ARTHRITIS & RHEUMATOLOGY Vol. 66, No. 1, January 2014, pp 213–217 DOI 10.1002/art.38193 © 2014, American College of Rheumatology
BRIEF REPORT
Spuriously Low Serum IgG4 Concentrations Caused by the Prozone Phenomenon in Patients With IgG4-Related Disease Arezou Khosroshahi,1 Lynn A. Cheryk,2 Mollie N. Carruthers,3 Judith A. Edwards,2 Donald B. Bloch,3 and John H. Stone3 Objective. To determine the frequency of the prozone effect in patients with IgG4-related disease (IgG4-RD). Methods. After identifying the prozone effect in an index patient with IgG4-RD, we examined additional samples to determine the frequency of this phenomenon. Thirty-eight serum samples obtained from patients with IgG4-RD whose results had been reported previously were retested. The serum IgG4 concentrations determined by this repeat analysis were compared with the originally reported values. Results. In 10 (26%) of 38 patients, the originally reported IgG4 values were falsely low; the prozone effect was identified in each of these 10 samples. Correction of the prozone effect by sample dilution led to revision of the mean serum IgG4 concentration in the 10 samples, from 26 mg/dl to 2,008 mg/dl (normal range 2.4–121 mg/dl). All 10 patients whose samples were affected by the prozone effect had active IgG4-RD. Failure to detect the elevated serum IgG4 concentrations had a direct impact on the decision not to institute treatment in these patients. Conclusion. The prozone effect may lead to major underestimations of IgG4 concentrations in patients with IgG4-RD and offers a potential explanation for the poor correlation observed between disease activity and serum IgG4 levels in some patients. This phenomenon should be considered if the serum IgG4 measurement
appears discordant with the clinicopathologic diagnosis and the clinical assessment of disease activity. IgG4-related disease (IgG4-RD) is a systemic fibroinflammatory condition of unknown etiology (1). The condition is defined by its unique pathologic features, which are similar across all affected organs (2). Patients with IgG4-RD also share certain clinical characteristics, e.g., the tendency to form tumefactive lesions, frequent elevations in the serum IgG4 concentration, and an excellent initial (but often unsustained) response to glucocorticoid treatment (1). Since the initial recognition of this condition, the elevated serum IgG4 concentrations and increased numbers of IgG4-positive plasma cells in tissue have been regarded as diagnostic hallmarks (3,4). Many clinicians give emphasis to the serum IgG4 concentration when evaluating patients with possible IgG4-RD. Some are inclined to reject the diagnosis of IgG4-RD if the serum IgG4 concentration is normal. Any phenomenon that contributes to an increase in the rate of false-negative results of serum IgG4 assays may lead to underdiagnosis or failure to recognize that disease once deemed to be quiescent might be active again. Delays in the initiation of therapy can be associated with substantial diseaserelated morbidity (5). We previously evaluated a patient with biopsyproven IgG4-RD within 4 days of an assessment by his primary care physician. The serum IgG4 value at the time of the patient’s appointment with us was reported to be 30.8 mg/dl (normal range 2.4–121 mg/dl). This report was not consistent with his known status of active disease within the lacrimal and parotid glands, both of which had been biopsied recently because of progressive enlargement. Moreover, the serum sample that had been tested at another laboratory only 4 days prior to his appointment with our group showed an elevated serum IgG4 concentration of 632 mg/dl. This finding led us to seek an explanation for the discrepant results.
1
Arezou Khosroshahi, MD: Emory University School of Medicine, Atlanta, Georgia, and Massachusetts General Hospital and Harvard Medical School, Boston; 2Lynn A. Cheryk, PhD, Judith A. Edwards, BS: Mayo Medical Laboratories, Andover, Massachusetts; 3 Mollie N. Carruthers, MD, Donald B. Bloch, MD, John H. Stone, MD, MPH: Massachusetts General Hospital and Harvard Medical School, Boston. Address correspondence to John H. Stone, MD, MPH, Massachusetts General Hospital, Rheumatology Unit, Yawkey 2, 55 Fruit Street, Boston, MA 02114. E-mail: [email protected]. Submitted for publication March 10, 2013; accepted in revised form September 5, 2013. 213
214
KHOSROSHAHI ET AL
PATIENTS AND METHODS Reexamination of the original serum sample. Discussions regarding the disparate laboratory findings raised the possibility that the low value had been the result of the prozone effect. Serial dilution of the stored original sample to 1:8,000 demonstrated that the serum IgG4 concentration on the day that we evaluated the patient was actually 655 mg/dl, not 30.8 mg/dl, as originally reported. Correction of the low value following dilution confirmed the prozone effect as the cause of the patient’s spuriously low serum IgG4 value. We then investigated other stored samples from our IgG4-RD Registry to determine the extent of this problem. Study cohort. Between July 2009 and April 2012, we obtained serum samples from patients with IgG4-RD. All patients were enrolled in the Massachusetts General Hospital IgG4-RD Registry and had a biopsy-confirmed diagnosis of IgG4-RD. Thirty-eight serum samples from patients in the IgG4-RD Registry whose original serum IgG4 concentration had been previously reported were retested. Assessment of disease activity. The disease activity of patients with IgG4-RD at the time when serum samples were obtained was assessed retrospectively, using the IgG4-RD Responder Index (6). Because an accurate assessment of the serum IgG4 concentration is important when performing a valid Responder Index assessment, the Index was calculated using both the initial and the amended reports of the serum IgG4 level, and the results were compared. Serum IgG assays. Total IgG was measured by immunonephelometry, using a Siemens BN II instrument and Siemens reagent. Concentrations of the IgG subclasses (IgG1, IgG2, IgG3, and IgG4) were measured by immunonephelometry, using a Siemens BN II instrument and reagent sets obtained from either Siemens or The Binding Site. The assays were performed at the Mayo Medical Laboratories, Andover, Massachusetts. Serial dilutions of serum were performed, and the IgG4 content was measured until the results obtained from 2 sequential dilutions either agreed within 20% or the reaction had reached its end point. During retesting of the serum specimens, the laboratory assessed the IgG subclasses using 2 sets of reagents, the IgG subclass reagent set from Siemens and the reagent set from The Binding Site. The rationale for doing so was the fact that the procedure of the laboratory for IgG subclass testing had changed over the course of our sample collection. Samples assayed from July 2009 to October 2010 had been evaluated using the Siemens IgG subclass reagents. After October 2010, reagents from The Binding Site were used. The technicians who performed the retesting were blinded to the patients’ clinical history and previously reported assay values.
