Sunitinib-induced macrocytosis in patients with metastatic renal cell carcinoma
Abstract
BACKGROUND.
Sunitinib and sorafenib are small molecules that inhibit the vascular endothelial growth factor and related receptors with substantial clinical activity reported in metastatic renal cell carcinoma (RCC). Cytopenia and macrocytosis have been described in patients treated with these agents.
METHODS.
A retrospective review of all patients with metastatic RCC who were treated with sunitinib or sorafenib for at least 3 months at the Cleveland Clinic Taussig Cancer Institute was undertaken. Complete blood count (CBC) data including red blood cell indices were recorded at baseline, after 3 months of therapy, and at the end of treatment.
RESULTS.
A total of 61 patients were treated with sunitinib and 37 patients were treated with sorafenib with available CBC data. In patients treated with sunitinib, the median corpuscular volume (MCV) increased significantly at 3 months compared with baseline (median increase of 5.1 femtoliters [fL]; P < .001) and continued to increase throughout treatment. Patients who developed hypothyroidism had a larger MCV increase at 3 months than patients who remained euthyroid (P = .06), although macrocytosis was observed in patients without hypothyroidism. Ten patients discontinued sunitinib therapy, and the MCV decreased in all patients within 2 to 4 months, without further intervention. Bone marrow analysis of 4 patients revealed a hypocellular bone marrow with trilineage hematopoiesis and no evidence of metastasis. There was no evidence of folate or vitamin B12 deficiency. In contrast to sunitinib, there was no change in the MCV for patients treated with sorafenib.
CONCLUSIONS.
Macrocytosis was a common occurrence after treatment with sunitinib but not sorafenib in patients with metastatic RCC. Sunitinib-induced macrocytosis is reversible with drug discontinuation. Cancer 2008. © 2008 American Cancer Society.
The underlying biology of renal cell carcinoma (RCC) supports the importance of the vascular endothelial growth factor (VEGF) pathway in tumor pathogenesis.1 As such, several therapeutic agents targeted against various aspects of the VEGF pathway have demonstrated robust clinical effects, largely supplanting cytokine therapy as a standard of care for advanced RCC. Sunitinib and sorafenib are small molecule inhibitors of a family of receptor tyrosine kinase inhibitors including the VEGF receptor.2, 3 The hematologic toxicities of these agents have been reported, including neutropenia, lymphopenia, thrombocytopenia, and anemia.4, 5 Macrocytosis, defined as an increase in the mean corpuscular volume (MCV) of the red blood cells (RBCs), has also been noted with sunitinib therapy.6, 7 Based on these observations, a comprehensive review of the hematologic and RBC indices of a large group of patients with metastatic RCC treated with either sunitinib or sorafenib was undertaken.
MATERIALS AND METHODS
The medical records of patients with metastatic RCC who were treated with sunitinib or sorafenib and who were enrolled onto institutional review board-approved clinical trials at the Cleveland Clinic Taussig Cancer Institute between October 2003 and January 2006 were reviewed. Sunitinib was administered at a dose of 50 mg orally daily for the first 4 weeks of a 6-week cycle. Sorafenib was administered orally twice daily at a dose of 400 mg. Dose interruptions and/or reductions were permitted for toxicity and undertaken according to the guidelines for each individual study. Eligibility criteria for the trials were fairly uniform and generally included the following: histologic documentation of clear cell RCC; clinical or biopsy evidence of metastatic disease; measurable disease; an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1; adequate renal, hepatic, and bone marrow function; absent or previously treated central nervous system metastasis; no prior history of cancer (except basal cell carcinoma or carcinoma in situ of the cervix); the absence of significant cardiac disease; and no recent surgery. Patients not receiving at least 3 months of treatment because of toxicity, death, or disease progression were excluded from the analysis.
Patient characteristics (including age and sex), complete blood count (CBC) indices (hemoglobin [in g/dL], white blood cell and neutrophil counts, platelet counts, and MCV), and thyroid function tests (TFT) were recorded at baseline (within 2 weeks of the initiation of starting therapy), at 3 months, and at the end of therapy.
