Assessment of Serum Free Light Chain Assays for Plasma Cell Disorder Screening in a Veterans Affairs Population

Transcription

1 Available online at Annals of Clinical & Laboratory Science, vol 36, no. 2, Assessment of Serum Free Light Chain Assays for Plasma Cell Disorder Screening in a Veterans Affairs Population Jude M. Abadie and Daniel D. Bankson Department of Laboratory Medicine, University of Washington Medical School, and Pathology and Laboratory Medicine Service, Veterans Affairs Puget Sound Health Care System, Seattle, Washington Abstract. This study evaluated serum κ and l free light chain (FLC) concentrations in a Veterans Affairs (VA) population. We hypothesized that our older, mostly male, population should not differ in serum FLC ranges from levels previously established for younger male and female populations and that the assay would improve our screening protocol for plasma cell dyscrasias (PCD). Serum κ and l FLC were assayed in 312 consecutive serum samples collected during a 16-week period from veterans whose clinical presentation indicated a need for serum protein electrophoresis (SPEP) analysis. We reviewed our laboratory information system (LIS) files to evaluate the patients diagnoses and treatment status in conjunction with serum FLC levels. All assays and validation studies were conducted using an immunoturbidimetric method with a Roche/Hitachi 911 modular analytical system. The intra-assay variability (CV) was <5%, based on 13 replicate assays of 4 control samples and 1 blank sample. Of the 312 patients, the SPEP results were normal in 235 and abnormal in 77. Of the 235 patients with normal SPEP results, 37 had abnormal FLC values and 20 of these were diagnosed as PCD. Of the 77 patients with abnormal SPEP results, only 9 had diagnoses unrelated to PCD. Using the FLC assay in conjunction with retrospective reviews of medical records, we obtained an 86% detection rate of PCD. This detection rate increased to 100% when both SPEP and FLC results were considered. In conclusion, this study documents an important role for serum FLC assays in diagnosing and monitoring PCD in a VA population. Our results support previously established serum FLC reference ranges that were obtained in younger, male and female populations. Using the serum FLC results in conjunction with SPEP results improves the sensitivity and specificity for managing VA patients whose clinical presentation indicates the need to evaluate PCD. Keywords: light chains, monoclonal gammopathy, myeloma, electrophoresis, plasma cell dyscrasia Introduction Multiple myeloma, a malignant plasma cell dyscrasia (PCD) associated with bone marrow plasmacytosis, is characterized by abnormal monoclonal immunoglobulin production. The neoplastic products of malignant plasma cells cause pathologic processes that include bone pain/fractures, renal failure, anemia, infection, hyperviscosity, and hypercalcemia [1,2]. A serum or urine monoclonal Address correspondence to Jude Abadie M.D., Ph.D.,Veterans Affairs Puget Sound Health Care System, 1660 S. Columbian Way (S-113-Lab), Seattle, Washington 98108, USA; fax ; judeabadie@medscape.com. component identified by immunofixation is the most common characteristic feature of monoclonal gammopathies [3]. Serum or urine immunoglobulin light chains may be the only finding in approximately 20% of myeloma patients [4]. While 24-hr urine samples contain high concentrations of light chains, their traditional use in clinical chemistry is often cumbersome, inaccurate, and difficult to measure by electrophoresis. Furthermore, changes in urinary M protein have been shown to correlate strongly with changes of serum free light chain (FLC) concentrations [4]. Therefore, serum FLC measurements may be better indices of disease progression than tests for Bence-Jones protein in urine /06/ $ by the Association of Clinical Scientists, Inc.

