Abstract
Background: Bevacizumab (BV) is widely used in chemotherapy for metastatic colorectal cancer (mCRC). Although specific adverse events have been observed, their risk factors have not been clarified. Patients and Methods: 178 mCRC patients who underwent chemotherapy were retrospectively examined and correlations between possible risk factors and adverse events were analyzed. Results: 87 out of 178 patients were treated with BV-containing chemotherapy. Possible risk factors for BV-related adverse events were: remaining primary tumor, current bleeding, history of arterial thromboembolism (ATE), hypertension, and proteinuria, and these were observed in 22%, 2%, 7%, 16%, and 8% of patients, respectively. Patients with hypertension prior to chemotherapy developed significantly worse hypertension (p=0.018). Gastrointestinal bleeding occurred in 3 out of 18 patients with residual primary tumor (16.7%) and 6 out of 63 patients with no primary tumor (8.7%) (p=0.385). Conclusion: Pre-existing hypertension appears to be a risk factor for BV-related deterioration of hypertension.
Systemic chemotherapy has been developed as a standard therapy against metastatic colorectal cancer (mCRC), and therapeutic outcomes have since improved. In addition to cytotoxic agents, inhibition of vascular endothelial growth factor (VEGF) has been one of the key strategies for several types of solid tumors whose growth depends on neovascularization (1). VEGF is a glycoprotein physiologically-regulating vascular permeability and neovascularization. Moreover, tumor cells often produce large amounts of VEGF and depend on VEGF-induced neovascularization supplying oxygen and nutrition.
Bevacizumab (BV) is a recombinant, humanized, monoclonal antibody against human VEGF-A that inhibits the binding of VEGF-A to VEGFR and subsequent growth of endothelial cells. It has also been reported that normalization of abnormal tumor vasculature by BV contributes to maintaining interstitial pressure of tumor tissue and effectively delivering drugs to the tumor cells (2-4). BV has been well-investigated for its therapeutic effects on mCRC in clinical trials (5-8). In combination with fluorouracil-based regimens, BV has been reported to provide longer progression-free survival and overall survival in large-scale, randomized, clinical trials (5-7). Based on these results, BV was approved for the first-line and second-line treatment of mCRC in the United States in 2004 and in Japan in 2007.
Although BV shows remarkable clinical benefits, specific adverse events including not only frequently-observed proteinuria and hypertension, but also more serious conditions including gastrointestinal (GI) perforation, bleeding, and arterial thromboembolism (ATE), have been reported (5-7). Thus, the risks for the occurrence of adverse events should be assessed. BRiTE, a prospective, observational cohort study of BV-based chemotherapy for mCRC, showed the possible risk factors for GI perforation and ATE: age above 65 years, intact primary tumors, and prior adjuvant radiotherapy were related to GI perforation, and performance status greater than 1, hypertension, and past history of arterial disease were related to ATE (9).
Little information regarding the effects and adverse events of BV-containing chemotherapies in Japanese mCRC patients is available, because few prospective, large-scale, clinical studies using BV-containing regimens for mCRC have been performed in Japan. Therefore, predictive factors of BV-related adverse events and precise guidelines for administration of BV to Japanese mCRC patients are not well-established. In order to clarify the relationship between predictive factors and adverse events of BV-containing chemotherapy in Japanese patients, patients' background characteristics and frequencies of adverse events with systemic chemotherapy for mCRC were retrospectively investigated.
Patients and Methods
Patients. The present study retrospectively investigated 178 patients with unresectable, recurrent, or metastatic colorectal adenocarcinoma who started systemic chemotherapy during the period from June 11, 2007 to August 31, 2008 in three Institutions: the Department of Hematology and Oncology of Kyushu University Hospital, the Gastrointestinal and Medical Oncology Division of National Kyushu Cancer Center, and the Department of Medical Oncology in Hamanomachi Hospital, Fukuoka, Japan. Patients who were administered BV prior to this period were excluded.
Information of all cases was obtained from the participating Hospitals' medical records. Patients' background characteristics included age, sex, Eastern Cooperative Oncology Group (ECOG) performance status (PS), primary site of the tumor, and possible risk factors for serious adverse events of BV including remaining primary tumor, current bleeding, past and present history of arterial thromboembolism (ATE), hypertension, and proteinuria. In patients who were treated with BV-containing regimens during this survey period (BV group), chemotherapy regimens and their duration, best therapeutic effects, adverse events, and reasons for withdrawal of BV administration were investigated. In patients who were treated with chemotherapy without BV (CT group), the reasons for the primary decision to not use BV-containing regimens when chemotherapy was started were examined. Adverse events that occurred in the CT group were also observed in the above period. The present study was carried out according to the regulations of local Ethics committee in each Hospital and the Declaration of Helsinki.
