Evaluation of Liquid Scintillation Counting Solutions for Radioimmunoassay of Hepatitis-associated Antigen (HBsAg) E. J. HICKS,T. 0. OEI,C. D. NORDSCHOW,A N D B. J. HUGHES From rhe Deparrnienr ojClinica1 Parho1og.v. Indiana University Medical Cenrer, Indianapolis. Indiana
Instagel) for radioimmunoassay of hepatitis associated antigen (HBsAg) and assesses the relative frequency of false positive and false negatives for each of these liquid scintillation counting solutions. By so doing we hoped to ascertain which liquid scintillation counting solution is the most effective for routine estimation of iodine-I25 for HBsAg detection in hospital patients and donors. In addition, this investigation described the possibilities of using liquid scintillation measurement of iodine-I25 as an alternative counting system for HBsAg detection in hospital patients and donors. I f feasible such a procedure would allow the use of an already available automatic counting facility (Beta counter). Automatic gamma counters have become available, but the demand for one in some small hospitals is usually not as great as is the demand, for a liquid scintillator counter. T h e liquid scintillator counter has the added advantage of measuring iodine-I25 and most isotopes of interest in the clinical laboratory, e.g., tritium and carbon-I4 which a r e b e t a emitters. Whereas, t h e g a m m a counter only measures gamma emitters.
Liquid scintillation counting solutions for radioimmunoassay of hepatitis associated antigen (HBsAg) are evaluated and discussed. Data is also presented which indicates that liquid scintillation counting could be carried out by placing cut-off Ausria-125 test tubes in counting vials containing 10 ml of either Brays, Unogel, or Instagel solutions. The data suggest that Brays was the least effective and Unogel and Instagel were the most effective liquid scintillation counting solutions. This is a fundamental consideration if a laboratory is contemplating an initial investment in counting equipment. The acquisition of a liquid scintillation counter provides additional capabilities for the clinical laboratory interested in radioisotope measurements. In addition, if a laboratory already has a liquid scintillation counter (Beta counter), the director of the clinical laboratory may elect not to purchase additional counting equipment (gamma counter) since he already has the added advantage of measuring iodine125 with the Beta counter. The authors propose liquid scintillation ( R I A ) as an alternative counting system for HBsAg detection in hospital patients and donors.
S E V E R A LA L T E R N A T I V E COUNTING have been proposed for the Abbott Ausria-TM-I25 RIA procedure for HBsAg detection. Simon and Langhofer described an alternative method,* in which the bottom of the Ausria-125 test tube is cut off 1.5 crn from the bottom and placed in a liquid scintillation for gamma counting of iodine-125. Another method has been described by Jordan and Spiehler:' This method involves solubilization of the tube-bound radiolabeled complex with glacial acetic acid, transfer of the eluate to a counting vial of choice, and counting by either liquid scintillation or gamma scintillation. Our present investigation systematically evaluates liquid scintillation counting solutions (Brays, Unogel, and METHODS
Received for publication April 22, 1976; accepted September 5 , 1976.
Materials and Methods Samples
Between 1973 and 1974 serum samples were received in the clinical laboratory for routine testing for HBsAg and antibody. Samples were also received from the blood bank for routine screening for both HBsAg and anti-HBs. The samples were tested immediately or stored in the freezer (-20 C ) until ready for testing. Most of the samples were tested within one or two weeks. All samples were tested by three liquid scintillation counting solutions (Brays, Instagel, and Unogel) and
503 Transfusion Sept.-Oct. 1977
Volume 17
Number 5
HICKS ET AL.
t h r e e o t h e r methods: radioimmunoassayIn (gamma counter), crossoverelectrophoresis,g and agar gel immunodiffusion.6 A total of 340 samples were included in this investigation. Antigens Hepatitis B surface antigens used in the testing of specimens were obtained from Abbott Laboratories, South Pasadena, California, and Hyland Laboratories, Costa Mesa, California. An tisera The antisera employed were also obtained from Abbott Scientific Products Division, South Pasadena, California and Hyland Laboratories, Costa Mesa, California. Both commercial sources of antisera gave lines of identity in our test system. Each commercial s o u r c e was used throughout these studies. Agar Gel Immunodijiision Agar gel immunodiffusion (AGD) was performed in plates supplied by Abbott Scientific Products Division, South Pasadena, California. The plates contained agarose gel with six peripheral wells and one center well. For HBsAg and anti-HBs testing the center well was filled with antigen, and the top and bottom peripheral wells were filled with antibody. Positive and negative antigen controls were included in each test system. T h e plates were incubated at room temperature in a moist chamber and scanned for precipitins after 18.48, and 72 hours. Crossoverelectrophoresis Crossoverelectrophoresis (CEP) was performed in plates supplied by Hyland Laboratories, Costa Mesa, California. The procedure for CEP was that recommended by the manufacturer. The electrophoresis was performed at 30 ma for 60 minutes. Plates were examined for precipitins after 60 minutes and 24 hours. Radioimmunoassay The A ~ s r i a - 1 2 5I kits ~ ~ were purchased from Abbott Laboratories, South Pasadena, California. The radioimmunoassay (RIA) technic used was that recommended by the manufacturer. The radioimmunoassay technic (Ausria-125) is a "sandwich" principle using guinea pig '"1 labeled antibody as the second antibody. The serum sample under evaluation (0.1 ml) was incubated with first antibody coated on the tube for approximately 16 hours, and then the second antibody was added. After another 90 minutes at room temperature, the radioactivity o f the bound fraction was counted in a Packard Auto-Gamma
Transfusion Sep1:Ocl. 1977
