Exposure of oral mucosa to bioactive milk factors reduces severityof chemotherapy-induced mucositis in the hamster

Julie Clarkea, Ross Butlerb, Gordon Howartha, Leanna Reada, Geoff Regestera,*aCooperative Research Centre for Tissue Growth and Repair, Child Health Research Institute, North Adelaide, South Australia 5006, Australia

bDepartment of Gastroenterology, Women’s and Children’s Hospital, North Adelaide, South Australia 5006, Australia

Received 28 August 2001; accepted 12 September 2001

Abstract

A biologically active extract containing bovine whey proteins, whey growth factor extract-A (WGFE-A) was administered topi-

cally to the oral mucosa of hamsters and its ability to prevent and treat chemotherapy-induced oral mucositis investigated. Oralmucositis was induced in Syrian golden hamsters through a combination treatment of the antimetabolite chemotherapy drug 5-fluorouracil (5-FU), and mild abrasion of the cheek pouch. WGFE-A administered to the oral mucosa via hydrogel and liquidtreatments, pre and concurrent to 5-FU therapy, resulted in significantly reduced mucosal ulceration. The protective effect was dose

dependent with greatest benefit from WGFE-A doses applied at 4.2 mg/ml gel and 14 mg/ml mouthwash (P<0.01). The protectiveactivity of WGFE-A also appeared related to mode of delivery. Administration of WGFE-A from an alternate vehicle Orabase1

did not alleviate mucositis compared to WGFE-A applied in hydrogel. When administered continuously after the chemotherapy

schedule, WGFE-A failed to reduce ulcer area when applied over a 12-day period. In a separate study, cell cycle staining indicatedthat cheek pouch mucosal epithelial cells pre-exposed to WGFE-A in-vivo showed a reduced rate of proliferation, measured as a21% reduction in the bromodeoxyuridine (BrdU) cell labelling index (P<0.04). This was consistent with a protective mode ofWGFE-A action against anti-metabolites such as 5-FUwhich target rapidly dividing cells. The results were also consistent with recentin vitro data showing protective properties from WGFE-A administered to epithelial cells given pre/concurrent to chemotherapyexposure. WGFE-A is known to contain mitogens which stimulate cells of mesenchymal origin and inhibit epithelial cell growth in

culture. Several WGFE-A constituents are likely to confer protective effects on the cheek mucosa, including anti-proliferative, anti-apoptotic and anti-microbial factors. WGFE-A provides a potentially valuable source of topically delivered proteins for clinicalapplication in preventing severe oral mucositis caused by chemotherapy. # 2002 Elsevier Science Ltd. All rights reserved.

Keywords:Mucositis; Ulcer; Whey protein; Hamster; Chemotherapy; Growth factors

1. Introduction

Oral mucositis is a major dose limiting side effect ofchemotherapy for which there is presently no effectivepreventative therapy. Currently 800,000 patients receivechemotherapy in the United States each year [1], a highpercentage of which develop oral and/or intestinalmucositis. Chemotherapy dose has been identified as amajor predictor of clinical outcome in different cancertreatment regimens [2]. Oncologists recognise that tointensify drug therapy and improve treatment prognosis,it will be necessary to overcome associated toxicities

arising from indiscriminate chemotherapy targeting ofrapidly dividing cells lining the mouth and gastro-intestinal tract.The severity of oral mucositis is partly determined by

cycling status of the epithelial cell layer [3]. Paediatricpatients with a higher epithelial cell turnover generallyexperience a higher incidence of mucositis comparedwith those over the age of 60 [4]. Furthermore, increas-ing rates of cell proliferation in the oral and intestinaltract using growth factors such as EGF markedlyincreases the severity of mucositis in animal models [5].Biologically active factors are now being considered fortheir potential efficacy in preventing and/or treatingmucositis. Granulocyte-colony stimulating factor (GM-CSF] [6], keratinocyte growth factor (KGF) [7], inter-leukin-11 (IL-11) [8] and transforming growth factor-b3(TGF-b3) [9,10] reportedly reduce the severity of

1368-8375/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.

PI I : S1368-8375(01 )00107-5

Oral Oncology 38 (2002) 478–485

www.elsevier.com/locate/oraloncology

* Corresponding author. Present address: GroPep Limited, PO Box

10065 BC, Adelaide 5000 Australia. Tel.: +61-8-8354-7717; fax: +61-

8-8354-7788.

