The potential role of sunitinib in gastrointestinal cancers other than GIST

2.1. Mode of action 

Sunitinib, previously known as SU-011248, has been identified as a potent inhibitor of VEGFR-1, -2, and -3, PDGFR-α, and -β, c-KIT, RET, and FLT-3. The inhibitory activity of sunitinib was evaluated in biochemical and clinical assays, demonstrating a competitive inhibition in the nanomolar range of all the mentioned receptors. In fact, the IC50 values of sunitinib against the VEGFR-2, PDGFR, and other receptors are comparable to or lower than that of many similar compounds in development (see Table 1) [8]. Sunitinib is a small molecule with a chemical structure similar to the ATP, hence competes with the ATP for binding to the ATP domain of the intracellular portion of the tyrosine kinase membrane receptor. Therefore, from the rational point of view, sunitinib can inhibit angiogenesis in a more complete sense, not only by blocking all the VEGFRs known (-1, -2, and -3), but also by blocking the pericyte cells through PDGFR-β (Fig. 2). Sunitinib can also directly inhibit the growth or survival of selected tumor types with dysregulated or overexpressed tyrosine kinase receptors involved in regulation of cell proliferation or cell survival conferring sunitinib a dual antiangiogenic and antitumor activity [9].

Table 1. Pharmacological enzyme 50% inhibitory concentration (IC50) in nanomolar range of the principal multitargeted angiogenesis inhibitors under development.

	Agent	VEGFR-1	VEGFR-2	VEGFR-3	PDGFR-β	c-Kit	Raf	FLT-3
	PTK787	77	37	660	580	730	–	–
	Sunitinib	2	9	17	2	22	–	250
	Sorafenib	–	90	20	57	68	6	–
	Vandetanib	1600	40	110	1100	20,000	–	–
	Axitinib	1	0.25	0.29	2.5	2	–	–
	Cediranib	5	1	3	5	2	–	1
	AEE788	59	77	330	320	790	2800	730

VEGFR: vascular endothelial growth factor receptor; PDGFR: platelet-derived growth factor receptor; Flt-3: FMS-like tyrosine kinase 3.

View Large Image Download to PowerPoint

Fig. 2. Multitargeted mode of action of sunitinib.

2.2. Pharmacokinetics 

Sunitinib is primarily metabolized by cytochrome CYP3A4, resulting in the formation of a pharmacologically active N-desethyl metabolite, SU12662, that is equipotent to sunitinib activity [10]. Pharmacokinetic data indicate good oral absorption, a prolonged half life for sunitinib (40h) and its active metabolite, SU12662 (80h) and linear kinetics at the doses administered. Across all dose levels, the time to plasma peak (Tmax) was generally observed at 4–6 and 8–12h, for both sunitinib and its metabolite. Sunitinib can be administered with or without food because no effect on drug exposure was seen in a food effect study in 16 healthy subjects [11]. Drug plasma protein binding rate is estimated to be 95% with a largest volume of distribution of 2230L. There is no reduction in the variability of exposure by normalizing to body surface area. Sunitinib is primarily excreted in the faeces (70–84%) and only 16% of the drug is excreted in urine. Pharmacokinetic/Pharmacodynamic data from animal studies showed that target plasma concentrations of sunitinib plus SU12662 capable of inhibiting PDGFR-β and VEGFR-2 phosphorylation were established in the range of 50–100mg/ml. The dosing studies demonstrated that maintenance of target plasma concentrations above 50–100ng/ml for at least 12h on a daily dosing regimen was potentially important for efficacy [8]. Sunitinib accumulates 3–4-fold with repeated daily administration and SU-12662 accumulates 7–10-fold. Steady-state concentrations of sunitinib and its primary active metabolite are achieved within 10–14 days. Population pharmacokinetic analyses of demographic data indicate that there are no clinically relevant effects of age, body weight, creatinine clearance, race, gender or performance status on the pharmacokinetics of sunitinib or the primary active metabolite [12].

