WE PRESENT a patient with non insulindependent

RENAL BIOPSY TEACHING CASE Monoclonal Gammopathy in a Type II Diabetic: A Case of Determined Significance Nancy J. Gritter, MD, Simin Goral, MD, and Agnes Fogo, MD INDEX WORDS: Monoclonal gammopathy; nephrotic syndrome; type 2 diabetes; membranous glomerulonephritis. WE PRESENT a patient with non insulindependent diabetes mellitus (NIDDM), nephrotic-range proteinuria, paraproteinemia, Bence-Jones proteinuria, and renal insufficiency. The diagnostic challenge was resolved by renal biopsy. This case demonstrates the differential diagnoses of significant proteinuria in patients with NIDDM, which include diabetic nephropathy, other glomerulopathies, and hypertensive renal lesions. CASE REPORT A 54-year-old black man was referred for evaluation of renal insufficiency, nephrotic-range proteinuria, and serum and urine immunoglobulin G (IgG) lambda monoclonal paraprotein. His medical history was significant for NIDDM diagnosed 5 years previously without documented retinopathy or neuropathy, hypertension treated medically for 9 years, hypercholesterolemia, gout, and Dukes B2 adenocarcinoma of the colon treated with a right hemicolectomy 3 years previously; the colon cancer had been monitored with colonoscopies and carcinoembryonic antigen measurements, with negative results. The patient denied bone pain, weakness, fatigue, or anorexia. He had no complaint of easy bruisability, epistaxis, or parasthesias. He did report a weight gain of 50 lb over the preceding 12 months with increasing lower-extremity edema. His sister also had NIDDM and resulting end-stage renal disease requiring hemodialysis. The patient was a retired police officer with a 40-pack/yr tobacco-smoking habit. On physical examination, the patient was noted to be 6 0 tall, to weigh 284 lb, and to have a blood pressure of 140/80 mm Hg. Bedside funduscopy did not reveal diabetic retinopathy. His lungs were clear to auscultation; cardiac examination revealed a systolic ejection murmur, an S4 gallop, and no audible friction rub. He had 2 to 3 pitting edema of the lower extremities. No neurologic motor or sensory deficit was elicited. From the Division of Nephrology, Vanderbilt University, Nashville, TN. Received and accepted as submitted February 11, 1998. Address reprint requests to Nancy J. Gritter, MD, Division of Nephrology, S-3223 MCN, Vanderbilt University Medical Center, 1161 21st Ave S & Garland, Nashville, TN 37232-4823. E-mail: agnes.fogo@mcmail.vanderbilt.edu 1998 by the National Kidney Foundation, Inc. 0272-6386/98/3106-0025$3.00/0 A 24-hour urine collection demonstrated 11 g of protein with a creatinine clearance of 46 ml/min. Urinalysis and urinary sediment examination were significant for 4 dipstick proteinuria, occasional white blood cells and red blood cells, and no casts. Serum immunoelectrophoresis and urine immunoelectrophoresis demonstrated a monoclonal IgG lambda paraprotein; the quantitative IgG level was 1,660 mg/dl (normal, 635 to 1,654 mg/dl). A complete cell blood count showed a white blood cell count of 6,390/µL, a hematocrit of 37%, and a normal platelet count. Complement levels were normal; hepatitis B and C serologies were negative. A chemistry profile revealed normal liver enzymes, cholesterol 262 mg/dl, serum albumin 2.5 g/dl, and total protein 6.2 g/dl. His serum electrolytes, blood urea nitrogen, creatinine, and glucose were measured: sodium 136 mmol/l, potassium 4.6 mmol/l, chloride 115 mmol/l, bicarbonate 22 mmol/l, blood urea nitrogen 29 mg/dl, creatinine 2.4 mg/dl, and glucose 118 mg/dl. His right and left kidneys measured 10.5 cm and 13.3 cm, respectively, by ultrasound. The echogenicity and corticomedullary differentiation were normal. A bone marrow biopsy was not diagnostic of multiple myeloma but revealed moderate