E-mail address: geoff.regester@gropep.com.au (G. Regester).

mucositis when applied as pre-treatments to chemo-therapy and/or in the repair phase. While most of thisdata has been demonstrated using in-vivo models, theevidence of human efficacy with these factors is limited.Widespread application of these therapies may also berestricted due to prohibitive treatment costs associatedwith commercial manufacture of recombinant proteins.Bovine milk and colostrum are rich sources of natural

anti-bacterial factors, cytokines and growth factors withbiological activity required for development of thenewborn and support of the gastrointestinal tract [11–15]. Milk also contains survival factors such as IGF-I,which regulate cell death by apoptosis [16–18]. The useof a cheese whey protein fraction as a source of growthfactors for sustaining mammalian cells in culture hasbeen reported [19]. This whey protein extract hasalso been identified as an important source of mito-gens (IGF-I and II, FGF, PDGF and TGF-b) formesodermal-derived cells in culture, while being inhibi-tory toward the growth of epithelial cell lines includingrat, canine and mink lung epithelial cells [20,21].A recent examination of in vitro cell survival activity

has demonstrated the whey extract to be cytoprotectiveagainst at least two chemotherapy drugs when adminis-tered onto epithelial cells before and concurrent withdrug exposure [22]. This study provided evidence of cellsurvival activity conferred by WGFE and demonstratedthat TGF-b2, a principle milk growth factor, was partlyresponsible for this protection. This effect was alsoassociated with marked reduction in DNA synthesis andpartial cell cycle blockage. Interestingly TGF-b2 couldnot account fully for the protection shown in culturedcells suggesting other WGFE factors were involved.This work also supported earlier studies with an invivo model of intestinal epithelial damage whichdemonstrated oral administration of WGFE amelio-rated crypt ablation by the chemotherapy agentmethotrexate [23].These studies provided a basis for the present examina-

tion which considers two treatment schedules using asimilar mixture of topically administered whey proteinsfor prevention and repair of oral mucositis in anestablished animal model. The work also highlights thepotential opportunities for clinical application of thisgrowth factor mixture as a topically delivered chemo-protective agent.

2. Material and methods

2.1. Preparation of whey growth factor extract-A(WGFE-A)

The method of enrichment of bioactive growth factorsin cheese whey has been described previously [19].Briefly, pre-clarified whey was applied to a strong

cation-exchange resin, SP-Sepharose ‘‘Big Beads’’(Amersham-Pharmacia, Castle Hill, NSW) and thebound protein fraction containing growth factors elutedin 0.5 M NaCl containing 10 mM sodium citrate, pH6.5. Ultrafiltration of the eluate provided a final proteinconcentration of 40 mg/ml. To facilitate dissociation oflatent protein complexes, and to release growth factorfrom binding proteins, the extract was acidified usingHCl, then re-adjusted to pH 6.5 with 1M NaOH. Thewhey extract contains amixture of growth factors includingInsulin-like growth factor I and II, Transforminggrowth factor-b and Fibroblast growth factor [19–21].

2.2. Oral mucositis model

Use of the hamster was based on the experimentalmucositis model of Sonis et al. [24]. Young castratedmale Syrian golden hamsters (157�2.8 g) were pur-chased from the University of Otago, New Zealand.They were caged individually and fed commercialrodent diet (Ridley Agriproducts Pty Ltd, MurrayBridge SA, Australia, New Joint Stock Ration).Chemotherapy treatment involved two intraperitonealinjections of 5-fluorouracil (5-FU) (Delta West1 Pty Ltd,Bentley, WA, Australia) administered at 90 and 60 mg/kgon days 0 and 2 of the trial. In combination with 5-FU, theleft cheek pouches were everted and lightly scratched witha small wire brush on day 1 and again on day 2 imme-diately prior to the second chemotherapy treatment.Hamsters were anaesthetised with isoflurane (Forth-

ane1, Abbott Australasia, Kurnell NSW, Australia)before treatments were applied. WGFE-A was adminis-tered as a liquid (0.3 ml solution/application across adose range of 4, 14, 24, 40 and 120 mg/ml) and incor-porated into hydrogel (Intrasite1, Smith and NephewPty Ltd, Clayton Vic, Australia) for application as atopical smear (0.2 ml gel/application across a dose rangeof 1.2, 4.2, 7.2, 12 and 120 mg/ml). Mouthwash wasadministered into the left cheek pouch and remained incontact with the abraded area for 2 min. The cheekpouch was then drained, everted and the gel applieddirectly onto the abraded area. Hamsters in the controlgroups were given mouthwash and gel containingradioimmunoassay grade bovine serum albumin (Sigma,Castle Hill NSW, Australia, product number A7888) innormal saline at equivalent protein concentrations.Assessment of the ulcers and animal condition have

been described in more detail previously [25]. Briefly,the hamsters were weighed daily, and ulcers assessedevery day from day 4 or 5 until day 15. Assessmentincluded tracing the margin of the ulcer for calculationof ulcer area by image analysis and photographing thecheek pouch. Each cheek pouch was given a visual score(ulcer score) graded on a 0–10 scale. This was based onan assessment of ulcer size and the degree of bruising,swelling and scarring (0–3 scale for each parameter).

