared with control hydrocolloid) as measured in 10
healthyvolunteers.
Animals
Five female domestic piglets oftheYorkshirestrain(18
to 21 kg) were used. Animals were housed in separate
cagesand given a zinc-adequate diet (86mg of zinc/kg)
and tap water as needed. Theanimals were anesthetized
with halothane.

221

222 Agren et al.
Partial-thickness wounds
The dose-dependency of zincoxideon epithelialization
was studied by adding to the hydrocolloid 6 concentra-
tions of zincoxide (0.1 %, 0.25%, 0.5%, 1.0%, 2.0%, and
6.00/0 as calculated by weight of the adhesive). The stan-
dard hydrocolloid dressing served as a control treatment.
In vitro experiments showed that zincionswere released
to HEPES-buffered (pH 7.4) salinein proportion to the
square root of the zincoxide concentration.
Sixteen partial-thickness wounds (2.2 X 2.2 em and
400 /lm deep) weremadewith an electrokeratome in the
paravertebral and thoracic area in each of two pigs. 2, 6
The proportion of epidermal to total thickness of excised
formalin-fixed skin was 40%. The zinc oxide dressings
wererandomly allocated to one pair of adjacent wounds
and the control dressing was randomly allocated to two
pairs of adjacent wounds in each pig. The reproducibility
of our resultswasinvestigated in an additional pig treated
with the 6% zinc oxide and the control hydrocolloids.
After 48 hours of treatment without intermediate
dressing changes the wounds were excised, fixed in 4%
formaldehyde, and embedded in paraffin. Epithelialcov-
erage wasevaluated morphometrically in eightrandomly
selected histologic sections stained with hematoxylin-
eosin per wound as described by Chvapil et al.6 The
thickness of the new epithelium coverage was measured
at five equidistant points alongthe wound in the control
group and in the 2% and 6% zinc oxide groups. The in-
flammatory response was assessed according to the den-
sityof inflammatorycells in the dermison a 2-pointscale
as follows: 1 = slight; 2 = moderate. All histologic eval-
uations were carried out in a blindedfashion.
Full-thickness wounds
Four full-thickness skin wounds (2.5 X 4.5 em) ex-
tending to the muscular layer were made with a scalpel
in each oftwo pigs. Twowounds on eachsideof the spine
were treated with the same dressing in each pig. Hy-
drocolloid dressings were trimmedwith scissors to fit the
woundcavities, applied to thewound surface,and covered
with gauzeswabsand an adhesive polyurethane film.The
hydrocolloid dressings were changed on postoperative
days 2, 4, 7, and 9.
The wound area was measured planimetrically from
tracings made on a transparent plastic film of the outer
margins of the wounds. The wound volume was deter-
mined by measuring the amount of sterilesaline needed
to fillthe cavityto the level of the surrounding skin.?Per-
centagechangefrom the initialvalueswascalculatedand
used for comparison between the two groups.
Bacterial counts and alkaline phosphatase were mea-
sured in twoadjacent 4-mmpunchbiopsy specimens ob-
tained from the centerof the wounds 4 and 11 days post-
operatively. The mean wet weight of each specimen was
50 mg. The specimen designated for bacterial culturing
Journal of the American Academy of Dermatology
August 1993
was firstdippedin absoluteethanoland flamedto remove
superficial bacteria.f Tenfolddilutions of biopsy specimen
homogenate (5 ml) were made in salineand 200 /ll of the
homogenate was plated on nutrient agar (Microbio,
Tempe, Ariz.). The number of colony-forming units
(CFUs) werecounted after 48 hours of incubation in air
at 37 C.
The tissuealkalinephosphataseactivity(ALP),used as
a marker for inflammation," was determined after ho-
mogenizing the specimens in 1 ml of 0.02 mol/L TRIS
as described by Prasad et al.lO with kit No. 245 from
Sigma Chemical Co. (St. Louis, Mo.).
Histologic examination was performed blindly on
transverse sections through the central part of the wound
in biopsyspecimens taken on day 11 and stained with he-
matoxylin and eosin. Epithelialization from the wound
edges was determined and was measured as the percent-
age of .the total wound surface length covered with
epithelium. The tissuewasrated accordingto the density
of inflammatory cells on a 3-point scale as follows:
1 =. slight; 2 = moderate; 3 = pronounced. The volume
densityof foam cells was assessed by point counting in a
250/lm wide and 2.5 mm long zone 100 /lm below the
woundsurface. An eyepiece mounted square point grid
with 100 points at 25 um distances from each other was
used.
Zinc measurements
Serum zincwasdetermineddirectlyafter wounding, at
dressing changes, and when the animals were killed.? To
investigate whether zinc migrated from the zinc-treated
partial-thickness wounds, 4 mm punch biopsyspecimens
weretaken from the skinbetweenthe wounds. Superficial
contaminating zinc was tangentially removed (lao /lm)
fromfrozenbiopsy specimens witha microtome. The skin
and wound tissues were delipidated in chloroform:meth-
anol3:1 (vol/vol) and driedat 105 C to constantweight.
The dried samples were dissolved in 12N nitric acid, de-
colored and disintegrated with 30% hydrogen peroxide
solution, evaporated to dryness, and rehydrated with 10
ml of Milli-Q water.
The zincconcentration of the solutions wasdetermined
witha flameatomic absorption spectrometer. A reference
material (NBS bovine liver 1577a, Washington, D.C.)
processed simultaneously gave 122.6 2.0 /lgl gm
(n = 6) compared with the certified value of 123 8
/lg/gm
Statistics
Duncan's multiple range test was applied to the
epithelialization data. The bacterial counts (CFUs per
gram) were logarithmically transformed before being
subjectedto statistical analysis. Student's t test for paired
observations was used to analyze the data from the full-
thickness wounds. A probability of lessthan 0.05 was ac-
 Journal of the American Academy of Dermatology
Volume 29, Number 2, Part I Agren et al. 223

