RENAL INVOLVEMENT IN MULTIPLE MYELOMA: A 10-YEAR STUDY

Abstract

Renal involvement in 204 cases with multiple myeloma admitted over a 10-year period to this tertiary care center in north India was retrospectively examined. Renal involvement occurred in 55 cases (27%); the vast majority of whom (94.5%) had presented with renal failure and 7.3% had nephrotic syndrome. The diagnosis of multiple myeloma was made after admission in 51 of the 55 (92.7%) cases. Oliguria was seen in 23.6% and two-third patients required dialysis. Factors precipitating renal failure were identified in 53% and included dehydration (33%), hypercalcemia (24%), nephrotoxic drugs (16%), sepsis (9%), recent surgery (5%) and contrast media (2%). Severe anemia, hypercalcemia, Bence Jones proteinuria and skeletal abnormalities were more frequent in those with renal involvement. Patients with renal involvement were more likely to have a high tumor burden. The myeloma was of light chain type in 68% of those with renal involvement whereas IgG myeloma was commonest (57%) in those without evidence of renal disease. Renal histology was studied in 27 cases with myeloma cast nephropathy seen in over 60%. Tubulointerstitial nephritis was seen in 14% cases, 11% had amyloidosis, 7% had acute tubular necrosis and 3.6% each had nodular glomerulosclerosis and plasma cell infiltration. In 8 cases (14.6%), renal biopsy provided the first clue to the diagnosis of myeloma. Renal function improved in 33% cases. Only 22% of patients on dialysis survived over 6 months. Median survival in those with renal involvement was only 4 months. Development of unexplained renal failure in an elderly individual with normal sized kidneys, in association with disproportionate anemia even in the absence of skeletal lesions should alert the physician to the diagnosis of multiple myeloma.

• Plasma cell dyscrasia
• Renal failure
• Myeloma cast nephropathy

INTRODUCTION

Multiple myeloma is characterized by neoplastic proliferation of a single clone of plasma cells. It exerts its effects locally by producing osteolytic bone lesions and systemically by unregulated production of homogeneous immunoglobulins or light chains. The kidneys are involved in a significant proportion of patients with myeloma who can present with acute or chronic renal failure, nephrotic syndrome, non-nephrotic proteinuria or tubular function defects [1-2]. Azotemia often provides a clue to the diagnosis of myeloma, poses a major management problem, and is an adverse prognostic factor [3-5]. Renal failure is often irreversible [6-9] and the second commonest cause of death after infections [[6]], [[10]]. Even though improvement in supportive care and availability of potent antibiotics and chemotherapeutic agents has resulted in an overall improvement in their survival [[11]], the incidence of renal failure amongst myeloma patients has not altered over the last two decades. The precise mechanisms whereby renal dysfunction develops in multiple myeloma are incompletely understood [1-2], [[10]].

In this retrospective study, we looked at the clinical and laboratory features associated with renal disease in patients with multiple myeloma seen at our institute over a 10-year period. We analyzed the clinical spectrum and renal histological abnormalities and studied the prognostic implications of renal failure in a large group of previously untreated patients seen at a single referral hospital in a developing country.

MATERIAL AND METHODS

The records of all patients admitted to our institute with a diagnosis of multiple myeloma between 1987 and 1996 were reviewed. The diagnosis of myeloma was based on the presence of 2 of the following 3 criteria (1) marrow plasmacytosis >10%, (2) lytic lesions on skeletal radiography and (3) a monoclonal peak on serum and/or urine electrophoresis. Renal failure was defined by a serum creatinine persistently >2.0 mg/dL after correction of volume deficit or urinary tract obstruction. Oliguria was defined by documentation of urine output < 400 mL/day for 2 consecutive days despite correction of dehydration.

Details of history and physical examination were recorded. Investigations included a complete hemogram and serum biochemistry. Serum and urine electrophoreses were performed by agarose gel electrophoresis. The type of monoclonal protein was determined by immunoelectrophoresis. Facilities for differentiation of urinary light chains were not available in this Institute until very recently and data on this was therefore not available in vast majority of our patients. The presence and extent of bony lesions on skeletal survey was recorded and lesions were classified into generalized osteoporosis, lytic lesions or a combination of the two. Renal histology was studied in 27 cases (twice in 1 case). The tissue was obtained by percutaneous kidney biopsy or at autopsy. Biopsy was performed if the cause of renal dysfunction was not clear on clinical grounds and consent could be obtained.

