Prevalence, clinical aspects, and natural history of IgM MGUS

Abstract

Background:

Waldenström macroglobulinemia (WM) and chronic lymphocytic leukemia (CLL) are related B-cell cancers that share several clinical and biological features. Both WM and CLL have associated precursor conditions: monoclonal gammopathy of undetermined significance (MGUS) of immunoglobulin M (IgM) type and monoclonal B-cell lymphocytosis (MBL), respectively. Recently, a case of MBL with an IgM MGUS was reported, suggesting a close biological relationship between these entities. While much is known about MGUS overall, investigations of IgM MGUS specifically have been fragmentary.

Methods:

In this paper, we review data on the prevalence, clinical aspects and natural history of IgM MGUS, and focus on identifying gaps in our understanding of the complex relationships among B-cell malignancies and their precursors.

Results:

There appears to be marked heterogeneity in the prevalence of IgM MGUS across populations. However, studies have varied in definition, design, laboratory methods, and endpoints. IgM MGUS differs from non-IgM MGUS in certain respects, including prevalence across racial groups, rate of progression and pattern of malignant outcomes. There are limited data regarding the coincident occurrence of IgM MGUS and MBL.

Conclusions:

Future studies incorporating both protein electrophoresis and flow cytometry are needed to define the underlying spectrum and causes of precursor development, risk factors for progression, and markers that distinguish low- and high-risk precursor patients.

Introduction

Waldenström macroglobulinemia (WM) and chronic lymphocytic leukemia (CLL) are related B-cell cancers. Clinically, WM and CLL share common features, including an asymptomatic stage, autoimmune manifestations, and indolent behavior. WM and CLL have been shown to co-aggregate in families (1), and both are incurable with standard therapy. Biologically, they appear to share similar gene expression profiles (2), and both have associated precursor conditions. The precursor for WM is monoclonal gammopathy of undetermined significance (MGUS) of immunoglobulin M (IgM) type (3). For CLL, the precursor state is monoclonal B-cell lymphocytosis (MBL), which has been characterized as the cellular counterpart of MGUS (4,5). MBL has been demonstrated in about 3% of the general adult population aged 50–60 years or older, and in up to 15% of adults from families with multiple affected CLL cases (6–9). Recently, one case of MBL with an IgM MGUS was reported (see Shim et al., this issue) (9, 10). While there are copious data addressing aspects of MGUS overall, much less information is available regarding IgM MGUS. This review will focus on the prevalence, clinical aspects, and natural history of IgM MGUS, with emphasis on those features that are distinctive for IgM.

Diagnostic and clinical aspects

Few malignancies have phenotypic markers associated with cancer risk. MGUS is the prototypical example of a precursor condition for hematolymphopoietic cancers. Based on current classifications, MGUS (including IgM MGUS) is defined by the presence of a monoclonal immunoglobulin in serum at a concentration of 3 g/dL or less; a proportion of bone marrow plasma cells of 10% or less; and the absence of end-organ manifestations attributable to neoplastic proliferation of monoclonal plasma cells (11).

Serum protein electrophoresis (SPE) followed by immunoelectrophoresis (IEP) or immunofixation electrophoresis (IFE) is integral to the recognition, diagnosis and clinical follow-up of MGUS. Methods to identify and type abnormal bands have evolved over time. Early studies relied on low resolution techniques applied initially to paper and later to cellulose acetate and agarose substrates. These assays were supplanted by high resolution methods, which improved detection of M-proteins of small size or that migrate outside the gamma region. The current method of choice to type monoclonal bands is either IEP or IFE. IFE avoids some of the limitations of IEP for identification of IgM (12). Moreover, IFE is more sensitive than IEP (13) for the detection of small monoclonal bands, but the clinical and biological significance of these small bands is unclear (12). Spontaneous disappearance of a measurable M-protein is a rare event (14). Nonetheless, the M-protein associated with MGUS is usually small and frequently below the densitometric threshold for measurement (15, 16). Such small bands may be transient (17, 18); however, progression to B-cell malignancies has been documented in patients with M-proteins <0.5 g/dL (19).

Once diagnosed, MGUS progresses to myeloma or a related malignant condition at a rate of about 1% per year (14). In one large study, the estimated cumulative excess risk of myeloma was similar for whites and African Americans, suggesting that the observed excess risk of myeloma in African Americans is due to the increased frequency of MGUS rather than an increased rate of progression (20). Natural history studies have shown that the monoclonal protein level may remain stable or increase either gradually or precipitously prior to diagnosis of myeloma. Several prognostic markers indicating risk for progression to myeloma have been proposed, including M-protein isotype (14), M-protein size (14), serum free light chain ratio (21), Bence Jones proteinuria (22), reduction of uninvolved polyclonal immunoglobulins (22), and proportion and atypia of bone marrow plasma cells (22, 23).

