Procedures for the Evaluation of Monoclonal Immunoglobulins

Migration of proteins in the electrophoretic field depends on several factors, including the isoelectric point of the molecule, the concentration of electrolytes and pH of the buffer, the temperature of the gel during the electrophoresis, the characteristics of the gel, and the amount of current applied.

The support medium (cellulose acetate, agarose, or the inner lining of capillaries used in capillary zone electrophoresis [CE]) has a negative charge relative to the buffer. The buffer cations flow towards the cathode creating an endosmotic flow. Because of this reaction, some proteins that are weakly negatively charged and that would be expected to migrate toward the anode are found cathodal to the origin after electrophoresis is complete. Indeed, in CE, the endosmotic flow is responsible for pulling the proteins past the photodetector.

Following electrophoresis in gel-based systems, the proteins are fixed by an acid solution, washed, and then stained by a protein dye. Several dyes may be used, including ponceau S, amido black, and Coomassie blue. After a series of destaining washes to remove background staining, the gels are dried and a densitometric scan may be performed to quantify the proteins present in the protein fractions that can be separated by these methods.

Low-resolution serum protein electrophoretic techniques are those that result in a 5-band electrophoretic pattern. This pattern does not have precise separation of the beta-1 (transferrin) and beta-2 (C3) bands. Typically, the anodal and cathodal sides of the bands are less distinct than in the high-resolution techniques. Support media has usually consisted of cellulose acetate, but some commercial agarose methods also provide only low resolution of serum protein bands. Low-resolution gels are a poor choice. Equipment and reagents from several commercial sources are available to provide these low-resolution electrophoretic results using either cellulose acetate or agarose gels. There is usually little or no control of the temperature during the electrophoresis. As the temperature of the apparatus rises during electrophoresis, the random motion of molecules increases and the resolution of the individual bands worsens.

These techniques are usually relatively easy to set up. The typical gel accommodates between 8 and 12 samples, although this number varies among devices provided by different manufacturers. Performance of the assays is relatively straightforward, and technologists are protected from exposure to electrodes with current in all methods (Table 1).

High-resolution techniques are those that provide crisp separation of transferrin and C3 bands in the beta region. The degree of resolution not only permits a better opportunity to detect beta-migrating monoclonal proteins (M-proteins), but also permits detection of small M-proteins that may migrate in the alpha-2 or gamma regions as well. The improved resolution results from a combination of factors, including control of temperature, the use of calcium lactate in the buffer, control of the thickness of the gel, and the method of sample application.

Some systems separate 10 to 12 fractions of proteins, including visualization of transthyretin (prealbumin), albumin, alpha lipoprotein, α₁-acid glycoprotein (orosomucoid), α₁-antitrypsin, α₂-macroglobulin, haptogloblin, transferrin, beta lipoprotein, C3, fibrinogen, C-reactive protein (when markedly increased), and gamma globulin (largely representing immunoglobulin [Ig] G).1 Yet, for practical purposes, many of these proteins are either not seen as discrete bands or are present in quantities too low for reliable use in clinical diagnosis. Transthyretin and C-reactive protein are present in such small quantities that they are better quantified by other assays, such as nephelometry. Alpha lipoprotein produces a broad diffuse band that usually migrates between albumin and α₁-antitrypsin. In patients receiving heparin therapy or patients who have activation of their lipoprotein lipase, alpha lipoprotein migrates together with or may be anodal to the albumin band. Beta lipoprotein is often an irregularly shaped band that migrates anywhere from the alpha-2 to the beta-2 region. There are better tests available for evaluating lipid abnormalities. α₂-Macroglobulin and haptoglobin usually overlap too much to allow distinct comments about each entity.

