Context.—In plasma cell dyscrasias, involvement of the distal tubules is frequent and well characterized. In contrast, proximal tubules have only rarely been reported to show diagnostic pathology such as intracytoplasmic crystals.
Objective.—To look for additional morphologic features that might be helpful in the diagnosis of proximal tubulopathy associated with an underlying plasma cell dyscrasia.
Design.—We examined patients presenting with nonspecific renal symptoms who were found to have light chain restriction limited to proximal tubular epithelium by immunofluorescence. We correlated these results with light microscopy, electron microscopy, and the clinical findings.
Results.—By immunofluorescence, 5 patients had light chain restriction in proximal tubular epithelium. By light microscopy, only 1 patient had focal rhomboid crystals in the proximal tubular epithelium; all other biopsies failed to show any discernible pathology within the proximal tubules or elsewhere in the kidney. By electron microscopy, proximal tubules from 2 patients showed crystals with a latticelike structure, whereas the remaining 3 patients had only prominent phagolysosomes. However, by immunoelectron microscopy, the lysosomal content showed light chain restriction (in 2 cases studied). Post–kidney biopsy, all patients were diagnosed with multiple myeloma or plasma cell dyscrasia. One patient developed renal failure and had recurrence of crystals in the allograft.
Conclusions.—Light chain proximal tubulopathy may be associated with the presence of crystals or with the presence of phagolysosomes with light chain restriction as the sole abnormality. Both κ and λ light chains may be involved. The prognosis is variable and the pathology may recur in transplants.
Plasma cell dyscrasias (PCDs) are characterized by the excessive and abnormal production of immunoglobulin molecules and very often include production of free light chains.1,2 They are associated with a wide range of renal pathologies that include light chain cast nephropathy, monoclonal immunoglobulin deposition disease, light and heavy chain amyloidosis, cryoglobulinemia, acute tubular necrosis–like tubulopathy, and tubulointerstitial nephritis.1–7
Acquired proximal tubular dysfunction is a poorly characterized entity that is rarely diagnosed and reported. This entity emerged when it was recognized that adult patients might develop acquired Fanconi syndrome (FS) in association with PCD.8 It was noted that the clinical picture of acquired FS was associated with the presence of intracellular crystals in the proximal tubular epithelium; similar crystals could also be seen within plasma cells in the patient's bone marrow. By immunofluorescence (IF), these crystals showed light chain restriction, almost exclusively κ. In most reported cases, FS preceded the diagnosis of an underlying PCD. However, it was also noted that the diagnosis of FS often went unrecognized for several years in many patients presenting with proteinuria, bone pain, or renal failure.8
The aim of this study was to identify morphologic features that may be helpful in the early diagnosis of proximal tubulopathy associated with an underlying PCD. We sought to look for evidence of PCD in patients presenting with nonspecific renal symptoms who were found to have light chain restriction in proximal tubular epithelium by IF as a sole abnormality. We sought to correlate these results with light microscopic, electron microscopic, and clinical findings.
MATERIALS AND METHODS
We identified 5 patients who, by IF studies, were found to have light chain restriction limited to the proximal tubular epithelium. We correlated the IF results with light microscopic, electron microscopic, and clinical findings.
Five patients were included: 3 women (47–62 years) and 2 men (56 and 62 years) (Table). Patient 2 underwent native kidney biopsy and 2 kidney allograft biopsies. The first allograft biopsy was done within 1-month posttransplantation (examined only by light microscopy), and the second was done 4 years posttransplantation.
The light microscopic features were recorded for both paraffin as well as plastic sections. For light microscopic examination, the slides were stained with hematoxylin-eosin, periodic acid–Schiff, toluidine blue, and Congo red. Immunofluorescence studies were performed on frozen sections and included the following antibodies: immunoglobulin (Ig) G, IgA, IgM, κ light chain, λ light chain, C3, C1q, and fibrinogen. Electron microscopic studies were performed as previously reported.9 In addition, immunogold staining was performed as per Herrera's technique in cases 4 and 5.10
The clinical features at the time of kidney biopsy and postbiopsy were obtained from the attending physician and from a review of the patient's clinical records.