2 158 Annals of Clinical & Laboratory Science, vol 36, no. 2, 2006 A new quantitative nephelometric assay for serum FLCs has been used in several clinical investigations of PCD. The assay detects circulating monomeric and dimeric κ and l FLCs. These light chains are free in that they are unbound to a heavy chain immunoglobulin. Serum levels of κ and l FLCs, in conjunction with the FLC κ/l ratio, have been used with a high degree of sensitivity and specificity to identify and monitor FLC disease. Additional clinical use of the FLC assay includes monitoring disease stage in systemic primary amyloidosis, nonsecretory multiple myeloma, light chain deposition disease, and light chain multiple myeloma [5]. In such diseases, the underlying PCD can be difficult to detect or quantify by serum protein electrophoresis [6]. While immunofixation is the current gold standard for detecting monoclonal FLCs, that assay is qualitative and has sensitivities that vary among laboratories. A sensitive and specific quantitative serum FLC assay may prove to be an important guide for monitoring and treating patients with light chain disease. The clinical strength of the serum FLC assay as a pre-neoplastic marker may lie in its ability to assess the risk of progression from monoclonal gammopathy of undetermined significance (MGUS) to clinical PCD. This rate of progression is about 1% per year [7]. Specifically, an abnormal κ/l ratio has been reported to be the major independent risk factor for disease progression from MGUS; patients with normal ratios have low risk and do not require long-term monitoring [8,9]. We evaluated the serum FLC assay in a population of mostly male Veterans Affairs (VA) patients and compared our findings to results from studies using slightly younger, mixed-sex populations. Most published studies include a population based on all-comers and therefore reflect a younger mean age group and contain more females than would be observed in a VA population. While PCD is generally classified as a disease of the elderly, serum FLC reference ranges that have been used clinically in many investigations were established in a population of 282 normal subjects (men and women) from 20 to 90 years old [10]. Although both κ and l FLC levels increase with age, the κ/l ratio remains constant as a function of age. Increases in the levels of both κ and l FLC with advancing age have been attributed to declining renal function, inasmuch as significant increases in cystatin C were observed in the same patients. Increased FLC levels are also seen in conjunction with the incidence of MGUS, which has been reported to be ~1% in the general population >50 years of age and 3% in those >70 years of age [11]. We conducted this study to assess the utility of a serum FLC assay in our population of veterans whose clinical presentation led to a serum protein electrophoresis (SPEP) analysis. We decided to use the established reference ranges for serum κ and l FLC levels, hypothesizing that our older male population should not differ from other populations with respect to disease monitoring and test utility. Additionally, we evaluated the serum FLC assay on a Roche Hitachi 911 analyzer, an instrument not previously used in published studies describing the clinical utility of this assay. Materials and Methods This study was conducted using patients sera obtained from the Seattle division of the Veterans Affairs Puget Sound Health Care System hospital. During a 16-week period from 1 October 2004 through 1 February 2005, 312 consecutive patient serum samples (304 males and 8 females), which were routinely being evaluated by SPEP, were stored at -20 C and later assayed to determine levels of κ and l FLC. A previous investigation reported no significant differences in levels of κ or l serum FLC content or κ/l ratio between fresh and frozen samples [12]. We queried our laboratory information system (LIS) for relevant medical record information that included SPEP results, sex, and age. If a patient had more than one SPEP during the collection time period, only the initial sample was used to represent the results presented in this study. We also used the LIS to determine patient diagnoses and treatment status. All procedures were conducted in accordance with the ethical standards established by the University of Washington Medical Center and the Seattle VA Hospital Institutional Review Board. At the Veterans Affairs Puget Sound Health Care System assays of serum total protein concentrations were performed on the Hitachi Modular P analyzer (Roche Diagnostics, Indianapolis, IN) using the biuret technique. The clinical laboratory cites a reference range of g/dl for serum total protein concentration. Serum protein electrophoresis was performed using agarose gels and acid blue stain on the REP 1 system (Helena Laboratories, Beaumont, TX). Densitometric scans of serum protein electrophoresis gels were evaluated by two observers. Reference values for serum protein electrophoresis were: albumin g/dl; alpha-1 globulin g/dl; alpha-2 globulin g/dl; beta