Scintillation Spectrometer, model 5 130. This gamma counting system consisted of a sodium iodide crystal. The crystal is activated with thallium and produces light photons when subjected to gamma radiation. It is optically coupled to a photomultiplier tube by a "window" of quartz. The scintillation crystal is hermetically sealed in aluminum and further shielded by antimagnetic protector. Additionally, the temperature in the chamber was maintained at approximately 5 C. Liquid Scintillation counting of '''1 was carried out using the following liquid scintillation counting solutions: Brays, Unogel, and Instagel. All determinations were counted in the same manner as before except that the radioactivity of the bound fraction was counted in a Packard Tri Carb Liquid Scintillation Spectrometer, model 3375. This counting unit has five independent channels o f pulse-height analyzers. The pulses o f both photomultipliers are summed to obtain one pulse representing a radioactive disintegration. The dual photomultiplier tube detector is housed in a lighttight shielded detecting chamber. The temperature was maintained at approximately 5 C. In all instances the tubes were counted on the gamma scintillator counter first and then counted on the liquid scintillation spectrometer. All determinations were carried out in duplicate. Seven negative controls and three positive controls were included with each run. The presence or absence of HBsAg was ascertained by relating counts/minute o f the unknown sample to counts/minute of the negative control mean times the factor 2. I . The calculations used to determine the cut off value were those recommended by the m anu factu rer . Liquid Scin t illators The Brays scintillation counting solution was prepared in our laboratory as follows:
Naph t ha1en e P PO POPP Methanol (absolute) Ethylene glycol p- Dioxan e
240gm I6gm 0.8gm 400m I 80ml to make 4 liters T h e liquid scintillation counting solution described above was prepared as recommended by Bray.' However, both Unogel, (Schwarz Mann) and Instagel (Packard Instrument Company, Inc.) were purchased from commercial sources. Both commercial solutions were used throughout these studies. In all instances the Ausria-I25 test tubes were cut off 1.5 cm from the bottom and placed in a counting vial containing 10 ml o f liquid scintillation solution (Brays, Unogel, or Instagel) for counting according to the method described by Simon and Langhofer.s
Volume 17 Number 5
SCINTILLATION COUNTING SOLUTIONS
hospital patients tested, 57 (17 per cent) were positive when assayed by liquid scintillation (Brays and Agar Gel irnrnunodiffusion solution) and 54 (16 per cent) were positive for HBsAg when counted in Unogel liquid scintillaPositive tion counting solution. In addition, 55 (16 per cent) Test Samples No. Positive,% were positive HBsAg when counted in Instagel liquid scintillation counting solution. RIA 340 57 17 Table 1 also indicates the results obtained with 0 Counter the gamma scintillation counter. Of the 340 hosBrays pital patients tested 55 (16%) were positive for RIA 340 54 16 HBsAg in unconcentrated samples and 61 (18%) j3 Counter were positive for HBsAg in concentrated samples. Unogel Additionally, 29 (9 per cent) CEP and 17 (5 per RIA 340 55 16 cent) AGD HBsAg positive samples were dej3 Counter tected by these methods. The data also indicate Instagel that the various liquid scintillation counting soluRIA 340 55 16 ' tions (Brays, Unogel, and Instagel) and gamma y Counter scintillation are comparable for detecting HBsAg Unconcentrated in unconcentrated samples and less comparable if the samples are concentrated and then counted by Ria 340 61 18 gamma scintillation. Concentrating the samples y Counter After increases the sensitivity of RIA with the gamma Concentration counter. Thus, the number of positive samples detected for HBsAg are increased for 16 to 18 per CEP 340 29 9 cent after sample_ concentration. Furthermore, AG D 340 17 5 the data indicate-'that C E P and AGD are much less sensitive than liquid scintillation (Brays, Unogel, and Instagel) for H BsAg detection. Neutralization The liquid scintillation data (Brays, Instagel, The AusriaR-125 Confirmatory Neutralization and Unogel) were compared with the gamma scinkits were purchased from Abbott Laboratories, tillation data in unconcentrated samples. There South Pasadena, California. The neutralization were nine Brays and seven Unogel discrepancies technic was performed according to the recombetween the data obtained with liquid scintillation mendations of the manufacturer. The procedure is counting and the gamma scintillation counting essentially t h e s a m e a s t h e A ~ s r i a - ' ~ . - 1 2 5 (unconcentrated samples); whereas, 8 discreradioimmunoassay technic except that 0.1 ml of pancies were detected between the results obhuman serum containing antibodies to HBsAg was tained with liquid scintillation (Instagel counting incubated for 90 minutes at room temperature solution) and t h e g a m m a counter i n unwith 0.1 ml of the HBsAg positive serum sample. concentrated versus concentrated samples for The interpretations and calculations used to deHBsAg (gamma counter). Additionally, there termine whether a sample was HBsAg positive were six discrepancies between unconcentrated were those recommended by the manufacturer. A versus concentrated samples for HBsAg (gamma sample was considered HBsAg positive when its counter ). reactivity was neutralized by the addition of conThe liquid scintillation data (Brays, Instagel, firmatory antibody and the reduction was 50 per and Unogel) were also compared with the gamma cent or more. scintillation data in concentrated samples. There were 12 discrepancies with Brays counting soluConcentration tion and 1 I discrepancies with Unogel counting soAll serum samples were concentrated accordlution. Only 10 discrepancies were obtained when ing to the procedure recommended by Ashcavai the samples were counted in Instagel and comand Peters' using LyphogelR (Gelman Instrument pared to those obtained in concentrated samples Company, Ann Arbor, Michigan). with the gamma scintillation counter. The frequencies of false negatives and false Results positives are compared in Table 2 for liquid scintillation counting solutions (RIA) against gamma Table I shows the comparative results of scintillation in concentrated samples. The freradioimmunoassay ( p and y scintillation count\ quencies of false negatives are as follows in uning), crossoverelectrophoresis and agar gel imconcentrated samples: Brays 8 (13 per cent), munodiffusion in the detection of HBsAg. Of 340 Table 1.
