|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cancer Diagnostics |
1 Departments of Clinical Chemistry and2 Hematology, Ulleval University Hospital, Oslo, Norway.
aAddress correspondence to this author at: Ulleval University Hospital, 0407 Oslo, Norway. Fax +47 22 11 81 89; e-mail armin.piehler{at}medisin.uio.no.
| Abstract |
|---|
|
|
|---|
Methods: We measured SFLCs in 3818 sera received for SPE over a 1-year period when patient symptoms or biochemical findings suggested myeloma-related tissue damage (n = 1067). We reviewed SPE and SFLC results from 489 patients together with their final diagnoses obtained from the hospital information technology department.
Results: SFLC measurement, combined with SPE and clinical information, allowed identification of 95% of patients (38 of 40) with previously undiagnosed MM, macroglobulinemia, or primary amyloidosis. Additionally, we identified 45 patients with monoclonal gammopathy of undetermined significance (MGUS) and 4 with plasmacytoma. Of patients followed at our hospital in whom SFLCs were not measured, only 1 patient was diagnosed with MM. This patient had anemia and was mistakenly not tested for SFLCs. An abnormal
/
ratio was found in 26 of 29 patients with MM but also in 36 of 203 patients with renal impairment, polyclonal immunoresponse, or other nonhematological diagnoses. None of the 203 patients with nonhematological disease had a
/
ratio <0.05 or >10.
Conclusions: The combined use of SPE, SFLC measurements, and clinical criteria allows MM to be efficiently diagnosed or excluded based on serum measurements only.
| Introduction |
|---|
|
|
|---|
Based on studies of well-defined patient groups, serum free light chain (SFLC) measurement has become an important tool in the diagnosis and monitoring of the monoclonal gammopathies, light chain multiple myeloma (LCMM) (5)(6), nonsecretory myeloma (7)(8), and primary amyloidosis (AL) (9)(10). However, measurement of serum free light chains has not yet become part of the general guidelines for diagnosis of multiple myeloma.
The monoclonal light chain disorder LCMM, which accounts for up to 15% of all new cases of multiple myeloma (11), and AL, which is one-fifth as common as multiple myeloma (11), may easily be overlooked in serum protein electrophoresis (SPE), where these disorders may not exhibit a monoclonal band. Typically they are diagnosed by the observation of increased concentrations of
or
light chains in urine. Symptoms and general findings in patients with LCMM or AL, however, are often not specific for the disease, and therefore urine samples may not be collected and sent for analysis. For this and other reasons, several authors have suggested that serum should replace urine for measurement of SFLCs (12)(13)(14).
Measurement and interpretation of SFLCs is not without problems, including deciding which reference range or decision limits to use (15)(16). In addition, a recent study using commonly agreed-upon decision limits (12) reported a considerable number of false-positive
/
ratios when all sera sent for SPE analysis were screened in a general hospital population.
SFLCs have been measured in our laboratory since 2004. We receive samples for SPE from hospitalized patients, the hospitals outpatient departments and surrounding hospitals, and general practitioners outside the hospital. In addition to SPE, our screening approach for diagnosing multiple myeloma includes the measurement of SFLC concentrations, but only when the laboratory is made aware of symptoms or there are biochemical findings indicating myeloma-related organ or tissue damage. Because immunofixation electrophoresis (IFE) can be a laborious method, it is not recommended as a general screening test (4) and, therefore, is not routinely performed in our laboratory on samples sent for protein electrophoresis. Given the reported high frequency of false-positive
/
ratios when SFLCs were measured in all sera sent for serum protein electrophoresis (SPE) (60% of all abnormal
/
ratios were false positive, corresponding to 4% of the whole study population) (12), the aim of our study was to evaluate an approach in which the diagnosis of multiple myeloma was made based on serum tests only (including SPE, SFLCs, and other biochemical parameters) and to assess the diagnostic value of SFLC measurements in a subgroup of patients at higher risk for myelomatosis. For this purpose, we reviewed the SPE and SFLC results in all patients in whom SPE was performed over a 1-year period, combined with their final diagnoses from the hospital information technology department.
| Materials and Methods |
|---|
|
|
|---|
and
, and immunoglobulins IgG, IgA, and IgM in samples (n = 1067) from patients meeting at least 1 of the criteria listed in Table 1
/
ratio, and clinical information, we divided patients into several subgroups (Fig. 1
/
ratio (15), and we applied the reference ranges of 3–19 mg/L and 6–26 mg/L (15) for
and
free light chain concentrations, respectively.
|
|
We carried out SPE using a Hydrasys instrument using Hydragel 15 HR gels (Sebia) and measured
and
SFLC concentrations by immunoturbidimetry on a Hitachi 917 analyzer (Roche) using reagents from the Binding Site and following the vendors protocol for the Hitachi 917 analyzer. To confirm visual hypo- or hyperimmunoglobulinemia on SPE and as part of estimating concentrations of monoclonal immunoglobulins present on SPE, we quantified immunoglobulins IgG, IgA, and IgM using a Hitachi 917 analyzer with reagents from Dako, according to the manufacturers protocols.
