A Phase II Study Evaluating the Safety and Efficacy of Sunitinib Malate in Combination with Weekly Paclitaxel Followed by Doxorubicin and Daily Oral Cyclophosphamide Plus G-CSF as Neoadjuvant Chemotherapy for Locally Advanced or Inflammatory Breast Cancer

Abstract

INTRODUCTION

Neoadjuvant chemotherapy is standard treatment for locally advanced (LABC) or inflammatory breast cancer (IBC). We hypothesized adding sunitinib, a tyrosine kinase inhibitor with anti-tumor and anti-angiogenic activity, to an anthracycline and taxane regimen would improve pathologic complete response (pCR) rates to a prespecified endpoint of 45% in patients with HER2 negative LABC or IBC.

PATIENTS AND METHODS

We conducted a multicenter, phase II trial of neoadjuvant sunitinib with paclitaxel (S+T) followed by doxorubicin and cyclophosphamide plus G-CSF for patients with HER2 negative LABC or IBC. Patients received sunitinib 25 mg PO daily with paclitaxel 80 mg/m² IV weekly x12 followed by doxorubicin 24 mg/m² IV weekly + cyclophosphamide 60 mg/m² PO daily with G-CSF support. Response was evaluated using pCR in the breast and the Clinical-pathologic scoring + estrogen receptor (ER) and grade (CPS+EG) score.

RESULTS

Seventy patients enrolled and 66 were evaluable for efficacy. Eighteen patients (27%) had pCR in the breast (10 had ER+ disease and 8 had triple negative disease). When defining response as pCR and/or CPS+EG score ≤ 2, 47% were responders. In ER+ patients, 23 (64%) were responders. The most common toxicities were cytopenias and fatigue.

CONCLUSIONS

Neoadjuvant S+T followed by AC+G-CSF was safe and tolerable in LABC and IBC. The study did not meet the prespecified endpoint for pCR. However, 47% were responders using pCR and/or CPS+EG score ≤2. ER+ patients had the highest response rate (64%). The addition of sunitinib to neoadjuvant chemotherapy may provide promising incremental benefit for ER+ patients.

MICROABSTRACT

Pathologic complete response (pCR) following neoadjuvant chemotherapy is associated with improved survival in locally advanced and inflammatory breast cancer. Neoadjuvant sunitinib + paclitaxel followed by doxorubicin + cyclophosphamide with G-CSF therapy resulted in a 27% pCR rate in this single-arm, phase II trial. ER+ patients had higher response rates (64%) using the CPS+EG score and pCR suggesting promising incremental benefit.

INTRODUCTION

Neoadjuvant chemotherapy remains the treatment of choice for locally advanced breast cancer (LABC) and inflammatory breast cancer (IBC). Unfortunately, pathologic complete response (pCR), which is associated with longer disease-free survival (DFS) and overall survival (OS), is attained only in a minority of patients treated with standard anthracycline and taxane regimens.^1–5 This highlights the need for more effective treatment strategies.

Incorporating anti-angiogenic biologic agents into breast cancer therapy has demonstrated improvements in response rate and DFS when compared to chemotherapy alone.^6–8 Sunitinib malate is an oral, multi-target tyrosine kinase inhibitor with anti-tumor and anti-angiogenic activity that is approved for treatment of metastatic renal cell carcinoma (RCC), advanced pediatric neuroendocrine tumors, and gastrointestinal stromal tumors (GIST) with progression or intolerance on imatinib.⁹ It functions as a potent broad specificity kinase inhibitor targeting VEGFR-1, 2, 3, FLT3, KIT, PDGFRα, and PDGFRβ.¹⁰ Activity against VEGFR-3 and dysregulation of pericyte homeostasis through interference of the PDGF/PDGFR pathway are potential anti-angiogenic activities which are unique to sunitinib, as compared to other angiogenesis inhibitors such as bevacizumab.¹¹ A multi-targeted approach has several proposed benefits including improved targeting of orthogonal tumor growth pathways and decreased potential for resistance development.¹² Moreover, preclinical models with dual inhibition of VEGF and PDGF signaling pathways were found to be synergistic in inhibiting tumor growth.^13,14