Comparative Results of Radioimmunoassay
(j3 and y Scintillation Counting), Crossoverelectrophoresis
~
505
~~
~
506
Transfusion
HICKS ET AL. Table 2.
Sep1:Ocl. 1977
Comparative Results for False Negative and False Positive Test for HBAgs Detection False Positive
False Negative No.
%
8
13
1.4
RIA ,!3 Counter Unogel
8
13
0.7
RIA
8
13
0.7
RIA y Counter Unconcentrated
6
10
0.4
RIA
0
0
.4
CEP
32
53
0.0
AG D
45
75
0.0
Test RIA
No.
%
,!3 Counter Brays
j3 Counter lnstagel
y Counter After . Concentration
NOTES: False negatives-total false negativegall H8Ags positive with neutralization test against anti-HEs sera; False positives-total false positives per test/all negatives with neutralization test and RIA test: Total negatives= 2 8 0 ; Total positives by neutralization test = 60
Unogel 8 (13 per cent) and lnstagel 8 (13 per cent). Six (10 per cent) false negatives were obtained by gamma scintillation ( R I A ) in u n concentrated samples. Table 2 also shows that 32 (53 per cent) false negatives were obtained with CEP and 45 (75 per cent) with AGD in unconcentrated samples. All falsely negative tests were confirmed with the neutralization test. Table 2 also shows the frequencies of false positives for the various liquid scintillation counting solutions in unconcentrated samples. They are as follows: Brays, 4 (1.4 per cent), Unogel, 2 (0.7 per cent) and Instagel, 2 (0.7 per cent). Additionally, I (0.4 per cent) false positive was obtained by g a m m a scintillation i n unconcentrated and concentrated samples. No false positives were obtained with CEP and AGD. Similarly all false positive samples were confirmed with the neutralization test.
Discussion
In our exploratory studies we also observed that counting of radiolabeled gamma emitting isotopes in a liquid scintillation counter is an alternative counting method for
the Ausria-I25 RIA kit.".' O u r experience with the liquid scintillation counting method is in agreement with those described by Simon and Langhofer,8 in which the bottom 1.5 cm of the Ausria-125 tube is cut off and placed in a liquid scintillation solution for counting. In general, there was very good agreement between liquid scintillation and gamma scintillation test results for HBsAg in u ncon cen t r a t ed s am pl es. However, t h e r e was very poor agreement between radioimmunoassay (beta and gamma scintillation counting) and crossoverelectrophoresis and agar gel immunodiffusion. For example, only 29 (9 per cent) C E P and 17 ( 5 per cent) A G D HBsAg positive samples were detected by these methods, whereas by liquid scintillation counting Brays solution 57 (17 per cent), Unogel 54 (16 per cent) and lnstagel 55 (16 per cent) were positive for HBsAg. Additionally, 55 (16 per cent) were positive for HBsAg in unconcentrated samples and 61 (18 per cent) were positive in concentrated
Volume 17 Number 5
SCINTILLATION COUNTING SOLUTIONS
s a m p l e s by g a m m a scintillation. T h u s , measurement by either a gamma scintillation o r a liquid scintillation (Brays, Unogel and I n s t a g e l ) s y s t e m c a n b e a d a p t e d for measurement of iodine-I 25. T h e frequencies of false negatives and false positives a r e also used to assess the efficiencies of Brays, Unogel and Instagel. The relative proportion of falsely negative results for the various liquid scintillation counting s o l u t i o n s w e r e s i m i l a r . F o r example, 8 ( I 3 per cent) false negatives were obtained with Brays, Unogel, and Instagel; 6 (10 per cent) false negatives were obtained with gamma scintillation in unconcentrated samples and no falsely negative results were o b s e r v e d with g a m m a scintillation in concentrated samples. Additionally, we observed 4 ( I .4 per cent) falsely positive results with Brays and only 2 (0.7 per cent) falsely positive results with lnstagel and Unogel. One false positive (0.4 per cent) was obtained with gamma scintillation counting in unconcentrated with concentrated samples. However, no false positives were obtained with AGD and C E P . In our studies, lnstagel and Unogel were found to be the most effective liquid scintillation counting solutions for measurement of iodine-125. Only 8 discrepancies were observed. With Brays, 12 discrepancies were observed when compared with results obtained in concentrated samples by gamma scintillation and neutralization. From these findings, it appears that lnstagel and Unogel scintillation counting solutions would be the most useful liquid scintillators for routine estimation of iodine- I25 in the clinical laboratory for HBsAg detection in hospital patients and donors. Several HBsAg positive samples were detected in concentrated samples by gamma scintillation, whereas, in unconcentrated samples six false negatives were obtained by gamma scintillation. This is probably due to the low level of HBsAg within the sera. I n a previous paper we reported that barely reactive samples by AGD and C E P and samples with low c o u n t s ( R I A ) should b e
