A 90-year-old man who had a history of transurethral prostate resection for benign prostatic hyperplasia was admitted for left hip pain. A hip radiography examination revealed a large osteolytic lesion in the left greater trochanter, and a bone scintigraphy evaluation showed increased activity in the same area. Because of the patient's history, metastasis from prostate carcinoma was suspected, and bone biopsies were performed the same week. Ten days later, the patient was hospitalized in the emergency department after a fall that fractured the femoral neck. At admission, plasma sodium, chloride, and potassium concentrations were all within their respective reference intervals. The plasma concentration of total protein was 5.0 g/dL (50 g/L) [reference interval, 6.0–8.0 g/dL (60–80 g/L)], and that of albumin was 1.5 g/dL (15 g/L) [reference interval, 3.0–4.5 g/dL (30–45 g/L)]. The albumin-corrected calcium concentration was 10.8 mg/dL (2.69 mmol/L) [reference interval, 9.0–10.6 mg/dL (2.25–2.65 mmol/L)]. Hematology tests showed mild normocytic (mean corpuscular volume, 95.1 fL; reference interval, 80–97 fL]), aregenerative [reticulocyte count, 54 × 103/μL (54 × 109/L); reference interval, 25–80 × 103/μL (25–80 × 109/L)] anemia, with a low hemoglobin concentration [9.9 g/dL (99 g/L); reference interval, 13.0–17.0 g/dL (130–170 g/L)] and leukopenia [2.5 × 103/μL (2.5 × 109/L); reference interval, 4.0–10.0 × 103/μL (4.0–10.0 × 109/L)]. The plasma creatinine value was 1.1 mg/dL (97 μmol/L) [reference interval, 0.71–1.20 mg/dL (62–106 μmol/L)], and the urea nitrogen concentration was 35.0 mg/dL (12.5 mmol/L) [reference interval, 7.8–19.6 mg/dL (2.8–7.0 mmol/L)]. The urine protein value was 540 mg/24 h (reference interval, 0–150 mg/24 h). Urine protein electrophoresis (UPEP)4 showed the presence of an unusual spike in the β region (Fig. 1A). We measured an increased urine concentration of free κ light chain [1.4 mg/dL (14 mg/L); reference interval, 0–0.2 mg/dL (0–2 mg/L)], along with a normal concentration of urine β2-microglobulin.
The patient underwent surgical repair of his left femur. In the meantime, our laboratory asked for serum to complete the investigation of the abnormal UPEP pattern. Investigations of the serum sample revealed a decreased albumin concentration [1.45 g/dL (14.5 g/L); reference interval, 3.0–4.5 g/dL (30–45 g/L)]. Immunoglobulin quantification revealed an increased IgG value [2950 mg/dL (29.5 g/L); reference interval, 700–1000 mg/dL (7–10 g/L)]. IgA and IgM were within their respective reference intervals. Serum protein electrophoresis (SPEP) showed an increased α1 region, along with decreased albumin and γ regions. Moreover, we found an increased β region (19.1 g/L, quantified from the electrophoresis trace and the total protein concentration), along with a loss of separation between the β1 and β2 regions because of a narrow spike (Fig. 1B). With a normal concentration of β-globulins of approximately 300 mg/dL (3 g/L), we estimated the monoclonal protein concentration at approximately 1600 mg/dL (16 g/L). Serum immunofixation was performed with antibodies specific for heavy chains (G, A, M) and light chains (κ, λ). A band was present in the IgG lane with a β electrophoretic mobility, with no corresponding band for the light chain (Fig. 1C). Serum free light chains included a decreased free κ chain [0.06 mg/dL (0.6 mg/L); reference interval, 0.33–1.94 mg/dL (3.3–19.4 mg/L)], a decreased free λ chain [0.045 mg/dL (0.45 mg/L); reference interval, 0.57–2.63 mg/dL (5.7–26.3 mg/L)], and a normal κ/λ ratio of 1.33 (reference interval, 0.26–1.65). The results of IgG subclass quantification were as follows: IgG1, 768 mg/dL (7.68 g/L) [reference interval, 500–800 mg/dL (5–8 g/L)]; IgG2, <9.0 mg/dL (<0.09 g/L) [reference interval, 90–300 mg/dL (0.9–3 g/L)]; IgG3, 23.0 mg/dL (0.23 g/L) [reference interval, 10–80 mg/dL (0.1–0.8 g/L)]; and IgG4, 1.0 mg/dL (0.01 g/L) [reference interval, 10–60 mg/dL (0.1–0.6 g/L)].
QUESTIONS TO CONSIDER
What are the potential causes of increased β-globulins?
What can explain the discrepancy between the presence of κ light chains in the urine and the presence of a monoclonal IgG without light chains in the serum?
What investigations should be performed to characterize the protein responsible for the spike in the β region of the UPEP results?
What can explain the observation that the sum of the 4 IgG subclasses [<800 mg/dL (<8 g/L)] was not equal to the total IgG concentration [2950 mg/dL (29.5 g/L)]?
Abnormal SPEP and UPEP patterns representing potential monoclonal peaks have to be characterized with additional tests. It is necessary to differentiate analytical artifacts from the presence of a monoclonal gammopathy. Interferences can arise from the presence of radiopaque imaging agents or antibiotics. We verified the absence of fibrinogen (found with plasma tube or incomplete blood clotting) and hemoglobin in the patient's serum (no hemolysis in the sample), which are the most common confounding factors leading to apparent monoclonal peaks in the β-globulin region of an SPEP analysis. Confirmation of a monoclonal protein can be reliably done only by immunofixation or immunotyping. Serum immunofixation electrophoresis revealed the presence of a monoclonal γ band with no corresponding light chain (Fig. 1C), a finding consistent with the diagnosis of γ–heavy chain disease (γ-HCD).
