Analysis of Ig heavy chain/light chain pairs (Hevylite™)

From Wikilite

32.1. Introduction: limitations of immunoglobulin measurements

Figure 32.1 Target epitopes (in black) for HLC antibodies are on the constant regions (CH1 and CL) between the heavy and light chains of immunoglobulin molecules.

Figure 32.2 Heavy chain/light chain pairs of IgG, IgA and IgM molecules showing the target epitopes for HLC immunoassays in black.

Typical analytical tests for monoclonal gammopathies are serum protein electrophoresis (SPE) with scanning densitometry together with serum free light chain (sFLC) immunoassays. While SPE is a simple, cheap test, it is not particularly sensitive, and quantification of proteins at low concentrations (1-3g/L) is inaccurate. This is particularly apparent for monoclonal IgA, since its anodal electrophoretic migration positions it over other bands such as transferrin. Improved sensitivity is achieved with immunofixation electrophoresis (IFE) but it is a non-quantitative assay. Nephelometry is also used for immunoglobulin measurements and is analytically accurate to low concentrations. However, patients’ samples also contain non-tumour polyclonal immunoglobulins of both κ and λ types which make results clinically inaccurate at normal serum concentrations. Furthermore, assessments of monoclonal IgG are unreliable due to variable catabolism, as FcRn recycling receptors become saturated or reduced by chemotherapy (Chapter 10).

In contrast, one of the great diagnostic benefits of sFLC analysis is the κ/λ ratio. This is because:

1. It provides a quantitative assessment of FLC clonality.
2. It has typical high immunoassay sensitivity.
3. Clinical ranges are wide due to immunosuppression of the non-tumour FLCs.
4. There is automatic compensation for variable renal metabolism and changes in blood volume (Chapter 10).

Immunoglobulin heavy chain/light chain immunoassays - “Hevylite” (HLC), have similar analytical advantages. This chapter provides early data on these new immunoglobulin reagents.

32.2. Concept: immunoglobulin heavy chain/light chain assays

Intact immunoglobulin molecules contain unique junctional epitopes between the heavy chain (CH1) and light chain (CL) constant regions (Figure 32.1). These are the target of HLC antibodies. Hence, they can separately identify the different light chain types of each immunoglobulin class, i.e. IgGκ, IgGλ, IgAκ, IgAλ, IgMκ and IgMλ (Figure 32.2). These molecules are then measured in pairs, e.g., IgGκ/IgGλ, to produce ratios of monoclonal immunoglobulin/background polyclonal immunoglobulin concentrations, in the same manner as sFLC κ/λ ratios.

32.3. Antibody specificity

Figure 32.3 Interference data for IgG and IgA HLC assays ^[1]. Data points are the mean values of 3 analyses with 95% confidence intervals shown as bars. Changes in the measured Ig concentrations upon the addition of potentially interfering substances are shown. Concentration of added substances (× normal median serum values) were as follows: IgGκ 11.0 g/L (×1.5); IgGλ 10.5 g/L (×2.5); IgAκ 8.7 g/L (×7); IgAλ 11.6 g/L (×13), IgMκ 4 g/L (×6); IgMλ 4 g/L (×8), FLCκ 45 mg/L (×6); FLCλ 51 mg/L (×4); Chyle 282 formazin turbidity units (FTU) (×25); haemoglobin 4.8 g/L (×100); bilirubin 200 mg/L (×30).

Figure 32.4 A. IgG HLC tests on 146 blood donor sera (95% range) and 245 IgG presentation sera (166 IgGκ, black circles and 79 IgGλ, red circles) from the IFM 2005-01 MM trial ^[2]. Immunosuppression of the non-tumour HLC from hypercatabolism is apparent in high concentration samples. (Courtesy of H. Avet-Loiseau). B. IgA HLC tests on 146 blood donor sera (95% range) and 94 IgA presentation sera (60 IgAκ, black circles and 34 IgAλ, red circles) from the IFM 2005-01 MM trial ^[2]. All MM samples were abnormal by HLC, while 31 of the samples could not be quantified by SPE because of overlying protein bands or low concentration. (Courtesy of H. Avet-Loiseau).