507
concentrated and retested before the results a r e reported.:' In all instances, specificity analysis was performed prior to concentration utilizing either anti-HBs o r normal guinea pig serum. I n our studies, one of 61 radioimmunoassay positive specimens failed to give positive results after preincubating with either anti-HBs o r normal guinea pig serum. This sample was also positive by liquid scintillation (Brays, lnstagel and Unogel) and gamma scintillation. All other radioimrnunoassay positive s e r a showed clear-cut neutralization results against antiHBs. Our findings confirm data reported by Sgouris that certain human sera contain antibodies which react with guinea pig globulin.' T h e neutralization technic was used as our criterion of specificity since it was the only method we have available in our laboratory. Specificity analyses were carried out prior to serum concentration and after serum concentration utilizing t h e neutralization test. However, in t h e cases where t h e concentrated samples were positive by R I A and negative by CEP, A G D and positive by the neutralization test prior to and after serum concentration we assume that the neutralizing antibodies were due to anti-HBs subtypes d o r y antibodies.5 Liquid scintillation counting solutions such as Brays, Unogel and lnstagel a r e adaptable for routine measurement of iodine-I25 for HBsAg detection in hospital patients and donors. In general, Brays was the least effective liquid scintillator counting solution since four false positives were obtained as compared with one false positive by gamma scintillation. Based on these findings, lnstagel and Unogel were the most effective liquid scintillation counting solutions. In addition to t h e s e c o n s i d e r a t i o n s , liquid scintillation counting could be an economical adv,antage for those laboratories contemplating an initial investment o r working from a small budget. References I.
Ashcavai, M . , and R . L . Peters: Hepatitis associated antigen: Improved sensitivity in detection. Am. J. Clin Patho1.55: 262, 1971.
HICKS ET AL 2.
3.
4.
5. 6.
7. 8.
9.
Bray, G. A,: A simple efficient liquid scintillator for counting aqueous solutions in a liquid scintillation counter. Anal. Biochem. 1:279, 1960. Hicks, E. J., and B. J. Hughes: Comparative sensitivities o f radioimmunoassay, crossoverelectrophoresis and agar gel immunodiffusion for HBAg detection. Am. J. Clin. Pathol. 65: 540, 1976. Jordan, C. W., V. Spiehler, R. Haendiges, and E. Z. Helman: Evaluation of alternative counting methods for radioimmunoassay of hepatitisassociated antigen (HBAg). Clin. Chem. 20: 733, 1974. Le Bouvier, G. L.: The heterogeneity of australia antigen. J . Infect. Dis. 123:671, 1971. Prince, A. M. An antigen detected in the blood during the incubation period of serum hepatitis. Proc. Nat. Acad. Sci. 60: 814, 1968. Sgouris, J . T.: Limitations of t h e radioimmunoassay for hepatitis B antigen. N. Engl. J. Med. 288: 160, 1973. Simon, R. G., L. A. Langhofer, Jr., and E. J . Hendricks: Australia antigen content o f commercial quality-control sera. Clin. Chem. 19: 221, 1973. Tripodi, D., J. Hawk, D. J. Gocke, A. Redeker, and N. A. Starkovsky: Detection o f antibody to
Transfusion Sep1:Ocl. 1977
hepatitis-associated antigen by indirect counterimmunoelectrophoresis. Am. J . Clin. Pathol. 59549, 1973. 10. Walsh, J. H., R. Yalow, and S. A. Berson: Detection of australia antigen and antibody by means of radioimmunoassay techniques. J. Infec. Dis. 121: 550. 1970.
Edward J. Hicks, Ph.D., Associate Professor, Department of Clinical Pathology, I100 West Michigan Street, Indiana University Medical Center, Indianapolis, Indiana 46202. Tjien 0. Oei, M.D., Associate Professor, Department of Clinical Pathology, I100 West Michigan Street, Indiana University Medical Center, Indianapolis, Indiana 46202. Carleton D. Nordschow, M.D., Ph.D., Professor and Chairman of the Department ofClinical Pathology, I100 West Michigan Street, Indiana University Medical Center, Indianapolis, Indiana 46202. Blossie J . Hughes, M.S., M.T. (ASCP), Surburan Hospital, Instructor, 8600 Old Georgetown Road, Bethesda, Maryland.