SFLC measurements have been performed as part of the routine set-up at Ulleval University Hospital since September 2004. All SFLC results were available to the clinician when a diagnosis of monoclonal gammopathy was established. Monoclonal gammopathies were diagnosed and classified according to the criteria of the International Myeloma Working Group (2).
For the period July 1, 2005, to December 31, 2006, the Department of Information Technology of the Ulleval University Hospital provided a complete list of all patients who were examined at the hospital and had been diagnosed with monoclonal gammopathy, including multiple myeloma [International Classification of Diseases, 10th revision (ICD-10) C90.0], Waldenstrom macroglobulinemia (ICD-10 C88.0), plasma cell leukemia (ICD-10 C90.1), extramedullary plasmacytoma (ICD-10 C90.2), monoclonal gammopathy of undetermined significance (MGUS) (ICD-10 D47.2), osteoporosis in multiple myeloma (ICD-10 M82.0), or glomerular disorders in multiple myeloma (ICD-10 N08.1). We compared the patients with a final diagnosis of monoclonal gammopathy to all patients in the study group. In some patients, we retrieved additional clinical information from the electronic hospital journal. Some of the sera were sent from other hospitals or from practitioners outside the hospital. In those samples, we lacked a reliable diagnosis unless the patient was later referred to our hospital (by April 2007).
The study received approval from the local ethics committee.
| Results |
|---|
|
|
|---|
SFLCs were not measured in 2751 sera (72% of all sera included) from 2302 patients because neither the SPE pattern nor clinical or other biochemical findings suggested monoclonal gammopathy (Table 1
). According to the list of patients diagnosed with monoclonal gammopathy in the study period that we obtained from the hospital information technology department, however, 1 patient in this group had multiple myeloma. This patient exhibited osteolysis, hypercalcemia, renal impairment, and anemia and had 25% plasma cells in the bone marrow. According to our inclusion criteria (Table 1
), SFLCs and immunoglobulins in serum should have been measured in this case.
In all, 545 patients met at least 1 criterion suggestive for monoclonal gammopathy (Table 1
), and 1067 sera from these patients had SFLC measurements. Fifty-six patients were excluded because we lacked precise clinical information. These patients were treated at other hospitals or by general practitioners and did not attend our hospital during blood sampling or in the subsequent 6–18 months. The remaining patients (n = 489) were divided into 3 groups based on the SPE results of the first sample drawn: group 1, patients without monoclonal component at SPE (n = 249); group 2, patients with a newly identified monoclonal component at SPE (n = 83); and group 3, patients with previously identified monoclonal gammopathy (n = 157).
Group 3 included patients with previously diagnosed MGUS or MM. Many of the MM patients in this group received treatment that likely influenced the production of immunoglobulin free light chains and thus the
/
ratios. As the study primarily focused on the evaluation of a diagnostic approach combining SPE, SFLC measurements, and clinical information, this group was not further examined.
group 1: patients without monoclonal component at spe
Group 1 included 5 patients with LCMM, 7 patients with other forms of monoclonal gammopathy, 34 patients with other malignant hematological diseases possibly affecting free light chain production (e.g., leukemia and lymphoma) (17), and 203 patients with other diagnoses (Table 2
). Fig. 2
shows the SFLC concentrations of group 1 patients. All 5 patients with LCMM had a markedly increased concentration of 1 type of free light immunoglobulin chain and thus markedly abnormal
/
ratios (Table 2
; Fig. 2A
). More variable
/
ratios were found in the 7 patients with other forms of monoclonal gammopathy (Fig. 2A
).