Statistical Analysis
Data were summarized descriptively as frequency counts, medians, and ranges. Both absolute and relative changes in the hematologic parameters after 3 months of treatment and at the end of treatment were analyzed; however, the results were found to be similar and therefore only the absolute changes are reported here. The Wilcoxon signed rank test was used to assess these changes within each treatment group. The Fisher exact test and the Wilcoxon rank sum test were used to compare the 2 groups with respect to patient characteristics and baseline hematologic parameters. All statistical tests were 2-sided and a P value <.05 was considered statistically significant. All analyses were conducted using SAS statistical software (version 8; SAS Institute Inc., Cary, NC).
RESULTS
Patient Characteristics
A total of 119 patients with metastatic RCC who were treated with either sunitinib or sorafenib were initially considered for inclusion in the current study. After excluding 21 patients because they did not receive at least 3 months of treatment due to disease progression or discontinuation of therapy due to toxicity, hematologic and patient data from 61patients who were treated with sunitinib and 37 patients who were treated with sorafenib were available for analysis. The study population was typical of RCC patients entered onto clinical trials (Table 1). Consistent with previous reports,8, 9 the majority of sunitinib-treated patients and some sorafenib-treated patents developed hypothyroidism (as defined by ≥1 TFT biochemical abnormalities consistent with hypothyroidism occurring at some point during therapy). Patients treated with sorafenib were somewhat older than the patients treated with sunitinib (median age of 64 years vs 59 years, respectively; P = .03); however, the distributions of sex and pretreatment hematologic parameters were found to be similar (P > .15 in all cases).
| Characteristic | Sunitinib-treated Patients (n=61) | Sorafenib-treated Patients (n=37) |
|---|---|---|
| Gender | ||
| Male | 47 (77%) | 30 (81%) |
| Female | 14 (23%) | 7 (19%) |
| Age, y | ||
| Median (range) | 59 (24-76) | 64 (35-87) |
| Treatment duration, mo | ||
| Median | 9 | 8 |
| Range | 3-23 | 3-16 |
Change in MCV and Other Hematologic Parameters
Among patients treated with sunitinib, the MCV increased significantly at 3 months compared with baseline (Table 2 (median increase of 5.1 femtoliters [fL]; P < .001). A total of 41 patients (67%) had an MCV >100 fL at some point after the initiation of sunitinib therapy. In addition, platelet and WBC counts decreased significantly, whereas hemoglobin remained fairly stable. Similar patterns of change were noted when comparing patients who received less than compared with those who received greater than 12 months of therapy (data not shown). Figure 1 depicts a representative peripheral blood smear of a sunitinib-treated patient who developed macrocytosis. Because of the retrospective nature of this review, serum studies to rule out other causes of macrocytosis were only able to be drawn on a subset of patients who were still receiving therapy of variable duration at the time of analysis. The median serum vitamin B12 level was 422 pg/mL (12 patients; range, 220 pg/mL–1249 pg/mL [laboratory normal range, 221 pg/mL–700pg/mL]) and the median serum folate level was 15.4 ng/mL (10 patients; range, 10.4 ng/mL->18 ng/mL [laboratory normal range, 2.8 ng/mL-18 ng/mL]). Available data indicated a normal reticulocyte count in the subset of patients for whom data were available and suggests that compensatory reticulocytosis in response to drug-induced anemia is an unlikely cause for macrocytosis. Similarly, spurious macrocytosis was unlikely because of the absence of RBC agglutination in several of the peripheral blood smears available. Bone marrow analysis of 4 patients who had an MCV increase while receiving sunitinib demonstrated a hypocellular bone marrow (<5% [2 patients], 10%, and 10%-20%) with trilineage hematopoiesis and no evidence of metastasis. One of these patients had undergone a bone marrow examination 5 years prior for unexplained neutropenia, with normal morphology and 30% cellularity reported (Figs. 2a and 2b). In addition, in all patients there was normoblastic maturation of erythroid precursors and no evidence of hypersegmented neutrophils.
