METHOD FOR PREDICTING THE OUTCOME OF A TREATMENT WITH AFLIBERCEPT OF A PATIENT SUSPECTED TO SUFFER FROM A CANCER BY MEASURING THE LEVEL OF A PLASMA BIOMARKER

Abstract

The present invention concerns the use of VCAM-1, ICAM-1 and/or PIGF as biomarkers for predicting the outcome of the treatment with aflibercept, or ziv-aflibercept of a patient suspected to suffer from a cancer.

Claims

1-3. (canceled)

4. A method for treating a patient, wherein the patient has colon cancer, a colorectal cancer or a rectal cancer, comprising administering a therapeutically effective amount of aflibercept or ziv-aflibercept to the patient, wherein a level of a biomarker selected from the group consisting of VCAM-1, ICAM-1, and PIGFdetected in a biological sample obtained from the patient is lower than a reference level of expression of the biomarker.

5. A method for treating a patient, wherein the patient has colon cancer, a colorectal cancer or a rectal cancer, comprising administering a therapeutically effective amount of aflibercept or ziv-aflibercept to the patient, wherein levels of VCAM-1 and ICAM-1 detected in a biological sample obtained from the patient are lower than respective reference levels of expression of VCAM-1 and ICAM-1.

6. (canceled)

7. A method according to claim 4, wherein the reference level of expression of VCAM-1 is between 406 and 577 ng/mL.

8. A method according to claim 4, wherein the reference level of expression of VCAM-1 is 553 ng/mL.

9. A method according to claim 4, wherein the reference level of expression of ICAM-1 is between 92 and 145 ng/mL.

10. A method according to claim 4, wherein the reference level of on expression of ICAM-1 is 144 ng/mL.

11. A method according to claim 4, wherein the reference level of expression of PIGF is between 12 and 19 pg/mL.

12. A method according to claim 4, wherein the reference level of expression of PIGF is 17 pg/mL.

13. A method according to claim 4, wherein the biological sample is selected from the group consisting of blood, serum, and plasma.

14. (canceled)

15. A method according to claim 4 wherein the colorectal cancer is a metastatic colorectal cancer.

16. A method according to claim 4, wherein the VCAM-1, ICAM-1 and/or PIGF level which is determined is a circulating level.

17. A method according to claim 4, comprising administering to said patient therapeutically effective amounts of aflibercept, or ziv-aflibercept, folinic acid, 5-fluorouracil (5-FU) and irinotecan.

18. (canceled)

19. A kit for predicting whether a patient suspected to suffer from cancer is a candidate for aflibercept, or ziv-aflibercept therapy, which kit comprises: a) means for measuring the level of a biomarker selected from the group consisting of VCAM-1, ICAM-1 and PIGF; and b) optionally, a label giving instructions for the use of said kit in predicting whether a patient suspected to suffer from cancer is a candidate for aflibercept, or ziv-aflibercept therapy.

20. An article of manufacture comprising: a) a packaging material; b) means for measuring the level of a biomarker selected from the group consisting of VCAM-1, ICAM-1 and PIGF; and c) a label giving instructions for the use of said kit in predicting whether a a patient suspected to suffer from cancer is a candidate for aflibercept, or ziv-aflibercept therapy.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0083] FIG. 1 represents Kaplan-Meier curves for PFS endpoint for sensitive and non-sensitive populations defined with VCAM-1.

[0084] FIG. 2 represents Kaplan-Meier curves for OS endpoint for sensitive and non-sensitive populations defined with VCAM-1.

[0085] FIG. 3 represents Kaplan-Meier curves for PFS endpoins for sensitive and non-sensitive populations defined with ICAM-1.

[0086] FIG. 4 represents Kaplan-Meier curves for OS endpoint for sensitive and non-sensitive populations defined with ICAM-1.

[0087] FIG. 5 represents Kaplan-Meier curves for PFS endpoint for sensitive and non-sensitive populations defined with VCAM-1 and ICAM-1.

[0088] FIG. 6 represents Kaplan-Meier curves for OS endpoint for sensitive and non-sensitive populations defined with VCAM-1 and ICAM-1.

[0089] FIG. 7 represents Kaplan-Meier curves for PFS endpoint for sensitive and non-sensitive populations defined with PIGF.

DETAILED DESCRIPTION OF THE INVENTION

Example: Predictive Effect of VCAM-1, ICAM-1 and/or PIGF on PFS in the AFLAME Study

Study EFC11338 (AFLAME)

[0090] EFC11338 was designed as a Multinational, Randomized, Double-Blind Study of Aflibercept Versus Placebo with Irinotecan/5-FU Combination (FOLFIRI) in Patients with Metastatic Colorectal Cancer (MCRC) After Failure of an Oxaliplatin Based Regimen.