monotypic plasmacytosis staining with IgG lambda. Congo red staining for amyloid was negative. The patient was given the hematologic diagnosis of monoclonal gammopathy of undetermined significance (MGUS). The clinical differential diagnosis of the renal disease included a light chain related renal lesion manifested by the serum and urine gammopathy, proteinuria, and renal insufficiency. Diabetic nephropathy was considered less likely given the absence of retinopathy and the short interval since the diagnosis of NIDDM. Other nondiabetic glomerulopathies presenting with nephrotic-range proteinuria were also considered, as was a hypertensive renal lesion, although this was much less likely. An open-renal biopsy was performed. Renal Biopsy The biopsy specimen comprised a large sample of cortex containing approximately 90 glomeruli, of which 15 were globally sclerosed. Remaining glomeruli were mildly enlarged with a mild increase in mesangial matrix and cells with a lobular pattern (Fig 1). The glomerular basement membrane showed extensive holes in tangential sections and small spikes in cross-section with occasional areas of splitting (Fig 2). In areas without spikes or holes, the basement membrane was thick. There was segmental nodular sclerosis in 23 of the glomeruli; there were only small adhesions with overlying visceral epithelial cell proliferation, indicative of a 1058 American Journal of Kidney Diseases, Vol 31, No 6 (June), 1998: pp 1058-1064

MONOCLONAL GAMMOPATHY IN A TYPE II DIABETIC 1059 Fig 1. Segmental nodular glomerulosclerosis and moderate patchy tubulointerstitial fibrosis and atrophy are present. (Periodic acid-schiff stain; magnification 200.) more active sclerosing process. No areas of proliferation, crescents, fibrin, or necrosis were present. Tubules showed moderate patchy atrophy in a zonal distribution with thickened tubular basement membranes and occasional proteinaceous casts without surrounding cell reaction. There was one small focus of intratubular calcification. The interstitium showed patchy moderate, zonal fibrosis with associated interstitial lymphocytic infiltrate, which in some areas was slightly more dense than expected from the degree of fibrosis, and a single focus with occasional polymorphonuclear leukocytes and cell debris in a tubule. Occasional areas showed up to five or six eosinophils per high-power field, but without interstitial edema or a granulomatous inflammation. Arterioles showed mild-to-moderate hyalinization. No sections with both afferent and efferent arterioles were identified. Interlobular arteries showed moderate medial thickening, and large arteries showed mild thickening. Immunofluorescence studies showed very fine granular trace intensity IgG along the capillary basement membrane superimposed on a linear accentuation of the basement membrane (Figs 3 and 4). Trace C3 granular staining was present in a more segmental capillary wall distribution. There was no staining for IgA, IgM, or C4. Kappa and lambda showed equal, very trace, very finely granular, respectively, staining along the capillary basement membrane and no staining of tubular basement membranes. Electron microscopy showed that the glomerular basement membrane was markedly thickened, with a marked increase of the lamina densa (up to three or four times normal). In addition, there were extensive small- to mediumsized dense deposits in a subepithelial distribution with early surrounding basement membrane spike formation (Figs 5 and 6). No endothelial deposits were present. Foot processes were completely effaced. The mesangial matrix was moder- Fig 2. Thickened glomerular basement membrane and small spikes (large arrowhead) and holes in tangential cuts (small arrowheads) are evident. (Jones silver stain, oil immersion; magnification 1,000.)