/
Ratios within the decision limits were found in 1 patient with nonsecretory multiple myeloma, 1 with plasma cell leukemia, and 2 with plasmacytoma (Table 2
). Urine protein electrophoresis was not indicative for monoclonal gammopathy in any of these 4 patients (data not shown). An abnormal
/
ratio was found in 12 of 34 patients with other hematological diseases (Table 2
; Fig. 2A
) and in 36 of 203 patients with other diagnoses (Table 2
; Fig. 2B
). Among the 36 patients with other diagnoses and an abnormal
/
ratio, 23 showed increased serum creatinine levels, indicating renal impairment and/or a polyclonal increase in immunoglobulins as confirmed by immunoglobulin quantification.
|
|
group 2: patients with a newly identified monoclonal component at spe
The clinical diagnoses of the 83 patients in group 2 are detailed in Table 2
; Fig. 3
shows the SFLC results. All but 1 of the 21 patients with multiple myeloma had an abnormal
/
ratio (Fig. 3
; Table 2
). An abnormal ratio was more variably present in the patients with other forms of monoclonal gammopathy, including Waldenstrom macroglobulinemia (6 of 9), plasmacytoma (0 of 2), MGUS (24 of 44), or other hematological diseases (5 of 6) (Table 2
). Altogether, an abnormal
/
ratio was found in 56 of the 83 patients (67%).
|
/
ratios and absolute serum free light chain concentrations in group 1 and 2 patients
Taking group 1 and 2 together, an abnormal
/
ratio was found in all patients with LCMM (7 of 7) and AL (2 of 2) and most patients with other forms of multiple myeloma (19 of 22) and Waldenstrom macroglobulinemia (6 of 9), but less frequently in MGUS (25 of 45), other hematological disorders (17 of 40), and other diagnoses (36 of 203) (Table 2
). The 4 patients with plasmacytoma had a normal
/
ratio. Excluding the other hematological diagnosis group, the overall positive predictive value (PPV) of an abnormal
/
ratio was thus 62% (59 of 95).
The degree of abnormality in the
/
ratio varied substantially between the different diagnosis groups (Table 3
). A markedly abnormal
/
ratio, <0.05 or >10.00, was found almost exclusively in patients with monoclonal gammopathies, whereas a more moderately abnormal
/
ratio was found in all patient groups except LCMM. Among the 36 patients with an abnormal
/
ratio in the other diagnosis group, 34 showed an increased value (Table 3
) and 18 showed increased serum concentrations of both
and
free light chains. A similar predominance of high
free light chains and a high
/
ratio was found in patients with other hematological diseases.
|
Taken together, the use of SPE and measurement of SFLCs in combination with clinical and biochemical information detected 95% of all patients (38 of 40) with any form of multiple myeloma, Waldenstrom macroglobulinemia, or primary amyloidosis (Table 2
).
| Discussion |
|---|
|
|
|---|
An initial evaluation of the SPE pattern, combined with the absence of clinical and biochemical signs indicating multiple myeloma, correctly ruled out multiple myeloma and other severe monoclonal gammopathies in 72% of all sera, with 1 exception due to a mistake in our routine. As clinical information was not available in about 10% of the patients, we cannot exclude the possibility that other patients were diagnosed with multiple myeloma at other institutions. Because we consistently looked for data on biochemical and/or clinical evidence of monoclonal gammopathy, however, the number of multiple myelomas overlooked should be small. Therefore, the present screening process, which included SPE and internationally recommended criteria (2)(4), functioned very efficiently, and we were able to reduce the number of SFLC analyses to about one-quarter of all sera in which SPE had been requested, resulting in substantial cost savings.
During the 1-year period, 7 new cases of light chain multiple myeloma and 22 new cases of other forms of multiple myeloma were diagnosed. The combination of SPE and SFLC measurements detected 95% of all patients (38 of 40) with any form of multiple myeloma, Waldenstrom macroglobulinemia, or primary amyloidosis. SFLC measurements alone detected 26 of 29 cases of multiple myeloma. A limiting factor of our study was that IFE was not performed routinely on sera in which a monoclonal component was not identified at SPE. Therefore, it is not known whether the 2 patients with plasmacytoma and the single patients with nonsecretory multiple myeloma and plasma cell leukemia, who had no monoclonal band upon SPE and a normal
/
ratio, would have been detected by IFE. Owing to lack of IFE results, it may also be that some patients with MGUS and low concentrations of monoclonal immunoglobulin in serum were missed in our screening process. However, it is well documented that MGUS patients with low concentrations of monoclonal immunoglobulin have a lower risk of progressing to malignancy (20). Therefore, SFLC measurements in combination with SPE greatly facilitated the diagnostic process of detecting multiple myeloma in our study. These findings are consistent with previous studies that quantitated SFLCs in all sera subjected to SPE without preselection (12)(18).