Treatment. Patients received fluoropyrimidine-based chemotherapy with or without BV until disease progression, occurrence of intolerable adverse events, or patient's refusal. The combined chemotherapy regimens were selected by the physicians with the patients' consent. The dose of BV was 5 mg/kg or 10 mg/kg every 2 weeks. BV was administered intravenously in 100 ml of saline for 30-90 min. In cases of adverse events, the BV dose was not reduced, but administration of BV was suspended until recovery from the adverse events.
Assessments. At every bi-weekly visit, patients were assessed for adverse events by physical examination, urinalysis, blood cell counts, and serum chemistry. All adverse events were evaluated according to the Common Terminology Criteria for Adverse Events (CTC-AE) version 3.0. The most severe grades of adverse events during chemotherapy were recorded.
Statistics. Analyses of correlations between the occurrence of BV-specific adverse events and their possible risk factors were performed using the Fisher's exact test.
Results
Patients' background characteristics and treatments. Overall, 87 of 178 patients were treated with BV-containing chemotherapy (BV group) during the observation period, and 91 patients were treated with non-BV-containing chemotherapy (CT group). No significant differences in sex, age, and primary site were identified between the two groups (Table I). Patients who had primary tumors tended to be treated with chemotherapy without BV. Bleeding risks were significantly more frequently observed in the CT group. No significant differences between the two patient groups were observed in the other factors.
In the BV group, BV-containing regimens were used in the first (17%), second (35%), and third line (29%) of chemotherapy (Table II). Combined chemotherapy regimens were oxaliplatin-based FOLFOX (48%) and irinotecan-based FOLFIRI or IFL (44%). The median number of BV administrations was 7, and the median duration of BV-containing chemotherapy was 8.1 months. In 71 regimens of BV-containing chemotherapy in which the therapeutic effect was confirmed until August 31, 2008, the response rate and the disease control rate were 18% and 78% in all lines of treatment, respectively, and 27% and 82% in first-line treatment, respectively (Table III). There were 15 patients who were treated with two BV-containing chemotherapy regimens. In 10 of these 15 patients, BV was administered after disease progression in their prior chemotherapy with BV. The other 5 patients stopped chemotherapy because of oxaliplatin-induced allergies (3 patients), diarrhea (1 patient), and patient's refusal (1 patient), but BV was continued for the subsequent chemotherapies including irinotecan-based or oxaliplatin-based regimens, which were not used in the prior therapy. The confirmed responses of the second BV-containing chemotherapy in 5 patients were 3 stable diseases and 2 progressive diseases.
Safety. Therapy-related toxicities in the BV group were assessed according to the Common Terminology Criteria of Adverse Events version 3.0 (CTC-AE ver.3.0) (Table IV). In terms of hematological toxicities, neutropenia was observed in 87% of patients (grade 3/4; 41%), anemia in 79% (grade 3/4; 5%) and thrombocytopenia in 17% (grade 3/4; 2%). Febrile neutropenia occurred in 8%. Concerning non-hematological toxicities, general fatigue was observed in 71%, anorexia in 76%, nausea and vomiting in 61%, and peripheral neuropathy in 64%. All grades of diarrhea occurred in 54% of patients, and more than grade 3 was seen in 5%. In BV-related toxicities, bleeding was observed in 23% of patients (10 cases of nasal bleeding, 1 case of gingival bleeding, 9 cases of gastrointestinal bleeding), hypertension in 22%, proteinuria in 15%, and delayed wound healing in 4%. There were no treatment-related deaths.
Baseline characteristics of all patients.
Characteristics of the systemic chemotherapy regimens of the BV group.
Efficacy of chemotherapy in the BV group.
Risk factors for BV-related adverse events. The relationships between BV-related adverse events and possible risk factors were analyzed. Gastrointestinal perforation did not occur in patients with or without primary tumor lesions. As mentioned above, 9 patients had gastrointestinal bleeding, including 4 with melena, 4 with bleeding from the stoma, and 1 proctorrhagia, and no correlation was observed between remaining primary tumors and bleeding (Table V). No arterial thrombosis events occurred in the patients with or without a present or past history of thrombotic events. On the other hand, patients with hypertension before starting BV-containing chemotherapy had significant deterioration of hypertension (p=0.02) (Table VI). Therapy-related deterioration of proteinuria was not correlated with presence of proteinuria before chemotherapy (Table VII).
Adverse events in the BV group.
BV-containing regimen-related gastrointestinal bleeding in patients with or without primary tumor.
Forty-eight patients in the BV group continued to the end of this surveillance period in August 2008, and 39 patients had stopped BV-containing chemotherapy because of progressive disease (20 patients) and toxicity (11 patients).
