Posology,Efficacy,and Safety of Epidermal Growth Factor Eye Drops in 305 Patients: Logistic Regression and Group-Wise Odds of Published Data

Abstract

The aim of this article is to investigate clinical research on indications, posology, efficacy, and safety of epidermal growth factor (EGF) eye drops in the treatment of some human corneal disorders. Methods used include systematic search and selection of series of cases and clinical trials in Medline database up to January 2012, kappa index (K) to validate retrieval information, cumulative Mantel-Haenszel-stratified meta-analysis, 2×2 contingency table of randomized EGF-vehicle-controlled treated groups, and statistical program SPSSv12. Our results indicate that EGF eye drops appear to be a very effective treatment of acute heterogeneous corneal diseases, without significant adverse effects, with a 86.8% clinical efficacy reported by authors, a 98% (P<0.05) probabilistic expected efficacy, and 51.3 (17.4–148.7 confidence interval 95%; P<0.05) odds ratio EGF/vehicle. However, clinical trials are scarce, with low sample sizes and serious inconsistencies in EGF posology. EGF eye drops (50–1,000 ng, 2–3 times/day) could be a useful treatment for promoting postoperative refractive surgery, reversing cases of keratopathy secondary to systematic EGF receptor inhibitors, diabetic keratopathy, and other corneal and conjunctival disorders.

Introduction

Epidermal growth factor (EGF) is a potent mitogen and probably the most relevant growth factor constitutively present in animal and human tears. In the cornea EGF acts by binding to its saturable-high specific transmembrane tyrosine-kinase receptor (EGF receptor [EGFR]), increasing cell growth, proliferation, differentiation, stratification, migration, and wound healing. EGFR activation is crucial in accelerating corneal wound repair following injury and it seems that murine recombinant EGF (mrEGF) and human recombinant EGF (hrEGF) can be used interchangeably in animals and patients with very similar results, both being effective and well tolerated.^1–3

It has been previously proved in the laboratory that topical hrEGF treatment of cancerous epithelial cells in culture with concentrations as low as 1 ng/mL was sufficient to modify the autocrine response of this factor.^4,5 This is in accordance with EGF constitutive secretion found in human corneal epithelium cell cultivation supernatants (about 2.4 ng/mL),⁶ and also with physiological EGF concentrations found in human tears (about 0.2–10 ng/mL).^7–11 It has been demonstrated that EGF drops at sustained concentrations ranging between 0.7 and 50 ng/mL induce mitotic and reparative activity in the corneal epithelium, keratocytes, and endothelial cultured cells.^12–17 A single maximal EGF exposure (10 ng/mL) induces EGFR internalization and increases extracellular signal-regulated kinases phosphorylation (ERK) in human lens cells.¹⁵ EGF eye drops (5 ng/mL) stimulate corneal epithelial cell proliferation significantly as well as wound healing in vitro.¹³ Dosimetric studies indicate that EGF promotes proliferation of corneal endothelial cells, demonstrating a linear dose-dependent effect when its concentration is lower than 10 ng/mL. This effect decreases when its concentration is higher than 30 ng/mL.¹⁴ EGF eye drops (50 ng/mL) are effective in corneal keratocyte and myofibroblast differentiation and in modulation of wound healing and scar tissue formation.¹² hrEGF (10 ng/mL) promotes endothelial wound healing in corneas from senior donors preserved in organ culture medium.¹⁶ Recently, using a human corneal wound healing model, reepithelialization rate was optimally accelerated by EGF at concentration of 25 ng/mL.¹⁷

During the 1980s–1990s optimal and safe posology of EGF eye drops in animals was established in 50,000 ng/mL/2–3 times a day. Such concentrations have shown occasional adverse effects, including significant delay in corneal wound healing both in rabbits and horses.^18,19 Other in vivo experiments with animal eyes have shown that EGF eye drops (100–1,000 ng/8–12 h/10 days) are an efficacious agent for accelerating and improving the restorative process after acute-experimental corneal damage.^20–23 EGF eye drops have demonstrated efficacy in different animal models of cornea injuries at diverse doses. For instance, (1) in rabbits postpenetrating keratoplasty, EGF eye drops (1,000 ng/8 h/14 days) increase the wound strength up to 600-fold compared with the control eyes—results demonstrated by tensile strength and histopathological studies (P<0.05).²⁰ (2) In full-thickness corneal wounded rabbit eyes, EGF eye drops (1,000 ng/12 h/7 days) accelerate wound strength and tensile strength (P<0.05).²¹ (3) In diabetic mice, EGF eye drops (100 ng/12 h/15 days) accelerate and improve caustic wound healing, accelerating reduction of wound diameter (P<0.001).²² (4) EGF eye drops (100 ng/12 h/15 days) accelerate and improve the healing of alkali-burned nondiabetic mice corneal wounds—results demonstrated by wound diameter reduction and exhaustive histopathological study during the treatment (P<0.05).²³