Preliminary results from an earlier trial demonstrated that continuous daily cyclophosphamide and weekly doxorubicin (AC) plus scheduled G-CSF followed by paclitaxel as initial therapy in LABC and IBC was superior to standard, intermittent AC and with significant improvement in pCR (31% versus 19%, p=0.02).¹⁵ Though the final results ultimately showed no significant advantage in pCR in the investigational arm³, the initial results prompted the design for this trial and we hypothesized that adding sunitinib to a similar regimen may further improve pCR. Based upon data showing decreased tumor interstitial fluid pressure and improved oxygenation in patients treated with neoadjuvant weekly paclitaxel followed by dose dense doxorubicin and cyclophosphamide¹⁶ and the potential for dose-limiting neutropenia with sunitinib in combination with doxorubicin, we combined sunitinib (S) with weekly paclitaxel (S+T) prior to continuous AC. Phase I studies of combination S+T previously demonstrated this regimen was well tolerated in patients with locally advanced or metastatic breast cancer and there was preliminary evidence of antitumor activity.^12,17

Herein we report outcomes regarding safety and efficacy of daily sunitinib in combination with weekly paclitaxel followed by weekly doxorubicin and daily cyclophosphamide plus G-CSF as neoadjuvant therapy for hormone receptor positive or negative, HER2 negative, LABC and IBC.

PATIENTS AND METHODS

Patient eligibility

Eligible patients had histologically confirmed, locally advanced or inflammatory HER2 negative breast cancer. Locally advanced was defined as clinical stages IIB, IIIA, IIIB, and IIIC or disease which was judged to be primarily unresectable by an experienced breast surgeon. Inflammatory breast cancer was defined by clinical criteria with erythema and peau d’orange of the breast with a histologic diagnosis of breast cancer. Additionally, patients were required to have an adequate performance status, adequate renal and kidney function, no significant neutropenia or thrombocytopenia, and normal ejection fraction measured within 3 months prior to enrollment.

Patients were excluded if they had any evidence of distant metastases, or, if they had received any prior chemotherapy/endocrine therapy, radiation therapy or definitive surgery for breast cancer. Additional exclusion criteria included significant medical comorbidities such as heart disease or a cerebrovascular event and contraindications to sunitinib including uncontrolled hypertension or thyroid abnormality.

This multicenter study was conducted at University of Washington/Seattle Cancer Care Alliance (SCCA), SCCA Network sites, and University of Arizona Cancer Center. Patients were enrolled between September 2007 and February 2012. All patients signed informed consent prior to any study-related procedures. This trial is registered with ClinicalTrials.gov, NCT00513695.

Drug administration

Patients were treated with paclitaxel 80 mg/m² IV weekly and sunitinib 37.5 mg daily PO for 12 weeks. The dose of sunitinib was reduced to 25 mg from 37.5 mg in March 2008 by protocol amendment based on preliminary data from a phase I study of S+T which showed increased myelosuppression and neutropenia.^17,18 Five patients received the 37.5 mg dose prior to the protocol amendment. After 12 weeks patients without significant toxicity that precluded further treatment then received doxorubicin 24 mg/m² IV weekly and cyclophosphamide 60 mg/m² PO daily for 15 weeks combined with growth factor support of G-CSF 5 ug/kg SC on days 2–7 for 15 weeks (Figure 1). Prior to starting this treatment, ANC had to be greater than 1,500/mm³ and platelets had to be greater than 100,000/mm³. Patients received pneumocystis carinii pneumonia prophylaxis with trimethoprim/sulfamethoxazole 1 single strength tablet daily for 15 weeks during doxorubicin and cyclophosphamide therapy per standard practice at the time.^19–24 After completion of neoadjuvant treatment patients were offered definitive surgical resection.

FIGURE 1: Treatment Schema.

Schema for a Phase II trial of daily sunitinib in combination with weekly paclitaxel followed by weekly doxorubicin and daily cyclophosphamide plus G-CSF as neoadjuvant therapy, NCT00513695

Toxicity

Adverse events (AEs) were evaluated at each study visit for the duration of the trial. During investigational treatment laboratory toxicities were assessed at each infusion and clinical toxicity was assessed at each visit. Following completion of treatment, patients were clinically evaluated no less than quarterly for 2 years, then every 6 months for three years, and annually thereafter. For treatment related toxicity and AE reporting, the study used the NCI CTC (Common Terminology Criteria) Version 3.0 (http://ctep.cancer.gov/protocolDevelopment/electronic_applications/docs/ctcaev3.pdf). Doses were adjusted or delayed per protocol according to the system showing the greatest degree of treatment related toxicity. AEs that meet severity grade 2 or greater were collected and reported.