The PPV reported for an abnormal
/
ratio varies substantially depending on the study population (8)(12)(18), with PPVs of up to 100% in selected patient groups (10). In our study, we found a PPV for an abnormal
/
ratio of 62%. As IFE was not performed routinely to exclude a small monoclonal band in the other diagnoses groups with an abnormal ratio, some patients with MGUS may have been missed and the PPV thus may be slightly underestimated. However, an abnormal
/
ratio was a sensitive but not very specific marker of monoclonal gammopathy. A recent study by Hill et al. (12) reported a similar PPV for an abnormal
/
ratio. As in their study, we identified polyclonal immunoresponse and renal impairment as the main causes of false-positive
/
ratios. Also consistent with the study by Hill et al., we found that most of the false-positive
/
ratios in patients without a monoclonal component at SPE were borderline abnormal results (1.66–3.00), and none of them exhibited a
/
ratio >10.00 or <0.05.
Present guidelines require that both serum and urine samples should be examined during the initial diagnostic process for possible monoclonal gammopathy (4). Despite our prior efforts, urine is infrequently sent to our laboratory in addition to serum. In our study, we identified nearly all patients with multiple myeloma by clinical information and serum tests exclusively. In the 4 patients with clinical disease (1 patient with nonsecretory MM, 1 with plasma cell leukemia, and 2 with plasmacytoma) that was not detected by SPE or SFLC measurements, urine protein electrophoresis was not indicative of monoclonal gammopathy either. These findings challenge the position of requiring urine protein electrophoresis as an initial screening procedure. This notion is also supported by a recent study by Katzmann et al. (14), who compared urine protein electrophoresis and urine immunofixation with serum studies alone (i.e., SPE, IFE, SFLC) in 428 patients with monoclonal gammopathy and urinary monoclonal immunoglobulin. In their study, serum measurements alone were able to detect all but 2 cases of monoclonal gammopathies with a monoclonal component in urine, and those 2 cases required no further treatment.
In conclusion, our study indicates that the combination of SPE and SFLC is capable of identifying nearly all patients with clinically relevant monoclonal gammopathy. Moreover, our data suggest that, in the diagnostic process for detecting multiple myeloma, it is sufficient to quantify SFLCs in patients suspected of plasma cell dyscrasia from clinical, biochemical, or SPE findings. It should kept in mind, however, that a substantial number of false-positive borderline
/
ratios require, as is often the case in medicine, the consideration of additional clinical and laboratory patient information to correctly interpret SFLC results.
| Acknowledgments |
|---|
Authors Disclosures of Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
Acknowledgments: We thank Bjørn Warhuus at the Department of Information Technology, Ulleval University Hospital, for providing the list of patients with monoclonal gammopathy and acknowledge the expert technical assistance of the protein group of the Department of Clinical Chemistry.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
R. B Fulton and S. L Fernando Serum free light chain assay reduces the need for serum and urine immunofixation electrophoresis in the evaluation of monoclonal gammopathy Ann Clin Biochem, September 1, 2009; 46(5): 407 - 412. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. J. Bakker, A. Bierma-Ram, C. Elderman-van der Werf, M. L. Strijdhaftig, and J. J. van Zanden Quantitation of Serum Free Light Chains Clin. Chem., August 1, 2009; 55(8): 1585 - 1587. [Full Text] [PDF] |
||||
![]() |
J. A. Katzmann, R. A. Kyle, J. Benson, D. R. Larson, M. R. Snyder, J. A. Lust, S. V. Rajkumar, and A. Dispenzieri Screening Panels for Detection of Monoclonal Gammopathies Clin. Chem., August 1, 2009; 55(8): 1517 - 1522. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. P. Piehler and P. Urdal In Reply Clin. Chem., August 1, 2009; 55(8): 1587 - 1588. [Full Text] [PDF] |
||||
![]() |
E. J.D. Robson, J. Taylor, C. Beardsmore, S. Basu, G. Mead, and T. Lovatt Utility of Serum Free Light Chain Analysis When Screening for Lymphoproliferative Disorders: The Experience at a District General Hospital in the United Kingdom Lab Med, June 1, 2009; 40(6): 325 - 329. [Full Text] [PDF] |
||||
![]() |
D. Siegel, E. Bilotti, and K. H. van Hoeven Serum Free Light Chain Analysis for Diagnosis, Monitoring, and Prognosis of Monoclonal Gammopathies Lab Med, June 1, 2009; 40(6): 363 - 366. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Palladini, P. Russo, T. Bosoni, L. Verga, G. Sarais, F. Lavatelli, M. Nuvolone, L. Obici, S. Casarini, S. Donadei, et al. Identification of Amyloidogenic Light Chains Requires the Combination of Serum-Free Light Chain Assay with Immunofixation of Serum and Urine Clin. Chem., March 1, 2009; 55(3): 499 - 504. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. A. Katzmann and A. Dispenzieri Screening Algorithms for Monoclonal Gammopathies Clin. Chem., November 1, 2008; 54(11): 1753 - 1755. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |