Indian J Med Res 126, July 2007, pp 34-38
Antinuclear antibodies by indirect immunofluorescence : Optimum screening dilution for diagnosis of systemic lupus erythematosus P. Ghosh, S. Dwivedi, Sita Naik, Vikas Agarwal, Anupam Verma*, Amita Aggarwal & Ramnath Misra
Departments of Immunology & *Transfusion Medicine, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India
Received October 9, 2006
Background & objectives: Antinuclear antibodies (ANA) are serological hallmark of systemic lupus erythematosus (SLE). Conventionally, the test is carried out on human epithelial cells (HEp2) by indirect immunofluorescence (IIF) technique. Since culturing and maintaining HEp2 cells in the laboratory are labour intensive, in-house assays have given way to kits manufactured by commercial companies. The reference screening dilutions provided by the manufacturers are based on different ethnic population than ours. Therefore, it becomes mandatory for every laboratory to have its own screening dilutions for the local population that distinguishes best between healthy and diseased state. As, there is paucity of such data, we aimed to define the optimum screening dilution that distinguishes the patient with SLE from healthy individuals. Methods: Sera of patients fulfilling ACR criteria for diagnosis of SLE, idiopathic inflammatory polymyositis/dermatomyositis (PM/DM) and rheumatoid arthritis (RA), and age and sex matched healthy individuals were tested for ANA by IIF using a commercial kit (Euroimmun, Germany) at 5 dilutions, namely 1:40, 1:80, 1:160, 1:320 and 1:640. Receiver operator characteristics (ROC) curve were constructed to define the optimum dilution that distinguished healthy sera from the diseased ones. Results: Test was performed on 213 sera from 94 healthy individuals, and 43 SLE, 37 RA and 39 DM/PM patients. In healthy individuals, ANA at dilutions 1:40, 1:80, 1:160, 1:320 and 1:640 was positive in 13.8, 4.3, 2.1, 2.1 and 0 per cent respectively, whereas in SLE it was positive in 95.3, 95.3, 65.1, 53.5 and 23.3 per cent respectively. Interpretation & conclusion: ROC curves analysis showed that at 1:40 dilution, sera of 95.3 per cent of SLE and 13.8 per cent of normal individuals were (ANA) positive, whereas at 1:80 dilution it was 95.3 per cent for SLE and 4.3 per cent for healthy individuals. A fluorescent intensity of ³2 was more specific for SLE. The best discrimination between healthy individuals and the SLE patients was found at screening dilution of 1:80 and fluorescent intensity of ³2 in our laboratory. Key words Antinuclear antibodies - indirect immunofluorescence - screening dilutions - systemic lupus erythematosus
Antinuclear antibodies (ANA) are the serological hallmark of systemic lupus erythematosus (SLE), originally described in 1957 using an
immunofluorescent assay with rodent liver tissue as the substrate 1. With the increasing awareness among physicians of autoimmune connective tissue diseases, 34
GHOSH et al : ANA SCREENING DILUTION
this test is frequently used by clinicians since it is the mainstay of diagnosis of SLE. Since its original description, the technique has undergone considerable methodological modifications and the rodent tissue substrate has been replaced with human epithelial cell lines such as the human epithelial (HEp2) cells. This has the advantages of increased sensitivity, easier identification of different patterns of reactivity and recognition of antibody specificities that react with antigens associated with dividing cells2. The culture and maintenance of cell lines in the laboratory is labour intensive and requires considerable expertise. Hence, ‘in house’ assays have been replaced by commercial kits. These kits contain HEp2 coated slides, antibody conjugates and reference sera. There have also been recent advances in the technology of coating HEp2 cells onto glass slides that is being adopted by some manufacturers and this adds to the variability in the sensitivity of different kits. Any increase in sensitivity usually compromises specificity and this is particularly relevant in the case of ANA detection since the test is also positive in a large variety of connective tissue diseases besides SLE, such as mixed connective tissue disorder, primary Sjogren’s syndrome, rheumatoid arthritis and in malignancies and infectious diseases3. ANA positivity is also seen in normal individuals, the frequency increasing with age4. Currently all the kits being used in Indian laboratories are imported. The standardization of these kits has been done using normal and diseased sera of non Indian origin. Since both genetic and environmental factors are known to influence autoantibody production5, it may be expected that the prevalence in healthy individuals will be different in our country. This is especially important in view of the high infection load in our population. Hence, it is relevant that the imported kits, which are being used for diagnostic tests in our population, should be validated using serum samples from local population with a view to define the screening dilutions. We therefore carried out this study to define the dilution to be used for distinguishing the patient with SLE from normal individuals and establish the frequency of ANA positivity in well defined patients with other connective tissue diseases. Material & Methods The study was conducted at the department of Immunology, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, between July 2004 to December 2005. Healthy volunteers from among the staff of the hospital and consecutive patients who fulfilled