Instagel) for radioimmunoassay of hepatitis associated antigen (HBsAg) and assesses the relative frequency of false positive and false negatives for each of these liquid scintillation counting solutions. By so doing we hoped to ascertain which liquid scintillation counting solution is the most effective for routine estimation of iodine-I25 for HBsAg detection in hospital patients and donors. In addition, this investigation described the possibilities of using liquid scintillation measurement of iodine-I25 as an alternative counting system for HBsAg detection in hospital patients and donors. I f feasible such a procedure would allow the use of an already available automatic counting facility (Beta counter). Automatic gamma counters have become available, but the demand for one in some small hospitals is usually not as great as is the demand, for a liquid scintillator counter. T h e liquid scintillator counter has the added advantage of measuring iodine-I25 and most isotopes of interest in the clinical laboratory, e.g., tritium and carbon-I4 which a r e b e t a emitters. Whereas, t h e g a m m a counter only measures gamma emitters.
Liquid scintillation counting solutions for radioimmunoassay of hepatitis associated antigen (HBsAg) are evaluated and discussed. Data is also presented which indicates that liquid scintillation counting could be carried out by placing cut-off Ausria-125 test tubes in counting vials containing 10 ml of either Brays, Unogel, or Instagel solutions. The data suggest that Brays was the least effective and Unogel and Instagel were the most effective liquid scintillation counting solutions. This is a fundamental consideration if a laboratory is contemplating an initial investment in counting equipment. The acquisition of a liquid scintillation counter provides additional capabilities for the clinical laboratory interested in radioisotope measurements. In addition, if a laboratory already has a liquid scintillation counter (Beta counter), the director of the clinical laboratory may elect not to purchase additional counting equipment (gamma counter) since he already has the added advantage of measuring iodine125 with the Beta counter. The authors propose liquid scintillation ( R I A ) as an alternative counting system for HBsAg detection in hospital patients and donors.
S E V E R A LA L T E R N A T I V E COUNTING have been proposed for the Abbott Ausria-TM-I25 RIA procedure for HBsAg detection. Simon and Langhofer described an alternative method,* in which the bottom of the Ausria-125 test tube is cut off 1.5 crn from the bottom and placed in a liquid scintillation for gamma counting of iodine-125. Another method has been described by Jordan and Spiehler:' This method involves solubilization of the tube-bound radiolabeled complex with glacial acetic acid, transfer of the eluate to a counting vial of choice, and counting by either liquid scintillation or gamma scintillation. Our present investigation systematically evaluates liquid scintillation counting solutions (Brays, Unogel, and METHODS
Received for publication April 22, 1976; accepted September 5 , 1976.
Materials and Methods Samples
Between 1973 and 1974 serum samples were received in the clinical laboratory for routine testing for HBsAg and antibody. Samples were also received from the blood bank for routine screening for both HBsAg and anti-HBs. The samples were tested immediately or stored in the freezer (-20 C ) until ready for testing. Most of the samples were tested within one or two weeks. All samples were tested by three liquid scintillation counting solutions (Brays, Instagel, and Unogel) and
503 Transfusion Sept.-Oct. 1977
Volume 17
Number 5
HICKS ET AL.
t h r e e o t h e r methods: radioimmunoassayIn (gamma counter), crossoverelectrophoresis,g and agar gel immunodiffusion.6 A total of 340 samples were included in this investigation. Antigens Hepatitis B surface antigens used in the testing of specimens were obtained from Abbott Laboratories, South Pasadena, California, and Hyland Laboratories, Costa Mesa, California. An tisera The antisera employed were also obtained from Abbott Scientific Products Division, South Pasadena, California and Hyland Laboratories, Costa Mesa, California. Both commercial sources of antisera gave lines of identity in our test system. Each commercial s o u r c e was used throughout these studies. Agar Gel Immunodijiision Agar gel immunodiffusion (AGD) was performed in plates supplied by Abbott Scientific Products Division, South Pasadena, California. The plates contained agarose gel with six peripheral wells and one center well. For HBsAg and anti-HBs testing the center well was filled with antigen, and the top and bottom peripheral wells were filled with antibody. Positive and negative antigen controls were included in each test system. T h e plates were incubated at room temperature in a moist chamber and scanned for precipitins after 18.48, and 72 hours. Crossoverelectrophoresis Crossoverelectrophoresis (CEP) was performed in plates supplied by Hyland Laboratories, Costa Mesa, California. The procedure for CEP was that recommended by the manufacturer. The electrophoresis was performed at 30 ma for 60 minutes. Plates were examined for precipitins after 60 minutes and 24 hours. Radioimmunoassay The A ~ s r i a - 1 2 5I kits ~ ~ were purchased from Abbott Laboratories, South Pasadena, California. The radioimmunoassay (RIA) technic used was that recommended by the manufacturer. The radioimmunoassay technic (Ausria-125) is a "sandwich" principle using guinea pig '"1 labeled antibody as the second antibody. The serum sample under evaluation (0.1 ml) was incubated with first antibody coated on the tube for approximately 16 hours, and then the second antibody was added. After another 90 minutes at room temperature, the radioactivity o f the bound fraction was counted in a Packard Auto-Gamma
Transfusion Sep1:Ocl. 1977
