TY - JOUR
AU - Kyle, Robert, A
AB - Abstract Background: The detection of monoclonal free light chains (FLCs) is an important diagnostic aid for a variety of monoclonal gammopathies and is especially important in light-chain diseases, such as light-chain myeloma, primary systemic amyloidosis, and light-chain-deposition disease. These diseases are more prevalent in the elderly, and assays to detect and quantify abnormal amounts of FLCs require reference intervals that include elderly donors. Methods: We used an automated immunoassay for FLCs and sera from a population 21–90 years of age. We used the calculated reference and diagnostic intervals to compare FLC results with those obtained by immunofixation (IFE) to detect low concentrations of monoclonal κ and λ FLCs in the sera of patients with monoclonal gammopathies. Results: Serum κ and λ FLCs increased with population age, with an apparent change for those >80 years. This trend was lost when the FLC concentration was normalized to cystatin C concentration. The ratio of κ FLC to λ FLC (FLC K/L) did not exhibit an age-dependent trend. The diagnostic interval for FLC K/L was 0.26–1.65. The 95% reference interval for κ FLC was 3.3–19.4 mg/L, and that for λ FLC was 5.7–26.3 mg/L. Detection and quantification of monoclonal FLCs by nephelometry were more sensitive than IFE in serum samples from patients with primary systemic amyloidosis and light-chain-deposition disease. Conclusions: Reference and diagnostic intervals for serum FLCs have been developed for use with a new, automated immunoassay that makes the detection and quantification of monoclonal FLCs easier and more sensitive than with current methods. The serum FLC assay complements IFE and allows quantification of FLCs in light-chain-disease patients who have no detectable serum or urine M-spike. Monoclonal gammopathies are characterized by the clonal expansion of plasma cells. The monoclonal immunoglobulin secreted by these cells is an indicator of clonal proliferation and can be quantitatively measured to monitor disease course (1). The monoclonal gammopathies include multiple myeloma (MM),1 light-chain myeloma, Waldenstrom macroglobulinemia, nonsecretory myeloma (NSMM), smoldering multiple myeloma, monoclonal gammopathy of undetermined significance, primary systemic amyloidosis (AL), and light-chain-deposition disease (LCDD) (2). The monoclonal light-chain diseases (light-chain myeloma, AL, and LCDD) and NSMM often do not have sufficiently high concentrations of serum monoclonal light chains to be detected by serum protein electrophoresis (SPEP) or immunofixation (IFE) (3). In addition, when IFE detects a monoclonal light chain, the amount of protein may be too low to be quantified and monitored by SPEP. Sensitive nephelometric assays that are specific for κ and λ free light chains (FLCs) but that do not recognize light chains bound to immunoglobulin heavy chains have recently been described (4). These automated assays are reported to be more sensitive than IFE for the detection of monoclonal FLCs. The nephelometric assays were used to evaluate serum samples from NSMM patients whose serum and urine samples were negative for FLCs by IFE, and these assays detected excess serum κ or λ FLCs in 19 of 28 patients (5). In a subset of NSMM patients, serial serum samples were analyzed for FLCs, whose quantities were correlated with disease activity. Because IFE does not quantify FLCs and is not sufficiently sensitive to detect small amounts of monoclonal FLCs in all patients with light-chain plasma-cell dyscrasias, it is important to evaluate the utility of FLC assays to diagnose and monitor the light-chain diseases. These diseases affect mainly the elderly, and we designed this study to determine the 95% reference intervals of κ and λ FLCs, as well as the diagnostic interval for the ratio of κ FLC to λ FLC (FLC K/L) in a population 21–90 years of age. In addition, we applied these intervals to a group of patients with light-chain diseases and compared the ability of the nephelometric assay with that of the IFE assay to detect abnormal light chains. Participants and Methods participants Fresh sera from 127 healthy donors 21–62 years of age [68 (54%) women and 59 (46%) men] were obtained from a pool of donors who had typical FLC values (Mayo Clinic). Frozen sera from 155 donors 51–90 years of age [78 (50%) women and 77 (50%) men] were obtained from the serum bank of an epidemiologic study that surveyed the incidence of monoclonal gammopathies in Olmsted County, MN. These samples were used under a minimum-risk protocol approved by the Mayo Clinic Institutional Review Board. Both fresh and frozen sera were used in the study to obtain a sampling from a wide age range. For donors 51–62 years of age, there were 25 fresh samples and 47 frozen samples. No significant differences in κ FLC content, λ FLC content, or FLC K/L were found between the 25 fresh and 47 frozen samples in this overlapping age group. All sera were assessed by SPEP and IFE for an M-spike or a restricted migration pattern that would suggest the presence of a monoclonal protein. No abnormalities were detected. Twenty-five polyclonal hypergammaglobulinemia serum samples were obtained from the clinical electrophoresis laboratory. As