1060 GRITTER, GORAL, AND FOGO Fig 3. Glomerular basement membrane staining with areas of linear accentuation and segmental granular appearance. (Anti-IgG immunofluorescence; magnification 400.) ately increased without deposits or interposition. No tubular basement membrane deposits were present. Diagnosis The finding of nodular glomerulosclerosis, arteriolar hyalinization, and diffuse, marked thickening of the glomerular basement membrane are characteristic of diabetic nephropathy. In addition, there is early stage II membranous glomerulonephritis, evidenced by small spikes and diffuse, granular, capillary wall IgG deposits, confirmed by the small subepithelial immune complex-type deposits using electron microscopy. No evidence of a paraprotein-related component of renal disease was present; kappa and lambda staining were trace and equal in intensity. No amorphous glomerular or tubular deposits as seen in light chain deposition disease were present; there were no myeloma casts or amyloid fibrillary deposits. There was also evidence of localized acute tubular damage, possibly related to pyelonephritis. Clinical Course After the biopsy, the patient was not treated with prednisone or other immunosuppressant agents. He was started on an angiotensin-converting enzyme inhibitor. His blood pressure and glycemic control were aggressively managed. Three months after the diagnostic biopsy, however, his 24-hour urine protein remained markedly elevated (18 g) and the creatinine clearance was reduced to 35 ml/min. DISCUSSION We present a patient with type II diabetes in whom a renal biopsy was conducted in search of evidence of a nondiabetic renal disease. Specifically, the clinical suspicion was that a light chain related renal lesion would be uncovered. Indeed, a diagnosis other than diabetic nephropa- Fig 4. Oil immersion view of granular capillary wall deposits, typical of membranous glomerulonephritis. (Anti-IgG immunofluorescence; magnification 1,000.)

MONOCLONAL GAMMOPATHY IN A TYPE II DIABETIC 1061 Fig 5. Thickened glomerular basement membrane with small subepithelial deposits and mesangial expansion without mesangial deposits. (Transmission electron micrograph; magnification 5,000.) Fig 6. Close-up of the glomerular basement membrane reveals markedly thickened lamina densa, typical of diabetes, and small subepithelial deposits with small, surrounding basement membrane reaction, giving rise to the spikes and holes seen on Jones stain by light microscopy (see Fig 2). The membranous glomerulonephritis contributes only a minor degree to the overall basement membrane thickening. (Transmission electron micrograph; magnification 11,000.)

1062 GRITTER, GORAL, AND FOGO thy was found; however, neither light chain casts, light chain deposition disease, nor amyloid lesions were manifest. Instead, interestingly, we found yet another histologic process: membranous glomerulonephritis superimposed on a background of moderately advanced diabetic nephropathy. In examining this case, several questions might be addressed. First, what is the prevalence of nondiabetic renal diseases in patients with NIDDM? Second, which patients, if any, should undergo biopsy in search of another diagnosis, and does finding another diagnostic entity impact on the clinical course? Third, what is the determined significance from a renal standpoint of a monoclonal gammopathy of undetermined significance? Previous investigators have attempted to determine how often nondiabetic renal disease occurs in diabetes mellitus. Parving et al 1 conducted a prospective study in NIDDM patients aged younger than 66 years to determine the prevalence and causes of persistent albuminuria. Fifty patients with albuminuria 300 mg/24 hr were identified and a kidney biopsy was performed in 35 (70%). Eight (23%) patients had nondiabetic glomerular lesions. These patients were characterized clinically by absence of retinopathy, low prevalence of neuropathy, and a short time interval between onset of diabetes and onset of albuminuria. 