Scintillation Spectrometer, model 5 130. This gamma counting system consisted of a sodium iodide crystal. The crystal is activated with thallium and produces light photons when subjected to gamma radiation. It is optically coupled to a photomultiplier tube by a "window" of quartz. The scintillation crystal is hermetically sealed in aluminum and further shielded by antimagnetic protector. Additionally, the temperature in the chamber was maintained at approximately 5 C. Liquid Scintillation counting of '''1 was carried out using the following liquid scintillation counting solutions: Brays, Unogel, and Instagel. All determinations were counted in the same manner as before except that the radioactivity of the bound fraction was counted in a Packard Tri Carb Liquid Scintillation Spectrometer, model 3375. This counting unit has five independent channels o f pulse-height analyzers. The pulses o f both photomultipliers are summed to obtain one pulse representing a radioactive disintegration. The dual photomultiplier tube detector is housed in a lighttight shielded detecting chamber. The temperature was maintained at approximately 5 C. In all instances the tubes were counted on the gamma scintillator counter first and then counted on the liquid scintillation spectrometer. All determinations were carried out in duplicate. Seven negative controls and three positive controls were included with each run. The presence or absence of HBsAg was ascertained by relating counts/minute o f the unknown sample to counts/minute of the negative control mean times the factor 2. I . The calculations used to determine the cut off value were those recommended by the m anu factu rer . Liquid Scin t illators The Brays scintillation counting solution was prepared in our laboratory as follows:
Naph t ha1en e P PO POPP Methanol (absolute) Ethylene glycol p- Dioxan e
240gm I6gm 0.8gm 400m I 80ml to make 4 liters T h e liquid scintillation counting solution described above was prepared as recommended by Bray.' However, both Unogel, (Schwarz Mann) and Instagel (Packard Instrument Company, Inc.) were purchased from commercial sources. Both commercial solutions were used throughout these studies. In all instances the Ausria-I25 test tubes were cut off 1.5 cm from the bottom and placed in a counting vial containing 10 ml o f liquid scintillation solution (Brays, Unogel, or Instagel) for counting according to the method described by Simon and Langhofer.s
Volume 17 Number 5
SCINTILLATION COUNTING SOLUTIONS
hospital patients tested, 57 (17 per cent) were positive when assayed by liquid scintillation (Brays and Agar Gel irnrnunodiffusion solution) and 54 (16 per cent) were positive for HBsAg when counted in Unogel liquid scintillaPositive tion counting solution. In addition, 55 (16 per cent) Test Samples No. Positive,% were positive HBsAg when counted in Instagel liquid scintillation counting solution. RIA 340 57 17 Table 1 also indicates the results obtained with 0 Counter the gamma scintillation counter. Of the 340 hosBrays pital patients tested 55 (16%) were positive for RIA 340 54 16 HBsAg in unconcentrated samples and 61 (18%) j3 Counter were positive for HBsAg in concentrated samples. Unogel Additionally, 29 (9 per cent) CEP and 17 (5 per RIA 340 55 16 cent) AGD HBsAg positive samples were dej3 Counter tected by these methods. The data also indicate Instagel that the various liquid scintillation counting soluRIA 340 55 16 ' tions (Brays, Unogel, and Instagel) and gamma y Counter scintillation are comparable for detecting HBsAg Unconcentrated in unconcentrated samples and less comparable if the samples are concentrated and then counted by Ria 340 61 18 gamma scintillation. Concentrating the samples y Counter After increases the sensitivity of RIA with the gamma Concentration counter. Thus, the number of positive samples detected for HBsAg are increased for 16 to 18 per CEP 340 29 9 cent after sample_ concentration. Furthermore, AG D 340 17 5 the data indicate-'that C E P and AGD are much less sensitive than liquid scintillation (Brays, Unogel, and Instagel) for H BsAg detection. Neutralization The liquid scintillation data (Brays, Instagel, The AusriaR-125 Confirmatory Neutralization and Unogel) were compared with the gamma scinkits were purchased from Abbott Laboratories, tillation data in unconcentrated samples. There South Pasadena, California. The neutralization were nine Brays and seven Unogel discrepancies technic was performed according to the recombetween the data obtained with liquid scintillation mendations of the manufacturer. The procedure is counting and the gamma scintillation counting essentially t h e s a m e a s t h e A ~ s r i a - ' ~ . - 1 2 5 (unconcentrated samples); whereas, 8 discreradioimmunoassay technic except that 0.1 ml of pancies were detected between the results obhuman serum containing antibodies to HBsAg was tained with liquid scintillation (Instagel counting incubated for 90 minutes at room temperature solution) and t h e g a m m a counter i n unwith 0.1 ml of the HBsAg positive serum sample. concentrated versus concentrated samples for The interpretations and calculations used to deHBsAg (gamma counter). Additionally, there termine whether a sample was HBsAg positive were six discrepancies between unconcentrated were those recommended by the manufacturer. A versus concentrated samples for HBsAg (gamma sample was considered HBsAg positive when its counter ). reactivity was neutralized by the addition of conThe liquid scintillation data (Brays, Instagel, firmatory antibody and the reduction was 50 per and Unogel) were also compared with the gamma cent or more. scintillation data in concentrated samples. There were 12 discrepancies with Brays counting soluConcentration tion and 1 I discrepancies with Unogel counting soAll serum samples were concentrated accordlution. Only 10 discrepancies were obtained when ing to the procedure recommended by Ashcavai the samples were counted in Instagel and comand Peters' using LyphogelR (Gelman Instrument pared to those obtained in concentrated samples Company, Ann Arbor, Michigan). with the gamma scintillation counter. The frequencies of false negatives and false Results positives are compared in Table 2 for liquid scintillation counting solutions (RIA) against gamma Table I shows the comparative results of scintillation in concentrated samples. The freradioimmunoassay ( p and y scintillation count\ quencies of false negatives are as follows in uning), crossoverelectrophoresis and agar gel imconcentrated samples: Brays 8 (13 per cent), munodiffusion in the detection of HBsAg. Of 340 Table 1.