Clinical Response:

The primary endpoint was pCR which was defined as no evidence of microscopic invasive tumor present in the surgical specimens including the primary tumor site (breast). Ellis et al. previously reported a pCR rate of 31% for patients treated with continuous doxorubicin and cyclophosphamide followed by paclitaxel in patients with inflammatory and locally advanced breast cancer.¹⁵ The pre-specified endpoint was selected based on this data and we hypothesized that incorporating sunitinib into this regimen would improve pCR to a rate of 45%. In addition to assessing pCR in the breast at the time of surgery, the MDACC Clinical-pathologic scoring +ER and grade (CPS+EG) was calculated. This scoring system incorporates pretreatment clinical stage, grade, and ER status with the final pathologic stage to more accurately stratify patients according to prognosis after administration of neoadjuvant chemotherapy.^25,26 There is robust data to support the use of the CPS+EG score which has been validated multiple times both in internal and external patient cohorts.^25–35

Secondary endpoints included clinical complete response, relapse rate, DFS, and OS. Prior to neoadjuvant chemotherapy, axillary nodes were assessed by sentinel node biopsy or image-guided biopsy. The optimal imaging method (mammogram, ultrasound or breast MRI) for assessing tumor volume was determined pre-study and then repeated every 4 – 8 weeks during investigational treatment. Radiologic test during the follow-up period was at the discretion of the treating physician.

Statistical Analysis

The study sample size of 70 patients (anticipating 64 evaluable) was determined to have 80% power to identify a 45% pCR rate as superior to a 30% null rate¹⁵, with a one-sided Type I error of 0.05. Baseline demographic and clinical characteristics were evaluated. For continuous variables, the median and range were calculated; for categorical variables, the number and percentage were calculated. For AE, the number of subjects under each AE category was summarized by different grades and by total, respectively.

For survival analysis, OS was defined as time from day 1 of protocol treatment to death and DFS was defined as time from day 1 of protocol treatment to disease progression or death whichever occurred first. DFS and OS were evaluated between different groups (e.g., CPS+EG score ≥ 3 vs CPS+EG score <3 vs indeterminate score) using Kaplan-Meier curves and Log-Rank test.

RESULTS

Patient and disease characteristics

From September 2007 to February 2012, a total of 70 patients provided informed consent and were enrolled from academic centers and community network oncology practices. Three patients were screen fails and 67 patients received protocol directed therapy. Patient and disease characteristics are shown in Table 1. Tumor characteristics were taken from diagnostic breast tumor biopsy prior to enrollment. Of the 67 patients treated, 59 (88%) had infiltrating ductal carcinoma and 5 (7%) had infiltrating lobular carcinoma. There were 37 patients with ER positive and/or PR positive disease and 30 patients with TNBC. Six patients had IBC of which 3 had ER positive and/or PR positive and 3 had TNBC.

TABLE 1:

Patient Characteristics

	Overall (N=67)
Age
Median (Range)	50 (33, 79)
Menopausal Status
Premenopausal	34 (51%)
Perimenopausal	3 (4%)
Menopausal	26 (39%)
Missing	4 (6%)
Study Site
UWMC/SCCA	31 (46%)
SCCA Network	8 (12%)
UA/Arizona Cancer Center	28 (42%)
Histology
Ductal	59 (88%)
Lobular	5 (7%)
Other	3 (4%)
Hormone Receptor Status
ER and/or PR positive	37 (55%)
TNBC	30 (45%)
Ki67
Median (range)	50 (5, 100)
Missing	16
Grade
1	2 (3%)
2	13 (19%)
3	43 (64%)
Missing	9 (13%)
Clinical Stage
IIB	9 (13%)
IIIA	40 (60%)
IIIB	10 (15%)
IIIC	8 (12%)
Inflammatory Breast Cancer	6 (9%)
ER and/or PR positive	3 (4%)
TNBC	3 (4%)

Selected patient and tumor characteristics from 67 patients

Safety analysis included all 67 patients who started therapy per protocol. The efficacy analysis included 66 patients as 1 patient withdrew consent after completing all therapy. Of the 66 patients, there were 61 patients who had surgery and were evaluable for pathologic response. The remaining 5 patients who did not complete surgery were considered non-responders for the primary endpoint of pCR. Of these 5 patients, 3 patients had progressive disease, 1 patient went off study due to toxicity, and 1 patient completed all therapy but elected not to have surgery.

