Monoclonal gammopathy of renal significance (MGRS) is a relatively new concept for patients with renal monoclonal protein deposition (RMPD) (except monoclonal cast nephropathy) and has been used as a reason for nephrologists to obtain a bone marrow biopsy (BMB). It takes a team of pathologists and clinicians to determine when RMPD at our institution can be defined as MGRS.
To identify the proportion of various subtypes of tentative MGRS diagnosed by renal biopsy that can be confirmed as final MGRS after BMB.
One hundred thirty kidney biopsies with variants of RMPD were identified during the past 10 years. Biopsy cases with known myeloma, B-cell lymphoma, or monoclonal cast nephropathy were separated as a heavy-burden group. The remaining biopsies with RMPD were considered tentative MGRS. Their BMB and clinical indices were further analyzed to determine the final percentage of MGRS diagnoses.
Among the 130 renal paraprotein deposition cases, 44 (33.8%) were categorized as the heavy-burden group. In the remaining 86 cases, 33 (38.4%) with subsequent identification of myeloma (>10% of monoclonal plasma cells) or lymphoma in BMB were further considered as heavy-burden cases. Eighteen cases (18 of 86; 20.9%) did not receive follow-up BMB; thus, no further analysis was performed. BMBs diagnosed as either nonmalignant (no plasma cells; 8 of 86 cases; 9.3%) or premalignant (<10% plasma cells; 27 of 86 cases; 31.4%) were confirmed to be final MGRS (35 of 86; 40.7%).
The data indicate that BMB is an important element in the confirmation of MGRS.
Myeloma occurs mainly in elderly individuals. The mean ages for men and women to develop multiple myeloma are 67 and 70 years old, respectively.1,2 Older individuals often have other factors such as hypertension and/or diabetes leading to compromised renal function. Some benign monoclonal gammopathies of undetermined significance are often associated with different types of renal dysfunction or significant proteinuria, which can result from monoclonal immunoglobulin and light-chain deposition—a condition called monoclonal gammopathy of renal significance (MGRS).3–6 In myeloma or lymphoma combined with monoclonal renal disease, monoclonal cast nephropathy, and even Waldenström macroglobulinemia nephropathy, the renal diseases are usually severe. As a result, these diseases are considered heavy-burden diseases beyond the category of MGRS. The remaining types of monoclonal nephropathy can be potentially classified as MGRS.3–6
The diagnosis of MGRS is a multidiscipline clinicopathologic correlation. A patient with either renal failure or proteinuria is initially evaluated by a nephrologist for detecting monoclonal protein in the blood and/or urine. A kidney biopsy is then evaluated by a renal pathologist. If the renal biopsy is positive for one of the MGRS entities, the nephrologist will refer the patient to a hematologist or directly order a bone marrow biopsy (BMB). The BMB is assessed by a hematopathologist, who determines not only the percentage of monoclonal plasma cells but also the morphology of tumor cells, distribution, amyloid, bone changes, vessels, etc. Finally, the nephrologist and the hematologist can decide how to treat the patient with one variant of MGRS after the confirmation by BMB. The goal of this study was, therefore, to provide a flow chart from tentative MGRS diagnosed by kidney biopsy to final MGRS by BMB. This provides members of the physician team (nephrologist, nephropathologist, hematopathologist, and hematologist) a disease landscape regarding what percentage of tentative MGRS can be confirmed as final MGRS following BMB. Proliferative glomerulonephritis with monoclonal immunoglobulin deposits (PGNMID) often lacks a tight correlation between the kidney biopsy findings of monoclonal protein deposition and the serum monoclonal gammopathy and/or the bone marrow confirmation of monoclonal proteins.7–9 PGNMID cases were therefore individually listed to depict various presentations of monoclonal protein deposition in the kidney.
