Neoadjuvant chemotherapy with or without nintedanib for advanced epithelial ovarian cancer: Lessons from the GINECO double-blind randomized phase II CHIVA trial.

AIM: The oral anti-angiogenic therapy nintedanib prolongs progression-free survival (PFS) when combined with chemotherapy after primary surgery for advanced epithelial ovarian cancer. The randomized phase II CHIVA trial evaluated the impact of combining nintedanib with neoadjuvant chemotherapy (NACT) for epithelial ovarian cancer. METHODS: Patients with newly diagnosed unresectable FIGO stage IIIC-IV epithelial ovarian cancer received 3-4 cycles of carboplatin plus paclitaxel every 3 weeks as NACT before interval debulking surgery (IDS), followed by 2-3 post-operative cycles. Patients were randomized 2:1 to receive either nintedanib 200 mg twice daily or placebo on days 2-21 every 3 weeks during NACT (omitting peri-operative cycles), and then as maintenance therapy for up to 2 years. The primary endpoint was PFS. RESULTS: Between January 2013 and May 2015, 188 patients were randomized (124 to nintedanib, 64 to placebo). PFS was significantly inferior with nintedanib (median 14.4 versus 16.8 months with placebo; hazard ratio 1.50, p = 0.02). Overall survival (OS) was also inferior (median 37.7 versus 44.1 months, respectively; hazard ratio 1.54, p = 0.054). Nintedanib was associated with increased toxicity (grade 3/4 adverse events: 92% versus 69%, predominantly hematologic and gastrointestinal), lower response rate by RECIST (35% versus 56% before IDS), and lower IDS feasibility (58% versus 77%) versus placebo. CONCLUSIONS: Adding nintedanib to chemotherapy and in maintenance as part of NACT for advanced epithelial ovarian cancer cannot be recommended as it increases toxicity and compromises chemotherapy efficacy (IDS, PFS, OS). CLINICALTRIALS: govregistration: NCT01583322.

Paclitaxel with or without pazopanib for ovarian cancer relapsing during bevacizumab maintenance therapy: The GINECO randomized phase II TAPAZ study.

BACKGROUND: Anti-angiogenic rechallenge with bevacizumab plus chemotherapy is effective in recurrent ovarian cancer (rOC); however, data are limited on tyrosine kinase inhibitors after progression on maintenance bevacizumab. METHODS: In the randomized phase II TAPAZ trial, patients with rOC during the first year of bevacizumab maintenance therapy were assigned 2:1 to either weekly paclitaxel 65 mg/m2 plus pazopanib 600-800 mg daily or standard weekly paclitaxel 80 mg/m2. The primary endpoint was 4-month progression-free survival (PFS) rate. RESULTS: Overall, 116 patients were randomized and treated: 79 with combination therapy and 37 with single-agent paclitaxel. Median follow-up was 13.1 months. There was no difference between treatment arms in 4-month PFS rate (61% [95% CI, 51-73%] with the combination versus 68% [95% CI, 54-85%] with paclitaxel alone), median PFS (4.9 [95% CI, 4.1-6.1] versus 5.8 [95% CI, 4.8-7.4] months, respectively) or median overall survival (13.6 versus 12.9 months, respectively). The combination was associated with more grade 3/4 toxicities (87% versus 70%, respectively) and toxicity-related paclitaxel discontinuations (22% versus 11%). Pazopanib was discontinued for toxicity in 44% of patients, most commonly for gastrointestinal and vascular events. There were two treatment-related deaths, both in the combination arm (pulmonary embolism and gastrointestinal perforation). At month 4, patient-reported outcomes deteriorated from baseline in the combination arm, particularly for abdominal/gastrointestinal symptoms, which showed a clinically important difference versus paclitaxel alone. CONCLUSIONS: In rOC progressing during maintenance bevacizumab, adding pazopanib to paclitaxel did not improve efficacy, increased toxicity, and compromised chemotherapy delivery. CLINICALTRIALS: govregistration:NCT02383251.

Regorafenib or Tamoxifen for platinum-sensitive recurrent ovarian cancer with rising CA125 and no evidence of clinical or RECIST progression: A GINECO randomized phase II trial (REGOVAR).

OBJECTIVE: To evaluate the efficacy and safety of regorafenib versus tamoxifen in platinum-sensitive ovarian cancer biological recurrence, defined by CA-125 increase without radiological (RECIST criteria) or symptomatic evidence of progression. PATIENTS AND METHODS: 116 patients with platinum-sensitive ovarian cancer presenting an isolated increase of CA-125 were planned to be randomized. Regorafenib was administered orally at 160 or 120 mg daily, 3 weeks on/1 week off or tamoxifen at 40 mg daily, until disease progression or development of unacceptable toxicity. The primary endpoint was Progression-Free Survival, assessed by progression according to RECIST 1.1 or death (by any cause). Secondary endpoints included Overall Survival, Best Response and CA-125 response rate. RESULTS: 68 patients were randomized. Median age was 67 years (range: 30-87). Primary site of cancer was ovarian for most patients (92.6%). Tumors were predominantly serous / (89.7%), high grade (83.6%) and initial FIGO staging was III for 69.6% of the patients. Most (79.4%) patients were included after the first line of platinum-based treatment. After a median follow-up of 32 months, there was no difference of progression-free survival (PFS) between regorafenib and tamoxifen groups (p = 0.72), with median PFS of 5.6 months (CI 90%: 3.84-7.52) for the tamoxifen arm and 4.6 months (CI 90%: 3.65-7.33) for the regorafenib arm. There was also no difference in term of overall survival, best response or CA-125 response, delay to next therapy. Regorafenib presented a less favorable safety profile than tamoxifen, with grade 3/4 events occurring for 90.9% of the patients compared to 54.3% for tamoxifen. The most frequent were cutaneous, digestive, and biological events. Notably, hand-foot syndrome occurred in 36.4% of these patients. CONCLUSION: Regorafenib presented an unfavorable toxicity profile compared to tamoxifen, with no superior efficacy in this population of patients.