HCDs are rare proliferative B-cell disorders characterized by the production of monoclonal proteins with incomplete heavy chain components and without associated light chains (1). HCDs involving the 3 main immunoglobulin classes (IgA, IgM, IgG) have been described. The most frequent HCD is α-HCD, whereas μ-HCD is very rare; γ-HCD is of intermediate incidence. The median age at the time of γ-HCD diagnosis is approximately 60 years. Fatigue, weakness, and lymphadenopathy are the most frequent initial symptoms, and hepato- and splenomegaly are the most common physical findings, which are often accompanied by anemia. The clinical course can range from an asymptomatic state to a rapid progression leading to death within a few weeks (2). γ-HCD can also be associated with other lymphoplasma cell proliferative diseases or autoimmune disorders. Because of the strong clinical heterogeneity and the varying SPEP patterns with an inconstant presence of a monoclonal peak, γ-HCD is thought to be an underdiagnosed disease.
At admission, the patient's hip pain and subsequent fractured femur neck suggested prostate cancer; however, the associated proteinuria and anemia also suggested a gammopathy. With only these clinical features, the γ-HCD diagnosis relied on laboratory tests. In accordance with other reports, a monoclonal peak with a β mobility was present after both SPEP and UPEP. This feature is the most common of γ-HCD electrophoresis patterns (3), although this peak can be found in α-globulin or γ-globulin regions.
The concentrations of serum free light chains were very low, supporting the diagnosis of γ-HCD. The presence of polyclonal free κ light chains in the patient's urine (Fig. 2, lane 11) seemed independent of γ-HCD, because κ light chains in the urine are due to both physiological degradation of polyclonal immunoglobulins and renal excretion of serum free κ light chains (Fig. 2, lane 5). Furthermore, such urine free light chains cannot explain the total proteinuria we measured. To investigate which protein was mainly responsible for the monoclonal peak observed after UPEP (4), we performed co-immunofixation of both serum and urine proteins (Fig. 2). Using γ antiserum, we found a monoclonal component in serum with the same β mobility as that of urine, confirming the presence of the monoclonal IgG heavy chain without the corresponding light chain.
γ-HCD proteins have a lower molecular weight than normal IgGs (4). Effectively, this feature is due not only to the absence of light chains but also to the presence of deletions in either the variable or the constant domain. These deletions can produce broad truncated regions (5) that can reach 50%–60% of the length of a typical heavy chain (4). Because of their small molecular size, these γ-HCD monoclonal proteins are usually present in both serum and urine (6), as we observed in our case. To assess the molecular weight of the γ-HCD protein, we purified it and subjected the fraction to SDS-PAGE (data not shown). A normal γ heavy chain has a molecular weight of 51 kDa, and the size of the truncation of the patient's monoclonal γ heavy chain was approximately 25%, because we estimated the protein's molecular weight to be between 37.5 and 40 kDa from the molecular weight markers and standard purified IgGs. The markedly increased IgG1 value and the very low concentrations of the 3 other subtypes probably identifies the M protein, although various explanations are possible for the discrepancy between the total-IgG measurement and the sum of the concentrations of the 4 IgG subtypes. First, nephelometric techniques can suffer from antigen excess (7) (also known as the high-dose hook effect), but an appropriate serial-dilution assay excluded this possibility. The observed underestimation of the monoclonal protein could then be explained by self-aggregation of truncated heavy chains. This self-aggregation is well recognized for IgM and IgA, but cases of covalent polymers of IgG have also been described (8). It is also possible that the antisera we used for typing the subclasses could have contained antibodies that recognized domains that were absent from the patient's truncated IgG, because these antisera were obtained from sheep that had been immunized against human IgG1, IgG2, IgG3, and IgG4.
Because an association of γ-HCD with other hematologic malignancies, such as malignant lymphoma and other lymphoplasma cell proliferative disorders, has been described, our patient was referred to a hematologist-oncologist. Analysis of a bone marrow biopsy revealed a double population of B lymphocytes. The first clone was of small size, expressed IgM κ, and was negative for CD5, CD23, and CD10. These results are consistent with B-cell non-Hodgkin lymphoma. The second clone of B lymphocytes expressed an IgG without a corresponding light chain, which is consistent with our earlier findings. The pathology laboratory examined the femoral biopsy sample and reported the presence of undifferentiated cells of unknown origin. These cells were not of hematologic lineage, and a prostate origin was thus suspected. A computed tomography examination of the chest and pelvis revealed numerous nodules in the lungs, liver, and bones, suggesting metastatic dissemination. Unfortunately, the diagnosis of γ-HCD had no impact on the clinical course and the treatment of the patient, because his health deteriorated rapidly and he died 3 weeks later.
POINTS TO REMEMBER
γ-HCD is an underdiagnosed and atypical lymphoproliferative process with high variability in the clinical picture; the laboratory often plays a central role in the diagnosis.
γ-HCD therapy depends on the underlying clinicopathologic features rather than on the presence of the monoclonal protein.
The monoclonal IgG found in γ-HCD presents no light chains, and the heavy chains are frequently truncated, yielding an abnormal protein with a low molecular weight.
Electrophoresis and immunofixation of proteins from both serum and urine samples are necessary to establish a γ-HCD diagnosis and should be performed in all cases of lymphoplasma cell proliferative disease, especially if proteinuria or an abnormal SPEP result are present.
↵4 Nonstandard abbreviations:
- urine protein electrophoresis;;
- serum protein electrophoresis;;
- γ–heavy chain disease.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors' Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
- Received for publication May 12, 2010.
- Accepted for publication September 20, 2010.
- © 2011 The American Association for Clinical Chemistry