3 Serum free light chain concentrations in a population of veterans 159 Table 1: Intra-assay performance of the serum FLC assay using a Roche/Hitachi 911 modular analytical system, based on 13 sets of replicate analyses of control and blank samples. Control & Mean CV ±2 SD range Blank samples (mg/l) (%) (mg/l) 27.7 mg/l l mg/l l mg/l k mg/l k Blank l Blank k globulin g/dl; and gamma globulin g/dl. The presence of one or more distinct peaks not corresponding to the usual protein bands in the gamma or beta globulin regions indicated the possible presence of a monoclonal or polyclonal gammopathy. Samples containing monoclonal peaks were further evaluated by immunofixation (Sebia, Norcross, GA). Serum κ and l FLCs were measured by immunoturbidimetry on a Hitachi 911 modular analytical system (Roche Diagnostics, Indianapolis, IN) using the FREELITE human κ and l kits (The Binding Site, San Diego, CA). The assay involves two separate measurements that result in the quantification of κ and l FLCs. Serum κ/l ratios are also computed. Patients with ratios >1.65 or <0.26 are identified as producing clonal κ or l FLCs, respectively. The normal serum reference ranges are mg/l for κ and mg/l for l. None of the results from this study were used for patient care. The assays were conducted during a 3-week period after all of the samples had been collected and after the validation studies had been completed (Table 1). Group means were compared by the t-test. Sensitivity, specificity, positive predictive values, and negative predictive values were calculated for results obtained by SPEP, the serum FLC assay, and the combination of both tests. Table 2. Incidences of true or false positive and true or false negative results of screening for plasma cell disorders, based on SPEP and FLC assays of 312 consecutive serum samples. SPEP assays FLC assays (n and %) (n and %) True positive 27 (9%) 33 (11%) True negative 220 (71%) 271 (87%) False negative 15 (5%) 4 (1%) 12 myeloma 2 myeloma on 1 Waldenstrom s treatment* 1 AL amyloidosis 2 myeloma 1 lymphoma False positive or MGUS 50 (16%) 4 (1%) Total 312 (100%) 312 (100%) * Because the half-life of FLC is shorter than that of IgG, FLC concentrations can become normal in advance of IgG concentrations during treatment. Therefore, these 2 values may not represent false negative results. Table 3. Sensitivity, specificity, and predictive values of SPEP and FLC assays, and of the combination of both assays, in screening 312 serum samples for plasma cell disorders. Parameter SPEP FLC SPEP alone alone + FLC Sensitivity 64% 88% 100% Specificity 81% 98% 99% Positive predictive value 35% 88% 89% Negative predictive value 94% 98% 100% Results Table 1 lists the mean values, CVs, and ±2 SD ranges for 4 different control samples and 1 blank sample, based on 13 sets of replicate assays. The 17.8 mg/l κ control gave the smallest CV at 2.39%, and the 54.0 mg/l λ control gave the highest CV at 4.90%. The blank sample, which contained albumin (7 g/dl), yielded concentrations of 1.98 mg/l for κ and 2.74 mg/l for l, with CVs <5.0% for each. Long-term inter-assay CVs did not differ from the intra-assay results (data not shown). Table 2 lists the SPEP and FLC results for the 312 consecutive serum samples. In comparison to SPEP, the FLC assay results demonstrate higher percentages of true positives and true negatives and lower percentages of false positives and false negatives. Table 3 lists the overall sensitivities, specificities, and predictive values for SPEP, serum FLC, and both tests combined. Of the 312 consecutive specimens analyzed by SPEP and subsequently assayed for serum FLC, 235 were classified as normal and 77 were classified as abnormal based on SPEP results. The ages (mean ± SD) for the normal group (67.5 ± 9.8 yr) were not significantly different from the abnormal group (66.5 ± 10.7 yr). Of the 235 patients with normal SPEP results, 198 (84.3%) had normal κ and l values and ratios in conjunction with unremarkable clinical history or disease state. These 198 samples are not identified in Fig. 1, but the values would fit within the indicated box. However, 37 (15.7%) of the 235 patients with normal SPEP results were found to have abnormal κ/l ratio, abnormal FLC levels, or both. These 37 data points are plotted in Fig. 1. The log concentrations of serum κ and l levels

4 160 Annals of Clinical & Laboratory Science, vol 36, no. 2, 2006 Fig. 1. Abnormal serum FLC results in patients with normal SPEP results (n =37). Symbols: open circles = myeloma with abnormal κ/l ratio; solid circles = myeloma with normal κ/l ratio; open squares = non-myeloma with abnormal κ/l ratio; solid squares = polyclonal FLC. Fig. 2: Abnormal FLC results among patients with abnormal SPEP results (n = 41). Symbols: see Fig. 1.