Toxicity profile

Grade 2 or higher events as reported using CTCAE 3.0 were observed in 64 patients (96%) during S+T and 51 (88%) during AC+G-CSF (Tables 3 and 4). The most common toxicities of any grade during S+T included neutropenia (52 events), leukopenia (44), fatigue (22), and anemia (17). Seven patients developed grade 2 neuropathy. The most common toxicities of any grade during AC+G-CSF included leukopenia (22 events), neutropenia (21), anemia (19), mucositis (15), fatigue (14), and nail changes (14). No grade 5 toxicities were reported. These toxicities were similar in distribution and severity to other reports evaluating these agents and we report no new or unexpected toxicities.

TABLE 3:

Toxicity During Sunitinib+Paclitaxel

Adverse Events during S+T	Grade 2	Grade 3	Grade 4	Total
Neutropenia	17	28	7	52
Leukopenia	26	17	1	44
Fatigue	19	3	0	22
Anemia	14	3	0	17
Diarrhea	6	4	0	10
Mucositis	6	3	0	9
Hypertension	7	0	0	7
LFT abnormality	6	1	0	7
Neuropathy	7	0	0	7
Rash	6	1	0	7
Allergic reaction	5	1	0	6
Heartburn	6	0	0	6
Pain	5	1	0	6
Infection	3	2	0	5
Nausea/Vomiting	5	0	0	5
Alopecia	4	0	0	4
Mood alteration	4	0	0	4
Thrombocytopenia	3	1	0	4
Weakness	4	0	0	4
Hot flashes	3	0	0	3
Nail changes	3	0	0	3
Anorexia	2	0	0	2
Fever	0	2	0	2
Hand-foot	1	1	0	2
Lymphopenia	1	1	0	2
Taste alteration	2	0	0	2
Nose bleed	0	1	0	1
Total Events	165	70	8	245
Total Patients	58 (87%)	42 (63%)	8 (12%)	64 (96%)

Most common adverse events during sunitinib plus paclitaxel phase. There were 245 total grade 2 or higher events that occurred in 64 (96% of patients). Neutropenia was most common grade 3/4 event and G-CSF 5 mcg/kg SC permitted. There were no grade 5 events.

TABLE 4:

Toxicity During AC+G-CSF

Adverse Events during AC	Grade 2	Grade 3	Grade 4	Total
Leukopenia	13	6	3	22
Neutropenia	7	10	4	21
Anemia	15	4	0	19
Mucositis	6	9	0	15
Fatigue	13	1	0	14
Nail changes	12	2	0	14
Nausea/Vomiting	10	1	0	11
Hand-foot	3	3	0	6
Alopecia	5	0	0	5
Infection	5	0	0	5
Mood alteration	5	0	0	5
Pain	5	0	0	5
Constipation	4	0	0	4
Lymphopenia	3	0	1	4
Neuropathy	3	1	0	4
Hot flashes	3	0	0	3
Diarrhea	2	0	0	2
Heartburn	2	0	0	2
Irregular menses	2	0	0	2
Nasal/paranasal reactions	2	0	0	2
Total Events	120	37	8	165
Total Patients	50 (86%)	27 (47%)	5 (9%)	51 (88%)

Most common adverse events during doxorubicin and oral cyclophosphamide plus G-CSF phase. There were 165 total grade 2 or higher events which occurred in 51 (88%) of patients. Leukopenia was most common grade 3/4 event. There were no grade 5 events.

The median dose delivery by percent for sunitinib was 100% (range 54–115) based on the value after the dose reduction. The median dose for paclitaxel by percent was 100% (range 50–100). A total of 42 (63%) patients required dose modifications or a hold during the course of S+T, and 36 (62%) patients required dose modifications or a hold during the course of AC+G-CSF (Table 2). The most common reasons for dose modifications/holds during S+T included neutropenia (25 events), leukopenia (7), and diarrhea (3). Other reasons for dose modifications/holds during S+T included anorexia (2), fatigue (2), mucositis (2), rash (2), fever (1), hypertension (1), LFT abnormalities (1), epistaxis (1), hand-foot syndrome (1), and thrombocytopenia (1). The most common reasons for dose modifications/holds during AC+G-CSF included neutropenia (10 events), mucositis (6), leukopenia (6), and hand-foot syndrome (3). Other reasons for dose modifications/holds during AC+G-CSF included anemia (2), alopecia (1), constipation (1), infection (1), nail changes (1), neuropathy (1), and pain (1).

