Correspondence *Ohio State University Nephrology Division, N210 Means Hall, 1654 Upham Drive, Columbus, OH 43210, USA

Email rovin.1@osu.edu

The case

A 59-year-old white male with a past medical history significant only for well-controlled hypertension and type II diabetes mellitus presented with progressive fatigue, abdominal pain and weight loss. An abdominal CT scan revealed two large left renal masses indicative of malignancy. The patient underwent a left radical nephrectomy. Examination of the tumor demonstrated conventional clear cell renal carcinoma staged as T3aN0M0.

Three months after nephrectomy, a restaging investigation showed evidence of metastatic disease in the lungs. Treatment with interferon 2b plus bevacizumab was initiated. Bevacizumab is a humanized monoclonal anti-vascular endothelial growth factor (anti-VEGF) antibody,¹ thought to inhibit tumor growth by blocking the angiogenesis that is supported by VEGF-induced endothelial cell proliferation and survival.² Corticosteroids were not used. A detailed chronology of changes in the patient's kidney function, proteinuria, and blood pressure is presented in Table 1. At the start of therapy, the patient's serum creatinine level was 115 mol/l and urinalysis was unremarkable—specifically there was no proteinuria or hematuria. Four months later, the patient's serum creatinine level had increased to 141 mol/l and his angiotensin-converting-enzyme inhibitor (quinapril) was discontinued. His serum creatinine level remained elevated. Nine months after therapy was started, serial dipstick evaluations of the patient's urine demonstrated the development of grade 1–2+ proteinuria. A 24-hour urine collection at this time contained 1,836 mg of protein. Between 9 months and 14 months after the start of therapy, proteinuria remained in the range 600–1,500 mg/day; however, it then increased significantly, despite initiation of an angiotensin-receptor blocker (telmisartan) at 13 months (Table 1). Interferon 2b was discontinued after 15 months because of neuropsychiatric symptoms including anxiety and depression. Bevacizumab was stopped after 15.5 months as a result of the proteinuria and elevated serum creatinine level. The patient achieved complete remission of metastatic renal cell carcinoma and remains disease-free 25 months after therapy was started.

Table 1 Summary of the main case

Full tableFigures & Tables indexDownload Power Point slide (202K)

As a result of continued nephrotic-range proteinuria and renal insufficiency after stopping interferon 2b and bevacizumab, the patient was referred to a nephrology clinic for evaluation approximately 2 months after these therapies were discontinued (Table 1). Physical examination found a thin, but not cachectic, white male in no distress. His blood pressure was 142/80 mmHg. He had no cervical lymphadenopathy, findings from his cardiopulmonary exam were normal, and his abdomen showed a well-healed scar, no masses, and no enlargement of the liver or spleen. He had mild to moderate bilateral lower extremity edema to the level of the midshin. Analysis of the patient's urine sediment showed 3–4 white blood cells per high power field, 0–2 red blood cells per high power field, no dysmorphic red blood cells, and an occasional white blood cell cast. Hepatitis, autoimmune, and HIV serologies were all negative and complement levels were normal. Serum and urine immunofixation did not show a monoclonal gammopathy. There was no indication of a hypercoagulable state, as evidenced by normal values for INR (international normalized ratio), partial thromboplastin time, D-dimer, and anticardiolipin (IgG and IgM) antibodies, and a negative lupus anticoagulant test. During the course of therapy with interferon 2b and bevacizumab, the patient developed a mild microcytic anemia (mean corpuscular volume 77 fl) with a nadir hemoglobin of 105 g/l, leukopenia with a nadir white blood cell count of 2.6 10⁹/l, and mild thrombocytopenia with a nadir platelet count of 126 10⁹/l. Cardiac function showed a normal ejection fraction. An ultrasound-guided renal biopsy was performed about 2 months after interferon 2b and bevacizumab were discontinued.