100

~ 80
----
(1)
0>
(1j
..... 60
CD
::-
0
o
E 40
.~
Q5
.!::
+-'
'5. 20
LU

a
o 0.1 0.25 0.5 1 2 6
Zinc oxide concentration (%, w/w)
Fig. 1. Effectof zinc oxide(0 [control], 0.1 %, 0.25%, 0.5%, 1.0%, 2.0%, and 6.0% wt/wt)
on the epithelial resurfacing of partial-thickness wounds after 48 hours of treatment
(mean SEM). Each mean valueis basedon measurement ofat least 24 histologic sections
from at least 4 wounds. Asterisk above bar indicates significant difference (p < 0.05) com-
pared with control.

cepted as significant. Numeric data are presented as
mean standard error of the mean.
RESULTS
Partial-thickness wounds
In partial-thickness wounds 1% and lower con-
centrations of zinc oxide retarded epithelialization,
whereas 2% and 6% zinc oxide were equal to the
control hydrocolloid dressing (Fig. 1). No significant
difference (p> 0.05) was found between the 6%
zinc oxide and control hydrocolloids when the ex-
periment was repeated; the zinc oxide-treated
wounds wereepitheJialized to 68.1% 4.3% (n = 4)
and the control-treated wounds to 74.7% 2.0%
(n = 4).
The thickness of new wound epithelium did not
differ significantly (p> 0.05) between the control-
treated (38.6 2.5 ,urn, n = 8), 2% zinc oxide-
treated (40.6 4.2 ,urn, n = 4), and 6% zinc oxide-
treated wounds (37.9 3.2 ,urn, n = 4).
The inflammatory response did not show any ap-
parent difference between controls (mean score
1.75) and zinc-treated wounds irrespective of the
zinc oxide concentration (mean score 1.5to 2.0). No
foam cells were found in the partial-thickness
wounds.
The serum zinc level was 0.69 ,uglml before and
0.70 ,ug/ml after treatment. The zinc concentration
of the skin adjacent to the zinc oxide-treated wounds
was 29.0 2.4 JLg/gm of dry weight (n = 7) com-
pared with 33.0 2.4,ug/gmofdryweight(n = 6)
for skin adjacent to control-treated wounds. These
results indicate that zinc was not transported later-
ally between the wounds nor was it absorbed sys-
temically in any appreciable amount.
Full-thickness wounds
Wound healing. The initial wound area was
11.8 0.4 cm 2 and volume was 6.1 0.2 ml
(n = 8). Wound area started to decrease on day 4,
and the wounds were filled with new granulation
tissue to the level of the skin by day 7 (Fig. 2). No
differences in either wound area or volume were
found between the two treatments. As determined
histologically, the control-treated wounds were epi-
thelialized by 63.0% 7.3% and the zinc-treated
wounds by 51.5% 3.0% (n ='4 in both groups), a
nonsignificant difference (p > 0.05).
Bacteria. Bacteria were recovered from all eight
wound biopsy specimens on day 4, whereas on day
11 bacterial growth was found in five of the eight
specimens. Significantly (p < 0.05) fewer bacterial
colonies were recovered from the zinc oxide-treated
wounds. On day 4 the difference in the bacterial
 Journal of the American Academy of Dermatology
224 Agren et al. August 1993

100 Volume

80
Area
60

40

20

o ..,...c::=--.....rr
-20 +----.----..,.---..,---..,.---,.--
o 2 4 6 8 10
Postoperative day
Fig. 2. Effect of zinc oxide (6%) on closure of full-thickness wounds as measured by area
(lower pair ofcurves) and volume (upper pair ofcurves) changes. Variability is omitted for
reasons of clarity (coefficient of variation was typically 15% to 20%). Dotted line, Control-
treated wounds; solid line, zinc oxide-treated wounds.