Therefore, even when using systems that may allow demonstration of 10 to 12 bands, most laboratories scan only 5 or 6 regions by densitometry. Most laboratories do not include transthyretin in the densitometric information. Although C3 and transferrin are separated by these techniques, C3 is a relatively labile protein that breaks down into components that tend to migrate in the beta-1 region. Therefore, unless the storage conditions are well controlled, the beta-2 region may be decreased and the beta-1 region increased owing to this factor. Even with these concerns, gels with better resolution provide an advantage when evaluating sera that contain small monoclonal gammopathies, or that contain monoclonal gammopathies that migrate in the beta or alpha regions.

These techniques are somewhat more demanding in terms of sample application and temperature control than those using low-resolution gels. The typical gel accommodates 8 samples. Performance of the assays is relatively straightforward, although great care must be exercised with the initial application and details of temperature regulation require attention. Yet, detecting low levels of M-proteins is so important for the differential diagnosis of plasma cell dyscrasias that high-resolution techniques are recommended.

At least 2 automated gel-based electrophoresis systems are available and currently approved by the Food and Drug Administration for use with serum protein electrophoresis, namely, the Helena Rep (Helena Laboratories, Beaumont, Tex) and the Sebia Hydrasis (Sebia, Issy-les-Moulineaux, France) systems.

Both techniques use a Peltier cooling device to regulate temperature. Much of the operation of both systems, from the application of the sample to the agarose, through the staining and washing of the gels, is performed by the instruments without significant intervention by the operator. After staining, both systems have densitometric devices with convenient software. The Helena Rep Unit provides a fair degree of separation, whereas the Sebia Hydrasis has 2 types of gels that provide a pattern similar to the high-resolution gels.

The Beckman CZE 2000 is the only currently available CE system that is approved by the Food and Drug Administration for performing serum protein electrophoresis.2 For CE, serum is placed into plastic sample wells that accommodate 7 sera. These are aspirated into a narrow-bore (about 50-μm) capillary tube. By virtue of strong endosmotic flow in these narrow, negatively charged capillaries, the proteins migrate toward the cathode during electrophoresis. The protein fractions are quantified using the optical density at about 200 nm (peptide bond absorbance), and an electropherogram is produced. Since there is no staining or destaining, the operator is only required to put reagents on the instrument, review the electropherograms, and perform maintenance on the instrument. About 40 samples can be processed in an hour. The instrument allows the operator to convert the electropherograms into virtual gel images that resemble the high-resolution gel-patterns and provide crisp separation of the beta-1 and beta-2 regions (Figure 1). The gamma region is relatively short on this system compared with the high-resolution gel-based systems, and reports have appeared of monoclonal gammopathies being missed when they migrate in the slow gamma region.3,4 Capillary electrophoresis is relatively new for general application, so its advantages and disadvantages will become more apparent as peer-reviewed articles appear in the next few years. Figure 1 shows the same sample assayed by CE and beta 1,2 agarose system by Sebia. The CE technique is cost-efficient in laboratories that process at least 40 serum samples per day. Because it is not yet approved for urine or cerebrospinal fluid, it may not be as practical for smaller laboratories.

With the manual techniques, the low-resolution systems require less technical time but produce less precise separation of components than the high-resolution systems. Why would one prefer the more complex, more costly high-resolution gels?

Although both systems are comparable in their ability to detect the large monoclonal spikes that are present in most cases of multiple myeloma or Waldenström’s macrogloblinemia, they differ in their capacity to detect small quantities of M-proteins, especially IgM and IgA M-proteins that migrate in the beta region.

Some objective information exists to support the use of the high-resolution systems for detection of monoclonal gammopathies. A survey by the College of American Pathologists in 19915 sent a specimen of serum with a subtle, but distinct M-protein to multiple participants (Table 2). Only 2 groups of laboratories in that survey were using high-resolution techniques. When their results were combined, 25 (96.2%) of the 26 laboratories using 1 of the 2 high-resolution techniques detected this monoclonal protein. The lower resolution systems accounted for the majority of the results, which included considerable variation in detection of the M-protein. None approached the sensitivity of the high-resolution system. Indeed, only 28% of laboratories using one of the low-resolution gels were able to detect this monoclonal protein. The best performance by low-resolution systems had 66% of their laboratories detecting the M-protein.