RESULTS Demographic features of the patients. Sera from 38 patients were retested for total IgG and its subclasses. The baseline characteristics of the patients are shown in Supplementary Table 1 (available on the Arthritis & Rheumatology Web site at http://onlinelibrary.wiley.com/ doi/10.1002/art.38193/abstract). The average age of the
patients with IgG4-RD was 59 years (range 25–83 years); 23 of the patients were male, and 15 were female. Clinical manifestations of IgG4-RD. Among the 38 patients, the manifestations of IgG4-RD covered the full range of organ system involvement. In 22 patients (57%), multiple organs were affected by IgG4-RD. In 16 (42%) of the 38 patients, only 1 organ was affected by IgG4-RD. Thirty-two patients (84%) had active disease at the time of their serum evaluation. When the uncorrected serum IgG4 value was used, the mean IgG4-RD Responder Index score among these 32 patients was 5.6 (range 2–12), compared with a score of 6.7 (range 3–12) when the corrected serum IgG4 value was used. In the remaining patients, disease was in clinical remission at the time their serum was sampled, and all of them had an IgG4-RD Responder Index score of zero. Eighteen (47%) of the 38 patients had been treated for IgG4-RD with glucocorticoids and/or immunosuppressive medications before measurements of their serum IgG4 concentration. However, 24 (63%) of the 38 patients were receiving no treatment at the time their serum samples were obtained. The remaining 14 patients (37%) were receiving treatment with glucocorticoids, methotrexate, azathioprine, or mycophenolate mofetil (or glucocorticoids in combination with one of these other agents) at the time of serum sampling. Serum IgG4 concentrations. At the time of initial laboratory testing, only 42% of the patients with IgG4-RD (16 of 38) had elevated serum IgG4 concentrations. We observed a consistent 2-fold difference between the 2 reagents: the IgG4 values determined using Siemens reagents were twice the values reported using The Binding Site reagents. This finding was observed across the spectrum of values in these patients and resulted in reduced sensitivity using The Binding Site reagents. According to the package insert for each reagent set, the reagents appear to be calibrated to the same International Federation of Clinical Chemistry standard material. Therefore, this does not appear to be an explanation for the observed bias between the 2 reagent sets. Any observed difference in the result ⬎2-fold between the 2 reagents was considered clinically significant. When The Binding Site reagents were used, falsely low IgG4 values were reported in 10 (26%) of 38 patients (Table 1). Upon retesting, the prozone phenomenon was demonstrated in 100% of these patients. Probing for the prozone effect by performing repeated dilutions until the reported concentration could be reproduced by more than one dilution step or the reaction reached its end point led to a correction in the
THE PROZONE EFFECT IN IgG4-RELATED DISEASE
Table 1. Characteristics of the patients with IgG4-related disease in whom the originally reported IgG4 value was significantly different from the retest value IgG4 value, mg/dl Case no. 1 2 3 4 5 6 7 8 9 10
Original report*
Retest†
Fold increase after dilution
10.3 28.4 29.3 59.8 12.7 17.6 8.0 43.9 14.4 37.5
2,470 941 219 337 5,340 1,850 5,160 1,030 1,910 819
247 33 7.5 5.6 420 105 645 23 132 22
No. of affected organs
Active disease
Presence of prozone effect
1 3 1 1 3 4 7 4 7 1
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
* Using the Binding Site reagent. † Using Siemens reagent.
