APPUED MICROBIOLOGY, Nov. 1969, p. 861-868 Copyright © 1969 American Society for Microbiology Vol. 18, No. 5 Printed in U.S.A. Automated Instrument for the Fluorescent Treponemal Antibody-Absorption Test and Other Immunofluorescence Tests GERALD F. BINNINGS, MEL J. RILEY, MERRITT E. ROBERTS, RICHARD BARNES, AND THOMAS C. PRINGLE Medical Instrumentation, Chemical & Biological Department, Space Division, Aerojel General Corporation, El Monte, California 91734 An automated diagnostic test instrument and its development program are described. The instrument automates the fluorescent treponemal antibody-absorption test for syphilis to the extent that only 4 hr of technician time is required to conduct approximately 200 tests daily. Evaluation to date suggests its efficacy. In addition, preliminary studies indicate the feasibility of detecting antibodies to Toxoplasma gondii, Plasmodium malariae, and nucleoprotein (antinuclear factor). The instrument would seem to have broad application for routine and research immunofluorescence testing. Two elements comprise the instrument: a slide processor and a microscope attachment. The slide processor is an electro-pneumatically actuated device which automatically feeds special laboratory slides, on which antigen or other reagents are prefixed, through a series of operations which provide reagent application, incubation, washing, drying, and stacking of the finished slides for readout. The instrument provides flexibility in that incubation time and temperature as well as point, sequence, and duration of reagent application can be varied to accommodate a variety of immunofluorescence techniques. The microscope attachment can be fitted to all conventional dark-field fluorescence microscopes and makes possible the reading of three to six slides per minute. The reacted slides from the processor are injected sequentially onto the stage of the microscope by movement of a lever. As injected, slides are automatically in visual focus; fine focus is occasionally required. Scanning of the reacted field is accomplished by means of the normal microscope controls. A buffered glycerol coupling is maintained between the darkfield condenser substage lens and the slide cover glass by means of a pushbuttonactuated feed system. Aerojet-General Corp., Chemical & Biological Division, under contract to Department of the Army, Rapid Warning Office, Fort Detrick, Frederick, Md., developed a number of immunofluorescence systems for the specific detection of biological agents in air samples. Certain techniques developed during that time appeared applicable to the automation of current, manually conducted immunofluorescent diagnostic tests. A survey of such tests was made which led to the selection of a first candidate test for automation. Factors considered in the survey included costs associated with the tests, levels of skill required to conduct the tests, total number of tests conducted, and the need or desirability to increase the number of tests that could be performed. The fluorescent treponemal antibody-absorption (FTA-ABS) test (2) was selected as the first 861 candidate for automation primarily because it appeared to us that there was a need for a syphilis screening test more sensitive and specific than the present, extensively used Venereal Disease Research Laboratory (VDRL) slide test. It is estimated that there were approximately 38 million standard tests for syphilis (STS) conducted in 1968 in the United States, of which approximately 90 to 95% were the VDRL slide test. Data reported (1) indicate that correlation of test results with diagnostic categories is 77% for the VDRL and 95%0 for the FTA-ABS. The cost and time required to conduct the FTA-ABS test have prevented its use for large-volume testing; however, it appeared to us that, if automated, it might be found suitable for such screening, thus permitting more accurate case finding and diagnosis and subsequent improvement in control of the disease. Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. Received for publication 14 August 1969 862 BINNINGS ET AL. APPL. MICROBIOL. showed that the agreement between the manual and automated FTA-ABS tests in this study was 87%. Several changes were made in the AFTA test procedure so that it would more nearly approximate the manual procedure. Subsequent test agreement on 201 specimens in clinically defined categories increased to 90%, with the manual FTA-ABS test showing slightly higher sensitivity. Additional refinements are being made which are expected to raise the sensitivity of the AFTA test and to improve the agreement with the manual procedure (G. W. Stout et al., American Public Health Association Annual Meeting, Detroit, Mich., 1968). Simultaneously with this test program, 15 instruments of the production design were fabricated, and a program was formulated by the National Communicable Disease Center for further field evaluations at California State Department of Public Health Laboratory, Berkeley; Venereal Disease Research Laboratory, National Communicable Disease Center, Atlanta, Ga.; Georgia State Department of Public Health Laboratory, Atlanta; and New Jersey State Department of Public Health Laboratory, Trenton. As of this writing, the Berkeley and National Communicable Disease Center programs are completed and the Georgia and New Jersey programs have been initiated. The results of these test lished (3, 4). Under phase II of the program, 1 December programs will be the subject of individual papers 1967 to 9 August 1968, a "breadboard" instru- to be published by each testing agency. ment was developed and tested. A prototype INSTRUMENT model was established and four prototype instruThe automated instrumentation, known as the ments were designed and fabricated. The prototype systems were tested at Aerojet-General SeroMatic System, is comprised of two major Corp., at the Venereal Disease Research Labora- items: a processor (Fig. 1) and a microscope attory, National Communicable Disease Center, at tachment (Fig. 2). The processor is an electro-pneumatically conSan Bernardino County Health Laboratory, San Bernardino, Calif., and at Long Beach Memorial trolled laboratory test instrument which autoHospital, Long Beach, Calif., during the period 26 mates the performance of the FTA-ABS test on as January 1968 to 9 August 1968. A total of 4,500 many as 200 test specimens per day (7.5 hr of sera were tested. Results were utilized to make continuous processor operation). functional improvement during the test program. Specimens are processed sequentially, the first Phase III of the program, 10 August 1968 to 21 specimen being fully processed and ready for October 1968, comprised the application of ex- microscope examination 85 min, 45 sec after perience from phase II into a production design processor startup. Subsequent specimens are fully and the fabrication of one production unit. This processed at intervals of 1 min, 45 sec thereafter. unit was then used in a study by the Venereal Figure 3 shows the process sequence, and Fig. 4 Disease Research Laboratory, National Com- and 5 are internal views of the processor. The processor is provided with two reservoirs municable Disease Center. In the study, the automated fluorescent treponemal antibody-absorp- for containing the wash fluids used during specition (AFTA) test was performed on 859 speci- men processing and with a set of 20 reusable 10mens from clinically defined donor groups, and cavity serum-sorbent applicator blocks. Processor specification. The processor is 35 the results were compared with those of the manual FTA-ABS test. The two tests were per- inches wide X 25.5 inches deep X 22 inches high formed independently by different groups of (89 X 65 X 56 cm) and weighs approximately 120 serologists at the Venereal Disease Research lb (54.5 kg). Utility requirements are electrical Laboratory. A comparison of the results obtained power (120 v, 60 Hz at 10 amp maximum) and air Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. Even with present patterns of use of serological tests, an automated version of the FTA-ABS test would provide a more ready means for detecting patients having false-positive reactivity in cardiolipin tests for syphilis and would enhance the detection of patients with primary, latent, or late syphilis [clinical stages of the disease in which present routine cardiolipin tests are not as sensitive as the FTA-ABS test (2) ]. DEVELOPMENT PROGRAM A development program for the automation of the FTA-ABS test was undertaken in 1967 by Aerojet-General Corp., with technical assistance and liaison from the Verneral Disease Research Laboratory, National Communicable Disease Center, Atlanta, Ga. During the first phase of the program, 3 January 1967 to 30 November 1967, the manual FTA-ABS test was studied to determine the effects of variations in protocol and the quality of commercially available reagents, and to develop techniques for reagent production. From this study, a protocol for the automated test was established, a process for the production of high quality reagents was proven in pilot lots, and simulated automated tests were conducted successfully. Certain aspects of the reagent development work have been pub- AUTOMATED INSTRUMENT FOR FTA-ABS TEST pressure (40 to 50 psig at 2 ft3/min). A floor sink or 10-gal (38-liter) reservoir is required for collection of process fluid waste. Acceptable ambient temperature and relative humidity limits are 20 to 30 C and 30 to 90%, respectively. 863 Reagents required to perform the automated FTA-ABS test are as follows. Antigen: one REDIFIX antigen slide per specimen. (The slide is comprised of a 25 X 75 x 1 mm thick styrene plastic frame into which is mounted a 22 X 40 mm FIG. 1. Processor. FIG. 2. Microscope attachment. Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. VOL. 18, 1969 864 BINNINGIs ET AL. APPL. MICROBIOL. no. 1/2 cover glass. On the cover glass is a meth- specimen. PBS:approximately 20 ml per specianol-fixed 2.5-mm spot of Nichols T. pallidum men. Distilled water:approximately 160 ml per antigen. After fixing, the antigen density is 50 to specimen. Process sequence. The proces sequence is 100 organisms per high power field.) Sorbent: 0.1 ml per specimen. Patient or reference serum: graphically illustrated in Fig. 3 and is described not less than 0.025 ml per specimen. Titered anti- below by following one slide through the entire human gamma globulin conjugate labeled with sequence. At the load station (station no. 1), a REDI-FIX fluorescein isothiocyanate [diluted in 2% Tween 80 in phosphate-buffered saline (PBS)]:0.1 ml per antigen slide is inserted, by the loader mechanism, PBS WASH -10 SECONDS PBS SOAK -7 MINUTES 1ST D/W WASH 30 SECONDS -' 1st DRY -20 SE APPLY % CONJUGATE % 'FINAL DRY -7 MINUTES *2nd D/W WASH - 30 SECONDS TOTAL PROCESSING TIME - 85 MIN., 45 SEC. FIG. 4. Internal top view. Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. FIG. 3. Processing sequence. VOL. 18, 1969 AUTOMATED INSTRUMENT FOR FTA-ABS TEST 865 from a slide-handling magazine into one of 50 slide-holding positions which are equally spaced around the periphery of a circular rotating table. The rotating table is advanced (indexed) 1/50 revolution (1 station) counter-clockwise every 1 min, 45 sec. Subsequent slides are inserted into adjacent slide-holding positions as these positions are moved into line with station no. 1. At 3.5 min after a slide has been inserted into a slide holder, it becomes positioned under the serum-sorbent applicator (station no. 3) where the serum-sorbent mixture is forced from a cavity of a serum-sorbent block onto the antigen spot on the slide. This station is provided with a slide-sensing interlock to maintain synchronization of serum specimens with slides. The slide is then indexed through the first incubator. The first incubator is 17 stations in length; therefore, the slide will remain in the incubator 29 min, 45 sec, during which time the antigen-serum reaction takes place. Incubator temperature is maintained at 37 C. The first incubation is terminated by indexing the slide out of the incubator and under the PBS wash nozzle. This nozzle applies PBS, pH 7.2, onto the slide for a period of 10 sec. Upon termination of the wash, a portion of the PBS wash fluid remains on the slide and in contact with the serum-reacted antigen spot. During the remainder of the time which the slide remains at this station, as well as during the time it spends at the next three stations (approximately 7 min total), PBS soak occurs. When the table is next indexed (this now being 42 min from the time the slide was inserted into its holder), the slide is brought to station no. 25 where the slide is deflected downward and rinsed with distilled water for a period of 30 sec to remove the PBS and the particulate matter not related to the antigen-serum reaction. Upon termination of the distilled water wash, the air dry nozzle (also located at this station) is activated for 20 sec to remove all distilled water from the top surface of the slide cover glass. Next, the table is indexed to the conjugate applicator station where a 0.1-ml drop of conjugate is applied to the antigen spot on the slide cover glass. The slide advances through the second incubator, remaining in the incubator for a period of 29 min, 45 sec. Incubator temperature is maintained at 37 C. The slide is next indexed to the second distilled water wash station (station no. 44), where a 30sec distilled water wash is applied to remove residual conjugate. The slide is then indexed through four stations where large droplets of distilled water are wiped Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. FiG. 5. Bottom view. 866 BINNINGS ET AL. APPL. MICROBIOL. from the side and warm air from the final dry functions are interlocked to occur during a given blower is forced across the slide to remove final timer cycle only if a slide is present at the station traces of moisture. where the function is to occur. The slide is finally indexed to the unloader staFigure 6 graphically illustrates the operation of tion (station no. 49) where the slide is ejected from all timer-controlled functions where the "home" the rotating table onto an unload platform. The position of the cam timer is specified as reference. unload mechanism carries the slide from this plat- Those functions shown as solid lines occur each form into a slide unload magazine. time cycle. Those functions shown as dashed lines The process sequence is terminated (85 min, 45 occur only when a slide is present at the stations sec after the slide was inserted into a slide holder), where such functions are to occur. and the slide is ready for examination with a The timer is of a design that permits rapid and fluorescence microscope assembly. simple alteration of the timer cycle program, Processor control. All time-related operations where desired. This includes changing the point in are controlled by a rotating cam timer which is time, during a timer cycle, that a function occurs attached to the underside of the instrument deck and, in the case of the three washes, the duration and accessible through a hinged panel on the front of the washes. face of the instrument. Microscope attachment. The microscope attachThe cam timer is driven by a 1 rev/min syn- ment, when installed onto a dark-field fluorescence chronous motor and is appropriately geared to microscope, permits REDI-FIX antigen slides to rotate one full revolution every 1 min, 45 sec. be visually examined at a rate consistent with the A total of 14 cams are attached to the timer maximal daily capacity of the processor by minishaft. Eleven of these cams are operational, the mizing the handling of slides and minimizing the remaining three are spares. As the shaft rotates, necessary manipulation of microscope controls by each cam either opens or closes an associated elec- the operation. Figure 7 shows a typical installatrical switch or pneumatic valve at a preset time tion of the attachment onto a microscope. [The and for a preset duration. These switches and following combination of microscope equipment valves, in turn, control the various mechanical has been found suitable for use in examination of functions of the processor. processed slides: illumination system consisting Certain functions of the process occur during of an Osram HBO-200 ultraviolet light source and each revolution of the timer cycle, such as table 6-v 30-w tungsten light source; excitation filtration indexing and slide loading. Serum-sorbent appli- consisting of a 2-mm thick KG 1 (heat absorpcation, fluid washes, and conjugate application tion) filter, a 2-mm thick BG-12 (or equivalent) L CAM ONE REVOLUTION OF SEQUENCE PROGRAMMER (ONE CYCLE) 0 FUNCTION A TIMER HOME SWITCH (ELECT.) B TABLE INDEX C SLIDE LOAD/UNLOAD D SLIDE EJECT E SERUM INDEX F SERUM DROP G CONJUGATE DROP H 1st AIR DRY 1st DISTILLED WATER WASH J 2nd DISTILLED WATER WASH K SPARE 1 L SPARE M SPARE N PBS WASH L LEGEND: ...J OPERATES EVERY CYCLE _ _ FIG. 6. Timing control diagram. -L _ OPERATES ONLY IF INTERLOCK ENGAGED Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. TIME (SEC) VOL. 18, 1969 AUTOMATED INSTRUMENT FOR FTA-ABS TEST 867 filter and a 2-mm thick BG-38 (or equivalent) filter; a 2-mm thick OG-1 (or equivalent) barrier filter, an immersion-type dark-field condenser having a numerical aperture of 1.20; a 40 x high dry objective with a numerical aperture of 0.65; and a 10 x eyepiece.] The attachment accepts a processor unloader magazine containing