What is the significance of the detection of a small M-protein? Whereas the most common cause of small M-proteins is monoclonal gammopathy of undetermined significance (MGUS),6 it is often difficult for the laboratory to determine whether there is an underlying pathological condition that requires more immediate clinical attention. In one case seen in our laboratory (Figure 2), only a tiny monoclonal component was present in the beta region, yet, after finding this, I recommended a urine study, which disclosed massive monoclonal free light chain (Figure 2). Therefore, multiple myeloma with light chain production may have minimal or even no significant findings in the serum. Even the presence of a low-level peak has potential implications for diagnosis of lymphoma (if IgM), amyloidosis, or asymptomatic myeloma. Therefore, high-resolution systems have a major clinical advantage and are recommended for routine serum protein electrophoresis.

A urine screen for monoclonal free light chain should always accompany the study of serum in the workup of a patient suspected of having multiple myeloma. The ability to detect urine changes that encourage further investigation may be crucial to the diagnosis of an obscure disorder.

The type of study that would be useful in clarifying the question of the importance of high-resolution electrophoretic systems to detect clinically relevant small M-proteins should compare a low-resolution with a high-resolution system in the same laboratory on the same samples for an extended period of time. There must be clinical information on patients with monoclonal gammopathies. It may take a decade or longer for some MGUS patients to progress to a clinically significant plasma cell proliferative or lymphoproliferative process. In other cases, further clinical follow-up is needed to determine if there is a neuropathy associated with the gammopathy, amyloidosis (which may have a small M-protein), or a more subtle lymphoproliferative disorder that deserves immediate attention.

Until results from such a study are available, the best information we have about the ability of different electrophoretic systems to detect M-proteins comes from survey information collected by organizations like the College of American Pathologists and from our own experience. I recommend using a high-resolution technique that permits the detection of small M-proteins.

There is also no clear-cut answer to the question of whether the detection of extremely low levels of M-proteins is clinically useful. Clearly, the high-resolution systems are able to detect tiny oligoclonal bands that may occur during a variety of infectious or autoimmune diseases. In my own laboratory, when we see a band that is about half of the size of the normal α₁-antitrypsin band (therefore about 50 mg/dL [0.5g/I]), I note that a tiny restriction is present. I also note that the significance of such tiny restrictions is not known, and that a urine study to rule out monoclonal free light chain should be performed. I also state whether the gamma region is increased, decreased, or normal in quantity. A decreased gamma region is more likely to be associated with a lymphoproliferative process than a normal quantity of gammaglobulin. A polyclonal increase in gammaglobulin with a tiny restriction suggests that the restriction is likely part of a polyclonal process. In any case, however, I suggest follow-up studies on serum and urine with evaluation for a monoclonal free light chain.

High-resolution electrophoresis by itself is not the answer. The details about the technique and the experience of the interpreter are important in the detection of monoclonal gammopathies. All electrophoretic systems have unique concerns. For instance, the alpha-lipoprotein region of some gels is quite dense and may interfere with the detection of an α₁-antitrypsin deficiency.7 On some high-resolution electrophoresis gels, the beta-lipoprotein band may simulate or, alternatively, may obscure an M-protein. The C-reactive protein band may resemble a small M-protein in cases with a vigorous acute-phase reaction. Although the resolution in the new CE systems is excellent, there have been troubling reports that even large monoclonal components that migrate in the slow gamma region are missed. By changing buffer conditions, it is possible to detect these proteins.3,4 Unfortunately, the system is not routinely run with more than 1 buffer condition. Furthermore, inexplicable false positives have been reported with CE due, presumably, to the presence in some sera of substances that also absorb at the wavelength used to detect peptide bonds. If a strong suspicion of a plasma cell dyscrasia exists, the clinician should request immunofixation even when the high-resolution electrophoresis screening is negative. Note that immunosubtraction (see “Characterization of Monoclonal Gammopathies: Immunosubtraction”) would be inadequate to increase sensitivity in such cases and is not recommended.