Inevitably, one of the most demanding aspects of HLC assay production is ensuring good specificity. As for FLC immunoassays, the reagents are polyclonal antibodies produced in sheep. Immunisation and subsequent purification techniques are designed to ensure no cross-reactivity. For example, IgGκ reagents do not cross-react with kappa, either free or bound to other heavy chain classes, or IgGλ. Cross-reactivity with non target immunoglobulins and interference is tested as part of the validation process (Figure 32.3). There are 4 HLC epitope regions per molecule - one on each side of the heavy chain / light chain contact regions. Because there are 4 per molecule, immune complexes readily form to produce good homogeneous immunoassays that are suitable for nephelometry and turbidimetry. Latex enhancement is not necessary for IgG and IgA HLC assays, but is required for IgM and IgD assays and may be required for analysis of cerebrospinal fluid samples.

32.4. Normal ranges of Hevylite assays

Table 32.1 Normal concentration ranges of HLC immunoglobulins and HLC ratios in blood donors. These data were generated using the Binding Site SPAPLUS analyser.

Figure 32.5 Pearson rank correlations of IgG, IgA and IgM to summation of their respective Hevylite pairs; Figures A, B and C respectively.

Intact immunoglobulin concentrations are normally controlled within narrow limits, as are their HLC κ and λ subsets ^[1][3]. The results from testing blood donor panels are shown in Table 32.1 and Figures 32.4A and 32.4B. Pearson rank correlations for summation of IgGκ+IgGλ Hevylite samples to total IgG was ~ 0.9 (p<0.01), IgAκ+IgAλ to total IgA was 0.9 (p<0.01) and IgMκ+IgMλ to total IgM was 0.9 (p<0.001) (Figure 32.5).

Ranges that include older individuals, hospital populations and patients with chronic infections and autoimmune diseases are required. Initial studies have indicated that HLC κ/λ ratios in diseases with raised polyclonal immunoglobulins are maintained within the narrow normal limits observed for blood donors (as with sFLC κ/λ ratios).

32.5. Clinical sensitivity of Hevylite assays for monoclonal gammopathies

Table 32.2. Concentrations of serum Igs and other proteins in 339 patients with MM showing median values, 95% limits and total ranges.

Figure 32.6 Plot of IgMκ v IgMλ for blood donor sera (solid black squares), 8 IgM IFE-positive amyloid patients (solid red circles SPE positive) ^[4]. The parallel lines indicate the 95% range for IgMκ / IgMλ ratios. (Courtesy of A. Wechalekar).

Figure 32.7 Plot of IgAκ v IgAλ for 146 blood donor sera (solid black squares), 14 IgA IFE-positive amyloid patients (solid red circles SPE positive, hollow black circles SPE negative) and 5 IgA amyloid patients where clonality was only detectable using IgAκ / IgAλ ratios (solid blue circles) ^[4]. The parallel lines indicate the 95% range for IgAκ / IgAλ ratios. Numbers correspond to positions on the SPE gel in Figure 32.9. (Courtesy of A. Wechalekar).

Figure 32.8 Plot of IgGκ v IgGλ for 146 blood donor sera (solid black squares), 58 IgG IFE-positive amyloid patients (solid red circles SPE positive, hollow black circles SPE negative) and 4 IgG amyloid patients where clonality was only detectable using IgGκ / IgGλ ratios (solid blue circles) ^[4]. The parallel lines indicate the 95% range for IgGκ / IgGλ ratios. Numbers correspond to positions on the SPE gel in Figure 32.9. (Courtesy of A. Wechalekar).

Figure 32.9 SPE of National Amyloid Centre samples ^[4]. Normal human serum controls (lanes 1 and 16), 7 IgA (lanes 2-8) and 4 IgG (lanes 9-12) IFE-positive patients whose monoclonal band was not accurately quantifiable by SPE densitometry and 9 IFE-negative patients (lanes 13–22). (Courtesy of A. Wechalekar).