Representative peripheral blood smear of a sunitinib-treated patient who developed macrocytosis demonstrating anisocytosis, numerous macrocytes, and normochromasia.
Representative bone marrow (a) before therapy with trilineage hematopoiesis and cellularity appropriate for age and (b) after 14 months of sunitinib therapy, demonstrating a hypocellular bone marrow with trilineage hematopoiesis and no abnormal predominance of any of the myeloid lineages.
| Mean Corpuscular Volume (Normal Reference Range, 80-100 fL) | ||
|---|---|---|
| Median, fL (range) | P | |
| Baseline | 90.2 (72.8-108) | — |
| End of 3 mo | 97.1 (77.7-117.1) | — |
| End of treatment | 102.2 (84.1-120.1) | — |
| Absolute change from baseline at 3 mo | 5.1 (-3.1-20.8) | <.001 |
| Absolute change from 3 mo to end of treatment* | 4.9 (-1.7-16.2) | <.001 |
| Hemoglobin (Normal Reference Range, 13.5-17.5 g/dL) | ||
| Median, g/dL (range) | P | |
| Baseline | 13.4 (8.5-18.0) | — |
| End of 3 mo | 12.8 (8.6-17.5) | — |
| End of treatment | 11.8 (9.3-15.6) | — |
| Absolute change from baseline at 3 mo | -0.2 (-4.7-3.4) | .31 |
| Absolute change from 3 mo to end of treatment* | -0.6 (-4.0-0.7) | .01 |
| Platelets (Normal Reference Range, 150-400 K/uL) | ||
| Median, K/uL (range) | P | |
| Baseline | 199 (77-508) | — |
| End of 3 mo | 117 (22-353) | — |
| End of treatment | 165 (60-419) | — |
| Absolute change from baseline at 3 mo | -80 (-396-81) | <.001 |
| Absolute change from 3 mo to end of treatment* | 7.5 (-76-99) | .97 |
| White Blood Cell Count (Normal Reference Range, 4-11 K/uL) | ||
| Median, K/uL (range) | P | |
| Baseline | 6.5 (2.9-14.0) | — |
| End of 3 mo | 3.8 (1.9-8.7) | — |
| End of treatment | 4.0 (2.4-15.2) | — |
| Absolute change from baseline at 3 mo | -2.6 (-10.5-1.6) | <.001 |
| Absolute change from 3 mo to end of treatment* | 0.3 (-1.7-5.7) | .05 |
- RCC indicates renal cell carcinoma; fL, Femotliters.
- * Restricted to 18 patients who received ≥12 months of treatment.
In patients who received sunitinib for ≥12 months (18 patients), the MCV continued to increase throughout treatment. Ten patients in whom the MCV increased during sunitinib treatment had laboratory data available approximately 2 months to 3 months after sunitinib was discontinued. Among these patients, the MCV values tended to decrease relative to their final on-treatment assessment (median decrease of 9.7 fL; P = .002), hemoglobin tended to remain stable (median decrease of 0.1 g/dL; P = .56), and the platelet and WBC counts tended to increase (median increases of 30.5 K/μL [P = .06] and 1.9 K/μL [P = .01], respectively). These patients did not receive subsequent treatment at the time of laboratory assessment while not receiving sunitinib.
There was no correlation reported between the development of hypothyroidism and pretreatment MCV, hemoglobin, platelet counts, or WBC counts. Patients who were euthyroid tended to have smaller increases in the MCV at 3 months compared with patients who developed hypothyroidism (median increase of 3.9 fL vs 6.2 fL, respectively; P = .03). A total of 28 patients (46%) who received sunitinib developed hypothyroidism (as defined by ≥1 TFT biochemical abnormalities consistent with hypothyroidism noted at some point during therapy). Twenty-one patients (51%) who developed macrocytosis (MCV >100 fL) also developed hypothyroidism. There was no association noted between other specific hematologic toxicities and other toxicities with sunitinib and no association between MCV changes and clinical outcomes such as objective response or time to disease progression was observed.