Kidney biopsy findings

Up to seven enlarged glomeruli were present in each of the sections prepared for light microscopy. Methenamine silver staining showed that these glomeruli had markedly thickened capillary walls with double contours (Figure 1). One arteriole was occluded by amorphous eosinophilic material (Figure 2) and an adjacent artery showed mucoid intimal thickening. Mild focal arterial fibrous intimal thickening was present. Neither glomerular nor arteriolar fibrin thrombi were observed, and no fragmented red blood cells were seen. The interstitium displayed very mild patchy fibrosis and tubular atrophy, involving approximately 10% of the cortex.

Figure 1 Light microscopy of the kidney biopsy

Thickened glomerular capillary walls are shown by (A) periodic acid-Schiff and (B) methenamine silver staining. The silver stain also shows double contours in some capillary loops (arrows; original magnification of both images 400.)

Full figure and legend (83K)Figures & Tables indexDownload Power Point slide (199K)

Figure 2 Light microscopy of the kidney biopsy

Amorphous eosinophilic debris in an arteriole (arrow) and glomerular capillary hyalinosis (arrowhead) are shown (hematoxylin and eosin stain; original magnification 200).

Full figure and legend (61K)Figures & Tables indexDownload Power Point slide (176K)

Frozen sections prepared for direct immunofluorescence microscopy contained up to five glomeruli, and showed strong diffuse granular IgA deposits in the glomerular capillary walls and mesangial areas (Figure 3). These deposits also demonstrated moderate fluorescence for kappa and lambda light chains, and mild fluorescence for IgG and IgM. The glomerular capillary walls and mesangial regions showed mild, segmental granular staining for the complement component C3, and only trace staining for C1q. There was moderate to prominent segmental glomerular capillary wall and mesangial fibrinogen staining.

Figure 3 Immunofluorescence microscopy of the kidney biopsy

The image shows bright diffuse granular glomerular capillary wall and mesangial staining for IgA (original magnification 400).

Full figure and legend (25K)Figures & Tables indexDownload Power Point slide (140K)

By electron microscopy, the glomerular capillary loops showed prominent amorphous subendothelial widening with mesangial cell interposition as well as electron-dense immune-type deposits (Figure 4A). Electron-dense immune-type deposits were also identified in the mesangium (Figure 4B), and some areas showed striking mesangiolysis with fibrin strands (Figure 4C). Focally, the widened subendothelial space and the loosened-up mesangium showed a vague fibrillary transformation. The glomerular endothelium was swollen and had lost its fenestration. Moderate foot process effacement was noted. No endothelial tubuloreticular inclusions were identified.

Figure 4 Electron microscopy of the kidney biopsy

(A) A glomerular capillary loop shows subendothelial widening with amorphous material (asterisk), electron-dense immune-type deposits (labeled 'D'), and swollen endothelial cells with loss of fenestrae (labeled 'E'). The glomerular basement membrane is indicated by 'B' and the capillary lumen is labeled 'L' (original magnification 7,000). (B) Mesangial electron-dense immune-type deposits (labeled 'D'; original magnification 3,000). (C) Lifting off of the glomerular basement membrane (labeled 'B') from the mesangium (labeled 'M') in an area of mesangiolysis (original magnification 4,400).

Full figure and legend (31K)Figures & Tables indexDownload Power Point slide (146K)

The kidney biopsy findings were interpreted as an unusual IgA predominant immune-complex glomerulonephritis associated with changes of thrombotic microangiopathy. The clinical history and morphology, however, were not truly characteristic of IgA nephropathy. Additionally, the light microscopic glomerular changes of thickened capillary walls with double contours resembled membranoproliferative glomerulonephritis, but neither the clinical history nor the findings on immunofluorescence or electron microscopy were consistent with this diagnosis. There was no evidence of diabetic kidney disease.

To determine whether any of these biopsy findings were present before the patient started chemotherapy, archived tissue from the patient's original nephrectomy specimen was processed for light microscopy, immunofluorescence microscopy, and electron microscopy examination. Non-neoplastic areas of the cortex were examined. Light microscopy only revealed changes consistent with mild nephrosclerosis. The majority of the glomeruli were unremarkable (Figure 5A). Immunofluorescence microscopy was performed on paraffin sections following protease pretreatment. No glomerular capillary wall or mesangial IgA deposits were identified (not shown). Mild segmental glomerular fibrinogen staining was seen in the mesangium. Ultrastructural examination was performed on a section of cortex that had been deparaffinized and reprocessed for electron microscopy. This examination showed only rare indistinct mesangial densities in the three glomeruli inspected. No glomerular capillary electron-dense deposits were seen (Figure 5B) and no subendothelial widening of the glomerular capillaries or mesangiolysis was observed. The glomerular basement membrane was normal in thickness and texture. The foot processes did not show significant effacement.

Figure 5 Sections of non-neoplastic areas of the cortex from the original specimen obtained from partial nephrectomy for renal cell carcinoma

(A) Glomeruli and renal parenchyma appear normal by light microscopy (hematoxylin and eosin stain; original magnification 200). (B) Glomerular capillary wall appears ultrastructurally normal in a section reprocessed for electron microscopy (original magnification 7,000).

Full figure and legend (58K)Figures & Tables indexDownload Power Point slide (168K)

Discussion of diagnosis

Targeting VEGF in cancer treatment has recently received considerable attention. Anti-VEGF antibody therapy has mainly been used in clinical trials for metastatic colorectal cancer and renal cell carcinoma,^{3, 4, 5} but recent data indicate that these antibodies also have activity in breast cancer and lung cancer.^{6, 7} The role of VEGF in human renal physiology is not known; however, VEGF is constitutively expressed by podocytes, and its receptors are found on normal glomerular capillary endothelial cells.⁸ Anti-VEGF therapy has been associated with the development of proteinuria in 23–38% of patients with colorectal cancer, and in up to 64% of patients with renal cell carcinoma.^{3, 5} Although this proteinuria has largely been asymptomatic and low-grade, it exceeded 3.5 g/day in 6.5% of all renal cell carcinoma patients in one large study.⁵ These findings indicate that VEGF might be involved in the maintenance of the glomerular capillary permeability barrier. To date, there have been no descriptions of kidney histology from patients receiving anti-VEGF therapies. In this Case Study, we report the renal pathologic findings of a patient treated with bevacizumab. The kidney biopsy demonstrated glomerular capillary endothelial injury indicative of a thrombotic microangiopathy, a finding consistent with the notion that VEGF is important for glomerular capillary integrity. The additional observation of an IgA-predominant immune-complex glomerulonephritis was surprising, and is difficult to explain in the context of VEGF's currently known biologic activities. Interestingly, a partial nephrectomy specimen from a second patient treated with bevacizumab was also consistent with endothelial injury secondary to a thrombotic microangiopathy, but lacked any evidence of immune-complex glomerulonephritis (Box 1 and Figure 6).

Figure 6 Electron micrograph of a glomerular capillary loop from a partial nephrectomy specimen from the patient described in Box 1

The specimen was taken after treatment with bevacizumab plus interferon 2b for 9 months. Subendothelial electron-lucent widening with amorphous material (asterisk) is shown (original magnification 12,000).

Full figure and legend (38K)Figures & Tables indexDownload Power Point slide (153K)

To explain the development of a thrombotic microangiopathy during anti-VEGF antibody treatment, it is reasonable to postulate that blocking VEGF prevents the repair of endothelial damage that probably occurs daily in blood vessels. Anti-VEGF therapy also commonly causes hypertension, which could contribute to such endothelial injury.^{3, 5} This endothelial injury might create a thrombogenic endothelium and predispose to intravascular clotting. Relevant to this hypothesis, arterial thrombosis is a recognized side effect of bevacizumab therapy.³ Furthermore, decreased levels of VEGF have been associated with sirolimus-induced thrombotic microangiopathy in kidney transplant recipients, and administering VEGF in experimental thrombotic microangiopathies improved glomerular capillary healing.⁸ A caveat in the cases presented is that interferon therapy has also been associated with the development of renal thrombotic microangiopathy.⁹ It will, therefore, be important to look at kidney pathology in patients treated with anti-VEGF antibodies in the absence of treatment with interferon .