A limited number of clinical cases and clinical trials in patients indicate the use of EGF eye drops as a particularly effective and safe treatment for accelerating healing and reducing the reepithelialization time after acute corneal damage. However, conflictive results have been reported and main indications and optimal posology of EGF eye drops in clinical practice remain unclear. In these circumstances, we have meta-analyzed all articles referenced in Medline that provide data about indications, posology, efficacy, and safety of EGF eye drops in patients. The study has been designed to investigate the theoretical indications of EGF eye drops and to compare the efficacy and safety of EGF eye drops versus vehicle eye drops containing identical ophthalmic solution except EGF.

Methods

We performed a systematic search and selection of abstracts in Medline database up to January 2012, by means of recommended strategies for scientific information retrieval.^24–26 Simple crosses and Boolean operators were used to combine the following key words and descriptors: epidermal growth factor/EGF/cornea/corneal/treatment/topical/eye drops/eyedrops/human/patients. Search and selection of articles was conducted by 2 independent investigators, with the following inclusion criteria:

(1) EGF eye drops are used to treat some type of corneal damage or disorder in patients.

(2) An objective method is used to measure and evaluate the results.

Search and selection results were validated using the Cohen's kappa index (K). Selected abstracts were requested in full-text version and they were meta-analyzed after translation from German, Chinese, and French to English. To summarize and quantify current evidence, we conducted a cumulative Mantel-Haenszel meta-analysis through stratification of comparative/combinable groups and tabulation of variables. The advantage of this method is that it allows to deduce, and occasionally demonstrate, sources of homogeneity and discrepancies in the results between different studies.^26–28

The following data were collected: murine/human EGF, indication of use, concentration of EGF used, number of daily applications, maximal duration of the treatment, number of eyes treated with EGF, number of homologous vehicle-controlled treated eyes, methods used to measure the results, follow-up period after the end of the treatment, clinical efficacy (CE) reported by authors, and adverse effects found. Logistic regression only included comparable groups with significant efficacy (P<0.05), no significant efficacy (P>0.05), or significant damage (P<0.05). A 2×2 contingency table was established to compare the effect of sample size and strength of association between EGF eye drops and vehicle eye drops (containing identical drop components, except EGF).

Statistical results were calculated using the computer program SPSSv.12 for Windows, including mean (m), range (r), percentages (%), odds ratio [OR, confidence interval (CI) 95%], probability of predictable efficacy of EGF eye drops (P), and statistical significance (P). The limitations of the method and the interpretation of the statistical results were based on logical-deductive-critical reasoning.^29,30

Results

The search/selection strategy yielded 19 articles from the title and abstract content (K=0.76), which were exhaustively analyzed in full text.^31–49 Four articles were excluded because they did not use EGF eye drops.^39,45,48,49 Two articles were excluded because they described a unique case reported to occur in the same patient, which was uncontrolled and what is more, the German-written version failed to indicate the doses used.^46,47 Lastly, an important article was excluded because it was not a piece of original research but a review of clinical trials made in China, which will be discussed later on.⁴²

The results of the remaining 12 articles (305 EGF-treated eyes) were stratified in 18 comparable groups and meta-analyzed in Table 1. Methods and parameters used to evaluate the efficacy of EGF eye drops are shown in Table 2. Group-wise odds between binary data (CE inefficacy; EGF vehicle; 10 groups, 206 eyes) are calculated in Table 3. In the reviewed articles the concentration of EGF eye drops has been expressed in the orders of mg/mL, μg/mL, and ng/mL. To unify, we have indicated EGF eye drops concentration in nanograms per milliliter.

Table 1.

