Document wqBe9NEbdL03Vw8R02j3mJ6M6

A * A 6 _ /7/ PERSONAL AND CONFIDENTIAL November 24, 1980 TO: FROM: R. J. B U R G E R C. R. C A M P B E L L D. A. E R D M A N PAUL THISTLETON COMMUNICATING RESULTS OF BLOOD'ANALYSES Details of recent blood analyses are given in m y 11/ 1 9 / 8 0 l e t t e r to R. J. B u r g e r (copy attac h e d ) . R e s u l t s of August 1979 samples have been multiplied by 1.25 in Table I for comparison with August 1980 results obtained by the C-8/GC method (compare lined columns). We believe that this adjustment is required for proper comparison. People sampled in 1979 were given results in . Columns 1 and 2 using standard medical cards. We plan to report the C-8/GC results in column 6 in the same way. Some explanation of the increased values resulting from the C-8/GC analyses will be required. It appears that this can best be handled by Dr. Power on an individual basis. He may use the attached statement for background but it will not be distributed. ,> When the results in column 2 are multiplied by 1 . 2 5 (see c o l u m n 3) t h e r e is g e n e r a l l y g o o d a g r e e m e n t w i t h the recent C-8/GC results (column 6). Perhaps there is a s i g n i f i c a n t i n c r e a s e for No. 16 (a fine p o w d e r d r y e r o p e r a t o r ) . The value reported for No. 17 in 1979 was recognized to be unusually low and may have been inaccurate. Three of the FEP p e o p l e (Nos. 19, 23 an d 24) s h o w lit t l e chan g e b e t w e e n the A u g u s t 1979 (column 2) a n d A u g u s t 1 980 C - 8 / G C r e s u l t s (column 6). Attachments PT/nsw AJP001326 EID080726 000234 ! CC: R. J. B u r g e r J . F . Doughty T. L. Schrenk N o v e m b e r 19, 1980 TO: FROM: DR. Y. L. P O W E R PAUL THISTLETON RESULTS OF BLOOD ANALYSES The blood sampling program proposed in my 8/1/80 letter is essentially complete. All samples have been analyzed by the C-8/GC method which was recommended for routine u s e in m y 1 1 / 1 2 / 8 0 l e t t e r t o R. J. Burg e r . M o s t of t h e samples have been analyzed by the Torch method at ESL and some have been measured by the Torch method at Jackson Laboratory (JL). Samples were sent to JL because of delays in demonstrating satisfactory Torch performance at ESL. . R e s u l t s a r e g i v e n in T a b l e I. It i n c l u d e s the 1979 data which was reported to the people sampled. Names of people sampled in 1980 and identification numbers used in T a b l e I ar e g i v e n in the e n c l o s e d l i s t (Dr. Y. L. Power, o n l y ) . One person is omitted from Table I because results were variable and a resample is being requested. The ESL and JL Torch results agree very well. The C-8/GC results are about 125% o f the To r c h results (see Figure of my 11/12/80 letter). The difference may result from incomplete recovery in the Torch meth o d and this is being checked at ESL and JL. 1 My 11/12/80 letter recommended that only C-8/GC results should be reported to employees. We believe that they are the best available measurements of organic fluorine in blood samples. The August, 1979, results given in Table I have been multiplied by 1.25 which is suggested.as the basis for comparing these results w i t h c u r rent C -8/GC results (see my 11 / 1 7 / 8 0 l e t t e r to R. J. B u r g e r , c o p y a t t a c h e d ) . In m o s t c a s e s the numbers are very similar and it is doubtful if any of the apparent changes are statistically significant. This can be established when the total sampling program is completed and more 1979/1980 