Nephelometric measurement of IgG, IgA, IgM, and κ and λ light chains has been advocated as a method to detect monoclonal gammopathies when used with high-resolution gel electrophoresis.8 This method is valid only in the presence of an obvious M-protein on the electrophoretic gel. Measurement of immunoglobulins is not an adequate screen in the absence of a monoclonal spike in the serum. It should not be used in the absence of an abnormal serum pattern on serum protein electrophoresis and should be considered only as an adjunctive method to characterize the monoclonal protein, never as a screening test.

Immunoglobulins may be quantified by nephelometry or by radial immunodiffusion. Nephelometry is preferred because it is a much more rapid method that provides more objective and reproducible information about the quantity of immunoglobulins present. Radial immunodiffusion requires visual measurement of a precipitin ring by the technologists. Either method, however, can have problems in accurately measuring the quantity of M-protein. Manufacturers have standardized their kits to be used in the equivalence range of antibody-antigen reactions. However, although some manufacturers claim that their kits do not produce false positives or negatives due to nonspecificity of reagents, or antigen and antibody excess, these problems do occur because the kits are standardized against polyclonal immunoglobulin.9 Monoclonal proteins may occur in huge quantities and may have antigenic determinants that are not well represented in the antisera created against polyclonal immunoglobulins. For this reason, it is important that the technologist performing the quantification and the interpreter of the combined electrophoretic and quantitative information understand immunochemistry. If there is any question about the quantity of an analyte, the serum study should be repeated using a more dilute sample to overcome an antigen excess problem. If the quantity of protein in a monoclonal spike as measured by densitometry does not correlate with the nephelometric results, the nephelometric results are usually due to an antigen excess problem.

Immunofixation electrophoresis (IFE) is the method of choice to identify monoclonal components. It can be performed in about 4 hours.10,11 Immunofixation requires placement of a dilution of the patient’s sample in application slits near the cathodal end of the electrophoretic gel. Unlike immunoelectrophoresis, the serum should be diluted prior to placing it on the gel. Some manufacturers have a standard recommended dilution for each analyte. Commonly, they employ a 1:10 dilution for IgG, 1:2 or 1:3 for IgA and IgM, 1:6 for κ light chain and 1:3 for λ light chain. Although these are useful dilutions for most samples, customizing the dilution for each patient optimizes the results. Methods to customize dilutions have been described previously.11 After the appropriate dilution is applied to lanes on the IFE gel, the proteins are separated by electrophoresis. Following electrophoresis, reagent antisera against immunoglobulins of interest are applied to the appropriate lanes and incubated for about 30 minutes. Nonprecipitating proteins are washed away with buffer, and the precipitin bands are stained with an appropriate protein stain.

Because of the geometry of IFE, even small monoclonal proteins are readily recognized. The main difficulty with IFE is determination of the optimal dilution at which to detect the presence of a small M-protein or a massive increase in immunoglobulin (monoclonal or polyclonal). A small M-protein could be missed if too large a dilution of the patient’s serum was used. If a massive increase of immunoglobulin was present, one could have an antigen excess effect where small immune complexes would be washed away. Optimal immunoprecipitin reactions occur with most commercial antisera when the immunoglobulin in the patient’s serum is diluted to about 100 mg/dL. This is a key reason that I recommend performing serum protein electrophoresis before IFE. If a large monoclonal spike is present, the quantity is easily estimated by the densitometric scan. Then an appropriate dilution can be made before IFE is performed. Similarly, if a small M-protein is seen in the gamma region of a sample from a patient with a lymphoproliferative disorder and hypogammaglobulinemia, the serum may need to be diluted only 1:2 instead of the 1:10 often used. By using appropriate dilutions, one will detect subtle M-proteins and not have problems with antigen excess effects.

