Correspondence SKIN TEST ANTIGENS: AN EVALUATION WHOSE TIME HAS COME
To the
Editor:
T h e recent editorial by Sbarbarol points out some of the gaps in our knowledge and the need for further studies. Sbarbaro served as chairman of the Panel on Review of Skin Test Antigens. The Panel is to be commended for an excellent job, despite the complex nature of the antigens and the varying responses to them. As with many other biologic products, there is need for improvement in skin test antigens and their standardization. However, flaws that exist are not, as Sbarbaro states, "fatal," and skin test antigens have provided useful information for many years. T h e publication of the recommendations2 underscores the need to continue the efforts that have been made towards the improvement in the use and understanding of these diagnostic aids. Changes in the final recommendations on the basis of comments submitted to the hearing clerk are likely. Variation in production continues to be a problem despite efforts to prepare reproducible batches. This is why the Food and Drug Administration has repeatedly stressed the advantages of preparing a large batch of each skin test antigen that can be stored under conditions favorable to stability and characterizing each, both in animals and in humans. From such batches, numerous lots of final product can be prepared for many years. Coccidioidin. Sbarbaro questioned the quantitation methods for standardization of coccidioidin products as not having been done appropriately. However, he apparently accepts the data on which deterioration of coccidioidin is claimed to have occurred, even though it was based on the same kind of data. The data on stability should not be construed as accurate while the data on standardization are considered unacceptable. Batches of coccidioidin are being characterized and standardized on the basis of dose-response curves. This is a difficult task involving the use of human volunteers, a great expenditure of time and money, and often results in data that are difficult to interpret. Histoplasmin. Histoplasmin multiple-puncture devices have not been marketed for many years because problems attendant to their use were recognized. T h e only available histoplasmin skin test on i Sbarbaro, J. A.: Am Rev Respir Dis, 1978, 118, 1. 2 Federal Register, September 30, 1977, vol. 42, no. 90,42674.
the market today is one that has been prepared as a large batch and characterized in animals and human beings by means of dose-response curves. Data on stability are being evaluated. Tuberculins. Guidelines for the interpretation of reactions to the 5 T U dose of tuberculin, PPD administered by the Mantoux® procedure are the most meaningful available for skin test products. However, these guidelines contain inadequacies and fail to correlate to significance of reaction size and the nature of the mycobacteria endemic to various geographic locations. Nor are the guidelines adequate in the consideration of subjects who have been in contact with a tuberculosis patient. As screening tests, multiple-puncture devices are expected to err in a false-positive direction relative to a 5 T U tuberculin, PPD Mantoux test. Even the Mantoux test is, on repeated or simultaneous testing, inconsistent 5 per cent or more of the time. It may be unrealistic to expect that multiple-puncture devices de teet 100 per cent of all persons who are known to have had bacteriologically confirmed tuberculosis and who are tuberculin positive by a simultaneous Mantoux test with 5 T U of PPD. T h e Panel was, perhaps, a bit overzealous in recommending that a bulk be acceptable only if it performs perfectly in bacteriologically confirmed Mantoux positive populations. The precise amount of tuberculin deposited on the skin by multiple-puncture devices is unknown, especially for those on which tuberculin is dried or semi-dried, because this can vary with texture, thickness, and moisture of the skin, period of contact with the skin, and the nature of the tuberculin itself. Recent statistical observations with tuberculin products administered with multiple-puncture devices indicate that closer correlations with reactions for Mantoux tests can be achieved. However, this effect is probably achieved with some loss in the false-positive rate and a corresponding increase in the false-negative rate. Recommendations are more easily proposed than they are executed. The state of the art of skin testing and interpretation of skin testing is far from "ideal." Cooperation is required from multiple sectors including professional organizations, universities, hospitals, and government. New restrictions on the use of human subjects for experimentation and the need for detailed consent forms hamper accumulation of the required skin test data. A great deal of planning time, money, and paper work are involved in efforts to conduct clinical studies properly. T h e Bureau of Biologics is preparing standards for the control of skin test antigens and their evalua-
AMERICAN REVIEW OF RESPIRATORY DISEASE, VOLUME 118, 1978
1109
1110
CORRESPONDENCE