Plasma cell tumor load was defined for each patient as high, intermediate or low according to criteria described by Durie and Salmon [[12]]. Patients received dialysis according to standard clinical criteria. Patients who did not show recovery of renal function after 4 months were referred back to their local hospitals for further dialysis. Chemotherapy consisted of monthly cycles of melphalan (8 mg/m2) and prednisolone 60 mg/m2 for 4 days. The dose of melphalan was reduced by 50% in patients with serum creatinine > 3 mg/dL. Nine patients received a combination of vincristine (0.4 mg/day) and doxorubicin (9 mg/m2/d) for 4 days along with dexamethasone (20 mg/m2/d) for 4 days beginning days 1,9 and 17. In 4 patients, doxorubicin was substituted by cyclophosphamide (3 mg/kg/day for 2 weeks).

Comparison between groups with and without renal involvement was done by Student's t test for paired and unpaired data and Chi-square test with Yates' correction for ordinal data. Survivals with-in groups were estimated using the Kaplan-Meier method.

RESULTS

Out of a total of 204 cases, aged 57.1 ± 10.8 yrs (range 30–82 yrs) with multiple myeloma admitted during the study period, renal involvement was documented in 55 (26.9%) cases. Fifty-two patients, 94.5% of those with renal involvement, presented with renal failure, of which 36 (65.5%) had a serum creatinine >6 mg/dL. Renal failure was oliguric in 23.6% cases. Thirty-six patients required dialysis for intervals ranging from 2 weeks to 11 months. Other manifestations included nephrotic syndrome and renal glycosuria in 4 patients (7.3%) each and non-nephrotic proteinuria in 3 patients (5.5%).

Diagnosis

The diagnosis of multiple myeloma was made after admission to our hospital in 51 of the 55 (92.7%) cases. Three of them had pre-existing hypertension and one had diabetes mellitus, but none had evidence of any pre-existing renal disease. Thirty-six (65.5%) of these patients had been referred primarily for management of rapidly progressive renal failure (serum creatinine >6 mg/dL) of unknown cause. Renal failure developed 3–48 months after the diagnosis of myeloma in the remaining 4 cases. The kidneys, imaged on ultrasonography, were normal sized in 47 and enlarged in 8 cases.

Clinical Features

The most frequent clinical abnormalities in patients with renal involvement were anemia (85%) and bone pains (68%). A comparative analysis of clinical and laboratory parameters (Table 1) revealed that patients with renal involvement were significantly more likely to have anemia and hypercalcemia. These patients also showed skeletal lesions with greater frequency. A serum M-component was detected with equal frequency in patients with or without renal involvement, but Bence Jones proteinuria was present more often in those with renal involvement. The myeloma was of light-chain type far more frequently in patients with renal involvement (68% vs. 22%, p < 0.01) whereas IgG myeloma was the commonest in those who did not show any renal disease (57% vs. 25%, p < 0.01). In the subgroup of patients with histologically proven cast nephropathy, light chain myeloma was documented in 82.4%. Patients with renal involvement were also more likely to have a higher tumor burden (Table 2).

Table 1. Comparison of various clinical and laboratory parameters in patients with and without renal involvement

	With renal involvement (n = 55)	No renal involvement (n = 149)	p value
Age (years)	58.4 ± 11.4	56.2 ± 10.4	>0.05
Sex (M:F)	39:16	107:42	>0.05
Hb < 8 g/dL	47 (85.5)	89 (59.7)	0.001
Hypercalcemia	18 (32.7)	28 (18.7)	0.035
Hyperuricemia	43 (78.2)	100 (67.1)	>0.05
Skeletal abnormalities	46 (83.6)	97 (65.1)	0.01
Lytic lesions	13 (23.6)	23 (15.4)	>0.05
Osteoporosis	17 (30.9)	40 (26.8)	>0.05
Both	16 (29.1)	34 (22.8)	>0.05
Bence Jones proteinuria	42 (76.4)	72 (48.3)	<0.001
Serum M-band	53 (96.4)	140 (93.9)	>0.05

Figures in parentheses are percentages.