35
American College of Rheumatology (ACR) criteria for diagnosis of SLE6 were included in the study. Healthy volunteers between the ages of 18 and 65 yr, and who gave no history of any illness and had no physical or mental disability were selected from among the staff of the hospital and members of their family. Five ml of blood was collected; and serum separated within 4 h and stored at -20°C till use. Stored sera (-80°C) from patients with dermatomyositis/polymyositis (DM/PM)7 and rheumatoid arthritis (RA)8 who fulfilled the diagnostic criteria of each of these diseases were also tested. The study was not submitted to the institutional ethics committee for the approval because this was a validation exercise undertaken to improve the quality of the test. However, a written informed consent was undertaken from all healthy volunteers and SLE patients who participated in the study. All the kits (Euroimmun, Germany) were purchased and tests were carried out without any knowledge or obligation to the manufacturer. The study was performed with the same batch of the kit from the manufacturer and as per the instructions provided. The kit contained HEp2 slides, sample diluents, antibody conjugate, control sera and wash buffer. The assay was performed according to manufacturer’s instruction. Briefly, 25 ml of diluted sera (1:40) were layered on HEp-2 cell spots and incubated for 30 min. The slide was washed by flush of phosphate buffer saline (PBS)–Tween and then immersed in PBS for 5 min. Later 20 ml of fluorescein-labeled (goat) antihuman globulin was added to each spot and slide incubated for another 30 min. Again the slide was washed in PBS-Tween for 5 min as described earlier, embedded with 10 ml of embedding medium and read under the fluorescent microscope (Olympus; Japan). Two independent observers (SN and VA) experienced in reading ANA slides, blinded to the diagnosis and the interpretation of the other observer read the slides. Prior to the actual test both the investigators had participated in several sessions of reading ANA slides to agree on a consensus. ANA was defined as positive at a particular dilution if it was read positive by both the observers at that dilution. Both positive and negative control sera were provided by the manufacturer. All sera that were positive at the 1:40 dilution were tested at 1.80, 1:160, 1:320 and 1:640 dilutions. The screening dilution recommended by the manufacturer was 1:100. The criterion for assigning intensity of fluorescence was as follows: 4+ (very bright green), 3+ (bright green), 2 + (green), 1 + (faint green).
36
INDIAN J MED RES, JULY 2007
Statistical analysis : All statistical analyses were carried out on SPSS software (11.5 version). Interclass correlation coefficient was used to look for inter-observer variability. The chi-square test of significance was used to find out any difference in ANA positivity in different age groups of healthy controls. The receiver operator characteristic (ROC) analysis was used to determine the cut-off level that differentiates between healthy and diseased individuals. While generating ROC curves larger test results were taken to indicate more positive test and standard error of area with distribution assumption as nonparametric and confidence levels of 95 per cent were used. ROC analysis was also done to assess which definition of the fluorescent positivity (³1 or ³2) gives better cut-off to differentiate healthy individuals from patients with SLE. Results Test for ANA was performed on 213 serum samples collected from 94 healthy controls (59 females, median age 35 yr, range 18-61), 43 SLE (35 females, median age 33 yr, range 23-47), 37 RA (26 females, median age 35 yr, range 25-65) and 39 DM/PM patients (27 females, median age 44 yr, range 35-64). The age distribution of healthy controls (Table I) shows that the number of subjects in the 56-65 yr group was less than in other age groups. Variability : Inter-observer variability gave an interclass coefficient of 0.938, [95% confidence interval (95% CI) 0.92-0.95] at 1:40 dilution, 0.964 (95%CI, 0.953- 0.972) at 1:80 dilution and 0.905 (95%CI, 0.821- 0.95) at 1:160 dilution. Thus, the inter-observer variability was low at all the three dilutions. ANA in healthy subjects : There was no difference in ANA positivity among different age groups (Table I). Overall, ANA was positive in 13.8, 4.3, 2.1, 2.1 and 0 per cent of healthy controls at dilutions 1:40, 1:80, 1:160, 1:320 and 1:640 respectively.