TABLE 2:

Dose Delivery

	S+T Phase (N=67)	Post S+T (N=64)	AC Phase (N=58)	Post AC (N=54)
Treatment summary
Completed without modification	22 (33%)		18 (31%)
Completed with modifications	42 (63%)		36 (62%)
Off study during of S+T	3 (4%)
Off study after completing S+T		6 (9%)
Off study during AC			4 (7%)
Off study after completing AC				4 (7%)
Reason for discontinuation
Progressive disease	1 (33%)	3 (50%)	0 (0%)	2 (50%)
Toxicity (by protocol specifications)	2 (67%)	0 (0%)	1 (25%)	0 (0%)
Patient choice	0 (0%)	1 (17%)	3 (75%)	1 (25%)
MD decision	0 (0%)	1 (17%)	0 (0%)	0 (0%)
Other	0 (0%)	1 (17%)	0 (0%)	1 (25%)

Dose delivery of neoadjuvant therapy. G-CSF 5 mcg/kg days 2–7 each wk administered per protocol during AC+G-CSF period. Sunitinib (S) dose was reduced from 37.5 to 25 mg PO daily by protocol amendment March 2008.

Of the 67 patients who received protocol therapy, a total of 3 patients discontinued therapy during treatment with S+T (2 due to toxicity and 1 due to progressive disease). Another 6 patients went off therapy after completion of all S+T, but prior to receiving AC (3 due to progressive disease, 1 due to patient choice, 1 due to physician choice, and 1 due to a femur fracture unrelated to treatment). A total of 4 patients went off study during AC (3 due to patient wishes and 1 due to toxicity). There were 4 patients who went off study after completion of all therapy, but prior to surgery (2 due to progressive disease, 1 due to patient wishes, and 1 patient withdrew consent). In the 3 patients who discontinued treatment due to toxicity, the AE resulting in termination of study treatment included hyponatremia, infusion reaction to paclitaxel during S+T, and neutropenia during AC+G-CSF.

Efficacy data

Of the 66 patients in the efficacy cohort, 18 (27%) had pCR in the breast and 15 (23%) had pCR in the breast and axilla, with similar pCR rates for patients with ER/PR+ disease and TNBC (Table 5, Chi-square test of independence p=0.99 for both). None of the 6 patients with IBC had a pCR. Patients were also evaluated with CPS+EG scores (Table 6). Twenty patients (30%) had a CPS+EG scores ≤ 2 and 38 (58%) had a CPS+EG score ≥ 3. There were 8 patients for whom CPS+EG was not available (3 left the study before surgery due to progressive disease, 2 left due to patient choice, 1 left due to toxicity, and 2 had insufficient information to calculate scores). Of the patients with indeterminant scores, 7 had TNBC disease and 1 had ER/PR+ disease. Of ER/PR+ patients, 18 (50%) had CPS+EG scores ≤ 2. Two (7%) patients with TNBC had a score ≤ 2. Responders were defined as patients with pCR in the breast and/or CPS+EG score ≤ 2. Overall, 31 (47%) patients were responders. Within the ER/PR+ cohort 23 patients (64%) were responders and within the TNBC cohort 8 (27%) were responders (p=0.006).

TABLE 5:

Pathologic Complete Response

	Evaluable Patients (N=66)	ER and/or PR positive (N=36)	TNBC (N=30)
pCR in breast
Yes	18	10	8
No	48	26	22
pCR rate, n (%)	18 (27%)	10 (28%)	8 (27%)
pCR in breast and axilla
Yes	15	8	7
No	51	28	23
pCR in breast and axilla rate, n (%)	15 (23%)	8 (22%)	7 (23%)

Efficacy assessment of pathologic complete response in breast, breast and axilla for evaluable patients.