Reasons for non-BV chemotherapy in the CT group. The reasons why BV was not administered in the CT group were also determined (Table VIII). The most frequent reason for this was the risk of bleeding (11 patients, 12%); in detail, 5 patients in the CT group had active bleeding from the primary site, 3 had bleeding from sites of tumor invasion, and 2 had bleeding from invasion of genital organs and other sites. Eleven patients (12%) were not administered BV because due to patients' refusal. BV was not used in 10 patients with a current or past history of arterial thrombotic events and in 8 patients with severe peritoneal dissemination. Since 7 patients had surgery immediately prior to chemotherapy, the use of BV was reduced.
BV-containing regimen-related hypertension in patients with or without hypertension before therapy.
Discussion
BV has been widely used in systemic chemotherapy for mCRC. Various large-scale, phase III clinical studies, mainly conducted in Western countries, have shown the efficacy and safety profiles of BV-containing chemotherapy regimens (5-8). Tamiya et al. (10) reported on a retrospective review of adverse events in 65 Japanese patients with mCRC treated with BV-containing regimens and showed similar safety profiles to those of previous prospective trials. Although these reports encouraged the use of BV as one of the standard chemotherapeutic agents for mCRC, selection of patients who could be safely treated with BV-containing therapy is required in the clinical setting. It is, thus, important to identify the risk factors for predicting adverse events specifically caused by BV in Japanese patients.
The present study analyzed clinical data from patients with mCRC who had been treated with systemic chemotherapy during the period of 15 months from the approval of BV in Japan. The efficacy and safety of BV-containing chemotherapies and the possible risk factors related to BV-specific adverse events were identified. Since it had been known that bleeding occurred in around 3% of patients with BV-containing chemotherapies (5-8), mCRC patients with possible bleeding risks were excluded from BV administration in the CT group. Two patients, one with a history of hemorrhage from a gynecological organ and another with a remaining primary tumor, were treated with a BV-containing regimen and had no bleeding episodes during the treatment period. The significant risk factors for bleeding could not be identified because no hemorrhagic events occurred in either the BV group or the CT group. These facts suggested that active bleeding or highly hemorrhagic lesions might be appropriate risk factors for BV-related severe bleeding, and avoidance of BV administration in this population might be appropriate. It may be important to assess the bleeding risk in advance to prevent severe bleeding, and BV might be administered safely to patients with a low bleeding risk.
BV-containing regimen-related deterioration of proteinuria in patients with proteinuria before therapy.
The incidence of arterial thrombosis has been reported to be 1.7% in a cytotoxic chemotherapy-alone group and 3.8% in a BV-containing chemotherapy group, in a meta-analysis including clinical trials for mCRC, non-small cell lung cancer, and breast cancer (11). The multivariate analysis revealed that age over 65 years (p=0.01) and a history of arterial thrombosis (p<0.001) were independent risk factors for ATE (11). In the present study, BV was not given to 10 patients with any risk of thrombosis. Given the risk factors demonstrated in the previous report, 339 cases were >65 years of age, and 89 cases had a history of thrombosis. In contrast, 6 patients who had had possible risks for thrombosis, such as a history of brain infarction in 4 patients, arteriosclerosis obliterans in 1, and ischemic heart disease in 1, were successfully treated in the present study. Importantly, the disease activities of thrombosis in all 6 patients were diagnosed as minimal prior to chemotherapy, and no severe thrombotic events occurred in these patients. Establishment of precise criteria for administering BV in patients with a history of thrombotic events is also required.
In the present study, hypertension prior to BV-containing therapy was identified as a significant risk factor for the induction of worse hypertension by BV. On the other hand, in the BRiTE study, de novo hypertension appeared in 22% of BV-treated patients without baseline hypertension, and deterioration of hypertension in BV-treated patients with baseline hypertension was similarly observed in 22% (9). Ethnic differences might explain for the difference between present findings and those of the previous study. It has been reported that the sensitivity to sodium differs between Asians and Caucasians or Africans, suggesting that molecular mechanisms of BV-induced hypertension may differ among races (12).
Identification of risk factors has two roles: prevention of severe toxicities induced by chemotherapy, and avoidance of losing the chance for patients to be treated with appropriate chemotherapy. Because patients harboring possible risks were often excluded from enrollment in clinical trials, prospective trials focusing on that population might be important. In addition, retrospective analysis might be helpful to reveal the safety and effectiveness of BV in such patient populations. The present study demonstrated that BV-containing therapy in Japanese mCRC patients showed a similar toxicity profile to the one previously reported. It was suggested that suitable assessment of possible risk factors could lead to safe treatment with BV. Since Japanese mCRC patients with hypertension had a higher possibility of worse hypertension with BV, careful follow-up is needed.
Reasons for choosing non-BV chemotherapy.
Acknowledgements
The Authors would like to thank Dr. Risa Tanaka for providing patient data. The Authors would also like to thank the medical staff of each institution who contributed to treatment of the patients.
- Received January 5, 2014.
- Revision received January 15, 2014.
- Accepted January 17, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.