Although antigen excess is a real problem, I do not wish to overstate its occurrence. In practice, massive M-proteins usually still give a precipitin band, although the center of the band is often washed away when inappropriate dilutions are used (Figure 3). These potential problems may be avoided by a custom dilution of the patient’s serum.1 The simplest and most economical way to estimate the dilution is to look at the serum protein electrophoresis result. The gamma region gives a reasonable estimate of IgG for these purposes. Since acceptable precipitation is produced by most commercial systems when the immunoglobulin concentration is about 100 mg/dL, if the gamma region is 1200 mg/dL a dilution of 1:12 will usually give good precipitation. Most of the light chains are bound to IgG. Since two thirds of the IgG is κ light chain and one third is λ light chain, a 1:8 dilution should be used for κ, and a 1:4 should be used for λ. Lastly, since IgA and IgM tend to be present in concentrations less than 300 mg/dL, I use a 1:2 dilution of the patient’s serum for those analytes unless there is a large spike in the beta region. Then I use the densitometric estimate of that spike as described above for IgG.

The results from IFE should identify a band seen by serum protein electrophoresis. If the reagent antisera used on the initial IFE do not adequately identify the monoclonal component, additional antisera, such as anti-IgD and anti-IgE, should be used. In addition, other dilutions of the serum and reagent antisera may help

Screening for monoclonal components can be performed by IFE using a pentavalent (reactivity with IgG, IgA, IgM, and κ and λ light chain) reagent and by looking for the presence of any monoclonal band. A subsequent IFE with isotype-specific reagents would be required to identify the band found in screening. Similarly, one may screen by applying the patient’s sample to 2 lanes, one for κ and the other for λ. Once again, a band in either lane would require further study to identify the heavy chain associated with the monoclonal band.

Immunofixation does not suffer from the umbrella effect and is readily able to identify small monoclonal gammopathies and double (or biclonal) gammopathies. Indeed, one problem with IFE is how to interpret very small monoclonal gammopathies. In my interpretations, those bands that are smaller than about 50 mg/dL generate a report that advises the clinician about the small size of the band. The report also notes that the significance of such small monoclonal proteins is not always clear and recommends both a urine evaluation for monoclonal free light chain and a follow-up serum study to determine if the identified band declines or increases.

One concern about IFE is that there are no control sera run on the same gel to compare with the precipitation of the immunoglobulins in the patient’s sera. This is another reason why the importance of using the proper dilution is stressed. When the appropriate dilution is used, a diffuse homogeneous precipitate of the normal polyclonal immunoglobulins should occur for each of the major heavy- and light-chain classes. If there is no precipitate, how does one know that the correct reagent (or any reagent for that matter) was applied to the gel? For quality control, the position of the normal immunoglobulins is helpful. Immunoglobulin A should precipitate in the beta region, IgM near the origin, and IgG mainly in the gamma region.

Immunofixation electrophoresis is more sensitive than protein electrophoretic screening techniques. This is why we routinely screen urine samples with both electrophoresis on concentrated urine and by IFE with anti-κ and anti-λ light chains. When the reagent sera react with the M-protein, the amount of protein present in the monoclonal spike doubles or quadruples. This reaction, plus the added contrast when the washing step removes all nonprecipitated protein, greatly enhances the test’s ability to detect relatively small quantities of monoclonal proteins.

False-positive results may occur when reagent antisera contain reactivity against other proteins. For instance, if an antibody to IgM has reactivity against C3, a band will appear in the beta-2 region that could be mistaken for an M-protein. Therefore, all new reagent sera should be tested against normal control serum (and against plasma to be sure that there is no anti-fibrinogen activity).12 

All serum samples with a request for IFE should have serum protein electrophoresis performed, preferably with an electrophoretic system that provides a high-resolution result. If no monoclonal component is seen on the electrophoresis, more clinical information should be requested to assess whether a plasma cell dyscrasia is suspected. If not, IFE is unlikely to yield useful information. Dilutions to use for κ and λ light chain IFE are based on the gamma region densitometric or electropherogram information assuming a 2:1 κ:λ light chain ratio. If no band is identified, no further study is performed on the serum, but a urine study is requested to rule out monoclonal free light chain. If a band is seen on either κ or λ, further IFE studies are performed with antisera against IgG, IgA, and IgM to identify the band. If none of these studies identifies the band, antisera against IgD and IgE are used. Note that although IgD constitutes about 1% of cases of myeloma, IgE myeloma is vanishingly rare. As of 1991, only about 30 cases had been reported in the medical literature worldwide.13 

