A PHASE II STUDY OF GEMCITABINE AND CAPECITABINE IN PATIENTS WITH ADVANCED RENAL CELL CANCER (RCC): SOUTHWEST ONCOLOGY GROUP STUDY S0312

Abstract

Background

Gemcitabine plus capecitabine has modest efficacy in patients with advanced RCC but has considerable toxicity. We evaluated the efficacy and toxicity of a modified dose-schedule of this doublet in patients with advanced unresectable or metastatic RCC.

Methods

Chemotherapy-naïve patients were treated with gemcitabine at 900mg/m² on days 1,8,15 and capecitabine at 625mg/m² twice daily on days 1 through 21, every 28 days. Eligible patients must have adequate performance status and end-organ function. The primary endpoint was tumor response rate (RR). No further evaluation of this regimen would be pursued if the RR was ≤ 5%. In an exploratory manner using archival specimens, we also evaluated potential markers of prognosis and treatment response including thymidylate synthase (TS) gene polymorphisms and tumor expression of p53, PTEN, pAKT, pmTOR, and ERCC1.

Results

Of 43 patients registered, 1 was ineligible and 2 were not analyzable. There was 1 confirmed complete response (CR) and three unconfirmed partial responses (PR), for an overall response rate of 10% (95% CI: 3, 24). Nineteen patients (48%) had stable disease (SD). The six-month freedom-from-treatment-failure and overall survival rates were 20% (95% CI: 8, 32) and 75% (95% CI: 62, 88), respectively. Median survival time was 23 months (95% CI: 10, 37). One patient each experienced Grade 4 neutropenia, fatigue, thrombocytopenia and hemolysis with renal failure. The most common Grade 3 toxicities were neutropenia (12 patients), fatigue (5), and leucopenia (4). Patients with a best response of stable disease or better were more likely to have a decrease in expression of PTEN and an increased expression of pmTOR.

Conclusions

Gemcitabine plus capecitabine at this reduced dose-schedule benefits a small percentage of patients with RCC with an acceptable toxicity profile. The combination of gemcitabine and capecitabine may serve as a base regimen for combination therapy with targeted agents in select RCC patients.

INTRODUCTION

In 2008, an estimated 54,360 people in the United States will be diagnosed with renal cell cancer (RCC) and more than 13,000 will die from this disease.¹ Approximately 20% of patients will have locally advanced disease at diagnosis and 25% will have metastatic disease at diagnosis.^2,3

Historically, attempts to treat locally advanced or metastatic renal cell cancer with chemotherapy have been unsuccessful with combined response rates of 4 - 6% for most single agent or combination regimens. The response to cytokine-based immunotherapy has been varied, and, while the administration of high-dose bolus interleukin-2 (IL-2) has produced durable responses in a small percentage of patients (7-10%), toxicity limits IL-2's utility to a small number of patients. Response rates to interferon-alpha are 5 - 15%, but these are primarily partial responses and are of limited duration.^4,5 Although patients with metastatic disease who are treated with nephrectomy plus interferon have an improved survival over patients treated with interferon alone, all patients will ultimately have disease progression.⁶

Mutations of the von Hippel-Lindau (vHL) pathway occur in up to 70% of patients with renal cell carcinoma and our understanding of the downstream effects of this mutation has increased dramatically in recent years. vHL encodes an E3 ligase that triggers proteasome-mediated degradation of the α–subunits of the hypoxia inducible factors HIF-1, -2 and -3. Loss of vHL in RCC results in stabilization of HIFα, leading to increased production of multiple tumor-promoting factors including vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF). The target of rapamycin, mTOR, also participates in HIFα expression by regulating translation of HIFα mRNA.⁷ VEGF receptor, PDGF receptor and mTOR are the targets of recently approved agents which include sorafenib, sunitinib, and temsirolimus.^8-12 Upstream of mTOR, pAKT has been identified as a potential treatment target, and loss of the tumor suppressor PTEN has poor prognostic implications.¹³ Although the development of these targeted therapies represents a major advance in the treatment of RCC, they are still limited in that they result in a non-durable partial remission and are effective primarily in RCC of the clear cell subtype.