1 This clinical portrait is similar to the case we present here. Other investigators 2-4 who have examined kidney biopsy specimens performed on NIDDM patients suggest that the prevalence rate of nondiabetic disease is even higher (approximately 30%). However, these retrospective studies used selection criteria that had the potential to overestimate nondiabetic kidney disease. Olsen and Mogensen 5 contend that this selection bias inflates the estimated prevalence and that even Parving et al s 1 figure should be closer to 10%. In their Danish study, the rate of nondiabetic renal disease complicating NIDDM is approximately 9%, although the 33 NIDDM biopsy cases analyzed were generated retrospectively from cases referred for a clinical course that were suspicious for nondiabetic renal disease (nephrotic syndrome, increasing creatinine or proteinuria, or uremia.) Another more recent study by Gambara et al 6 reported the heterogeneity of renal lesions in NIDDM. All patients in their series of 52 NIDDM patients were referred because of clinical proteinuria (range, 0.9 to 9.2 g/24 hr) and biopsies were conducted in all of them. Seventeen of the 52 (33%) biopsy specimens showed other glomerular disease superimposed on diabetic glomerulosclerosis; 19 (37%) had lesions typical of diabetic nephropathy; and 16 (31%) showed chronic and nonspecific changes. There were no differences in multiple clinical features among the three groups, including age, duration of diabetes, renal function, urinary protein excretion, and mean arterial pressure. The investigators speculated that the different patterns demonstrated by biopsy may translate into separate categories of disease progression, prognosis, and management strategy among NIDDM patients. The literature supports the potential for dual diagnoses or diagnoses other than diabetic glomerulosclerosis in patients with NIDDM. How do we determine who needs or would benefit from a biopsy? Amoah et al 3 considered multivariate predictors of other renal diseases in patients with both type I and type II diabetes mellitus. Considering only the type II diabetic patients, late age at onset of diabetes (55.1 3.4 years), absence of neuropathy, and white race were all statistically significant multivariate predictors of nondiabetic renal disease. Olsen and Mogensen 5 suggest that those patients with nephrotic-range proteinuria occurring at an unexpected point in the natural course of their diabetes, those patients with a sudden increase in the amount of urinary protein secretion, those patients with rapidly progressive renal failure, and those with other specific nephrologic or urologic symptoms should be considered for biopsy. In type I diabetes mellitus, 90% to 95% of the proteinuric patients will have diabetic retinopathy. The absence of retinopathy would then suggest a diagnosis other than diabetic nephropathy. In type II diabetes mellitus, however, lack of retinopathy is a poor predictor of nondiabetic kidney disease because even in confirmed cases of diabetic nephropathy, retinopathy is not consistently present. In fact, upward of 60% of proteinuric NIDDM patients with biopsy-proven diabetic nephropathy lack retinopathy. 7 For albuminuric NIDDM patients without retinopathy, Parving et al 1 suggest that the chance for diabetic or nondiabetic glomerulopathy is approximately fifty-fifty. The deci-

MONOCLONAL GAMMOPATHY IN A TYPE II DIABETIC 1063 sion of whether to perform a biopsy in type II diabetic patients probably should be made based on other non-eye grounds. The decision of whether to perform a biopsy in search of a nondiabetic renal disease in a NIDDM patient should take into consideration the potential of making an intervention that will alter that disease and improve patient outcome. In their study, Amoah et al 3 looked at treatable nondiabetic diagnoses in NIDDM patients and the immediate and 12-month outcome. Seven of 17 patients with NIDDM and a biopsy-proven nondiabetic renal lesion who were then treated showed clinical improvement (increase in creatinine clearance by 25% and/or decrease in 24-hour urinary protein 50%) initially and at 12 months (one relapse at 10 months). High-dose prednisone was the intervention for a wide variety of diagnoses, including one patient with membranous glomerulonephritis. No clear inference can be made as to the most commonly encountered nondiabetic diagnosis or the efficacy of therapy. Certainly, the finding of membranous glomerulonephritis in this patient with a previous history of colon cancer is intriguing. There is a known association between the presence of nephrotic syndrome and malignancy. 8 Moreover, membranous glomerulopathy appears to be the pathologic abnormality most commonly associated with malignancy, specifically carcinoma, including adenocarcinoma of the colon. 