When CE is used to detect monoclonal gammopathies, immunosubtraction can be used to identify the monoclonal protein.14 For immunosubtraction, one must mix the serum with beads that are coated with specific antibody, then CE is repeated on the absorbed serum. This is performed for each major heavy-chain isotype and κ and λ light chains. The monoclonal protein is removed (subtracted) by the appropriate reagent. For example, after a serum containing an IgG-κ monoclonal protein is incubated with beads coated with anti-κ, the monoclonal spike will be absent when CE is performed on the absorbed serum. However, when the same serum is incubated with beads coated with anti-λ, the monoclonal spike will still be present.

Immunosubtraction performed on CE is an automated technique that requires minimal training or technical time to perform. It is excellent at identifying obvious monoclonal spikes that are detected by CE. Unfortunately, it is not more sensitive than CE, while IFE is more sensitive than serum protein electrophoresis. Therefore, when a small monoclonal component is suspected in a patient with a sensorimotor neuropathy, immunosubtraction on CE does not add information to a negative CE study. Because background is not subtracted out, there is no improvement in contrast by immunosubtraction.

Immunosubtraction is limited by availability of reagents. Reagents are not generally available for identifying an IgE or IgD M-protein. With IFE this can be readily performed with many commercial reagents. Furthermore, the laboratory can readily alter the dilution of reagents for IFE, but with immunosubtraction the conditions are usually immutable.

Immunoelectrophoresis is no longer the preferred method to identify M-proteins in serum or urine. Although it was the original technique used prior to IFE or immunosubtraction to identify a monoclonal component, it is less sensitive, slower, and often more difficult to interpret. Immunofixation or immunosubtraction is recommended to characterize monoclonal components seen on serum or urine protein electrophoresis.

Rarely, one may need to use immunoselection (ISE) to identify patients with heavy-chain disease. In heavy-chain disease, the immunoglobulin heavy chains are produced in excess and do not have light chains attached. The technique was originally used in an immunoelectrophoresis format for detecting cases of alpha heavy-chain disease. This uncommon disease occurs mainly in the Middle East and Mediterranean regions.

Classic ISE incorporated anti-κ and anti-λ light-chain reagents into the agarose when it was in liquid form, but below 60°C. Then, the agarose containing these antibodies was poured onto the supporting membrane. When immunoelectrophoresis was performed on serum or urine, intact molecules of immunoglobulin that contained light chains precipitated around the well of origin. Only molecules of free heavy chain were able to migrate from the well of origin. Following electrophoresis, antisera against the heavy chain of interest were placed in the trough and allowed to diffuse overnight toward the electrophoresed serum. The presence of a precipitin arc identified the free heavy chain.

In Western countries, heavy-chain disease occurs most frequently with gamma heavy chain, namely, Franklin’s disease. It is most commonly seen in serum from patients with B-cell lymphoma or leukemia. Sun et al15 reported a modified ISE procedure that can be used in an IFE format to identify the heavy chain. For this procedure, anti-κ and anti-λ light-chain reagent is placed on an IFE gel in the region of the application. After allowing the antisera to diffuse into the gel, antisera specific to the heavy chain of interest is applied to a more anodal portion of the gel. The patient’s sample is applied in a well at the origin and the serum is electrophoresed. As with ISE in the immunoelectrophoresis format, any immunoglobulin containing either κ or λ light chain precipitate near the origin. Any free heavy chains are able to migrate toward the anode, where they precipitate as a second arc.

Both forms of ISE are complex and require considerable technical training and expertise in clinical immunology.