comparisons are possible. AJP001327 EID080727 000235 1 DR. Y. L. P O W E R 2 N O V E M B E R 19, 1980 In the meantime I conclude that there has been no significant decrease in organic fluorine in blood samples between August, 1979 and August, 1980. This may be because many of our corrective measures were functioning for only a small part of the year. Our 1980 data using the C-8/GC method, which is specific, should provide a good basis for comparing data to be obtained in 1981. Attachment PT/nsw AJP001328 EID080728 000236 II TABLE I COMPARISON OF BLOOD ANALYSES 1979 SAMPLES BOMB ANALYSIS - JL&. IDENTIFICATION NUMBER ppm Organic Fluorine AUGUST JUNE AUGUST X 1.25(2) -1- -2- -3- No Direct Exposure 1` 2 -- ---- Professionals 3 4 5 6 -- - 0.45 -- - 0.56 Monomer Operators 7 8 (3) 9 (4) 0.39 5.3 6.7 0.49 6.6 8.4 Fine Powder Dispersion Zone 6 10 22.2 21.2 26.5 12 10.6 8.7 10.9 13 15.0 13.8 17.3 14 20.8 26.0 15 1.8 2.3 Zone 4 16 1.8 2.3 AUGUST 1980 SAMPLES TORCH ANALYSIS - C-8/GC ppm Organic Fluorine ESL JL ESL -4- -5- -6- 0.24 -- 0.022 0.015 0.03 0.44 - 0.3 0.22 0.40 0.19 0.52 0.8 5.2 6.5 6.7 0.78 6.4 ' 8.2 20.3 9.7 16.5 3.3 21.0 22.9 3.8 24.0 13.0 21.0 29.0 4.6 4.6 4.6 5.6 Granular Zone 6 17 0.47 0.59 1.4 1.7 1.9 AJP001329 EID080729 000237 IDENTIFICATION NUMBER FEP Polymerization 18 19 ' 20 21 22 23 24 25 26 FEP Service 27 Research Semiworks 28 2 1979 .SAMPLES BOMB ANALYSIS - ppm Organic Fluorine AUGUST JUNE AUGUST X 1.25(2) 2- 1 - -- -3- AUGUST 1980 SAMPLES TORCH ANALYSIS - C-8/GC ppm Organic Fluorine ESL -4-- JL -5- ESL 6- - 1.36 3.61 0.99 3.7 1.99 4.96 4.14 4.52 2.71 4.64 0.91 1.2 4.6 2.5 6.2 5.2: 5.7: 3.4 5.8 l.l: 1.1 3.4 2.9 5.6 3.7 4.1 2.1 6.4 0.9, 1.1 2.9 0.87 1.5 4.0 3.7 6.6 5.5 4.9 2.9 7.8 1.2 1.1 0.72 0.5 0.32 0.26 (1) JL = Jackson Laboratory (2) August, 1979 borrib results increased by 25%. This is the factor recommended to allow comparison of 1979 and 1980 results. ESL C-8/GC results are about 125% of ESL and JL Torch results for the August, 1980 saitples. Equivalence of Torch and Bomb results was demonstrated in a study reported by Erik Kissa, Jackson Laboratory, 6/13/80. (3) Monomer Operator 21 months, 16 years Polymerization Service. (4) Monomer Operator 32 months, 15 years Polymerization Service. AJP001330 000238 FT 11/19/80 E1D080730 N o v e m b e r 24, 1980 RESULTS OF BLOOD ANALYSES A sample of your blood was taken in August for a test program to compare two analytical methods for measuring organic fluorine in blood. The Torch method burns the blood in a special torch and the combustion products are scrubbed and analyzed for fluorine. It measures organic fluorine plus inorganic fluorocompounds that burn in the torch. The C - 8 / G C m e t h o d m e a s u r e s t h e C - 8 b y gas c h r o m a t o g r a p h y (GC) which separates the C-8 from other fluorocompounds. We believe that C-8/GC results are the best measurements of organic fluorine in blood samples. We plan to use the C-8/GC method for analyzing blood samples because it measures C-8 and is less subject to interference than the Torch method. We are reporting the C-8/GC measurement for your blood sample expressed as ppm organic fluorine. This method gives results about 25% higher than the method used for the 1979 samples. The difference may result from incomplete recovery of organic fluorine in the 1979 analyses. If you have questions please contact Medical Division. Y. L. POWER, M.D. 000239 EID080731 mm ' tsut:i}nsi E. I. d u P o n t d e N e m o u r s & C o m p a n y INCeftPOKATCD W i l m i n g t o n , D