- Email: [email protected]
Incidence and risk of hypertension with sorafenib in patients with cancer: a systematic review and meta-analysis
Articles
Incidence and risk of hypertension with sorafenib in patients with cancer: a systematic review and meta-analysis Shenhong Wu, John J Chen, Andrzej Kudelka, Janice Lu, Xiaolei Zhu
Summary Background Sorafenib is used in patients with advanced renal-cell carcinoma (RCC) or hepatocellular cancer, and its application in other types of cancers is also undergoing extensive clinical assessment. Hypertension is one of the major side-effects of this drug, and reported incidences vary substantially between clinical trials. We did a systematic review and meta-analysis of published clinical trials to establish the incidence of hypertension associated with sorafenib. The aim of this study is to gain a better understanding of the overall risk of hypertension in patients with cancer who receive sorafenib. Methods Databases, including Medline (July, 1966, to July, 2007), and Web of Science, and abstracts presented at the American Society of Clinical Oncology annual meetings from 2004 to 2007 were searched to identify relevant studies. Eligible studies were prospective clinical trials of patients with cancer assigned single-drug sorafenib at 400 mg twice daily with data on hypertension available. Incidence and relative risk (RR) of hypertension were calculated using a random-effects or fixed-effects model, depending on the heterogeneity of the included studies. Findings Nine studies published between January, 2006, and July, 2007, which included a total of 4599 patients with RCC or other solid tumours, were selected from 223 articles screened for analysis. For patients assigned sorafenib, the overall incidence of all-grade and high-grade (ie, grade 3 or 4) hypertension were 23·4% (95% CI 16·0–32·9%) and 5·7% (2·5–12·6%), respectively. No significant difference was noted between patients with RCC or a non-RCC malignancy (all grade: RR 1·03 [95% CI 0·73–1·45], p=0·89; high-grade: RR 1·23 [0·76–1·99], p=0·40) who were assigned sorafenib. Sorafenib was associated with a significantly increased risk of all-grade hypertension in patients with cancer with an RR of 6·11 (2·44–15·32], p<0·001) compared with controls. Interpretation Patients with cancer assigned sorafenib have a significant risk of developing hypertension. Appropriate monitoring and treatment is strongly recommended to prevent cardiovascular complications.
Introduction Sorafenib is a multikinase inhibitor with a broad spectrum of antitumour activity on cancer-cell proliferation and angiogenesis.1 It targets: Raf kinase; vascular endothelial growth factor receptors (VEGFR) 1, 2, and 3; platelet-derived growth factor receptor β (PDGFR β); FMS-like tyrosine kinase 3 (FLT-3); c-KIT; and RETreceptor tyrosine kinase.1 In animals, the antitumour effect of sorafenib is mainly mediated by its inhibition of angiogenesis.2 The clinical benefit of sorafenib was initially seen in a phase II randomised discontinuation trial, which showed a lengthened progression-free survival with sorafenib compared with placebo in patients with metastatic renal-cell carcinoma (RCC).3 Subsequently, a phase III randomised controlled trial (Treatment Approaches in Renal cancer Global Evaluation Trial [TARGET]) confirmed this finding.4 Additionally, sorafenib has clinical activity in advanced hepatocellular carcinoma.5,6 The efficacy of sorafenib in non-small-cell lung cancer, prostate cancer, and melanoma is currently undergoing assessment in a randomised phase II trial7 and in single-arm phase II clinical studies.8,9 As with many therapeutic drugs, sorafenib is associated with substantial side-effects. Diarrhoea, rash, fatigue, and hand and foot skin reactions were the most common adverse events associated with sorafenib in the TARGET http://oncology.thelancet.com Vol 9 February 2008
Lancet Oncol 2008; 9: 117–23 Published Online January 22, 2008 DOI:10.1016/S14702045(08)70003-2 See Reflection and Reaction page 86 Division of Medical Oncology, Department of Medicine, State University of New York, Stony Brook, NY, USA (S Wu MD, A Kudelka MD, J Lu MD); Department of Preventive Medicine, State University of New York, Stony Brook, NY, USA (J J Chen PhD); and Kidney Specialists of Long Island, Port Jefferson Station, NY, USA (X Zhu MD) Correspondence to: Dr Shenhong Wu, Division of Medical Oncology, Department of Medicine, State University of New York, Stony Brook, NY 11794-9447, USA shenhong.wu@stonybrook. edu
trial.4 Additionally, hypertension is a major side-effect that has been noted in trials, with its incidence ranging from 16·0–42·6%.3,4,7–11 The recognition and management of hypertension is an important issue in patients treated with sorafenib because poorly controlled hypertension can lead to serious cardiovascular events. In the TARGET trial, cardiac ischaemia and infarction were more common in patients assigned sorafenib than in those assigned placebo.4 Additionally, the use of sorafenib might be associated with reversible posterior leucoencephalopathy syndrome, a clinical event characterised by headache, seizures, impaired vision, and acute hypertension.12 Because of the limited number of patients in trials, the overall risk of hypertension with sorafenib is unclear. Therefore, we did a systematic review of the published work and a meta-analysis to assess this issue.
Methods Data source We did an independent review of citations from Medline (July, 1966, to July, 2007). Our search strategy included the terms “sorafenib”, “BAY 43-9006”, “cancer”, and “hypertension” and was restricted to human studies published in English. Additionally, we manually searched all abstracts containing the term “sorafenib” that were presented at the American Society of Clinical Oncology 117
Articles
Trial
Trial Design
Patients Sample size Median age, enrolled, n years
Underlying malignancy
Ratain et al3
Randomised phase II (placebo vs sorafenib)
202
202
58
RCC
Escudier et al4
Randomised phase III (placebo vs sorafenib)
903
451
58
RCC
Eisen et al7
Randomised phase II (placebo vs sorafenib)
502
37
53
Melanoma
Gatzemeier et al8
Single-arm phase II
52
52
62
NSCLC
Wu et al9
Single-arm phase II
22
22
64
Prostate
Figlin et al10
Expanded-access programme
2502
2502
63
RCC
Szczylik et al11
Randomised phase II (interferon vs sorafenib)
189
97
62
RCC
D’Adamo et al17
Single-arm phase II
134
111
55
Non-GIST sarcoma
Hobday et al18
Single-arm phase II
93
93
59
Neuroendocrine tumours
RCC=renal-cell carcinoma. NSCLC=non-small-cell lung cancer. GIST=gastrointestinal stromal tumour.
Study
Number of events Sample size 4
Escudier (2007) Eisen (2006)7 Ratain (2006)3 Szczylik (2007)11 Gatzemeier (2006)8 Wu (2006)9 Figlin (2007)10 Overall
76 6 86 24 12 7 401 612
Test of heterogeneity: Q=85·873, p<0·001, I²=93·013
B
Study Escudier (2007)4 Eisen (2006)7 Ratain (2006)6 Szczylik (2007)11 Gatzemeier (2006)8 Wu (2006)9 Figlin (2007)10 D'Adamo (2007)17 Hobday (2007)18 Overall
Incidence (95% CI)
451 37 202 97 52 22 2502 3363
0·169 (0·137–0·206) 0·162 (0·075–0·317) 0·426 (0·359–0·495) 0·247 (0·172–0·343) 0·231 (0·136–0·364) 0·318 (0·160–0·534) 0·160 (0·146–0·175) 0·234 (0·160–0·329)
0
0·50
1·00
Number of events Sample size 16 5 62 2 2 1 125 3 4 220
Test of heterogeneity: Q=167·076, p<0·001, I²=95·212
Incidence (95% CI)
451 37 202 97 52 22 2502 111 93 3567
0·035 (0·022–0·057) 0·135 (0·057–0·286) 0·307 (0·247–0·374) 0·021 (0·005–0·079) 0·038 (0·010–0·141) 0·045 (0·006–0·261) 0·050 (0·042–0·059) 0·027 (0·009–0·080) 0·043 (0·016–0·109) 0·057 (0·025–0·126) 0
0·50
1·00
Incidence (1=100%)
Figure 1: Annotated forest plot for meta-analysis of incidence of hypertension in patients with cancer who were assigned sorafenib Summary incidences of all-grade (A) and high-grade (B) hypertension are calculated using a random-effects model. Size of squares is directly proportional to amount of information available.
(ASCO) annual meetings and ASCO Prostate Cancer Symposiums from 2004 to 2007. An independent search by use of the citation database Web of Science was also done to ensure that no studies were missed. Details on trial design, patient characteristics, treatment information, findings, and follow-up were used from these 118
Study selection Sorafenib has been approved for use in patients with advanced RCC and hepatocellular carcinoma as a single drug at the starting dose of 400 mg twice daily. To ensure clinical significance, we assessed the risk of hypertension with sorafenib at this dose. Therefore, phase I trials were excluded from the analysis due to variations in dosage. We included phase II and III clinical trials, and expandedaccess programmes, in which sorafenib as a single drug was given at the standard dose. Trials that met the following criteria were chosen for analysis: prospective clinical trials in patients with cancer; patients assigned to treatment with sorafenib as a single drug at a starting dose of 400 mg twice a day; data available for the events or incidences of hypertension.
Clinical endpoints
Table 1: Characteristics of clinical trials and patients included in the meta-analysis
A
trials. The entire selection process (ie, search, selection, and exclusion of studies) was done by SW.
Clinical endpoints were selected from the safety profile of each trial. Hypertension was recorded according to versions 2 or 3 of the Common Terminology Criteria for Adverse Events (CTCAE),13 each of which are the same regarding the grading of hypertension as follows: grade 1—asymptomatic, transient (<24 h) increase in blood pressure by more than 20 mm Hg (diastolic) or to more than 150/100 mm Hg if previously within normal limit, intervention not indicated; grade 2—recurrent or persistent (>24 h) or symptomatic increase by more than 20 mm Hg (diastolic) or to more than 150/100 mm Hg if previously within normal limit, monotherapy might be indicated; grade 3—more than one drug needed for treatment or more intensive treatment than used previously; grade 4—life-threatening consequences (eg, hypertensive crisis). We included the incidence of hypertension of grade I or above in our analysis.
Statistical analysis All statistical analyses were done using version 2 of the Comprehensive MetaAnalysis programme (Biostat, Englewood, NJ, USA). The numbers of patients with hypertension, both all-grades and high-grades (grade 3 and 4), were summarised from the extracted data on all patients assigned sorafenib at 400 mg twice daily in the clinical trials included for analysis. For each trial, the proportion of patients with hypertension was calculated and the 95% CI was derived. For studies with a control group, the relative risk (RR) of hypertension in patients assigned to sorafenib was also calculated, and was compared only with those assigned to a control treatment in the same trial. For the meta-analysis, both the fixed-effects model (weighted with inverse variance) and the random-effects model were considered.14,15 For each meta-analysis, the Cochrane’s Q statistic was first calculated to assess the heterogeneity of the included trials. For p values less than 0·1, the assumption of homogeneity was deemed http://oncology.thelancet.com Vol 9 February 2008
Articles
invalid,16 and the random-effects model was used only after substantial efforts were made to explore the possible reasons for the heterogeneity. Otherwise, data were assessed using both the fixed-effects model and the random-effects models. A two-tailed p value of less than 0·05 was deemed statistically significant. All the statistical analyses were done by SW and JJC.
A
http://oncology.thelancet.com Vol 9 February 2008
76 86 24 401 587
Escudier (2007) Ratain (2006)3 Szczylik (2000)11 Figlin (2007)16 Overall
B
Study
Incidence (95% CI)
451 202 97 2502 3252
Test of heterogeneity: Q=82·596, p<0·001, I²=96·368
0·169 (0·137–0·206) 0·426 (0·359–0·495) 0·247 (0·172–0·343) 0·160 (0·146–0·175) 0·236 (0·143–0·365) 0
0·50
1·00
Number of events Sample size 4
Escudier (2007) Ratain (2006)3 Szczylik (2007)11 Figlin (2007)10 Overall
16 62 2 125 205
Incidence (95% CI)
451 202 97 2502 3252
0·035 (0·022–0·057) 0·307 (0·247–0·374) 0·021 (0·005–0·343) 0·050 (0·042–0·059) 0·065 (0·018–0·211) 0
Test of heterogeneity: Q=159·079, p<0·001, I²=98·114
Results Our search yielded a total of 223 articles on sorafenib from the published work. After reviewing each publication, we identified 16 original studies including randomised controlled trials, single-arm phase II studies, and case reports. From these 16 articles, three clinical trials matched our inclusion criteria.3,4,7 Of note, the phase II and III trials of sorafenib in patients with hepatocellular carcinoma were excluded from the analysis because of no available data regarding hypertension.5,6 From the abstracts published during recent ASCO annual meetings and ASCO Prostate Cancer Symposiums (2004–2007), we identified 68 abstracts that were related to sorafenib. After reviewing each abstract, we included six additional trials in our meta-analysis (table 1).8–11,17,18 Of the nine trials included, three were blinded3,4,11 and the rest were openlabelled. Three trials were sponsored by Bayer Pharmaceuticals (West Haven, CT, USA) and Onyx Pharmaceuticals (Emeryville, CA, USA).3,4,11 One trial in prostate cancer was supported by the US National Cancer Institute.9 Support for the other trials was not specified. A total of 4599 patients were available for analysis, with 3567 patients assigned treatment with sorafenib as a single drug. Baseline characteristics of these patients from the nine clinical trials are listed in table 1. Hypertension was not mentioned as a pre-existing condition in any of these trials. Baseline Eastern Cooperative Oncology Group performance status for most patients was between 0 and 1. Underlying metastatic malignancies included RCC, melanoma, non-small-cell lung cancer, prostate cancer, non-gastrointestinalstromal-tumour (GIST) sarcoma, and neuroendocrine tumours. For the randomised discontinuation trial by Eisen and co-workers,7 a total of 502 patients with various tumour types were enrolled, but the publication only focused on the 37 treated patients with advanced melanoma. Treatment options were randomly assigned in four clinical studies including two randomised controlled trials and two randomised discontinuation trials.3,4,7,11
Number of events Sample size 4
Role of the funding source This study was partially funded by the Research Foundation of the State University of New York, Stony Brook, NY, USA. This foundation played no role in the design, collection, analysis, interpretation, and writing of the study. The corresponding author had access to all the raw data and had the final responsibility to submit the manuscript for publication.
Study
0·50
1·00
Incidence (1=100%)
Figure 2: Annotated forest plot for meta-analysis of incidence of sorafenib-associated hypertension in patients with renal-cell carcinoma Summary incidences of all-grade (A) and high-grade (B) hypertension are calculated using a random-effects model. Size of squares is directly proportional to amount of information available.
A
Study
Number of events Sample size 7
Eisen (2006)
Gatzemeier (2006)8 Wu (2006)9 Overall
6
37
0·162 (0·075–0·317)
12
52
0·231 (0·136–0·364)
7
22
0·318 (0·160–0·534)
25
111
0·230 (0·160–0·319)
Test of heterogeneity: Q=1·897, p=0·387, I²<0·001
B
Study
Incidence (95% CI)
0
0·50
1·00
Number of events Sample size
Eisen (2006)7 Gatzemeier (2006)8
5 2
Wu (2006)9 D'Adamo (2007)17 Hobday (2007)18 Overall Test of heterogeneity: Q=6·376, p=0·173, I²=37·263
Incidence (95% CI)
37 52
0·135 (0·057–0·286) 0·038 (0·010–0·141)
1
22
0·045 (0·006–0·261)
3 4
111 93
0·027 (0·009–0·080) 0·043 (0·016–0·109)
15
315
0·053 (0·027–0·100) 0 0·50 1·00 Incidence (1=100%)
Figure 3: Annotated forest plot for meta-analysis of incidence of sorafenib-associated hypertension in patients with non-renal-cell-carcinoma malignancies Summary incidences of all-grade (A) and high-grade (B) hypertension are calculated using the fixed-effects model. Size of squares is directly proportional to amount of information available.
Data for all-grade hypertension was available for analysis from a total of 3363 patients enrolled in seven trials who had various solid tumours and were assigned sorafenib. The incidence of all-grade hypertension ranged between 16·0% and 42·6%, with the lowest incidence noted in the expanded access programme,10 and the highest in the phase II placebo-controlled randomised discontinuation trial in patients with metastatic RCC3 Meta-analysis showed the heterogeneity of the included studies (Q=85·9, p=0·0000). We explored the potential source of heterogeneity by analysing blinded and open-label trials separately. Our analysis showed that heterogeneity exists 119
Articles
Study
Number of events/sample size Sorafenib Control
Escudier (2007)4 Szczylik (2007)18 Overall
76/451 24/97
8/451 6/90
100/548
14/541
Risk ratio and 95% CI
0·01
Test of heterogeneity: Q=2·755, p=0·097, I²=63·705
0·1 Control
1
10 Sorafenib
Risk ratio (95% CI)
Z-value
p-value
9·5000 (4·6393–19·4533) 3·7113 (1·5905–8·6603)
6·1564 3·0333
0·0000 0·0024
6·1082 (2·4357–15·3183)
3·8577
0·0001
100
Figure 4: Relative risk of sorafenib-associated hypertension versus control from the two randomised controlled trials of patients with renal-cell carcinoma Summary relative risk (RR) was calculated using the random-effects model. Size of squares is directly proportional to amount of information available.
Study
Number of events Sample size
Johnson (2004) Kabbinavar (2003) Kindler (2005) Miller (2005) Yang (2003) Overall Test of heterogeneity: Q=3·933, p=0·415, I²<0·001
6 9 13 54 14 96
Incidence (95% CI)
34 32 47 229 39 381
0·176 (0·081–0·341) 0·281 (0·153–0·458) 0·277 (0·168–0·420) 0·236 (0·185–0·295) 0·359 (0·225–0·519) 0·254 (0·213–0·301) 0
0·50
1·00
Incidence (1=100%)
Figure 5: Annotated forest plot for meta-analysis of incidence of high-dose bevacizumab-associated hypertension in patients with cancer Studies included in the analysis were derived from our previous study for high-dose bevacizumab.19 Summary incidence of all-grade hypertension is calculated using the fixed-effects model. Size of squares is directly proportional to amount of information available.
in the three blinded studies,3,4,7 but not in the four openlabel studies.8–11 We were unable to find a compelling reason to exclude these blinded studies from the final analysis. As analysed by the random-effects model, the overall incidence of all-grade hypertension was 23·4% (95% CI 16·0–32·9) in patients who were assigned sorafenib (figure 1A). High-grade hypertension needs to be treated with more than one drug or with a more intense treatment than used previously (grade 3), or is associated with lifethreatening consequences (grade 4). It can result in substantial morbidity, dose reduction, or discontinuation of sorafenib. Data for high-grade hypertension was available for analysis from the total 3567 patients enrolled in the nine trials. The incidence of grade 3 or 4 hypertension ranged from 2·1% to 30·7%, with the lowest incidence noted in the phase II trial in patients with RCC,11 and the highest incidence again noted in the phase II randomised discontinuation trial in patients with RCC.3 The overall incidence of high-grade hypertension was 5·7% (95% CI 2·5–12·6), as determined by the random-effects model (figure 1B). Patients with RCC can be more susceptible to developing hypertension because of previous nephrectomy and renal dysfunction. Therefore, we further analysed the incidence of hypertension in patients with RCC compared with those with non-RCC malignancies, regardless of treatment asssignment. For patients with RCC, the overall incidence of all-grade and high-grade hypertension was 23·6% 120
(14·3–36·5) and 6·5% (1·8–21·1) respectively (figure 2), as determined by the random-effects model; for those patients with non-RCC malignancies, the overall incidences of all-grade and high-grade hypertension were 23·0% (16·0–31·9) and 5·3% (2·7–10·0) respectively, as analysed by the fixed-effects model (figure 3). No significant difference was detected in the incidence of sorafenib-associated all-grade hypertension (RR 1·03 [95% CI 0·73–1·45], p=0·89) and high-grade hypertension (RR 1·23 [95% CI 0·76–1·99], p=0·40) between patients with RCC and those with a non-RCC malignancy. A meta-analysis of the RR for hypertension with sorafenib in patients with metastatic RCC compared with the control group was done for the two randomised controlled trials.4,11 One trial used a placebo as the control,4 and the other used interferon.11 In both trials, the incidence of hypertension was low in the control group (1·8% and 6·7%, respectively). The overall RR was 6·11 for sorafenib versus control for all-grade hypertension (95% CI 2·44–15·32, p<0·0001) as calculated using the random-effects model (figure 4). Thus, sorafenib was associated with a significantly increased risk of hypertension in patients with RCC compared with patients with RCC not assigned sorafenib. To understand the risk of hypertension with sorafenib compared with other anti-angiogenesis drugs, we assessed the incidence of hypertension with bevacizumab, a humanised monoclonal antibody against vascular endothelial growth factor (VEGF). Bevacizumab is associated with a significantly increased risk of hypertension. We have previously shown that the RR for hypertension with bevacizumab is 3·0 (95% CI 2·2–4·2; p<0·001) for low doses (≤7·5 mg/kg per dose) and 7·5 (4·2–13·4; p<0·001) for high doses (10 or 15 mg/kg per dose).19 In that study, we did a meta-analysis of the clinical studies that were included in our previous study, and calculated that the overall incidence of hypertension associated with high-dose bevacizumab is 25·4% (95% CI 21·3–30·1) by use of the fixed-effects model (figure 5).