Comparative Results of Radioimmunoassay
(j3 and y Scintillation Counting), Crossoverelectrophoresis
~
505
~~
~
506
Transfusion
HICKS ET AL. Table 2.
Sep1:Ocl. 1977
Comparative Results for False Negative and False Positive Test for HBAgs Detection False Positive
False Negative No.
%
8
13
1.4
RIA ,!3 Counter Unogel
8
13
0.7
RIA
8
13
0.7
RIA y Counter Unconcentrated
6
10
0.4
RIA
0
0
.4
CEP
32
53
0.0
AG D
45
75
0.0
Test RIA
No.
%
,!3 Counter Brays
j3 Counter lnstagel
y Counter After . Concentration
NOTES: False negatives-total false negativegall H8Ags positive with neutralization test against anti-HEs sera; False positives-total false positives per test/all negatives with neutralization test and RIA test: Total negatives= 2 8 0 ; Total positives by neutralization test = 60
Unogel 8 (13 per cent) and lnstagel 8 (13 per cent). Six (10 per cent) false negatives were obtained by gamma scintillation ( R I A ) in u n concentrated samples. Table 2 also shows that 32 (53 per cent) false negatives were obtained with CEP and 45 (75 per cent) with AGD in unconcentrated samples. All falsely negative tests were confirmed with the neutralization test. Table 2 also shows the frequencies of false positives for the various liquid scintillation counting solutions in unconcentrated samples. They are as follows: Brays, 4 (1.4 per cent), Unogel, 2 (0.7 per cent) and Instagel, 2 (0.7 per cent). Additionally, I (0.4 per cent) false positive was obtained by g a m m a scintillation i n unconcentrated and concentrated samples. No false positives were obtained with CEP and AGD. Similarly all false positive samples were confirmed with the neutralization test.
Discussion
In our exploratory studies we also observed that counting of radiolabeled gamma emitting isotopes in a liquid scintillation counter is an alternative counting method for
the Ausria-I25 RIA kit.".' O u r experience with the liquid scintillation counting method is in agreement with those described by Simon and Langhofer,8 in which the bottom 1.5 cm of the Ausria-125 tube is cut off and placed in a liquid scintillation solution for counting. In general, there was very good agreement between liquid scintillation and gamma scintillation test results for HBsAg in u ncon cen t r a t ed s am pl es. However, t h e r e was very poor agreement between radioimmunoassay (beta and gamma scintillation counting) and crossoverelectrophoresis and agar gel immunodiffusion. For example, only 29 (9 per cent) C E P and 17 ( 5 per cent) A G D HBsAg positive samples were detected by these methods, whereas by liquid scintillation counting Brays solution 57 (17 per cent), Unogel 54 (16 per cent) and lnstagel 55 (16 per cent) were positive for HBsAg. Additionally, 55 (16 per cent) were positive for HBsAg in unconcentrated samples and 61 (18 per cent) were positive in concentrated
Volume 17 Number 5
SCINTILLATION COUNTING SOLUTIONS
s a m p l e s by g a m m a scintillation. T h u s , measurement by either a gamma scintillation o r a liquid scintillation (Brays, Unogel and I n s t a g e l ) s y s t e m c a n b e a d a p t e d for measurement of iodine-I 25. T h e frequencies of false negatives and false positives a r e also used to assess the efficiencies of Brays, Unogel and Instagel. The relative proportion of falsely negative results for the various liquid scintillation counting s o l u t i o n s w e r e s i m i l a r . F o r example, 8 ( I 3 per cent) false negatives were obtained with Brays, Unogel, and Instagel; 6 (10 per cent) false negatives were obtained with gamma scintillation in unconcentrated samples and no falsely negative results were o b s e r v e d with g a m m a scintillation in concentrated samples. Additionally, we observed 4 ( I .4 per cent) falsely positive results with Brays and only 2 (0.7 per cent) falsely positive results with lnstagel and Unogel. One false positive (0.4 per cent) was obtained with gamma scintillation counting in unconcentrated with concentrated samples. However, no false positives were obtained with AGD and C E P . In our studies, lnstagel and Unogel were found to be the most effective liquid scintillation counting solutions for measurement of iodine-125. Only 8 discrepancies were observed. With Brays, 12 discrepancies were observed when compared with results obtained in concentrated samples by gamma scintillation and neutralization. From these findings, it appears that lnstagel and Unogel scintillation counting solutions would be the most useful liquid scintillators for routine estimation of iodine- I25 in the clinical laboratory for HBsAg detection in hospital patients and donors. Several HBsAg positive samples were detected in concentrated samples by gamma scintillation, whereas, in unconcentrated samples six false negatives were obtained by gamma scintillation. This is probably due to the low level of HBsAg within the sera. I n a previous paper we reported that barely reactive samples by AGD and C E P and samples with low c o u n t s ( R I A ) should b e