4.1. Sunitinib in colorectal cancer 

In colorectal cancer (CRC) patients, circulating VEGF levels are higher than in normal population and seem to be correlated with tumor stage: the higher circulating VEGF levels the higher stage [14]. Increased expression of VEGF in CRC has been associated with tumor vascularization, proliferation of cancer cells and metastasis [15]. It is suggested that VEGF expression can be more sensitive than traditional staging methods for predicting prognosis and risk of relapse [14].

Fluoropyrimidines, mainly 5-fluorouracil (5FU), have been the key agents to treat CRC for decades. In the 90s, irinotecan and oxaliplatin were added to the armamentarium and recently new selective targeted agents (bevacizumab, cetuximab, and panitumumab) have demonstrated an improved response rate (RR) and/or PFS. Both hepatic resection techniques and chemotherapy improvement have contributed to prolong overall survival in CRC patients [16]. Despite of these new drugs, the 5-year survival rate for metastatic CRC (mCRC) remains <10% and additional active agents are needed [17].

Sunitinib has shown activity in colon cancer cell lines and xenografts models. In these preclinical models using HT-29 and Colo-205 cell lines, sunitinib has significantly inhibited tumor growth or even induced marked regression of largely established tumors [18], [19]. Saltz et al. [17] conducted a phase II trial with sunitinib as single agent in 84 mCRC patients who had previously failed to standard therapy. All patients received prior treatment with irinotecan, oxaliplatin, and fluoropyrimidines. A total of 63 patients had received treatment with another therapeutic agent and 46 of them did receive either cetuximab or bevacizumab or both. One patient achieved a confirmed PR and eleven (28%) patients in the bevacizumab-naïve cohort experienced stable disease (SD) for at least 22 weeks. Median time to progression (mTTP) was 2.2 and 2.5 months, and median OS (mOS) was 7.1 and 10.2 months in prior bevacizumab and bevacizumab-naïve cohorts, respectively. Safety profile was similar to other studies with sunitinib in other tumor types. The most often AEs were fatigue (13%), vomiting (7%), nausea (4%), and anorexia. No treatment-related grade 4 events were reported in more than one patient. Overall, sunitinib dosing was delayed or reduced in 38 patients as a result of AEs.

Although sunitinib failed to demonstrate a significant single-agent objective response rate in heavily pretreated patients, the mechanism of action, mild toxicity, and preclinical synergy with chemotherapy served to conduct two phase I trials combining sunitinib plus chemotherapy. A phase I trial by Leong et al. [20] investigated the combination of a fixed dose of FOLFOX6 [oxaliplatin 85mg/m2, leucovorin (LV) 400mg/m2, bolus 5FU 400mg/m2, and 46h-continuous infusion 5FU 2400mg/m2 on day 1 of a 2 weeks cycle] plus escalating doses of sunitinib. The maximum tolerated dose (MTD) of sunitinib plus FOLFOX6 was 50mg/day in 2 weeks “on” followed by 2 weeks “off” (2/2) schedule. This regimen was safe and tolerable. Thrombocytopenia and neutropenia were the DLTs and other common AEs were fatigue, gastrointestinal effects, and oxaliplatin neuropathy.

Another phase I trial was conducted in 27 patients treated with FOLFIRI (irinotecan 180mg/m2, LV 200mg/m2, bolus 5FU 400mg/m2, and 46h infusion of 5FU 2400mg/m2) plus escalating dosing of sunitinib [21]. Two out of 21 patients treated in the 37.5-mg level in the 4/2 schedule had DLTs (febrile neutropenia and grade 4 neutropenia). Therefore, the MTD for sunitinib 4/2 schedule in combination with FOLFIRI was 37.5mg/d. Confirmed PR were observed in 8 of the 19 evaluable patients treated with at MTD and, in addition, 10 patients achieved SD. This combination regimen appears safe and generally well tolerated. However, it has recently been communicated in a press release that a randomized phase III trial comparing FOLFIRI versus FOLFIRI plus sunitinib has been stopped due to futility according to Data Safety Monitoring Committee (DSMC) decision [22].