Serum samples from patients with multiple myeloma (MM) enrolled onto the IFM 2005 trial (courtesy of H Avet-Loiseau and the IFM, France; Figures 32.4A, 32.4B and Table 32.2) and AL amyloidosis (courtesy of P Hawkins and the UK, National Amyloidosis Centre; Figures 32.6 to 32.9) were tested using the IgG, IgA and IgM HLC assays. The results are shown as HLC κ/λ plots (as for FLC assays) alongside normal (blood donor) samples to show normal ranges.

As with FLC assays, in the majority of MM patients the isotype specific monoclonal protein production level was greater than the upper limit of the normal range (Table 32.2). In addition, all of the patients tested had the appropriate abnormal HLC κ/λ ratio and matched IFE sensitivity. The suppression of the non-tumour isotype matched immunoglobulins (eg. suppression of IgGκ by an IgGλ tumour) and so abnormal ratios were present even when immunoglobulin concentrations were within the normal range. The degree of suppression varied greatly in individual patients although suppression was generally greater with IgA-producing tumours compared with IgG tumours (Table 32.2). However, the correlation between suppression and production was greater in IgG patients (IgGκ r = -0.456; p=8.7x10^-10, IgGλ patients r = -0.310; p=0.005) than in IgA patients (IgAκ r = -0.28; p=0.031, IgAλ r = -0.33; p=0.05). The correlation was greater for IgG patients than IgA patients due to saturation of the FcRn receptor; thus, greater concentrations of monoclonal IgG directly result in more rapid catabolism of any polyclonal IgG. Approximately 33% (31/94) of IgA patients could not be accurately quantified by SPE because of their co-migration with other serum proteins or low concentration, but all had abnormal HLC ratios.

Serum samples were also tested in patients with AL amyloidosis to assess sensitivity compared with IFE ^[4]. Patients were categorized as having a detectable monoclonal protein in serum and/or urine, or no detectable monoclonal protein and normal sFLC ratio. Initially 8 IgM, 14 IgA and 58 IgG IFE-positive sera were tested. All eight patients had IgM proteins quantifiable by SPE, all of whom had an abnormal IgM HLC ratio (Figure 32.6). Seven of 14 IgA patients had monoclonal proteins quantifiable by SPE, whilst a further 7/14 patients had monoclonal proteins that were hidden/non quantifiable; the IgA HLC ratio was abnormal in all 14 cases (Figure 32.7). Fifty-four IgG patients had monoclonal proteins quantifiable by SPE, and in 4 patients the concentration was below the sensitivity of SPE. Fifty-four of 58 patients had abnormal IgG HLC ratios, whilst in 4/58 patients the HLC ratio was normal; in all cases the monoclonal protein level was not quantifiable by SPE densitometry (Figure 32.8). The normal ratio here is likely to be because of very low monoclonal protein production (below the sensitivity of SPE) against a normal polyclonal background. Among 46 AL amyloidosis patients with no detectable serum (Figure 32.9) or urinary bands and a normal sFLC ratio, the HLC ratio was abnormal in 9 cases (19%) identifying 2 IgAκ, 3 IgAλ, and 4 IgGκ clones (Figures 32.7 and 32.8).

These results show that HLC assays can quantify monoclonal proteins and can identify monoclonality (by abnormal HLC ratios) even when SPE and in some instances IFE are negative. This high sensitivity should help clarify disease status in many patients with subtle monoclonal gammopathies.

32.6. Hevylite assays for monitoring monoclonal gammopathies

Figure 32.10 Monitoring a patient with IgGλ MM using IgG Hevylite ratios (IgGλ/IgGκ: pink), scanning densitometry (SD: brown) and nephelometry (N: blue). NR: Normal Range, CVAMP: cyclophosphamide and vincristine, adriamycin, melphalan, methyl-prednisolone.

Figure 32.11 Monitoring a patient with IgGλ MM using IgGλ Hevylite (blue) and IgG Hevylite ratios (IgGλ/IgGκ: pink). Comparison with SPE is shown and see Figure 32.10 for other details. NR: Normal Range.