Discussion This study shows the overall risk of hypertension associated with sorafenib in patients with cancer. We noted a high overall incidence of all-grade hypertension http://oncology.thelancet.com Vol 9 February 2008
Articles
Molecular target
Incidence of hypertension (95% CI)
Relative risk of hypertension (95% CI)
Ref
Sorafenib
VEGFR-2 and VEGFR-3*; Raf; PDGFR; C-KIT; FLT-3; RET
23·4% (16·0–32·9)
6·1 (2·4–15·3)
This study
Sunitinib
VEGFR-1 and VEGFR-2*; PDGFR; C-KIT; FLT-3; RET
22·5% (19·5–25·9)
3·9 (2·6–5·9)
Zhu X, unpublished data
Bevacizumab
VEGF*
25·4% (21·3–30·1)†
7·5 (4·2–13·4)‡
19
AG013736
VEGFR-1, VEGFR-2, and VEGFR-3*; PDGFR; C-KIT
57·7%
NA
24
VEGF Trap
VEGF*
31·6%
NA
25
VEGFR=vascular endothelial growth factor receptor. PDGFR=platelet-derived growth-factor receptor. *Targets directly involved in angiogenesis. †Incidence is calculated from patients receiving high-dose bevacizumab (10 or 15 mg/kg per dose) by a meta-analysis using the fixed-effects model. ‡Relative risk is derived from patients receiving highdose bevacizumab (10 or 15 mg/kg per dose) by use of fixed-effects model. NA=not available.
Table 2: Risk of hypertension with angiogenesis inhibitors
(23·4% [95% CI 16·0–32·9]) with sorafenib and an overall incidence of high-grade hypertension of 5·7% (2·5–12·6). Adequate and aggressive management of moderate hypertension can be essential for many patients, because hypertension is an independent risk factor of renal and cardiovascular events.20–23 As seen in a phase III trial for patients with metastatic RCC,4 cardiac ischaemia or infarction were significantly more frequent in the sorafenib group (n=12 [3%]) than in the placebo group (n=2 [0·4%]; p=0·01). Furthermore, poorly controlled sorafenib-associated hypertension can lead to reversible posterior leucoencephalopathy syndrome.12 Finally, severe hypertension has led to the permanent discontinuation of sorafenib.4 Hypertension associated with angiogenesis inhibitors is an emerging issue in patients receiving these therapeutic drugs. In addition to sorafenib, several other angiogenesis inhibitors, such as bevacizumab, sunitinib, AG013736, and VEGF trap have been associated with the development of hypertension (table 2). In a meta-analysis, sunitinib, another multi-tyrosine-kinase inhibitor, was shown to be associated with hypertension with an incidence of 22·5% (95% CI 19·5–25·9) and an RR of 3·89 versus control (Zhu X, unpublished data). Our study shows that sorafenib is associated with a significantly increased risk of hypertension with an incidence of 23·4% (16·0–32·9) and an RR of 6·11 (95% CI 2·44–15·32; p<0·0001); high-dose bevacizumab (10 or 15 mg/kg per dose) is associated with an incidence of hypertension of 25·4% (21·3–30·1; figure 5) and an RR of 7·5 (4·2–13·4) compared with controls.19 When considering these combined findings, the risks of hypertension associated with these angiogenesis inhibitors seem considerable. The association of sorafenib with hypertension might be directly related to its inhibitory effect on the VEGF receptor. Possible mechanisms include: impaired angiogenesis leading to a decrease in the density of microvessels (a process known as rarefaction), endothelial dysfunction associated with a decrease in nitric-oxide production and an increase in oxidative stress; and changes in neurohormonal factors or the renin-angiotensin-aldosterone system. In a small study by Veronese and colleagues26 in http://oncology.thelancet.com Vol 9 February 2008
which 20 patients received 400 mg of sorafenib twice a day, there were no significant changes in humoral factors, including serum total catecholamines, epinephrine, norepinephrine, endothelin I, urotensin II, renin, and aldosterone after 3 weeks of treatment, even though an evident increase in systolic blood pressure (≥10 mm Hg) was noted in 75% of patients. Therefore, the role of these humoral factors in the development of sorafenib-associated hypertension might be limited, which implies that nonhumoral factors, such as impaired angiogenesis or endothelial dysfunction, play an important part. Patients with RCC might have a greater risk of hypertension than those with non-RCC malignancies as a result of previous nephrectomy and renal dysfunction.27–29 However, our study showed that the incidence of hypertension is not significantly different in patients with RCC compared with those with a non-RCC malignancy when they are treated with sorafenib (all grade: RR 1·03 [95% CI 0·73–1·45]; high-grade: RR 1·23 [0·76–1·99]). A possible explanation for this finding is that the increase in blood pressure and hypertension induced by sorafenib is so prominent that the risk associated with RCC is not evident in this setting. This notion is supported by the high incidence of all-grade hypertension with sorafenib (23·4% [95% CI 16·0–32·9) noted in this study and the high incidence of increased blood pressure in about 75% of patients receiving sorafenib in the study by Veronese and colleagues.26 Another explanation could be that the combined effect of VEGF from normal tissue and tumour tissue on bloodpressure regulation is comparable between patients with RCC and non-RCC malignancies, even though RCC can be associated with a higher VEGF secretion as a result of its biology;30 thus, the effect of sorafenib on hypertension might be similar. Additionally, sorafenib is mainly metabolised by the liver,31 and renal dysfunction associated with RCC might not affect the concentration of sorafenib in the blood in a substantial way. The management of sorafenib-associated hypertension is still under debate. According to the manufacturer package insert for sorafenib, hypertension can occur early in the course of treatment and, therefore, blood pressure should be monitored weekly during the first 121
Articles
6 weeks of treatment; thereafter hypertension should be monitored and treated in accordance with standard medical practice. In cases of severe or persistent hypertension despite the initiation of antihypertensive treatment, temporary or permanent discontinuation of sorafenib should be considered.31 In most patients, hypertension can be controlled with standard antihypertensive medications.3 However, the biological effect of these antihypertensive medications on angiogenesis and its implications should be considered, and need to be fully understood. Both enalapril (an angiotensin-convertingenzyme [ACE] inhibitor) and candesartan (an angiotensin receptor blocker) can inhibit myocardial angiogenesis induced by VEGF.32 Additionally, candesartan has been shown to have an antiangiogenesis effect in a xenograft model of bladder cancer.33 However, nifedipine (a calciumchannel blocker) has been shown to induce VEGF secretion.34 Thus, the possibility exists that some antihypertensive medications are more effective in treating anti-VEGF-associated hypertension and have less toxic effects when used in conjunction with sorafenib. Additionally, certain antihypertensive medications can interfere with the antineoplastic activity of sorafenib by modulating VEGF expression and reducing its antiangiogenic effect. Drug–drug interactions are also important issues. Sorafenib is metabolised mainly in the liver, where it undergoes an oxidative reaction mediated by the cytochrome p450 enzyme system—mainly by CYP3A4.31 Therefore, the metabolism of sorafenib can be potentially changed by certain antihypertensive drugs that are inhibitors of CYP3A4. For example, nondihydropyridine calcium-channel blockers, such as verapamil and diltiazem are CYP3A4 inhibitors, and can substantially increase the plasma concentration of sorafenib. However, ketoconazole, a potent CYP3A4 inhibitor, does not change the mean area under the curve of a single oral 50 mg dose of sorafenib in healthy volunteers when administered once a day at 400 mg for 7 days.31 Therefore, predicting the effect of CYP3A4 inhibitors on the metabolism of sorafenib is difficult. In the absence of available data from clinical studies, dihydropyridines, such as amlodipine and nifedipine, are the preferred class of calcium-channel blockers. Nondihydropyridine calcium-channel blockers should be used cautiously in conjunction with sorafenib. Alternatively, ACE inhibitors or angiotensin-receptor blockers are reasonable choices with the additional benefit of improving endothelial function and microcirculatory density.35 Other antihypertensive medications, such as diuretics, alpha blockers, and beta blockers, which have been used in the setting of bevacizumab-associated hypertension,36 can also be considered for use in this setting. Because sorafenibinduced hypertension might be related to reduced nitric-oxide production secondary to the anti-VGFR effect, the use of phosphodiesterase inhibitors or nitrate 122