Table 2. Tumor burden in patients with and without renal failure

Tumor burden (×10^12)	With renal failure	Without renal failure
<0.6	1 (1.8)	12 (8.1)
0.6–1.2	8 (14.6)	41 (27.5)
>1.2	46 (83.6)^*	96 (64.4)^*

^*p = 0.025

Precipitating Factors

Factors that could have precipitated the development of renal failure were identified in 52.7% cases (Table 3). More than one factor appeared to be responsible in 34.5%. The commonest were dehydration (32.7%) and hypercalcemia (23.6%). Renal failure was precipitated by contrast administration in one case with Bence Jones proteinuria. No precipitating cause, however, was evident In 47.3% cases.

Table 3. Precipitating factors for renal failure

Factors	All cases (n = 55)	BJP +ve (n = 42)	BJP −ve (n = 13)
Dehydration	18 (32.7)	15 (35.7)	3 (23.1)
Hypercalcemia	13 (23.6)	11 (26.2)	2 (15.4)
Nephrotoxic drugs	10 (18.1)	9 (21.4)	1 (7.7)
Aminoglycoside	5 (9.1)	4 (9.5)	1 (7.7)
NSAIDs	4 (7.2)	4 (9.5)	0
Contrast	1 (1.8)	1 (2.4)	0
Sepsis	5 (9.1)	2 (4.8)	3 (23.1)
Surgery	3 (5.4)	2 (4.8)	1 (7.7)
No cause	26 (47.3)	21 (50)	5 (38.4)

More than one factor responsible in 19 (34.5%)

Figures in parentheses are percentages

Histology

Renal histology (Table 4) was studied in 27 cases (twice in one patient). Tissue was obtained by percutaneous renal biopsy in 19 and at autopsy in 9 instances. In 8 cases, renal biopsy provided the first clue to the diagnosis of myeloma. The commonest histological lesion was myeloma cast nephropathy characterized by the presence of eosinophilic intratubular homogeneous casts surrounded by multinucleated giant cells. The casts were laminated or fractured in 14 cases. Varying degrees of flattening, necrosis and atrophy of tubular cells along with interstitial fibrosis were also noted. Tubulo-interstitial nephritis (TIN) alone without the typical casts was seen in 4 cases. The infiltrate consisted mainly of lymphomononuclear cells. Of the 4 cases with a nepthrotic syndrome, 3 showed amyloidosis and one revealed nodular glomerulosclerosis. The diagnosis of amyloidosis was made at autopsy in 2 of these cases. Other organs that showed amyloid deposits included the gastrointestinal tract, spleen, liver and bone marrow. Twelve patients with cast nephropathy and 2 with TIN were dialysis dependent. Renal function improved with treatment in 10 of these patients (Table 5).

Table 4. Histological findings in patients with renal involvement (n = 28^*)

Histology	All cases (n = 28)	BJP+ve (n = 17)	BJP−ve (n = 11)
Myeloma cast nephropathy	17 (60.7)	15 (88.2)	2 (18.2)
Tubular epithelial giant-cell reaction	13 (46.4)	12 (70.6)	1 (9.1)
Tubular atrophy and fibrosis	10 (35.7)	9 (52.9)	1 (9.1)
Interstitial mononuclear cell infiltration	2 (7.1)	2 (11.8)	0
Nephrocalcinosis	3 (10.7)	2 (11.8)	1 (9.1)
Tubulo-interstitial nephritis	4 (14.3)	1 (5.9)	3 (27.3)
Amyloidosis	3 (10.7)	0	3 (27.3)
Acute tubular necrosis	2 (7.1)^*	1 (11.8)	1 (9.1)
Nodular glomerulosclerosis	1 (3.6)	0	1 (9.1)
Plasma cell infiltration	1 (3.6)^*	0	1 (9.1)

^* renal histology studied twice in one patient

Figures in parentheses are percentages.

Table 5. Histology and renal function

Lesion	No. of Cases	Dialysis dependent	No. of cases showing improvement in renal function
Myeloma cast nephropathy	17 (15)	12 (10)	7 (5)
Tubulointerstitial nephritis	4 (1)	2 (1)	3 (1)
Acute tubular necrosis	2 (1)	2 (1)	1 (0)
Amyloidosis	3 (0)	0	0

Figures in parentheses indicate patients who showed Bence Jones proteinuria.