8 Surveillance measures had demonstrated no recurrence of disease to date in our patient, and we had no evidence of significant proteinuria preceding or present at the time of the diagnosis of the colon carcinoma. Another interesting question is whether the membranous glomerulopathy will accelerate the renal function deterioration in this patient and whether it will accelerate the renal function deterioration in this patient and others with dual diagnoses. Suzuki et al 9 examined Japanese patients with NIDDM and diabetes mellitus associated glomerulonephritis (membranous glomerulonephritis and IgA in their series) and suggested that the diabetic glomerular lesions and loss of renal function were independent of the diabetes mellitus associated glomerulonephritis. Finally, we must consider the finding of the MGUS, which prompted the renal biopsy in this case, and the concomitant NIDDM. Groop et al 10 reported the consistent finding of increased urinary excretion of kappa light chain in all types of diabetes mellitus; however, none of the patients had a serum paraprotein. Although the mechanism remains uncertain, the increased urinary excretion may have been an indication of tubular damage and impaired protein reabsorption, perhaps signaling a stage in the sequential development of diabetic nephropathy. 10 Our patient had both a serum and urine lambda paraprotein and had no evidence of a neoplastic disorder of plasma cells, ie, MGUS. This is a disorder whose frequency increases with age and whose prevalence among blacks is significantly greater than among whites for each age group older than 50 years. 11 Certainly, NIDDM is also a disease whose prevalence rate correlates with increasing age. Therefore, perhaps the coexistence of NIDDM and MGUS is simply a coincident finding in an age group in which both diseases are prevalent and no specific causal relationship exists. Moreover, patients in that same age group with renal insufficiency and proteinuria are commonly tested for the presence of serum and urine gammopathies. One might expect that renal disease and the paraprotein are also coincident without a causal link. In our experience (S.B. Shappell and A. Fogo, unpublished observations), in patients with a known paraprotein undergoing renal biopsies, over half have renal lesions with no apparent relationship to that monoclonal process. We expect that the association is causally disparate in our case as well. Importantly, the renal biopsy has determined that the monoclonal gammopathy is not of significance for this patient s renal disease. As in all cases of MGUS, however, the patient should be monitored closely, as reports reveal that approximately 20% of patients develop myeloma, macroglobulinemia, or lymphoma over a subsequent 10- to 15-year period of observation. 12 REFERENCES 1. Parving H-H, Gall M-A, Skott P, Jorgensen HE, Lokkegaard H, Jorgensen F, Nielsen B, Larsen S: Prevalence and causes of albuminuria in non insulin-dependent diabetic patients. Kidney Int 41:758-762, 1992 2. Chihara J, Takebayashi S, Taguchi T, Yokoyama K,

1064 GRITTER, GORAL, AND FOGO Horada T, Naito S: Glomerulonephritis in diabetic patients and its effect on the prognosis. Nephron 43:45-49, 1986 3. Amoah E, Glickman JL, Malchoff CD, Steergill BC, Kaiser DL, Bolton WK: Clinical identification of nondiabetic renal disease in diabetic patients with type I and type II disease presenting with renal dysfunction. Am J Nephrol 8:204-211, 1988 4. Taft JL, Billson VR, Nankervis A, Kincaid-Smith P, Martin FJR: A clinical-histological study of individuals with diabetes and proteinuria. Diabetic Med 7:215-221, 1990 5. Olsen S, Mogensen CE: How often is NIDDM complicated with non-diabetic renal disease? An analysis of renal biopsies and the literature. Diabetologia 39:1638-1645, 1996 6. Gambara V, Mecca G, Remuzzi G, Bertani T: Heterogeneous nature of renal lesions in type II diabetes. J Am Soc Nephrol 3:1458-1466, 1993 7. Schmitz A, Vaeth M: Microalbuminuria: A major risk factor in type 2 diabetes. A 10-year follow-up study of 503 patients. Diabetic Med 5:126-134, 1988 8. Eagan JW, Lewis EJ: Glomerulopathies of neoplasia. Kidney Int 11:297-306, 1977 9. Suzuki Y, Ueno M, Hayashi H, Nishi S, Satou H, Karasawa R, Inn H, Suzuki S, Maruyama Y, Arakawa M: A light microscopic study of glomerulosclerosis in Japanese patients with noninsulin-dependent diabetes mellitus: The relationship between clinical and histological features. Clin Nephrol 42:155-162, 1994 10. Groop L, Makipernaa A, Stenman S, DeFronzo RA, Teppo AM: Urinary excretion of kappa light chains in patients with diabetes mellitus. Kidney Int 37:1120-1125, 1990 11. Singh J, Dudley AW, Kulig KA: Increased incidence of monoclonal gammopathy of undetermined significance in blacks and its age-related differences with whites on the basis of a study of 397 men and one woman in a hospital setting. J Lab Clin Med 116:785-789, 1990 12. Kyle RA: Benign monoclonal gammopathy: A misnomer? JAMA 251:1849-1854, 1984