J. Clarke et al. / Oral Oncology 38 (2002) 478–485 479

In a separate experiment, another topical vehiclefor WGFE-A was compared with the hydrogel carrier.WGFE-A was incorporated into Kenalog1 in Ora-base1 (Bristol-Myers Squibb Pharmaceuticals, NoblePark, Vic, Australia) and applied at the protective dosefound for hydrogel (4.2 mg/ml) in combination withliquid WGFE-A at 14 mg/ml.All animal experimentation was approved by the

Animal Ethics Committee of the Women’s and Chil-dren’s Hospital, Adelaide and complied with the Aus-tralian Code of Practice for the Use of Animals inResearch and Teaching (1990).

2.3. Time course of oral mucositis

A time-course study was undertaken to examine theeffects of 5-FU chemotherapy on white cell counts(WCC) and healing of the cheek pouch epithelium.Untreated hamsters were sacrificed on day 0 as a base-line group. The remaining hamsters received BSAtreatments to the cheek pouch three times over the 28 hprior to their first injection of 5-FU and twice daily untilimmediately after the second injection of 5-FU wasadministered. Three hamsters were killed every secondday from days 5 to 15 inclusive of the time course.Blood samples were submitted for complete bloodcounts (Veterinary Pathology Services, Adelaide, Aus-tralia) and the ulcer was assessed as described aboveduring the time-course study.

2.4. Prophylactic WGFE-A administration

A total of five prophylactic doses were compared inthe hamster model ranging between 0.4 and 120 mgWGFE-A/ml of mouthwash and between 0.12 and 120mg WGFE-A/ml of hydrogel. Selection of these doseswas based on the range and extent of inhibitory activityin cultured epithelial cell lines. The combined treatmentof WGFE-A fortified mouthwash and gel was selectedto give optimal exposure time. During the experimentsthe cheek pouch treatments were applied three timesover the 28 h prior to the first injection of 5-FU, andwere repeated twice daily until immediately after thesecond injection of 5-FU was administered. Cheekpouches were not treated after day 2. The number ofscratches to the cheek pouch and severity of the initialwound was identical for all hamsters.

2.5. Post chemotherapy WGFE-A administration

A separate trial was performed to determine the effectof WGFE-A treatment administered during the healingphase of the mucositis model. There were six animals inthe vehicle group and five survived in the WGFE-Agroup. Administration of WGFE-A after chemotherapywas applied at 12 mg/ml hydrogel and 40 mg/ml

mouthwash with treatments given daily from days 3 to14 inclusive. A higher dose of WGFE-A was applied inthe post chemotherapy schedule, providing higher levelsof epithelial cell mitogens necessary for restoration ofdamaged buccal mucosa. These factors, primarilyepithelial cell mitogens, are generally found at low con-centration in bovine milk and whey.The severity and number of scratches applied to cheek

pouches on day 1 was identical for all hamsters in thepost chemotherapy trial, whereas on day 2 the number ofscratches was adjusted to produce ulcers of similar size.

2.6. Epithelial proliferation

To examine the effects of WGFE-A on epithelial pro-liferation, WGFE-A or control treatments (eight ham-sters/treatment group) were applied three times duringthe 30 h prior to sacrifice at �30, �24 and �6 h. Thecheek pouches of all hamsters were lightly abradedimmediately prior to the initial WGFE-A application.WGFE-A was applied at a uniform dose of 4 mg/mlmouthwash and 1.2 mg/ml gel to all hamsters. Thehamsters were injected intraperitoneally with 50 mg/kgbromodeoxyuridine (BrdU; Sigma, Castle Hill NSW,Australia, product number A7888) 2 h prior to sacrifice.At sacrifice the cheek pouches were collected and tissuesfixed and orientated to provide linear sections for his-tology. Sections taken through the abraded region wereprepared and stained for BrdU as previously described[26]. Slides were scored with a Video Pro Image Analy-sis package and Zeiss light microscope. The labellingindex was determined by counting the number of BrdUlabelled cells/total epithelial cells over five fields of view(magnification �250) on each side of the ulcer, andexpressed as a percentage. For each ulcer 20 paraffinsections of 5 mm thickness were counted. Preliminarystudies using this protocol showed a moderate varia-bility with a coefficient of variation (CV) of 17.8%. Theproliferative zone from the ulcer was measured bycounting the number of graticule units in the ocular lensfrom the edge of the ulcer, to the end of the proliferativezone. The end of the proliferative zone was defined asthe point where two consecutive fields of view had novisible labelled epithelial cells. These measurementsallowed an assessment of the degree of proliferation,and using the size of the proliferative zone, the extent ofthe response to the applied agents.