As with sunitinib, the WBC count decreased after 3 months of therapy with sorafenib (P = .04) (Table 3). However, in contrast to sunitinib, there was no change in MCV noted (P = .30), hemoglobin increased (P = .02), and the platelet count remained fairly stable (P = .62).
| Mean Corpuscular Volume (Normal Reference Range, 80-100 fL) | ||
|---|---|---|
| Median, fL (range) | P | |
| Baseline | 90.7 (77.0-106.3) | — |
| End of 3 mo | 90.6 (79.7-99.0) | — |
| End of treatment | 89.5 (74.9-98.6) | — |
| Absolute change from baseline at 3 mo | -0.3 (-16.9 -5.2) | .30 |
| Absolute change from 3 mo to end of treatment* | 1.3 (-9.9 -4.9) | .36 |
| Hemoglobin (Normal Reference Range, 13.5-17.5 g/dL) | ||
| Median, g/dL (range) | P | |
| Baseline | 13.2 (7.4-17.9) | — |
| End of 3 mo | 14.0 (9.7-18.2) | — |
| End of treatment | 12.8 (8.8-17.9) | — |
| Absolute change from baseline at 3 mo | 0.4 (-2.2 –7.7) | .02 |
| Absolute change from 3 mo to end of treatment* | -0.7 (-5.0 -0.6) | <.001 |
| Platelets (Normal Reference Range, 150-400 K/uL) | ||
| Median, K/uL (range) | P | |
| Baseline | 198.0 (64.0-386.0) | — |
| End of 3 mo | 198.0 (109.0-356.0) | — |
| End of treatment | 206 (115-409) | — |
| Absolute change from baseline at 3 mo | 0.5 (-115 –101) | .62 |
| Absolute change from 3 mo to end of treatment* | 14 (-60 –169) | .06 |
| White Blood Cell Count (Normal Reference Range, 4-11 K/uL) | ||
| Median, K/uL (range) | P | |
| Baseline | 7.0 (4.1-12.4) | — |
| End of 3 mo | 6.8 (3.8-11.6) | — |
| End of treatment | 7.2 (3.5-12.4) | — |
| Absolute change from baseline at 3 mo | -0.5 (-3.5 –5.1) | .04 |
| Absolute change from 3 mo to end of treatment* | 0.7 (-2.8 -5.4) | .08 |
- RCC indicates renal cell carcinoma; fL, Femtoliters.
- * Restricted to 26 patients who received ≥6 months of treatment.
DISCUSSION
Sunitinib and sorafenib are targeted anticancer agents that are in common use in metastatic RCC and currently are being investigated in a wide variety of malignancies. This retrospective review describes the hematologic effects of sunitinib and sorafenib in metastatic RCC patients treated for at least 3 months. The MCV of RBCs increased throughout sunitinib therapy and was associated with a nonsignificant decline in hemoglobin levels. Although the MCV increase was greater in those patients who developed biochemical hypothyroidism, macrocytosis was not restricted to this subset. Furthermore, no alternative etiology of macrocytosis, including alternative drugs or vitamin B12/folate deficiency, was identified and sunitinib-induced macrocytosis was found to be reversible with drug discontinuation. These observations support sunitinib as a cause of drug-induced macrocytosis.