Table I. Bacterial growth and alkaline phosphatase activity (ALP) in granulation tissue from
full-thickness pig wounds
Bacterial growth (loglO CFU/gm) ALP (TU/g)

Control
I Zinc oxide Control
I Zinc oxide

Day 4 6.6 0.6 4.8 0.4* 180 58 340 40

Day II 2.8 1.0 1.7 1.0 65 12 96 16
Data are presented as mean standard error of the mean.
*p < 0.05 compared with control.

count between the two groups was 1.8 log units (Ta-
ble I).
Alkaline phosphatase. The activity decreased sig-
nificantly from day 4 to day 11 in both groups but
was higher in the zinc oxide group at both time
points (Table I).
Histologic features. The inflammatory reaction
was more marked in 6% zinc oxide-treated wounds
that always showed a pronounced inflammation
(mean score 3.0), whereas the inflammation in con-
trols was slight to moderate (1.75 0.25). Foam
cells were seen underneath the wound surface in all
specimens but were more frequent in zinc-treated
wounds (Fig. 3). The volume density of foam cells
as assessed morphometrically was 42.7 9.2 in
zinc-treated woundsand5.8 4.5 in control wounds
(n = 4 in both groups), a significant difference
(p < 0.01). Vacuoles were found in both groups but
no granulomata, giant cells, or birefringent crystals
were found in the granulation tissue of either group.
Zinc analyses. Serum zinc levels did not change
during the course of treatment (Fig. 4). Zinc con-
centration in the wounds treated with the plain hy-
drocolloid was significantly higher (p < 0.001) than
in adjacent uninjured skin (78.8 3.9 ,ug/gm of dry
weight versus 21.4 3.0 ug] gm of dry weight). In
zinc-treated wounds the zinc concentration varied
greater (range 440 to 3490 ,ug/gm of dry weight)
because of the presence of zinc oxide aggregates.
Attempts to extract tissue-bound zinc with EDTA,
1,1O-phenanthroline in ethanol or trypsin, proved to
be unsuccessful. For example, treatment of a wound
biopsy specimen with 0.25% (wt/vol) trypsin in
TRIS-HCI (pH 9.0) for 1 hour at 37 C resulted in
incomplete solubilization (45%) of available zinc. In
addition, more than 65% of known zinc oxide quan-
Journal of the American Academy of Dermatology
Volume 29, Number 2, Part 1 Agren et al. 225

Fig. 3. Photomicrographs ofgranulationtissueapproximately 100 J.tm belowwoundsurface

of 6% zinc oxide-treated (A) and control-treated wound (B). Pronounced inflammation and
numerousfoam cellsare seen in zincoxide-treated wound (A). No foam cells are seen in pho-
tomicrograph from control-treated wound (B), and only scattered leukocytesappear in gran-
ulation tissue. (Hematoxylin-eosin stain; X310.)

tities added to the wound tissue were solubilized with
the trypsin treatment.
DISCUSSION
In previous studies, zinc oxide enhanced healing
of skin wounds when incorporated in a gauze vehi-
cle. 1,2 In the present study another vehicle, a hy-
drocolloid occlusive dressing, was investigated. No
stimulatory action of zinc oxide in the hydrocolloid
vehicle was found on the healing of superficial or
deep skin wounds.
The serum zinc level did not change during the
course of treatment. However, because of the rela-
tively small wound area receiving topical zinc ( < 1%
of the total body surface area), the absorbed zinc
could probably not be detected in serum. Studies in
rats have shown that zinc is' absorbed through
wounds from the 6% zinc oxide hydrocolloid and
even more so than any other zinc oxide preparation
we have examined. I I
The only beneficial effect of zinc oxide was the
reduced bacterial growth in the granulation tissue,
demonstrated previously in wounds in rats and
guinea pigs,S, 9 The lower total bacterial counts in