2.7. Statistics

To assess the area under the curve for ulcer score andarea values, a curve was fitted to the ulcer area and cheekpouch score data for each hamster using Table Curve(Jandel San Rafael CA, USA). Lowess smoothing wasapplied at a level of 40%, and an asymmetric transitionfunction equation was used for all ulcers. The equation

480 J. Clarke et al. / Oral Oncology 38 (2002) 478–485

was used to calculate area under the curve for eachhamster’s ulcer area and score. The time at which the firstderivative minimum, or the time at which maximumhealing occurred, was recorded for the hamsters whichhad been given the optimal WGFE-A dose. Data werecompared using either unpaired two-tailed t-tests, or ifindicated, Mann–Whitney non-parametric tests.

3. Results

The mortality rate in the trials described in this paperwas 3.1%. The animal losses were unrelated to theirtreatments, and included two from vehicle and one fromWGFE-A groups. The data from these animals wereexcluded from the analysis.

3.1. Time course of oral mucositis

Haematological analysis showed the dose of 5-FUadministered resulted in severe immunosuppression byday 5 (Fig. 1). Hamsters responded with a strongregenerative leukocytosis, which peaked at day 9. Manyanimals had activated lymphocytes, and several animalswere thrombocytopenic, had reduced pack cell volumesand red blood cell counts, indicating they had under-gone acute haemolytic crises (data not shown). Theulcer area of BSA (vehicle) treated hamsters from theprophylactic trials reported in this paper is superimposedon the same graph (Fig. 1), and shows restoration ofintact mucosa by day 15. Typically, the rate of ulcerhealing plateaus between days 5 and 9 before a period ofmaximum healing between days 9 and 13.

3.2. Prophylactic WGFE-A administration

Administration of WGFE-A prior to and duringtreatment with chemotherapy covered a dose range

between 0.4 and 120 mg WGFE-A/ml of mouthwash,and between 0.12 and 120 mg WGFE-A/ml of hydrogel.Vehicle groups were treated with the equivalent BSAprotein concentration in each trial.The severity of mucositis was measured as the area

under ulcer area and ulcer score curves. The vehicletreated groups responded similarly in all trials with nosignificant difference between trials in either the ulcerarea or score across the five doses (P>0.05 by ANOVA,Table 1). Compared with the vehicle groups, a statisti-cally significant benefit in ulcer area and score wasrecorded from treatment at the mid-range WGFE-Adose, 14 mg/ml mouthwash and 4.2 mg/ml gel. Therewere no significant benefits from the higher or lowerWGFE-A concentration treatments (40–120 mg/mlmouthwash, 12–120 mg/ml gel; and 0.4–4 mg/ml mouth-wash, 0.12–1.2 mg/ml gel, respectively).The dose ranging trials (Table 1) suggested that 14

mgWGFE-A/ml mouthwash in combination with 4.2 mgWGFE-A/ml hydrogel were the most protective treat-ment concentrations. Three independent experimentsadministered WGFE-A at this dose. Combined ulcerarea data for the three trials at this dose is presentedgraphically in Fig. 2. Treatment with WGFE-A sig-nificantly reduced the ulcer areas on days 8, 10, 11 and12 (P<0.05 to P<0.01). It also reduced the total areaunder the healing curve (P<0.007). The day of maxi-mum healing for WGFE-A treated hamsters occurredon day 10.2, which was significantly earlier than the dayof maximum healing for the vehicle treated group at day11.3 (P<0.05). Photographs of cheek pouch ulcersrepresentative of WGFE-A and vehicle treated hamsterson day 10 of the trials are shown in Fig. 3. Loss of bodyweight during the trials was less for hamsters treatedwith WGFE-A compared to hamsters treated with BSA

Fig. 1. Mean ulcer areas (n=37) and white blood cell counts

(WCC;�109/l; n=2 or 3) of 5-FU treated hamsters treated prophy-

lactically with BSA (mean�S.E.M.). Triangles: ulcer area; squares:

white cell count.

Fig. 2. Ulcer areas of hamsters injected with 5-FU and treated pro-

phylactically with WGFE-A (n=19 WGFE-A, n=18 vehicle). The bar

chart shows the mean AUC of ulcer area for the same two groups of

animals. Levels of significance: *P<0.05; **P<0.01.