mean IgG4 concentration reported, from 26 mg/dl to 2,008 mg/dl (normal range 2.4–121 mg/dl). Samples assayed with the Siemens reagent were automatically tested for antigen excess; the appropriate dilutions were performed either automatically by the instrument or manually as the result of a “flag” associated with the value, thereby avoiding the prozone effect. The Binding Site assay gave no indication that antigen excess might be present. All 10 patients whose samples were affected by the prozone effect had active IgG4-RD (mean IgG4-RD Responder Index 7.7 [range 6–12]). In 6 of those patients, multiple organs were affected by IgG4-RD. A medical records review indicated that the original clinical decision regarding further evaluation (e.g., tissue biopsy) or treatment might have been different if the correct value had been known to the clinician. Other IgG subclass abnormalities. One other irregularity in IgG subclass testing was identified in this study. Serum IgG2 concentrations were spuriously high in 9 patients (24% of all samples) when The Binding Site reagent was used. The patients whose sera were affected by problems with IgG2 concentration measurement tended to be those who also had extremely high serum IgG4 concentrations (i.e., ⬎500 mg/dl). Six of these patients were among those whose initial serum IgG4 concentration measurements were affected by the prozone phenomenon, leading to deceptively low results. DISCUSSION IgG4-RD was identified as an emerging disease entity less than a decade ago, and recognition of this
215
condition has increased quickly in recent years (7–9). Although details of the pathophysiology of this condition have not been elucidated, infiltration of high numbers of IgG4-bearing plasma cells into the tissue of involved organs is an important feature of this disease (2). Moreover, the majority of patients have elevated serum concentrations of IgG4, and these concentrations correlate approximately with disease activity (10). In patients with multiple organ involvement, elevations of the serum IgG4 level are striking, occasionally reaching 30–40 times the upper limit of normal. Because many clinicians consider elevated serum IgG4 concentrations to be linked tightly with the diagnosis of IgG4-RD, factors that cause spuriously low serum IgG4 concentrations may lead to a failure to diagnose IgG4-RD. In this study, we demonstrated that the prozone effect led to falsely low serum IgG4 concentrations in 26% of patients tested. A point of even greater concern stems from the general relationship between active disease and higher serum IgG4 concentrations. The prozone effect is more likely to occur in patients with active disease—precisely the subset of patients for whom an accurate diagnosis is most critical. Diagnostic delays in the setting of IgG4-RD can lead to cirrhosis, pancreatic failure, aneurysms of the thoracic or abdominal aorta, advanced renal dysfunction, and many other complications (5,11). The failure of serum IgG4 assays to accurately identify patients with possible IgG4-RD may also lead to additional morbidity and costs from unnecessary diagnostic tests in pursuit of the wrong diagnosis. Nephelometric assays, which are commonly used today for the measurement of serum IgG, measure the amount of light scattered at a given wavelength by a particulate suspension. The amount of light scatter is dependent on the size and amount of aggregates formed by the binding of antibodies present in the reagent to antigens present in the patient sample (12); in the present case, the antigen consists of the IgG subclass molecule being assayed. The combination of antibodies and antigens leads to formation of lattices large enough to scatter light (12). The antigen–antibody aggregation and lattice formation increase to a maximum. If the antigen is present in substantial excess of the antibody available, then lattice formation and the light scatter signal begin to decrease, a phenomenon known as the “hook” or prozone effect (13). To detect the prozone effect, samples are tested both before and after dilutions are performed. If the result obtained when dilutions are performed is greater than that obtained in the undiluted sample, then the
216
KHOSROSHAHI ET AL
prozone phenomenon is present. The prozone effect can be circumvented by diluting the test sample to allow the antibody concentration to remain in excess of the antigen concentration. As knowledge of IgG4-RD has broadened and clinicians have more often assessed the serum IgG4 concentrations of their patients, recognition of the potential shortcomings of serum IgG4 assays as biomarkers for diagnosis and disease activity assessments has developed. Wide variability exists in the reported frequency of elevated serum IgG4 levels (14–18), and the reliability of the serum IgG4 concentration as a biomarker for diagnosis and disease activity has been questioned. Although the prozone effect appears to explain a substantial number of false-negative results of serum IgG4 assays, the cumulative literature and our own experience suggest the existence of patients with biopsy-proven IgG4-RD and normal serum IgG4 concentrations. Our analysis identifies the interaction of IgG4 with other IgG subclasses (in this case, IgG2) as another possible reason for erroneous results of IgG4 assays. IgG4 has unique structural and functional properties that lead to Fc–Fc interactions between different IgG4 molecules and between the Fc portions of IgG4 and those of other IgG subclasses. These phenomena have been reported by other investigators (19), focusing in particular on the interaction between IgG4 and IgG1 (20,21). To our knowledge, this is the first report of possible interaction between IgG4 and IgG2. Deceptively high IgG2 levels were reported in no less than 24% of all samples tested in this cohort. We have also observed potential IgG4–IgG2 interaction in other patients with IgG4-RD (Stone J, et al: unpublished observations). We have demonstrated that the prozone effect is an important explanation for the disparity that sometimes exists between the presence of clinical features of IgG4-RD and the results of serum IgG4 assays. Clinicians should consider the possibility of the prozone effect when the reported serum IgG4 measurement appears discordant with other evidence for a diagnosis of IgG4-RD or the overall assessment of disease activity in a patient with known disease. Other causes of spurious results of IgG subclass testing may be identified as awareness of IgG4-RD grows and investigations of serum IgG subclass concentrations become more common. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved
the final version to be published. Dr. Stone had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Khosroshahi, Cheryk, Carruthers, Edwards, Bloch, Stone. Acquisition of data. Khosroshahi, Cheryk, Edwards, Bloch, Stone. Analysis and interpretation of data. Khosroshahi, Cheryk, Edwards, Bloch, Stone.