METHODS
General Design
The protocol for the retrospective study was approved by the Institutional Research Board of Beaumont Health System, Royal Oak, Michigan. In total, 130 kidney biopsies with variants of renal monoclonal protein deposition (RMPD) were identified in our medical center out of 3958 kidney biopsies (3.3% of overall cases) during the past 10 years. Biopsy cases with known myeloma, B-cell lymphoma, or monoclonal cast nephropathy were separated as a heavy-burden group (group 1). The remaining biopsies with RMPD were considered tentative MGRS, including monoclonal light/heavy chain deposition disease (L/HCDD), monoclonal amyloidosis, PGNMID, monoclonal proximal tubulopathy, monoclonal cryoglobulinemic glomerulopathy, monoclonal fibrillary glomerulopathy, and immunotactoid glomerulopathy. Their BMBs and clinical indices were reviewed to determine the final counts of MGRS.
Determination of Monoclonality in BMB
EDTA or sodium heparin anticoagulated bone marrow specimens were routinely washed 3 times with 0.1% bovine serum albumin/phosphate-buffered saline wash buffer. After appropriate adjustment of cell concentration, aliquots were incubated with a panel of fluorochrome-labeled antibodies routinely applied in our laboratories for diagnosis. Red cells were removed with ammonia chloride lysis buffer. After the initial staining with appropriate surface markers, cells were washed and exposed to the reagents in the Fix and Perm Permeabilization Kit (Invitrogen, Waltham, Massachusetts) according to the manufacturer's instructions to study for cytoplasmic Ig light chains (κ and λ). Cells were subsequently incubated with monoclonal antibodies for κ and λ light chains. All antibodies were incubated for 15 minutes in dark conditions at room temperature. After washing with PBS, cells were analyzed with a 10-color flow cytometer (Gallios, Beckman Coulter). The routine immunophenotypic analysis for myeloma workup used a panel of antibodies (Beckman Coulter Life Sciences, Indianapolis, Indiana) against surface markers CD3 (clone UCHT1), CD19 (clone HD237), CD20 (clone B9E9), CD38 (clone LS198-4-3), CD45 (clone J33), CD56 (clone 901), CD117 (clone 104D2D1), and CD138 (B-A38). The panel also included light-chain antibodies (BD Biosciences, San Jose, California) murine anti-human κ-FITC (TB28-2) and anti-human λ-PE (1-155-2) to study the cytoplasmic light-chain expression. Based on the expression patterns of surface markers (CD19, CD20, CD45, CD38, CD56, CD117, and CD138) aberrant plasma cells were identified and differentiated from normal plasma cells. Plasma cells or their subsets were also evaluated for their light-chain expression patterns (monotypic or polytypic). Light-chain expression restricted to κ (κ:λ ratio >3:1) or λ (κ/λ ratio <1:3) was defined as a monotypic pattern. The sensitivity of detecting monotypic plasma cells was 0.1%. The percentage of plasma cells infiltrate of bone marrow cellularity was assessed based on CD138 staining by immunohistochemistry on bone marrow tissue (biopsy core and clot). Colorimetric in situ hybridization for κ and λ analysis was performed to help with the assessment of the percentage of monotypic plasma cell infiltrate in the bone marrow when admixed polytypic and monotypic plasma cells were identified by flow cytometry.
Recalculation Following BMB
When BMB revealed either myeloma (≥10% of monoclonal plasma cells) or lymphoma, tentative MGRS cases became the second heavy-burden group (group 2). The tentative MGRS cases without follow-up bone marrow biopsies were not further analyzed. The remaining tentative MGRS cases with either no monoclonal plasma cells (nonmalignant) or monoclonal plasma cells less than 10% (premalignant) were counted as the final MGRS group (group 3). Each category of RMPD, such as monoclonal amyloidosis, is further calculated based on BMB results. Finally, all PGNMID cases were individually listed to show various pathologic and clinical manifestations of this RMPD at our institution.
Statistics
The median age was calculated for each group. In addition, serum creatinine levels were expressed as the mean ± standard error of the mean (SEM) for statistical analysis. Data from 3 groups were compared using a 1-way analysis of variance. A P value < .05 was considered statistically significant.
