Antiangiogenic therapies target the VEGF molecule, its receptor, or downstream pathways. FDA-approved antiangiogenic agents include bevacizumab, a recombinant humanized monoclonal antibody that binds and sequesters the VEGF molecule, [4] and multi-targeted TKIs , small molecules that inhibit the vascular endothelial growth factor receptor (VEGFR) intracellular intrinsic kinase activity, such as sunitinib, sorafenib, axitinib, and pazopanib [5]. TKIs are not entirely specific for VEGFR2, they also have inhibitory activity against other receptor tyrosine kinases, such as the platelet-derived growth factor receptor or c-Kit. The use of these medications has since expanded to many different solid tumors, with numerous clinical trials of newer formulations of the medications underway. Antiangiogenic therapies are now the first-line therapies for cancers such as metastatic renal cell carcinoma, which accounts for 2.5 % of all new cancer diagnoses. With increased use of these medications and many highly potent formulations in development, there is an increasing clinical need to understand how to diagnose and manage VEGF inhibitor toxicities.

Hypertension is a very common toxicity of antiangiogenic therapy. Hypertension occurs in 19–24 % of patients receiving FDA-approved therapies [6, 7], but it can occur in up to 80 % of patients on the newer experimental forms of these medications [8]. Nearly all patients taking these drugs experience a rise in BP, even if not to hypertensive levels. With high-potency TKIs, BP rise can be rapid—within days. On the other hand, with biologic therapies such as bevacizumab, the rise can be slower, over weeks or months, as in the patient in Case #1. In one study examining sorafenib, among 54 patients initiating therapy, 93 % had a rise in BP by day 6, and most experienced a rise in BP over the first 24 h of therapy, as assessed by ambulatory blood pressure monitoring (ABPM) [9, 10]. BP typically falls when treatment is interrupted. We reported a very rapid rise in BP among women initiating cediranib therapy, a highpotency small-molecule VEGF-targeted therapy , with 67 % of patients developing hypertension over the first 3 days of therapy, and 87 % by the end of the study [8] .

The binding of VEGF to its receptor VEGFR2 activates the intrinsic tyrosine kinase activity of VEGFR2, which normally couples with eNOS and increases synthesis of vasodilatory NO [11]. Based on this, inhibition of VEGF signaling decreases NO bioavailability, causing vasoconstriction. Indeed, VEGF inhibition is associated with decreased urinary nitrite/nitrate excretion and decreased serum levels of NO metabolites in humans [12, 13], although no difference in flow-mediated dilation, a surrogate for NO bioavailability, has also been reported [13]. Treating mice with an anti-VEGFR2 antibody causes a rise in BP and reduces kidney eNOS and nNOS [14]. On the other hand, in a swine model of sunitinib-induced hypertension, NO bioavailability does not appear to contribute to sunitinib-induced hypertension [15]. Despite some conflicting evidence in the literature, most evidence to date does implicate increased peripheral vascular resistance in the pathophysiology of antiangiogenic therapy-induced hypertension .

Another important factor underlying antiangiogenic therapy-induced hypertension is ET-1, a potent vasoconstrictor. Sunitinib induces a rise in circulating levels of ET-1 in rodents, as well as in humans [16]. ET-1 levels rise rapidly after starting regorafenib, a TKI, and they normalize just as rapidly after discontinuation [17]. Pig models suggest that the rise in BP induced by antiangiogenic therapy can be prevented ET receptor antagonists [15, 18]. Since antiangiogenic therapies induce generalized endothelial dysfunction, which itself is a known trigger of ET-1 secretion [19], it is likely that this also plays a role in the underlying pathophysiology.

A third factor in antiangiogenic therapy-induced hypertension is microcapillary rarefaction. Although the data supporting a role for this are less strong, several observations suggest it may be playing a role. Rodents treated with VEGF inhibitors experience regression of tracheal capillary networks by up to 30 % at 21 days of therapy, and this process reverses with antiangiogenic therapy discontinuation [20]. In humans, capillary density reduction has been measured in patients taking either bevacizumab or TKIs, of the order of 10–20 % reduction [21, 22]. It is important to note, however, that increasing peripheral vascular resistance by 5 % requires rarefaction of 40 % of the microcapillary bed—more than what has been observed in humans [16, 23]. At the present time, more research is needed to better define the role of capillary rarefaction in antiangiogenic therapy-induced hypertension .