As per the study design, in all 5 patients (6 biopsies) IF studies demonstrated light chain restriction limited to proximal tubular epithelium as the sole abnormality: κ light chain in 4 biopsies (cases 1, 2, and 4) and λ light chain in 2 patients (patients 3 and 5) (Table). The stain was positive in virtually all proximal tubular epithelial cells within the sections (Figure 1). In patient 2, similar IF findings were seen in the native kidney and in 1 allograft biopsy. In contrast, all other stains, including the heavy chains, λ light chain (cases 1, 2, and 4), κ light chain (cases 3 and 5), and complements, were negative.
Examination of the paraffin sections in case 1 showed the focal presence of intracytoplasmic rhomboid crystals within the proximal tubular epithelium (Figure 2). In paraffin sections, these crystals were somewhat ill defined and very subtle. A slightly better resolution was seen in Epon plastic sections (not shown). However, the focality of their presence by light microscopy was in stark contrast to the diffuse presence of light chain restriction by IF. In the remaining 4 patients, no specific light microscopic pathology was appreciated within the proximal tubular epithelium. Although on occasion proximal tubules showed some apical disruption, such changes are difficult to distinguish from preservation artifacts. The glomeruli were unremarkable in all cases, and in cases 3 to 5 there was no evidence of any appreciable tubulointerstitial damage. Only in cases 1 and 2 was there a picture of chronic tubulointerstitial nephritis with tubular atrophy and interstitial expansion with a low-grade mononuclear infiltrate; these were more pronounced in patient 2, who also had a long history of nonsteroidal anti-inflammatory drug use. The Congo red stain was negative for amyloid in all biopsies.
Electron microscopy confirmed the presence of intracytoplasmic crystals in case 1 (Figure 3, A). In addition, in patient 2, electron microscopy demonstrated intracytoplasmic crystals, which were not discernible by light microscopy, even on retrospective examination. By electron microscopy, the rhomboid crystals had a latticelike structure (Figure 3, B). These crystals were seen in most of the proximal tubular epithelial cells paralleling the distribution of the κ light chain stain by IF.
Routine electron microscopic studies in cases 3 to 5 showed only nondiagnostic changes in the form of somewhat prominent, on occasion atypical but otherwise nonspecific, phagolysosomes within the proximal tubular epithelium. However, by immunoelectron microscopy the lysosomal content showed light chain restriction in 2 of 3 cases studied: for κ and λ light chains in cases 4 and 5, respectively (Figure 4). No other diagnostic abnormalities were seen. Specifically, there was no evidence of amyloid or monoclonal immunoglobulin deposition disease and no crystals were seen elsewhere in the kidney biopsy.
Clinicopathologic Correlation and Follow-up
At the time of kidney biopsy, only patient 3 had a prior diagnosis of multiple myeloma and known light chain proteinuria with apparently normal renal function. The kidney biopsy was obtained before planned chemotherapy for multiple myeloma. In patients 1 and 2 renal failure was the original indication for a kidney biopsy, whereas patients 4 and 5 had some proteinuria not otherwise specified; with the exception of patient 3, none had a prior diagnosis of multiple myeloma or PCD. Following kidney biopsy, all patients were diagnosed with some form of PCD, including multiple myeloma in 2 additional patients (patients 1 and 4), and all patients were diagnosed with proteinuria: exclusively light chain proteinuria (patients 2 and 3) or light chain proteinuria with some albuminuria (patients 1 and 5). No follow-up data are available on patient 4 who initially was referred with “proteinuria.” Although patient 2 had no detectable proteinuria at the time of her native kidney biopsy, at subsequent biopsy of her renal allograft that was performed about 8 years later, she also had isotypic light chain proteinuria.
Patient 1, who had crystals detectable focally by light microscopy, had clinical features of acquired FS with low serum uric acid levels despite renal dysfunction and significantly increased levels of amino acids in his urine. His serum κ free light chain was elevated (14 976 mg/L) and he was diagnosed with multiple myeloma. He underwent plasmapheresis and intensive multiagent chemotherapy followed by high-dose chemotherapy and autologous bone marrow transplantation. Despite this aggressive therapy, his bone marrow showed persistent disease. Posttransplant, he was found to have isotypic (κ light chain) AL amyloid in gastrointestinal biopsies and died 6 months after a diagnosis of light chain proximal tubulopathy (LCPT) associated with crystals. Patient 3 who had a pre– kidney biopsy diagnosis of IgD-λ light chain myeloma with λ light chain proteinuria received chemotherapy and a bone marrow transplant. She continues to do well 10 years postbiopsy. Her renal function remains normal; currently, there is no proteinuria. Patient 4 was diagnosed with isotypic multiple myeloma within 2 months of her kidney biopsy, and chemotherapy was planned at another institution. However, we have no additional follow-up data. Patient 2, who had crystals detectable only by electron microscopy also had a long history of analgesic use and a clinical diagnosis of analgesic nephropathy. Although the original biopsy demonstrated light chain restriction within proximal tubular epithelium, no evidence of an underlying PCD could be demonstrated clinically at that time. This patient, however, did develop renal failure, for which she underwent renal transplantation 4 years after the initial diagnosis of LCPT associated with crystals. Recurrence of LCPT with crystals was seen in a subsequent allograft biopsy performed 4 years after transplant. Her bone marrow biopsies, however, have never shown any increase in plasma cells. Only recently, 8 years after the original diagnosis of LCPT with crystals, has her serum immunofixation shown a small IgA-κ paraprotein and discernible κ light chain band. The patient has been treated for the past 4 years with dexamethasone and later combinations including dexamethasone, bortezomib, and cyclophosphamide. She has responded well, with serum creatinine levels generally showing a consistent inverse relationship to her κ light chains. She has a partial acquired FS, with low serum uric acid levels despite renal dysfunction. Patient 5 was found to have a small band of monoclonal λ light chain on serum protein electrophoresis with immunofixation. No further follow-up data are available for this patient.
Although the presence of crystals was associated with κ light chains and renal failure, among patients without crystals, both κ or λ light chains were found and no renal failure was apparent. Electron microscopic examination of bone marrow biopsies in patients 1 and 2 demonstrated abundant crystals which, however, were not discernible in paraffin sections, even on retrospective examination (not shown).
In our study, the detection of light chain restriction by IF studies on frozen sections was the most sensitive method for the diagnosis of isolated LCPT associated with underlying PCD. Although this finding correlated very well with the clinical evidence of some form of PCD in all our patients, the underlying PCD was not otherwise apparent at the time of kidney biopsy in all but 1 patient in our series. Moreover, in 1 patient, detection of light chain restriction within proximal tubular epithelium preceded the diagnosis of PCD by several years.
In our series, crystals within the renal proximal tubular epithelium were demonstrated in only 1 case by light microscopy and in 2 cases by electron microscopy. These crystals were apparent only focally and could potentially be easily overlooked. In the remaining 3 cases, light microscopy failed to show specific pathology, whereas routine electron microscopy, on retrospective examination, revealed somewhat more prominent and at times atypical phagolysosomes as a sole feature. Previously, Herrera et al6 and Sanders et al11 demonstrated that LCPT may be associated with an acute tubular necrosis–like picture with prominence of the lysosomal system, including the presence of atypical lysosomes. This pathology was also reproduced experimentally in rats by perfusing light chains extracted and purified from urine of patients with renal failure and proximal tubulopathy.12 Thus, although not entirely specific, these features may be helpful in directing diagnostic suspicion toward underlying PCD. These considerations notwithstanding, diagnosis of LCPT not associated with crystals is challenging by light microscopy and routine electron microscopy. Although immunogold labeling is diagnostic, this technique is not routinely applied and is available only in specialized laboratories. Although IF remains the single, most sensitive screening method for the detection of light chain restriction, the issue of its sensitivity should also be addressed. Data thus far available demonstrate that the light chains associated with LCPT are partially digested and may contain predominantly the V region, which is not recognized by commercial antibodies.13–15 Thus, rare cases of proximal tubulopathy have been reported that were not detectable by IF.8 Gu et al10 postulated that immunoelectron microscopy might be more sensitive than IF in detecting LCPT; immunoelectron microscopy has also been shown to be more sensitive in the detection of truncated light chains in amyloid.16