REFERENCES 1. Stone JH, Zen Y, Deshpande V. IgG4-related disease. N Engl J Med 2012;366:539–51. 2. Deshpande V, Zen Y, Chan JK, Yi EE, Sato Y, Yoshino T, et al. Consensus statement on the pathology of IgG4-related disease. Mod Pathol 2012;25:1181–92. 3. Hamano H, Kawa S, Horiuchi A, Unno H, Furuya N, Akamatsu T, et al. High serum IgG4 concentrations in patients with sclerosing pancreatitis. N Engl J Med 2001;344:732–8. 4. Kamisawa T, Funata N, Hayashi Y, Eishi Y, Koike M, Tsuruta K, et al. A new clinicopathological entity of IgG4-related autoimmune disease. J Gastroenterol 2003;38:982–4. 5. Stone JH, Patel VI, Oliveira GR, Stone JR. Case records of the Massachusetts General Hospital. Case 38-2012. A 60-year-old man with abdominal pain and aortic aneurysms. N Engl J Med 2012; 367:2335–46. 6. Carruthers MN, Stone JH, Deshpande V, Khosroshahi A. Development of an IgG4-RD Responder Index. Int J Rheumatol 2012;2012:259408. 7. Chen G, Cheuk W, Chan JK. IgG4-related sclerosing disease: a critical appraisal of an evolving clinicopathologic entity. Zhonghua Bing Li Xue Za Zhi 2010;39:851–68. In Chinese. 8. Stone JH. IgG4-related disease: nomenclature, clinical features, and treatment. Semin Diagn Pathol 2012;29:177–90. 9. Umehara H, Okazaki K, Masaki Y, Kawano M, Yamamoto M, Saeki T, et al, the Research Program for Intractable Disease by Ministry of Health, Labor and Welfare (MHLW) Japan G4 team. A novel clinical entity, IgG4-related disease (IgG4RD): general concept and details [review]. Mod Rheumatol 2012;22:1–14. 10. Kamisawa T, Okamoto A, Funata N. Clinicopathological features of autoimmune pancreatitis in relation to elevation of serum IgG4. Pancreas 2005;31:28–31. 11. Khosroshahi A, Bloch DB, Deshpande V, Stone JH. Rituximab therapy leads to rapid decline of serum IgG4 levels and prompt clinical improvement in IgG4-related systemic disease. Arthritis Rheum 2010;62:1755–62. 12. Feldkamp CS. Immunochemical techniques. In: Kaplan LA, Pesce AJ, Kazmierczak SC, editors. Clinical chemistry: theory, analysis, correlation. 4th ed. St. Louis: Mosby; 2003. p. 227–45. 13. Kricka LJ. Principles of immunochemical techniques. In: Burtis CA, Ashwood ER, Bruns DE, editors. Tietz textbook of clinical chemistry and molecular diagnostics. St. Louis: Elsevier Saunders; 2006. p. 219–43. 14. Raina A, Yadav D, Krasinskas AM, McGrath KM, Khalid A, Sanders M, et al. Evaluation and management of autoimmune pancreatitis: experience at a large US center. Am J Gastroenterol 2009;104:2295–306. 15. Ghazale A, Chari ST, Smyrk TC, Levy MJ, Topazian MD, Takahashi N, et al. Value of serum IgG4 in the diagnosis of autoimmune pancreatitis and in distinguishing it from pancreatic cancer. Am J Gastroenterol 2007;102:1646–53. 16. Frulloni L, Lunardi C, Simone R, Dolcino M, Scattolini C, Falconi M, et al. Identification of a novel antibody associated with autoimmune pancreatitis. N Engl J Med 2009;361:2135–42. 17. Morselli-Labate AM, Pezzilli R. Usefulness of serum IgG4 in the
THE PROZONE EFFECT IN IgG4-RELATED DISEASE
diagnosis and follow up of autoimmune pancreatitis: a systematic literature review and meta-analysis. J Gastroenterol Hepatol 2009; 24:15–36. 18. Yamamoto M, Takahashi H, Ohara M, Suzuki C, Naishiro Y, Yamamoto H, et al. A new conceptualization for Mikulicz’s disease as an IgG4-related plasmacytic disease. Mod Rheumatol 2006;16:335–40. 19. Rispens T, Ooievaar-De Heer P, Vermeulen E, Schuurman J, van der Neut Kolfschoten M, Aalberse RC. Human IgG4 binds to
217
IgG4 and conformationally altered IgG1 via Fc-Fc interactions. J Immunol 2009;182:4275–81. 20. Kawa S, Kitahara K, Hamano H, Ozaki Y, Arakura N, Yoshizawa K, et al. A novel immunoglobulin-immunoglobulin interaction in autoimmunity. PLoS One 2008;3:e1637. 21. Van der Neut Kolfschoten M, Schuurman J, Losen M, Bleeker WK, Martinez-Martinez P, Vermeulen E, et al. Anti-inflammatory activity of human IgG4 antibodies by dynamic Fab arm exchange. Science 2007;317:1554–7.