derivatives to increase nitric oxide levels has been suggested and might be a treatment option depending on the findings of further clinical studies.37 This study has several limitations. As with any metaanalysis, the findings described here are affected by the limitations of the individual clinical trials that are included in the analysis. These trials might have underestimated the incidence of sorafenib-associated hypertension because of the imperfection of the CTCAE version 2 or 3 for recording adverse events.38 In both these versions, patients are considered hypertensive only if diastolic pressure is increased by more than 20 mm Hg or blood pressure is greater than 150/100 mm Hg. In our study, this grading criteria would have missed many patients with hypertension according to the standard criteria for the diagnosis of hypertension (140/90 mm Hg). Additionally, the grading criteria do not clearly differentiate between grade 2 and 3 hypertension, and our findings of high-grade hypertension could be affected by the overlap between these two grades. Furthermore, the prevalence of baseline hypertension was not described in the included trials, which might have led to an overestimation of the incidence of sorafenib-associated hypertension. However, the presence of baseline hypertension would probably have been low at about 2–4%, as estimated from patients who received placebos in the control groups. We have kept the likelihood of bias to a minimum by calculating the RR by use of randomised controlled clinical trials for the direct comparison of hypertension with and without sorafenib. Additionally, there seems to be no significant variation in the incidence of baseline hypertension in patients with metastatic cancers, as reflected in the control groups of many clinical trials.19 The incidence of hypertension in the control groups described here was 1·8%4 and 6·7%,11 consistent with that of the control patients in phase III bevacizumab trials (range 0–8·3%).19 Additionally, the patients in this study had metastatic cancers and were mainly selected from clinical trials. Therefore, our findings were noted mainly in academic centres and major research institutes, and might not entirely apply to patients with cancer who are treated in the community. However, this analysis also included many patients from the expanded-access programme,10 which did contain a substantial number of patients from the community setting, although the recording of hypertension and other data might be inadequate in this setting. Finally, even though we have shown that there is no significant difference in the incidence of sorafenib-associated hypertension between patients with RCC and non-RCC malignancies (RR 1·03 [95% CI 0·73–1·45], p=0·89), this finding might be limited by the small sample size of patients with a non-RCC malignancy who were assigned sorafenib (n=111 for all-grade hypertension; n=315 for high-grade hypertension analysis). This study has shown that sorafenib is associated with a significant risk of developing hypertension. Early detection and effective management of hypertension http://oncology.thelancet.com Vol 9 February 2008
Articles
might allow for safer use of this drug. The hypertensive and cardiovascular side-effects of sorafenib need thorough postmarketing surveillance and reporting, and future studies will be needed to identify the mechanism and appropriate treatment of sorafenibinduced hypertension. Contributors SW and XZ conceived the study, and SW was responsible for the data search. SW and JJC produced the figures and all authors contributed to the writing of the manuscript. Conflicts of interest SW is supported by the Research Foundation of the State University of New York at Stony Brook (NY, USA), and is a consultant for Onyx Pharmaceuticals and a speaker for Pfizer Inc. Acknowledgments The authors would like to thank Stefan Madajewicz for critically reading the manuscript. References 1 Wilhelm SM, Carter C, Tang L, et al. BAY 43-9006 exhibits broad spectrum oral antitumor activity and targets the RAF/MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis. Cancer Res 2004; 64: 7099–109. 2 Chang YS, Adnane J, Trail PA, et al. Sorafenib (BAY 43-9006) inhibits tumor growth and vascularization and induces tumor apoptosis and hypoxia in RCC xenograft models. Cancer Chemother Pharmacol 2007; 59: 561–74. 3 Ratain MJ, Eisen T, Stadler WM, et al. Phase II placebo-controlled randomized discontinuation trial of sorafenib in patients with metastatic renal cell carcinoma. J Clin Oncol 2006; 24: 2505–12. 4 Escudier B, Eisen T, Stadler WM, et al. Sorafenib in advanced clearcell renal-cell carcinoma. N Engl J Med 2007; 356: 125–34. 5 Abou-Alfa GK, Schwartz L, Ricci S, et al. Phase II study of sorafenib in patients with advanced hepatocellular carcinoma. J Clin Oncol 2006; 24: 4293–300. 6 Llovet J, Mazzaferro V, Hilgard P, et al. Sorafenib improves survival in advanced hepatocellular carcinoma (HCC): results of a phase III randomized, placebo-controlled trial (SHARP trial). Proc Am Soc Clin Oncol 2007; 25: (abstr LBA1). 7 Eisen T, Ahmad T, Flaherty KT, et al. Sorafenib in advanced melanoma: a phase II randomised discontinuation trial analysis. Br J Cancer 2006; 95: 581–86. 8 Gatzemeier U, Fosella F, Simantov R, et al. Phase II trial of singleagent sorafenib in patients with advanced non-small cell lung carcinoma. Proc Am Soc Clin Oncol 2006; 24: (abstr 7002). 9 Wu S, Posadas E, Scripture C, et al. BAY43-9006 (sorafenib) can lead to improvement of bone lesions in metastatic androgenindependent prostate cancer despite rises in serum PSA levels. 2006 Prostate Cancer Symposium. San Francisco, CA, USA. Feb 24–26, 2006: (abstr 259). 10 Figlin R, McDermott D, Gabrail N, et al. The Advanced Renal Cell Carcinoma Sorafenib (ARCCS) expanded access trial in North America: safety and efficacy. Proc Am Soc Clin Oncol 2007; 25: (abstr 5011). 11 Szczylik C. Demkow T, Staehler M, et al. Randomized phase II trial of first-line treatment with sorafenib versus interferon in patients with advanced renal cell carcinoma: final results. Proc Am Soc Clin Oncol 2007; 25: (abstr 5025). 12 Govindarajan R, Adusumilli J, Baxter DL, El-Khoueiry A, Harik SI. Reversible posterior leukoencephalopathy syndrome induced by RAF kinase inhibitor BAY 43-9006. J Clin Oncol 2006; 24: e48. 13 US National Cancer Institute. Common Toxicity Criteria, Version 2.0. http://ctep.cancer.gov/reporting/ctc_archive.html (accessed Jan 7, 2007). 14 Lewis S, Clarke M. Forest plots: trying to see the wood and the trees. BMJ 2007; 322: 1479–80. 15 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–88. 16 Lau J, Ioannidis JP, Schmid CH. Quantitative synthesis in systematic reviews. Ann Intern Med 1997; 127: 820–26.
http://oncology.thelancet.com Vol 9 February 2008
17
18
19
20 21
22
23 24
25
26
27
28 29
30
31 32
33
34
35
36
37
38
D’Adamo DR, Keohan M, Schuetze S, et al. Clinical results of a phase II study of sorafenib in patients (pts) with non-GIST sarcomas (CTEP study #7060). Proc Am Soc Clin Oncol 2007; 25: (abstr 10001). Hobday TJ, Holen K, Picus J, et al. A phase II trial of sorafenib in patients with metastatic neuroendocrine tumors (NET): a phase II consortium (P2C) study. Proc Am Soc Clin Oncol 2007; 25: (abstr 4504). Zhu X, Wu S, Dahut WL, Parikh CR. Risks of proteinuria and hypertension with bevacizumab, an antibody against vascular endothelial growth factor: systematic review and meta-analysis. Am J Kidney Dis 2007; 49: 186–93. Kaplan NM, Opie LH. Controversies in hypertension. Lancet 2006; 367: 168–76. Peterson JC, Adler S, Burkart JM, et al. Blood pressure control, proteinuria, and the progression of renal disease. The Modification of Diet in Renal Disease Study. Ann Intern Med 1995; 123: 754–62. Wachtell K, Ibsen H, Olsen MH, et al. Albuminuria and cardiovascular risk in hypertensive patients with left ventricular hypertrophy: the LIFE study. Ann Intern Med 2003; 139: 901–06. Weisstuch JM, Dworkin LD. Does essential hypertension cause endstage renal disease? Kidney Int Suppl 1992; 36: S33–37. Rixe O, Bukowski RM, Michaelson MD, et al. Axitinib treatment in patients with cytokine-refractory metastatic renal-cell cancer: a phase II study. Lancet Oncol 2007; 8: 975–84. Nguyen QD, Shah SM, Hafiz G, et al. A phase I trial of an IVadministered vascular endothelial growth factor trap for treatment in patients with choroidal neovascularization due to age-related macular degeneration. Ophthalmology 2006; 113: 1522e1–e14. Veronese ML, Mosenkis A, Flaherty KT, et al. Mechanisms of hypertension associated with BAY 43-9006. J Clin Oncol 2006; 24: 1363–69. McKiernan J, Simmons R, Katz J, Russo P. Natural history of chronic renal insufficiency after partial and radical nephrectomy. Urology 2002; 59: 816–20. Melman A, Grim CE, Weinberger MH. Increased incidence of renal cell carcinoma with hypertension. J Urol 1977; 118: 531–22. Shirasaki Y, Tsushima T, Nasu Y, Kumon H. Long-term consequence of renal function following nephrectomy for renal cell cancer. Int J Urol 2004; 11: 704–08. Schips L, Dalpiaz O, Lipsky K, et al. Serum levels of vascular endothelial growth factor (VEGF) and endostatin in renal cell carcinoma patients compared to a control group. Eur Urol 2007; 51: 168–73. Bayer. NEXAVAR Prescribing Information. http://www.univgraph. com/bayer/inserts/nexavar.pdf (accessed Jan 7, 2008). Siddiqui AJ, Mansson-Broberg A, Gustafsson T, et al. Antagonism of the renin-angiotensin system can counteract cardiac angiogenic vascular endothelial growth factor gene therapy and myocardial angiogenesis in the normal heart. Am J Hypertens 2005; 18: 1347–52. Kosugi M, Miyajima A, Kikuchi E, Horiguchi Y, Murai M. Angiotensin II type 1 receptor antagonist candesartan as an angiogenic inhibitor in a xenograft model of bladder cancer. Clin Cancer Res 2006; 12: 2888–93. Miura S, Fujino M, Matsuo Y, Tanigawa H, Saku K. Nifedipineinduced vascular endothelial growth factor secretion from coronary smooth muscle cells promotes endothelial tube formation via the kinase insert domain-containing receptor/fetal liver kinase-1/NO pathway. Hypertens Res 2005; 28: 147–53. Agabiti-Rosei E. [Structural and functional changes of the microcirculation in hypertension: influence of pharmacological therapy]. Drugs 2003; 63 (Spec No 1): 19–29. Mares JE, Worah S, Mathew SV, et al. Increased rates of hypertension (HTN) among patients with advanced carcinoid treated with bevacizumab. Proc Am Soc Clin Oncol 2005: 23: (abstr 4087). Izzedine H, Rixe O, Billemont B, Baumelou A, Deray G. Angiogenesis inhibitor therapies: focus on kidney toxicity and hypertension. Am J Kidney Dis 2007; 50: 203–18. National Cancer Institute. CTC v 2.0 and Common Terminology Criteria for Adverse Events Criteria V3.0 (CTCAE). http://ctep. cancer.gov/reporting/ctc.html (accessed Jan 7, 2007).