Clinical trials of 5-fluorouracil (5FU) and its oral analog, capecitabine, have shown modest activity for these agents in RCC.^14,15 Other studies have similarly shown modest activity of gemcitabine in RCC.^16,17 These trials led to a series of phase I/II studies testing the combination of oral capecitabine and weekly IV gemcitabine with single institution studies reporting response rates similar to that seen with the infusional 5-FU.^18,19,20 Subsequently, in order to confirm these results in a multi-institutional setting, concurrent phase II trials were developed within the Southwest Oncology Group (SWOG) and the Cancer and Leukemia Group B (CALGB). The CALGB trial was previously reported, demonstrating a modest response rate of 11% and a median survival of 14.5 months.²¹ However, the investigators concluded that the “associated toxicity would not support further evaluation in a Phase III trial of unselected patients”. Here we report the results of the SWOG trial employing a reduced dose-schedule of the gemcitabine-capecitabine doublet.

MATERIALS AND METHODS

Patients

Patients were eligible to enroll if they had histologically or cytologically confirmed renal cell carcinoma that was metastatic (M1) or with an unresectable primary tumor (M0). Patients were required to have measurable disease which was defined according to standard Response Evaluation Criteria in Solid Tumors (RECIST).²² Patients may have had 1 or 2 prior immunotherapy treatments with either interferon and/or interleukin-2. Prior radiation therapy was allowed provided treatment was to less than 25% of bone marrow. Patients may have had prior resection of their primary tumor provided they had adequately recovered from their surgery. No prior chemotherapy for RCC was allowed. Patients must have had a Zubrod performance status of 0-2 along with adequate hepatic, hematologic, and renal function. Pregnant or nursing women were excluded as were patients with brain metastases. Patients with any uncontrolled intercurrent illness (e.g., uncontrolled diabetes mellitus) including, but not limited to, ongoing or active infection, symptomatic congestive heart failure, unstable angina pectoris, cardiac arrhythmia, or psychiatric illness/social situations that would limit compliance with study requirements were not eligible. No other prior malignancy was allowed except for the following: adequately treated basal cell or squamous cell skin cancer, in situ cervical cancer, adequately treated Stage I or II cancer from which the patient was currently in complete remission, or any other cancer from which the patient has been disease-free for 2 years. All patients were informed of the investigational nature of this study and gave written informed consent in accordance with institutional and federal guidelines. The study protocol also requested that all patients submit a serum and white cell buffy coat specimen and an archival pathologic specimen for correlative studies.

Treatment Plan and Dose Modifications

Patients received 900 mg/m² gemcitabine intravenously on days 1, 18, and 25 and were to take 625 mg/m² oral capecitabine twice a day (for a total of 1,250 mg/m²/day) on days 1-21. All capecitabine doses were calculated on the basis of milligrams of drug per square meter of body surface area with the dose rounded upward to the closest whole tablet combination of 500 and 150 mg tablets. Treatment doses were based on prior phase I studies and preliminary results of the CALGB study which suggested significant toxicity at higher dose levels.²¹ Treatment cycles were every 28 days and patients continued treatment until disease progression. Study requirements also included prophylactic antiemetic therapy of ondansetron 8 mg IV or equivalent plus dexamethasone 10 mg IV or orally plus prophylaxis of the hand/foot syndrome with pyridoxine. Toxicities were graded using the National Cancer Institute Common Toxicity Criteria version 2.0. Since there is a potential interaction between capecitabine and coumarin-based anticoagulants, patients requiring therapeutic concomitant coumarin-based anticoagulants were eligible but required monitoring of INR at more frequent intervals. Prophylactic G-CSF was not allowed and patients with grade 4 neutropenia underwent a dose reduction.

Dose adjustments were made for selected Grade 3 and 4 toxicities according to the system showing the greatest degree of toxicity and the suspected etiology agent (i.e. gemcitabine dose was not decreased for the hand/foot syndrome). Dose level -1 and -2 adjustments for gemcitabine were 720mg/m² and 575mg/m² and for capecitabine were 1000mg/m² and 750 mg/m² respectively. Treatment could be delayed no more than three weeks to allow recovery from toxicity or the patient was removed from protocol treatment.

Study Endpoints and Patient Follow-up

The primary endpoint was tumor response. Response was assessed using standard RECIST criteria.²⁰ Patients were assessed for response after every second treatment cycle (every 8 weeks). Additional assessments at least 4 weeks apart were used for confirmation of response. Secondary endpoints were 6-month time to treatment failure rate, overall survival, and toxicity (National Cancer Institute Common Toxicity Criteria version 2.0). Another secondary endpoint was an examination of the relationship between biologic markers (e.g. p53, PTEN, pAKT, pmTOR, vHL, ERCC1) and RECIST response in the subset of patients with available pathologic specimens.