Prevalence of IgM MGUS in various populations

MGUS overall has a prevalence of about 3% among white individuals 50 years of age or older and is strongly age-dependent with a consistent male predominance (16). The true population prevalence of IgM MGUS is uncertain. No studies have addressed IgM MGUS prevalence specifically. Instead, prevalence and frequency data are derived from studies of monoclonal gammopathy and/or MGUS overall (Table 1), in which IgM M-protein frequency is usually expressed as a relative percent distribution of immunoglobulin isotopes across all MGUS in a given study population.

Table 1.

Studies reporting the proportional distribution of IgM among cases of MGUS

Study	Year	Location	Population	Target Entity	No. of subjects	Age range	No. with MGUS		No. MGUS with IgM
							n	%	n	%
Hospital/Lab-based
Ameis et al. (28)	1976	Canada	All patients referred for EP for suspected M-protein		1255	all	426		63	14.8
Pick et al. (29)	1979	Israel			30000	all	170		0	0.0
Malacrida et al. (30)	1987	Italy	Hospitalized patients having EP	MG & MGUS	102000*	n.r.	261	0.3	32	12.3
Crawford et al. (31)	1987	USA	Well residents of retirement home		111	62–95
Singh et al. (25)	1990	USA	In- and outpatients with quantitatively normal SPE	MGUS	398	≥20	40	10.1	6	15.0
			White		270		21	5.3	5	23.8
			African American		128		19	14.8	1	5.3
McMurdo et al. (32)	1990	Scotland	Emergency admissions to a geriatric ward	MG & MGUS	192	65–99	9	4.7	0	0.0
Blade et al. (33)	1992	Spain	All patients with MGUS diagnosed in hospital	MGUS	128	30–85	128		14	10.9
Bowden et al. (34)	1993	Japan	Well elderly visitors to community center		146	63–95	—
Lolin et al. (35)	1996	Hong Kong	Patients having EP	MG & MGUS	1600	n.r.	87	5.4	9	10.3
Ong et al. (36)	1997	Netherlands	Patients with M-spike	MG & MGUS	1275	all*	713	55.9	34	4.8
Pasqualetti et al. (37)	1997	Italy	All patients with MG and follow-up		263	40–89	263		26	9.9
Vuckovic et al. (38)	1997	Croatia	Patients with MGUS	MGUS	87	median 60	87		5	5.7
Anagnostopoulos et al. (39)	2002	Greece	Patients with no hematologic disorder having EP		1564	>50	60		4	6.7
Roberts-Thomson et al. (40)	2002	South Australia	All patients with new M-protein on EP		613	all	411		87	21.2
Bergon et al. (41)	2007	Spain	Patients having EP within regional healthcare district	MG & MGUS	n.r.	all	290	0.1	33	11.4
Population-based
Axelsson et al. (42)	1966	Sweden	Residents of 4 parishes	MG & MGUS	6995	>25	59	0.8	5	8.5
Carrell et al. (43)	1971	New Zealand	Adult Caucasian residents of Rangoria	MG & MGUS	2192	>21	7		2	28.6
Fine et al. (44)	1972	France	Blood donors	MG & MGUS	13400	25–60	16	0.1	3	18.8
		Paris			10300		7	0.1	0	0.0
		St. Nazaire			3100		9	0.3	3	33.3
Kyle et al. (45)	1972	USA	Southeast Minnesota residents aged ≥ 50 yrs	MG & MGUS	1200	≥50	15	1.3	3	20.0
Pezzoli et al. (46)	1980	Italy	Healthy blood donors from Trieste region	MG & MGUS**	3800	18–65	13	0.34	2	15.4
Saleun et al. (47)	1982	France	Adult residents of Finistère	MG & MGUS	30279	>18	60***	54.1	13	21.7
Cohen et al. (48)	1998	USA	stratified sampling of elderly from 5 counties; blacks oversampled	MG & MGUS**	1732	≥70	n.r.		n.r.	10.0
Study	Year	Location	Population	Target Entity	No. of subjects	Age range	No. with MGUS		No. MGUS with IgM
							n	%	n	%
Kyle et al. (16)	2006	USA	Residents of southeast Minnesota aged ≥ 50 yrs	MGUS	21463	≥50	694	3.2	n.r.	17.2
Landgren et al. (26)	2007	Ghana	Population-based sample with no history of cancer	MGUS	917	50–74	54	5.9	3	5.6
Iwanaga et al. (27)	2007	Japan	A-bomb survivors in Nagasaki	MG & MGUS	52781	>40	1088	2.1	82	7.5
Landgren et al. (49)	2010	USA	Population-based sample with no history of cancer	MGUS	1000 whites	40–79	39	3.9	3	7.7
					996 blacks	40–79	21	2.1	4	19.1