123
Incidence and risk of hypertension with sorafenib in patients with cancer: a systematic review and meta-analysis Shenhong Wu, John J Chen, Andrzej Kudelka, Janice Lu, Xiaolei Zhu
Summary Background Sorafenib is used in patients with advanced renal-cell carcinoma (RCC) or hepatocellular cancer, and its application in other types of cancers is also undergoing extensive clinical assessment. Hypertension is one of the major side-effects of this drug, and reported incidences vary substantially between clinical trials. We did a systematic review and meta-analysis of published clinical trials to establish the incidence of hypertension associated with sorafenib. The aim of this study is to gain a better understanding of the overall risk of hypertension in patients with cancer who receive sorafenib. Methods Databases, including Medline (July, 1966, to July, 2007), and Web of Science, and abstracts presented at the American Society of Clinical Oncology annual meetings from 2004 to 2007 were searched to identify relevant studies. Eligible studies were prospective clinical trials of patients with cancer assigned single-drug sorafenib at 400 mg twice daily with data on hypertension available. Incidence and relative risk (RR) of hypertension were calculated using a random-effects or fixed-effects model, depending on the heterogeneity of the included studies. Findings Nine studies published between January, 2006, and July, 2007, which included a total of 4599 patients with RCC or other solid tumours, were selected from 223 articles screened for analysis. For patients assigned sorafenib, the overall incidence of all-grade and high-grade (ie, grade 3 or 4) hypertension were 23·4% (95% CI 16·0–32·9%) and 5·7% (2·5–12·6%), respectively. No significant difference was noted between patients with RCC or a non-RCC malignancy (all grade: RR 1·03 [95% CI 0·73–1·45], p=0·89; high-grade: RR 1·23 [0·76–1·99], p=0·40) who were assigned sorafenib. Sorafenib was associated with a significantly increased risk of all-grade hypertension in patients with cancer with an RR of 6·11 (2·44–15·32], p<0·001) compared with controls. Interpretation Patients with cancer assigned sorafenib have a significant risk of developing hypertension. Appropriate monitoring and treatment is strongly recommended to prevent cardiovascular complications.
Introduction Sorafenib is a multikinase inhibitor with a broad spectrum of antitumour activity on cancer-cell proliferation and angiogenesis.1 It targets: Raf kinase; vascular endothelial growth factor receptors (VEGFR) 1, 2, and 3; platelet-derived growth factor receptor β (PDGFR β); FMS-like tyrosine kinase 3 (FLT-3); c-KIT; and RETreceptor tyrosine kinase.1 In animals, the antitumour effect of sorafenib is mainly mediated by its inhibition of angiogenesis.2 The clinical benefit of sorafenib was initially seen in a phase II randomised discontinuation trial, which showed a lengthened progression-free survival with sorafenib compared with placebo in patients with metastatic renal-cell carcinoma (RCC).3 Subsequently, a phase III randomised controlled trial (Treatment Approaches in Renal cancer Global Evaluation Trial [TARGET]) confirmed this finding.4 Additionally, sorafenib has clinical activity in advanced hepatocellular carcinoma.5,6 The efficacy of sorafenib in non-small-cell lung cancer, prostate cancer, and melanoma is currently undergoing assessment in a randomised phase II trial7 and in single-arm phase II clinical studies.8,9 As with many therapeutic drugs, sorafenib is associated with substantial side-effects. Diarrhoea, rash, fatigue, and hand and foot skin reactions were the most common adverse events associated with sorafenib in the TARGET http://oncology.thelancet.com Vol 9 February 2008
Lancet Oncol 2008; 9: 117–23 Published Online January 22, 2008 DOI:10.1016/S14702045(08)70003-2 See Reflection and Reaction page 86 Division of Medical Oncology, Department of Medicine, State University of New York, Stony Brook, NY, USA (S Wu MD, A Kudelka MD, J Lu MD); Department of Preventive Medicine, State University of New York, Stony Brook, NY, USA (J J Chen PhD); and Kidney Specialists of Long Island, Port Jefferson Station, NY, USA (X Zhu MD) Correspondence to: Dr Shenhong Wu, Division of Medical Oncology, Department of Medicine, State University of New York, Stony Brook, NY 11794-9447, USA shenhong.wu@stonybrook. edu
trial.4 Additionally, hypertension is a major side-effect that has been noted in trials, with its incidence ranging from 16·0–42·6%.3,4,7–11 The recognition and management of hypertension is an important issue in patients treated with sorafenib because poorly controlled hypertension can lead to serious cardiovascular events. In the TARGET trial, cardiac ischaemia and infarction were more common in patients assigned sorafenib than in those assigned placebo.4 Additionally, the use of sorafenib might be associated with reversible posterior leucoencephalopathy syndrome, a clinical event characterised by headache, seizures, impaired vision, and acute hypertension.12 Because of the limited number of patients in trials, the overall risk of hypertension with sorafenib is unclear. Therefore, we did a systematic review of the published work and a meta-analysis to assess this issue.
Methods Data source We did an independent review of citations from Medline (July, 1966, to July, 2007). Our search strategy included the terms “sorafenib”, “BAY 43-9006”, “cancer”, and “hypertension” and was restricted to human studies published in English. Additionally, we manually searched all abstracts containing the term “sorafenib” that were presented at the American Society of Clinical Oncology 117
Articles
Trial
Trial Design
Patients Sample size Median age, enrolled, n years
Underlying malignancy
Ratain et al3
Randomised phase II (placebo vs sorafenib)
202
202
58
RCC
Escudier et al4
Randomised phase III (placebo vs sorafenib)
903
451
58
RCC
Eisen et al7
Randomised phase II (placebo vs sorafenib)
502
37
53
Melanoma
Gatzemeier et al8
Single-arm phase II
52
52
62
NSCLC
Wu et al9
Single-arm phase II
22
22
64
Prostate
Figlin et al10
Expanded-access programme
2502
2502
63
RCC
Szczylik et al11
Randomised phase II (interferon vs sorafenib)
189
97
62
RCC
D’Adamo et al17
Single-arm phase II
134
111
55
Non-GIST sarcoma
Hobday et al18
Single-arm phase II
93
93
59
Neuroendocrine tumours
RCC=renal-cell carcinoma. NSCLC=non-small-cell lung cancer. GIST=gastrointestinal stromal tumour.
Study
Number of events Sample size 4
Escudier (2007) Eisen (2006)7 Ratain (2006)3 Szczylik (2007)11 Gatzemeier (2006)8 Wu (2006)9 Figlin (2007)10 Overall
76 6 86 24 12 7 401 612
Test of heterogeneity: Q=85·873, p<0·001, I²=93·013
B
Study Escudier (2007)4 Eisen (2006)7 Ratain (2006)6 Szczylik (2007)11 Gatzemeier (2006)8 Wu (2006)9 Figlin (2007)10 D'Adamo (2007)17 Hobday (2007)18 Overall
Incidence (95% CI)
451 37 202 97 52 22 2502 3363
0·169 (0·137–0·206) 0·162 (0·075–0·317) 0·426 (0·359–0·495) 0·247 (0·172–0·343) 0·231 (0·136–0·364) 0·318 (0·160–0·534) 0·160 (0·146–0·175) 0·234 (0·160–0·329)
0
0·50
1·00
Number of events Sample size 16 5 62 2 2 1 125 3 4 220
Test of heterogeneity: Q=167·076, p<0·001, I²=95·212
Incidence (95% CI)
451 37 202 97 52 22 2502 111 93 3567
0·035 (0·022–0·057) 0·135 (0·057–0·286) 0·307 (0·247–0·374) 0·021 (0·005–0·079) 0·038 (0·010–0·141) 0·045 (0·006–0·261) 0·050 (0·042–0·059) 0·027 (0·009–0·080) 0·043 (0·016–0·109) 0·057 (0·025–0·126) 0
0·50
1·00
Incidence (1=100%)
Figure 1: Annotated forest plot for meta-analysis of incidence of hypertension in patients with cancer who were assigned sorafenib Summary incidences of all-grade (A) and high-grade (B) hypertension are calculated using a random-effects model. Size of squares is directly proportional to amount of information available.
(ASCO) annual meetings and ASCO Prostate Cancer Symposiums from 2004 to 2007. An independent search by use of the citation database Web of Science was also done to ensure that no studies were missed. Details on trial design, patient characteristics, treatment information, findings, and follow-up were used from these 118
Study selection Sorafenib has been approved for use in patients with advanced RCC and hepatocellular carcinoma as a single drug at the starting dose of 400 mg twice daily. To ensure clinical significance, we assessed the risk of hypertension with sorafenib at this dose. Therefore, phase I trials were excluded from the analysis due to variations in dosage. We included phase II and III clinical trials, and expandedaccess programmes, in which sorafenib as a single drug was given at the standard dose. Trials that met the following criteria were chosen for analysis: prospective clinical trials in patients with cancer; patients assigned to treatment with sorafenib as a single drug at a starting dose of 400 mg twice a day; data available for the events or incidences of hypertension.
Clinical endpoints
Table 1: Characteristics of clinical trials and patients included in the meta-analysis
A
trials. The entire selection process (ie, search, selection, and exclusion of studies) was done by SW.
Clinical endpoints were selected from the safety profile of each trial. Hypertension was recorded according to versions 2 or 3 of the Common Terminology Criteria for Adverse Events (CTCAE),13 each of which are the same regarding the grading of hypertension as follows: grade 1—asymptomatic, transient (<24 h) increase in blood pressure by more than 20 mm Hg (diastolic) or to more than 150/100 mm Hg if previously within normal limit, intervention not indicated; grade 2—recurrent or persistent (>24 h) or symptomatic increase by more than 20 mm Hg (diastolic) or to more than 150/100 mm Hg if previously within normal limit, monotherapy might be indicated; grade 3—more than one drug needed for treatment or more intensive treatment than used previously; grade 4—life-threatening consequences (eg, hypertensive crisis). We included the incidence of hypertension of grade I or above in our analysis.