TABLE 6:

CPS+EG Scores

	Evaluable patients (N=66)	ER and/or PR positive (N=36)	TNBC (N=30)
CPS+EG score
≤ 2	20 (30%)	18 (50%)	2 (7%)
≥ 3	38 (58%)	17 (47%)	21 (70%)
Indeterminate	8 (12%)	1 (3%)	7 (23%)
Responder = pCR and/or CPS+EG ≤ 2	31 (47%)	23 (64%)	8 (27%)
Non-responder = non-pCR or CPS+EG score ≥ 3	35 (53%)	13 (36%)	22 (73%)

Efficacy assessment by CPS+EG score. Responders included patients with pCR in the breast and/or CPS+EG score ≤ 2. Non-responders were patients with no pCR or CPS+EG score ≥ 3.

Secondary endpoints included DFS and OS (Figure 2). At 5 years, 45 patients (67%) were disease-free. DFS was not reached in both the ER+ and TNBC subsets and was statistically similar between the two groups (p=0.068). Median DFS was significantly longer in patients with CPS+EG scores ≤ 2 (DFS not reached for CPS+EG ≤ 2 vs 8.23 years for scores ≥ 3 vs 1.02 years for indeterminate scores, p=0.0035). Patients who were responders had significantly better DFS compared to those who were non-responders. The median DFS was not reached for responders vs 3.03 years for non-responders (p=0.00013).

Figure 2: Disease Free Survival and Overall Survival.

(A) DFS and OS based on hormone receptor status. Median OS and median DFS were not reached for ER/PR+ or TNBC disease. (B) DFS and OS based on CPS+EG score. Median DFS was 1.02 years for those with indeterminate CPS+EG, 8.23 years for patients with CPS+EG ≥ 3, and was not reached for patients with CPS+EG ≤ 2 (p=0.0035). Median OS for patients with indeterminate CPS+EG score 2.41 years and was not reached for patients with CPS+EG ≤ 2 and CPS+EG ≥ 3 (p=0.0029). (C) DFS and OS based on response to treatment. Patients were deemed responders if they had a pathologic complete response and/or had CPS+EG ≤ 2. Median DFS for non-responders was 3.03 years and was not reached for responders (p=0.00013). Median OS for non-responders was 4.73 years and was not reached for responders (p=<0.0001). N=66.

At 5 years, 31 (86%) percent of patients with ER/PR+ disease were still alive and 16 (63%) of patients with TNBC were alive. Overall, 47 (71%) of patients were alive. Median OS was not reached but was significantly better in the ER+ group compared to TNBC (p=0.014). Median OS was also significantly longer in patients with CPS+EG scores ≤ 2 (OS not reached for both CPS+EG scores ≤ 2 and ≥ 3 vs 2.41 years for indeterminant scores, p=0.0029). Responders had significantly better OS compared to non-responders: the median OS was not reached for responders vs 4.73 years for non-responders (p=<0.0001)

DISCUSSION

We found that the addition of S+T followed by AC+G-CSF was safe and tolerable as neoadjuvant therapy for LABC and IBC. The majority of patients (81%) completed protocol therapy. The primary endpoint for the study was pCR: overall pCR rate was 27%, 27% for TNBC, and 28% for hormone receptor-positive disease. Though the observed pCR rate did not meet the pre-specified criteria to reject the null hypothesis, the pCR rate observed in patients with ER+ LABC exceeded rates previously reported in randomized clinical trials.¹⁵ Additionally, when response was defined by pCR and/or CPS+EG ≤ 2, 31 patients (47%) were responders. Interestingly in the SWOG S0800 trial, which examined bevacizumab in combination with nab-paclitaxel followed by dose-dense AC compared to nab-paclitaxel followed or preceded by AC as neoadjuvant treatment for HER2-negative LABC or IBC, they found the addition of the anti-angiogenic agent bevacizumab significantly increased the pCR rate overall (36 vs 21 %; p = 0.019) and in TNBC (59 vs 29 %; p = 0.014), but not in hormone receptor-positive disease (24 vs 18%; p = 0.41).³⁶