Patients were removed from protocol therapy for disease progression, intercurrent illness that prevented further administration of protocol treatment, or unacceptable toxicity. Study participants could withdraw from the study at any time for any reason or for changes in the patient's condition that rendered the patient unacceptable for further treatment in the judgment of the treating physician. Patients were to be followed for a maximum of 3 years after registration.

Statistical Methods

The primary endpoint of this study was probability of response (confirmed and unconfirmed complete and partial responses). A one-stage design was used for patient accrual due to the experience of rapid accrual to studies of renal cell cancer in SWOG. The chemotherapy regimen described would have been deemed worthy of further study if the observed data were consistent with a true probability of response of >=20%, and the regimen would not have been considered of interest if the observed data were consistent with a true probability of response of<=5%. With 38 patients, five or more observed responders were needed to warrant further study of this treatment provided toxicity was acceptable. This design had a power of 90% and a significance level of 0.04 (using a one-sided test).

Secondary objectives were to examine the 6-month probability of treatment failure, overall survival, and a toxicity profile for this regimen. Time to treatment failure was defined as the time to progression, death, symptomatic deterioration, or early discontinuation of treatment. Patients who were not known to have failed treatment were censored at the date of last contact. For the survival endpoint, patients who were not known to have died were censored at the date of last contact. Kaplan-Meier estimates were used to calculate the 6-month probabilities of treatment failure and overall survival. With 38 patients, the probability of occurrence of any specific toxicity, 6-month treatment failure probability, and 6-month survival probability could be estimated to within 16% (95% confidence interval).

In a subset of patients with available pathologic specimens, we evaluated possible associations between biologic markers and RECIST response. Assessment of TS gene polymorphisms and ERCC1 expression were specified in the initial protocol. Additional studies were also conducted to examine the percentage of cells expressing PTEN, pAKT, pmTOR, as well as mutations in the vHL gene. ERCC1 expression was reported as the Mean Normalized Expression.²³

Immunohistochemistry

Pretreatment archival tumor tissue from consenting patients was evaluated by a pathologist blinded to the clinical data. Tumor was confirmed on histologic sections and each immunostained slide scored for reactivity by light microscopy. Reactivity was determined by the percentage of positive tumor cells. Immunohistochemistry was by standard techniques previously reported.^24,25 Previously characterized cell lines were formalin fixed and paraffin embedded for antibody validation and use as IHC positive and negative controls. For p-mTOR (ser 2448), LNCaP cells treated with or without rapamycin were used as controls (antibody: 2976, Cell Signaling, Danvers, MA); for PTEN staining, MCF7 cells were used as a positive control and LNCaP cells as a negative control (antibody: 138G6 9559, Cell Signaling); for pAKT (ser 473), LNCaP cells untreated or treated with LY294002 (antibody: 4051, Cell Signaling).

VHL Mutation Analysis

DNA was extracted from paraffin-embedded tissue following manual microdissection using the QIAamp DNA Mini Kit (Qiagen, Valencia, Ca). Four different PCR reactions were used to amplify VHL exons 1-3, with two separate groups of primers used to amplify exon 1, one pair (1A) includes the first 53 codons while the second group (1B) encompasses the second initiation site. For each exon, 4.0 ul of DNA was amplified in a 24 ul reaction that included 1xPCR buffer, 20 uM dNTP, 0.3 uM primers and 2.0 U of Cloned Pfu polymerase (Stratagene, Cedar Creek, Texas). For exon 1B and exon 2, 3% dimethyl sulfoxide was also used. PCR conditions for each exon were as previously reported.²⁶ Following initial amplification, 2.0 ul of PCR product was re-amplified under the same conditions. PCR products were run on a 2% agarose gel and stained with ethidium bromide. Bands were purified with the QIAquick Gel Extraction Kit (Qiagen, Valencia, Ca), and sequenced using an ABI 3730 Capillary Electrophoresis Genetic Analyzer ABI with BigDye Terminator v3.1 Cycle Sequencing (Applied Biosystems, Foster City, Ca). All PCR reactions were repeated and subsequently sequenced to verify results.

