image

11.5.2. Comparison of Hevylite and immmunoglobulin measurements by SPE

Chapter 11

In an in-house study, the IgG iHLC concentration was compared with the monoclonal concentration, determined by SPE scanning densitometry, for 160 IgG MM patients. In the majority of patients, there was good agreement in the concentration determined by the two methods but some samples did show variance (Figure 11.6). Lopez-Anglada et al. [918] reported similar results. Possible explanations for the variance include inaccurate SPE measurements, for example if the monoclonal protein co-migrated with other serum proteins (e.g. transferrin), or dye saturation underestimated the concentration of IgG monoclonal proteins. This is discussed in more detail in Section 17.4. In addition, when a high concentration of polyclonal immunoglobulins is present, Hevylite measurements may be less sensitive than serum electrophoretic techniques for detecting a monoclonal protein (e.g. MGUS; Chapter 13).

Katzmann et al. [38] compared the iHLC concentration with the monoclonal protein concentration determined by SPE for 114 IgG and 41 IgA MM serum samples. Each sample had a band present on SPE (M-spike) and a corresponding abnormal HLC ratio. There was a linear correlation between the iHLC concentrations and the M-spike for both IgG and IgA patients (Figure 11.7 and 11.4B). Other studies have also reported a good correlation of IgA iHLC concentration with the SPE M-spike [918][897][880]. Similar findings were reported by Boyle et al. [917], who compared IgG or IgA M-Spike concentrations with the corresponding dHLC value. Katzmann et al. [38] noted that the correlation of iHLC with total IgA (Figure 11.4A; r=0.97) was better than the correlation of iHLC with the SPE M-spike (Figure 11.4B; r=0.87), and suggested that this was most likely to be due to the difficulty of quantitating β-migrating monoclonal proteins that co-migrate with other serum proteins (Section 17.4). Katzmann et al. [38] study also compared the relative changes in the iHLC and serum M-spike concentration for 13 IgG patients for whom diagnostic and four follow-up samples were available. The relative changes of the mean iHLC and mean M-spike were not statistically different (p>0.75; Figure 11.8).

Boyle et al. [901] compared monoclonal IgM concentrations determined using the involved IgM HLC concentration or SPE densitometry for 66 WM patients with a quantifiable M-spike. The agreement between the two assays was poor, and a higher value was reported by the IgM HLC assays in most cases (Linear regression R2 = 0.49; Passing-Bablok regression y=-4.4 + 1.98x; Bland-Altman systematic bias = 10.1 g/L, Figure 11.9). Similar findings were reported by Manier et al. [681] and Sarto et al. [936]. The lack of agreement between electrophoretic techniques and automated IgM immunoassays has been reported previously [369][902], and was also demonstrated by Boyle et al. [901]. One explanation for these findings is the polymeric nature of IgM molecules, which causes an overread by nephelometric assays [369]. It should be noted that whilst the numerical agreement between IgM HLC and SPE densitometry is poor, there is an excellent correlation between the responses assigned using the two methods (Section 32.4.2).

Figures

References