Macrocytosis is defined as an MCV of > than 100fL. Macrocytosis has been reported in 2% to 4% of unselected patients, with the most common causes being therapeutic agents, alcoholism, liver disease, hypothyroidism, and the megaloblastic anemias including vitamin B12 and folate deficiency.10 RBCs enter the circulation from the bone marrow as reticulocytes, which are macrocytic with an MCV ranging from 100fL to 125 fL until remodeling leads to reductions in the cell water content and membrane. Macrocytic anemia is divided into 2 categories: megaloblastic and nonmegaloblastic. Nonmegaloblastic causes of normoblastic macrocytosis are alcoholism, liver disease, hypothyroidism, primary bone marrow disorders, and certain drugs, most notably zidovudine, azathioprine, methotrexate, and capecitabine. None of these alternative etiologies were identified in the current study cohort. Excessive reticulocytes in the peripheral blood from a variety of causes including anemia can also lead to an overall increase in MCV. Although hemoglobin levels in the sunitinib-treated cohort decreased over the duration of treatment, this decline was nonsignificant and there was no evidence of reticulocytosis leading to the MCV increase observed. Megaloblastic anemia results from defective DNA synthesis leading to ineffective or dysplastic erythropoiesis, most commonly from vitamin B12 or folate deficiency. Although data were not available for the entire cohort, no evidence of nutritional deficiency was identified in any of the patients assayed. Furthermore, the decline in the MCV to normal levels when patients were withdrawn from sunitinib therapy argues against an alternative underlying etiology.
The results of the current study are in contrast to recent reports of sunitinib-induced macrocytosis, which implicated vitamin B12 deficiency in the etiology of macrocytosis.6, 7 These studies also observed similar degrees of macrocytosis, but were limited by smaller sample sizes and their retrospective nature. One of the studies noted that vitamin B12 deficiency preceded macrocytosis and persisted after vitamin B12 replacement therapy, thus implicating additional or alternative mechanisms to sunitinib-induced macrocytosis.6 An additional recent preliminary report of RCC patients receiving sunitinib supports the observations made in the current study cohort, with a 48% incidence of macrocytosis and a lack of identifiable alternative etiologies.11
The current study is limited by the retrospective nature of the review. Although all patients who received sunitinib or sorafenib for at least 3 months were included, we were unable to undertake any prospective examination to more thoroughly investigate additional etiologic factors that might contribute to macrocytosis. Similarly, the measurement of serum factors such as vitamin B12 and folate was undertaken in only a subset of patients. However, to our knowledge, the current study represents the largest number of sunitinib-treated RCC patients with a comprehensive blood count analysis presented to date.
Although the mechanism of sunitinib-induced macrocytosis is not clear, data from a report of imatinib treatment in patients with gastrointestinal stromal tumors (GISTs) may provide insight.12 GISTs prominently exhibit gain-of-function mutations in the stem cell factor receptor (c-KIT) gene, resulting in constitutive activation of the c-KIT protein and its downstream signals. Imatinib is a small molecule that has antitumor effect in GISTs through inhibition of c-KIT. Macrocytosis was noted in 42% of GIST patients receiving imatinib with no other etiologies identified. The common mechanism for imatinib and sunitinib that may lead to macrocytosis is c-KIT inhibition.2 It is interesting to note that sorafenib has much weaker inhibitory activity against c-kit, and macrocytosis was not observed with sorafenib treatment in this cohort.3 In support of the hypothesis that sunitinib could cause macrocytic anemia through inhibition of c-kit, the bone marrow of the patients treated with sunitinib was found to be hypocellular, with decreased trilineage hematopoiesis, indicating a possibly low proliferation rate in thestem cell compartment. The precise mechanisms of c-kit–mediated macrocytosis requires further investigation.
In conclusion, sunitinib treatment leads to RBC macrocytosis in patients with metastatic RCC. The correlation between macrocytosis and the development of anemia and hypothyroidism, and levels of nutritional factors such as vitamin B12 and folate have not been fully elucidated, but data presented in the current study indicate sunitinib as a direct cause of macrocytosis. The mechanism by which sunitinib leads to increases in RBC volume may be mediated by c-kit inhibition. The return of the MCV to normal after the discontinuation of sunitinib without other intervention supports the idea that this phenomenon is temporary and reversible. The occurrence of macrocytosis with sunitinib treatment does not preclude the continuation of treatment and does not appear to impact clinical efficacy or toxicity. However, this phenomenon may compromise blinding in placebo-controlled sunitinib studies and could subsequently lead to unintended bias in the treatment or interpretation of clinical outcomes such as objective response. Routine monitoring of CBC and RBC indices during sunitinib therapy is warranted, and sunitinib should be considered as a possible cause for otherwise unexplained macrocytosis.