BRIEF REPORT
Spuriously Low Serum IgG4 Concentrations Caused by the Prozone Phenomenon in Patients With IgG4-Related Disease Arezou Khosroshahi,1 Lynn A. Cheryk,2 Mollie N. Carruthers,3 Judith A. Edwards,2 Donald B. Bloch,3 and John H. Stone3 Objective. To determine the frequency of the prozone effect in patients with IgG4-related disease (IgG4-RD). Methods. After identifying the prozone effect in an index patient with IgG4-RD, we examined additional samples to determine the frequency of this phenomenon. Thirty-eight serum samples obtained from patients with IgG4-RD whose results had been reported previously were retested. The serum IgG4 concentrations determined by this repeat analysis were compared with the originally reported values. Results. In 10 (26%) of 38 patients, the originally reported IgG4 values were falsely low; the prozone effect was identified in each of these 10 samples. Correction of the prozone effect by sample dilution led to revision of the mean serum IgG4 concentration in the 10 samples, from 26 mg/dl to 2,008 mg/dl (normal range 2.4–121 mg/dl). All 10 patients whose samples were affected by the prozone effect had active IgG4-RD. Failure to detect the elevated serum IgG4 concentrations had a direct impact on the decision not to institute treatment in these patients. Conclusion. The prozone effect may lead to major underestimations of IgG4 concentrations in patients with IgG4-RD and offers a potential explanation for the poor correlation observed between disease activity and serum IgG4 levels in some patients. This phenomenon should be considered if the serum IgG4 measurement
appears discordant with the clinicopathologic diagnosis and the clinical assessment of disease activity. IgG4-related disease (IgG4-RD) is a systemic fibroinflammatory condition of unknown etiology (1). The condition is defined by its unique pathologic features, which are similar across all affected organs (2). Patients with IgG4-RD also share certain clinical characteristics, e.g., the tendency to form tumefactive lesions, frequent elevations in the serum IgG4 concentration, and an excellent initial (but often unsustained) response to glucocorticoid treatment (1). Since the initial recognition of this condition, the elevated serum IgG4 concentrations and increased numbers of IgG4-positive plasma cells in tissue have been regarded as diagnostic hallmarks (3,4). Many clinicians give emphasis to the serum IgG4 concentration when evaluating patients with possible IgG4-RD. Some are inclined to reject the diagnosis of IgG4-RD if the serum IgG4 concentration is normal. Any phenomenon that contributes to an increase in the rate of false-negative results of serum IgG4 assays may lead to underdiagnosis or failure to recognize that disease once deemed to be quiescent might be active again. Delays in the initiation of therapy can be associated with substantial diseaserelated morbidity (5). We previously evaluated a patient with biopsyproven IgG4-RD within 4 days of an assessment by his primary care physician. The serum IgG4 value at the time of the patient’s appointment with us was reported to be 30.8 mg/dl (normal range 2.4–121 mg/dl). This report was not consistent with his known status of active disease within the lacrimal and parotid glands, both of which had been biopsied recently because of progressive enlargement. Moreover, the serum sample that had been tested at another laboratory only 4 days prior to his appointment with our group showed an elevated serum IgG4 concentration of 632 mg/dl. This finding led us to seek an explanation for the discrepant results.
1
Arezou Khosroshahi, MD: Emory University School of Medicine, Atlanta, Georgia, and Massachusetts General Hospital and Harvard Medical School, Boston; 2Lynn A. Cheryk, PhD, Judith A. Edwards, BS: Mayo Medical Laboratories, Andover, Massachusetts; 3 Mollie N. Carruthers, MD, Donald B. Bloch, MD, John H. Stone, MD, MPH: Massachusetts General Hospital and Harvard Medical School, Boston. Address correspondence to John H. Stone, MD, MPH, Massachusetts General Hospital, Rheumatology Unit, Yawkey 2, 55 Fruit Street, Boston, MA 02114. E-mail: [email protected]. Submitted for publication March 10, 2013; accepted in revised form September 5, 2013. 213
214
KHOSROSHAHI ET AL
PATIENTS AND METHODS Reexamination of the original serum sample. Discussions regarding the disparate laboratory findings raised the possibility that the low value had been the result of the prozone effect. Serial dilution of the stored original sample to 1:8,000 demonstrated that the serum IgG4 concentration on the day that we evaluated the patient was actually 655 mg/dl, not 30.8 mg/dl, as originally reported. Correction of the low value following dilution confirmed the prozone effect as the cause of the patient’s spuriously low serum IgG4 value. We then investigated other stored samples from our IgG4-RD Registry to determine the extent of this problem. Study cohort. Between July 2009 and April 2012, we obtained serum samples from patients with IgG4-RD. All patients were enrolled in the Massachusetts General Hospital IgG4-RD Registry and had a biopsy-confirmed diagnosis of IgG4-RD. Thirty-eight serum samples from patients in the IgG4-RD Registry whose original serum IgG4 concentration had been previously reported were retested. Assessment of disease activity. The disease activity of patients with IgG4-RD at the time when serum samples were obtained was assessed retrospectively, using the IgG4-RD Responder Index (6). Because an accurate assessment of the serum IgG4 concentration is important when performing a valid Responder Index assessment, the Index was calculated using both the initial and the amended reports of the serum IgG4 level, and the results were compared. Serum IgG assays. Total IgG was measured by immunonephelometry, using a Siemens BN II instrument and Siemens reagent. Concentrations of the IgG subclasses (IgG1, IgG2, IgG3, and IgG4) were measured by immunonephelometry, using a Siemens BN II instrument and reagent sets obtained from either Siemens or The Binding Site. The assays were performed at the Mayo Medical Laboratories, Andover, Massachusetts. Serial dilutions of serum were performed, and the IgG4 content was measured until the results obtained from 2 sequential dilutions either agreed within 20% or the reaction had reached its end point. During retesting of the serum specimens, the laboratory assessed the IgG subclasses using 2 sets of reagents, the IgG subclass reagent set from Siemens and the reagent set from The Binding Site. The rationale for doing so was the fact that the procedure of the laboratory for IgG subclass testing had changed over the course of our sample collection. Samples assayed from July 2009 to October 2010 had been evaluated using the Siemens IgG subclass reagents. After October 2010, reagents from The Binding Site were used. The technicians who performed the retesting were blinded to the patients’ clinical history and previously reported assay values.