Descriptive Results in 305 Epidermal Growth-Factor-Treated Eyes (18 Comparative, No-Combinable Groups)

Group	Ref.	Language	Indication	EGF origin	ng/mL	t (days)	No. of times a day	Eyes EGF	Eyes vehicle	CE reported by authors	P value
1	31	E	HCD	NS	2,000,000	UWH	6	23	0	Yes	NSA
2	31	E	HCD	NS	2,000,000	UWH	6	19	19	Yes	NSA
3	31	E	HCD	NS	2,000,000	10	6	17	17	Yes	NSA
4	31	E	Herpetic dendritic ulcers	NS	2,000,000	10	6	10	0	Yes	NSA
5	31	E	Bullous keratoplasty	NS	2,000,000	UWH	6	4	0	No	NSA
6	32	E	Penetrating keratoplasty	Murine	2,000,000	4	6/2	8	8	No	NSA
7	32	E	Penetrating keratoplasty	Murine	1,000,000	6	6/2	9	10	No	0.029
8	33	E	Severe burns	Murine	10,000	10	4	10	10	Yes	NSA
9	34	E	Epikeratoplasty	Murine	10,000	8	2	14	42	Yes	≤0.001
10	35	E	Traumatic epithelial defects	Murine	10,000	6	6	47	57	Yes	≤0.01
11	36	F	Epikeratoplasty	Murine	10,000	8	6	8	9	Yes	NSA
12	37	E	Traumatic ulcers	Murine	10,000	7	4	20	20	Yes	≤0.03
13	38	G	Perforating keratoplasty	NS	100	6	6	9	7	Yes	0.05
14	40	G	Perforating keratoplasty	Human	30,000	10	6	9	9	No	0.71
15	40	E	Perforating keratoplasty	Human	100,000	10	6	9	9	No	0.14
16	41	E	Herpetic corneal ulcers	Murine	10,000	12	4	20	20	Yes	0.013
17	43	Ch	Excision of pterigium	NS	20,000	7	4	49	51	Yes	≤0.05
18	44	Ch	Excision of pterigium	Human	20,000	8	4	20	20	Yes	≤0.01

Ref, number of reference; E, English; F, French; G, German; Ch, Chinese; HCD, heterogeneous corneal disorder; UWH, until wound healing; NS, not specified; CE, clinical efficacy; P, statistical significance; NSA, not statistical analysis; EGF, epidermal growth factor.

Table 2.

Methods and Parameters Used to Quantify the Results in Each Group of Table 1

Methods and parameters	Groups
Clinical evolution analyzed by the same investigators	1–18
Healing time/wound closure/reepithelialization time	1–18
Fluorescein staining/planimetry/epithelial wound size	1–7, 9, 12
Healing curve: epithelial wound area (mm²)/time	1–7, 15
Epithelial wound size on a semiquantitative grade rating scale	9, 10, 12, 16–18

Table 3.

2×2 Contingency Table Performed to Measure the Effect of Size in 206 Epidermal Growth-Factor-Treated Eyes Versus 245 Vehicle-Treated Eyes from the 10 Homogeneous and Combinable Groups Presented in Table 1

	Applied eye drops
Therapeutic result	EGF	Vehicle	Total
CE	179	28	207
NCE	27	217	244
Total	206	245

NCE, no increased clinical efficacy; CE, increased clinical efficacy.

Table 1 shows the descriptive results of 305 EGF-treated eyes from 18 comparative not combinable groups—m=mean, r=range (lowest-highest value). Principal results are as follows:

Language of publication: English (69%), Chinese (22%), German (6%), and French (3%), percentage related to the 305 EGF-treated eyes.

CE reported by authors: 266/305=87.2%.

Significant adverse effects: 2.8%. Tolerability of EGF eye drops was excellent in all treated eyes. There have been reports of mild conjunctival hyperemia and asymptomatic conjunctival injection in both eyes treated with EGF at high doses and those treated with the respective vehicle. The only adverse effect demonstrated statistically is the delayed wound healing when using EGF eye drops at doses of 1,000,000 ng/mL.³²

EGF concentration in eye drops (ng/mL): m=735,000, r=100–2,000,000.

Vehicle composition: Vehicle solution is usually a balanced salt solution (BSS) with diverse excipients, stabilizers, and conservants, frequently including mannitol, albumin, and other common components in artificial tears. Often conventional treatments are maintained including antibiotics and corticoids. In some cases, a possible synergistic effect with other active substances, such as fibronectin, collagen shields, and vitamin A, is studied separately.^33,38

Application (times/day): m=4.8, r=2–6.

Maximum duration of treatment (days): m=6.7, r=4–12.