Statistical analysis All statistical analyses were done using version 2 of the Comprehensive MetaAnalysis programme (Biostat, Englewood, NJ, USA). The numbers of patients with hypertension, both all-grades and high-grades (grade 3 and 4), were summarised from the extracted data on all patients assigned sorafenib at 400 mg twice daily in the clinical trials included for analysis. For each trial, the proportion of patients with hypertension was calculated and the 95% CI was derived. For studies with a control group, the relative risk (RR) of hypertension in patients assigned to sorafenib was also calculated, and was compared only with those assigned to a control treatment in the same trial. For the meta-analysis, both the fixed-effects model (weighted with inverse variance) and the random-effects model were considered.14,15 For each meta-analysis, the Cochrane’s Q statistic was first calculated to assess the heterogeneity of the included trials. For p values less than 0·1, the assumption of homogeneity was deemed http://oncology.thelancet.com Vol 9 February 2008
Articles
invalid,16 and the random-effects model was used only after substantial efforts were made to explore the possible reasons for the heterogeneity. Otherwise, data were assessed using both the fixed-effects model and the random-effects models. A two-tailed p value of less than 0·05 was deemed statistically significant. All the statistical analyses were done by SW and JJC.
A
http://oncology.thelancet.com Vol 9 February 2008
76 86 24 401 587
Escudier (2007) Ratain (2006)3 Szczylik (2000)11 Figlin (2007)16 Overall
B
Study
Incidence (95% CI)
451 202 97 2502 3252
Test of heterogeneity: Q=82·596, p<0·001, I²=96·368
0·169 (0·137–0·206) 0·426 (0·359–0·495) 0·247 (0·172–0·343) 0·160 (0·146–0·175) 0·236 (0·143–0·365) 0
0·50
1·00
Number of events Sample size 4
Escudier (2007) Ratain (2006)3 Szczylik (2007)11 Figlin (2007)10 Overall
16 62 2 125 205
Incidence (95% CI)
451 202 97 2502 3252
0·035 (0·022–0·057) 0·307 (0·247–0·374) 0·021 (0·005–0·343) 0·050 (0·042–0·059) 0·065 (0·018–0·211) 0
Test of heterogeneity: Q=159·079, p<0·001, I²=98·114
Results Our search yielded a total of 223 articles on sorafenib from the published work. After reviewing each publication, we identified 16 original studies including randomised controlled trials, single-arm phase II studies, and case reports. From these 16 articles, three clinical trials matched our inclusion criteria.3,4,7 Of note, the phase II and III trials of sorafenib in patients with hepatocellular carcinoma were excluded from the analysis because of no available data regarding hypertension.5,6 From the abstracts published during recent ASCO annual meetings and ASCO Prostate Cancer Symposiums (2004–2007), we identified 68 abstracts that were related to sorafenib. After reviewing each abstract, we included six additional trials in our meta-analysis (table 1).8–11,17,18 Of the nine trials included, three were blinded3,4,11 and the rest were openlabelled. Three trials were sponsored by Bayer Pharmaceuticals (West Haven, CT, USA) and Onyx Pharmaceuticals (Emeryville, CA, USA).3,4,11 One trial in prostate cancer was supported by the US National Cancer Institute.9 Support for the other trials was not specified. A total of 4599 patients were available for analysis, with 3567 patients assigned treatment with sorafenib as a single drug. Baseline characteristics of these patients from the nine clinical trials are listed in table 1. Hypertension was not mentioned as a pre-existing condition in any of these trials. Baseline Eastern Cooperative Oncology Group performance status for most patients was between 0 and 1. Underlying metastatic malignancies included RCC, melanoma, non-small-cell lung cancer, prostate cancer, non-gastrointestinalstromal-tumour (GIST) sarcoma, and neuroendocrine tumours. For the randomised discontinuation trial by Eisen and co-workers,7 a total of 502 patients with various tumour types were enrolled, but the publication only focused on the 37 treated patients with advanced melanoma. Treatment options were randomly assigned in four clinical studies including two randomised controlled trials and two randomised discontinuation trials.3,4,7,11
Number of events Sample size 4
Role of the funding source This study was partially funded by the Research Foundation of the State University of New York, Stony Brook, NY, USA. This foundation played no role in the design, collection, analysis, interpretation, and writing of the study. The corresponding author had access to all the raw data and had the final responsibility to submit the manuscript for publication.
Study
0·50
1·00
Incidence (1=100%)
Figure 2: Annotated forest plot for meta-analysis of incidence of sorafenib-associated hypertension in patients with renal-cell carcinoma Summary incidences of all-grade (A) and high-grade (B) hypertension are calculated using a random-effects model. Size of squares is directly proportional to amount of information available.
A
Study
Number of events Sample size 7
Eisen (2006)
Gatzemeier (2006)8 Wu (2006)9 Overall
6
37
0·162 (0·075–0·317)
12
52
0·231 (0·136–0·364)
7
22
0·318 (0·160–0·534)
25
111
0·230 (0·160–0·319)
Test of heterogeneity: Q=1·897, p=0·387, I²<0·001
B
Study
Incidence (95% CI)
0
0·50
1·00
Number of events Sample size
Eisen (2006)7 Gatzemeier (2006)8
5 2
Wu (2006)9 D'Adamo (2007)17 Hobday (2007)18 Overall Test of heterogeneity: Q=6·376, p=0·173, I²=37·263
Incidence (95% CI)
37 52
0·135 (0·057–0·286) 0·038 (0·010–0·141)
1
22
0·045 (0·006–0·261)
3 4
111 93
0·027 (0·009–0·080) 0·043 (0·016–0·109)
15
315
0·053 (0·027–0·100) 0 0·50 1·00 Incidence (1=100%)
Figure 3: Annotated forest plot for meta-analysis of incidence of sorafenib-associated hypertension in patients with non-renal-cell-carcinoma malignancies Summary incidences of all-grade (A) and high-grade (B) hypertension are calculated using the fixed-effects model. Size of squares is directly proportional to amount of information available.
Data for all-grade hypertension was available for analysis from a total of 3363 patients enrolled in seven trials who had various solid tumours and were assigned sorafenib. The incidence of all-grade hypertension ranged between 16·0% and 42·6%, with the lowest incidence noted in the expanded access programme,10 and the highest in the phase II placebo-controlled randomised discontinuation trial in patients with metastatic RCC3 Meta-analysis showed the heterogeneity of the included studies (Q=85·9, p=0·0000). We explored the potential source of heterogeneity by analysing blinded and open-label trials separately. Our analysis showed that heterogeneity exists 119
Articles
Study
Number of events/sample size Sorafenib Control
Escudier (2007)4 Szczylik (2007)18 Overall
76/451 24/97
8/451 6/90
100/548
14/541
Risk ratio and 95% CI
0·01
Test of heterogeneity: Q=2·755, p=0·097, I²=63·705
0·1 Control
1
10 Sorafenib
Risk ratio (95% CI)
Z-value
p-value
9·5000 (4·6393–19·4533) 3·7113 (1·5905–8·6603)
6·1564 3·0333
0·0000 0·0024
6·1082 (2·4357–15·3183)
3·8577
0·0001
100
Figure 4: Relative risk of sorafenib-associated hypertension versus control from the two randomised controlled trials of patients with renal-cell carcinoma Summary relative risk (RR) was calculated using the random-effects model. Size of squares is directly proportional to amount of information available.
Study
Number of events Sample size
Johnson (2004) Kabbinavar (2003) Kindler (2005) Miller (2005) Yang (2003) Overall Test of heterogeneity: Q=3·933, p=0·415, I²<0·001
6 9 13 54 14 96
Incidence (95% CI)
34 32 47 229 39 381
0·176 (0·081–0·341) 0·281 (0·153–0·458) 0·277 (0·168–0·420) 0·236 (0·185–0·295) 0·359 (0·225–0·519) 0·254 (0·213–0·301) 0
0·50
1·00
Incidence (1=100%)
Figure 5: Annotated forest plot for meta-analysis of incidence of high-dose bevacizumab-associated hypertension in patients with cancer Studies included in the analysis were derived from our previous study for high-dose bevacizumab.19 Summary incidence of all-grade hypertension is calculated using the fixed-effects model. Size of squares is directly proportional to amount of information available.
in the three blinded studies,3,4,7 but not in the four openlabel studies.8–11 We were unable to find a compelling reason to exclude these blinded studies from the final analysis. As analysed by the random-effects model, the overall incidence of all-grade hypertension was 23·4% (95% CI 16·0–32·9) in patients who were assigned sorafenib (figure 1A). High-grade hypertension needs to be treated with more than one drug or with a more intense treatment than used previously (grade 3), or is associated with lifethreatening consequences (grade 4). It can result in substantial morbidity, dose reduction, or discontinuation of sorafenib. Data for high-grade hypertension was available for analysis from the total 3567 patients enrolled in the nine trials. The incidence of grade 3 or 4 hypertension ranged from 2·1% to 30·7%, with the lowest incidence noted in the phase II trial in patients with RCC,11 and the highest incidence again noted in the phase II randomised discontinuation trial in patients with RCC.3 The overall incidence of high-grade hypertension was 5·7% (95% CI 2·5–12·6), as determined by the random-effects model (figure 1B). Patients with RCC can be more susceptible to developing hypertension because of previous nephrectomy and renal dysfunction. Therefore, we further analysed the incidence of hypertension in patients with RCC compared with those with non-RCC malignancies, regardless of treatment asssignment. For patients with RCC, the overall incidence of all-grade and high-grade hypertension was 23·6% 120
(14·3–36·5) and 6·5% (1·8–21·1) respectively (figure 2), as determined by the random-effects model; for those patients with non-RCC malignancies, the overall incidences of all-grade and high-grade hypertension were 23·0% (16·0–31·9) and 5·3% (2·7–10·0) respectively, as analysed by the fixed-effects model (figure 3). No significant difference was detected in the incidence of sorafenib-associated all-grade hypertension (RR 1·03 [95% CI 0·73–1·45], p=0·89) and high-grade hypertension (RR 1·23 [95% CI 0·76–1·99], p=0·40) between patients with RCC and those with a non-RCC malignancy. A meta-analysis of the RR for hypertension with sorafenib in patients with metastatic RCC compared with the control group was done for the two randomised controlled trials.4,11 One trial used a placebo as the control,4 and the other used interferon.11 In both trials, the incidence of hypertension was low in the control group (1·8% and 6·7%, respectively). The overall RR was 6·11 for sorafenib versus control for all-grade hypertension (95% CI 2·44–15·32, p<0·0001) as calculated using the random-effects model (figure 4). Thus, sorafenib was associated with a significantly increased risk of hypertension in patients with RCC compared with patients with RCC not assigned sorafenib. To understand the risk of hypertension with sorafenib compared with other anti-angiogenesis drugs, we assessed the incidence of hypertension with bevacizumab, a humanised monoclonal antibody against vascular endothelial growth factor (VEGF). Bevacizumab is associated with a significantly increased risk of hypertension. We have previously shown that the RR for hypertension with bevacizumab is 3·0 (95% CI 2·2–4·2; p<0·001) for low doses (≤7·5 mg/kg per dose) and 7·5 (4·2–13·4; p<0·001) for high doses (10 or 15 mg/kg per dose).19 In that study, we did a meta-analysis of the clinical studies that were included in our previous study, and calculated that the overall incidence of hypertension associated with high-dose bevacizumab is 25·4% (95% CI 21·3–30·1) by use of the fixed-effects model (figure 5).