tion. T h e development of spherule-phase coccidioidin has been funded by the Bureau. Criteria for standardization recommended by the Panel for this product had been applied before licensure of the product. Similarly, the production of a yeast-phase histoplasmin skin test is being funded by the Bureau of Biologics and is currently being evaluated in human beings. In both of these studies, standardization has been accomplished by means of dose-response curves. With regard to coccidioidin and histoplasmin, the Panel did not consider whether the classification of a reaction of 5 mm or more of induration as a positive reaction is too arbitrary. Does a reaction size of 5 mm have the same meaning as one of 35 mm? In contrast to tuberculin PPD, guidelines for interpretation of large versus small reactions to fungal skin test antigens are not available. Such data need to be generated. New standard reagents for these products have been prepared and will be evaluated. The value of these reagents awaits the accumulation of data concerning their use. There are now 6 products licensed for the administration of tuberculin by means of multiple-puncture devices. T h e Bureau of Biologics has initiated, with the cooperation of the Center for Disease Control, a contract for the evaluation of each of these products in a randomized and double-blind manner according to recommended guidelines. It is anticipated that approximately 35,000 test subjects will be required to provide the data necessary for interpretation of the efficacy of these skin test products. T h e Bureau of Biologics has undertaken many o ? these studies because these responsibilities have largely been ignored by professional groups and manufacturers. Perhaps the most important message in Sbarbaro's editorial is that more professional groups should be active in the pursuit of "ideal" biologic products. Further cooperation among government regulatory agencies such as the FDA, professional organizations, and manufacturers is needed not only in the definition of problems but also in the implementation of the necessary improvements. SOTIROS D . CH AP ARAS
Director, Mycobacterial and Fungal Antigens Branch Division of Bacterial Products Bureau of Biologics Food and Drug Administration U. S. Public Health Service Bethesda, Md. 20014 August 8, 1978
From the
Author:
In characteristic style, Dr. Chaparas has underscored some of the areas of serious concern faced by the Panel on Skin Test Antigens. Although the editorial
was based entirely on the opinion of the Panel, the response to Dr. Chaparas's observations will, by necessity, reflect only the opinion of its author. Coccidioidin. Dr. Chaparas confirms what the Panel noted—that we do not know the qualitative and quantitative status of products presently on the market, and if the manufacturers are allowed to continue their present methods of standardization, we will continue to have clinicians throughout the United States relying on materials of unknown effectiveness. There is no question that the standardization of test materials is a difficult task; it involves the use of human volunteers and indeed, requires both time and money. However, these features do not justify the continued sale or products that may mislead and confuse the clinician, unless it is made quite clear to the clinician that he is dealing with a substance that at one time in the past, or sometime in the future, may or may not provide him with accurate information. The Panel did not accept the data on stability as accurate, while discounting the manufacturer's data on standardization as unacceptable. Rather, the combined information merely confirmed the confusion surrounding the presently marketed product. Histoplasmin. Three companies applied for continued licensure of their product: the Parke, Davis Company, Lederie (for its histoplasmin Tine® test), and the Michigan Department of Public Health. Obviously, Dr. Chaparas is referring to the Parke, Davis product, and to quote the Panel's review of Parke, Davis's application concerning effectiveness: No specific data are furnished on the ability of Parke Davis histoplasmin to identify all individuals who have previous immunologic experience with histoplasma. A study by Furcolow in 1950 submitted to the manufacturer, demonstrated that the results with Parke Davis Lot H - l l are comparable to those obtained with the reference standard preparation H-42 in approximately 160 persons. T h e raw data in terms of average size of skin reaction were not made available and one cannot evaluate the conclusions of the author. New production batches in subsequent lots are standardized against U.S. Reference Standard No. 1 by a 4 X 4 design on three to five guinea pigs. There is no standardization of these batches or lots in the human model. Therefore, I must take issue with Dr. Chaparas's statement that the present, marketable histoplasmin has been characterized in human beings by means of dose-response curves. Dr. Chaparas attended many sessions of the Panel and if he had access to such data, we were not made aware of it. Both Dr. Chaparas and the Parke, Davis Company were aware that the Panel was placing the product in Category 3A, i.e., "available data are insufficient," and no objections were heard. In faet, the Panel's recommendation calls for the FDA to obtain data comparing