When following the progress of patients with monoclonal components, the quantity of the monoclonal component provides an objective means with which to evaluate the course of the disease or response to therapy. Quantification is best accomplished by a densitometric scan of the M-protein. One may also quantify the type of monoclonal protein by nephelometry or radial immunodiffusion. These tests will provide information about the total isotype and not just the M-protein. It is important to use the same technique while monitoring a patient. Densitometric scan of the monoclonal spike is the preferred method with which to monitor a patient with a serum spike that has been identified as an M-protein.

We recommend not repeating the IFE on the serum or urine if the monoclonal spike has not changed its migration. Any change in migration, however, or appearance of another spike should result in a reevaluation of the sample by IFE.

For urine samples, the monoclonal spike (or spikes in the case of both monomeric and dimeric monoclonal light chains) representing the monoclonal free light chain should be followed using a 24-hour sample of urine.16 Do not include the intact immunoglobulin that may also be present in the urine specimen in the scan. The clinical relevance lies with the quantity of free monoclonal light chain excreted, not with the intact monoclonal immunoglobulin. To calculate the amount of free monoclonal light chains excreted each day, an accurate 24-hour collection with a densitometric scan of the spike representing the free monoclonal light chains is critical.

When evaluating a patient for the presence of a monoclonal component, all of the laboratory information and any additional clinical information that is available should be reviewed. The immunoglobulin levels should be correlated with the electrophoretic pattern and the IFE result. Some laboratories perform these functions in different locations.17 However, with modern computer systems, information concerning hemoglobin, calcium, and renal function should be readily available to the individual who interprets the electrophoretic patterns. Any inconsistencies in the information require repeat analysis. For instance, if there is a large spike on the serum protein electrophoresis, but the immunoglobulins are all within normal limits, a repeat analysis of immunoglobulins should be performed using dilutions to avoid antigen excess effect. One must also be certain to review simple logistical problems, such as checking the name on the tubes used for analysis. When a discrepancy cannot be resolved, a fresh sample should be obtained for restudy rather than reporting discrepant results.

Any previous immunologic and electrophoretic studies on that patient’s serum and urine should be reviewed to determine if there has been a change.18 Record any increase or decrease in the M-protein since the previous sample. In our laboratory, I like to see a 20% change in the M-component before I note an increase or decrease from the previous sample. If the uninvolved gammaglobulin in the serum has declined to less than the normal range for the laboratory since the previous study, that should be noted as well. If the monoclonal band has changed its migration, or if a second monoclonal band has appeared, a repeat IFE should be performed to identify that band. If the band has disappeared entirely, suggesting a “complete remission” of disease, an IFE should be performed because IFE is more sensitive than serum protein electrophoresis. Furthermore, if the pattern looks dissimilar to the previous result, the clinician should be contacted for a repeat analysis. One may have 2 patients with similar names who have different M-proteins. Alternatively, a sample may be mislabeled. Contact with the clinician may also reveal therapy that explains the result.

A key component of internal quality assurance is direct contact with the clinician about any case that has unusual features or discrepancies in the results. The degree of uncertainty in a case will vary with the clinical situation and the experience of the interpreter. However, the presence of unexplained bands (especially when seen with the new CE techniques), a change in the electrophoretic pattern, or complete disappearance of a large monoclonal band should generate a conversation with the clinician. Contact between the laboratory and the clinician is one of the most overlooked mechanisms of assuring that the laboratory is performing the correct assay on the correct specimen and making the correct interpretation.

Finally, all laboratories should subscribe to an external program to compare their test results against other laboratories using comparable and different methodology. The subjective nature of some of the tests described herein make this type of comparison all the more important. By observing results of other laboratories in surveys (such as the 1991 College of American Pathologists Survey Report EC-075), objective information about differences in methodologies will become apparent. I encourage agencies performing surveys to include at least 1 small M-protein (between 50 and 100 mg/dL) annually to allow laboratories a close look at the sensitivity of their assays for detecting the subtle M-proteins seen in some plasma cell proliferative and lymphoproliferative disorders.