Document 6Rn1d54mMm67ZKrQpXa99rwbo

AR226-2904 cc; Haskell Laboratory Report N6. ApS-Sl October 6, 1981 EVALUATION OF OCULAR IRRITANCY POTENTIAL: SOURCES OF IMTRALABORATORY VARIABILITY AND EFFECT OF DOSAGE VOLUME 1 - Company SMHSzedL Does not contain TSCA CBf Haskell Laboratory Report N6. 685-81 October 6, 1981 EVALUATION OF OCULjftRIRRITAHCY POTENTIAI '. SOURCES OF INTRALABORATORY VARIABILITY AND EFFECT OF DOSAGE VOLUME Study Periforaed by: ' c^y<w^^fu^' tyrmS. Ford Technician ^ John ^. E. ^^/ Henry 1 Technician Beporc Prepared by: ^sapPn^.^Ula^ Research Toxicologist Study Director Apjprovedby; ^^ft,i ^y k^_^.i- \ Garald L. Kennedy, Jn. | Chief, Acute Toxicology SJW; ic.-WP;3.A : Stud) nictated/Completed: 2/7/81-4/8/81 J)ate I&sued; Octobq" 6, 1981. 2 - Company Sanitized. Does not contain TSCA CBI Haskell Laboratory Report No. 485-81 October 6, 1981 EVALUATION OF OCVU& IRRiIANCY POTENTIAL: SOURCES OF INTRALABORATORY VARIJABILITYAND EFFECT OF DOSAGE VOLUME ACKNOWLEDGEMENTS The author would like Co acknoujiLedge the excellent technical asslstence of John E. Henry and Lynn S. Ford vand thank G. Jay Graepel for consultation on the design of these studies and for statistical analysis of the data. U ^ - 3 - Company Sanitized. Does not contain TSCA CBI Haskell Laboratory Report No. 485-81 October 6, 1981 EVALUATION OF OCufAK IRRITANCY POTEW-'IAL; SOURCES OF IKTRALABORATORY VARIABILITY AND EFFECT OF DOSAGE VOLUME INTRODUCTION Current standard methods for assessing the ocular irritancy potential of chemicals have been the subject of criticism from a number of directions. The best documented criticism its a study (1) which demonstrated the high degree of intralaboratory and ,'lnterlaboratory variability of results from such tests using a 'standardized' methodology. This test has also been criticized based on the c.atis'tical invalidity of adding semi-quantitative scores ob^-iined from three separate ocular tissues. Moreover, it has been suggested that a dosage volume) of 0.1 oL, the standard liquid volume used, is extremely high relative to thej surface area of the rabbit eye. This fact, togeeher with known physiologic differences between rabbit and human eyes as f ' well as hwian experience with eye irritants, has led to the conclusion chat the current rabbit eye test maf be an overly-sensitive indicator of human risk (2). The most recent criticism of this test has been on humane grounds. v- j PURPOSE { The present investigation has been designed to study the following: 1. The sources of intralaboratory variability (reader, rabbit or test group) using standard (laskell testing procedures. 2. The dose-effect relationship using dose volumes of 0.10 mL and 0.01 aL of test cospounds. ' 3. The sensi'ivity and reproducibillty of the test using reduced dose volume. PROCEDURE : Seven compounds (Appendices A and B) 'previously tested using standard Haskell Laboratory procedures were chosen. These compounds were selected on the basis of being liquids seat neutral pH with observed, graded irritancy potential. Two were rated previously as severe irritants, 2 as moderate and 3 as oild. Five mL of liquid W transferred to a labelled foil-covered test tube and given a letter code (A-H). Compound G was the same as compound B. u Company Sanitized. Does not contain TSCA CB1 Thirteen rabbles were used each week for 6 week* (72 rabbice total). Ten microllters each of compounds A-H, excluding G, were applied directly to the corneal surface of the right eye of 1 of 7 male albii. rabbiCB. The left eye was uncreated and served as tine concrol. The rabbit was releead and the eye ^Hby was not washed or manipulated! in any way. Compound G has instilled in a similar manner into the right eye of 6 rabbits, but a volume of 0.10 oL was used. Compound G served as the internal control, and eyes dosed with it provided the data base for sajBesding sources of variability. Eyes were then .scored at I and 4 hours and at I, 2, 3, 7, 16, and 21 days (Haskell SOPpfkMj^ 2 Independent and expe^iencfcJ readers (J. E. Henry and L. S. PordT^^^J ^^oservationa of the cornea, iris and conjunctiva were made with an ophthalaoacope. Fluor-1-strip* staining was used at each examination after the day of treatment:. Scores; for conjunctiv'al, comeal and iritic changes were recorded separately. To characterize the sources of variability, each of the Indices (total corneal;'score or total conjunctival score) was analyzed independently using analyals of variance (3). Because of the magnitude of the variability change over time, each of the time periods was analyzed independently. RESULTS Sources of Variability C t o The 36 rabbits dosed with^compound G (0.1 ml.) served as internal controls elucidate the sources of variability in eye irrlcancy scoring. Total scores for conjunctiva and coifnea only are used in these determinations, since this compound did not have any Iritic effect. The analysis of variance data Is given in Appendices C and 0. The results demonstrate that rabbit to rabbit differences are the greatest source of variability at any given time. At no time are the test groups! a source of variability, suggesting chat test group results are consistent and therefore reproducible. The only significant reader variability was measured at 4 hours post dosing. This is apparent for both comeal and eonjunctival scores. This difference might be attributed to an artifact of the protocol. The study director was not aware that 1 reader consistently read before the other. At 4 hours post treatment the effects, in most cases, arts reaching a maximum and may be changing rapidly. Consequently, the difference observed is probably due to a time effect, not reade varlabilityi. Animal strain: New Zealand White Rabbles Supplier: B & H Babbitry, Bockvllle, HO 5 - Company Sanitized. Does not contain TSCA CBI Dose-effect Relationship Appendices E and F show the individual scores (cornea! and conjunctival) ^ ^ f H ^ for rabbite dosed with 0.01 am 0*10 oL of Che^saae compound. For Chis purpose, the first rabble dosec with'0.1 aL ofUBtfBBfO'^11 weekwaauaed for comparison with che rabbiti dosed each weetTwith 0.01 inL The comparison between dose grcups was made ueing the Mann-WhItneytetTT). Each rabbit was assigned the mean score for the 2 readers ac each time. The endpolnts used for comparison were maximum score and cine to recovery for the cornea! and conjunctiva! index.; Both ^ndpoints for each index were significantly different at the jp< 0.05 level. Ley-volume Eye Teat The median scores for corneal and conjunctival effects are plotted against time in Appendices G and H. By inspection, it is apparent that there is a graded response to this series of chemicals. The irrlcancy ratings If which appear for each curve are those assigned to the respective compound when each was studied previously at a dose of 0.10 aL. A more rigorous analysis of these effeccs ca: be made the time to recovery and maximum score are chosen as endpolnts for comparison. By calculating the rank correlation at each end point bjetween these scores and the qualitative scores obtained from previous tests, aj quantitative measure of correlation between the 2 dose volumes can be derived. The correlation coefficients between the 2 dose ^levels In the cornea foe time to recovery and maximum scores are 0*80 and 0.78,. respectively. The correlation coefficients are 0.78 and 0.80 in the conjunctiva. All of these correlations are aigaificanc at the p < .05 level. Finally, since a clear ^one-affect rclatloirhip h&& been demonstrated. It Is possible thjat In some cases, mildly irricacing compounds which would be detectable at a dose of 0.10 mL may be missed at the low volume. However, all 3 mild compounds tested had measurable corneal and conjunctival effects. : SUMMARY j Reader bias aad test group to test group differences do not contribute significantly Co the variability in scores obtained using a seml-quantitiative e> ale for assessing ocular irrifancy. Moat of the observed differences can be attributed to rabbit variability. However, the eye test is used routinely at H|askell only for effect/no effecC determination ' of ocular Irrifancy potential. Therefore, the number of rabbits used for an eye test study should QOC be Increased from 2 to 6 or more in order to Increase the sensltl?'ty of the test. When tested at a dosage vol>mc of 0.01 or 0.10 asL, a dose-effect relationship was discernible fo^jjmmj^using either time to recovery or maximal effect as endpoints. 6 - Company Sanitized. Does not contain TSCA CB1 I j^BHBHiiR^f When tested coiapoundi which at 0.01 aL, t ie saae graded response was obtained for However, based have been tea :d previoucly gc on che doac-efect relationship 0.10 aL of pure material. based on daoaging previous Haakell -lati, all of Cheae demonstrated in this study and coapounds are consistently less reducing when applied ac a vo uu of 0.01 the voluae Applied t> the eye oL. In light of theae reaulcs, observed with most coopounda ind any be aappmaorreentrley amliosdteicrataepsprthoaechrestoponses assessing risk of eye irritation in huaans. i a dose voluae^ of 0.01 aL or ijts weight It is therefore recommended chat: cheaicfils f^r eye equivalent be used in routine irritancy potential at Hsskell testing tboratory. ^ : 7 - Company Sanitized. Does not contain TSCA CBI APPENDIX A - TEST COMPOUNDS Letter Code Chemicals 6 Jackson Lab 0 v.' APPENDIX B CHARACTERISTICS OF TEST COMPOUNDS Composition: ^ - 9 Company Sanitized. Does not contain TSCA CB1 ApALYSIS 01' VARIANCE CONpROL CORNI!AL SCORES Tine Post-Dosing 1 Hr. 4 Hr. 1 Day 2 Day 3 Day 7 Day 14 Day 21 Day Reader NS0 0.25 NS NS NS NS NS NS \ Parlance Group Rabbit NS 0.91 NS 0.25 NS 0.33 NS 0.30 NS i ss 3.03 1.71 NS 1.73 NS 1.75 Residual 1.10 0.67 0.54 0.86 0.62 0.93 0.38 0.05 a) each rabbles. value Is based upon tl^e total corneal score obtained 36 b) group to group differences 'were from rabbles/week for 6 weeks assessed on the basis of 6 c) Not significant : 10 Company Sanitized. Does not contain TSCA CBI Time Post-Dosing 1 Hr. 4 Hr. 1 Day 2 Day 3 Day 7 Day 14 Day 21 Day Reader NS' 0.99 NS NS NS NS NS NS APPENDIX D AIMILYSXS OF VARIANCE CONTROL - CONJUNCT'IVAL SCORESa Varl ance Group NS NS NS ; NS NS . cS '; NS ' NS :' Rabbits 1.60 0.75 1.43 3.00 3.01 1.45 0.64 0.04 Residual 0.95 0.53 0.73 0.62 0.48 0.39 0.19 0.05 2 Of Variance Due To Rabbits 61.1 33.0 66.2 82.9 86.0 78.8 77.1 30.8 a) 36 each value rabbits Is baaed upon the total con-luctival score obtained b) group to group differences were assessed on the basis of 6 rabbits/week for 6 weeks c) not significant from 11 - Company Sanitized. Does not contain TSCA CBI APPENDIX E Cornea! Scores - Dose-Effect Relationship Hours Time Post-Dosmg Days 12 - Company Sanitized. Does not contain TSCA CBI APPENDIX F Conjunctival Scores - Dose-Effect Relationship Tlnr Post-DosIno .L. 13 Company SanHized. Does not contain TSCA CBI 0 0 C a SCORE bJLbob-.Mb^b.fb.UblObl ^S ^ 5H; ob ,L,,,,,,,,,,i,..,,,..,,,..-----------''--' r- o.^ <* xi Hs ob m2: o > ui b o 0 ^3 S m > sg 0 C/) s o0 0) ;D m ui b (/) o s p b M Ul b 0  ^ 00 . II I SCORE a +1> i u 53 u b i go go yo-o^ ^o^bybc b n ^ b o x i i;Eg p o ui b | ui s m c- ^ <- 0 i? f5 LO CO <- ui b C3 S 00 M o b t> is Ul b APPENDIX I Literature Cited , (1) Well, C. S. and Scala, R. A' (1971), "Study of Intra-and Interlaboratory Varlabllllty in the Results of Rabbit Eye and Skin Irritation Tests". ToxleoL Appl.^haraiacol^, 19:276-360. (2^ Griffith, J. F., Nixon, G.|A., Sruce. R. D, Reer, P. J., Bannan, E. A., (1980). "Dose-Response Studies with Chemical Irritants In the Albino-Rabbit Eye as a Basis for Selecting Optimum Testing Conditions for Predicting Hazard to the Hmoan Eye". Toxleol. Appl. Pharaacol., 55:501-513. j ---------------- (3) Snedeeor, G. W. and Cochrai^, W. G., (1980) Iowa State University Pres4, Amefl, Iowa. Statistical Methods, The 16 - Company Sanitized. Does not contain TSCA CB1