RESULTS

Accrual and Data Submission

Forty-three patients from 12 SWOG institutions were registered to this study. One patient was ineligible due to having an interferon treatment within 28 days prior to registration. Two eligible patients are not analyzable and were not included in this analysis. The first patient was not given any protocol treatment, and the second patient refused protocol treatment. Thus, forty eligible patients were available for the analysis.

Patient Characteristics

Patient characteristics are listed in Table 1. Males accounted for 80% of the patients, and the median age was 61.3 years (range: 36.0 to 80.3 years). The racial distribution was 90% white, 7% African-American, and 3% Native American. One patient (3%) was recorded as being of Hispanic origin with this status recorded as unknown in five other patients (12%). The predominant histologic subtype was clear cell carcinoma in 33 patients (82%), with two patients (5%) having a papillary carcinoma, one patient (3%) having adenocarcinoma, and four patients (10%) having other (e.g. sarcomatoid) or unknown histology. Thirty patients (75%) had a prior nephrectomy and twenty-two (55%) had prior immunotherapy.

TABLE 1.

Patient Characteristics

	Gemcitabine + Capecitabine
No. Eligible Patients	40
Median patient age (range)	61.3 yrs (36.0, 80.3)
No. Sex (%)
Male	32 (80)
Female	8 (20)
No. Race (%)
White	36 (90)
African-American	3 (7)
Native American	1 (3)
No. Hispanic Origin (%)
Yes	1 (3)
No	34 (85)
Unknown	5 (12)
No. Performance Status (%)
0	17 (42)
1	21 (53)
2	2 (5)
No. Prior Therapy (%)
Immunotherapy	22 (55)
Nephrectomy	30 (75)
Radiotherapy	3 (7)
No. Tumor Histology (%)
Clear Cell	33 (82)
Papillary	2 (5)
Adenocarcinoma	1 (3)
Other (e.g. sarcomatoid)	3 (7)
Unknown	1 (3)
No. of Metastatic Sites (%)
1	6 (15)
2	12 (30)
≥3	22 (55)

Efficacy

Median follow-up time was 14.9 months (range: 1.3 to 27.3). One confirmed complete response (CR) and three unconfirmed partial responses (UPR) were observed, for an overall response probability of 10% (95% CI: 3% to 24%). Four patients submitted disease assessment data that were inadequate for response assessment. These patients were assumed to be non-responders and were included in the calculation of the response probability. Nineteen patients (48%) had stable disease. One patient with an UPR had papillary carcinoma and all other responders had clear cell carcinoma.

The 6-month probability of being free from treatment failure was 20% (95% CI: 8% to 32%). The probability of survival at 6 months was 75% (95% CI: 62% to 88%). Median survival time was estimated to be 23 months (95% CI: 10, 37). These data are summarized in Figures 1 and 2.

Figure 1.

Time to Treatment Failure for eligible patients.

Figure 2.

Overall Survival for eligible patients.

Toxicity

Common toxicities are listed in Table 2. There were no Grade 5 toxicities among patients on this study. Four patients (10%) were described with Grade 4 toxicities. One patient experienced Grade 4 lymphopenia, leukopenia, and neutropenia which were judged to be definitely related to treatment. Grade 4 fatigue was reported for another patient (possibly related to treatment). A third patient experienced Grade 4 thrombocytopenia (definitely related to treatment), and a fourth patient was reported as having Grade 4 hemolysis and renal failure (both possibly related to treatment).

TABLE 2.

Selected Adverse Events (Grade 2-4) on S0312

Adverse Event	Grade 2 (Moderate)#	Grade 3 (Severe)	Grade 4 (Life threatening)
	Number of patients (%)	Number of patients (%)	Number of patients (%)
Blood/Bone Marrow
Hemoglobin	12	3	0
Leukocytes	10	4	1
Lymphopenia	8	0	1
Neutrophils	8	12	1
Platelets	4	0	1
Hemolysis	0	0	1
Cardiovascular
Cardiac Ischemia/infarction	0	1	0
Thrombosis/embolism(vascular accident)	1	0	0
Thrombosis/embolism	1	1	0
Phlebitis	2	0	0
Constitutional Symptoms
Fatigue	16	5	1
Weight Loss	2	0	0
Sweating	2	0	0
Fever	3	0	0
Insomnia	2	0	0
Dermatology/Skin
Rash: hand-foot skin reaction	5	2	0
Gastrointestinal
Diarrhea	4	1	0
Anorexia	8	0	0
Nausea	4	2	0
Vomiting	4	2	0
Mucositis (Oral)	1	1	0
Dehydration	2	0	0
Heartburn	1	1	0
Metabolic/Laboratory
AST	2	0	0
Hyperglycemia	1	1	0
Hypoalbuminemia	5	0	0
Hypokalemia	0	2	0
Neurology/Musculoskeletal
Neuro Pain: head/headache	1	1	0
Musculoskeletal-other	0	1	0
Pulmonary
Cough	2	0	0
Lung Pain: chest/thorax	2	0	0
Dyspnea	0	1	0
Renal
Renal failure	0	0	1
Maximum grade of any adverse event
Number	9	26	4