Figure 32.12 Monitoring a patient with IgGλ MM using IgGκ Hevylite (blue) and IgG Hevylite ratios (IgGλ/IgGκ: pink). Comparison with SPE is shown and see Figure 32.10 for other details. NR: Normal Range.

There are several reasons (listed below) why HLC assays might be useful for monitoring patients with monoclonal gammopathies:

In order to determine these features of HLC assays, we assessed 9 patients with IgG MM and 5 with IgA MM who were undergoing treatment in the UK MRC VII trial. For the IgG patients (4 IgGκ and 5 IgGλ), 25 samples were available. In 4/4 who did not achieve complete response, the ratio remained abnormal throughout. In 3/5 patients achieving complete remission, IgGκ/λ ratios became abnormal and indicated relapse earlier than IFE measurements.

For the IgA patients (4 IgAκ and 1 IgAλ), all of the 26 samples that were positive by IFE had abnormal IgAκ/λ ratios. In 2/5 patients, abnormal ratios indicated residual disease when IFE was negative. For one patient the IgA monoclonal protein was obscured by another protein in the SPE gel but could be monitored by HLC ratios. In a second patient, abnormal HLC ratios indicated a slow relapse more than a year before IFE became positive.

One IgG MM patient was studied in detail during 2 remissions and relapses, and illustrates the main features of HLC assays (Figures 32.10 to 32.12) ^[1]. HLC ratios had a greater range of values than IgG quantitation by scanning densitometry or nephelometry and were more sensitive during remissions, identifying relapse earlier than other methods. Of particular interest were the HLC results during the second course of chemotherapy. IgG measurements indicated a tumour response but HLC ratios did not change, indicating no selective tumour cell killing. This was in contrast to the first course of chemotherapy that had huge selective tumour cell killing. Indeed, the patient terminally relapsed after the second round of chemotherapy. This suggested that the HLC ratio provided the correct interpretation of the lack of response to the chemotherapy. The discrepancy between total IgG measurements and IgG HLC ratios may, in part, be due to inhibition of the FcRn receptor by the chemotherapy. This would cause a fall in total IgG (because of faster turnover) but IgG HLC ratios would be unaffected.

Measurement of HLC IgGλ (Figure 32.11) did not provide more information than total IgG. IgGκ quantitation showed the functional activity of the bone marrow plasma cells, the response to chemotherapy and the subsequent tumour relapse (Figure 32.12). However, it was the HLC IgG κ/λ ratio that provided the most interesting results.

32.7. Prognostic value of Hevylite assays in monoclonal gammopathies

Figure 32.13 Kaplan Meier survival curves for 308 MM patients. Heavy/light chain values were measured in presentation sera and progression free survival (PFS) calculated ^[5]. (Courtesy of H. Avet-Loiseau). A. Patients with involved monoclonal protein values in upper tertile (red, n = 101) were compared with those in the lower two tertiles (blue, n = 207). Increased concentrations of involved monoclonal protein were weakly associated with shorter PFS (p=0.039). B. Patients with heavy/light chain ratios in the upper tertile (red, n = 101) were compared to those in the lower two tertiles (blue, n = 207). Increased heavy/light chain ratios were significantly associated with shorter PFS (p=0.0002). C. Effect of increasingly abnormal IgG ratios on the relative risk of progression free survival. Increasingly abnormal Hevylite ratios were associated with decreased PFS (p=0.0001). D. Patients with uninvolved polyclonal protein values in the lower tertile (red, n = 101) were compared to those in the upper two tertiles (blue, n = 207). Decreased concentrations of polyclonal isotype specific protein were associated with shorter PFS (p=0.002).

Figure 32.14 Kaplan Meier survival curves for 79 IgA and 208 IgG MM patients ^[5]. Patients with no immunoparesis (blue) and with systemic immunoparesis (red) were compared. Systemic immunoparesis was defined as a reduction of the immunoglobulin measurement 33% below the normal range. A and B: decreased IgG or IgM concentrations in IgA MM patients were not associated with shorter progression free survival (PFS) (p=0.169 and p=0.477 respectively). C and D: decreased IgA or IgM concentrations in IgG MM patients were not associated with shorter PFS (p=0.952 and 0.977 respectively). (Courtesy of H. Avet-Loiseau).