However, in our study the addition of sunitinib to neoadjuvant chemotherapy was not associated with improvement in pCR for TNBC patients compared to historical controls alone while patients with ER or PR+ disease had higher response rates than previously reported (64% when response was defined by pCR and/or CPS+EG ≤ 2). This suggests that the addition of sunitinib to neoadjuvant chemotherapy may result in promising incremental benefits for patients with ER+ LABC. The biological rationale for why sunitinib might be more effective in ER+ patients is not fully understood. VEGF and the estrogen receptor are co-expressed in many breast tumors and several studies show estrogen can modulate VEGF expression in multiple tissues.^37,38 Heer et al. reported elevated VEGF levels immediately prior to surgery for patients with ER+ tumors while those with lobular carcinoma and ER− tumors had serum VEGF levels comparable to healthy controls.³⁹ Therefore it is possible that the tumorigenesis of some ER+ positive breast cancers may also be dependent on VEGF signaling. However, other studies have shown that VEGF levels are higher in TNBC.⁴⁰ As sunitinib is a broad specificity kinase inhibitor, it is also possible the effect we observed in the ER+ subtype is due to sunitinib’s inhibition of a pathway other than VEGF. For example, sunitinib is also known to be an inhibitor of RET. RET and ER expression have been shown to be significantly correlated in prior studies.^41–44 In a preclinical in vivo model Spanheimer et al. showed inhibition of RET via sunitinib reduced growth of luminal breast cancer cells primarily via increased apoptosis.⁴⁵ Overall, additional studies are needed to further elucidate why the ER+ subtype may be more sensitive to sunitinib.

The most common toxicities during treatment were neutropenia, leukopenia, fatigue, and anemia with no new safety signals observed. A total of 42 (63%) patients required dose modifications or a hold during the course of S+T treatment and 36 (62%) required dose modifications or a hold during the course of AC+G-CSF treatment. No grade 5 toxicities occurred. The number of AEs reported was relatively high, perhaps in part due to frequent assessment at weekly administration of dose-dense chemotherapy. Overall, the treatment combination was safe and tolerable.

One of the limitations of this work is that the continuous AC regimen that forms the backbone of the study is not the current standard of care. Selection of this regimen was based on data from SWOG 9625 and SWOG 0012 as well as institutional experience.^15,46 In SWOG 0012 Ellis et al. compared standard AC followed by paclitaxel to a continuous dosing schedule with daily oral cyclophosphamide and weekly doxorubicin followed by weekly paclitaxel in the neoadjuvant setting for patients with LABC or IBC. Initial results showed improved pCR rates in the continuous arm (31% in the continuous arm versus 19% pCR in the standard arm, p=0.02). At the time, this was highest reported pCR for this patient population. This regimen was consequently chosen as our backbone and the pre-specified endpoint for pCR was based on these results. Our study was powered to detect improvement in pCR when compared to this regimen. Additionally, the final results of the Ellis et al. study showed no difference in pCR including in the hormone receptor positive subgroup which had a pCR rate of 15.3% (n=98) in the continuous arm and 10.7% (n=84) in the standard arm. This supports that the difference observed in our trial was primarily driven by addition of sunitinib. Our study provides a provocative observation particularly in the setting of ER+ disease which has traditionally had unimpressive pCR rates following neoadjuvant chemotherapy. This highlights the continued need to better understand which patients may derive benefit from treatment with anti-angiogenic agents.

There have since been several trials evaluating the use of sunitinib in advanced breast cancer. A phase II study by Burstein et al showed that sunitinib was active and safe in patients with heavily pretreated metastatic breast cancer with an objective response rate (ORR) of 11%.⁴⁷ However, multiple studies have shown minimal benefit in terms of survival or response, but have raised concern for increased toxicity. A phase II study by Bergh et al showed sunitinib plus docetaxel improved ORR but did not prolong either DFS or OS compared with docetaxel alone in HER2 negative advanced breast cancer.⁴⁸ Similarly, there have been two phase III trials evaluating sunitinib in combination with capecitabine in advanced breast cancer both of which showed no improvement in progression free survival, OS, or ORR though the frequency and severity of toxicity was increased in patients receiving sunitinib.^49,50

In the neoadjuvant setting, a phase I/II trial of sunitinib administered with weekly paclitaxel/carboplatin in patients with locally advanced TNBC showed a pCR rate of 35%. While these results were comparable to other trials, substantial toxicity was observed with resulting dose reductions/omissions including discontinuation of sunitinib in 27% of patients.⁵¹