Objectives:

[0091] To identify potential predictive biomarkers for response to treatment on efficacy endpoints (Progression Free Survival (PFS), Overall Survival (OS) and Overall Response Rate (ORR))

[0092] To identify potential prognostic biomarkers for efficacy endpoints (PFS, OS and ORR)

[0093] To identify potential correlation between biomarkers and other baseline characteristics

[0094] To identify potential correlation of longitudinal plasma measurements with clinical endpoints

[0095] To identify potential groups of homogeneous individuals based on their molecular profiles in an unsupervised way and to estimate correlation with clinical outcomes (PFS, OS and ORR)

[0096] To assess the safety profile of the population identified by prognostic/predictive biomarkers

Dosage and Schedule of Administration:

[0097] Patients were administered either aflibercept or placebo, depending on arm assigned. Immediately after, patients received irinotecan, 5-FU and leucovorin (FOLFIRI regimen). This treatment was repeated every 2 weeks.

Aflibercept/Placebo

[0098] Arm A, aflibercept: 4 mg/kg was administered IV over 1 hour on Day 1, every 2 weeks, OR

[0099] Arm B, placebo: 4 mg/kg was administered IV over 1 hour on Day 1, every 2 weeks.

FOLFIRI Regimen

[0100] Immediately after aflibercept/placebo administration, all the patients received: [0101] Irinotecan 180 mg/m2 IV infusion in 500 mL D5W over 90 minutes and dl leucovorin* 400 mg/m2 IV infusion over 2 hours, at the same time, in bags using a Y-line, followed by: [0102] 5-FU 400 mg/m2 IV bolus given over 2-4 minutes, followed by: [0103] 5-FU 2400 mg/m2 continuous IV infusion in 500 mL D5W (recommended) over 46-hours.

Duration of Treatment:

[0104] Patient was treated until disease progression, unacceptable toxicity or patient's refusal

Duration of Observation:

[0105] Patients were followed when on study treatment and during follow up period until death or the study cut-off date for OS, whichever comes first. The cut-off date for OS is one year after the last patient enrolled.

Number of Subjects:

[0106] Intent-to-Treat (ITT) population: 332 (109 in the placebo group+223 in the aflibercept group)

[0107] Evaluable population for response rate: 295 (96 in the placebo group+199 in the aflibercept group)

[0108] Evaluable for Luminex biomarkersGroup 1 biomarkers: 295 (99 in the placebo group+196 in the aflibercept group)

[0109] Evaluable for Luminex biomarkersGroup 2 biomarkers: 108 (37 in the placebo group+71 in the aflibercept group)

[0110] Evaluable for ELISA biomarkers: 329 (107 in the placebo group+222 in the aflibercept group)

[0111] Due to an operational error, some patients randomized to the aflibercept arm have received placebo for one or several treatment cycles; or vice versa. A total of 198 patients received at least one misallocated treatment kit.

Criteria for Evaluation:

Biomarkers:

[0112] Different types of biomarker data were investigated in the current translational research proposal in the AFLAME study: [0113] 107 plasma angiogenic factors and inflammatory cytokines measured at baseline, during and after treatment (end of cycle 1 infusion, 48 h post aflibercept/placebo on cycle 2 or 3, 30 days after last aflibercept/placebo administration) with Luminex? technology [0114] Free VEGF-A and PIGF measured in plasma samples at baseline with ELISA (Enzyme-Linked Immunosorbent Assay) technology

Preparation of Samples and Analysis

[0115] Plasma biomarkers were measured either with Luminex? technology (bead-based multiplex assay) or with ELISA.

Plasma Biomarkers Measured with Luminex? Technology

[0116] Plasma angiogenic factors and inflammatory cytokines have been measured at baseline pre cycle 1.

[0117] Thirty proteins measured on all the samples were defined as Group 1 biomarkers and 77 proteins measured only for some samples were defined as Group 2 biomarkers. The biomarkers from Group 1 were angiogenic/inflammatory molecules and have been selected based on aflibercept mechanism of action (inhibiting 3 angiogenic factors and their receptors), key candidate biomarkers identified on independent aflibercept studies or literature/experts.

Plasma Biomarkers Measured with ELISA

[0118] In addition to plasma biomarkers measured with Luminex? technology, free VEGF-A and free PIGF concentrations of baseline plasma samples were measured with ELISA technology.

Results

Univariate Analysis

[0119] Biomarkers have been tested for predictive and prognostic effects for PFS.

[0120] VCAM-1 and ICAM-1 have been identified as potentially predictive with corrected Benjamini-Hochberg (BH) p-value?0.2 (295 subjects).

[0121] Then sensitive and non-sensitive populations have been defined using VCAM-1 and ICAM-1 (BH p-value?0.2).

[0122] PIGF has been identified as potentially predictive by ELISA (unadjusted p-value=0.075).

Identification of Sensitive and Non-Sensitive Populations with VCAM-1

[0123] The threshold of 6.32 corresponding to 553 ng/mL has been determined to define sensitive (197 individuals) and non-sensitive populations (98 individuals) respectively corresponding to individuals with low values for VCAM-1 and high values for VCAM-1.