507
concentrated and retested before the results a r e reported.:' In all instances, specificity analysis was performed prior to concentration utilizing either anti-HBs o r normal guinea pig serum. I n our studies, one of 61 radioimmunoassay positive specimens failed to give positive results after preincubating with either anti-HBs o r normal guinea pig serum. This sample was also positive by liquid scintillation (Brays, lnstagel and Unogel) and gamma scintillation. All other radioimrnunoassay positive s e r a showed clear-cut neutralization results against antiHBs. Our findings confirm data reported by Sgouris that certain human sera contain antibodies which react with guinea pig globulin.' T h e neutralization technic was used as our criterion of specificity since it was the only method we have available in our laboratory. Specificity analyses were carried out prior to serum concentration and after serum concentration utilizing t h e neutralization test. However, in t h e cases where t h e concentrated samples were positive by R I A and negative by CEP, A G D and positive by the neutralization test prior to and after serum concentration we assume that the neutralizing antibodies were due to anti-HBs subtypes d o r y antibodies.5 Liquid scintillation counting solutions such as Brays, Unogel and lnstagel a r e adaptable for routine measurement of iodine-I25 for HBsAg detection in hospital patients and donors. In general, Brays was the least effective liquid scintillator counting solution since four false positives were obtained as compared with one false positive by gamma scintillation. Based on these findings, lnstagel and Unogel were the most effective liquid scintillation counting solutions. In addition to t h e s e c o n s i d e r a t i o n s , liquid scintillation counting could be an economical adv,antage for those laboratories contemplating an initial investment o r working from a small budget. References I.
Ashcavai, M . , and R . L . Peters: Hepatitis associated antigen: Improved sensitivity in detection. Am. J. Clin Patho1.55: 262, 1971.
HICKS ET AL 2.
3.
4.
5. 6.
7. 8.
9.
Bray, G. A,: A simple efficient liquid scintillator for counting aqueous solutions in a liquid scintillation counter. Anal. Biochem. 1:279, 1960. Hicks, E. J., and B. J. Hughes: Comparative sensitivities o f radioimmunoassay, crossoverelectrophoresis and agar gel immunodiffusion for HBAg detection. Am. J. Clin. Pathol. 65: 540, 1976. Jordan, C. W., V. Spiehler, R. Haendiges, and E. Z. Helman: Evaluation of alternative counting methods for radioimmunoassay of hepatitisassociated antigen (HBAg). Clin. Chem. 20: 733, 1974. Le Bouvier, G. L.: The heterogeneity of australia antigen. J . Infect. Dis. 123:671, 1971. Prince, A. M. An antigen detected in the blood during the incubation period of serum hepatitis. Proc. Nat. Acad. Sci. 60: 814, 1968. Sgouris, J . T.: Limitations of t h e radioimmunoassay for hepatitis B antigen. N. Engl. J. Med. 288: 160, 1973. Simon, R. G., L. A. Langhofer, Jr., and E. J . Hendricks: Australia antigen content o f commercial quality-control sera. Clin. Chem. 19: 221, 1973. Tripodi, D., J. Hawk, D. J. Gocke, A. Redeker, and N. A. Starkovsky: Detection o f antibody to
Transfusion Sep1:Ocl. 1977
hepatitis-associated antigen by indirect counterimmunoelectrophoresis. Am. J . Clin. Pathol. 59549, 1973. 10. Walsh, J. H., R. Yalow, and S. A. Berson: Detection of australia antigen and antibody by means of radioimmunoassay techniques. J. Infec. Dis. 121: 550. 1970.
Edward J. Hicks, Ph.D., Associate Professor, Department of Clinical Pathology, I100 West Michigan Street, Indiana University Medical Center, Indianapolis, Indiana 46202. Tjien 0. Oei, M.D., Associate Professor, Department of Clinical Pathology, I100 West Michigan Street, Indiana University Medical Center, Indianapolis, Indiana 46202. Carleton D. Nordschow, M.D., Ph.D., Professor and Chairman of the Department ofClinical Pathology, I100 West Michigan Street, Indiana University Medical Center, Indianapolis, Indiana 46202. Blossie J . Hughes, M.S., M.T. (ASCP), Surburan Hospital, Instructor, 8600 Old Georgetown Road, Bethesda, Maryland.