RESULTS Demographic features of the patients. Sera from 38 patients were retested for total IgG and its subclasses. The baseline characteristics of the patients are shown in Supplementary Table 1 (available on the Arthritis & Rheumatology Web site at http://onlinelibrary.wiley.com/ doi/10.1002/art.38193/abstract). The average age of the
patients with IgG4-RD was 59 years (range 25–83 years); 23 of the patients were male, and 15 were female. Clinical manifestations of IgG4-RD. Among the 38 patients, the manifestations of IgG4-RD covered the full range of organ system involvement. In 22 patients (57%), multiple organs were affected by IgG4-RD. In 16 (42%) of the 38 patients, only 1 organ was affected by IgG4-RD. Thirty-two patients (84%) had active disease at the time of their serum evaluation. When the uncorrected serum IgG4 value was used, the mean IgG4-RD Responder Index score among these 32 patients was 5.6 (range 2–12), compared with a score of 6.7 (range 3–12) when the corrected serum IgG4 value was used. In the remaining patients, disease was in clinical remission at the time their serum was sampled, and all of them had an IgG4-RD Responder Index score of zero. Eighteen (47%) of the 38 patients had been treated for IgG4-RD with glucocorticoids and/or immunosuppressive medications before measurements of their serum IgG4 concentration. However, 24 (63%) of the 38 patients were receiving no treatment at the time their serum samples were obtained. The remaining 14 patients (37%) were receiving treatment with glucocorticoids, methotrexate, azathioprine, or mycophenolate mofetil (or glucocorticoids in combination with one of these other agents) at the time of serum sampling. Serum IgG4 concentrations. At the time of initial laboratory testing, only 42% of the patients with IgG4-RD (16 of 38) had elevated serum IgG4 concentrations. We observed a consistent 2-fold difference between the 2 reagents: the IgG4 values determined using Siemens reagents were twice the values reported using The Binding Site reagents. This finding was observed across the spectrum of values in these patients and resulted in reduced sensitivity using The Binding Site reagents. According to the package insert for each reagent set, the reagents appear to be calibrated to the same International Federation of Clinical Chemistry standard material. Therefore, this does not appear to be an explanation for the observed bias between the 2 reagent sets. Any observed difference in the result ⬎2-fold between the 2 reagents was considered clinically significant. When The Binding Site reagents were used, falsely low IgG4 values were reported in 10 (26%) of 38 patients (Table 1). Upon retesting, the prozone phenomenon was demonstrated in 100% of these patients. Probing for the prozone effect by performing repeated dilutions until the reported concentration could be reproduced by more than one dilution step or the reaction reached its end point led to a correction in the
THE PROZONE EFFECT IN IgG4-RELATED DISEASE
Table 1. Characteristics of the patients with IgG4-related disease in whom the originally reported IgG4 value was significantly different from the retest value IgG4 value, mg/dl Case no. 1 2 3 4 5 6 7 8 9 10
Original report*
Retest†
Fold increase after dilution
10.3 28.4 29.3 59.8 12.7 17.6 8.0 43.9 14.4 37.5
2,470 941 219 337 5,340 1,850 5,160 1,030 1,910 819
247 33 7.5 5.6 420 105 645 23 132 22
No. of affected organs
Active disease
Presence of prozone effect
1 3 1 1 3 4 7 4 7 1
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
* Using the Binding Site reagent. † Using Siemens reagent.
mean IgG4 concentration reported, from 26 mg/dl to 2,008 mg/dl (normal range 2.4–121 mg/dl). Samples assayed with the Siemens reagent were automatically tested for antigen excess; the appropriate dilutions were performed either automatically by the instrument or manually as the result of a “flag” associated with the value, thereby avoiding the prozone effect. The Binding Site assay gave no indication that antigen excess might be present. All 10 patients whose samples were affected by the prozone effect had active IgG4-RD (mean IgG4-RD Responder Index 7.7 [range 6–12]). In 6 of those patients, multiple organs were affected by IgG4-RD. A medical records review indicated that the original clinical decision regarding further evaluation (e.g., tissue biopsy) or treatment might have been different if the correct value had been known to the clinician. Other IgG subclass abnormalities. One other irregularity in IgG subclass testing was identified in this study. Serum IgG2 concentrations were spuriously high in 9 patients (24% of all samples) when The Binding Site reagent was used. The patients whose sera were affected by problems with IgG2 concentration measurement tended to be those who also had extremely high serum IgG4 concentrations (i.e., ⬎500 mg/dl). Six of these patients were among those whose initial serum IgG4 concentration measurements were affected by the prozone phenomenon, leading to deceptively low results. DISCUSSION IgG4-RD was identified as an emerging disease entity less than a decade ago, and recognition of this
215
condition has increased quickly in recent years (7–9). Although details of the pathophysiology of this condition have not been elucidated, infiltration of high numbers of IgG4-bearing plasma cells into the tissue of involved organs is an important feature of this disease (2). Moreover, the majority of patients have elevated serum concentrations of IgG4, and these concentrations correlate approximately with disease activity (10). In patients with multiple organ involvement, elevations of the serum IgG4 level are striking, occasionally reaching 30–40 times the upper limit of normal. Because many clinicians consider elevated serum IgG4 concentrations to be linked tightly with the diagnosis of IgG4-RD, factors that cause spuriously low serum IgG4 concentrations may lead to a failure to diagnose IgG4-RD. In this study, we demonstrated that the prozone effect led to falsely low serum IgG4 concentrations in 26% of patients tested. A point of even greater concern stems from the general relationship between active disease and higher serum IgG4 concentrations. The prozone effect is more likely to occur in patients with active disease—precisely the subset of patients for whom an accurate diagnosis is most critical. Diagnostic delays in the setting of IgG4-RD can lead to cirrhosis, pancreatic failure, aneurysms of the thoracic or abdominal aorta, advanced renal dysfunction, and many other complications (5,11). The failure of serum IgG4 assays to accurately identify patients with possible IgG4-RD may also lead to additional morbidity and costs from unnecessary diagnostic tests in pursuit of the wrong diagnosis. Nephelometric assays, which are commonly used today for the measurement of serum IgG, measure the amount of light scattered at a given wavelength by a particulate suspension. The amount of light scatter is dependent on the size and amount of aggregates formed by the binding of antibodies present in the reagent to antigens present in the patient sample (12); in the present case, the antigen consists of the IgG subclass molecule being assayed. The combination of antibodies and antigens leads to formation of lattices large enough to scatter light (12). The antigen–antibody aggregation and lattice formation increase to a maximum. If the antigen is present in substantial excess of the antibody available, then lattice formation and the light scatter signal begin to decrease, a phenomenon known as the “hook” or prozone effect (13). To detect the prozone effect, samples are tested both before and after dilutions are performed. If the result obtained when dilutions are performed is greater than that obtained in the undiluted sample, then the
216
KHOSROSHAHI ET AL
prozone phenomenon is present. The prozone effect can be circumvented by diluting the test sample to allow the antibody concentration to remain in excess of the antigen concentration. As knowledge of IgG4-RD has broadened and clinicians have more often assessed the serum IgG4 concentrations of their patients, recognition of the potential shortcomings of serum IgG4 assays as biomarkers for diagnosis and disease activity assessments has developed. Wide variability exists in the reported frequency of elevated serum IgG4 levels (14–18), and the reliability of the serum IgG4 concentration as a biomarker for diagnosis and disease activity has been questioned. Although the prozone effect appears to explain a substantial number of false-negative results of serum IgG4 assays, the cumulative literature and our own experience suggest the existence of patients with biopsy-proven IgG4-RD and normal serum IgG4 concentrations. Our analysis identifies the interaction of IgG4 with other IgG subclasses (in this case, IgG2) as another possible reason for erroneous results of IgG4 assays. IgG4 has unique structural and functional properties that lead to Fc–Fc interactions between different IgG4 molecules and between the Fc portions of IgG4 and those of other IgG subclasses. These phenomena have been reported by other investigators (19), focusing in particular on the interaction between IgG4 and IgG1 (20,21). To our knowledge, this is the first report of possible interaction between IgG4 and IgG2. Deceptively high IgG2 levels were reported in no less than 24% of all samples tested in this cohort. We have also observed potential IgG4–IgG2 interaction in other patients with IgG4-RD (Stone J, et al: unpublished observations). We have demonstrated that the prozone effect is an important explanation for the disparity that sometimes exists between the presence of clinical features of IgG4-RD and the results of serum IgG4 assays. Clinicians should consider the possibility of the prozone effect when the reported serum IgG4 measurement appears discordant with other evidence for a diagnosis of IgG4-RD or the overall assessment of disease activity in a patient with known disease. Other causes of spurious results of IgG subclass testing may be identified as awareness of IgG4-RD grows and investigations of serum IgG subclass concentrations become more common. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved
the final version to be published. Dr. Stone had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Khosroshahi, Cheryk, Carruthers, Edwards, Bloch, Stone. Acquisition of data. Khosroshahi, Cheryk, Edwards, Bloch, Stone. Analysis and interpretation of data. Khosroshahi, Cheryk, Edwards, Bloch, Stone.
REFERENCES 1. Stone JH, Zen Y, Deshpande V. IgG4-related disease. N Engl J Med 2012;366:539–51. 2. Deshpande V, Zen Y, Chan JK, Yi EE, Sato Y, Yoshino T, et al. Consensus statement on the pathology of IgG4-related disease. Mod Pathol 2012;25:1181–92. 3. Hamano H, Kawa S, Horiuchi A, Unno H, Furuya N, Akamatsu T, et al. High serum IgG4 concentrations in patients with sclerosing pancreatitis. N Engl J Med 2001;344:732–8. 4. Kamisawa T, Funata N, Hayashi Y, Eishi Y, Koike M, Tsuruta K, et al. A new clinicopathological entity of IgG4-related autoimmune disease. J Gastroenterol 2003;38:982–4. 5. Stone JH, Patel VI, Oliveira GR, Stone JR. Case records of the Massachusetts General Hospital. Case 38-2012. A 60-year-old man with abdominal pain and aortic aneurysms. N Engl J Med 2012; 367:2335–46. 6. Carruthers MN, Stone JH, Deshpande V, Khosroshahi A. Development of an IgG4-RD Responder Index. Int J Rheumatol 2012;2012:259408. 7. Chen G, Cheuk W, Chan JK. IgG4-related sclerosing disease: a critical appraisal of an evolving clinicopathologic entity. Zhonghua Bing Li Xue Za Zhi 2010;39:851–68. In Chinese. 8. Stone JH. IgG4-related disease: nomenclature, clinical features, and treatment. Semin Diagn Pathol 2012;29:177–90. 9. Umehara H, Okazaki K, Masaki Y, Kawano M, Yamamoto M, Saeki T, et al, the Research Program for Intractable Disease by Ministry of Health, Labor and Welfare (MHLW) Japan G4 team. A novel clinical entity, IgG4-related disease (IgG4RD): general concept and details [review]. Mod Rheumatol 2012;22:1–14. 10. Kamisawa T, Okamoto A, Funata N. Clinicopathological features of autoimmune pancreatitis in relation to elevation of serum IgG4. Pancreas 2005;31:28–31. 11. Khosroshahi A, Bloch DB, Deshpande V, Stone JH. Rituximab therapy leads to rapid decline of serum IgG4 levels and prompt clinical improvement in IgG4-related systemic disease. Arthritis Rheum 2010;62:1755–62. 12. Feldkamp CS. Immunochemical techniques. In: Kaplan LA, Pesce AJ, Kazmierczak SC, editors. Clinical chemistry: theory, analysis, correlation. 4th ed. St. Louis: Mosby; 2003. p. 227–45. 13. Kricka LJ. Principles of immunochemical techniques. In: Burtis CA, Ashwood ER, Bruns DE, editors. Tietz textbook of clinical chemistry and molecular diagnostics. St. Louis: Elsevier Saunders; 2006. p. 219–43. 14. Raina A, Yadav D, Krasinskas AM, McGrath KM, Khalid A, Sanders M, et al. Evaluation and management of autoimmune pancreatitis: experience at a large US center. Am J Gastroenterol 2009;104:2295–306. 15. Ghazale A, Chari ST, Smyrk TC, Levy MJ, Topazian MD, Takahashi N, et al. Value of serum IgG4 in the diagnosis of autoimmune pancreatitis and in distinguishing it from pancreatic cancer. Am J Gastroenterol 2007;102:1646–53. 16. Frulloni L, Lunardi C, Simone R, Dolcino M, Scattolini C, Falconi M, et al. Identification of a novel antibody associated with autoimmune pancreatitis. N Engl J Med 2009;361:2135–42. 17. Morselli-Labate AM, Pezzilli R. Usefulness of serum IgG4 in the
THE PROZONE EFFECT IN IgG4-RELATED DISEASE
diagnosis and follow up of autoimmune pancreatitis: a systematic literature review and meta-analysis. J Gastroenterol Hepatol 2009; 24:15–36. 18. Yamamoto M, Takahashi H, Ohara M, Suzuki C, Naishiro Y, Yamamoto H, et al. A new conceptualization for Mikulicz’s disease as an IgG4-related plasmacytic disease. Mod Rheumatol 2006;16:335–40. 19. Rispens T, Ooievaar-De Heer P, Vermeulen E, Schuurman J, van der Neut Kolfschoten M, Aalberse RC. Human IgG4 binds to
217
IgG4 and conformationally altered IgG1 via Fc-Fc interactions. J Immunol 2009;182:4275–81. 20. Kawa S, Kitahara K, Hamano H, Ozaki Y, Arakura N, Yoshizawa K, et al. A novel immunoglobulin-immunoglobulin interaction in autoimmunity. PLoS One 2008;3:e1637. 21. Van der Neut Kolfschoten M, Schuurman J, Losen M, Bleeker WK, Martinez-Martinez P, Vermeulen E, et al. Anti-inflammatory activity of human IgG4 antibodies by dynamic Fab arm exchange. Science 2007;317:1554–7.