Abbreviations: No., number; EP, electrophoresis; SPE, serum protein electrophoresis; MG, monoclonal gammopathy; MGUS, monoclonal gammopathy of undetermined significance

^*Sera

^**MGUS was inferred from asymptomatic status

^***Clinical data available on 111 patients

Among population-based studies, Kyle et al. reported the largest screening study investigating prevalence of MGUS based on current methodology to date (16). Among the predominately white residents of Olmstead County, MN, all consenting individuals older than 50 years were evaluated by two-stage SPE and IFE. IgM MGUS accounted for 17.2% of all MGUS, resulting in an estimated prevalence of 0.55 per 100 individuals. There was no difference in the frequency of the IgM isotype based on whether patients had a previous diagnosis of MGUS. The distribution of IgM MGUS by age or gender was not reported.

Studies in African American (24, 25), native African (26), and Japanese (27) populations have suggested racial variations in MGUS prevalence (Table 1) (50). In fact, MGUS overall and multiple myeloma have been shown to be 2-fold more common in African Americans compared to whites in the U.S.(50) In a recent study, MGUS was also shown to be more common in Ghanaian blacks compared to U.S. whites (26). Furthermore, a recent large population-based study found that the doubling of MGUS risk among African Americans (versus whites) was virtually unchanged when obesity, education status and income status were in the same multivariate model (49). This finding suggests that the racial difference is not an artifact of differences in socio-economic status, and strengthens the hypothesis that other factors are involved, perhaps including a role for susceptibility genes in myelomagenesis. In contrast to multiple myeloma, WM is only half as common in African Americans as in whites, based on U.S. registry data (51). Thus, there is considerable interest in determining whether racial differences also occur in the frequency of IgM MGUS (Table 1). In the previously mentioned population-based study in Ghana, the prevalence of MGUS overall was 5.8%. IgM accounted for only 5.6% of MGUS, compared to 14% of men of similar age in the U.S. In an earlier hospital-based study, Singh et al. investigated the occurrence of MGUS in 270 white males and 128 African Americans patients (127 male and 1 female) at a Veterans Administration hospital (25). All patients had both serum protein electrophoresis and IFE. IgM accounted for 20.0% (3 of 15) of MGUS among whites and 6.3% (1 of 16) among blacks. In both groups, IgM MGUS was not found before age 60 years. It is important to note that the real excess of IgM MGUS among whites compared to blacks is probably lower than the IgM frequency distributions suggest. In both studies, because the frequency of MGUS overall among African Americans was about twice that among whites, there was a net 1.4-fold excess of IgM MGUS among whites compared to African Americans.

IgM MGUS may be uncommon in Asian populations. A recent study reported a prevalence of 2.1% for MGUS overall, of which IgM comprised 7.5%, among 52,781 atomic bomb survivors in Nagasaki, Japan (27). These data yield an estimated prevalence of 0.16% for IgM MGUS in this population. Men (n=50, 8.7%) were affected 1.4-fold more commonly than women (n=32, 6.2%). Age-specific data for IgM MGUS were not reported. In contrast, a survey of 1600 mostly Chinese patients undergoing EP for suspected monoclonal gammopathy in Hong Kong found 10.3% of MGUS to be due to IgM in that setting (35). Clearly, more data are needed to describe the prevalence patterns of IgM MGUS among different racial and ethnic groups. However, taken together, these data support a role for racial factors, either genetic, environmental, or a combination, influencing development of IgM MGUS.

In addition to racial differences, there is also evidence supporting geographical variation in both population- and hospital-based studies (Table 1). Overall, IgM appears to constitute a larger proportion of MGUS in countries of western European descent, such as the U.S. (16), Canada (28), and southern Australia (40). This observation is not uniform, as much lower frequencies have been reported for Sweden (42) and, more recently, the Netherlands (36). These two reports illustrate the difficulty in generalizing across studies that use different methodologies or case definitions: the Swedish study was based on IEP rather than IFE for identification of IgM, and the Dutch study employed a restricted definition of MGUS compared to many other reports. Based on a limited number of reports, IgM MGUS appears to be substantially less frequent in Eastern Europe (38), Greece (39), and Israel (29) and intermediate in southern European countries such as Spain (41) and Italy (37). Geographical variation also appears within populations of western European descent (42, 44, 47). In particular, a study of blood donors in France demonstrated remarkable variation by region, ranging from 0% (in a group of mostly urban dwellers in Paris) to 33% (in residents of an agricultural district in western France) (44).