^#Grade 2 toxicities with only one reported adverse event and no grade 3 or 4 events were excluded.

Grade 3 toxicities were observed in 25 of the 40 (62.5%) of the eligible patients on study. The most common Grade 3 toxicities were neutropenia (12 patients), fatigue (5 patients), and leukopenia (4 patients). Six patients were removed from study treatment secondary to toxicity.

Correlative Studies

Twenty-six of the eligible patients consented to submit a pathologic specimen for analysis of tumor markers. Of these patients, 8 (31%) had a best response of progression, 16 (62%) had a best response of stable disease, and 2 (7%) were unconfirmed partial responders. Results from the correlative studies are summarized in Table 3.

Table 3.

Correlative studies for 25 available pathologic samples from patients on S0312, by categories of RECIST response

	RECIST best response
Tumor Marker	Progressive Disease (n(%))	Stable or better (n(%))
PTEN (% cells expressing)*
0	3/6 (50)	13/14 (93)
>0	3/6 (50)	1/14 (7)
pAKT (% cells expressing)*
0	5/6 (83)	13/14 (93)
>0	1/6 (17)	1/14 (7)
pmTOR (% cells expressing)*
0	4/6 (67)	4/14 (29)
>0	2/6 (33)	10/14 (71)
vHL**
mutation in Exon 1	2/4 (50)	3/9 (33)
mutation in Exon 2	0/4 (0)	2/9 (22)
mutation in Exon 3	1/4 (25)	0/9 (0)
TS Polymorphism***
-6/-6	1/6 (17)	3/15 (20)
+6/-6	3/6 (50)	8/15 (53)
+6/+6	2/6 (33)	4/15 (27)
ERCC1 Mean Normalized Expression
Minimum	0.00130	0.00043
Q1	0.00143	0.00087
Median	0.00199	0.00288
Q3	0.00348	0.00385
Maximum	0.00352	0.00858

^*sample viable for testing in 20 of 26 patients

^**sample viable for testing in 13 of 26 patients (see text)

^***sample viable for testing in 21 of 26 patients

****sample viable for testing in 23 of 26 patients. Mean Normalized Expression (MNE) calculated using method outlined by Simon ²⁵. MNE is a ratio comparing ERCC1 to a reference gene (beta-actin).

Expression of pTEN, and phosphorylation levels of AKT and mTOR were assessed by IHC. Although limited by the small number, half of the patients (3/6) with progressive disease showed expression of PTEN with percent tumor cell positivity ranging from 60% to 100%, compared to 7% of the patients (1/14) with stable disease or better (a single patient with 20% positivity). Tumor pTEN staining was exclusively cytoplasmic, and in contrast to tumor cells, blood vessel endothelium showed marked positivity in 55% of patients. Three tumors were of the papillary type, two of which showed strong pTEN staining. Of patients with progressive disease, 33% showed nuclear immunostaining of phosphorylated mTOR (pmTOR) compared to 71% of patients with stable disease or better. Expression of pAKT was only observed in 2 of 20 patients. The majority of patients (10/14) with stable disease or better were both pTEN negative and pmTOR positive; whereas 0/6 patients with progressive disease had this combination of markers (p=0.01). Representative images of PTEN, and pmTOR and pAKT staining are shown in figure 3.

Figure 3.

Representative IHC staining of PTEN and phospho-mTOR

Additional studies were conducted in manually microdissected tumor sections to identify mutations in the vHL gene and examine RNA levels of ERCC1, XRCC1 and TS. Polymorphisms in the TS gene were also examined in PBMCs. In 13 patients with sufficient material for sequencing, 7 were positive for vHL mutations in exons 1, 2 or 3. One patient harbored mutations in both exons 1 and 3. The absolute numbers of patients with vHL mutation were small in both groups. No significant differences were observed in ERCC1 or XRCC1 Mean Normalized Expression between patients with disease control versus those that progressed. ERCC1 had a wider range in patients with stable response or better compared to those with progressive disease. The distribution of polymorphisms in the TS gene is shown in Table 3.