Lack of a predictive biomarker has hindered the use of sunitinib in the treatment of breast cancer and other solid tumors.⁴⁸ Biomarker data collected during this trial was reported previously by Brown-Glaberman et al.⁵² These results validated circulating carbonic anhydrase IX (CAIX) levels as a biomarker of hypoxia and HIF-1α upregulation suggesting this as a potential identifier for tumors unlikely to respond to anti-angiogenic therapies. Pre-clinical studies have suggested that anti-angiogenic agents may overprune the tumor vasculature, causing tumor hypoxia and genetic drift which leads to more aggressive phenotypes.^53–55 Areas of hypoxia may promote resistance by upregulation of pro-angiogenic factors that allow tumors to circumvent anti-angiogenic agents such as bevacizumab and sunitinib.⁵⁶ The biomarker data collected during this trial is consistent with this observation and suggests that tumors adapted to a hypoxic environment, as determined by baseline VCAM and CAIX levels, may be less sensitive to anti-angiogenic therapy.⁵² Additionally, in vivo models have shown that intra-tumoral hypoxia generated by treatment with sunitinib or bevacizumab increases cancer stem cell populations via HIF-1α upregulation and Akt/β-catenin activation,⁵⁷ thus limiting effectiveness of anti-angiogenic agents. Therefore a combination therapy that additionally targets resistance pathways and/or cancer stem proliferation may be necessary.⁵⁷

Another theory proposed by Jain et al is the concept of vascular normalization, suggesting that judicious use of anti-angiogenic therapy may temporarily stabilize the vasculature and therefore enhance drug delivery.⁵⁸ Conversely prolonged therapy or high dose therapy would lead to blood vessel destruction which in turn would decrease drug delivery.^58,59 Wong et al implemented this strategy in a phase Ib/II of neoadjuvant doxorubicin and cyclophosphamide with or without short course sunitinib (12.5 mg given daily for 7 days prior to chemotherapy.⁶⁰ While the addition of sunitinib did not improve pCR, they observed evidence of vascular normalization on IHC which was accompanied by improved vascular perfusion seen on imaging. However, there is no study that addresses the optimal timing or duration of anti-angiogenic therapy in relation to chemotherapy which would be critical in order to implement a vascular normalization strategy.⁶⁰

CONCLUSIONS

Neoadjuvant treatment with S+T followed by AC+G-CSF was found to be safe and tolerable in patients with locally advanced breast cancer. Using a combined definition of response of pCR and/or CPS+EG score ≤2, 47% of patients had a response to treatment. Patients with ER positive disease had the highest response rate (64%). Our data provide evidence that the addition of sunitinib to neoadjuvant chemotherapy may result in promising incremental benefit for patients with ER positive breast cancer. While other trials in breast cancer have not shown significant benefit with the addition of sunitinib, the present study suggests that carefully selected patients may derive benefit. This work highlights the continued challenges of the use sunitinib and other anti-angiogenic agents which have shown varying degrees of benefit in patients with breast cancer.

CLINICAL PRACTICE POINTS:

• Neoadjuvant chemotherapy is the treatment of choice for locally advanced breast cancer and inflammatory breast cancer. Achieving a pathologic complete response (pCR) following treatment is associated with improved survival, but unfortunately pCR is achieved in a minority of patients with standard anthracycline and taxane regimens.^1–5

• Sunitinib is a tyrosine kinase inhibitor with anti-tumor and anti-angiogenic activity.

• This multicenter, phase II clinical trial showed neoadjuvant treatment with sunitinib plus weekly paclitaxel (S+T) followed by doxorubicin and daily oral cyclophosphamide plus G-CSF (AC+G-CSF) was safe and tolerable patients with HER2 negative locally advanced breast cancer and inflammatory breast cancer. The most common toxicities were uncomplicated neutropenia, leukopenia, fatigue, and anemia.

• While pathologic complete response (pCR) did not meet the pre-specified criteria of interest, it was similar to comparable trials.

• The MDACC Clinical-pathologic scoring +ER and (CPS+EG) system incorporates pretreatment clinical stage, grade, and ER status with the final pathologic stage to more accurately stratify patients according to prognosis after administration of neoadjuvant chemotherapy.^25,26

• When response to S+T followed by AC+G-CSF was evaluated using a combined definition of response of pCR and/or CPS+EG score ≤2, there were a significant number of responders among ER+ patients suggesting there may be in promising incremental benefits in the subset of patients with ER+ tumors.