5 Serum free light chain concentrations in a population of veterans 161 (mg/l) are plotted on the x- and y-axes, respectively. The boxed area represents the normal reference ranges for κ and l. The parallel lines represent the range of normal κ/l ratios. Of the 37 individuals with normal SPEP results and an abnormal FLC result, 17 had an abnormal κ/l ratio. These comprised 15 with multiple myeloma, 1 with lymphoma, and 1 with bladder transitional cell carcinoma. Of the 20 patients with a normal κ/l ratio, 3 were previously treated for myeloma, and the remaining 17 had diagnoses that were unrelated to plasma cell disorders. Fig. 2 shows the κ and l FLC levels for a portion of the 77 patients with an abnormal SPEP result. Of these 77 patients, 20 had both an abnormal κ/l ratio and a FLC level that was outside the reference ranges for κ and/or l. Of these 20, 12 had a diagnosis of multiple myeloma, and 2 were diagnosed with marginal cell lymphoma. The remaining 6 patients included Waldenstrom s macroglobulinemia, leiomyosarcoma, PCD osteomyelitis, or an unspecified diagnosis. Fifteen of the 77 abnormal SPEP patients had a normal κ/l ratio with a FLC level that was outside the reference ranges for κ and/or l. These patients were diagnosed with renal insufficiency, chronic anemia, smoldering myeloma with comorbidities, light chain disease, relapsed myeloma, or smoldering myeloma. The solid circles in Fig. 2 represent 6 of those 15 patients as well as 4 of the remaining 42 patients with a myeloma diagnosis but an abnormal SPEP and a normal κ/l ratio with FLC values that were within the reference ranges for both κ and l. The solid squares in Fig. 2 represent 9 of the 15 patients with a normal κ/l ratio but a κ and/or l FLC value outside the reference range. These included 3 patients with renal insufficiency, 1 with chronic anemia, and 5 with unspecified diagnoses. Discussion These results demonstrate the utility of the serum FLC assay for diagnosing and monitoring PCDs in a population of US military veterans. Previous studies have shown the value of this assay for diagnosis of light chain myeloma [6], nonsecretory myeloma [13], and primary amyloidosis [12]. In the results presented here, the serum free light chain assay was also valuable for detection of intact immunoglobulin myeloma. The detection rate for the serum free light chain assay in this study was 86% (25 of 29 PCDs). This is consistent with the only other report of the use of the serum free light chain assay in a clinical setting [5], in which the overall detection rate was 80.5%. Differences between the patient populations may account for the small difference in detection rate between that observed in the present study and that previously reported. Whereas the current study included primarily patients with intact immunoglobulin myeloma, 40% of the patients in the study by Katzmann et al [5] had MGUS or nonsecretory myeloma. It has been reported that excess FLCs are detected in 93% of patients with intact immunoglobulin myeloma, while 82% of nonsecretory patients produce excess levels of FLCs [12]. Therefore, the higher detection rate for the serum FLC assay in this study may reflect a different proportion of intact immunoglobulin myeloma patients, and signals the value of the serum FLC assay for detection and diagnosis of intact immunoglobulin myeloma. In contrast to the detection rate for the serum FLC assay, the detection rate for SPEP was only 48% (14 of 29 cases ofpcd). However, if SPEP and the serum FLC assay were used together, the detection rate for PCD was 100% (33 of 33 cases), albeit with a higher false positive rate. Thus an abnormal SPEP was found in 77 patients, but only 18% of these had a plasma cell disorder. On the other hand, there were 4 false positives for the serum FLC assay in this study. The specificity of the serum FLC assay in this study (99%) was clearly higher than for SPEP (77%). This indicates that the κ/l ratio is a sensitive and specific marker for monoclonal gammopathy. As the κ/l ratio is a measure of plasma cell clonality, it should not be surprising that an abnormal κ/l ratio is specific for monoclonal gammopathies. Because a number of disorders and diseases can increase production of immunoglobulins, there was a significant number of false positive SPEP results. At the same time, there were also several false negative SPEP results. There were fewer false positive and false negative results using the serum FLC assay. Furthermore, combining the serum