Discussion This study shows the overall risk of hypertension associated with sorafenib in patients with cancer. We noted a high overall incidence of all-grade hypertension http://oncology.thelancet.com Vol 9 February 2008
Articles
Molecular target
Incidence of hypertension (95% CI)
Relative risk of hypertension (95% CI)
Ref
Sorafenib
VEGFR-2 and VEGFR-3*; Raf; PDGFR; C-KIT; FLT-3; RET
23·4% (16·0–32·9)
6·1 (2·4–15·3)
This study
Sunitinib
VEGFR-1 and VEGFR-2*; PDGFR; C-KIT; FLT-3; RET
22·5% (19·5–25·9)
3·9 (2·6–5·9)
Zhu X, unpublished data
Bevacizumab
VEGF*
25·4% (21·3–30·1)†
7·5 (4·2–13·4)‡
19
AG013736
VEGFR-1, VEGFR-2, and VEGFR-3*; PDGFR; C-KIT
57·7%
NA
24
VEGF Trap
VEGF*
31·6%
NA
25
VEGFR=vascular endothelial growth factor receptor. PDGFR=platelet-derived growth-factor receptor. *Targets directly involved in angiogenesis. †Incidence is calculated from patients receiving high-dose bevacizumab (10 or 15 mg/kg per dose) by a meta-analysis using the fixed-effects model. ‡Relative risk is derived from patients receiving highdose bevacizumab (10 or 15 mg/kg per dose) by use of fixed-effects model. NA=not available.
Table 2: Risk of hypertension with angiogenesis inhibitors
(23·4% [95% CI 16·0–32·9]) with sorafenib and an overall incidence of high-grade hypertension of 5·7% (2·5–12·6). Adequate and aggressive management of moderate hypertension can be essential for many patients, because hypertension is an independent risk factor of renal and cardiovascular events.20–23 As seen in a phase III trial for patients with metastatic RCC,4 cardiac ischaemia or infarction were significantly more frequent in the sorafenib group (n=12 [3%]) than in the placebo group (n=2 [0·4%]; p=0·01). Furthermore, poorly controlled sorafenib-associated hypertension can lead to reversible posterior leucoencephalopathy syndrome.12 Finally, severe hypertension has led to the permanent discontinuation of sorafenib.4 Hypertension associated with angiogenesis inhibitors is an emerging issue in patients receiving these therapeutic drugs. In addition to sorafenib, several other angiogenesis inhibitors, such as bevacizumab, sunitinib, AG013736, and VEGF trap have been associated with the development of hypertension (table 2). In a meta-analysis, sunitinib, another multi-tyrosine-kinase inhibitor, was shown to be associated with hypertension with an incidence of 22·5% (95% CI 19·5–25·9) and an RR of 3·89 versus control (Zhu X, unpublished data). Our study shows that sorafenib is associated with a significantly increased risk of hypertension with an incidence of 23·4% (16·0–32·9) and an RR of 6·11 (95% CI 2·44–15·32; p<0·0001); high-dose bevacizumab (10 or 15 mg/kg per dose) is associated with an incidence of hypertension of 25·4% (21·3–30·1; figure 5) and an RR of 7·5 (4·2–13·4) compared with controls.19 When considering these combined findings, the risks of hypertension associated with these angiogenesis inhibitors seem considerable. The association of sorafenib with hypertension might be directly related to its inhibitory effect on the VEGF receptor. Possible mechanisms include: impaired angiogenesis leading to a decrease in the density of microvessels (a process known as rarefaction), endothelial dysfunction associated with a decrease in nitric-oxide production and an increase in oxidative stress; and changes in neurohormonal factors or the renin-angiotensin-aldosterone system. In a small study by Veronese and colleagues26 in http://oncology.thelancet.com Vol 9 February 2008
which 20 patients received 400 mg of sorafenib twice a day, there were no significant changes in humoral factors, including serum total catecholamines, epinephrine, norepinephrine, endothelin I, urotensin II, renin, and aldosterone after 3 weeks of treatment, even though an evident increase in systolic blood pressure (≥10 mm Hg) was noted in 75% of patients. Therefore, the role of these humoral factors in the development of sorafenib-associated hypertension might be limited, which implies that nonhumoral factors, such as impaired angiogenesis or endothelial dysfunction, play an important part. Patients with RCC might have a greater risk of hypertension than those with non-RCC malignancies as a result of previous nephrectomy and renal dysfunction.27–29 However, our study showed that the incidence of hypertension is not significantly different in patients with RCC compared with those with a non-RCC malignancy when they are treated with sorafenib (all grade: RR 1·03 [95% CI 0·73–1·45]; high-grade: RR 1·23 [0·76–1·99]). A possible explanation for this finding is that the increase in blood pressure and hypertension induced by sorafenib is so prominent that the risk associated with RCC is not evident in this setting. This notion is supported by the high incidence of all-grade hypertension with sorafenib (23·4% [95% CI 16·0–32·9) noted in this study and the high incidence of increased blood pressure in about 75% of patients receiving sorafenib in the study by Veronese and colleagues.26 Another explanation could be that the combined effect of VEGF from normal tissue and tumour tissue on bloodpressure regulation is comparable between patients with RCC and non-RCC malignancies, even though RCC can be associated with a higher VEGF secretion as a result of its biology;30 thus, the effect of sorafenib on hypertension might be similar. Additionally, sorafenib is mainly metabolised by the liver,31 and renal dysfunction associated with RCC might not affect the concentration of sorafenib in the blood in a substantial way. The management of sorafenib-associated hypertension is still under debate. According to the manufacturer package insert for sorafenib, hypertension can occur early in the course of treatment and, therefore, blood pressure should be monitored weekly during the first 121
Articles
6 weeks of treatment; thereafter hypertension should be monitored and treated in accordance with standard medical practice. In cases of severe or persistent hypertension despite the initiation of antihypertensive treatment, temporary or permanent discontinuation of sorafenib should be considered.31 In most patients, hypertension can be controlled with standard antihypertensive medications.3 However, the biological effect of these antihypertensive medications on angiogenesis and its implications should be considered, and need to be fully understood. Both enalapril (an angiotensin-convertingenzyme [ACE] inhibitor) and candesartan (an angiotensin receptor blocker) can inhibit myocardial angiogenesis induced by VEGF.32 Additionally, candesartan has been shown to have an antiangiogenesis effect in a xenograft model of bladder cancer.33 However, nifedipine (a calciumchannel blocker) has been shown to induce VEGF secretion.34 Thus, the possibility exists that some antihypertensive medications are more effective in treating anti-VEGF-associated hypertension and have less toxic effects when used in conjunction with sorafenib. Additionally, certain antihypertensive medications can interfere with the antineoplastic activity of sorafenib by modulating VEGF expression and reducing its antiangiogenic effect. Drug–drug interactions are also important issues. Sorafenib is metabolised mainly in the liver, where it undergoes an oxidative reaction mediated by the cytochrome p450 enzyme system—mainly by CYP3A4.31 Therefore, the metabolism of sorafenib can be potentially changed by certain antihypertensive drugs that are inhibitors of CYP3A4. For example, nondihydropyridine calcium-channel blockers, such as verapamil and diltiazem are CYP3A4 inhibitors, and can substantially increase the plasma concentration of sorafenib. However, ketoconazole, a potent CYP3A4 inhibitor, does not change the mean area under the curve of a single oral 50 mg dose of sorafenib in healthy volunteers when administered once a day at 400 mg for 7 days.31 Therefore, predicting the effect of CYP3A4 inhibitors on the metabolism of sorafenib is difficult. In the absence of available data from clinical studies, dihydropyridines, such as amlodipine and nifedipine, are the preferred class of calcium-channel blockers. Nondihydropyridine calcium-channel blockers should be used cautiously in conjunction with sorafenib. Alternatively, ACE inhibitors or angiotensin-receptor blockers are reasonable choices with the additional benefit of improving endothelial function and microcirculatory density.35 Other antihypertensive medications, such as diuretics, alpha blockers, and beta blockers, which have been used in the setting of bevacizumab-associated hypertension,36 can also be considered for use in this setting. Because sorafenibinduced hypertension might be related to reduced nitric-oxide production secondary to the anti-VGFR effect, the use of phosphodiesterase inhibitors or nitrate 122