determined by SPEP, all 25 sera had increased γ-globulin concentrations of 18–39 g/L. None of these sera contained a monoclonal protein as determined by IFE. The clinical laboratory also identified 47 serum samples that had given equivocal IFE results. These samples were either determined to have a monoclonal light chain after multiple IFE assays at different sample dilutions or were from patients who were serum negative but urine positive for a monoclonal light chain. These 47 frozen sera had been collected from AL, LCDD, or MM patients: 24 samples were from patients with a monoclonal κ light chain and 23 from patients with a monoclonal λ light chain. Sera from 19 patients with LCDD were obtained from the Dysproteinemia Clinic frozen serum bank. These LCDD serum samples were selected to represent patients with (a) monoclonal light chains detected by IFE in the serum, (b) monoclonal light chains detected by IFE in the urine but not the serum, or (c) a monoclonal population of bone marrow plasma cells but no IFE-detectable light chains in either serum or urine. analytic methods SPEP was performed on agarose gels with the Helena REP system (Helena Laboratories). Ponceau S stain was used to visualize the proteins, and the stained gels were scanned with a Helena Cliniscan 3 scanner. Serum total protein was determined with biuret reagent on a Hitachi 747 analyzer (Boehringer-Mannheim Corp.). The serum total protein value multiplied by the percentage of protein migrating in the gamma region was used to quantify the gamma fraction. IFE was performed with a Sebia HYDRAGEL 4IF reagent set on a Sebia HYDRASYS electrophoresis system and agarose gels. The IFE assay used antisera against γ, α, μ, κ, and λ to fix specific proteins after electrophoretic separation, and precipitated protein was visualized with acid-violet stain. The detection limits of the IFE assay for monoclonal proteins are 25–50 mg/L, depending on the position of the monoclonal protein band and the content of polyclonal immunoglobulin. Any samples that exhibited monoclonal light-chain staining but no corresponding monoclonal heavy-chain staining were also analyzed by the Ouchterlony method for δ and ε reactivity. This assay has a detection limit of 40 mg/L for IgD and IgE. When either δ or ε was detected, the IFE was repeated with antisera to κ, λ, δ, and ε to detect a monoclonal IgD or IgE protein. Nephelometry was performed on a Dade-Behring BNII. Quantification of total κ, total λ, and cystatin C used antibodies from Dade-Behring. FLCs were quantified with FREELITETM reagent sets from The Binding Site Ltd. These FLC assays use sheep antisera coated on polystyrene latex particles and are enhanced by the addition of polyethylene glycol to the reaction. Saline-diluted serum samples that contained either a monoclonal κ or a λ light chain were used to test assay linearity. We performed 20 replicate tests on polyclonal sera with typical FLC values to determine the CV of the assay. The κ FLC assay was linear to a minimum value of 0.5 mg/L; at 0.7 mg/L, the intraassay CV was 7.9%; and at 14 mg/L, the interassay CV was 8.7%. The λ FLC assay was linear to a minimum value of 0.6 mg/L; at 0.9 mg/L, the intraassay CV was 10%; and at 32 mg/L, the interassay CV was 7%. statistical analysis The SAS and S-PLUS statistical software packages were used to perform the analyses and create graphs. Reference intervals were determined with a method suggested by O’Brien and Dyck (6). This method accounts for differences in means and the variability across age and sex groups. Linear regression analysis was used to adjust for significant differences in means across age or sex groups (6). Significance was defined as both P <0.05 and R2 >0.10. For variables with a means adjustment, variability was adjusted separately for positive and negative residuals from the first regression, again only when significant. Z-scores were then created for each datum by subtracting the fitted value from the first regression and dividing the fitted value (corresponding to the appropriate second regression), depending on whether the residual from the first regression was positive or negative. The 2.5 and 97.5 percentiles were then determined from the Z-scores. These percentiles were back-transformed to the units of measure by reversing the process that had created the Z-scores, which produced reference intervals stratified by age and/or sex when significant. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were estimated for the FLC K/L on the basis of both the central 95% interval and a diagnostic range that captured 100% of the test data. Accuracy was calculated as the proportion of individuals classified correctly. Confidence intervals were calculated according to the exact binomial distribution for sensitivity, specificity, and accuracy and by bootstrap for PPV and NPV. We calculated both PPV and NPV after assuming a 15% prevalence of monoclonal proteins in the samples submitted for monoclonal protein studies. Results The γ fraction, total κ, and total λ were quantified for the 282 reference serum samples (data not shown). There was no age or sex dependence in samples from donors between the ages of 21 and 90 years. The median and central 95% interval for the γ fraction were 13.1 