[0124] FIG. 1 and FIG. 2 represent Kaplan-Meier curves for PFS and OS endpoints for sensitive and non-sensitive populations illustrating the better treatment effect for low VCAM-1 group compared to high VCAM-1 group. For OS, in addition to a better treatment effect compared to placebo, sensitive population showed globally a better prognostic (increased OS for low VCAM-1 group in placebo and aflibercept arms).

[0125] Table 1 shows the response rate for sensitive and non-sensitive populations by treatment group. There was an increased response rate (26%) for the aflibercept/folfiri treatment in sensitive population compared to the non-sensitive population (10%).

TABLE-US-00001 TABLE 1 Response Rate for non-sensitive/sensitive populations defined with VCAM-1 for PFS - Preprocessed data - VCAM-1 and RR evaluable population Population Placebo/Folfiri Aflibercept/Folfiri Sensitive VCAM-1 <= 6.32 3/68 (4%) 31/121 (26%) (N = 189) Non-sensitive VCAM-1 > 6.32 1/26 (4%) 7/68 (10%) (N = 94) Total 4/94 (4%) 38/189 (20%)
Identification of Sensitive and Non-Sensitive Populations with ICAM-1

[0126] The threshold of 5.04 corresponding to 144 ng/mL has been determined to define sensitive (205 individuals) and non-sensitive populations (90 individuals) respectively corresponding to individuals with low values for ICAM-1 and high values for ICAM-1.

[0127] FIG. 3 and FIG. 4 represent Kaplan-Meier curves for PFS and OS endpoints for sensitive and non-sensitive populations illustrating the better treatment effect for low ICAM-1 group compared to high ICAM-1 group. For OS, in addition to a better treatment effect compared to placebo, sensitive population showed globally a better prognostic (increased OS for low ICAM-1 group in placebo and aflibercept arms).

[0128] Table 2 shows the response rate for sensitive and non-sensitive populations by treatment group. There has been an increased response rate (25%) for the aflibercept/folfiri treatment in sensitive population compared to the non-sensitive population (9%).

TABLE-US-00002 TABLE 2 Response Rate for non-sensitive/sensitive populations defined with ICAM-1 for PFS - Preprocessed data - ICAM-1 and RR evaluable population Population Placebo/Folfiri Aflibercept/Folfiri Sensitive ICAM-1 <= 5.04 3/63 (5%) 33/134 (25%) (N = 197) Non-sensitive ICAM-1 > 5.04 1/31 (3%) 5/55 (9%) (N = 86) Total 4/94 (4%) 38/189 (20%)

Multivariate Analysis

[0129] The multivariate predictive score 0.089?ICAM?1+0.17?VCAM?1 has been dichotomized using quantile 10% to 90% as threshold.

[0130] Sensitive population showed a significant difference in PFS and in OS in favor of aflibercept over placebo (HR=0.47 for PFS and HR=0.66 for OS), which was increased compared to the non-sensitive population (HR=0.98 for PFS and HR=1.04 for OS).

[0131] Median PFS difference between aflibercept and placebo was equal to 2.59 months in sensitive population, showing a greater but moderate difference compared to non-sensitive population (0.5 month).

[0132] Median OS difference between aflibercept and placebo was equal to 3.75 months in sensitive population, showing a greater difference compared to non-sensitive population (?0.39 month). For non-sensitive population there was a decrease in median OS for placebo arm and treated arm (8.90 months and 8.51 months) compared to the other populations illustrating a potential prognostic effect of multivariate score in addition to the predictive effect.

[0133] In conclusion sensitive population showed a decreased HR compared to the non-sensitive population for PFS and OS with moderate gain in term of median.

[0134] FIG. 5 and FIG. 6 represent Kaplan-Meier curves for PFS and OS endpoints for sensitive and non-sensitive populations illustrating the better treatment effect for low score group compared to high score group. For OS, in addition to a better treatment effect compared to placebo, sensitive population showed globally a better prognostic (increased OS for low score group in placebo and aflibercept arms).

Identification of Sensitive and Non-Sensitive Populations with PIGF

[0135] The threshold of 2.82 corresponding to 17 pg/ml has been determined to define sensitive (230 individuals) and non-sensitive populations (99 individuals) respectively corresponding to individuals with low values for PIGF and high values for PIGF.

[0136] FIG. 7 which represents the Kaplan-Meier curves for PFS endpoint for sensitive and non-sensitive populations defined with PIGF illustrates the determination of the PIGF cut off for sensitive and non-sensitive populations.

CONCLUSIONS

[0137] VCAM-1 and ICAM-1 have been identified as potentially predictive biomarkers for PFS in a univariate framework (unadjusted p-value equal to 0.00017 for VCAM-1 and 0.0043 for ICAM-1).

[0138] The third biomarker that showed up to be potentially predictive was PIGF measured by ELISA (unadjusted p-value=0.075).

[0139] Linear combination of VCAM-1 and ICAM-1 has been identified as potentially predictive for PFS.