DISCUSSION

The results of this multi-institutional Phase II study of gemcitabine and capecitabine in advanced RCC further confirms results of the prior multi-center trial using of this drug combination with responses rates of 10% and 11%, respectively.²⁰ The median survival in the present study was 23 months compared to 14.5 months in the prior study; both of which are encouraging when compared prior multi-centered trials evaluating cytotoxic agents in advanced RCC. The 10% objective response rate observed in the present study, although greater than the null hypothesis of <5%, was not consistent with a true response rate of >20%, indicating the regimen is not worthy of further study in the phase III setting. Nonetheless, in our study 8 patients (20%) remained on treatment for greater than six months suggesting a small cohort of patients do derive benefit from this combination, a finding consistently present in all reported studies of gemcitabine and 5-FU analogs in RCC.^18-21

The evaluation of the biologic markers was preformed in an effort to further identify the group of patients that are most likely to benefit. TS polymorphisms are well described predictors of response to treatment and toxicity in gastrointestinal malignancies²⁷ and prior studies have suggested a higher level of TS activity correlates with renal cell carcinoma progression.²⁸ In this preliminary analysis, however, we found no correlation between TS polymorphisms and disease response. Secondly, we assessed the expression of ERCC1, a DNA-repair gene which has been evaluated in other malignancies using cisplatin, 5-FU and gemcitabine-based regimens. In one series assessing patients with non-small lung cancer who had received combination therapy with cisplatin and gemcitabine, the median survival was significantly prolonged in patients with lower levels of expression of ERCC1.²⁹Similar findings were seen in a study evaluating the efficacy of 5-FU and oxaliplatin in gastric cancer.³⁰ In this limited study, no correlation was observed between patient outcome and ERCC1 levels.

Patients with higher levels of pmTOR and absence of PTEN expression were more likely to derive benefit as evidence by a best response of stable disease or better. PTEN functions as a tumor suppressor gene, regulating cycle progression and cell survival through repression of AKT signaling. In RCC, reduced PTEN expression correlates with a higher proliferative rate.³¹ Although it requires confirmation in a larger series, our results suggest that the loss of PTEN expression, and the resultant cell cycle and proliferative effects, may increase tumor susceptibility to cytotoxic therapy. Potentially, cycling cells may be more vulnerable to nucleoside analogs such as gemcitabine and capecitabine which are most effective in S-phase. Of note, two of three tumors with papillary histology expressed PTEN. The status of PTEN and phospho-mTOR may have utility in assigning patients to chemotherapy versus targeted therapies.

The dose levels for this study (900 mg/m² gemcitabine IV on days 1, 18, and 25 and 625 mg/m² oral capecitabine twice a day for days 1-21) were determined after the initial reports from the CALGB study suggested significant treatment related toxicity.²¹ At the dose levels for our study, the gemcitabine and capecitabine combination was overall well tolerated and with minimal grade 4 and no grade 5 reported toxicities. The most common grade 3 toxicities were related to bone marrow suppression. Grade 2-3 fatigue was also a commonly reported adverse event but the frequency of the hand/foot syndrome was significantly less than the prior reported studies.^20,21 In addition, the response rate, stable disease rate, and median survival rate did not appear to be significantly compromised by the selected dosing schedule of the present study.

Although at present the FDA approved therapies targeting VEFGR, PDGRF, and mTOR for RCC, sunitinib, sorafenib and temsorlimus have encouraging response and/or progression-free survival rates, response rates are predominantly partial and are of a limited duration.^10,11 The combination of traditional chemotherapy with VEFGR, PDGFR, and epidermal growth factor receptor(EGFR)-targeted therapies is being successfully pursued in multiple solid tumors malignancies, with these combined approaches having indications which are becoming the standard of care in malignancies such as colon and non small cell lung cancer.^22-24

Finally, although the results of this study do not warrant testing in the phase III setting, this regimen provides benefit to selected patients and should also be consider as a base regimen for further combination therapies with VEGFR, PDGFR or mTOR inhibitors and/or as a treatment option for patients who have failed these targeted agents.