Blood pressure monitoring in patients receiving bevacizumab.

We have read with great interest the prospective study by Shah et al. [1] assessing the safety of short infusions of bevacizumab. The authors report similar rates of proteinuria and hypertension in patients receiving the standard infusion regimen (over 90, 60 then 30 min), compared with shorter administrations with an infusion rate of 0.5 mg/kg/min. Although we agree with Shah et al. that shorter infusions appear safe in terms of short-term reactions as well as cumulative vascular toxic effects [2, 3], we would like to point out potential weaknesses of this prospective study. First, the rhythm of assessment of blood pressure is not detailed. We have previously shown that assessing blood pressure only before each infusion detects significantly less cases of hypertension than daily blood pressure monitoring did [4]. Second, the authors do not mention the use of a validated blood pressure measurement device, as recommended in recent international recommendations [5]. Third, hypertension grading used in this study was based on NCI-CTC v3.0, with a modified definition for grade 4 hypertension. As well, we have previously shown that NCI-CTC v3.0 allowed the detection of bevacizumab-induced hypertension in significantly fewer cases than the ESH (European Society of Hypertension) classification did [5]. Notably, the recent NCICTC v4.0, which was not available at the time of the study by Shah et al., is quite similar to the ESH classification. As a consequence, we should keep in mind that the proportion of patients developing hypertension in both the prospective and retrospective arms of this study may be under-estimated. The use of intensive blood pressure monitoring and the recent NCI-CTC v4.0 classification for hypertension grading should be generalized in further studies of hypertension induced by anti-vascular endothelial growth factor (VEGF) agents, as stressed by recent recommendations. Moreover, these tools should also be implemented in routine clinical practice, in particular in real-life populations of patients, as seen in the study by Shah et al. (with up to 50% of patients receiving antihypertensive drugs before initiation of bevacizumab).