the processed slides. The slides are accurately positioned under the objective lens of the microscope by the operator. The design of the attachment is such that the antigen spot on the slide is in sufficient focus to require only very slight adjustment of the microscope fine focus control to obtain a sharp field image. A glycerol feed assembly is provided with the attachment (see Fig. 7). When a slide has been positioned under the objective, the operator depresses and releases the plunger of this assembly to glycerol-couple the slide cover glass to the darkfield condenser. After examining a slide, the operator positions the next slide under the objective, thereby pushing the previous slide clear. The unique feature of this method is that the REDI-FIX slides are placed in the attachment in an inverted position; that is, the reacted antigen spot is on the lower side of the slide. By handling slides in this manner, the need for installing a cover glass over the specimen (as with conventional slides in fluorescence microscopy) is eliminated. DISCUSSION The feasibility of utilizing the automated instrumentation in the performance of routine immunofluorescence tests has been demonstrated. This has been accomplished by successfully automating the repetitious, time-consuming, and often tedious steps, including those necessary during slide readout, associated with manually performed tests. Test results to date give general indication that the FTA-ABS test has been successfully automated. The subject instrument will permit an increase in the total number of tests which can be performed per unit operator time to a point where the test may be used more extensively for screening for syphilis. Preliminary experiments (Norins et al., to be published) have demonstrated the feasibility of using this instrumentation in the detection of antibodies to malaria parasites, toxoplasma organisms, and nucleoprotein (antinuclear factor). Experimentation continues in an effort to discover additional immunofluorescence tests which can be similarly automated. It is felt that the inherent functional flexibility of this instrument may be a further measure of its value, providing investigators with a fertile tool in the field of developmental research. ACKNOWLEDGMENTS The instrument development program was inspired by Leslie C. Norins of the Venereal Disease Research Laboratory, National Communicable Disease Center, Atlanta, Ga. In addition, G. W. Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. FIG. 7. Microscope with attachment and glycerol feed assembly. BINNINGS ET AL. Stout, V. Falcone, A. Wallace, W. Duncan, and J. Lewis of the Venereal Disease Research Laboratory provided valuable advice and assistance in the conduct of evaluation tests and furnished information helpful to the refinement of the design. Others who rendered valuable assistance in conducting evaluation test programs and making suggestions which were helpful for the improvement of the design were R. Wood, E. Coffey, and R. Jue, California State Department of Public Health Laboratory, Berkeley; S. Brandon, C. Gates, and J. Dalton, Georgia State Department of Public Health Laboratory, Atlanta; E. Thomas and C. Levy, New Jersey State Department of Public Health Laboratory, Trenton; F. Gettman and H. Davis, San Bernardino County Health Department Laboratory, San Bernardino, Calif.; and M. Mercer, Long Beach Memorial Hospital, Long Beach, Calif. We thank E. Mishuck of Aerojet-General Corp. for encouragement and counsel throughout the course of these studies. APPL. MICROBIOL. LITERATURE CITED 1. Deacon, W. E., J. B. Lucas, and E. V. Price. 1966. Fluorescent treponemal antibody-absorption (FTA-ABS) test for syphilis. J. Amer. Med. Assoc. 198:624-628. 2. Hunter, E. F., L. C. Norins, V. Falcone, and G. W. Stout. 1968. The fluorescent treponemal antibody-absorption (FTAABS) test, development, use and present status. Bull. World Health Organ. 39:873-881. 3. Roberts, M. E., J. N. Miller, and G. F. Binnings. 1968. Reduction of nonspecific background staining in the fluorescent treponemal antibody-absorption test. J. Bacteriol. 96:15001506. 4. Roberts, M. E., J. N. Miller, T. C. Pringle, and G. F. Binnings. 1968. Shelf life of fluorescent treponemal antibody-absorption test reagents. J. Bacteriol. 96:1507-1511. Downloaded from https://journals.asm.org/journal/am on 11 August 2021 by 3.238.203.22. 868