derivatives to increase nitric oxide levels has been suggested and might be a treatment option depending on the findings of further clinical studies.37 This study has several limitations. As with any metaanalysis, the findings described here are affected by the limitations of the individual clinical trials that are included in the analysis. These trials might have underestimated the incidence of sorafenib-associated hypertension because of the imperfection of the CTCAE version 2 or 3 for recording adverse events.38 In both these versions, patients are considered hypertensive only if diastolic pressure is increased by more than 20 mm Hg or blood pressure is greater than 150/100 mm Hg. In our study, this grading criteria would have missed many patients with hypertension according to the standard criteria for the diagnosis of hypertension (140/90 mm Hg). Additionally, the grading criteria do not clearly differentiate between grade 2 and 3 hypertension, and our findings of high-grade hypertension could be affected by the overlap between these two grades. Furthermore, the prevalence of baseline hypertension was not described in the included trials, which might have led to an overestimation of the incidence of sorafenib-associated hypertension. However, the presence of baseline hypertension would probably have been low at about 2–4%, as estimated from patients who received placebos in the control groups. We have kept the likelihood of bias to a minimum by calculating the RR by use of randomised controlled clinical trials for the direct comparison of hypertension with and without sorafenib. Additionally, there seems to be no significant variation in the incidence of baseline hypertension in patients with metastatic cancers, as reflected in the control groups of many clinical trials.19 The incidence of hypertension in the control groups described here was 1·8%4 and 6·7%,11 consistent with that of the control patients in phase III bevacizumab trials (range 0–8·3%).19 Additionally, the patients in this study had metastatic cancers and were mainly selected from clinical trials. Therefore, our findings were noted mainly in academic centres and major research institutes, and might not entirely apply to patients with cancer who are treated in the community. However, this analysis also included many patients from the expanded-access programme,10 which did contain a substantial number of patients from the community setting, although the recording of hypertension and other data might be inadequate in this setting. Finally, even though we have shown that there is no significant difference in the incidence of sorafenib-associated hypertension between patients with RCC and non-RCC malignancies (RR 1·03 [95% CI 0·73–1·45], p=0·89), this finding might be limited by the small sample size of patients with a non-RCC malignancy who were assigned sorafenib (n=111 for all-grade hypertension; n=315 for high-grade hypertension analysis). This study has shown that sorafenib is associated with a significant risk of developing hypertension. Early detection and effective management of hypertension http://oncology.thelancet.com Vol 9 February 2008
Articles
might allow for safer use of this drug. The hypertensive and cardiovascular side-effects of sorafenib need thorough postmarketing surveillance and reporting, and future studies will be needed to identify the mechanism and appropriate treatment of sorafenibinduced hypertension. Contributors SW and XZ conceived the study, and SW was responsible for the data search. SW and JJC produced the figures and all authors contributed to the writing of the manuscript. Conflicts of interest SW is supported by the Research Foundation of the State University of New York at Stony Brook (NY, USA), and is a consultant for Onyx Pharmaceuticals and a speaker for Pfizer Inc. Acknowledgments The authors would like to thank Stefan Madajewicz for critically reading the manuscript. References 1 Wilhelm SM, Carter C, Tang L, et al. BAY 43-9006 exhibits broad spectrum oral antitumor activity and targets the RAF/MEK/ERK pathway and receptor tyrosine kinases involved in tumor progression and angiogenesis. Cancer Res 2004; 64: 7099–109. 2 Chang YS, Adnane J, Trail PA, et al. Sorafenib (BAY 43-9006) inhibits tumor growth and vascularization and induces tumor apoptosis and hypoxia in RCC xenograft models. Cancer Chemother Pharmacol 2007; 59: 561–74. 3 Ratain MJ, Eisen T, Stadler WM, et al. Phase II placebo-controlled randomized discontinuation trial of sorafenib in patients with metastatic renal cell carcinoma. J Clin Oncol 2006; 24: 2505–12. 4 Escudier B, Eisen T, Stadler WM, et al. Sorafenib in advanced clearcell renal-cell carcinoma. N Engl J Med 2007; 356: 125–34. 5 Abou-Alfa GK, Schwartz L, Ricci S, et al. Phase II study of sorafenib in patients with advanced hepatocellular carcinoma. J Clin Oncol 2006; 24: 4293–300. 6 Llovet J, Mazzaferro V, Hilgard P, et al. Sorafenib improves survival in advanced hepatocellular carcinoma (HCC): results of a phase III randomized, placebo-controlled trial (SHARP trial). Proc Am Soc Clin Oncol 2007; 25: (abstr LBA1). 7 Eisen T, Ahmad T, Flaherty KT, et al. Sorafenib in advanced melanoma: a phase II randomised discontinuation trial analysis. Br J Cancer 2006; 95: 581–86. 8 Gatzemeier U, Fosella F, Simantov R, et al. Phase II trial of singleagent sorafenib in patients with advanced non-small cell lung carcinoma. Proc Am Soc Clin Oncol 2006; 24: (abstr 7002). 9 Wu S, Posadas E, Scripture C, et al. BAY43-9006 (sorafenib) can lead to improvement of bone lesions in metastatic androgenindependent prostate cancer despite rises in serum PSA levels. 2006 Prostate Cancer Symposium. San Francisco, CA, USA. Feb 24–26, 2006: (abstr 259). 10 Figlin R, McDermott D, Gabrail N, et al. The Advanced Renal Cell Carcinoma Sorafenib (ARCCS) expanded access trial in North America: safety and efficacy. Proc Am Soc Clin Oncol 2007; 25: (abstr 5011). 11 Szczylik C. Demkow T, Staehler M, et al. Randomized phase II trial of first-line treatment with sorafenib versus interferon in patients with advanced renal cell carcinoma: final results. Proc Am Soc Clin Oncol 2007; 25: (abstr 5025). 12 Govindarajan R, Adusumilli J, Baxter DL, El-Khoueiry A, Harik SI. Reversible posterior leukoencephalopathy syndrome induced by RAF kinase inhibitor BAY 43-9006. J Clin Oncol 2006; 24: e48. 13 US National Cancer Institute. Common Toxicity Criteria, Version 2.0. http://ctep.cancer.gov/reporting/ctc_archive.html (accessed Jan 7, 2007). 14 Lewis S, Clarke M. Forest plots: trying to see the wood and the trees. BMJ 2007; 322: 1479–80. 15 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986; 7: 177–88. 16 Lau J, Ioannidis JP, Schmid CH. Quantitative synthesis in systematic reviews. Ann Intern Med 1997; 127: 820–26.
http://oncology.thelancet.com Vol 9 February 2008
17
18
19
20 21
22
23 24
25
26
27
28 29
30
31 32
33
34
35
36
37
38
D’Adamo DR, Keohan M, Schuetze S, et al. Clinical results of a phase II study of sorafenib in patients (pts) with non-GIST sarcomas (CTEP study #7060). Proc Am Soc Clin Oncol 2007; 25: (abstr 10001). Hobday TJ, Holen K, Picus J, et al. A phase II trial of sorafenib in patients with metastatic neuroendocrine tumors (NET): a phase II consortium (P2C) study. Proc Am Soc Clin Oncol 2007; 25: (abstr 4504). Zhu X, Wu S, Dahut WL, Parikh CR. Risks of proteinuria and hypertension with bevacizumab, an antibody against vascular endothelial growth factor: systematic review and meta-analysis. Am J Kidney Dis 2007; 49: 186–93. Kaplan NM, Opie LH. Controversies in hypertension. Lancet 2006; 367: 168–76. Peterson JC, Adler S, Burkart JM, et al. Blood pressure control, proteinuria, and the progression of renal disease. The Modification of Diet in Renal Disease Study. Ann Intern Med 1995; 123: 754–62. Wachtell K, Ibsen H, Olsen MH, et al. Albuminuria and cardiovascular risk in hypertensive patients with left ventricular hypertrophy: the LIFE study. Ann Intern Med 2003; 139: 901–06. Weisstuch JM, Dworkin LD. Does essential hypertension cause endstage renal disease? Kidney Int Suppl 1992; 36: S33–37. Rixe O, Bukowski RM, Michaelson MD, et al. Axitinib treatment in patients with cytokine-refractory metastatic renal-cell cancer: a phase II study. Lancet Oncol 2007; 8: 975–84. Nguyen QD, Shah SM, Hafiz G, et al. A phase I trial of an IVadministered vascular endothelial growth factor trap for treatment in patients with choroidal neovascularization due to age-related macular degeneration. Ophthalmology 2006; 113: 1522e1–e14. Veronese ML, Mosenkis A, Flaherty KT, et al. Mechanisms of hypertension associated with BAY 43-9006. J Clin Oncol 2006; 24: 1363–69. McKiernan J, Simmons R, Katz J, Russo P. Natural history of chronic renal insufficiency after partial and radical nephrectomy. Urology 2002; 59: 816–20. Melman A, Grim CE, Weinberger MH. Increased incidence of renal cell carcinoma with hypertension. J Urol 1977; 118: 531–22. Shirasaki Y, Tsushima T, Nasu Y, Kumon H. Long-term consequence of renal function following nephrectomy for renal cell cancer. Int J Urol 2004; 11: 704–08. Schips L, Dalpiaz O, Lipsky K, et al. Serum levels of vascular endothelial growth factor (VEGF) and endostatin in renal cell carcinoma patients compared to a control group. Eur Urol 2007; 51: 168–73. Bayer. NEXAVAR Prescribing Information. http://www.univgraph. com/bayer/inserts/nexavar.pdf (accessed Jan 7, 2008). Siddiqui AJ, Mansson-Broberg A, Gustafsson T, et al. Antagonism of the renin-angiotensin system can counteract cardiac angiogenic vascular endothelial growth factor gene therapy and myocardial angiogenesis in the normal heart. Am J Hypertens 2005; 18: 1347–52. Kosugi M, Miyajima A, Kikuchi E, Horiguchi Y, Murai M. Angiotensin II type 1 receptor antagonist candesartan as an angiogenic inhibitor in a xenograft model of bladder cancer. Clin Cancer Res 2006; 12: 2888–93. Miura S, Fujino M, Matsuo Y, Tanigawa H, Saku K. Nifedipineinduced vascular endothelial growth factor secretion from coronary smooth muscle cells promotes endothelial tube formation via the kinase insert domain-containing receptor/fetal liver kinase-1/NO pathway. Hypertens Res 2005; 28: 147–53. Agabiti-Rosei E. [Structural and functional changes of the microcirculation in hypertension: influence of pharmacological therapy]. Drugs 2003; 63 (Spec No 1): 19–29. Mares JE, Worah S, Mathew SV, et al. Increased rates of hypertension (HTN) among patients with advanced carcinoid treated with bevacizumab. Proc Am Soc Clin Oncol 2005: 23: (abstr 4087). Izzedine H, Rixe O, Billemont B, Baumelou A, Deray G. Angiogenesis inhibitor therapies: focus on kidney toxicity and hypertension. Am J Kidney Dis 2007; 50: 203–18. National Cancer Institute. CTC v 2.0 and Common Terminology Criteria for Adverse Events Criteria V3.0 (CTCAE). http://ctep. cancer.gov/reporting/ctc.html (accessed Jan 7, 2007).
123