Follow-up after the conclusion of treatment with EGF eye drops (months): m=5.3, r=0–24. Follow-up periods of maximum 2 years after the end of acute treatment have not demonstrated significant differences in possible late complications between the eyes treated with EGF and those treated with the vehicle. Some authors insinuate, without statistical proof, that patients treated with EGF show fewer recurrences of primary problems.

Table 2 shows that the procedure for evaluating the results clearly implies a degree of subjectivity and does not include histopathological studies. However, trials on animals prove that the clinical improvement in vivo following EGF eye drops is clear and is closely related to histological improvements in vitro.^20–23

Table 3 presents a 2×2 contingency table to measure the effect of size in 206 EGF-treated eyes versus 245 vehicle-treated eyes. Dichotomous variables were increased CE versus no-increased CE and EGF eye drops versus vehicle eye drops. The analysis of these results shows a CE of 179/206=86.8% (P<0.05), an OR=51.3 (17.4–148.7 CI 95%; P=0.05), a 98% probability (P) to obtain an effective result following EGF eye drops, and finally 4.3% of significant adverse effects (delayed healing) (P=OR/[OR+1]).

Interpretation of the results

Logistic regression and group-wise odds of significant randomized controlled eyes indicate that EGF eye drops are about 50 times more effective than vehicle eye drops in obtaining an accelerated and successful corneal wound healing, independently of the clinical indication and posology used. CE reported by authors is nearly 90%. Theoretical probability of obtaining effective results using EGF eye drops in future clinical trials is 98% (P<0.05). It was not possible to demonstrate any type of relation/association between the nature of the EGF molecule used, the cause of treatment, the concentration used, the composition of the vehicle solution, the number of daily applications, and the duration of treatment.

Weaknesses of the method

Clinical trials are scarce, with low sample sizes and a high degree of subjectivity in measuring the results and serious inconsistencies in the posology. Simpson's paradox has been rejected but other sources of bias, iceberg phenomenon, data integrity, reliability, and fraud are not controlled by our study. Unfortunately, any good meta-analysis of badly executed or fraudulent studies will result in a false groupal statistics.^50,51

Discussion

Since EGFR inhibitors started being used as systemic chemotherapy to treat some cancers, there have been puzzling reports of characteristic dermatological and ophthalmic adverse side effects, particularly acneiform pustular rash, xerosis, pruritus, follicular pustules, acquired trichomegaly, and persistent corneal epithelial defects. Ophthalmic damage secondary to systemic EGF inhibitors include severe dry eye syndrome, extensive and persisting corneal erosions in both eyes, trichomegaly, tear film dysfunction, meibomitis, telangiectasia of eyelid margins, and hyperpigmented and tortuous eyelashes.^45,48,49,52 These facts show the huge importance of EGF in the human cornea and conjunctiva, especially since EGF eye drops seem to revert the specific corneal damage due to systemic EGFR inhibitors.^46,47

Medline database is the main source of scientific and clinical information in ophthalmology. Since 1979 there have been only 12 relevant publications that have studied the efficiency of EGF eye drops in acute corneal problems. Ten of these studies indicate EGF as a potent epithelializer of great clinical interest.^{31,33,34–38,41,43,44} Only 2 assays emphasize that the treatment is either inefficient or counterproductive when used in concentrations higher than 30,000 ng/mL, 6 or more times a day.^32,40 Our statistical analysis of clinical trials on patients shows that EGF eye drops are an effective and safe treatment of heterogeneous and acute corneal disorders that heal in the short term, specifically in cases of perforating keratoplasty, corneal ulcers, traumatic epithelial defects, epikeratoplasty, burns, excision of pterigium, and isolated cases of other acute corneal injuries (see Tables 1 and 3). Such results are in agreement with some articles written in Chinese and with other data from the Chinese Institute of Basic Medical Sciences (Beijing), including a randomized, double-blind multicentric clinical trial demonstrating the high efficacy and safety of EGF eye drops (20,000 ng/mL) in accelerating and enhancing the corneal wound healing in 200 cases of cornea transplantation, and 247 cases of nebulae, without showing adverse effects.^42,43 The Chinese Drug Administration approves EGF eye drops for ophthalmic treatment, while the U.S. Food and Drug Administration approves topical hrEGF only for dermatological treatments, but not for corneal disorders. However, current knowledge, compiled in the introduction section of this article, suggests that the results of most of these assays are not reliable enough, since the dosage used was enormous. Besides, some other interesting applications of EGF eye drops in clinical practice could be explored.

The huge concentrations used in the most part of the studies pose an even more serious design flaw. Recently, the pharmacokinetics and pharmacodynamics of topical EGF in the skin and in the cornea have been reviewed, as well as the potential clinical indications to successfully treat a wider number of corneal disorders. It has been concluded that EGF eye drops (50–200 ng/mL/3 times a day) could be a first choice treatment for regenerating corneal damage in the medium-long term, in cases of systemic use of EGFR inhibitors, diabetic keratopathy, and especially in the treatment of dry eye syndrome.⁵²

Trials on animals have demonstrated the clinical possibilities of EGF eye drops in regenerating the cornea after acute damage. In patients only the indications in Table 1 have been tried. There are not yet clinical trials that demonstrate the need to incorporate EGF eye drops in the postoperative stage of refractive surgery (laser in situ keratomileusis [LASIK] and photorefractive keratectomy [PRK]). On the other hand, there are good arguments to suggest that the use of EGF eye drops for 1–2 weeks following the surgery can minimize the residual astigmatism and dry eye syndrome, usually linked to these types of procedures.^52,53–55 Keeping in mind the variation of concentrations of EGF in tears after penetrating keratoplasty³⁹ and the exponential diminution of the concentrations of EGF just a few hours after applying the eye drops, therapeutic concentrations of EGF eye drops of 100–1,000 ng/mL, 3 times a day, could be right for short-term treatments (<2 weeks). In theory, a dosage superior to 30,000 ng/mL could be ineffective or inhibitory, just like Kandarakis et al.³² and Dellaert et al.⁴⁰ demonstrate. In our opinion concentrations over 20,000 ng/mL are not recommended.

It is worth mentioning the possibilities of EGF eye drops in the treatment of dry eye syndrome, a very common, chronic and irritating problem for the patient, that in advance stages causes dramatic corneal damage. Although in most cases artificial tears are necessary and effective, the ideal eye drops have not yet been discovered. It has been demonstrated that very elemental dilutors like BSS are efficient and necessary to hydrate and regenerate the corneal epithelium.^56,57 Nevertheless, artificial tears are insufficient when histological changes persist, like squamous metaplasia, loss of globet cells, and so on, which can progress to more severe injuries (corneal erosion, ulceration, scarring, thinning, and even perforation). EGF eye drops are theoretically very effective in preventing and treating these injuries, though no clinical trials are available. Moreover, release of EGF at the surface of the dry eye and its tear film is related to an autocrine regulation via EGFR ligand.^9,52,58 Actually autologous serum is used by some clinicians in dry eye syndrome to prevent severe ocular surface disorders and persistent corneal epithelial defects, but there is some controversy regarding the efficacy and risk of hepatitis and human immunodeficiency virus infections.^59–61 In our opinion, the growth factors needed in cornea surface are the ones found in tears, not in the serum. The composition of the autologous serum is excessively ectopic for the cornea. In other words, the serum concentration of the growth factors and the autologous serum are not complementary to the needs of the corneal receptors, at least with EGF, whose physiological concentrations in human tears are almost 20 times higher than in the human autologous serum. On the other hand, important concentrations of other growth factors found in autologous serum have not been detected in tears, or have been detected in tiny concentrations.⁵² We are convinced that EGF eye drops must be a better and more specific treatment than the autologous serum eye drops, but clinical trials to prove it are missing.

In conclusion, we agree with the potential interest of EGF eye drops in the treatment of corneal and periocular-conjunctival disorders but we propose a change of focus for future clinical trials of EGF eye drops in regards to indications, posology, and duration of treatment. We are conscious that our proposal asks for a radical change regarding clinical trials, which are meta-analyzed in this article.

Footnotes

Acknowledgments

The authors thank Mrs. Maria Antònia Riera (Alcon Labs, Barcelona) for her kindness in providing the articles noted in the references, and Mrs. Isabel Lou Bonafonte for her kindness in revising the English text. The report given in this study is registered in the academic activities of the Centro de Oftalmología Bonafonte and IMC-Investiláser, year 2009–2010.

Author Disclosure Statement

No competing financial interests exist.

References

Tripathi

B.J.

, Raja

S.C.

, Tripathi

R.C.

Prospects for epidermal growth factor in the management of corneal disorders. Surv. Ophthalmol., 34:457–462. 1990.

Feldman

S.T.

The effect of epidermal growth factor on corneal wound healing: practical considerations for therapeutic use. Refract. Corneal. Surg., 7:232–239. 1991.