RESULTS
Group 1 (initial heavy-burden), group 2 (BMB-proven heavy-burden), and group 3 (final MGRS) had similar median ages and relatively evenly distributed male and female patients (Table 1). Group 1 demonstrated significantly higher levels of serum creatinine (mean ± SEM, 5.78 ± 0.64 mg/dL [510.95 ± 56.58 μmol/L]; median, 4.85 mg/dL [404.87 μmol/L]) than the other 2 groups (Table 1). Despite being made up of heavy-burden cases, group 2 (mean ± SEM, 2.52 ± 0.29 mg/dL [222.77 ± 25.64 μmol/L]; median, 1.82 mg/dL [160.89 μmol/L]) had a similar mean level of serum creatinine to group 3 (mean ± SEM, 2.47 ± 0.35 mg/dL [218.35 ± 30.94 μmol/L]; median, 1.85 mg/dL [163.54 μmol/L]).
Clinical Indices Among Initial Heavy-Burden (HB) Group, Bone Marrow Biopsy (BMB)-Proven HB Group, and Final Monoclonal Gammopathy With Renal Significance (MGRS) Group

Among the 130 RMPD cases, 44 (33.8%) were categorized as the heavy-burden group (Figure 1). The remaining 86 cases with other variants of RMPD were further divided into 4 categories based on the BMB. Myeloma or lymphoma was found in 33 cases (38.4%) following the BMB. Eighteen cases (20.9%) did not show follow-up BMB and were thus excluded from any further analysis. BMBs diagnosed as nonmalignant (no monoclonal plasma cells in 8 of 86 cases; 9.3%) or premalignant (<10% monoclonal plasma cells in 27 of 86 cases; 31.4%) were confirmed to be final MGRS, for a total of 35 cases (35 of 86; 40.7%). Therefore, 35 of the total 130 cases (26.9%) with RMPD were ultimately categorized as final MGRS. Among the different categories of tentative MGRS, L/HCDD had the highest positive rate for myeloma found on the subsequent BMB (Table 2). Monoclonal amyloidosis had 48.3% (14 of 29) nonmalignant and premalignant diagnoses after BMB, although BMB also found 51.7% (15 of 29) of this renal entity to show more than 10% of monoclonal plasma cells. PGNMID had a high rate of nonmalignant or premalignant diagnosis by BMB (Table 2).
Flowchart from identification of renal monoclonal protein deposition to the final count of monoclonal gammopathy of renal significance (MGRS); see detailed explanations in the Results section. Abbreviation: f/u, follow-up.
Flowchart from identification of renal monoclonal protein deposition to the final count of monoclonal gammopathy of renal significance (MGRS); see detailed explanations in the Results section. Abbreviation: f/u, follow-up.
Variants of Outcome From Initial Monoclonal Gammopathy With Renal Significance (MGRS) Categories to Final MGRS After Bone Marrow Biopsies

When we counted all cases of PGNMID, 15 kidney biopsies from 13 patients with a diagnosis of PGNMID were identified (Table 3). The patients' ages ranged from 41 to 89 years, with a median of 64 years. Nine were women and 4 were men. Serum creatinine ranged from 0.91 to 9.72 mg/dL (80.44–859.25 μmol/L), with a mean of 2.97 mg/dL (262.55 μmol/L). The urine protein-to-creatinine ratio ranged from 0.23 to 20.01 mg/mg, and 8 of 13 patients had nephrotic range proteinuria (protein-to-creatinine ratio >3). Two of the PGNMID cases belonged to the heavy-burden group (cases 8 and 10; Table 3) because they were known to have either myeloma or lymphoma. Four of 13 cases (30.1%) had no monoclonal proteins in serologic studies, and therefore no bone marrow biopsies were conducted for these cases. Six of 13 cases (46.2%) were negative for any monoclonal plasma cells in their BMBs. Light microscopy revealed either a mesangial proliferative or membranoproliferative pattern with crescent formation in 3 cases (Figure 2, A through D). Most patients had immunoglobulin (Ig) G-κ type PGNMID (9 of 13; 69.2%); other variants included IgG-λ (2 of 13; 15.4%), IgG (1 of 13; 7.7%), and λ (1 of 13; 7.7%) types. The mesangial proliferative pattern of PGNMID showed dominant electron-dense deposits in the mesangial areas, and the membranoproliferative pattern was closely linked with subendothelial deposits by electron microscopy. Two