Treatment and Outcome

Six patients died before any specific treatment could be started. Of the remaining 49 cases, 36 received melphalan and prednisolone, 9 were treated with vincristine, adriamycin and dexamethasone and the remaining 4 received vincristine, cyclophosphamide and prednisolone. None received plasmapheresis. Twelve more patients died between 6 weeks and 6 months later. Improvement in renal function was noted in 18 (33%) cases, including 8 who were dialysis dependent. The renal function eventually normalized in 7, including two dialysis dependent cases. Renal biopsy had shown acute tubular necrosis in one and myeloma cast nephropathy in the other. A total of 23 cases continued to require dialysis despite chemotherapy at 4 months. Eighteen of these were subsequently lost to follow up and presumed to have died because of withdrawal from dialysis. One patient became dialysis independent after 4.5 months. Of the rest, one patient was put on continuous ambulatory peritoneal dialysis and 3 continued on hemodialysis. Median survival in patients with renal involvement was only 4 months. Amongst the dialysis dependent patients, the longest survival was 11 months, and only 5 survived for over 6 months. Survival in patients with milder degrees of renal failure and in those who experienced improvement in renal function was longer, but the difference did not reach statistical significance. Of the 22 patients in whom the cause of death was known, 11 patients (50%) died of infection, 7 (31.8%) of cancer cachexia, 2 (9.1%) of gastrointestinal bleeding and one patient each (4.5%) of myocardial infarction and pulmonary thromboembolism.

DISCUSSION

The incidence of clinical renal involvement in multiple myeloma varies from 7–55% [[3]], [[6]], [[8]], [[10]], [13-15] and renal histological abnormalities are seen in 50–72% patients [16-17]. Renal failure is the commonest mode of presentation in those with renal involvement [1-2], [[10]], [[16]]. In one report of 998 patients, 43% had a plasma creatinine concentration above 1.5 mg/dL [[2]]. This variation in the reported incidence is due to differences in the case-mix and definition of renal involvement in different series. Workers who have reported higher figures have tended to include patients with milder degrees of azotemia [[3]], [14-16] without taking into consideration whether factors like dehydration had been corrected. On applying a stricter definition of renal failure (serum creatinine > 2–2.5 mg/dL after correction of fluid deficit), however, the incidence comes down to 7–20% [[6]], [[8]], [[10]], [14-15], [[17]]. Patients with severe renal failure constitute a larger proportion of cases in series reported by renal units [[7]], [[9]], [[18]]. Over 65% of patients with renal failure (18% of all cases) in the present study required dialytic therapy. The proportion of myeloma patients who require dialysis has been reported to vary from 2 to 66% [3-4], [8-9], [19-21]. Renal failure was non-oliguric in over 75% cases in the present study. In one report of 45 cases with renal failure, oliguria was seen in 36% cases whereas in others, it has varied from 24–50%. In contrast to these reports, renal failure was non-oliguric in only 40% of cases reported by Cohen et al. [[21]].

The diagnosis of multiple myeloma was made after admission to our hospital in 92.7% of those with renal involvement indicating that renal involvement antedated or coincided with the diagnosis of myeloma in them. Other workers [4-5], [[7]], [[18]], [21-23] have also described this phenomenon. Pozzi et al. [[5]], diagnosed multiple myeloma along with or after the detection of renal failure in 78% cases. In other series, the figures vary from 50–76% [[4]], [[7]], [[9]], [[18]], [[21]], [[23]]. On the basis of data reported by Rayner et al. [[21]], Winearls [[2]] concluded that the absence of severe renal impairment at presentation predicted a low probability of developing renal failure subsequently.

The importance of precipitating factors in development of renal failure in myeloma has been stressed in many studies [[3]], [[5]], [[16]], [[20]]. The common factors are fluid depletion, hypercalcemia and nephrotoxic drug use. At least one precipitating factor was present in 52.7% of our cases. Extracellular fluid depletion was the commonest and frequently followed a minor gastroenteritic illness. The next commonest precipitating factor was hypercalcemia. The role of hypercalcemia in the development of renal failure in patients with myeloma has been emphasized in a number of studies [3-5], [7-8], [[10]], [[15]], [[22]]. The measured serum calcium value may be normal or even low in those with renal failure and/or hypoalbuminemia [[21]]. Infections, to which these patients are predisposed because of abnormalities in immunoglobulin production have been incriminated in the development of both early [[8]], [[14]], [[22]] and late [[21]] renal failure in the course of myeloma. Intravenous contrast administration was thought to be an important predisposing factor [[6]], but more recent studies have failed to substantiate this association. In a review of 7 studies consisting of 476 myeloma patients who underwent contrast studies, the prevalence of acute renal failure ranged from 0.6–1.25% in contrast to 0.15% in the general population [[24]]. In the present study renal failure was precipitated by contrast in only 1 (1.8%) patient. In recent years, nonsteroidal anti-inflammatory drug (NSAID) use has been incriminated in the precipitation of renal failure in a significant proportion of myeloma patients [[2]]. NSAID use could be documented only in 7.2% cases in the present series. However, since these drugs are available openly in the market as non-prescription agents it is possible that more patients could have consumed them.

Of the various clinical and laboratory features that were associated with renal involvement, the commonest was anemia, followed by skeletal lesions. Both these features are known to be associated with chronic renal failure of any cause. However, demonstration of normal sized kidneys is an important clue to the possibility of myeloma. Therefore any elderly patient presenting with unexplained renal failure along with a bland urinary sediment, anemia out of proportion to renal failure and normal sized kidneys should be investigated for multiple myeloma. Hypercalcemia in a patient with renal failure of undetermined etiology should also raise the suspicion of underlying myeloma.

Although IgG was the most common monoclonal protein overall, the myeloma was of light chain type in 68% cases with renal involvement. Bladé et al. [[25]] have shown that patients with light chain or IgD myeloma have a much higher incidence of renal failure and amyloidosis than those with IgG or IgA myeloma. Non-light chain proteinuria was detected in 7 cases. It was mild (non-nephrotic) in 3 cases. Three of them also showed renal glycosuria. Isolated renal glycosuria was noted in one case. Other features of proximal tubular dysfunction, such as phosphate and bicarbonate wasting were not specifically looked for. Four patients presented with a nephrotic syndrome. Non-light chain proteinuria has been described in 16–24% cases with multiple multiple myeloma [[15]] but nephrotic syndrome is uncommon and indicates glomerular involvement due to amyloidosis or light chain deposition disease (LCDD) [[2]], [16-17], [26-28]. LCDD has been observed in 5–21% of cases with multiple myeloma and can be present either alone or in association with myeloma cast nephropathy and renal amyloidosis [26-28]. The most striking light microscopic abnormality in LCDD is a nodular glomerulosclerosis similar to the diabetic Kimmelstiel-Wilson lesion. However this lesion is seen in only 35% of patients [[27]]. LCDD can be diagnosed with greater sensitivity on electron or immuno fluorescence microscopy. We did not employ these techniques and therefore could have underestimated the true frequency of LCDD. Amyloidosis has been observed in 5–11% of cases with multiple myeloma and is detected more frequently at autopsy than on renal biopsies [[16]], [[22]], [[24]]. The distribution of amyloid is often widespread.

The commonest histological finding in patients with renal failure in multiple myeloma is myeloma cast nephropathy, also called “myeloma kidney” [[5]], [[8]], [[16]], [[18]], [[20]]. This condition is characterized by the presence of numerous large laminated eosinophilic casts composed of light chains, Tamm-Horsfall protein and fibrin. The casts show fissures and are surrounded by inflammatory reaction including giant cells. Associated tubular atrophy and interstitial fibrosis is seen frequently. This characteristic histological appearance is not seen in renal disease unrelated to plasma cell dyscrasias. Quite often, the diagnosis of multiple myeloma is first suspected on discovery of the classical histological picture of cast nephropathy [7-8], [[29]]. In 14.5% of our cases, the first clue to the diagnosis of multiple myeloma was provided by the renal histology.

There is no agreement on the relative contribution of various histologic components of cast nephropathy to the pathogenesis and course of renal failure. Some workers have shown the severity of cast formation to be directly proportional to the degree of renal insufficiency and its reversibility [[5]], [[18]]. In contrast, others have emphasized the importance of tubulointerstitial damage as being predictive of a worse renal outcome [[20]]. In the present study, the serum creatinine was <5 mg/dL in all cases with pure tubulointerstitial nephritis. Of the 16 cases with cast nephropathy, 75% were dialysis dependent at initial presentation. Thus it appears that whereas cast formation is important in determining the severity of renal failure, absence of advanced lesions in the tubulointerstitial compartment identifies patients with a potential for renal functional recovery [[8]], [[20]].