The observed variability across studies has yet to be fully explained. Interpretation must be undertaken with caution, since these studies vary in design, study population, endpoint(s), and assay techniques. Several early studies were large surveys in clinically-defined populations, such as blood donors (42, 44, 46) that were pivotal in identifying the subgroup of monoclonal gammopathy that was ultimately designated MGUS. Most are series based on patients referred to tertiary care centers for electrophoretic studies during a diagnostic evaluation because of a known or suspected M-protein (29, 30, 39, 41). Others evaluate defined population of interest, such as atomic bomb survivors (27) or elderly residents in a retirement home (31, 34). A few represent unselected registry-based (36) or geographically-defined (16, 26, 43, 47, 48) populations.

There are also important differences in the definition and documentation of MGUS across studies resulting from changes in the diagnostic criteria over time and differences in access to clinical records. Some studies infer a diagnosis of MGUS based on the ambulatory status and apparent good health of the subjects (31, 34, 42, 48). Others employ a classical definition of MGUS based on specified criteria regarding M-protein level, distribution of bone marrow plasma cells, and absence of manifestations of overt plasma cell disease (16, 30, 38, 39). Still others further restrict their definition of MGUS by excluding M-proteins associated with a variety of underlying conditions, including non-lymphoid neoplasia, autoimmune/connective tissue disease, and infections and/or chronic inflammatory conditions (36).

The literature is further complicated by the evolution of assay methodology over time. The earliest studies were based on low-resolution electrophoresis (42, 44). The current clinical standard is a two-step approach in which serum is subjected to high-resolution protein electrophoresis; if an M-protein is detected, it is further characterized by either immunoelectrophoresis or immunofixation electrophoresis. Immunofixation electrophoresis is more sensitive than immunoelectrophoresis and avoids some well-known limitations of IEP with respect to the detection of monoclonal IgM. Accordingly, it is occasionally used as the primary assay for diagnosing MGUS, whether or not an M-protein is detected by serum protein electrophoresis (25). Finally, after assay methodology is taken into account, individual studies differ in collection and reporting of data elements by specific Ig class, such as age and/or gender.

It is interesting to note that the same concerns over study population, case definitions, and assay techniques are reflected in prevalence studies of MBL (10). To our knowledge, at this time there has been only limited investigation of the coexistence of MBL and monoclonal gammopathies (9, 10, 58).

Distinctive features of IgM MGUS and risk of transformation

Although details of IgM MGUS are not consistently reported, where such information is available, it suggests that IgM MGUS differs from non-IgM MGUS in certain respects. A striking contrast is the marked difference in racial distribution. As mentioned previously, IgM MGUS has now been shown to comprise a much smaller proportion of MGUS overall in Africans (26) and African Americans (25, 49) compared to whites. This relationship is reversed for non-IgM MGUS.

Most importantly, the pattern of outcomes for MGUS appears to differ by immunoglobulin class. Among prevalence studies that report outcome, numbers are small and follow-up is variable. However, whereas IgG and IgA MGUS overwhelmingly progress to myeloma, IgM MGUS has been reported to progress most often to WM or other lymphoproliferative disorders (Table 2). WM is the most common malignant outcome, but there are reports of IgM MGUS progressing to CLL and non-Hodgkin lymphoma (NHL) (40) (Table 2). In a large natural history study, Kyle et al. (3) followed 213 patients with carefully defined IgM MGUS to determine risk of progression to overt lymphoproliferative disease (Table 3). Nearly half of the IgM M-proteins were smaller than 1.0 g/dL. At a median of 6.3 years, patients with IgM MGUS had the highest risk for progression to WM and were also at increased risk for progression to CLL and NHL. Only 1 patient, who had developed an IgM/IgA biclonal gammopathy during the follow-up interval, was thought to have progressed to myeloma. These observations are also noteworthy because of an epidemiological study showing that WM co-aggregates significantly with CLL but not myeloma in families (1). The rate of progression for IgM MGUS was 1.5% per year, compared to 1.0% per year for MGUS overall. Despite these differences, there are also similarities across isotypes. For example, like MGUS overall, IgM MGUS has been shown to increase with age (47), has a male predominance (47), and most commonly presents with monoclonal proteins of small concentration, often less than 0.5 g/dL (15) (Table 3).