patients had 2 sequential biopsies. Case 1 had 2 native kidney biopsies, and both biopsies found PGNMID, IgG-κ type. Case 12 had 2 posttransplant biopsies, and both biopsies revealed PGNMID, IgG-κ type (Table 3). Case 12's native biopsy was reported as type 1 membranoproliferative glomerulonephritis with equal 2+ κ and 2+ λ staining a few years before the kidney transplantation. Six months posttransplantation, she developed nephrotic-range proteinuria. Two sequential posttransplant biopsies showed a membranoproliferative pattern of glomerulonephritis with positive IgG-κ staining; thus, a recurrent PGNMID, IgG-κ was considered. However, neither her serologic study nor the BMB found monoclonal protein or monoclonal plasma cells to further support that she had monoclonal protein disease, except in her posttransplant kidney biopsies. During the 10-year follow-up of the 13 patients with PGNMID, 11 patients were alive, but 2 patients were deceased.
An example of proliferative glomerulonephritis with monoclonal immunoglobulin deposits (PGNMID) (case 10 in Table 3). The patient had a serum creatinine of 3.93 mg/dL (347.41 μmol/L) and a urine protein-to-creatinine ratio of 20.01 mg/mg. Light microscopy (A) revealed a membranoproliferative pattern with focal crescent formation (white arrow) (periodic acid–Schiff staining). Electron microscopy (B) showed diffuse effacement of foot processes and thickened glomerular basement membranes with double-contoured features and subendothelial electron-dense deposits (orange arrow). Immunofluorescent stains revealed strong 3+ staining for immunoglobulin G (C) and 3+ staining for λ in glomeruli (D), whereas κ stained negatively in the glomeruli (not shown). The overall findings were consistent with a PGNMID, IgG-λ type (original magnifications ×400 [A, C, and D] and ×5700 [B]).
An example of proliferative glomerulonephritis with monoclonal immunoglobulin deposits (PGNMID) (case 10 in Table 3). The patient had a serum creatinine of 3.93 mg/dL (347.41 μmol/L) and a urine protein-to-creatinine ratio of 20.01 mg/mg. Light microscopy (A) revealed a membranoproliferative pattern with focal crescent formation (white arrow) (periodic acid–Schiff staining). Electron microscopy (B) showed diffuse effacement of foot processes and thickened glomerular basement membranes with double-contoured features and subendothelial electron-dense deposits (orange arrow). Immunofluorescent stains revealed strong 3+ staining for immunoglobulin G (C) and 3+ staining for λ in glomeruli (D), whereas κ stained negatively in the glomeruli (not shown). The overall findings were consistent with a PGNMID, IgG-λ type (original magnifications ×400 [A, C, and D] and ×5700 [B]).
DISCUSSION
A definite diagnosis of MGRS is determined by multiple factors such as serologic findings of monoclonal immunoglobulin, free light chains, RMPD, and BMB with monoclonal plasma cells or B lymphocytes. The diagnosis of MGRS can be a temporal or preceding process, as over time, some of these patients will develop myeloma (reclassified into a heavy-burden group) from their previous MGRS status. Therefore, the renal pathologic diagnosis of each monoclonal protein-associated nephropathy, such as “AL amyloidosis, λ type” or “monoclonal light chain deposition disease, κ type,” should be used in the pathologic report, whereas MGRS or heavy-burden status is usually pending additional clinical investigation such as BMB and the patient's follow-up. In our current study, we found that approximately one-third of patients were initially classified into the heavy-burden group as they were known to have medical histories of lymphoma or myeloma or were diagnosed with monoclonal cast nephropathy. In the remaining tentative MGRS cases, one-third were further categorized as a heavy-burden group, as their BMBs revealed more than 10% of monoclonal plasma cells or lymphoma. Approximately 40% of the tentative MGRS cases were confirmed to be final MGRS following BMB.