226 Agren et al.
1.0
0.8,-
0.6
o
c:
'N
0.4
E
2(J.)
(j)
0.2
0.0 +----~---,_---_r_---T_---~-
o 2
Postoperative day
Fig. 4. Serum zinc levels (mean of two pigs) during COurse of treatment of full-thickness
wounds.
zinc oxide-treated wounds did not seem to influence
healin g. Although most control-treated wounds on
postoper ative day 4 had bacterial counts exceeding
106 CFU/ gm, a threshold value for infection in hu-
mans, these wounds did not show clinical signs of
infection. Thus the wounds were probably only col-
onized by natural flora.
A hydrocolloid dressing fulfills many require-
ments for an ideal wound cover. However, its adhe-
sive disintegrates when applied to wounds; part of it
is absorbed into the wound tissue, giving rise to foam
celJs, that is, altered macrophages identified earlier
during treatment of full-thickness skin wounds in
rats, 12 and pigs. l3 , 14 Foam cells were more common
in zinc-treated wounds. One possible explanation
ma y be that zinc modifies one or more of the poly-
mers, rendering them more absorb able. The zinc
oxide-supplemented dressing also elicited a more
pronounced inflammatory reaction than the control
dressing in the full-thickness wounds, possibly be-
cause of deposition of zinc oxide in the granulation
tissue.
Although zinc oxide was found ineffective in ac-
celerating the healing rate of porcine wounds, the
incorporation of zinc oxide might be justified for
other reasons. Because our experimental animals
were not zinc-deficient, the wounds were probably
healing at a maximal rate under the control hy-
drocolloid dressing. Zinc-deficient patients, in con-
Journal of the American Academy of Dermatology
August 1993
---.....
468 10
trast, show delayed healing and may still respond to
the zinc oxide-medicated hydrocolloid dressing. An
ant ibacterial action may be advantageous in a com-
promised host, especially when using occlusive dress-
ings that increase bacterial growth in wounds. I5
Zinc oxide can also reduce skin maceration, 16 a side
effect seen sometimes after prolonged occlusion.!?
REFERENCES
1. Stromberg H-E, Agren MS. Topical zinc oxide improves
arterial and venous leg ulcers. Br J Dermatol 1984;
111 :461-8.
2. Agr en MS, Chvapil M, Franzen L. Enhancementof re-ep-
ithelialization with topical zinc oxide in porcine partial-
thickness wounds. J Surg Res 1991;50:101-5.
3. Kikta MJ, Schuler JJ, Meyer JP, et a!. A prospective, ran-
domized trial of Unna's boots versus hydroactive dressing
in the treatment of venous stasis ulcers. J Vase Surg
1988;7:478-83.
4. Eriksson G. Comparison of two occlusive bandages in the
treatmentofvenouslegulcers. BrJ Dermatol1986;114:227-
30.
5. Cordts PR, Hanrahan LM, Rodriguez AA, et al. A
prospective, randomized trial of Unna's boot versus Duo-
derm CGF hydroactive dressing plus compression in the
management of venous leg ulcers. J Vase Surg 1992;
15:480-6.
6. Chvapil M, Gaines JA, Chvapil T, et aJ. An optimal mor-
phometric methodfor quantitatingwoundepithelialization.
J Surg Res 1988;44:266-77.
7. Berg W, Traneroth C, Gunnarsson A, et al. A method for
measuring pressure sores. Lancet 1990;335: 1445-6.
8. Hallmans G, Elmros T. Zinc tape treatment of burns
infected by Pseudomonas aeruginosa, Scand J Plast Re-
constr Surg 1980; 14:221-7.
Journal of the American Academyof Dermatology
Volume 29, Number 2, Part I
9. AgrenMS, SOderberg TA, Reuterving C-O,et al. Effectof
topicalzincoxideon bacterial growthand inflammation in
full-thickness skinwounds in normal and diabetic rats. Eur
J Surg 1991;157:97-101.
10. Prasad AS, Oberleas 0, Miller ER, et al. Biochemical ef-
fectsof zincdeficiency: changes in activities of zinc-depen-
dent enzymes and ribonucleic acid and deoxyribonucleic
acid content of tissues. ] Lab Clin Med 1971;77:144-52.
II. Agren MS. Influence of two vehicles for zincoxideon zinc
absorption through intact skin and wounds. Acta Derm
Venereol (Stockh) 1991;71:153-6.
12. Agren MS, Franzen L. Foam cells after treatment with
hydrocolloid dressings [Letter]. Br J Dermatol 1991;
125:193.
13. Leek MD, Barlow YM. Tissue reactions induced by
hydrocolloid wounddressings. J Anat 1992;180:545-51.
Agren et al. 227
14. Young SR,Dyson M, Hickman R, et al. Comparison of the
effects of semi-occlusive polyurethane dressings and hy-
drocolloid dressings on dermal repair: 1. Cellular changes.
] Invest DermatoI1991;97:586-92.
15. SOderberg T, Agren M, Tengrup I, et al. The effects of an
occlusive zincmedicated dressing on the bacterial flora in
excised wounds in the rat. Infection 1989;17:81-5.
16. Auger P, Colin P, Joly J, et al, Treatment of cutaneous
candidosis in guinea pigs: effect of zinc oxide on the
antifungal efficacy of nystatin. Mycoses 1989;32:455-60.
17. Apelqvist J, Larsson J, Stenstrom A. Topical treatmentof
necrotic foot ulcers in diabetic patients: a comparative trial
ofDuoDerm and MeZinc. BrJ DermatoI1990;123:787-92.