To the
Editor:
T h e recent editorial by Sbarbarol points out some of the gaps in our knowledge and the need for further studies. Sbarbaro served as chairman of the Panel on Review of Skin Test Antigens. The Panel is to be commended for an excellent job, despite the complex nature of the antigens and the varying responses to them. As with many other biologic products, there is need for improvement in skin test antigens and their standardization. However, flaws that exist are not, as Sbarbaro states, "fatal," and skin test antigens have provided useful information for many years. T h e publication of the recommendations2 underscores the need to continue the efforts that have been made towards the improvement in the use and understanding of these diagnostic aids. Changes in the final recommendations on the basis of comments submitted to the hearing clerk are likely. Variation in production continues to be a problem despite efforts to prepare reproducible batches. This is why the Food and Drug Administration has repeatedly stressed the advantages of preparing a large batch of each skin test antigen that can be stored under conditions favorable to stability and characterizing each, both in animals and in humans. From such batches, numerous lots of final product can be prepared for many years. Coccidioidin. Sbarbaro questioned the quantitation methods for standardization of coccidioidin products as not having been done appropriately. However, he apparently accepts the data on which deterioration of coccidioidin is claimed to have occurred, even though it was based on the same kind of data. The data on stability should not be construed as accurate while the data on standardization are considered unacceptable. Batches of coccidioidin are being characterized and standardized on the basis of dose-response curves. This is a difficult task involving the use of human volunteers, a great expenditure of time and money, and often results in data that are difficult to interpret. Histoplasmin. Histoplasmin multiple-puncture devices have not been marketed for many years because problems attendant to their use were recognized. T h e only available histoplasmin skin test on i Sbarbaro, J. A.: Am Rev Respir Dis, 1978, 118, 1. 2 Federal Register, September 30, 1977, vol. 42, no. 90,42674.
the market today is one that has been prepared as a large batch and characterized in animals and human beings by means of dose-response curves. Data on stability are being evaluated. Tuberculins. Guidelines for the interpretation of reactions to the 5 T U dose of tuberculin, PPD administered by the Mantoux® procedure are the most meaningful available for skin test products. However, these guidelines contain inadequacies and fail to correlate to significance of reaction size and the nature of the mycobacteria endemic to various geographic locations. Nor are the guidelines adequate in the consideration of subjects who have been in contact with a tuberculosis patient. As screening tests, multiple-puncture devices are expected to err in a false-positive direction relative to a 5 T U tuberculin, PPD Mantoux test. Even the Mantoux test is, on repeated or simultaneous testing, inconsistent 5 per cent or more of the time. It may be unrealistic to expect that multiple-puncture devices de teet 100 per cent of all persons who are known to have had bacteriologically confirmed tuberculosis and who are tuberculin positive by a simultaneous Mantoux test with 5 T U of PPD. T h e Panel was, perhaps, a bit overzealous in recommending that a bulk be acceptable only if it performs perfectly in bacteriologically confirmed Mantoux positive populations. The precise amount of tuberculin deposited on the skin by multiple-puncture devices is unknown, especially for those on which tuberculin is dried or semi-dried, because this can vary with texture, thickness, and moisture of the skin, period of contact with the skin, and the nature of the tuberculin itself. Recent statistical observations with tuberculin products administered with multiple-puncture devices indicate that closer correlations with reactions for Mantoux tests can be achieved. However, this effect is probably achieved with some loss in the false-positive rate and a corresponding increase in the false-negative rate. Recommendations are more easily proposed than they are executed. The state of the art of skin testing and interpretation of skin testing is far from "ideal." Cooperation is required from multiple sectors including professional organizations, universities, hospitals, and government. New restrictions on the use of human subjects for experimentation and the need for detailed consent forms hamper accumulation of the required skin test data. A great deal of planning time, money, and paper work are involved in efforts to conduct clinical studies properly. T h e Bureau of Biologics is preparing standards for the control of skin test antigens and their evalua-
AMERICAN REVIEW OF RESPIRATORY DISEASE, VOLUME 118, 1978
1109
1110
CORRESPONDENCE