J. Clarke et al. / Oral Oncology 38 (2002) 478–485 481

(day 0 and day 15 body weights for WGFE-A andvehicle hamsters respectively were 184.8�5.7 and186.9�7.0, and 162.7�4.5 and 156.9�5.4 g). The dif-ference in body weights on day 15 was not significant(P=0.42).To gauge the efficiency of WGFE-A delivery to the

buccal cells from impregnated hydrogel, we testedthe effectiveness of WGFE-A in an alternate deliveryvehicle, Kenalog1 in Orabase1. The most protectiveWGFE-A concentration identified from the dose rang-ing experiments (Table 1) at 14 mg/ml mouthwash and4.2 mg/ml gel was used for comparison. Fig. 4a illus-trates that WGFE-A delivered in Kenalog1 did notsignificantly reduce the ulcer area, whereas WGFE-Aadministered in the hydrogel base (Fig. 4b) caused a

significant reduction in ulcer area on days 10(*P<0.03), 11 (**P<0.04) and 12 (***P<0.02).

3.3. Post Chemotherapy WGFE-A administration

The effect of WGFE-A administered post chemo-therapy to the cheek pouch was measured and is shownin Fig. 5. There were no significant effects of WGFE-Aapplications on ulcer size; the cheek pouch scores fol-lowed a pattern similar to ulcer area with no significanttreatment effects (data not shown). The body weights ofWGFE-A and BSA continuously treated groups werealso not significantly different during the trial (data notshown).

3.4. Epithelial proliferation

A 21% decrease in BrdU labelling per total countedcells expressed as labelling index%BrdU (P<0.04)was recorded in WGFE-A treated cheek pouch

Table 1

Mean areas under the curves of ulcer areas (AUC) and ulcer scores

(mm2) for hamsters injected with 5-FU and treated prophylactically

with doses of WGFE-A or vehicle (mean�S.E.M.)

Protein liquid

(mg/ml)

Protein gel

(mg/ml)

Ulcer areas AUC

(mm2)

Ulcer scores AUC

(mm2)

Vehicle WGFF-A Vehicle WGFF-A

0.4 0.12 184.0 185.8 29.4 25.6

27.2 24.0 4.8 3.1

n=5 n=6 n=6 n=6

4 1.2 231.3 197.1 31.9 25.5

50.7 24.0 6.2 3.3

n=5 n=5 n=5 n=6

4 1.2 184.0 191.2 29.4 26.9

27.2 21.7 4.8 2.9

n=5 n=6 n=5 n=6

14 4.2 184.0 144.7 29.4 16.9

27.2 35.1 4.8 1.6

n=5 n=5 n=6 n=5

14 4.2 258.8 174.6 34.9 24.6*

35.1 24.0 3.5 3.1

n=6 n=7 n=6 n=7

14 4.2 233.5 190.01* 33.0 26.5

9.3 17.1 1.2 3.0

n=7 n=7 n=4 n=6

24 7.2 266.1 218.4 36.7 33.3

33.5 19.9 3.5 2.5

n=9 n=9 n=9 n=9

24 7.2 256.7 171.1* 34.4 24.7

34.6 14.5 4.1 2.1

n=5 n=6 n=5 n=6

40 12 231.3 223.4 31.9 31.2

50.7 33.6 6.2 4.2

n=5 n=6 n=5 n=6

120 120 231.3 239.8 31.9 34.2

50.7 28.8 6.2 4.0

n=5 n=6 n=5 n=6

*WGFE-A and vehicle were significantly different (P<0.05).

Fig. 3. Photographs of cheek pouches of 5-FU injected hamsters on

day 10 of the trial after treatment with (a) WGFE-A and (b) BSA

preparations (14 mg protein/ml in liquid and 4.2 mg/ml in gel).

Treatments were applied using the prophylactic administration proce-

dure as described.

482 J. Clarke et al. / Oral Oncology 38 (2002) 478–485

(15.2�0.8%; mean�S.E.M.) compared to BSA treatedbuccal mucosa (19.3�1.4%). The proliferative zonemeasured by BrdU labelling at the margin of the ulcerwas examined in the WGFE-A treated animals (7.4�0.5units) and found to be reduced by 26% (P<0.02)compared with an equivalent BSA treatment (9.8�0.6units).