6 162 Annals of Clinical & Laboratory Science, vol 36, no. 2, 2006 FLC assay with SPEP resulted in significantly improved sensitivity, specificity, positive predictive value, and negative predictive value. These results indicate an important role for the serum FLC assay in screening for monoclonal gammopathies. The results of this study support our hypothesis that the established reference ranges for serum κ and l FLC levels remain valid for an older male population such as that at a Veterans Affairs Medical Center. We noted that the κ/l ratio, as computed from results of the serum FLC assay, is a sensitive and specific marker for PCDs. Addition of the serum FLC assay to a screening protocol that uses SPEP improves the sensitivity and specificity of the protocol. This will likely result in improved management of patients whose clinical presentation is suspicious for PCD. Such improved management could result in longer life and more productive quality of life for patients with PCD. Acknowledgements The authors thank Mr. James Garbin for technical assistance with FLC assays and Ms. Nancy Audino for LIS support. We also thank The Binding Site, Inc., for providing the FLC reagents and guidance regarding the FLC assay. References 1. Hallek M, Bergsagel P, Andersen P. Multiple myeloma: increasing evidence for a multistep transformation process. Blood 1998;91: Tricot G. New insights into the role of microenvironment in multiple myeloma. Lancet 2000;335: Kyle R. Multiple myeloma. Review of 869 cases. Mayo Clin Proc 1975;50: Abraham R, Clark R, Bryant S, Larson T, Kyle R, Katzmann J. Correlation of serum immunoglobulin free light chain quantification with urinary Bence Jones protein in light chain myeloma. Clin Chem 2002;48: Katzmann J, Abraham R, Dispenzier A, Lust J, Kyle R. Diagnostic performance of quantitative kappa and lambda free light chain assays in clinical practice. Clin Chem 2005;51: Lachmann H, Gallimore R, Carr-Smith H, Bradwell A, Pepys M, Hawkins P. Outcomes in systemic AL amyloidosis in relation to changes in concentration of circulating free immunoglobulin light chains following chemotherapy. Br J Haematol 2003;122: Bradwell A: Serum Free Light Chain Analysis, 3rd Ed. The Binding Site Ltd.; Birmingham, UK, 2005, pp Rajukumar S, Kyle R, Therneau T, Clark R, Bradwell A, Melton L, et al. Presence of monoclonal free light chains in the serum predicts risk of progression in monoclonal gammopathy of undetermined significance. Br J Haematol 2004;127: Bakshi N, Guilbranson R, Garstka D, Bradwell A, Keren D. Serum free light chain (FLC) measurement can aid capillary zone electrophoresis (CZE) in detecting subtle FLC M-proteins. Am J Clin Pathol 2005;124: Katzmann J, Clark R, Abraham R, Bryant S, Lymp J, Bradwell A, Kyle R. Serum and diagnostic ranges for free kappa and free lambda immunoglobulin light chains: relative sensitivity for detection of monoclonal light chains. Clin Chem 2002;48: Kyle R, Rajkumar S. Monoclonal gammopathies of undetermined significance. Hematol Oncol Clin North Am 1999;13: Drayson M, Tang L, Drew R, Mead G, Carr-Smith H, Bradwell A. Serum free light-chain measurements for identifying and monitoring patients with nonsecretory multiple myeloma. Blood 2001;97: Bradwell A, Carr-Smith H, Mead G, Harvey T, Brayson M. Serum test for assessment of patients with Bence Jones myeloma. Lancet 2003;361: