| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Clinical Chemistry and2 Unit of Investigation, University Hospital of Elche, Alicante, Spain;
aaddress correspondence to this author at: Department of Clinical Chemistry, University Hospital of Elche, Camino de la Almazara 11, 03203 Alicante, Spain; fax 34-966-679408, e-mail javiedma{at}telefonica.net
Bence Jones proteins [monoclonal free immunoglobulin light chains (FLCs)] are important tumor markers for identifying and managing patients with monoclonal plasma cell diseases, particularly multiple myeloma. In 
20% of such patients, urine FLCs are the only monoclonal component detected (1)(2)(3)(4). FLCs produced in excess and secreted by a clonal population of plasma cells are excreted into urine in variable amounts as monomers, dimers, polymers, or low–molecular-weight fragments, frequently accompanied by significant quantities of polyclonal FLCs (1)(2)(5)(6)(7). The excretion of monoclonal FLCs varies markedly according to tumor mass, renal function, and the molecular features of the light chains. Furthermore, the clinical significance of FLCs depends on the underlying disease and is assessed in relation to other biochemical, hematologic, clinical, and radiologic data. In particular, urine monoclonal FLCs, often present only at low concentrations, are of considerable clinical significance in the diagnosis of renal disease because of their association with primary (AL) amyloidosis or light chain deposition disease (1)(2)(8). Low concentrations of FLCs are also found in some patients with monoclonal gammopathies of undetermined significance (1)(2).
For some years it has been suggested that immunochemical methods for detection of 
and 
immunoglobulin light chains might be useful for detecting urine monoclonal FLCs (9)(10). More recently, immunochemical assays for FLCs have been developed, and 1 study showed that the ratio of urine free to in combination with urine protein electrophoresis was clinically useful (11). The purpose of our study was to compare the sensitivity of 2 automated nephelometric assays with immunofixation electrophoresis (IFE) for the detection of urinary monoclonal FLCs (12)(13).
Between January 2001 and May 2002, we analyzed sera and second-morning urine samples from 362 patients (204 male and 158 female) at the University Hospital of Elche (Spain). Samples were collected from patients with known diseases that included monoclonal gammopathies of undetermined significance (n = 104), multiple myeloma (n = 60) including light chain myeloma (n = 17), and Waldenström macroglobulinemia (n = 8); 204 samples were from patients with chronic inflammatory and autoimmune diseases characterized by polyclonal increases in serum immunoglobulins and patients with renal diseases that caused significant proteinuria. All patients gave informed, verbal consent to the studies.
Urine samples were tested immediately in most cases or stored at 4 °C and analyzed within 3 days of collection. Before analysis, all were centrifuged at 10 000g for 5 min. Of the 362 patients, 350 supplied only 1 urine sample each, 10 patients supplied 2 samples each, and 2 patients supplied 3 samples each.
Urine samples were usually concentrated to protein concentrations of 50–100 g/L by ultrafiltration on a polyethersulfone membrane with a 7.5 kDa cutoff (Vivapore VP 10/20; Vivascience Sartorius AG). Urine total protein was measured by the sodium dodecyl sulfate–pyrogallol red molybdate complex method (Biotrol urine proteins; Biotrol Diagnostic) (14).
We assayed sera and concentrated urines by high-resolution electrophoresis with commercially available agarose gel plates (REP SPE Hi-Res; cat. no. 3276) on a REP automated electrophoresis apparatus (Helena Laboratories). Coomassie Brilliant Blue R-250 stain was used to visualize the proteins.
IFE was performed with agarose plates and antisera from Paragon (IFE Kit nos. 444930 and 446390; Beckman Coulter) according to the manufacturer’s instructions. Polyclonal antisera (IgG fraction) against and that recognize both FLCs and light chains bound in intact immunoglobulins were used to identify FLC bands in the gel plates. Proteins on the gel were overstained with Coomassie Violet R-150. The detection limit of the IFE for monoclonal FLCs was 10 mg/L.
Quantitative determinations of FLCs and total light chains were made with 2 fully automated immunoassays on a BN II nephelometer analyzer (Dade Behring). We measured FLCs with the NSC assay (product code K.BNA.FRK.FRL; New Scientific Company), which uses highly purified polyclonal antisera against hidden determinants of FLCs. The assay protocols included a prereaction step for the detection of antigen excess.
Polyclonal rabbit antisera to human FLCs type (code OWHG 09) and type (code OWHH 09; Dade Behring), which recognize both exposed and hidden determinants of free FLCs, were used for the measurement of total light chains. The manufacturer assay protocol for total did not include a prereaction step to detect antigen excess.
Both immunochemical methods were considered to be acceptable on the basis of the results of detection limit determinations (5.0 mg/L), a total imprecision study (total CVs <10%), antigen excess detection (tested up to 50 g/L), analytical recovery near 100%, and parallelism between polyclonal and monoclonal FLCs within the analytical ranges. Antisera (FLC method) were highly specific and did not show cross-reactivity to intact immunoglobulins; thus we found no detectable concentrations of FLCs despite to the presence of high concentrations of intact immunoglobulins when we analyzed normal diluted-serum pools or urine samples from patients with glomerular nonselective proteinuria.
All urine samples (n = 376) were assessed blindly by the different methods. To avoid reviewer bias, the immunofixation gels were evaluated by 2 independent expert observers. Samples with discordant or unclear results were reassayed and reevaluated until agreement.
Statistical and ROC analyses were made with GraphROC (Ver. 2.0) software for Windows (15). Differences in unpaired comparisons of the areas under the ROC curves (AUCs) were considered statistically significant at P <0.05. Optimal cutoffs for the best classification of monoclonal and polyclonal FLCs were defined by ROC analyses of free and total and and / ratios.
IFE identified 175 monoclonal FLCs (115 and 60 ) in 166 (44.1%) urine samples, including biclonal FLCs in 9 samples. Characteristic polyclonal ladder bands were observed in 213 (56.6%) urines, mainly from patients with renal, chronic inflammatory, or connective autoimmune diseases, but also frequently in urines containing Bence Jones proteins (n = 97).
All detectable concentrations for both methods, FLC (n = 280; 74.4%) and total light chain (n = 312; 82.9%), were included in the data analyses. FLC and total light chain methods detected both and concentrations in 164 (43.6%) and 253 (67.3%) of the urine samples and neither nor in 96 (25.5%) and 64 (17.0%) of the urine samples. The methods detected alone in 108 (28.7%) and 50 (13.3%) of the urine samples and (monoclonal FLCs in all cases) alone in 8 (2.1%) and 9 (2.3%) of the urine samples.
In this studied population, the optimal cutoffs for the detection of monoclonal FLCs for were 43 mg/L (free ), 3.60 (free / ratio), 137 mg/L (total ), and 3.10 (total / ratio) and for were 70 mg/L (free ), 0.96 (free / ratio), 193 mg/L (total ), and 1.18 (total / ratio). The FLC and light chain methods gave false-positive results (/ ratio outside the cutoff limits) in 12 (5.7%) and 24 (11.4%) of 210 urine samples, respectively, whereas they misclassified monoclonal FLCs (, , or both below the cutoffs or undetectable; / ratio within the cutoff limits) in 56 (33.7%) and 78 (46.9%) of 166 urine samples, respectively.
Scatter plots of the results obtained with the and FLC (n = 164) and total light chain (n = 253) methods are shown in panels A and B, respectively, of Fig. 1
. Polyclonal FLCs values were inside and monoclonal FLCs outside the / ratio cutoff limits for the FLC method in 77 (86.5%) of 89 and 60 (80%) of 75 cases and for the total light chain method in 125 (83.9%) of 149 and 76 (73%) of 104 cases.
|
ROC curves for monoclonal FLC detection by the FLC and total light chain methods, with IFE as the comparison method, are shown in panels C and D of Fig. 1
. Overall, calculated free (n = 164) and total (n = 253) / ratios had the largest AUC. Thus, for identification of monoclonal FLCs, both free and total light chain / ratio measurements were significantly better than the individual free (P <0.0006) and total (P <0.000001) measurements. Likewise, for monoclonal FLCs identification, / ratio measurements were significantly better than the individual free (P <0.04) and total (P <0.003) measurements.
Recent studies have shown that measurement of serum FLCs by immunoassay is of considerable clinical value (16)(17). In particular, the assays were more sensitive than urine IFE for the diagnosis and management of patients with nonsecretory multiple myeloma and AL amyloidosis. Moreover, serum FLC concentrations have been correlated with changes in urinary FLC excretion (18), suggesting a role for serum FLCs as an additional laboratory test for identifying and monitoring monoclonal FLCs (19).
In conclusion, both nephelometric methods for urinary FLCs were objective and reproducible and identified those samples containing high concentrations of monoclonal FLCs without antigen excess problems. Both immunoassays had low sensitivity for detecting low concentrations of monoclonal FLCs and did not distinguish monoclonal from polyclonal FLCs at low concentrations because of overlap for and FLC concentrations and for / ratios. Otherwise, the frequent presence of polyclonal and monoclonal mixtures of FLCs may obscure the finding of monoclonal FLCs, particularly if they are present at low concentrations. Therefore, the presence of monoclonal FLCs should never be confirmed solely on the basis of the immunochemical FLC measurements. The FLC method had the highest detection rate for negative samples; it might be used in combination with good quality electrophoresis to exclude if monoclonal FLCs are present in selected patients without clinical suspicion of monoclonal gammopathy.
Acknowledgments
We thank Leonardo Massaro (New Scientific Company) for providing free of charge the necessary reagents and the BNA protocols for FLC determinations.
References
/ ratios to improve the diagnostic accuracy of urine protein electrophoresis and to reduce the volume required for immunoelectrophoresis. Clin Chim Acta 1997;262:121-130.[CrossRef][ISI][Medline]
[Order article via Infotrieve]
to in the diagnosis of Bence Jones proteinuria. Clin Chem Lab Med 2004;42:429-434.[CrossRef][ISI][Medline]
[Order article via Infotrieve]
and free immunoglobulin light chains: relative sensitivity for detection of monoclonal light chains. Clin Chem 2002;48:1437-1444.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, polyclonal; •, monoclonal; 
, subset of urine samples with very high concentrations of monoclonal FLCs and low concentrations of polyclonal FLCs near the detection limit of the assay for the opposite isotype. Dashed lines represent the optimal cutoff values to distinguish monoclonal from polyclonal FLCs for