Table I. ANA positivity in sera from healthy controls in different age groups Age groups (yr) 18-25 26-35 36-45 46-55 56-65 Total
No. of indivi- 1:40 dual 19 33 19 18 5 94
2 4 4 2 1 13
Dilutions 1:80 1:160 0 1 2 0 1 4
0 1 1 0 0 2
1:320
1:640
0 1 0 0 0 1
0 0 0 0 0 0
Fig. 1. ANA positivity at different dilutions in patients with SLE, DM/PM and RA.
ANA in systemic autoimmune disorders :The percentage of sera positive for ANA in systemic autoimmune diseases is shown in Fig. 1. Very few patients of RA had ANA positivity at 1:320 or 1:640 in contrast to SLE and DM/PM. Determining the dilution that distinguishes healthy individuals from SLE patients : The ROC curve (Fig. 2) showed that at a dilution of 1:40, 95.3 per cent of SLE patients and 13.8 per cent of healthy individuals were ANA positive. At a dilution of 1:80, 95.3 per cent of the SLE patients and 4.3 per cent of healthy individuals were ANA positive (Table II). This indicated that 1:80 dilution was the best screening dilution to distinguish SLE patients from healthy individuals. ROC analysis to determine whether fluorescent intensity of ³1 or ³2 is a better cut-off for ANA positivity : At a dilution 1:80 the sensitivity of ANA for diagnosis of SLE was no different between fluorescence intensity of ³1 or ³2 but there was better specificity at a fluorescence intensity of ³2 (95.7% vs 90.4%) as compared to intensity of ³1 (Fig. 2, Table II). Discussion This is perhaps the first study from our country to determine the optimum screening dilution for ANA using a commercial kit. This validation has important clinical implications as ANA is being widely used due to ease of availability as well as increasing awareness among medical fraternity. Our observation of 1:80 being the best screening dilution for diagnosis of SLE is in contrast to the
GHOSH et al : ANA SCREENING DILUTION
37
(a)
(b)
Fig. 2. ROC curves of ANA for diagnosis of SLE (a) taking intensity of fluorescent (IF) > 1, and (b) taking intensity of fluorescent > 2 as positive.( ) 1:40 positive, ( ) 1:80 positive, ( ) 1:160 positive, ( ) 1:320 positive, ( ) 1:640 positive). Table II. Sensitivity and specificity taking intensity of fluorescence (IIF)³1 and ³2 as positive Disease SLE
DM/PM
RA
Dilution 1:40 1:80 1:160 1:320 1:640 1:40 1:80 1:160 1:320 1:640 1:40 1:80 1:160 1:320 1:640
Sensitivity (%) IIF intensity³I
Specificity (%)
IIF intensity³2
95.3 95.3 88.4 69.8 53.5 51.3 48.7 46.2 38.5 23.1 45.9 13.5 10.8 8.1 2.7
93 95.3 65.1 53.5 23.3 51.3 46.2 41 30.8 17.9 43.2 5.4 8.1 8.1 2.7
IIF intensity³I 86.2 90.4 93.6 95.7 100 86.2 90.4 93.6 95.7 100 86.2 90.4 93.6 95.7 100
IIF intensity³2 89.4 95.7 97.9 97.9 100 89.4 95.7 97.9 97.9 100 89.4 95.7 97.9 97.9 100
SLE - Systemic lupus erythematosus DM/PM – Dermatomyositis/polymyositis RA – Rheumatoid arthritis
prevalent practice of using sera at 1:40 dilution. This is reflected in a survey9 by the college of American Pathologists which, found that 59.6 per cent of the laboratory use ³1:40 as a cut-off, 23.1 per cent use ³1:80, 14 per cent use ³1:160 and rest use other cutoffs. Tan et al10 have shown that 1:160 was the best dilution which distinguished SLE patients from healthy
individuals. The percentage of ANA positivity in our study at 1:40 was lower than that reported by Tan et al where 31.7 per cent of healthy persons were ANA positive. The lower percentage in our study could be explained by difference in ethnicity and environmental factors. However, our percentage was more that those reported from Middle East countries where it is reported
38
INDIAN J MED RES, JULY 2007