Pancholi

, Trullo

, Khaliq

, Foreman

, Boulton

The effects of growth factors and conditioned media on the proliferation of human corneal epithelial cells and keratocytes. Graefes Arch. Clin. Exp. Ophthalmol., 236:1–8. 1998.

Martínez-Carpio

P.A.

, Mur

, Rosel

, Navarro

M.A.

Constitutive and regulated secretion of epidermal growth factor and transforming growth factor-ß₁ in MDA-MB-231 cell line in 11-day cultures. Cell. Signal., 11:753–757. 1999.

Martínez-Carpio

P.A.

, Mur

, Rosel

, Navarro

M.A.

Secretion and dual regulation between epidermal growth factor and transforming growth factor-ß₁ in MDA-MB-231 cell line in 42-hour-long cultures. Cancer. Lett., 147:25–29. 1999.

Zhang

X.F.

, Liu

T.L.

, Sheng

W.H.

, Yang

J.C.

, Miao

J.C.

Impact of regenerated silk protein membrane of the cytokine expression of transfected human corneal epithelial cells. Zhonghua Yan Ke Za Zhi., 45:992–998. 2009.

Ohashi

, Motokura

, Kinoshita

et al. Presence of epidermal growth factor in human tears. Invest. Ophthalmol. Vis. Sci., 30:1879–1882. 1989.

van Setten

G.B.

, Viinika

, Tervo

et al. Epidermal growth factor is a constant component of normal human tear fluid. Graefes Arch. Clin. Exp. Ophthalmol., 227:184–187. 1989.

Rao

, Farley

W.J.

, Pflugfelder

S.C.

Association between high tear epidermal growth factor levels and corneal subepithelial fibrosis in dry eye conditions. Invest. Ophthalmol. Vis. Sci., 51:844–849. 2010.

10.

Pflugfelder

S.C.

, Jones

, Ji

, Afonso

, Monroy

Altered cytokine balance in the tear fluid and conjunctiva of patients with Sjögren's syndrome keratoconjunctivitis sicca. Curr. Eye. Res., 19:201–211. 1999.

11.

Sheardown

, Cheng

Y.L.

Tear EGF concentration following corneal epithelial wound healing creation. J. Ocular. Pharm. Ther., 12:239–243. 1996.

12.

, Bazan

H.E.

Epidermal growth factor synergism with TGF-beta1 via PI-3 kinase activity in corneal keratocyte differentiation. Invest. Ophthalmol. Vis. Sci., 49:2936–2945. 2008.

13.

Huo

, Qiu

W.Y.

, Pan

et al. Reactive oxygen species (ROS) are essential mediators in epidermal growth factor (EGF)-stimulated corneal epithelial cell proliferation, adhesion, migration and wound healing. Exp. Eye. Res., 89:876–886. 2009.

14.

Shao

, Hu

, Chen

A study on effect of bFGF, EGF and NGF on growth of cultured human corneal endothelial cells. Yan Ke Xue Bao., 24:9–12. 2008.

15.

Wang

, Wormstone

I.M.

, Reddan

J.R.

, Duncan

Growth factor receptor signalling in human lens cells: role of the calcium store. Exp. Eye. Res., 80:885–895. 2005.

16.

Hoppenreijs

V.P.

, Pels

, Vresen

G.F.

, Oosting

, Treffers

W.F.

Effects of human epidermal growth factor on endothelial wound healing of human corneas. Invest. Ophthalmol. Vis. Sci., 33:1946–1957. 1992.

17.

Carrier

, Deschambeault

, Talbot

et al. Characterization of wound reepithelization using a new human tissue-engineered corneal wound healing model. Invest. Ophthalmol. Vis. Sci., 49:1376–1385. 2008.

18.

Kopecky

E.A.

, Rootman

D.S.

Antimetabolite interactions with epidermal growth factor. Curr. Eye. Res., 15:247–253. 1996.

19.

Burling

, Seguin

M.A.

, Marsh

et al. Effect of topical administration of epidermal growth factor on healing of corneal epithelial defects in horses. Am. J. Vet. Res., 61:1150–1155. 2000.

20.

Szaflik

, Fryczkowski

A.W.

, Liberek

et al. Corneal wound healing after penetrating keratoplasty with EGF application. Experimental studies. Klin Oczna., 101:409–416. 1999.

21.

Gönül

, Kaplan

, Bilgihan

, Budak

M.T.