4. Discussion

In the present study, topical administration of bio-active milk proteins (WGFE-A) to the oral mucosa sig-nificantly reduced the severity of mucositis in hamsterswhen applied before and concurrent to chemotherapy(prophylactic protocol). In contrast, WGFE-A admi-

nistered continuously after chemotherapy through themucosal repair phase did not provide any benefit com-pared to applications of a control protein source.Oral mucositis results from non-specific drug and

radiation-induced damage to rapidly multiplying basaloral epithelial cells, the severity of which is partlydetermined by rate of cell cycling [3]. Evidence of a linkbetween rate of cell turnover and cell damage caused bychemotherapy comes from studies which demonstratesevere mucositis following treatment with epithelial cellproliferative agents such as EGF [5]; also by partialprevention of mucositis after topical treatment with acell cycle arresting agent TGF-b3 [9,10]. Other studieshave shown benefits from topical and systemic applica-tion of growth factors to oral and gastrointestinalmucositis, by stimulation of a localised immuneresponse [6,8]. Similarly, growth factors such as kerati-nocyte growth factor [7] or IGF-I [27] have beendemonstrated to enhance the rate of mucosal repairfollowing chemotherapy-induced injury. In the presentstudy, maximum ulcer area coincided with peakimmunosuppression between days 5 and 7 of the trial,supporting the relationship between low immune com-petence and mucositis severity.The fractionated whey protein extract WGFE-A

contains a cocktail of growth factors including TGF-b,IGF-I, IGF-II, acidic and basic FGF, and PDGF, withinhibitory and stimulatory activity on proliferation andfunction of different cell types [20,21]. We believe theactions of several WGFE-A components administeredprophylactically are responsible for protection of theoral mucosa against chemotherapy-induced damage.These results are partly consistent with a protectiveeffect shown with TGF-b in the same mucositis model[9,10]. TGF-b induced reduction in basal epithelial cell

Fig. 4. Ulcer areas of cheek pouches of hamsters injected with 5-FU

and treated with WGFE-A and vehicle using (a) Orabase1 and (b)

Hydrogel delivery systems. The bar charts show the mean AUC of

ulcer area for the two groups of animals given the same delivery sys-

tem. For (a) Orabase1, n=7 for vehicle and WGFE-A; for (b)

hydrogel n=7 for WGFE-A and n=6 for vehicle. Levels of sig-

nificance *P<0.05, **P<0.01.

Fig. 5. Ulcer areas of cheek pouches of hamsters treated post-

chemotherapy with WGFE-A or vehicle after injection with 5-FU

(n=6 for vehicle, n=5 for WGFE-A).

J. Clarke et al. / Oral Oncology 38 (2002) 478–485 483

proliferation at the time of chemotherapy has been citedas the protective mechanism in this animal model. Wealso demonstrated a significant inhibitory effect on basalepithelial cell proliferation by WGFE-A. A difference inthe proliferative response was also evident between theWGFE-A treated and control tissue, with a significantlyreduced proliferative zone from the ulcer margin inWGFE-A treated animals. Whether this was a directeffect related to diminution of the proliferative activityof epithelial cells by the action of TGF-b, an indirecteffect on the immune response or alteration to the com-ponents of the cellular matrix, remains to be elucidated.However the beneficial effect from WGFE-A treatmentmay not be solely related to TGF-b, given protectivedoses of recombinant TGF-b were 20-fold higher thanthose administered via WGFE-A in the present study[9,10]. Recent in vitro findings [22] using WGFE-A atlevels below its inhibitory concentration in epithelialcells have demonstrated protection against cytotoxicdrug damage from these samples, suggesting TGF-b isnot the sole chemoprotective factor in WGFE-A. Thiswas also confirmed through immuno-neutralisationstudies of TGF-b present in WGFE-A which removedall growth inhibitory activity but did not eliminate itscell survival activity against the chemotherapy agentvinblastine [22]. This study also suggested that a pre/concurrent drug administration schedule was necessaryfor survival activity in cell culture, consistent with thepresent in vivo findings.Other growth factors present in WGFE-A have been

shown to protect cells against chemotherapy-induceddeath by apoptosis. WGFE-A is rich in IGF-I (>20 ng/mg protein) [21], which is recognised as an anti-apoptoticfactor. IGF-I prevents apoptosis at low concentrations(10�9 M) in differentiated PC12 cells, by activation ofmultiple signal transduction pathways and throughupregulation of bcl-xL expression [16,17]. IGF-I hasalso been shown to modulate chemotherapy-inducedapoptosis in three dimensional culture systems withhuman breast cancer cells HBL001, providing arationale to develop serum IGF-I lowering drugs toimprove chemotherapy targeting [18]. Levels of IGF-Iadministered to the cheek pouch from WGFE-A in thepresent study calculated at micromolar concentration,fall within the biologically active ‘‘anti-apoptotic’’ rangedemonstrated in cultured cells, and may be responsiblein part for the protection observed.Other bioactive proteins in WGFE-A may play a role