e l a w a r e 19898 POLYMER PRODUCTS DEPARTMENT EXPERIMENTAL STATION W. C. Percival S. S. Stafford R. R. Twelves PRAL File 1C 353 269 353 January 23, 1981 TO: PAUL THISTLETON - PPD, Washington Works FROM: l '. J. PAPA TOr-TUg COMMENTS ON VALIDITY OF 1979 FLUORINE IN BLOOD RESULTS I have reviewed the 1979 blood fluorine results from Jackson Laboratory (JL) and investigated the entire situation to comment on the validity of the results. The situation is not simple because 1. The values reported were raw data and were not corrected for recovery. The data were obtained by the modified 3M bomb method (private com munication with E. Kissa). 2. From data appearing in Kissa*s report (CP-JL-80-14, p. 10) issued Sept. 1980, his recovery at that tine appears to be ^94%. 3. 3M published a recovery of 92+5%. Belisle and Hagen, Anal. Bio. * 87 545-555 (1978). 4. The bomb data correlated 1:1 with the torch data in Kissa's early work (CP-JL-80-14, p. 19-21B) - hence torch recoveries must have also been ^94%. 5. A reagent deteriorated in the modified bomb method causing the 8/79 values to be low by a factor of ^1.18 (Memos, E. Kissa to G. H. Patterson dated 10/30/79 and 11/29/79). This was not dicovered and communicated to WW until after the results were given to our employees. 6. A recent study by Kissa shows the torch method, which is allegedly equivalent to the bomb method, gives 83% recovery and is 80% of the value by GC (Ref: my recent letter to you dated 1/23/81). We are left with these facts A.- All of the results from August 1979 sampling should have been corrected by a factor of 1.18 to compensate for the deteriorated reagent. AJP001317 There's a world of things we re coing something about EID080717 000240 - 2- B. An additional correction is necessary to compensate for recovery. That factor is 1.06 if the 94% recovery of Kissa's early work is correct or 1.20 if the later 83% recovery is correct. I have no way of judging which recovery number is correct. If fact, they both may be correct. He could have started with 94% recovery and drifted to 83% recovery. However, the numbers do allow us to set up boundaries. With the bad reagent and a 94% recovery, the correction factor is 1.25. With a bad reagent and an 83% recovery, the correction factor is 1.41. Results obtained from JL other than the August to October 1979 period do not suffer from the bad reagent contribution and the 1.18 correction factor is not applicable. However, they must still be corrected for recovery. The correction is 1.06 if you believe the 94% recovery, 1.09 if you believe the 92% recovery or 1.20 if you believe the 83% recovery. I suggest you use the following set of corrections for any data you have in hand: Multiply by 1.09 for all data prior to August 1, 1979 - this uses 3M's recovery of 92% and was suggested by E. Kissa. %^ ^ For bomb data in the period of August 1, 1979 to October 30, 1979 use a factor of 1.28 - this assumes 92% recovery and corrects for the bad reagent. For torch data in this period use a factor of 1.09. __For period November 1, 1979 to early 1980 (1st quarter), assume the recovery was 92% and use a factor of 1.09. The rat dermal study blood analyses were performed in this time period. From early 1980 on, GC values or torch values corrected for 83% recovery are used so you have no corrections to make. I hope this letter helps to end the confusion and does not create more. If you have questions please contact me. -- i-i \-2j O-fi-cyw. c7 . -UXL. \LKsy>*~- AJP001318 E1D080718 000241 <WM PnjrCMi t > irs$t*wUui!hmtrr -tf:e:: E. I. du P o n t de N e m o u r s S C o m p a n y iMCOftPOftATce W il m in g t o n . D elaw are 19898 PCLYV.ER PRODUCTS DEPARTMENT- EXPERIMENTAL STATION S. S. Stafford - 269 R. R. Twelves - 353 PRAL File I.C. January 23, 1981 TO: PAUL THISTLETON - PPD, Washington Works FROM: L. J. PAPA SPECIFIC DETERMINATION OF PERFLUOROOCTANOIC ACID IN BLOOD BY. GAS CHROMATOGRAPHY AND COMPARISON TO TORCH METHOD S. S. Stafford has completed development and study of a gas chromatographic method to specifically determine perfluorooctanoic acid or its salts (including FC 143), hereby defined as Cg, in human or rat blood. The method is sensitive to 7 ppb fluorine and has a precision of MJ.0% throughout most of the concentration range of 7 ppb to 100 ppm although the precision falls off at the lower ppb range. The method gives comparable results to the Modified Wickbold Torch used by E. Kissa at Jackson Laboratory and duplicated at ESL by R. R. Twelves. The principle differences are the GC method is specific, easier to use, faster, cheaper and much more sensitive. You received a copy of this nethod on your last visit to our laboratory on January 8, 1981. DISCUSSION We compared the Cg-specific GC method to torch methods at ESL (Twelves) and at JL (Kissa) by analyzing 26 human blood samples obtained from Washington Works personnel. This allowed a simultaneous comparison of the two torch methods at ESL and JL. The data is listed in Table I and plotted in Figure 1. A least squares examination of this data (line shown in Figure 1) shows the two torch methods give comparable data and are 79% of the GC numbers. The GC numbers are corrected for recovery but a true recovery study had never ben performed on the torch method. I asked Kissa (JL) to perform a recovery study on this torch method. He later reported (by telephone) that he performed a 5 concentration calibration curve study in aqueous solution from 0.5 to 12.0 ppm fluorine. The slope of his line, or recovery, was 83%. He then spiked two blood samples with 10 ppm C8 and obtained recoveries of 80 and 84%. I conclude from these data that his recovery is 83%. R. R. Twelves has never performed such a study but indications are that he has a similar recovery. Table II lists the GC data again and the JL torch values corrected for 83% recovery - the agreement is now very good. AJP001319 There's a world of things we're coing something about EID080719 000242 -2S. S. Stafford later analyzed 7 rat blood samples from a Haskell Laboratory C8 inhalation study that were also analyzed at Jackspn Laboratory by the Torch Method (Kiss'a). These data are listed in Table III and again show good agreement. These data show that the discussed methods can and did give equivalent results on real blood samples when all are calibrated to compensate for recovery. It should be remembered that Interferences may be encountered in the future which could give erroneous answers by either method. This seems less likely with the Cg-specific GC method. For this reason as well as those mentioned in the first paragraph I think our decision to use the Cg-specific GC method is well founded. Attachments fmt EID080720 000243 AJP001320 TABLE I Comparison of Ca-Specific GC Method to Torch Methods' at ESL and Jackson Lab (JL) PRAL No. 80-63841 80-63838 80-63837 80-62921 80-63834 80-63835 80-62916 80-62915 80-63839 80-62912 80-62920 80-62919 80-63842 80-63843 80-63836 80-62922 80-62918 80-63844 80-62910 80-62913 80-62911 80-62917 C.4-33^ 80- i , 80-63845 80-62914 80-63846 _____ppa Fluorine ESL JLTorch_____ 1.1 -- 20.3 21.0 1.4 1.7 0.5 0.3 0.03 -- -- 0.3 2.1 2.6 ND 0.2 6.5 6.7 0.8 -- 1.0 0.87 6.4 6.4 3.4 -- 2.9 2.9 0.44 -- 3.3 3.8 22.4 22.9 5.6 -- 9.7 10.3 16.5 14.8 4.1 -- 4.6 4.6 0.24 -- 1.1 -- 5.2 5.0 3.7 Cr/GC 1.5 24. 1.9 0.26 0.22 0.52 2.9 0.015 8.2 0.78 1.2 7.8 4.0 3.7 0.40 4.6 29. 6.6 13. 21. 4.9 5.6 0.022 0.72 6.4 5.5 EID080721 000244 AJP001321 TABLE II Comparison of Cg-Specific GC Method to JL Torch - .Corrected for 83% Recovery PRAL No. 80-63841 80-63838 80-63837 80-62921 80-63834 80-63835 80-62916 80-62915 80-63839 80-62912 80-62920 80-62919 80-63842 80-63843 80-63836 80-62922 80-62918 80-63844 80-62910 80-62913 80-62911 80-62917 04-33O-] 80-63863r 80-63840 * 80-63845 80-62914 80-63846 ppm Fluorine JL Torch Cft/GC -- 1.5 25.3 24. 2.0 1.9 0.4 -- 0.26 0.22 0.4 0.52 3.1 2.9 0.2 0.015 8.1 8.2 -- 0.78 1.0 1.2 7.7 7.8 -- 4.0 3.5 -- 4.6 28. -- 3.7 0.40 4.6 29. 6.6 12.4 17.8 -- 5.5 -- -- -- 13. 21. 4.9 5.6 0.022 -- 0.72 6.0 6.4 -- 5.5 EID080722 000245 AJP001322 TABLE III Comparison of Cs-Specific GC Method to Torch Method (JL) on Rat Blood PRAL No. 80-67461 80-67462 80-67752 80-67756 80-67481 80-67460 80-67484 (Blank) Raw Data 6.9 10.5 9.5 9.0 1.8 1.1 0.76 ______ ppm Fluorine Torch Jackson Lab Corrected for 83% Recover 8.3 12.7 11.4 10.8 2.2 1.3 0.92 Cr /GC 7.7 13.1 11.4 8.6 1.3 1.1 <0.007 AJP001323 EID080723 000246 EID080724 O ^eioodfv G C /C 8 29. 0 .EiA OU lN/V l-i X U b U i. * n x i TO: FROM: DR. Y. L. P O W E R iPAUL THISTL E T O N CC: E. D. C h a m p n e y , J r . - Wilm. A. J. Dahl - ESL D. K. D u n c a n - Wilm. L. J. Papa - ESL J. W. Raines - Wilm. R. J . B u r g e r ' C. R. C ampbell J. F. D o u ghty * D. A. Erdman ' L. W. Go i n J. G. L o s c h i a v o T. L. Schrenk R. N. T a y l o r February 1 7 r 1981 CORRECTION OF ORGANIC FLUORINE IN BLOOD RESULTS FOR RECOVERY The recent comparison of Torch and C-8/GC results for 27 blood samples made at ESL indicated that C-8/GC results were about 25% above uncorrected Torch results. This led to a r e v i e w o f 1979 r e s u l t s (8 samples in J u n e a n d 78 samples in August, 1979) w h i c h is r e p o r t e d in L. J. Papa's January 23, 1980, letter "Comments on V a l i d i t y of 1979 Fluorine in Blood Results", (copy attached). Following Papa's recommendations the results obtained by Bomb and Torch methods have been corrected for recovery as indicated below. Bomb data obtained for August, 1979 samples require an additional correction factor of 1.18 to compensate for deterioration of the reagent (#5 in Papa's letter), c o m b i n i n g this w i t h 92% r e c o v e r y g ives correction factor of 1.28. a Sampling Period Analytical Method Correction Factor June, 1979 August, 1979 December, 1979 August, 1980 Bomb Bomb Bomb Torch 1.09 1.28 1.09 1.20 EID079382 000248 AJP002519 DR. Y. L. POWER 2 FEBRUARY 17, 1981 R e s u l t s are g i v e n in T a b l e I. O n l y th e 1979 samples that were repeated in 1980 are included. There is good agreement between Torch results {Columns 5 a n d 6) a n d 0 8 / G C r e s u l t s {Column 7) w h i c h w e r e obtained on the same samples. There is no obvious trend of results with time. A better comparison, including statistical analysis, will be possible whe n the rest of the 78 people sampled in 1979 have been tested. We expect that their blood will be analyzed only by the C-8/GC method. Their uncorrected 1979 results which were reported to them should be increased by 28% to allow comparison with C-8/GC results. Attachment PT/nsw EID079383 000249 TABLE I COMPARISON OF BLOOD ANALYSES IDENTIFICATION NUMBER No Direct Exposure 2 \S Professionals 3/ 4/ 65 */ Monomer Operators 1979 SAMPLES m BOMB ANALYSIS - J L U ' ppm Organic Fluorine JUNE AUGUST JUNE X 3X09 AUGUST X 1.28 -1- -2- -3- 0 4 9v Fine Powder Dispersion 1 Zone 6 ib/ 111243VV/, 15 / 22,2 10.6 ^.5.0 Zone 4 16 V AUGUST 1980 SAMPLES TORCH ANALYSIS C-8/GC ppm Organic Fluorine esl(2) JL ESL -5- -6- 0 0.29 0.2 00..0543 0.4 167...280 68..01 00..001252^^ 0 .2i 2 ^ 0-.1490^'--' S3 1? 0.78 ' 24.4 11.6 19.8 4.0 5.5 25.3 12.4 17.8 28.0 4.6 5.5 24. 0_, 13.0__ 21.0^ 29.0- 4.6_~- ;Vv# 5:6_ Granular Zone 6 i7 y 1.7 2.0 1.9 BEST COPY AVAILABLE EID079384 0002S0 AJP002521 IDENTIFICATION NUMBER FEP Polymerization 18/, 12 90 // 21/ 22/ 23/ 24 V 25 / 26/ - 2- 1979 SAMPLES m BOMB ANALYSIS - JL U ' ppm Organic Fluorine JUNE AUGUST JUNE X 1.09 AUGUST X 1.28 - 1- - 2- -3- -4- 1.36 3.61 1.5 3.9 0.99 3.7 1.99 4.96 4.14 4.52 2.71 4.64 0.91 1?$ 4 7 2.5 6.3 5.3 5.8 3.5 5.9 1.2 AUGUST 1980 SAMPLES TORCH ANALYSIS C-8/GC ppm Organic Fluorine ES L (2) JL ESL -5- -6- -7- 1.3 1 ? $ ^ 4.1 4.V*' 3.5 3.5 3.f> 6.7 6 . > 4.4- ;5,%x 4.9 ' % % s f 2.5 3.1 7.7 7.7 7 . X 1.2 1.0 1.2 FEP Service 27/ --7Research Semivrorks 1.1 0.72 0.5 0.4 0.26 (1) JL = Jackson Laboratory (2) ESL = Experimental Station Laboratory Notes on Columns -1-2-3- -4-5- -6-7- Part of first WW sampling (total of 8 sanples). Data in this column is enclosed in a block to indicate that it should not be compared with corrected data. Column -1- X 1.09 correction factor. Part of second W W sampling (total of 78 samples, including resample of original 8 samples). Data in this column is enclosed in a block to indicate that it should not be compared with corrected data. Column -3- X 1.28 correction factor. August 1980 sampling for comparison of analytical methods (total of 28 samples)* analyzed by Torch method at ESL. Results corrected for 83% recovery demonstrated for Torch method by E. Kissa, Jackson Laboratory. August 1980 sampling for conparison of analytical methods analyzed by Torch method at Jackson Laboratory. Results corrected for 83% recovery. August 1980 sampling for oonparisan of analytical methods analyzed by C-8/GC method at ESL. 100% recovery assumed. Correction factors and recovery are taken from L. J. Papa's letter "Comments on Validity of 1979 Fluorine in Blood Results", 1/23/81. * Sample No. 11 is not reported because unusually large variation in results was found. Analysis will be made on a new sample. . EID079385 2/17/81 000251 AJP002522 Z CC; E. D. Champney, Jr. - Wii A. J. D a h l - ESL D. K. D u n c a n - Wilro L. J . Papa -- ESL J. W. Raines - Wilm. R . J . Burger C, R. Campbell ; ,J. R. D o u g h t y j? D. A. Erdman L. W. Goin J. G. Loschiavo T. L. Schrenk R. N. Taylor M a r c h 6, 1981 TO: FROM: DR. Y. L. POWER PAUL THISTLETON 4 CORRECTION OF ORGANIC FLUORINE IN BLOOD RESULTS FOR RECOVERY Ref: Letter Thistleton to Power, same topic, dated 2/17/81. In the above referenced letter an error was made on page two of the attachment. Please let this letter serve to correct this error. Attachment - Page Two - Notes on Columns - -7- Nofte -7- s h o u l d r e a d as follows: -7- August 1980 sampling for comparison of analytical methods analyzed by C-8/GC method at ESL. Correction for recovery is included in the method with an internal standard? calculation giving 100 + or - 5% (relative standard deviation) for spiked samples. NOTE: New or corrected information has been underlined for clarity. PT/nsw rrczoodfv BID0793S6 I 00025a