and 7.9–19.3 g/L, respectively. The γ fraction should be mostly IgG, and approximately one-third of its mass should be immunoglobulin light chain (two-thirds κ and one-third λ) bound to the γ heavy chain. The total κ content was 2.52 g/L (1.55–3.78 g/L), and the total λ content was 1.43 g/L (0.89–2.03 g/L). The ratio of total κ to total λ (K/L) had a median value of 1.78 and a central 95% interval of 1.30–2.52. Quantification of κ and λ FLCs showed a trend of increasing values with increasing age (Fig. 1, A and B1). There was no relationship between FLC and sex. The κ and λ FLC values showed an increase that was most apparent for those >80 years of age (Table 1). Although the data for the FLCs tended to increase with increasing age, this increase was not significant. The P value and R2 result for κ FLC vs age were 0.004 and 0.06, respectively, whereas for λ FLC, the respective values were 0.06 and 0.03. A central 95% reference interval was defined without regard to age or sex (Table 2). FLC K/L did not vary with age (Fig. 1C). Cystatin C concentrations showed a relationship with age similar to that between κ and λ FLC concentrations and age (Fig. 2A). When the FLC results were expressed as a ratio of FLC to cystatin C, the trend with age was no longer apparent (Fig. 2, B and C2). When the FLC results were expressed as a ratio of FLC to creatinine, the age dependence was reduced but not eliminated (data not shown). Figure 1. Open in new tabDownload slide κ (A) and λ (B) FLC concentrations and calculated FLC K/L (C) in the 282 reference sera plotted against donor age. •, frozen sera; ○, fresh sera. There is an apparent age dependence for FLC quantification but not for the FLC K/L. Figure 1. Open in new tabDownload slide κ (A) and λ (B) FLC concentrations and calculated FLC K/L (C) in the 282 reference sera plotted against donor age. •, frozen sera; ○, fresh sera. There is an apparent age dependence for FLC quantification but not for the FLC K/L. Figure 2. Open in new tabDownload slide Cystatin C concentrations (A) and calculated κ FLC/cystatin C (B) and λ FLC/cystatin C (C) ratios in the 282 reference samples plotted against donor age. •, frozen sera; ○, fresh sera. The previously observed age dependence of the FLC measurements is no longer apparent when the FLC results are normalized with the cystatin C measurement. Figure 2. Open in new tabDownload slide Cystatin C concentrations (A) and calculated κ FLC/cystatin C (B) and λ FLC/cystatin C (C) ratios in the 282 reference samples plotted against donor age. •, frozen sera; ○, fresh sera. The previously observed age dependence of the FLC measurements is no longer apparent when the FLC results are normalized with the cystatin C measurement. Table 1. Median κ FLC, λ FLC, and FLC K/L values by decade.1 Age, years . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . 20–29 6.3 12.4 0.49 30–39 7.2 13.6 0.55 40–49 7.5 12.8 0.58 50–59 6.4 11.3 0.59 60–69 6.9 11.8 0.70 70–79 8.0 11.9 0.65 80–90 9.1 15.1 0.64 Age, years . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . 20–29 6.3 12.4 0.49 30–39 7.2 13.6 0.55 40–49 7.5 12.8 0.58 50–59 6.4 11.3 0.59 60–69 6.9 11.8 0.70 70–79 8.0 11.9 0.65 80–90 9.1 15.1 0.64 1 κ FLC and λ FLCs were measured for the 282 reference serum samples, and the FLC K/L was calculated. Open in new tab Table 1. Median κ FLC, λ FLC, and FLC K/L values by decade.1 Age, years . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . 20–29 6.3 12.4 0.49 30–39 7.2 13.6 0.55 40–49 7.5 12.8 0.58 50–59 6.4 11.3 0.59 60–69 6.9 11.8 0.70 70–79 8.0 11.9 0.65 80–90 9.1 15.1 0.64 Age, years . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . 20–29 6.3 12.4 0.49 30–39 7.2 13.6 0.55 40–49 7.5 12.8 0.58 50–59 6.4 11.3 0.59 60–69 6.9 11.8 0.70 70–79 8.0 11.9 0.65 80–90 9.1 15.1 0.64 1 κ FLC and λ FLCs were measured for the 282 reference serum samples, and the FLC K/L was calculated. Open in new tab Table 2. FLC reference intervals and diagnostic ranges.1 . 95% reference interval . Diagnostic range . κ FLC 3.3–19.4 λ FLC 5.7–26.3 FLC K/L 0.3–1.2 0.26–1.65 . 95% reference interval . Diagnostic range . κ FLC 3.3–19.4 λ FLC 5.7–26.3 FLC K/L 0.3–1.2 0.26–1.65 1 The reference intervals for the 282 reference serum samples were calculated as the central 95% interval. The FLC K/L diagnostic range includes 100% of the reference population. Open in new tab Table 2. FLC reference intervals and diagnostic ranges.1 . 95% reference interval . Diagnostic range . κ FLC 3.3–19.4 λ FLC 5.7–26.3 FLC K/L 0.3–1.2 0.26–1.65 . 95% reference interval . Diagnostic range . κ FLC 3.3–19.4 λ FLC 5.7–26.3 FLC K/L 0.3–1.2 0.26–1.65 1 The reference intervals for the 282 reference serum samples were calculated as the central 95% interval. The FLC K/L diagnostic range includes 100% of the reference population. Open in new tab The median FLC K/L was 0.59 and was substantially different from the median total K/L of 1.78. This difference has been attributed to the dimerization of λ light chains and the consequently slower clearance compared with that of κ light chains (5). The FLC K/L was not significantly related to age or sex. The FLC K/L central 95% reference interval was 0.3–1.2 (Table 2). By definition, 5% of the general population will have a FLC K/L outside this 95% reference interval. If the FLC K/L is used as a diagnostic test for monoclonal FLCs and the FLC diseases, a 5% false-positive rate is unacceptable. We therefore defined a FLC K/L diagnostic range that included all 282 reference sera tested in this study. The FLC K/L had a 100% range of 0.26–1.65 (Table 2). The FLC results for 25 serum samples with polyclonal hypergammaglobulinemia are shown in Table 3. Although a majority of the κ and λ FLCs were increased in this group of hypergammaglobulinemia sera, none of the FLC K/L values were abnormal. Table 3. Polyclonal hypergammaglobulinemia: FLC results (n = 25).1 . Median . Range . Abnormal, % . κ FLC, mg/L 19.6 4.3–273 52 λ FLC, mg/L 28.8 8.5–307 58 FLC K/L 0.55 0.38–1.18 0 . Median . Range . Abnormal, % . κ FLC, mg/L 19.6 4.3–273 52 λ FLC, mg/L 28.8 8.5–307 58 FLC K/L 0.55 0.38–1.18 0 1 Serum samples were polyclonal hypergammaglobulinemic as determined by SPEP and IFE. Open in new tab Table 3. Polyclonal hypergammaglobulinemia: FLC results (n = 25).1 . Median . Range . Abnormal, % . κ FLC, mg/L 19.6 4.3–273 52 λ FLC, mg/L 28.8 8.5–307 58 FLC K/L 0.55 0.38–1.18 0 . Median . Range . Abnormal, % . κ FLC, mg/L 19.6 4.3–273 52 λ FLC, mg/L 28.8 8.5–307 58 FLC K/L 0.55 0.38–1.18 0 1 Serum samples were polyclonal hypergammaglobulinemic as determined by SPEP and IFE. Open in new tab The κ and λ FLC values were determined for 47 patient serum samples that had been difficult to immunotype by IFE (Table 4). Twenty-four samples came from patients for whom IFE had detected a monoclonal κ light chain in the urine or who had a history of a previous urine sample that contained a monoclonal κ light chain. The serum IFE assays identified 21 positive (monoclonal κ) and 3 equivocal sera. Two of the patients with an equivocal IFE result for serum monoclonal κ protein had this protein detected in their urine by IFE. The third patient with an equivocal IFE result for serum monoclonal κ protein had no monoclonal κ in the concurrent urine sample. Interestingly, one of the patients with monoclonal serum κ protein, as detected by IFE, had no monoclonal κ protein detected in the concurrent urine by IFE. The κ FLC and the FLC K/L were abnormally increased in all 24 sera. The median κ FLC, median λ FLC, and median FLC K/L were 716 mg/L, 1.2 mg/L, and 539, respectively. Table 4. Sensitivity of IFE and FLCs.1 Diagnosis . Age, years . IFE . . Serum . . . . . Serum . Urine . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . Kappa patients  MM 64 κ κ 2230 3.4 666  MM 52 Equivocal κ 46 1.2 38  MM 53 κ κ 929 0.6 1548  MM 67 κ κ 873 6.3 139  MM 47 Equivocal κ 630 8.8 72  MM 70 κ κ 1620 0.6 2700  MM 84 κ κ 109 3 36  MM 53 κ κ 427 1.8 237  MM 60 κ κ 802 3.7 217  MM 58 κ κ 122 7.2 169  MM 45 κ κ 5670 7 810  MM 67 κ κ 2390 1.2 1992  MM 61 κ κ 479 0.2 2395  MM 54 κ κ 214 0.4 535  MM 68 κ κ 60 0.2 301  AL 60 κ κ 141 1.2 118  AL 44 Equivocal Negative2 273 0.2 1365  AL 56 κ κ 372 0.2 1860  AL 58 κ Negative2 939 2 470  AL 70 κ κ 3710 0.2 18 550  AL 64 κ κ 899 0.2 4495  AL 52 κ κ 1330 1.5 887  AL 47 κ κ 218 0.4 545  LCDD 33 κ κ 2950 6.4 466 87% positive 100% positive 100% positive Lambda patients  MM 66 λ λ 3 46 0.065  MM 45 λ λ 3 696 0.004  MM 77 λ λ 2 49 0.041  MM 54 λ λ 0.9 213 0.004  MM 78 λ λ 5 225 0.022  MM3 63 λ λ 3.1 626 0.005  MM4 51 λ λ 3.5 150 0.023  AL 54 Negative λ 0.5 41 0.012  AL 73 Negative λ 5 258 0.019  AL 59 λ λ 5.3 652 0.008  AL 62 λ λ 20.7 637 0.033  AL 47 λ λ 4.5 566 0.008  AL 51 λ Negative6 3.5 4660 0.001  AL 47 Negative λ 21.2 323 0.066  AL 75 λ NA56 13.6 716 0.019  AL 80 Negative λ 21.3 1680 0.013  AL 55 Negative Negative6 3 8.1 (negative) 0.37 (negative)  AL 52 Negative λ 7.4 490 0.015  AL 51 λ λ 21.9 213 0.103  AL 46 Negative λ 5.3 212 0.025  AL 68 λ λ 3.1 36 0.087  AL 71 Negative λ 4.9 57 0.087  AL 72 λ NA6 4.5 340 0.013 65% positive 96% positive 96% positive Diagnosis . Age, years . IFE . . Serum . . . . . Serum . Urine . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . Kappa patients  MM 64 κ κ 2230 3.4 666  MM 52 Equivocal κ 46 1.2 38  MM 53 κ κ 929 0.6 1548  MM 67 κ κ 873 6.3 139  MM 47 Equivocal κ 630 8.8 72  MM 70 κ κ 1620 0.6 2700  MM 84 κ κ 109 3 36  MM 53 κ κ 427 1.8 237  MM 60 κ κ 802 3.7 217  MM 58 κ κ 122 7.2 169  MM 45 κ κ 5670 7 810  MM 67 κ κ 2390 1.2 1992  MM 61 κ κ 479 0.2 2395  MM 54 κ κ 214 0.4 535  MM 68 κ κ 60 0.2 301  AL 60 κ κ 141 1.2 118  AL 44 Equivocal Negative2 273 0.2 1365  AL 56 κ κ 372 0.2 1860  AL 58 κ Negative2 939 2 470  AL 70 κ κ 3710 0.2 18 550  AL 64 κ κ 899 0.2 4495  AL 52 κ κ 1330 1.5 887  AL 47 κ κ 218 0.4 545  LCDD 33 κ κ 2950 6.4 466 87% positive 100% positive 100% positive Lambda patients  MM 66 λ λ 3 46 0.065  MM 45 λ λ 3 696 0.004  MM 77 λ λ 2 49 0.041  MM 54 λ λ 0.9 213 0.004  MM 78 λ λ 5 225 0.022  MM3 63 λ λ 3.1 626 0.005  MM4 51 λ λ 3.5 150 0.023  AL 54 Negative λ 0.5 41 0.012  AL 73 Negative λ 5 258 0.019  AL 59 λ λ 5.3 652 0.008  AL 62 λ λ 20.7 637 0.033  AL 47 λ λ 4.5 566 0.008  AL 51 λ Negative6 3.5 4660 0.001  AL 47 Negative λ 21.2 323 0.066  AL 75 λ NA56 13.6 716 0.019  AL 80 Negative λ 21.3 1680 0.013  AL 55 Negative Negative6 3 8.1 (negative) 0.37 (negative)  AL 52 Negative λ 7.4 490 0.015  AL 51 λ λ 21.9 213 0.103  AL 46 Negative λ 5.3 212 0.025  AL 68 λ λ 3.1 36 0.087  AL 71 Negative λ 4.9 57 0.087  AL 