in preventing severe mucositis. Oral administration ofbovine lactoferrin, an antibacterial milk protein [28],has been shown to improve intractable stomatitis and theimmune defence system in felines with immuno-deficiency virus [29]. This activity was attributable tolactoferrin mediated activation of neutrophil phago-cytosis. WGFE-A also contains trace amounts of lacto-ferrin (1–5% w/w protein) which may act to assist

prevention of mucositis. It also contains high levels oflactoperoxidase (>20% w/w protein), which has beendemonstrated to be an effective bacteriostatic agentthrough generation of thiocyanate SCN-oxidationproducts which attack sulfhydryl (SH) groups inmicrobial metabolic enzymes [29]. Application ofantimicrobial agents such as lactoferrin and lacto-peroxidase to the mucosal surface at time of damage,may be important in preventing secondary tissuedamage by infectious agents.WGFE-A delivered prophylactically at its effective

concentration in Orabase1 gel, did not provide anyprotection against mucositis. In contrast administrationof hydrogel containing WGFE-A at equivalent dose,effectively reduced the severity of mucositis. This alsosuggests that the choice of vehicle is important foreffective application of the bioactive factors.The lack of efficacy from continuous application of

WGFE-A suggests a mode of action which is probablynot due to stimulation of cell growth. Indeed, weobserved no increased rate of repair following repeatedapplication of WGFE-A after the mucosal injury. Thisin vivo finding also supports in vitro observationsmade previously with WGFE-A in cultured epithelialcells where it does not promote growth of this cell type[21].In summary, the protective mechanism of WGFE-A

administered prior to and during chemotherapy treat-ment probably represents the combined effect of differ-ent growth factors and bioactive proteins which offeranti-infectious, anti-proliferative and anti-apoptoticactivity. The results presented are important in light ofcurrent clinical demands for mucositis treatmentoptions. Administration of a natural milk extract at thetime of chemotherapy treatment could be used either toreduce symptoms of mucositis in standard cytotoxictherapy regimens, or alternatively, to increase thedosage of chemotherapy agents, thereby improvingpatient prognosis.

Acknowledgements

The authors wish to thank Leanne Srpek and BenEdwards for their contribution in conducting the animaltrials, Quang Doan for preparing the WGFE-A samples,and Joanne Pech for sectioning and staining of histolo-gical specimens. The Co-operative Research Centre forTissue Growth and Repair is also acknowledged for itssupport of this study.

References

[1] Mueller BA, Millheim ET, Farrington EA, Brusko C, Wiser TH.

Mucositis management practices for hospitalized patients:

National survey results. J Pain Symptom Manage 1995;10:510–8.

484 J. Clarke et al. / Oral Oncology 38 (2002) 478–485

[2] Quilitz R. Modulation of chemotherapy-induced mucositis. Can-

cer Control 1995;2(5):452–9.

[3] Sonis AL, Sonis ST. Complications of cancer chemotherapy in

pediatric patients. J Pedod 1979;3:122–8.

[4] Sonis ST. Oral complications of cancer therapy. In: DeVita VT,

Hellman S, Rosenberg SA, editors. Cancer: principles and prac-

tice of oncology. 4th ed. Philadephia, PA: J.B. Lippincott Co,

1993. p. 2385–94.

[5] Sonis ST, Costa JW, Evitts SM, Lindquist LE, Nicolson M.

Effect of epidermal growth factor on ulcerative mucositis in

hamsters that receive cancer chemotherapy. Oral Surg Oral Med

Oral Pathol 1992;74:749–55.

[6] Kannan V, Bapsy PP, Anantha N, Doval DC, Vaithianathan H,

Banumathy G, et al. Efficacy and safety of granulocyte macro-

phage-colony stimulating factor (GM-CSF) on the frequency and

severity of radiation mucositis in patients with head and neck

carcinoma. Int J Radiat Oncol Biol Phys 1997;37(5):1005–10.

[7] Farrell CL, Bready JV, Rex KL, Chen JN, DiPalma CR, Whit-

comb KL, et al. Keratinocyte growth factor protects mice from

chemotherapy and radiation-induced gastrointestinal injury and

mortality. Cancer Res 1998;58:1–7.

[8] Sonis ST, Van Vugt AG, McDonald J, Dotoli E, Schwertschlag

U, Szklut P, et al. Mitigating effects of interleukin 11 on con-

secutive courses of 5-fluorouracil-induced ulcerative mucositis in

hamsters. Cytokine 1997;9(8):605–12.