Other pathways may also be playing roles. A shift in the pressure–natriuresis curve, causing volume overload, has been reported by Facemire and colleagues [14]. Macrophage-derived VEGF-C has been proposed to regulate lymphangiogenesis and extracellular fluid buffering, and anti-VEGF therapies may interrupt this capacity [24]. The degree to which these pathways and others, such as prostaglandins or reactive oxygen species might contribute remains unclear [25]. Figure 5.1 summarizes these pathways .

Proteinuria after antiangiogenic therapy is most commonly the consequence of renal TMA .

In one study of 1850 patients receiving bevacizumab, the incidence of new proteinuria ascribable to bevacizumab was about 5–10 % and the risk was dose-dependent (RR, 1.4 with low-dose bevacizumab; 95 % confidence interval, 1.1–1.7; RR, 2.2 with high dose; 95 % CI, 1.6–2.9) [6]. This proteinuria probably reflects renal TMA . In a larger meta-analysis (12,268 patients), the incidence of any grade proteinuria associated with bevacizumab was found to be 13.3 % (95 % confidence interval, 7.7–22.1 %) [26]. The incidence of high-grade proteinuria (greater than 3.5 g per day) was 2.2 % (95 % CI 1.2–4.3 %), and the incidence of nephrotic syndrome was 0.8 % (95 % CI 0.4–1.8 %) with a relative risk of 7.78 (95 % CI: 1.80–33.62; P = 0.006) .

The kidney histology in patients who develop proteinuria after starting antiangiogenic therapy includes endotheliosis, focal foot process effacement, evidence of glomerular basement membrane damage, including double contours and mesangiolysis. In severe cases fibrin deposition and red blood cell entrapment is also seen. These are the histologic characteristics of renal TMA.

The pathophysiology of this syndrome involves VEGF expressed by podocytes and implicates podocyte–EC crosstalk [27]. Eremina et al. investigated the role of podocyte-derived VEGF in the maintenance of the glomerular filtration barrier. Using conditional mouse genetic models, they specifically ablated VEGF from podocytes alone, and showed that this was sufficient to cause renal TMA [28]. Thus, antiangiogenic therapies act by blocking the effect of podocyte-derived VEGF on adjacent glomerular EC, resulting in disruption of the glomerular filtration barrier with proteinuria, as well as enhanced clotting [29]. Numerous reports have now documented that renal TMA lesions are the most common pattern of glomerular injury found in humans receiving antiangiogenic therapy [28, 30–33]. However, other biopsy findings have also been reported, including immune complex glomerulonephritis [31, 34, 35] and allergic interstitial nephritis [36–38] .

Inhibition of the EGFR is a recognized cause of both hypomagnesemia and secondary hypokalemia. Monoclonal antibody inhibitors of the EGFR include cetuximab and panitumumab. Small-molecule EGFR inhibitors include gefitinib and erlotinib. The overall incidence of hypomagnesemia in patients treated with these drugs is 17 %. Compared to controls not receiving anti-EGFR therapy, the relative risk of hypomagnesemia for cetuximab was 3.87, and the relative risk for panitumumab was 12.55 [39]. Because renal magnesium handling is dependent on normal EGFR signaling, hypomagnesemia reflects adequate EGFR inhibition in vivo and may therefore serve as a surrogate of anticancer efficacy. Indeed, early hypomagnesemia could be a biomarker for superior cancer response. In a study of KRAS wild-type colorectal cancer patients, those that developed an early fall in magnesium > 50 % by day 28 had a higher tumor response rate (55.8 vs. 16.7 %, P < 0.0001) and a longer overall survival (11.0 vs. 8.1 months, P = 0.002) [40].