Effects of epidermal growth factor in artificial tear on vitamin C levels of corneal wounded eye tissues. Eye. (Lond)., 15:213–216. 2001.

22.

Gönül

, Koz

, Ersöz

, Kaplan

Effect of EGF on the corneal wound healing of alloxan diabetic mice. Exp. Eye. Res., 54:519–524. 1992.

23.

Gönül

, Erdogan

, Ozogul

et al. Effect of EGF dosage forms on alkali burned corneal wound healing of mice. Burns., 21:7–10. 1995.

24.

Dickersin

, Scherer

, Lefevre

Identifying relevant studies for systematic reviews. BMJ., 309:1286–1291. 1994.

25.

Soualmia

L.F.

, Dahamna

, Thirion

, Darmoni

S.J.

Strategies for health information retrieval. Stud. Health. Tecnol. Inform., 124:595–600. 2006.

26.

Martínez-Carpio

P.A.

, Bonafonte-Márquez

, Heredia-García

C.D.

, Bonafonte-Royo

Efficacy and safety of intravitreal injection of bevacizumab in the treatment of neovascular glaucoma: systematic review. Arch. Soc. Esp. Ophthalmol., 83:579–588. 2008.

27.

Mantel

, Haenszel

Statistical aspects of the analysis of data from retrospective studies of disease. J. Natl. Cancer. Inst., 22:719–748. 1959.

28.

Higgins

J.P.

, Thompson

S.G.

, Deeks

J.J.

, Altman

D.G.

Measuring inconsistency in meta-analyses. BMJ., 327:557–560. 2003.

29.

Sharar

A popperian perspective of the term “evidence based medicine.”

J. Eval. Clin. Pract., 3:109–116. 1997.

30.

Osborn

Observation, Sherlock Holmes, and evidence based medicine. Med. Secoli., 14:515–527. 2002.

31.

Daniele

, Frati

, Fiore

, Santoni

The effect of epidermal growth factor (EGF) on the corneal epithelium in humans. Albrecht Von Graefes Arch Klin Exp. Ophthalmol., 210:59–165. 1979.

32.

Kandarakis

A.S.

, Page

, Kaufman

H.E.

The effect of epidermal growth factor on epithelial healing after penetrating keratoplasty in human eyes. Am. J. Ophthalmol., 98:411–415. 1984.

33.

Reim

, Kehrer

, Lund

Clinical application of epidermal growth factor in patients with more severe eye burns. Ophthalmologica., 197:179–184. 1988.

34.

Caporossi

, Manetti

Epidermal growth factor in topical treatment following epikeratoplasty. Ophthalmologica., 205:121–124. 1992.

35.

Pastor

J.C.

, Calonge

Epidermal growth factor and corneal wound healing. A multicenter study. Cornea., 11:311–314. 1992.

36.

Ganem

, Mondon

, De Felice

G.P.

Local treatment by epidermal growth factor after epikeratoplasty. A double-blind clinical trial. J. Fr. Ophtalmol., 15:443–447. 1992.

37.

Scardovi

, De Felice

G.P.

, Gazzaniga

Epidermal growth factor in the topical treatment of traumatic corneal ulcers. Ophthalmologica., 206:119–124. 1993.

38.

Kammann

J.P.

, Vollenberg

Effect of corneal collagen shields and epithelial growth promoting factors on epithelization of the human cornea. Klin Monatsbl Augenheilkd., 205:201–209. 1994.

39.

Schüller

, Knorr

, Steuhl

K.P.

, Thiel

H.J.

Lacrimal secretion of human epidermal growth factor in perforating keratoplasty. Ger. J. Ophthalmol., 5:268–274. 1996.

40.

Dellaert

M.M.

, Casey

T.A.

, Wiffen

Influence of topical human epidermal growth factor on postkeratoplasty re-epithelisation. Br. J. Ophthalmol., 81:391–395. 1997.

41.

Cellini

, Baldi

, Caramazza

, De Felice

G.F.

, Gazzaniga

Epidermal growth factor in the topical treatment of herpetic corneal ulcers. Ophthalmologica., 208:37–40. 1994.

42.

Huang

B.R.

, Cai

L.W.

, Xiang

X.Z.

The basic and applied study on the epidermal growth factor. Zhongguo Yi Xue Ke Xue Yuan Xue Bao., 23:176–180. 2001.

43.