4.2. Sunitinib in gastric cancer 

VEGF plays an important role in gastroesophageal cancers. Its overexpression correlates with microvascular density, depth of tumor invasion, presence of lymph node and distant metastasis, recurrence and worse prognosis. When compared to normal mucosa, Barret's metaplasia or dysplasia, esophageal adenocarcinoma patients have significantly higher VEGF and VEGF messenger RNA circulating levels [23].

Five-year overall survival of gastric cancer (GC) patients is less than 30%. There is no standard of care for metastatic GC previously progressed to one platinum-based regimen. Therefore, there is an urgent need to find new active drugs. Sunitinib has demonstrated both antiangiogenic and antitumor activity in a preclinical model of GC [24]. In a phase II single arm trial conducted in metastatic GC patients, who had previously progressed to one prior chemotherapy regimen, sunitinib was given at the standard dose of 50mg/day in 4/2 schedule. In 42 patients evaluable at the time of the analysis, there were 2 PR (5%) and 15 (36%) had stable disease, with a duration of SD>3 months in 12 cases. Median TTP and mOS were 17.1 and 50.7 weeks, respectively. The most commonly AEs were stomatitis and skin discoloration (all grade 1 or 2). Toxicities grade ≥3 were neutropenia (31%), thrombocytopenia (29%), anemia (14%), anorexia (10%), hand-foot syndrome (10%), fatigue (10%), nausea (5%), vomiting (5%), mucositis (5%), and hypertension (2%). Eighteen patients (43%) required dose reduction or transient dose interruption [25].

As sunitinib demonstrated single agent activity in the refractory advanced GC and had acceptable tolerability, three phase I are currently ongoing in the first line setting combining sunitinib with cisplatin- and fluoropyrimidine-based chemotherapy (5FU, S1, and capecitabine). Future development of sunitinib in gastric cancer will depend on the safety and efficacy results obtained from these phase I trials.

4.3. Sunitinib in hepatocellular carcinoma 

Hepatocellular carcinoma (HCC) is a highly vascular cancer in which VEGFR and PDGFR play key roles in proliferation and angiogenesis. High VEGF expression is associated with inferior survival and circulating VEGF levels increase in advancing HCC stages [26]. Recently, sorafenib, another multitargeted agent, has become the first agent to demonstrate improved OS in advanced HCC when compared to placebo. However, prognosis is still poor and mOS is less than 1 year [27].

Two phase II trials with similar design were conducted in advanced HCC patients. In the European and Asian trial [28], 37 patients were treated with sunitinib at doses of 50mg/day in the 4/2 schedule. Clinical benefit rate, defined as partial response plus stable disease longer than 3 months, was 38% (14 patients) but the RR was very low (only 1 case; 2.7%). Interestingly, major tumor necrosis was observed in 48% patients. Median mOS was 8 months. Commonly observed grade 3 and 4 adverse events included thrombocytopenia (37.8%), neutropenia (24.3%), asthenia (13.5%), hand-foot syndrome (10.8%), and anaemia (10.8%). There were four deaths among the 37 patients (10.8%) that were possibly related to treatment. So, sunitinib showed pronounced toxicities at a dose of 50mg/day in patients with unresectable HCC. In the American study [29], 34 patients were treated with lower dose than in the previous trial (37.5mg/day 4/2 schedule). Six patients had previous systemic chemotherapy and/or chemoembolization. Only one patient (2.9%) achieved a PR, but the stable disease rate was high (50%). Median PFS and mOS were 3.9 and 9.8 months, respectively. Most common AEs, generally manageable, were neutropenia (62%), thrombocytopenia (65%), fatigue (62%), elevation of transaminases (59%), and hand-foot syndrome (15%).

There was a translational analysis of circulating biomarkers performed in both trials. DePrimo et al. [30] found that sunitinib was associated with increases in plasma VEGF-A, and decreases in plasma soluble VEGFR-2 and VEGFR-3, s-Kit, and VEGF-C. Baseline levels of VEGF-C previous to sunitinib treatment were higher in patients with longer TTP and OS, and with best objective response (PR or SD>3 months). These findings suggest that VEGF-C may represent a predictive or prognostic biomarker in HCC. On the other hand, greater decreases in s-KIT and VEGFR-2 levels during sunitinib treatment correlated with longer TTP and also with reduced tumor density. In the American trial, circulating molecular and cellular biomarkers were evaluated before and at six time points after sunitinib treatment. Sunitinib rapidly decreased vessel leakiness, and this effect was more pronounced in patients with delayed progression. When evaluated early (at baseline and day 14) as well as over three cycles of treatment, higher levels of inflammatory molecules (e.g., interleukin-6, stromal-derived factor 1 alpha, soluble c-KIT) and circulating progenitor cells were associated with a poor outcome [29].