Figure 32.15 A. Kaplan Meier survival curves for 310 MM patients ^[2]. International Staging System criteria (Stage I β₂M <3.5mg, albumin >35g/L, blue; Stage II not Stage I or Stage III, green; Stage III β₂M >5.5mg/L, red) were used. Patients with Stage I, II or III disease had significantly different progression free survival times (p=0.023). (Courtesy of H. Avet-Loiseau). B. Kaplan Meier survival curves for 310 MM patients ^[2]. Extreme HLC ratios (>200 or <0.01) were combined with β₂M >3.5mg/L to produce a three tiered risk stratification model. Patients with 0 (β₂M <3.5mg/L and HLC ratio ≥0.01 to ≤200: blue); 1 (β₂M >3.5mg/L or HLC ratio <0.01 or >200: green); or 2 (β₂M >3.5mg/L and HLC ratio <0.01 or >200: red) risk factors had significantly different progression free survival times (p=0.000013). (Courtesy of H. Avet-Loiseau).

Several recent studies have assessed the relationship between HLC ratios and outcome in MM, both at presentation and at maximum response. Avet Loiseau ^[2][5] et al. investigated the prognostic value of baseline HLC ratios in 339 patients (166 IgGκ/79 IgGλ; 60 IgAκ/34 IgAλ). The median observed ratios for the individual immunoglobulin isotypes are summarised in Table 32.2. During the study period 125 patients (37%) progressed and 46 patients (14%) died. Progression-free survival (PFS) and overall survival (OS) were used as the time dependent variables. Figure 32.13A shows the relationship between PFS and involved monoclonal protein production (measured by Hevylite) for both IgG and IgA patients. There was a weak correlation (p=0.039) in PFS when comparing patients with involved monoclonal protein concentrations in the upper tertile (red line) compared to those in the lower two tertiles (blue line). However, there was a stronger correlation between HLC ratios and PFS. Figure 32.13B compares patients within (blue line) or outside (red line) the ratio range <0.01 to >200 (p=0.0002); approximately two-thirds of the patients had ratios within this range. IgG and IgA HLC ratios were also analysed individually to predict the relative risk of progression. Increasingly abnormal IgG HLC ratios were associated with an increased risk of progression (Figure 32.13C, p<0.0001), but this did not apply to IgA HLC ratios (p=0.32). It is likely that the IgA HLC ratios would have been similarly associated with poorer survival if more patients had been included. HLC ratios are determined by the concentration of both the tumour-derived immunoglobulin and that of the non-tumour immunoglobulin of the same heavy chain class. Whilst the monoclonal, tumour immunoglobulin concentration was significant (p=0.039, Figure 32.13A) suppression of the uninvolved polyclonal immunoglobulins (p=0.002, Figure 32.13D) accounted for most of the association between HLC ratios and PFS. As there is no international guideline for describing systemic immunoparesis in this study it was defined as a reduction in the immunoglobulin concentration 33% below the normal range. Polyclonal levels of IgG and IgM in IgA MM patients and IgA and IgM in IgG MM were measured. Figure 32.14 shows the relationship between patients with no immunoparesis (blue lines) and those with systemic immunoparesis (red lines). Decreased concentrations of polyclonal immunoglobulins were not associated with shorter PFS in either IgA (IgG p=0.169, IgM p=0.477) or IgG (IgA p=0.952, IgM p=0.977) MM patents.

The current international staging system (ISS) for MM relies upon serum β₂-microglobulin (β₂M) and albumin measurements (Stage I β₂M<3.5mg/L, albumin>35g/L; Stage II not Stage I or III; Stage III β₂M>5.5mg/L). The correlation between these measurements and PFS is shown in Figure 32.15A (p=0.023). In a Cox multivariant regression model β₂M and extreme HLC ratios were the only independent risk factors identified. Figure 32.15B shows an alternative staging system where albumin is replaced by HLC ratios <0.01 or >200. Patients with 0, 1 or 2 risk factors had significantly different PFS (p=0.000013) ^[2].

