Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
A DEMONSTRATION OF THE USE OF THE CBCC SYSTEM IN THE FIELD OF ANTITHYROID COMPOUNDS George W. Anderson Chemotherapy Division, Stamford Research Laboratories American Cyanamid Company, Stamford, Connecticut
156 When this demonstration was planned, it was hoped that most of the antithyroid data in the literature would be recorded on CBCC punch cards and available for use before the Symposium. Thus, it would be possible to make correlations of chemical structure with anti- thyroid activity, utilizing the same basic information which was recently used in a review article.^ Advantages and disadvantages of the CBCC system should become apparent. However, revision of the detailed CBCC biological code caused sufficient delay that not all of the anti- thyroid literature could be abstracted and the data coded, punched and sorted for a complete â¢ tudy. The results to be described here were nevertheless sufficient to indicate that the CBCC system is workable, and can be used conveniently to answer specific questions in the antithyroid field. This demonstration is based on fifteen questions which were designed to disclose the most active antithyroid compounds, and to bring out correlations of chemical structure with anti - thyroid activity. A continuation of the process would, of course, yield more information, and other approaches could be used to get the same results. Before proceeding with a discussion of the machine operations used to obtain data from the punched cards, a few general remarks about the nature of the information there are in order. Antithyroid compounds, which may be defined as those compounds which depress the secretion of the thyroid hormone, have largely been developed since 1942. Something like eight hundred compounds have been tested and reported in the literature. Most of the testing has been done in rats. Generally, the compound was fed in the diet or drinking water to young animals for ten days to two weeks and the effect on the thyroid gland was then determined. An active compound will cause cellular hyperplasia, an increase in weight of the gland, and a decrease in iodine content. All of this arises because of an interference with the synthesis of the thyroid hormone. Organically bound iodine is not stored, and the gland increases in size in a compensatory action caused by the thyrotropic hormone from the pituitary. The antithyroid effect was measured in the early work by the hyperplasia, and in the later, more precise investigations, by the effect on thyroid weight and iodine content. The latest method depends upon the rate of uptake of radioactive iodine by the thyroid gland. This method has been of particular value in determining relative activities in man and monkey. Antithyroid activities relative to a standard, which is usually 2-thiouracil. have been reported in the literature for tests in rats, and also for the limited tests in man and monkey by the radioactive iodine method. Also, where relative activities have not been given, they can be estimated from dose-response data. The biological code of the CBCC is suited to this, since doses and responses are recorded as falling in ranges. Thus, most of the animal evaluation of antithyroid activities of chemical compounds has been done in rats. This information is either summarized in the form of the author's evaluation relative to 2-thiouracil as a standard, or can be put in this form by comparing dose-response data. Finally, these relative activities in rats can be compared to those reported for man as a test animal. With this background, the following questions were devised: (Group I. Designed to select the most active compounds in rat tests, and see which have been tested in man. ) 1. How many compounds tested in rats have an activity of 0. 05 or greater? (Thiouracil = 1) i/ Medicinal Chemistry, Vol. I, pp. 1-150, John Wiley and Sons, Inc. , N. Y. , 1951.
157 2. Which compounds tested in rats have an activity of 0. 5 or greater? (Thiouracil = 1) 3. How many compounds showed a +2 hyperplasia of the thyroid gland when administered orally to rats for a week or more in daily doses of 125 mg. /kg. /day or less ? (Should give compounds 0. 2 as active as thiouracil or greater, based on Astwood's (1943) results. ) 4. Which of the compounds in Question 3 showed a +2 hyperplasia or greater when tested at a dose of 25 mg. /kg. /day or less ? (Should give compounds as active as thiouracil or greater. ) 5. Which of the compounds in Questions 1 and 3 have been tested in man? 6. How many of the compounds in Questions 1 and 3 have not been tested in man ? 7. Which of the compounds in Question 5 have an activity of 0. 05 or greater in man? (Thiouracil = 1) 8. Which of the compounds in Question 5 have an activity of <0. 05 in man? 9. Which of the compounds in Questions 2 and 4 have not been tested in man? 10. Based on thyroid weights alone. how many compounds have an activity of Class 5 or higher (Gives a group of compounds which could be further analyzed as in the questions above. ) (Group II. To correlate chemical structure with activity. ) 11. Do compounds in Questions 1, 3 and 10 have any common chemical grouping ? 12. Are there chemical groupings which confer inactivity? 13. Are heterocycle compounds more active in general than nonheterocycles ? 14. Knowing that compounds containing a -SH or =S group may be active, are any -SR compounds active? (R is anything but H) 15. Are any compounds of molecular weight >500 active ? We can now proceed to the demonstration. The plan followed in the CBCC machine room by Mr. Ballard, supervisor, was to first select a basic group of cards containing the information desired and then classify these. The classification was actually done by two procedures, and a description of both will illustrate the possibilities of the CBCC system. Since demonstrations in both the antihistamine and antithyroid field were planned for this symposium, a group of cards containing data in both fields was sorted out. For Procedure I, 1432 cards were available. Later, the number of cards increased considerably as the work of abstracters and coders flowed in, and all of these were used in Procedure II. Results are given in Table I. '
158 TABLE I ANALYSIS OF BASIC CARDS Operation (machine) No. of Cards Procedure I (used on preliminary data) Start 1432 1. Separate cards in Field F-3 (general action), punched in 207 in Columns 73-75 (antithyroid) 862 Rejects 570 2. Divide the antithyroid cards according to the test organism (Field A-l); get A4 (frogs) 9 A6 (chickens) 4 A7 (mammals) 849 3. Breakdown of A7 in (2): A72 (carnivora) [2 dogs, 1 cat] 3 A73 (rodents) 824 A74 (primates) [all man] 22 4. Breakdown of A73 (rodents) A730401 (guinea pigs) 2 A730201 (mice) 21 A730202 (rats) 801 5. Cards in activity classes 3, 5, 7, 9 relative to thiouracil as a standard (Field H) Quantitative Summary (Fields T * G), for rats 39 Procedure II (simplified from 1; used on final data) 1. Cards in Field F-3, punched 207 (antithyroid) 2081 2. Number of cards in response classes 3, 5, 7 and 9 (all animals; some cards counted twice because of double punches) 1041 3. Activity breakdown for rats - total cards 935 Class 3 383 Class 5 328 Class 7 142 Class 9 154 (double punches are present) 1007
159 In Procedure I, the cards punched for antithyroid action were separated. These were then sorted by machine according to the test animal. Since the questions deal with tests in rats and man, the cards for these were selected for further study. As well as giving the desired cards for further study, operations 1-2 also disclose that most of the antithyroid testing has been done with mammals, and rats in particular, as was expected. Since the questions in general deal with quantitative responses, the cards for rats were further treated by operation 5. Field T of the CBCC system records responses, and Field G gives quantitative summaries in five classes (!, 3, S, 7 and 9). When activities are given relative to a standard (specified in Field H), activity ratios are classified according to: Class Ratio of Test Chemical Standard 1 <0. 05 3 . 0.051 - 0. 5 5 0.51-5 7 5.1-50 9 >50 Much of the antithyroid data in the literature is given as activities relative to thiouracil as a standard. A compound twice as active as thiouracil would fall in Class 5, for instance. Thus, operation 5 gave the separation of cards necessary to answer the Group I questions. When more cards were available, time was saved by using Procedure II (Table 1) to obtain the same grouping of cards resulting from operation 5 of Procedure I. The fact that the machine counted 1007 punches from 935 cards indicates "double punching" This is sometimes done in the CBCC coding because the information available shows a range of activity greater than any one digit in Field G indicates (see above Classes vs. Ratio of Test Chemical to Standard). A given activity may be recorded in two or more ways, depending upon the criteria used in the evaluation. For instance, the antithyroid activity of a compound may be a response (a) relative to thiouracil as a standard; (b) based on hyperplasia; (c) based on thyroid weight; (d) based on iodine content of the gland; or (e) miscellaneous minor responses. If the recorded response involves two criteria, both may be punched but on separate cards. Summarizing the basic procedures, (I) gave incidental information as well as the basic cards to answer the questions and is an exploratory type operation; (II) was a direct approach to the basic cards. This illustrates that information on punches cards can be extracted by various procedures. When a large number of cards is involved, the actual running time of the machine is important; a skilled operator can reduce it to a minimum using an efficient pre-filing system.
160 The answers to the Group I questions are given in Table 2. Questions 1 and 2 require no comment. Question 3 was chosen to include the'earlier investigations where the results were not expressed relative to thiouracil. From Astwood's early and later papers wherein thiouracil was tested by both methods, the estimate was made that the answer to Question 3 should give compounds which are two-tenths as active as thiouracil or greater (particularly for Astwood's results). The CBCC code given under Question 3, Table 2, involves a rating also of verbal statements of activity, and the answer here probably covers other papers, which is all the better. The eighteen compounds found can be considered a basic group, of interest comparable to those of Question 1. Question 4 selects the more active ones, and is to be compared with Question 2. It so happens that the three compounds turned up have all been retested relative to thiouracil, but only 6-methyl-2-thiouracil was as active as thiouracil. A check of the thirty-two compounds in Question 2 confirms this - it is the only one of the three listed there. Therefore, the basic assumption that the procedure should turn up compounds as active as thiouracil or better is a little too broad. Incidentally, the values found by Astwood were 0.4 for 1, 3-diethylthiourea, 0.3 for tetramethylthiourea, and 1 for 6-methylthiouracil. Also, all three have had clinical trial in man. 6-Methylthiouracil is twice as active as thiouracil clinically, and is widely used for the treatment of hyperthyroidism. The other two were not as good as thiouracil but have not been definitely rated. TABLE 2 ANSWERS TO GROUP I QUESTIONS Question 1 Compounds tested relative to Thiouracil . breakdown: Question 2 Question 3 Answer: 18 compounds Question 4 Answer: CBCC Number 801.456 801,602 900,059 Class 3 5 7 9 Total (answer) 42 24 7 âI 74 The cards in activity classes 5, 7 and 9 (32 in number) of the group separated for Question 1 are examined for CBCC code number, and this is looked up in the files to give the name. Composed of compounds of Classes 5, 7 and 9 in Field G, punch -01 (Author's evaluation) or punch -02 (Coder's or Abstracter's evaluation) of the response criteria (in Field T): Verbally stated in abstract 0 0 inactive â¢f slight or mild + + + active or moderate +++ ++ very active .f + + extremely active Name urea, 1, 3-Diethyl-2-thio- urea, Tetramethylthio- uracil, 6-Methyl-2-thio-
161 TABLE 2 (Cont. ) Question 5 Answer: CBCC Number 800,253 800,694 900,059 900,587 800,569 900,899 901.796 Question 6 Answer: 70 compounds Question 7 Answer: CBCC Number 800,253 800,569 900.899 901.796 Question 8 Answer: None Compound urea, Thio- 2 - Benzimidazolethiol uracil, 6-Methyl-2-thio- uracil, 2-Thio- 2-Thiazoline-2-thiol barbituric acid, 5, 5-Diethyl-2-thio- uracil, x-Methyl-2-thio- Compound urea, Thio- 2-Thiazoline-2-thiol barbituric acid, 5, 5-Diethyl-2-thio- uracil, x-Methyl-2-thio- Indicates that 800,694, 900,059 and 900, 587 in Question 5 were not rated as to relative activities. Question 9 Answer: Ouestion 10 Answer: 26 compounds CBCC numbers given for 24 compounds of activity class 5, seven in class 7, and one in class 9. The numbers then looked up for the structure of the compounds. Question 5 necessitated the use of the collator. For this purpose, the cards from Question 1 were recombined with those from Question 3. These were then reassembled by the collator in order of the CBCC numbers. Thus, this stack was ready for comparison with the cards indicating tests in man, which were obtained by an operation on all the residue cards of Procedure II. The machine was then set to select cards on the basis of identical CBCC numbers in both stacks. The numbers and compounds obtained are listed in Table 2. Question 6 is an easy one to answer, since it merely involves a count of the rejects from Question 5. The answer was seventy compounds. The purpose of this question was to indicate that there is a large group of compounds highly active in rats which might logically be investigated clinically in man. Question 7 shows that four of the seven compounds tested in man are active. Question 8 gave the answer of none. This indicates that the residual compounds in Question 5 were not rated as to relative activities. This was easily checked by observation of the cards. Question 9 involved use of the collator as done for Question 5. The machine operation also gave the activity classes, which from the nature of the activity were Class 5, 7 or 9. There were twenty-four 5's, seven 7's, and one 9. The purpose of the question was to make available a list of highly active compounds (in rats) which might logically be tried in man. Question 6 told us that there were seventy highly active compounds by the rat tests which had not been tested in man, and Question 9 extracted and listed the most active thirty-two. From the results of the above questions, particularly No. 4, we might safely assume that all of these thirty-two are at least one-half as active as thiouracil in rats. The last question concerned with activity tests is No. 10. It gives a group of active compounds not included in the above. These were excluded from the detailed treatment because