Our current data indicate that 15 of 18 monoclonal L/HCDD cases (83.3%) had more than 10% monoclonal plasma cells in the BMB; hence, they were reclassified into the heavy-burden group. A significant percentage (15 of 29 cases; 51.7%) of initially identified monoclonal amyloidosis was found to contain more than 10% of monoclonal plasma cells in the BMB. Our data were compatible with the findings of other studies based on a recent review of MGRS.5 We had 5 cases of monoclonal proximal tubulopathy; 2 were found to have monoclonal plasma cells more than 10% in BMB and the remaining 3 were classified as final MGRS. This entity has been classified into 4 subcategories, including proximal tubulopathy without cytoplasmic inclusions, tubulopathy associated with interstitial nephritis, proximal tubulopathy with cytoplasmic inclusions, and proximal tubulopathy with lysosome indigestion.10 The variants of monoclonal proximal tubulopathies have been supported by other studies.5 We also had limited cases of cryoglobulinemic glomerulopathies, monoclonal fibrillary glomerulopathies, and immunotactoid glomerulopathy; therefore, it is difficult to conclude anything specific regarding these variants.
Monoclonal immunoglobulin-associated glomerulonephritis was initially described11 in 11 patients in 1985. Formal recognition and a better understanding of PGNMID were established in recent years.7–9 Renal pathologists occasionally received puzzling feedback regarding the validity of the diagnosis of PGNMID from either nephrologists or hematologists, as one-third of the PGNMID cases had no monoclonal proteins identified in the serology tests, and half of these patients had no monoclonal plasma cells detected following BMBs. Certainly, our findings were consistent with others' findings on the monoclonal status of this entity.7–9,12 In addition, we had a recurrent PGNMID case following kidney transplantation, compatible with a known high recurrence rate for this entity.9,13 As the etiology of myeloma in humans remains unclear, several transgenic models of myeloma, targeting different genetic locations, have been developed in recent years.14–20 Among these mouse models, the XBP1s-transgenic model and the Vk*Myc model have monoclonal immunoglobulin deposition in the kidneys.16,17 The XBP1s transgenic mouse model spontaneously developed myeloma over time. There are deposits of IgG-κ but not λ in glomeruli, and no deposits were present in renal tubules.16 Taken together with the electron microscopic finding of subendothelial deposits into account, the kidney deposition pattern is most consistent with PGNMID,16,21 with a similar pattern to human PGNMID.
This study focused on the correlation of bone marrow and renal pathology findings. There are a few limitations in this retrospective study. There are no long-term follow-up data on those cases with confirmed MGRS diagnosis to see if they have developed any plasma cell dyscrasia or lymphoma. Second, radiologic data were not provided. Given that bone lesions of 5 mm or greater using imaging techniques is one of the recently introduced criteria for diagnosing myeloma,22 additional radiologic evaluation may be necessary in patients with iliac BMB showing plasma cells less than 10%. BMB, as a snapshot itself, without the aid of imaging studies, cannot totally rule out the possibility of lymphoma involving only lymph nodes or localized plasmacytoma, both of which would disqualify a diagnosis of MGRS.
In summary, our study laid out a flow chart to show that 35 of 130 renal biopsies (26.9%) with monoclonal protein deposition were eventually qualified as final MGRS following the BMB and follow-up. Monoclonal L/HCDD had a higher chance of correlation with more than 10% of monoclonal plasma cells in BMB. By contrast, PGNMID had a lower rate of having monoclonal protein in either blood tests or BMB.
References
Competing Interests
The authors have no relevant financial interest in the products or companies described in this article.