to vary between 4.2 to 7.6 per cent11,12. Besides ethnic factors, the variability could be related to inclusion of only male subjects in one study and variability in substrate in another study. Earlier studies using rodent tissues as substrate have reported a lower percentage of ANA (1-3%) in normal population from Malaysia and New Zealand13,14. Moreover, we did not find any difference in the ANA positivity among different age groups. However, the number of individuals in each age group was very small to reach to a definite conclusion. Since we found lower percentage positivity in our healthy controls which could be due to lower sensitivity of our assay, we analyzed its positivity in well defined disease groups. We found that the percentage in SLE, RA and PM/DM was compatible to that reported previously3. However, in contrast to SLE the sera from RA patients had low titre positivity. In clinical practice sometimes it is difficult to distinguish RA from SLE which could be useful in clinical practice and ANA is requisitioned. Our study suggested that 1:80 dilution could be used to distinguish SLE from RA. The intensity of fluorescence, which reflects the avidity and concentration of antibodies, may be a source of inter-observer variability. Our data suggest that using a cut-off of fluorescence intensity ³2 improves specificity for diagnosis of SLE. This is related to subjective nature of reading slides where negative and 1+ may be difficult to distinguish. At the IIF intensity of ³2 the sensitivity recorded in the SLE group was 95.3 at 1:80 dilution as against sensitivity of 93 at 1:40 dilution. The possible explanation could be the prozone phenomenon in these two sera15. Efforts are on to quantitate the signal using ELISA format or to have mixture of recombinant antigen based ELISAs. Arrays using purified antigens also hold promise for future16. In conclusion, we have addressed an important validation issue of diagnostic kits meant for ANA. In contrast to the kit’s recommendation of using 1:100 dilution, our data suggested that the use of screening dilution of 1:80 would be economical yet valid in distinguishing SLE from normal individuals. All laboratories using imported kits need to do validation exercise to determine the screening dilution.
References 1.
Holbrow EJ, Weir DM, Jhonson GD. A serum factor in lupus erythematosus with affinity for tissue nuclei. Br Med J 1957; 2 : 732-4.
2.
Lipscomb MF, Cope LD, Stephens GL, Deng JS, Gilliam JN. Comparison of substrates for the detection of antinuclear antibodies in normals and in patients with connective tissue and other diseases. Diagn Immunol 1984; 2 :181-7.
3.
Von Muhlen CA, Tan EM. Autoantibodies in the diagnosis of systemic rheumatic diseases. Semin Arthritis Rheum 1998; 41 : 1152-60.
4.
Hooper B, Whittingham S, Mathews JD, Machay IR, Crunow DH. Autoimmunity in a rural community. Clin Exp Immunol 1972; 12 : 79-87.
5.
Cainelli F, Betterle C, Vento S. Antinuclear antibodies are common in an infectious environment but do not predict systemic lupus erythematosus. Ann Rheum Dis 2004; 63 : 1707-8.
6.
Hoschberg MC. Updating the ACR revised criteria for the classification of SLE. Arthritis Rheum 1997; 40 :1725.
7.
Bohan A, Peter JB. Polymyositis and dermatomyositis. N Engl J Med 1975; 292 : 403-7.
8.
Arnett FC, Edworthy SM, Bloch DA, McShance DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31 : 315-24.
9.
Summary report CAP. Diagnostic Immunology S-A. Chicago: College of American Pathologists; 2001 p.14.
10. Tan EM, Feltkamp TE, Smolen JS, Butcher B, Dawkins R, Fritzler MJ, et al. Range of antinuclear antibodies in “healthy” individuals. Arthritis Rheum 1997; 40 : 1601-11. 11. Al-Jabri AA, Al Belushi MS, Nsanze H. Frequency and level of autoantibody in healthy adult Omanis. Ann Saudi Med 2003; 23 : 372-5. 12. Baig MM, Shere SJ. Prevalance of autoantibodies in Saudi population. J Med 1989; 20 : 286-90. 13. Azizah MR, Shahnaz M, Zulkifli MN, Azila MN, Nasuruddin BA. The occurrence of autoantibodies in sera of healthy pregnant women. Med J Malaysia 1997; 52 : 8-11. 14. Bullock JY, Booth RJ, Wilson JD. Tissue antibodies in a healthy New Zealand population. N Z Med J 1979; 89 : 11-3. 15. Butch AW. Dilution protocols for detection of hook effects/ prozone phenomenon. Clin Chem 2000; 46 : 1719-21. 16. Avaniss-Aghajani E, Berzon S, Sarkissian A. Clinical value of multiplexed bead-based immunoassays for detection of autoantibodies to nuclear antigens. Clin Vaccine Immunol 2007; 14 : 505-9.
Reprint requests: Prof. Ramnath Misra, Department of Clinical Immunology, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow 226014, India. e-mail :
[email protected]