The finding of glomerular IgA deposition in the main case presented, and its absence in the second patient (Box 1), is intriguing. The glomerular IgA deposition was not a pre-existing condition, a conclusion based on immunofluorescence and electron microscopy examinations of the patient's original nephrectomy specimen before any chemotherapy had been given. While systemic and renal VEGF expression might be altered in IgA nephropathy,¹⁰ there is no evidence that blocking the actions of VEGF predisposes to the development of IgA immune complexes. Similarly, interferon treatment has not been associated with de novo IgA nephropathy, although it has been suggested that the treatment exacerbates existing IgA nephritis.¹¹ It will be important to determine whether IgA immune-complex glomerulonephritis is found in other patients treated with anti-VEGF antibodies who develop renal insufficiency and high-grade proteinuria, but not in patients who develop only minor proteinuria. If the case described here does not represent an isolated finding, and if glomerular IgA deposition occurs in the absence of interferon therapy, then there could be implications for the understanding of the pathogenesis of IgA nephropathy.

Manipulation of the VEGF axis in experimental animals supports the importance of VEGF in glomerular capillary integrity. In normal mice, administration of antibodies to VEGF, or neutralization of VEGF with soluble VEGF receptor 1, caused a significant increase in proteinuria within 3 hours; this increase was associated with hypertrophy of glomerular endothelial cells, endothelial detachment from the basement membrane, and loss of slit diaphragms.¹² These animals also showed reduced glomerular expression of nephrin. The chronic administration of anti-VEGF receptor 2 antibodies to mice prone to systemic lupus erythematosus markedly increased mortality and accelerated the development of lupus nephritis and proteinuria.¹ Finally, podocyte-specific deletion of the VEGF gene caused proteinuria, glomerular endotheliosis, and the eventual loss of endothelial fenestrations in heterozygous mice, while VEGF-null mice failed to develop endothelial fenestrations at all.¹³ Interestingly, in human preeclampsia, fenestrated endothelium is especially vulnerable to injury, and the placenta overproduces a soluble VEGF receptor that acts as a VEGF antagonist.^{14, 15}

VEGF is also important in glomerular capillary repair after immune injury. Infusing VEGF into rats with crescentic antiglomerular basement membrane antibody nephritis or anti-Thy-1.1 mesangial proliferative glomerulonephritis enhanced recovery, decreased proteinuria, and improved glomerular histology.^{16, 17} Furthermore, steroid treatment of rats given anti-Thy-1.1 significantly attenuated the expected increase in VEGF expression (found in untreated anti-Thy-1.1 rats), and was associated with a threefold increase in proteinuria.¹⁸

Conclusions

The cases presented here are an important first step towards understanding the mechanisms underlying the renal toxicity of a cancer therapy that will be widely used in the future. The kidney biopsy findings in patients treated with bevacizumab plus interferon 2b for renal cell carcinoma, along with the data from experimental animals, indicate that this therapy, and in particular anti-VEGF antibodies, might cause glomerular capillary injury consistent with a thrombotic microangiopathy. This injury can be associated with severe or minor proteinuria, and impaired or normal renal function. Factors that affect the clinical manifestation of the renal injury remain to be determined, but it is intriguing to speculate that immune-complex deposition might be required for severe presentations. The pathologic findings described in these cases will need to be confirmed in other patients treated with anti-VEGF antibodies, and a large series of patients will be required for clinical correlations to be made.

References

1. Watanabe H et al. (2005) Anti-vascular endothelial growth factor receptor-2 antibody accelerates renal disease in the NZB/W F₁ murine systemic lupus erythematosus model. Clin Can Res 11: 407–409 | ChemPort |
2. Rini BI and Small EJ (2005) Biology and clinical development of vascular endothelial growth factor-targeted therapy in renal cell carcinoma. J Clin Onc 23: 1028–1043 | ChemPort |
3. Gordon MS and Cunningham D (2005) Managing patients treated with bevacizumab combination therapy. Oncology 69 (Suppl 3): S25–S33 | Article |
4. Rini BI et al. (2004) Cancer and leukemia group B 90206: a randomized phase III trial of interferon-alpha or interferon-alpha plus anti-vascular endothelial growth factor antibody (bevacizumab) in metastatic renal cell carcinoma. Clin Can Res 10: 2584–2586 | ChemPort |
5. Yang JC et al. (2003) A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic renal cancer. N Eng J Med 349: 427–434 | Article | ISI | ChemPort |
6. Ellis LM et al. (2006) Overview of anti-VEGF therapy and angiogenesis: angiogenesis inhibition in solid tumor malignancies. Clin Adv Hematol Oncol 4 (Suppl 1): S1–S12
7. Jain RK et al. (2006) Lessons from phase III clinical trials on anti-VEGF therapy for cancer. Nat Clin Pract Oncol 3: 24–40 | Article | PubMed | ISI | ChemPort |
8. Schrijvers BF et al. (2004) The role of vascular endothelial growth factor (VEGF) in renal pathophysiology. Kidney Int 65: 2003–2017 | Article | PubMed | ISI | ChemPort |
9. Badid C et al. (2001) Renal thrombotic microangiopathy induced by interferon-alpha. Nephrol Dial Transplant 16: 846–848 | PubMed | ChemPort |
10. De Vriese AS et al. (2001) Antibodies against vascular endothelial growth factor improve early renal dysfunction in experimental diabetes. J Am Soc Nephrol 12: 993–1000 | PubMed | ISI | ChemPort |
11. Gordon A et al. (2004) Combination pegylated interferon and ribavirin therapy precipitating acute renal failure and exacerbating IgA nephropathy. Nephrol Dial Transplant 19: 2155–2157 | PubMed |
12. Sugimoto H et al. (2003) Neutralization of circulating vascular endothelial growth factor (VEGF) by anti-VEGF antibodies and soluble VEGF receptor 1 (sFlt-1) induces proteinuria. J Biol Chem 278: 12605–12608 | Article | PubMed | ISI | ChemPort |
13. Eremina V et al. (2003) Glomerular-specific alterations of VEGF-A expression lead to distinct congenital and acquired renal diseases. J Clin Invest 111: 707–716 | Article | PubMed | ISI | ChemPort |
14. Maynard SE et al. (2003) Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 111: 649–658 | Article | PubMed | ISI | ChemPort |
15. Karumanchi SA et al. (2005) Preeclampsia: a renal perspective. Kidney Int 67: 2101–2113 | Article | PubMed |
16. Shimizu A et al. (2004) Vascular endothelial growth factor-165 resolves glomerular inflammation and accelerates glomerular capillary repair in rat anti-glomerular basement membrane glomerulonephritis. J Am Soc Nephrol 15: 2655–2665 | Article | PubMed | ISI | ChemPort |
17. Masuda Y et al. (2001) Vascular endothelial growth factor enhances glomerular capillary repair and accelerates resolution of experimentally induced glomerulonephritis. Am J Pathol 159: 599–608 | PubMed | ISI | ChemPort |
18. Ha SI et al. (2002) Glucocorticoid diminishes vascular endothelial growth factor and exacerbates proteinuria in rats with mesangial proliferative glomerulonephritis. Am J Kidney Dis 39: 1001–1010 | Article | PubMed | ISI | ChemPort |

Contact the journal about this article

Subject areas under which this article appears: Oncology | Other