There is currently a pair of ongoing randomized phase III trials in the first line setting, one of them compares sunitinib versus placebo in Asian countries, and the other one compares head to head sunitinib (37.5mg/d continuous) versus sorafenib worldwide only in patients with Child-Plough A status.

In conclusion, sunitinib is not considered as standard treatment for HCC patients and it has only shown some activity in patients with Child-Plough A status. There are concerns on the safety of sunitinib when given at full dose of 50mg/d in advanced HCC patients. Future role of circulating biomarkers predicting outcome with sunitinib treatment will be prospectively determined.

4.4. Sunitinib in pancreatic cancers 

Pancreatic adenocarcinoma (PAC) is a highly vascularised tumor with VEGF common overexpression that usually predicts for liver metastasis and poor prognosis [31]. Despite of advances in understanding of molecular and genetic basis of exocrine pancreatic cancer (PC), the outcome for this disease remains dismal. Gemcitabine, the standard chemotherapy, offers modest improvement of tumor-related symptoms and little advantage in survival. Addition of new targeted agents has marginal effect or does not increase significantly the OS but adding side effects. In a preclinical mouse model PAC cell lines, sunitinib given previously to the irradiation significantly delayed tumor growth versus either sunitinib or irradiation alone [32]. A phase II trial of sunitinib single agent in refractory advanced PAC was conducted in 77 patients by the CALGB Group [33]. ECOG PS≤2 was accepted. Surprisingly, one patient achieved a complete response and 14 patients (20%) had SD. Toxicity was generally mild, but there were two deaths not clearly drug-related (one due to GI perforation and one due to dyspnea). PS seems to be critical determinant of prognosis and possibly more stringent eligibility criteria (ECOG≤1 only) might provide clearer insights into a sub-population who could benefit for sunitinib in second-line therapy. There is no planned, at least at this moment, any large randomized phase III trial with sunitinib in patients with advanced exocrine PC.

Tumors derived from the endocrine portion of the pancreas, the pancreatic neuroendocrine tumors (PNET), also known as islet cell tumors, are highly vascularised and overexpress both VEGF and VEGFR. Moreover, elevated levels of circulating VEGF have been associated with tumor progression [34]. In the RIP1-TAG2 transgenic mouse model, a well established model for the study of islet cell carcinoma [35], sunitinib reduced tumor burden and increased animal survival by inhibiting tumor angiogenesis [36]. In the phase I conducted by Faivre et al. [9], one out of 4 patients with neuroendocrine tumors achieved a partial response by RECIST criteria and another one resulted in a minor response with a long disease stabilization.

Sunitinib was tested in a single arm, open label, phase II multicenter study that included both advanced PNET and carcinoid. A total of 107 patients were recruited, 66 of them with histologically or cytologically proven PNET not amenable to surgery, radiation, or combined modality therapy with curative intent. Patients received sunitinib 50mg/day for 4 weeks, followed by 2 weeks off. Overall objective response rate by RECIST criteria in the pancreatic cohort was 16.7% (gastrinoma, n=1; vasoactive intestinal peptide tumor, n=1, non-functioning PNET, n=9). By contrast, only one out of 14 patients with foregut carcinoid tumors had a confirmed response. Stable disease was achieved in 69% PNET and in 82% carcinoid patients. mTTP was 8 and 10 months, and 1-year OS rate was 81% and 83% in PNET and carcinoid patients, respectively. Toxicity profile was similar to that observed in other trials with sunitinib. Most common treatment-related toxicities were constitutional (fatigue and anorexia) or GI (diarrhea and nausea) [37]. In conclusion, sunitinib resulted in higher objective response in pancreatic endocrine tumors versus carcinoids, and TTP and OS mirrored the ones achieved by classical cytotoxic agents in this setting.

Raymond et al. have conducted a phase III randomized, double-blind study, to assess the efficacy and safety of sunitinib in patients with progressive advanced/metastatic pancreatic islet cell tumors [38]. The key inclusion criteria included: age≥18 years; histologically or cytologically proven, well differentiated pancreatic islet cell tumor according to WHO 2000 classification; local, locally advanced or metastatic disease; documented progression by computed tomography, magnetic resonance imaging or Octreoscan® within 12 months prior to baseline; disease not amenable to surgery, radiation, or combined modality treatment with curative intent; ECOG PS of 0 or 1; and adequate organ function. The key exclusion criteria were: poorly differentiated PNET tumors, according to WHO 2000 classification; prior treatment with any tyrosine kinase inhibitors or anti-VEGF angiogenic inhibitors (prior treatment with non-VEGF-targeted angiogenic inhibitors was permitted); myocardial infarction, severe/unstable angina, symptomatic congestive heart failure; cerebrovascular accident or transient ischemic attack, or pulmonary embolism within the 12 months prior to study drug administration; abnormal cardiac function, ongoing cardiac dysrhythmias of National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE) grade ≥2, atrial fibrillation of any grade, or prolongation of the QTc interval to >450ms for males or >470ms for females; symptomatic brain metastases, spinal cord compression, or new evidence of brain or leptomeningeal disease; and left ventricular ejection fraction (LVEF) ≤50% as measured by or either multigated acquisition (MUGA) scan or echocardiogram (ECHO). Patients were randomized 1:1 to receive either sunitinib or placebo treatment, each with best supportive care. Primary endpoint was PFS and secondary endpoints were OS, RR, duration of response, time to tumor response, safety and patient-reported outcomes (PRO). Sunitinib 37.5mg was administered in a continuous once-daily dosing regimen. Patients not experiencing a response according to RECIST and experiencing only ≤ grade 1 non-hematologic or ≤ grade 2 hematologic treatment-related adverse events during the first 8 weeks were permitted dose escalation to 50mg/day. Based on the type and severity of toxicity, dose reduction to 25mg/day was permitted with the option for subsequent re-escalation. Sunitinib treatment could be continued within the study until disease progression, unacceptable toxicity, or until the final analysis was performed. As of February 2009, 154 patients were enrolled in 11 countries in Europe, Asia and America. Safety and tolerability data are available for 57 patients treated with sunitinib. The most common treatment-emergent, all-causality, all-grade AEs in patients treated with sunitinib were diarrhea (63%), nausea (53%), vomiting (39%), asthenia and fatigue (35% each). Palmar–plantar erythrodysesthesia and hypertension occurred in 30% and 23% of patients, respectively.

Efficacy data are available for 154 patients; 75 patients treated with sunitinib and 79 patients who received placebo. Preliminary data demonstrated a substantial improvement in progression free survival with sunitinib [11.1 months vs. 5.5 months (HR: 0.39; 95% CI: 0.243, 0.649; p<0.001)]. Although it was planned to enrol a total of 340 patients, and based upon the DSMC recommendation, this study was terminated early due to differences in efficacy and good tolerability.

4.5. Sunitinib in billiary tract cancers 

Billiary tract carcinomas (BTC) are uncommon highly fatal malignancies in western countries that comprise gallbladder carcinoma and cholangiocarcinoma. Even with aggressive surgery, general outcome of patients with BTC is disappointing and unfortunately most of them are diagnosed at advanced stages when the tumor is unresectable and mOS is only few months. VEGF expression examined by immunohistochemistry is an independent negative predictor of response in extrahepatic BTCs and correlates with tumor progression in gallbladder cancer [39]. New active drugs are urgently needed in these tumors but, unfortunately, other antiangiogenic compounds like bevacizumab or sorafenib have failed in this setting [40], [41]. There are at least two ongoing phase II trials with sunitinib in BTC which results have not been reported yet.