[9] Sonis ST, Lindquist L, Van Vugt A, Stewart AA, Stam K, Qu GY,

et al. Prevention of chemotherapy-induced ulcerative mucositis

by transforming growth factor-b3. Cancer Res 1994;54:1135–8.[10] Sonis ST, Van Vugt AG, Brien JPO, Muska AD, Bruskin AM,

Rose A, et al. Transforming growth factor-b3 mediated modula-tion of cell cycling and attenuation of 5-fluorouracil induced oral

mucositis. Oral Oncol 1997;33(1):47–54.

[11] Donovan SM, Odle J. Growth factors in milk as mediators of

infant development. Annu Rev Nutr 1994;14:147–67.

[12] Schams D. Growth Factors in Milk. Endocr Rev 1994;28:3–8.

[13] Grosvenor CE, Picciano MF, Baumrucker CR. Hormones and

growth factors in milk. Endocr Rev 1993;14:710–28.

[14] Tome D, Debabbi H. Physiological effects of milk protein com-

ponents. Int Dairy Journal 1998;8:383–92.

[15] Kelly D, Coutts AGP. Early nutrition and the development of

immune function in the neonate. Pro Nutr Soc 2000;59:177–85.

[16] Parrizas M, Saltiel AR, LeRoith D. Insulin-like growth factor-1

inhibits apoptosis using the phosphatidylinositol 30-kinase and

mitogen-activated protein kinase pathways. J Biol Chem 1997;

272(1):154–61.

[17] Parrizas M, LeRoith D. Insulin-like growth factor-1 inhibition of

apoptosis is associated with increased expression of the bcl-xL

gene product. Endocrinology 1997;138(3):1355–8.

[18] Dunn SE, Hardman RA, Kari FW, Barrett JC. Insulin-like

growth factor-1 (IGF-I) alters drug sensitivity of HBL100 human

breast cancer cells by inhibition of apoptosis induced by diverse

anticancer drugs. Cancer Res 1997;7(13):2687–93.

[19] Francis GL, Regester GO, Webb HA, Ballard FJ. Extraction from

cheese whey by cation exchange chromatography of factors that sti-

mulate the growth of mammalian cells. J Dairy Sci 1995;78:1209–18.

[20] Belford DA, Rogers M, Regester GO, Francis GL, Smithers GW,

Liepe IJ, et al. Milk-derived growth factors as serum supplements

for the growth of fibroblast and epithelium cells. In vitro Cell

Dev Biol 1995;31:752–60.

[21] Belford DA, Rogers ML, Francis GL, Payne C, Ballard FJ,

Goddard C, et al. Platelet-derived growth factor, insulin-like

growth factors, fibroblast growth factors and transforming

growth factor b do not account for the cell growth activity pre-sent in bovine milk. J Endocrinol 1997;154:45–55.

[22] Taylor VL, Goddard C, Read LC. A milk growth factor extract

reduces chemotherapeutic drug toxicity in epithelial cells in vitro. In

vitroCellular andDevelopmental Biology (Animal) 2001 (in press).

[23] Howarth GS, Francis GL, Cool JC, Xu X, Byard RW, Read LC,

et al. Milk growth factors enriched from cheese whey ameliorate

intestinal damage by methotrexate when administered orally to

rats. J Nutr 1996;126:2519–30.

[24] Sonis ST, Tracey C, Shklar G, Jensen J, Florine D. An animal

model for mucositis induced by cancer chemotherapy. Oral Surg

Oral Med Oral Pathol 1990;69:437–43.

[25] Clarke JM, Edwards B, Srpek L, Regester G. Evaluation of bovine

lactoferrin as a topical therapy for chemotherapy-induced muco-

sitis in the golden Syrian hamster. Oral Oncol 1999;35:197–202.

[26] Howarth GS, Xian CX, Read LC. Predisposition to colonic dys-

plasia is unaffected by continuous administration of insulin-like

growth factor-I for twenty weeks in a rat model of chronic

inflammatory bowel disease. Growth Factors 2000;18:119–33.

[27] Howarth GS, Cool JC, Bourne AJ, Ballard FJ, Read LC. Insulin-

like growth factor-I (IGF-I) stimulates regrowth of the damaged

intestine in rats, when administered following, but not concurrent

with, methotrexate. Growth Factors 1998;15:279–92.

[28] Dionysius DA, Grieve PA, Milne JM. Forms of lactoferrin: their

antibacterial effect on enterotoxigenic Escherichia coli. J Dairy

Sci 1993;76:2597–606.

[29] Sato R, Inanami O, Tanaka Y, Takase M, Naito Y. Oral

administration of bovine lactoferrin for treatment of intractable

stomatitis in feline immunodeficiency virus (FIV)-positive and

FIV-negative cats. Am J Vet Res 1996;57(10):1443–6.

J. Clarke et al. / Oral Oncology 38 (2002) 478–485 485