The exact reason why renal failure develops only in some patients with multiple myeloma is not known. In the present study, patients with renal failure had more severe anemia, hypercalcemia and bone lesions on skeletal survey. Taken together, these features are indicative of advanced disease. This was further confirmed when patients were classified according to the schema of Salmon and Durie [[12]]. Alexanian et al. [[3]] also found a direct correlation of the tumor burden with the presence of renal failure.

The factor that was most significantly associated with renal failure was the presence of Bence Jones proteinuria. On electrophoresis, light chain myeloma was seen with a significantly greater frequency in those with renal failure. Light chains co-precipitate with Tamm-Horsfall protein in the distal tubule and collecting ducts to give rise to the characteristic casts. The relationship between light chain excretion and renal failure due to cast nephropathy has been demonstrated in clinical [4-5], [7-8], [[18]], [[21]] and experimental [[30]] studies. The frequency of light chain myeloma varies from 15–60% amongst patients with renal failure [[4]], [7-9], [[18]], [[21]], [31-33]. Pozzi et al. [[5]] showed that patients with light chain myeloma were less likely to recover renal function than those with other varieties.

The impact of renal involvement on the overall prognosis of multiple myeloma is controversial. Earlier reports suggested that the prognosis was uniformly poor once renal failure developed [[9]], [[19]] and it was futile to treat these patients with long term dialysis [[34]]. The mortality of myeloma associated renal failure is particularly high in the first few weeks [[4]], [[9]], [[19]], [[21]]. Most of the deaths in the acute stage are due to infections. Eleven percent of our patients died within the first 6 weeks. In one study, the actuarial survival was 45% at one year and 25–30% at 2–3 years amongst those who survived the first two months [[31]]. In the present study, the median survival was only 4 months. An important factor contributing to this dismal figure was the inability of many patients to continue dialysis beyond 4 months because of economic reasons. The adverse effect of renal failure on overall survival has been shown in a number of studies [[5]], [[12]]. In the series of Cohen et al. [[21]], 62% of those who recovered renal function survived for 1 year whereas only 24% of those who continued to have renal failure did so. In another series, the median survival of patients who experienced complete recovery of renal function was similar to that of myeloma patients without renal failure [[18]]. Low tumor mass and response to chemotherapy are associated with an improved survival [[3]], [[12]]. In the series of long term survivors of multiple myeloma reported by Alexanian et al. [[2]] none of the patients who survived for more than 10 years had renal failure.

The proportion of cases in whom renal failure reverses or improves with treatment varies from 0 to 82% [[3]], [[5]], [[7]], [18-19], [[21]], [31-32], [34-39], [[41]]. In the present study, renal function improved in only 17% of dialysis dependent cases. A comparative analysis of studies is complicated by the great heterogeneity of the patient population and treatment modalities. Only dialysis dependent patients have been included in some studies, whereas others have included patients with all grades of renal failure. Patients with minor elevations in serum creatinine associated with a reversible factor are more likely to show an improvement in renal function [[3]]. Still others have included only those patients who were thought to have “acute” renal failure, and therefore more likely to show a recovery ([[4]], [[5]], [[6]], [[21]], [[23]]. A number of studies have shown that the likelihood of renal functional recovery with chemotherapy alone is low in patients who are dialysis dependent [7-9], [[19]], [31-32], [[34]].

In conclusion, renal failure occurs in 27% of patients with multiple myeloma, and is the initial presenting feature in a large proportion of cases. Renal failure is more likely to occur in those with Bence Jones proteinuria and a high tumor burden. Development of unexplained renal failure in an elderly individual, especially in association with disproportionate anemia with or without bone pains should alert the physician to this diagnosis. Common precipitating factors include dehydration, hypercalcemia and infection. The renal failure is non-oliguric in a majority of cases. Failure of renal functional recovery confers a poor prognosis on these patients, especially in underprivileged areas where maintenance dialysis facilities are not available.