Almost all cases of crystal-associated LCPT have been associated with κ light chains, in particular the Vκ1 subgroup.13 It has been proposed that mutations in the CDR-L1 loop may be a factor in protease resistance, because these light chains are incompletely digested. It is postulated that these light chain fragments serve as a nidus for crystal formation in renal tubules8 and that the accumulation of crystals in lysosomes causes proximal tubular injury. To this end, in vitro studies have shown that the crystals interfere with a broad range of apical membrane transporters.17 Only very rare cases of LCPT associated with λ light chains have been reported (with or without crystals).18–20 In our study, both cases with intracytoplasmic crystals were associated with monoclonal κ light chains and both presented with renal failure. It must be stressed that not all crystals are associated with an underlying PCD, and other etiologies (drugs, toxins) must be considered in the differential diagnosis.21
Light chain proximal tubulopathy associated with underlying PCD is a rare entity. There are only 4 series of patients published to date,8,10,18,19,22 2 of which are from the same institution; other cases were published as single case reports.4,19,20,23 Most of the reports have stressed the presence of intracytoplasmic crystals and an association with FS.4,5,10,18–20,22,24 Our study shows that LCPT may be present in the absence of crystals and probably more frequently than is currently appreciated. In our series, only one patient had evidence of FS clinically and another had a partial acquired FS based on retrospective examination 8 years following the original diagnosis of LCPT and 4 years following renal transplant. It has been noted previously that FS may be underdiagnosed or diagnosed late into the disease process. Thus, in a series by Messiaen et al,8 the diagnosis of FS was often unrecognized for several years (mean, 3.1 years). Thus, if based on the recognition of FS, the diagnosis of an underlying PCD could often be missed or significantly delayed. Although it is difficult to estimate the “n” for these lesions in our material, awareness appears to play a role. During the preceding several years, we recognized only 2 lesions; within the last 2 years we have diagnosed 3 patients with LCPT.
Prior reports in the literature showed that most patients with LCPT appeared to do well in the absence of an overt malignancy,19,24 whereas in the presence of overt myeloma, the prognosis depends on the tumor load and the presence of amyloidosis.10,24 Accordingly, our patient 2 continued to be well for several years following the original diagnosis, whereas our patient 1, who developed chemotherapy-resistant multiple myeloma and amyloidosis, died of the disease within 6 months. However, as seen in our patient 3, successful treatment of the underlying multiple myeloma prevented progression of tubular injury. Light chain proximal tubulopathy can also be associated with other PCD pathologies such as light chain cast nephropathy or amyloid (our case 1). The association of crystal deposition in podocytes with glomerular dysfunction has also been reported.25,26 The presence of intracytoplasmic crystals in proximal tubular epithelium is frequently correlated with the presence of crystals in the bone marrow,8,27,28 which was also confirmed in our patients 1 and 2.
It should be stressed that a high index of suspicion for an underlying PCD is warranted in the evaluation of kidney biopsies, and an appropriate clinical work-up should always be recommended. Testing for circulating free light chains, a relatively recent test, should be recommended to substantiate the diagnosis.29 However, again, although the free light chain test has been very useful, its sensitivity has not yet been established.30 Early diagnosis is important because LCPT is reversible on suppression of nephrotoxic light chain secretion.31 Given the morphologic and clinical heterogeneity of proximal tubular involvement in underlying PCD, we propose the term LCPT as an inclusive term encompassing LCPT with or without crystals.
In summary, our study shows that neither intracytoplasmic crystals nor acquired FS are constant features of LCPT. Renal pathologists must keep in mind that LCPT can be very subtle, and a high index of suspicion is needed to ensure a correct diagnosis.
The authors have no relevant financial interest in the products or companies described in this article.
Reprints: Maria M. Picken, MD, PhD, Department of Pathology, Building 110, Room 2242, Loyola University Medical Center, 2160 S First Ave, Maywood, IL 60153 (firstname.lastname@example.org)