Table 2.

Malignant transformation in patients with IgM MGUS

Author	No. with IgM MGUS	Median follow-up	No. progressing to WM (%)	Median time to progression	# other LPD
All MGUS
Cesana (22)	130	67 mo	12 (9.2)	89 mo	—*
Blade (31)	14	56 mo	1 (7.1)	n.r.	—
Anagnostopoulos (37)	4	71 mo	0 (0.0)	n.a.	—
Colls (52)	3	4 yrs	2 (66.7)	3 yrs	—
van de Donk (53)	9	6.75 yrs	1 (11.1)	9.46 yrs	—
Veneri (54)	71	67 mo	8 (11.3)	n.r.	—
IgM MGUS
Kyle (3)	213	6.3 yrs	6 (2.8)	> 5 yrs**	17 NHL, 3 CLL, 3 AL, 1 MM***
Montoto (55)	52	5 yrs	5 (9.6)	3.6 yrs	1 NHL
Baldini (56)	217	56.1 mo	13 (6.0)	n.r.	2 NHL
Morra (57)	452****	49 mo	36 (8.0)	53 mo	2 NHL, 1 CLL, 1 AL, 1 MM

Abbreviations: MGUS, monoclonal gammopathy of undetermined significance; No., number; mo, months; yrs, years; n.r., not reported; n.a., not applicable

^*Cannot be determined from the data

^**For 5 of 6 patients

^***Patient developed a biclonal IgM/IgA gammopathy and subsequently developed IgA myeloma.

^****Included patients with asymptomatic IgM monoclonal gammopathy

Table 3.

Risk factors predicting progression of IgM MGUS to hematologic malignancy^*

	Factors Associated with Risk of Progression					Other Factors Studied and Found to be Not Associated with Risk of Progression
Author	MC size	Hgb	Sex	BM LPC	Other
Kyle (3)	+	−	−	n.d.	albumin	Age, light chain, immunoparesis, BJP
Montoto (55)				+	n.a.
Baldini (56)	+	+	+	−	n.a.	Age, LDH
Morra**	+	+	n.d	−	ALC	BJP

Abbreviations: MC, monoclonal component; Hgb, hemoglobin; BM LPC, bone marrow lymphoid and/or plasma cells; BJP, Bence Jones proteinuria; LDH, lactate dehydrogenase; ALC, absolute lymphocyte count; n.d., not done; n.a., not applicable

^*Table reflects results of multivariate analyses.

^**Study population included patients with asymptomatic IgM monoclonal gammopathy (i.e., some asymptomatic WM) and outcome was WM requiring therapy.

Future directions

Future studies are needed to define underlying causes of IgM MGUS development as well as risk factors for progression from IgM MGUS to full-blown malignancies. Such insights will hopefully facilitate the development of future efforts designed to delay and prevent IgM MGUS transformation into frank malignant processes. Clearly, there is need in the clinic to better define high-risk versus low-risk precursor patients. For example, access to better molecular markers would help physicians to provide more individualized follow-up and also they would guide clinicians regarding the need to consider early initiation of therapy in high-risk patients.

The lack of data regarding the coincident occurrence of MBL and monoclonal gammopathies (particularly IgM MGUS and abnormal free light chain (FLC) patterns) represents a major information gap that should be addressed by including both types of markers in studies whenever possible. The one reported case with detectable MBL and IgM MGUS (9, 10) was later given a clinical diagnosis of WM, but this was not confirmed histologically, and the patient was subsequently diagnosed with a diffuse large cell lymphoma. In stored prediagnostic blood from 109 persons who subsequently developed CLL, the prevalence of an abnormal FLC ratio, M-protein, and hypogammaglobulinemia before CLL diagnosis was 38% (95% confidence interval, 29%−47%), 13% (7%−21%), and 3% (1%−8%), respectively (58). In fact, M-proteins and abnormal FLC ratios were detected up to 9.8 years before CLL diagnosis in a total of 48 persons (44%). Hypogammaglobulinemia was not present until 3 years before the diagnosis of CLL. Among 37 patients with information on tumor cell immunophenotype, an association between immunophenotype and involved FLC (P = .024) was observed (58). Prospective studies on the natural history of co-existing MBL and monoclonal gammopathies should help illuminate the complex relationships among B-cell malignancies and their precursors.