Hong

, Zhang

, Liu

, Zhu

Effect of recombinant epidermal growth factor on corneal epithelial cells after excision of pterygium. Zhongguo Yi Xue Ke Xue Yuan Xue Bao., 23:199–201. 2001.

44.

, Lin

Y.S.

, Guo

, Li

D.M

, Zhang

H.Y.

Effect of recombinant epidermal growth factor on ocular surface re-epithelization following amniotic membrane transplantation in patients with pterygium excision. Di Yi Jun Yi Da Xue Xue Bao., 22:437–438. 2002.

45.

Zhang

, Basti

, Jampol

L.M.

Acquired trichomegaly and symptomatic external ocular changes in patients receiving epidermal growth factor receptor inhibitors: case reports and review of literature. Cornea., 26:858–860. 2007.

46.

Förster

C.G.

, Cursiefen

, Kruse

F.E.

Topical application of EGF for the therapy of persisting corneal erosion under cetuximab treatment. Ophthalmologe., 105:269–273. 2008.

47.

Förster

C.G.

, Cursiefen

, Kruse

F.E.

Persisting corneal erosion under cetuximab (Erbitux) treatment (epidermal growth factor receptor antibody)

Cornea., 27:612–614. 2008.

48.

Johnson

K.S.

, Levin

, Chu

D.S.

Persistent corneal epithelial defect associated with erlotinib treatment. Cornea., 28:706–707. 2009.

49.

Yeh

, Fine

H.A.

, Smith

J.A.

Corneal verticillata after dual anti-epidermal growth factor receptor and anti-vascular endothelial growth factor receptor 2 therapy (vandetanib) for anaplastic astrocytoma. Cornea., 28:699–702. 2009.

50.

Egger

, Scheider

, Smith

G.D.

Meta-analysis Spurious precision?

Meta-analysis of observational studies. BMJ., 316:40–144. 1998.

51.

Baerlocher

M.O.

, O'Brien

, Newton

, Gautam

, Noble

Data integrity, reliability and fraud in medical research. Eur. J. Inter. Med., 21:40–45. 2010.

52.

Bonafonte-Márquez

, De Ortueta

, Bonafonte-Royo

, Martínez-Carpio

P.A.

Epidermal growth factor receptor (EGFR) in corneal damage: update and new insights from recent reports. Cutan. Ocul. Toxicol., 30:7–14. 2011.

53.

Lohmann

C.P.

, Hoffmann

, Reischl

Epidermal growth factor (EGF) in tears in excimer laser photorefractive keratectomy. Responsible for postoperative refraction and “haze”?

Ophthalmologe., 95:80–87. 1998.

54.

Tomas-Barberan

, Schultz

G.S.

, Tarnuzzer

R.W.

, Fagerholm

Messenger RNA levels for genes involved in extracellular matrix from human corneal scrapings before and after photorefractive keratectomy. Acta. Ophthalmol. Scand., 76:568–572. 1998.

55.

Lohmann

C.P.

, Winkler von Mohrenfelds

, Reischl

, Muller

, Guell

J.L.

Screening of myopic LASIK patients with increased epithelial wound healing. Ophthalmologe., 98:460–465. 2001.

56.

Bonafonte

, Aguirre Vila-Coro

Scanning electron microscopic study: the effect of balanced salt solution on rabbit corneal endothelium. Cornea., 4:195–201. 1986.

57.

Bonafonte

, Fernández-del-Cotero

J.N.

, Aguirre Vila-Coro

Mineral analysis in experimental corneal scars. An EDAX study. Cornea., 7:122–126. 1988.

58.

Dartt

D.A.

Interaction of EGF family growth factors and neurotransmitters in regulating lacrimal gland secretion. Exp. Eye. Res., 78:337–345. 2004.

59.

Tsubota

, Goto

, Fujita

et al. Treatment of dry eye by autologous serum application in Sjögren's syndrome. Br. J. Ophthalmol., 83:390–395. 1999.

60.

Rolando

, Geerling

, Dua

H.S.

, Benítez del Castillo

J.M.

, Creuzot-Garcher

Emerging treatment paradigms of ocular surface disease: proceedings of the Ocular Surface Workshop. Br. J. Ophthalmol., 94:1–9. 2010.

61.

Weisbach

, Dietrich

, Kruse

F.E.

, Eckstein

, Cursiefen

HIV and hepatitis B/C infections in patients donating blood use as autologous serum eye drops. Br. J. Ophthalmol., 91:1724–1725. 2007.