162 some of the papers which report activities by the effect on thyroid weight alone are not of very high quality. The compounds in this group might well be analyzed as in the question above. In summary, the purpose of these ten questions was to test whether or not the CBCC system of recording and handling data on antithyroid compounds is useful in indicating compounds which might logically be further tested in man. The answer is - yes, it is useful. But since the field of antithyroid compounds is comparatively small, a critical study of the published papers would be possible for a clinician desiring to enter the field, and it would probably give better results. It is to be remembered that the CBCC system involves little criticism. Nevertheless, the results such as obtained here would be a good starting point for a newcomer in the field. The value of the CBCC system grows as the quantity of published data in a field increases. The second phase of this investigation on antithyroid compounds (Group II questions) is concerned with correlations of chemical structure with biological activity, and Questions 11 through 15 are associated. The procedure was to compare the cards obtained from Questions 1, 3 and 10 with the rotated chemical cards which were made up from all the basic antithyroid group including active and inactive compounds. You will recall that the rotated cards are obtained by duplicating each chemical card in such a way that each chemical group in turn appears first in the sequence of chemical code designations. Actually the Center had time to process only those cards from the Question 1 group. However, the procedure would be the same for the Question 3 and Question 10 cards. The result was a machine-typed listing of chemical code designations and their frequency of occurrence. That is, the number of cards for each chemical field was given, and these were also broken down according to all activity classes 1, 3, 5, 7 and 9. Question 11 asks for activity classes 3, 5, 7 and 9. Question 12 requires study of the activity class 1 cards, and Questions 13 and 14 use all activity classes. This sheet contained 180 compounds with chemical groups occuring 720 times. Duplicates of any one group in a given compound were excluded. Seventy compounds had antithyroid activity and 110 were inactive. These numbers are probably not great enough for detailed statistical analysis. i Question 1 1 was written to see if a chemical grouping such as -N-C-SH could be detected, since published reviews indicate that this group is found in most of the more active antithyroid compounds. This particular group does not appear in the CBCC code. It might appear from a detailed procedure of.comparing the most active CBC groups which do occur, and analyzing them for common structural features. When a larger basic group of cards on antithyroid compounds is available, this process will be worthwhile. For the moment, no common chemical sub-group- ing is apparent by inspection of the active compounds. Answer to 11: none apparent. Question 12 also got a negative answer. Question 1 3 - to answer this, a count of heterocycles in both the active and inactive classes was made. Sixty-four heterocycles, or 27 per cent of the 241 groups in the active classes, were found. There were eighty heterocyclies in 479 groups in the inactive class, or 17 per cent. One might conclude that heterocycle groups are favorable to antithyroid activity. Question 14 was answered by observing that five compounds containing sulfide, or R-S-R' groups, were active. If these had not been observed, an investigation of such groups as Code 825, which includes thioureas and pseudothioureas, would be necessary. The pseudothioureas are ]^>C-SR compounds, and the thioureas ^yC-S compounds. When Question 15 was written, the author did not know that molecular weights were not coded. Coding of physical data is not now done, but there is considerable interest in such at the Center. The staff kindly calculated the molecular weights for the one hundred and six active compounds and the three hundred and nineteen inactive ones. Only one turned up in the active group which had a molecular weight greater than 500. This was thyroxine, which did not belong there anyhow, and slipped in by a mistake in punching (although it is an indirect antithyroid compound). Four inactive compounds of molecular weight greater than 500 were found. This is not a significant number. Perhaps someone should test more high molecular weight compounds.
163 These questions were, of course, only examples of what can be handled by the CBCC ' ^ero. Many interesting things might well turn up on statistical analysis of the frequency of :1ccurr'nce of chemical groups as related to biological activity. As the quantity of data grows, the value of the work done by the CBCC will increase. Perhaps some day we shall call it indispens - able