IgG but not IgA HLC ratios have been shown to predict malignant transformation in MGUS patients ^[3]. Table 32.3 shows comparison of 105 IgG MGUS samples: 36 with stable disease, 30 initial samples from patients who subsequently progressed and 39 samples collected shortly before malignant transformation was diagnosed. HLC suppression was present in 22%, 53% and 90% respectively, whereas non-isotype suppression was present in 6%, 7% and 46%. Thus, IgG HLC pair suppression was more frequent than suppression of Igs from other heavy chain classes, and MGUS patients who eventually progressed had a 2-fold higher rate of isotype specific suppression than stable MGUS patients. In IgA MGUS patients there was no difference between the degree of non isotype suppression and isotype specific suppression (Table 32.3).

Systemic (or non-isotype matched) immunoparesis is not associated with shortened PFS in MM. The prognostic value of HLC ratios, therefore, supports the existence of separate “niches” within the bone marrow for plasma cells producing IgG or IgA.

Table 32.3 Prognostic value of Hevylite ratios in IgG and IgA MGUS patients ^[3]. Clonal IgG plasma cells appear to suppress other IgG producing non-clonal plasma cells more effectively than IgA or IgM secreting cells and this occurs more frequently with cells that will eventually undergo malignant transformation. * Suppression is defined as below the lower limit of the normal reference range. ** MGUS, no progression (or progression) are from Olmsted County, long-term study of MGUS (mean follow-up = 8 years). *** MGUS, last sample, pre-MM sample are from NIH PLCO cohort. (Courtesy of J. Katzmann).

32.8. Hevylite assays in non-Hodgkin lymphoma

Figure 32.16 A. HLC assays for IgA and B. IgM in NHL (Patient identification for A + B are the same).(Courtesy of G. Pratt)

Quantitative abnormalities of sFLC and/or HLC were identified in 45/93 (48%) patients with non-Hodgkin lymphoma (NHL) compared with 17/93 (18%) by SPE alone. The frequency of abnormalities varied markedly between disease i.e. 8/8 with Waldenström's Macroglobulinaemia, 13/20 (65%) with diffuse B cell lymphomas, 17/27 (63%) with marginal zone lymphomas and 5/17 (29%) with follicular lymphomas (Figure 32.16).

The most frequent abnormalities were in sFLC ratios (22/93) and IgDκ/IgDλ ratios (18/93). For both, the abnormalities predominantly indicated an excess production of κ clones (20/22 for κFLC and 17/18 for IgDκ), while only 1 patient had both sFLC and IgD abnormalities. HLC abnormalities were usually present (37/46 positive patients) in only one immunoglobulin class, as would be expected in monoclonal diseases. Abnormally low concentrations of IgM (with normal IgMκ/IgMλ ratios) were found in 28% (26/93) of the sera. This degree of immunoparesis was not seen with the other immunoglobulins or sFLCs. Further studies will determine whether sFLC ratios and HLC ratios have utility for prognosis or disease monitoring in NHL.

32.9. Hevylite assays for immunohistochemistry

Figure 32.17 Immunohistochemistry of a lymph node stained with peroxidase labelled IgGλ Hevylite reagent.

Figure 32.18 Immunohistochemistry of a lymph node stained with peroxidase-labelled IgMκ Hevylite reagent.

HLC antibodies may also find a use in immunohistochemistry (Figures 32.17 and 32.18) for assessing immunoglobulin light chain subsets. Commonly, lymph node, bone marrow and other tissue biopsies contain B-cells and plasma cells that need to be tested for clonality. Measurements of total light chain ratios are not always reliable and in any case, are not specific for each of the IgG-, A-, M-, or D-producing cells. These subsets can be evaluated using double staining techniques but such methods are cumbersome. HLC reagents labelled with peroxidase or fluorochromes could be used to provide sensitive analysis of clonality in a variety of immune derived tumours.

32.10. Publications on FcRn receptors and related issues

32.11. Publications on Hevylite assays

References