72 λ NA6 4.5 340 0.013 65% positive 96% positive 96% positive 1 Frozen serum samples from 47 patients were obtained from the Clinical Laboratory. There were 24 samples from patients with a monoclonal free κ protein and 23 from patients with a monoclonal free λ protein. These serum samples had been saved because the serum monoclonal FLC had been difficult to identify by IFE or because the serum was negative by IFE but the urine contained a monoclonal FLC detected by IFE. Two of the AL patients with a serum monoclonal λ FLC detected by IFE did not have concurrent urine samples available for testing. Four of the patients had a negative urine IFE on the concurrent sample, but these patients had a previous urine that had been positive for a monoclonal FLC by IFE. The serum IFE, urine IFE, serum FLC, and serum FLC K/L results are displayed. Patients with a history of a monoclonal κ or λ light chain had a previous urine sample that was positive for monoclonal light chain by IFE. 2 History of urine with monoclonal κ by IFE. 3 Indolent MM. 4 NSMM. 5 NA, not available. 6 History of urine with monoclonal λ by IFE. Open in new tab Table 4. Sensitivity of IFE and FLCs.1 Diagnosis . Age, years . IFE . . Serum . . . . . Serum . Urine . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . Kappa patients  MM 64 κ κ 2230 3.4 666  MM 52 Equivocal κ 46 1.2 38  MM 53 κ κ 929 0.6 1548  MM 67 κ κ 873 6.3 139  MM 47 Equivocal κ 630 8.8 72  MM 70 κ κ 1620 0.6 2700  MM 84 κ κ 109 3 36  MM 53 κ κ 427 1.8 237  MM 60 κ κ 802 3.7 217  MM 58 κ κ 122 7.2 169  MM 45 κ κ 5670 7 810  MM 67 κ κ 2390 1.2 1992  MM 61 κ κ 479 0.2 2395  MM 54 κ κ 214 0.4 535  MM 68 κ κ 60 0.2 301  AL 60 κ κ 141 1.2 118  AL 44 Equivocal Negative2 273 0.2 1365  AL 56 κ κ 372 0.2 1860  AL 58 κ Negative2 939 2 470  AL 70 κ κ 3710 0.2 18 550  AL 64 κ κ 899 0.2 4495  AL 52 κ κ 1330 1.5 887  AL 47 κ κ 218 0.4 545  LCDD 33 κ κ 2950 6.4 466 87% positive 100% positive 100% positive Lambda patients  MM 66 λ λ 3 46 0.065  MM 45 λ λ 3 696 0.004  MM 77 λ λ 2 49 0.041  MM 54 λ λ 0.9 213 0.004  MM 78 λ λ 5 225 0.022  MM3 63 λ λ 3.1 626 0.005  MM4 51 λ λ 3.5 150 0.023  AL 54 Negative λ 0.5 41 0.012  AL 73 Negative λ 5 258 0.019  AL 59 λ λ 5.3 652 0.008  AL 62 λ λ 20.7 637 0.033  AL 47 λ λ 4.5 566 0.008  AL 51 λ Negative6 3.5 4660 0.001  AL 47 Negative λ 21.2 323 0.066  AL 75 λ NA56 13.6 716 0.019  AL 80 Negative λ 21.3 1680 0.013  AL 55 Negative Negative6 3 8.1 (negative) 0.37 (negative)  AL 52 Negative λ 7.4 490 0.015  AL 51 λ λ 21.9 213 0.103  AL 46 Negative λ 5.3 212 0.025  AL 68 λ λ 3.1 36 0.087  AL 71 Negative λ 4.9 57 0.087  AL 72 λ NA6 4.5 340 0.013 65% positive 96% positive 96% positive Diagnosis . Age, years . IFE . . Serum . . . . . Serum . Urine . κ FLC, mg/L . λ FLC, mg/L . FLC K/L . Kappa patients  MM 64 κ κ 2230 3.4 666  MM 52 Equivocal κ 46 1.2 38  MM 53 κ κ 929 0.6 1548  MM 67 κ κ 873 6.3 139  MM 47 Equivocal κ 630 8.8 72  MM 70 κ κ 1620 0.6 2700  MM 84 κ κ 109 3 36  MM 53 κ κ 427 1.8 237  MM 60 κ κ 802 3.7 217  MM 58 κ κ 122 7.2 169  MM 45 κ κ 5670 7 810  MM 67 κ κ 2390 1.2 1992  MM 61 κ κ 479 0.2 2395  MM 54 κ κ 214 0.4 535  MM 68 κ κ 60 0.2 301  AL 60 κ κ 141 1.2 118  AL 44 Equivocal Negative2 273 0.2 1365  AL 56 κ κ 372 0.2 1860  AL 58 κ Negative2 939 2 470  AL 70 κ κ 3710 0.2 18 550  AL 64 κ κ 899 0.2 4495  AL 52 κ κ 1330 1.5 887  AL 47 κ κ 218 0.4 545  LCDD 33 κ κ 2950 6.4 466 87% positive 100% positive 100% positive Lambda patients  MM 66 λ λ 3 46 0.065  MM 45 λ λ 3 696 0.004  MM 77 λ λ 2 49 0.041  MM 54 λ λ 0.9 213 0.004  MM 78 λ λ 5 225 0.022  MM3 63 λ λ 3.1 626 0.005  MM4 51 λ λ 3.5 150 0.023  AL 54 Negative λ 0.5 41 0.012  AL 73 Negative λ 5 258 0.019  AL 59 λ λ 5.3 652 0.008  AL 62 λ λ 20.7 637 0.033  AL 47 λ λ 4.5 566 0.008  AL 51 λ Negative6 3.5 4660 0.001  AL 47 Negative λ 21.2 323 0.066  AL 75 λ NA56 13.6 716 0.019  AL 80 Negative λ 21.3 1680 0.013  AL 55 Negative Negative6 3 8.1 (negative) 0.37 (negative)  AL 52 Negative λ 7.4 490 0.015  AL 51 λ λ 21.9 213 0.103  AL 46 Negative λ 5.3 212 0.025  AL 68 λ λ 3.1 36 0.087  AL 71 Negative λ 4.9 57 0.087  AL 72 λ NA6 4.5 340 0.013 65% positive 96% positive 96% positive 1 Frozen serum samples from 47 patients were obtained from the Clinical Laboratory. There were 24 samples from patients with a monoclonal free κ protein and 23 from patients with a monoclonal free λ protein. These serum samples had been saved because the serum monoclonal FLC had been difficult to identify by IFE or because the serum was negative by IFE but the urine contained a monoclonal FLC detected by IFE. Two of the AL patients with a serum monoclonal λ FLC detected by IFE did not have concurrent urine samples available for testing. Four of the patients had a negative urine IFE on the concurrent sample, but these patients had a previous urine that had been positive for a monoclonal FLC by IFE. The serum IFE, urine IFE, serum FLC, and serum FLC K/L results are displayed. Patients with a history of a monoclonal κ or λ light chain had a previous urine sample that was positive for monoclonal light chain by IFE. 2 History of urine with monoclonal κ by IFE. 3 Indolent MM. 4 NSMM. 5 NA, not available. 6 History of urine with monoclonal λ by IFE. Open in new tab There were also 23 serum samples from patients for whom IFE had detected monoclonal λ light chain protein in the urine or who had a history of monoclonal λ light chain protein detected in urine. The serum IFE assays identified 15 positive and 8 negative sera. Seven of the patients whose sera were λ protein negative by IFE had concurrent urine samples that were positive for monoclonal λ protein by IFE. One patient with a negative serum result by IFE had a concurrent urine sample that was also negative for monoclonal λ protein by IFE. The λ FLC concentration was abnormally increased in 22 sera, and the FLC K/L was abnormally low in 22 sera. The patient with negative FLC λ results also had negative serum and urine results by IFE. As in the κ group, there was one patient whose serum was positive for monoclonal λ by IFE and FLC, but whose urine was λ protein negative by IFE. The median κ FLC, median λ FLC, and median FLC K/L were 4.5 mg/L, 258 mg/L, and 0.019, respectively. The results for 19 LCDD patients are listed in Table 5. These four patient groups included (a) 9 patients whose sera were monoclonal κ positive by IFE; (b) 3 whose sera were monoclonal λ positive by IFE; (c) 4 whose sera were IFE negative but whose urine samples were positive for monoclonal κ by IFE; and (d) 3 whose sera and urine samples were both negative by IFE but whose bone marrow stains were restricted to κ. Among these 19 patients, 12 had a serum monoclonal protein detected by IFE (63%) and 17 had a serum monoclonal protein detected by FLC K/L (89%). Six patients with a negative serum IFE had a positive FLC K/L, and one patient with a positive serum IFE had a negative FLC K/L. Table 5. LCDD: Serum FLC results.1 . Abnormal/Sera tested, n . . . FLC . FLC K/L . Serum, IFE κ positive 8/9 8/9 Serum, IFE λ positive 2/3 3/3 Serum, IFE negative; urine, IFE κ positive 4/4 4/4 Serum, IFE negative; urine, IFE negative; bone marrow, κ positive 1/3 2/3 . Abnormal/Sera tested, n . . . FLC . FLC K/L . Serum, IFE κ positive 8/9 8/9 Serum, IFE λ positive 2/3 3/3 Serum, IFE negative; urine, IFE κ positive 4/4 4/4 Serum, IFE negative; urine, IFE negative; bone marrow, κ positive 1/3 2/3 1 The κ FLC, λ FLC, and FLC K/L were determined for 19 serum samples from patients with LCDD. The fractions of patients with abnormal serum FLC and FLC K/L results are listed. Open in new tab Table 5. LCDD: Serum FLC results.1 . Abnormal/Sera tested, n . . . FLC . FLC K/L . Serum, IFE κ positive 8/9 8/9 Serum, IFE λ positive 2/3 3/3 Serum, IFE negative; urine, IFE κ positive 4/4 4/4 Serum, IFE negative; urine, IFE negative; bone marrow, κ positive 1/3 2/3 . Abnormal/Sera tested, n . . . FLC . FLC K/L . Serum, IFE κ positive 8/9 8/9 Serum, IFE λ positive 2/3 3/3 Serum, IFE negative; urine, IFE κ positive 4/4 4/4 Serum, IFE negative; urine, IFE negative; bone marrow, κ positive 1/3 2/3 1 The κ FLC, λ FLC, and FLC K/L were determined for 19 serum samples from patients with LCDD. The fractions of patients with abnormal serum FLC and FLC K/L results are listed. Open in new tab The results from the 25 polyclonal hypergammaglobulinemia sera (Table 3) and 282 reference sera were used to calculate the specificity of FLC K/L, and the results from the 66 patients with AL, MM, or LCDD (Tables 4 and 5) were used to calculate the sensitivity in this selected patient group (Table 6). As expected, use of the diagnostic range for FLC K/L significantly improved the specificity of the FLC assay. Table 6. FLC K/L: Comparison of reference intervals and diagnostic ranges.1 . Reference interval (0.3–1.2) . . Diagnostic range (0.26–1.65) . . . Estimate . 95% CI2 . Estimate . 95% CI . Sensitivity, % 98 91–100 97 89–100 Specificity, % 95 92–98 100 98–100 PPV, % 78 65–89 100 91–100 NPV, % 100 98–100 99 97–100 Accuracy, % 96 93–98 99 98–100 . Reference interval (0.3–1.2) . . Diagnostic range (0.26–1.65) . . . Estimate . 95% CI2 . Estimate . 95% CI . Sensitivity, % 98 91–100 97 89–100 Specificity, % 95 92–98 100 98–100 PPV, % 78 65–89 100 91–100 NPV, % 100 98–100 99 97–100 Accuracy, % 96 93–98 99 98–100 1 Sera from the 282 reference individuals and 25 polyclonal hypergammaglobulimia patients as well as the 66 sera from the AL, LCDD, and MM patients were used to calculate the sensitivity, specificity, PPV, NPV, and accuracy of the FLC K/L ratio. The PPV and NPV assumed a prevalence of 15% in the test population. Both the 95% central reference interval and the diagnostic range were evaluated. 2 CI, confidence interval. Open in new tab Table 6. FLC K/L: Comparison of reference intervals and diagnostic ranges.1 . Reference interval (0.3–1.2) . . Diagnostic range (0.26–1.65) . . . Estimate . 95% CI2 . Estimate . 95% CI . Sensitivity, % 98 91–100 97 89–100 Specificity, % 95 92–98 100 98–100 PPV, % 78 65–89 100 91–100 NPV, % 100 98–100 99 97–100 Accuracy, % 96 93–98 99 98–100 . Reference interval (0.3–1.2) . . Diagnostic range (0.26–1.65) . . . Estimate . 95% CI2 . Estimate . 95% CI . Sensitivity, % 98 91–100 97 89–100 Specificity, % 95 92–98 100 98–100 PPV, % 78 65–89 100 91–100 NPV, % 100 98–100 99 97–100 Accuracy, % 96 93–98 99 98–100 1 Sera from the 282 reference individuals and 25 polyclonal hypergammaglobulimia patients as well as the 66 sera from the AL, LCDD, and MM patients were used to calculate the sensitivity, specificity, PPV, NPV, and accuracy of the FLC K/L ratio. The PPV and NPV assumed a prevalence of 15% in the test population. Both the 95% central reference interval and the diagnostic range were evaluated. 2 CI, confidence interval. Open in new tab Discussion Although FLCs in serum are usually associated with monoclonal gammopathies, polyclonal FLCs have been detected at low concentrations in healthy serum (4)(7)(8)(9)(10)(11). In monoclonal gammopathies associated with FLCs, these proteins may be present in serum in small amounts and thus be difficult to detect by IFE and often impossible to quantify by an M-spike on SPEP. Unlike IFE, the assessment of FLCs by nephelometry is a quantitative measurement, and the described sensitivity and specificity of the nephelometric FLC assay (4) may allow quantification and monitoring of monoclonal light chains in serum. The ability to detect abnormal amounts of FLCs and an abnormal FLC K/L is dependent, however, on accurately determined reference intervals, so that the specificity of disease detection remains high. The reference intervals reported herein are close to those described in the original report of this nephelometric FLC immunoassay (4). That study, however, did not include older healthy donors. The monoclonal gammopathies are more prevalent in older populations. Several epidemiologic studies have reported the incidence of monoclonal gammopathy of undetermined significance as ∼1% in the population >50 years and 3% in the population >70 years of age (12). The κ and λ serum FLCs showed a trend for increased values with increasing age in our study, and there was a substantial increase in these values in individuals >80 years of age. The FLC K/L ratio, however, normalized the age-dependent increases in FLC. Because total κ and λ concentrations do not show an increase with age and because the FLC K/L normalizes the increase in κ and λ FLC values, the most likely explanation for the observed increase in serum FLC concentrations is a decrease in renal clearance with advancing age. Measures of renal clearance show an age-related decrease in renal function that begins in the third decade (13)(14). Cystatin C is a sensitive indicator of renal clearance, and the increase in FLC content with age is reflected by an increase in cystatin C. Dividing the FLC result by the cystatin C result eliminates the apparent dependence of FLC on age. The increase in FLC values is therefore likely attributable to reduced kidney function and not to age per se. Because many patients with monoclonal gammopathies also have decreased renal function and proteinuria, these could be confounding factors in interpreting FLC measurements. The FLC K/L, however, is not affected by renal function and is therefore the most straightforward representation of the data for diagnostic testing. In the group of 25 samples from patients with polyclonal hypergammaglobulinemia, the FLC values were also increased. The increment in this patient group was presumably not attributable to reduced kidney function but to increased immunoglobulin synthesis. The FLC K/L in this group of samples also normalized the κ and λ FLC increases, such that no samples had an abnormal FLC K/L. Using the reference intervals and diagnostic ranges developed in this study, we assessed the relative sensitivity of the FLC assay for detecting monoclonal κ and λ FLC concentrations in sera of a cohort of patients who had a variety of monoclonal gammopathies. In this small, selected group of samples, the FLC measurements had a higher sensitivity than did IFE for detecting small concentrations of monoclonal light chains in serum. Interestingly, those serum samples that were negative by IFE had FLC concentrations that were similar to those in samples with positive IFE results. The lower sensitivity of the serum IFE assay in this group may be attributable to polyclonal immunoglobulins that obscure the small, monoclonal FLC band. The binding of the FLC assay is reported to be >10 000-fold higher for FLCs in comparison with the light chains bound to heavy chains in intact immunoglobulin (4). This binding preference may allow detection of a small increase in FLC concentration when polyclonal immunoglobulins are present. Alternatively, the serum IFE assay may not detect FLCs in some of these samples because of polymerization of monoclonal light chains. The polymerization of some light chains yields complexes that electrophorese in very broad patterns that are not recognized as monoclonal light chains (5). The increased sensitivity of the FLC K/L compared with serum IFE makes the FLC method a useful diagnostic assay. Primary systemic amyloidosis and LCDD are often difficult to diagnose, and the presence of a monoclonal FLC is an important differential diagnostic clue. The sensitivity of the IFE method in serum or urine is ∼70% for AL and ∼90% when both serum and urine assays are performed. The enhanced diagnostic sensitivity may be useful for disease detection and may be an additional laboratory assessment for patients suspected of having a light-chain disease. The FLC data for the 19 LCDD serum samples demonstrate the differences between the FLC and IFE assays. Seven of these samples were negative for a serum monoclonal light chain by IFE, and 6 of these had an abnormal FLC K/L. Surprisingly, the FLC K/L was not increased in one of nine sera with a monoclonal κ light chain detected by IFE. This serum sample was retested, and the IFE and FLC results were reproducible. The κ FLC value in this sample was at the low end of the usual reference interval and was much lower than the other eight samples in this group. The assay selectivity for cryptic light-chain sites that are “hidden” when bound to heavy-chain proteins most likely depends on reactivity with very few sites on the light chain. We speculate that the lack of reactivity with the monoclonal light chain from this patient may have been attributable to truncation of the light chains and the consequent loss of antigenic sites. The increased sensitivity of the FLC assay compared with the serum IFE assay, coupled with the inability of the FLC assay to detect certain light chains, suggests that these two assays be used as complementary diagnostic tests. In addition to the increased sensitivity and diagnostic potential of the FLC assays, the ability to quantify monoclonal FLCs may be useful to monitor the disease process in light-chain diseases. The disease course of AL can be difficult to assess and is currently monitored by evaluating organ function (e.g., kidney function by urine protein measurements). Changes in organ function, however, may take a long time to manifest, and therefore, direct measurements of the serum or urine M-spike provide a real-time marker for monitoring AL. Fewer than one-half of AL patients have a measurable M-spike, and the quantification of FLC may provide a more universal and timely assessment of disease activity. Studies correlating disease activity and FLC quantification remain to be done. 1 " Nonstandard abbreviations: MM, multiple myeloma; NSMM, nonsecretory multiple myeloma; AL, primary systemic amyloidosis; LCDD, light-chain-deposition disease; SPEP, serum protein electrophoresis; IFE, immunofixation; FLC, free light chain; FLC K/L, ratio of κ FLC to λ FLC; PPV, positive predictive value; and NPV, negative predictive value. This work was supported in part by Research Grant CA 62242 from the NIH (Bethesda, MD). The Binding Site Ltd. (Birmingham, England) provided the immunoassay reagent sets for FLC quantification. Dr. Timothy Larson (Mayo Clinic Renal Laboratory, Rochester, MN) provided the cystatin C assay. References 1 Kyle RA. Sequence of testing for monoclonal gammopathies: serum and urine assays. Arch Pathol Lab Med 1999 ; 123 : 114 -118. PubMed 2 Kyle RA. Evaluation of patients with monoclonal gammopathies. Cancer Res Ther Control 1999 ; 9 : 249 -259. 3 Kyle RA, Gertz MA. Primary systemic amyloidosis: clinical and laboratory features in 474 cases. Semin Hematol 1995 ; 32 : 45 -59. PubMed 4 Bradwell AR, Carr-Smith HD, Mead GP, Tang LX, Showell PJ, Drayson MT, et al. Highly sensitive, automated immunoassay for immunoglobulin free light chains in serum and urine. Clin Chem 2001 ; 47 : 673 -680. Crossref Search ADS PubMed 5 Drayson M, Tang LX, Drew R, Mead GP, Carr-Smith H, Bradwell AR. Serum free light-chain measurements for identifying and monitoring patients with nonsecretory multiple myeloma. Blood 2001 ; 97 : 2900 -2902. Crossref Search ADS PubMed 6 O’Brien PC, Dyck PJ. Procedures for setting normal values. Neurology 1995 ; 45 : 17 -23. Crossref Search ADS PubMed 7 Sölling K. Free light chains of immunoglobulins in normal serum and urine determined by radioimmunoassay. Scand J Clin Lab Invest 1975 ; 35 : 407 -412. Crossref Search ADS PubMed 8 Brouwer J, Otting-van de Ruit M, Busking-van der Lely H. Estimation of free light chains of immunoglobulins by enzyme immunoassay. Clin Chim Acta 1985 ; 150 : 257 -274. 9 Nelson M, Brown RD, Gibson J, Joshua DE. Measurement of free κ and λ chains in serum and the significance of their ratio in patients with multiple myeloma. Br J Haematol 1990 ; 81 : 223 -230. 10 Abe M, Goto T, Kosaka M, Wolfenbarger D, Weiss DT, Solomon A. Differences in kappa and lambda (κ:λ) ratios of serum and urinary free light chains. Clin Exp Immunol 1998 ; 111 : 457 -462. Crossref Search ADS PubMed 11 Wakasugi K, Suzuki H, Imai A, Konishi S, Kishioka H. Immunoglobulin free light chain assay using latex agglutination. Int J Clin Lab Res 1995 ; 25 : 211 -215. Crossref Search ADS PubMed 12 Kyle RA, Rajkumar SV. Monoclonal gammopathies of undetermined significance. Hematol Oncol Clin North Am 1999 ; 13 : 1181 -1202. Crossref Search ADS PubMed 13 Slack TK, Wilson DM. Normal renal function: CIN and CPAH in healthy donors before and after nephrectomy. Mayo Clin Proc 1976 ; 51 : 296 -300. PubMed 14 Peters AM, Henderson BL, Lui D. Indexed glomerular filtration rate as a function of age and body size. Clin Sci 2000 ; 98 : 439 -444. Crossref Search ADS PubMed © 2002 The American Association for Clinical Chemistry This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Serum Reference Intervals and Diagnostic Ranges for Free κ and Free λ Immunoglobulin Light Chains: Relative Sensitivity for Detection of Monoclonal Light Chains
JO - Clinical Chemistry
DO - 10.1093/clinchem/48.9.1437
DA - 2002-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/serum-reference-intervals-and-diagnostic-ranges-for-free-and-free-WPerX4Ok8g
SP - 1437
VL - 48
IS - 9
DP - DeepDyve
ER -