tion. T h e development of spherule-phase coccidioidin has been funded by the Bureau. Criteria for standardization recommended by the Panel for this product had been applied before licensure of the product. Similarly, the production of a yeast-phase histoplasmin skin test is being funded by the Bureau of Biologics and is currently being evaluated in human beings. In both of these studies, standardization has been accomplished by means of dose-response curves. With regard to coccidioidin and histoplasmin, the Panel did not consider whether the classification of a reaction of 5 mm or more of induration as a positive reaction is too arbitrary. Does a reaction size of 5 mm have the same meaning as one of 35 mm? In contrast to tuberculin PPD, guidelines for interpretation of large versus small reactions to fungal skin test antigens are not available. Such data need to be generated. New standard reagents for these products have been prepared and will be evaluated. The value of these reagents awaits the accumulation of data concerning their use. There are now 6 products licensed for the administration of tuberculin by means of multiple-puncture devices. T h e Bureau of Biologics has initiated, with the cooperation of the Center for Disease Control, a contract for the evaluation of each of these products in a randomized and double-blind manner according to recommended guidelines. It is anticipated that approximately 35,000 test subjects will be required to provide the data necessary for interpretation of the efficacy of these skin test products. T h e Bureau of Biologics has undertaken many o ? these studies because these responsibilities have largely been ignored by professional groups and manufacturers. Perhaps the most important message in Sbarbaro's editorial is that more professional groups should be active in the pursuit of "ideal" biologic products. Further cooperation among government regulatory agencies such as the FDA, professional organizations, and manufacturers is needed not only in the definition of problems but also in the implementation of the necessary improvements. SOTIROS D . CH AP ARAS
Director, Mycobacterial and Fungal Antigens Branch Division of Bacterial Products Bureau of Biologics Food and Drug Administration U. S. Public Health Service Bethesda, Md. 20014 August 8, 1978
From the
Author:
In characteristic style, Dr. Chaparas has underscored some of the areas of serious concern faced by the Panel on Skin Test Antigens. Although the editorial
was based entirely on the opinion of the Panel, the response to Dr. Chaparas's observations will, by necessity, reflect only the opinion of its author. Coccidioidin. Dr. Chaparas confirms what the Panel noted—that we do not know the qualitative and quantitative status of products presently on the market, and if the manufacturers are allowed to continue their present methods of standardization, we will continue to have clinicians throughout the United States relying on materials of unknown effectiveness. There is no question that the standardization of test materials is a difficult task; it involves the use of human volunteers and indeed, requires both time and money. However, these features do not justify the continued sale or products that may mislead and confuse the clinician, unless it is made quite clear to the clinician that he is dealing with a substance that at one time in the past, or sometime in the future, may or may not provide him with accurate information. The Panel did not accept the data on stability as accurate, while discounting the manufacturer's data on standardization as unacceptable. Rather, the combined information merely confirmed the confusion surrounding the presently marketed product. Histoplasmin. Three companies applied for continued licensure of their product: the Parke, Davis Company, Lederie (for its histoplasmin Tine® test), and the Michigan Department of Public Health. Obviously, Dr. Chaparas is referring to the Parke, Davis product, and to quote the Panel's review of Parke, Davis's application concerning effectiveness: No specific data are furnished on the ability of Parke Davis histoplasmin to identify all individuals who have previous immunologic experience with histoplasma. A study by Furcolow in 1950 submitted to the manufacturer, demonstrated that the results with Parke Davis Lot H - l l are comparable to those obtained with the reference standard preparation H-42 in approximately 160 persons. T h e raw data in terms of average size of skin reaction were not made available and one cannot evaluate the conclusions of the author. New production batches in subsequent lots are standardized against U.S. Reference Standard No. 1 by a 4 X 4 design on three to five guinea pigs. There is no standardization of these batches or lots in the human model. Therefore, I must take issue with Dr. Chaparas's statement that the present, marketable histoplasmin has been characterized in human beings by means of dose-response curves. Dr. Chaparas attended many sessions of the Panel and if he had access to such data, we were not made aware of it. Both Dr. Chaparas and the Parke, Davis Company were aware that the Panel was placing the product in Category 3A, i.e., "available data are insufficient," and no objections were heard. In faet, the Panel's recommendation calls for the FDA to obtain data comparing