Predicting tumor response and outcome of second-look surgery with 18F-FDG PET/CT: insights from the GINECO CHIVA phase II trial of neoadjuvant chemotherapy plus nintedanib in stage IIIc-IV FIGO ovarian cancer.

BACKGROUND: This ancillary study aimed to evaluate 18F-FDG PET parameter changes after one cycle of treatment compared to baseline in patients receiving first-line neoadjuvant anti-angiogenic nintedanib combined to paclitaxel-carboplatin chemotherapy or chemotherapy plus placebo and to evaluate the ability of 18F-FDG PET parameters to predict progression-free survival (PFS), overall survival (OS), and success of second-look surgery. MATERIALS AND METHODS: Central review was performed by two readers blinded to the received treatment and to the patients' outcome, in consensus, by computing percentage change in PET metrics within a volume of interest encompassing the entire tumor burden. EORTC and PERCIST criteria were applied to classify patients as responders (partial metabolic response and complete metabolic response) or non-responders (stable metabolic disease and progressive metabolic disease). Also analyzed was the percentage change in metabolic active tumor volume (MATV) and total lesion glycolysis (TLG). RESULTS: Twenty-four patients were included in this ancillary study: 10 received chemotherapy + placebo and 14 chemotherapy + nintedanib. PERCIST and EORTC criteria showed similar discriminative power in predicting PSF and OS. Variation in MATV/TLG did not predict PFS or OS, and no optimal threshold could be found for MATV/TLG for predicting survival. Complete cytoreductive surgery (no residual disease versus residual disease < 0.25 cm/0.25-2.5 cm/> 2.5 cm) was more frequent in responders versus non-responders (P = 0.002 for PERCIST and P = 0.02 for EORTC criteria). No correlation was observed between the variation of PET data and the variation of CA-125 blood level between baseline sample and that performed contemporary to the interim PET, but a statistically significant correlation was observed between ΔSULpeak and ΔCA-125 between baseline sample and that performed after the second cycle. CONCLUSION: 18F-FDG PET using EORTC or PERCIST criteria appeared to be a useful tool in ovarian cancer trials to analyze early tumor response, and predict second-look surgery outcome and survival. An advantage of PERCIST is the correlation of ΔSULpeak and ΔCA-125, PET response preceding tumor markers response by 1 month. Neither MATV nor TLG was useful in predicting survival. TRIAL REGISTRATION: NCT01583322 ARCAGY/ GINECO GROUP GINECO-OV119, 24 April 2012.

First-line bevacizumab and eribulin combination therapy for HER2-negative metastatic breast cancer: Efficacy and safety in the GINECO phase II ESMERALDA study.

PURPOSE: Combining bevacizumab with paclitaxel significantly improves progression-free survival (PFS) versus paclitaxel alone in HER2-negative metastatic breast cancer (MBC). Eribulin is active and tolerable in pretreated MBC. To assess whether eribulin may offer a more tolerable yet effective combination partner for bevacizumab, we evaluated a bevacizumab/eribulin combination regimen as first-line therapy for MBC. METHODS: In this single-arm phase II study, patients with histologically confirmed HER2-negative MBC and no prior chemotherapy for MBC received eribulin 1.23 mg/m2 on days 1 and 8 every 3 weeks for ≥6 cycles plus bevacizumab 15 mg/kg on day 1 every 3 weeks until disease progression. The primary endpoint was non-progression rate at 1 year. Secondary endpoints included objective response rate (ORR), PFS, and safety. RESULTS: The median age of the 61 treated female patients was 59 years, 16% had triple-negative MBC, 30% had ≥3 metastatic sites, and 71% had received prior (neo)adjuvant chemotherapy. Patients received a median of six eribulin and nine bevacizumab cycles. The non-progression rate at 1 year was 32% (95% confidence interval [CI]: 20-43%), ORR was 47% (95% CI: 34-60%), and median PFS was 8.3 months (95% CI: 7.0-9.6 months). The only grade ≥3 clinical adverse events in >5% of patients were hypertension (39%), neutropenia (26%), thrombosis (10%), and paresthesia/dysesthesia (7%). CONCLUSION: First-line eribulin/bevacizumab combination therapy showed interesting activity in MBC with an acceptable safety profile, including a particularly low incidence of high-grade neuropathy.

Comparative analysis of predictive values of the kinetics of 11 circulating miRNAs and of CA125 in ovarian cancer during first line treatment (a GINECO study).

OBJECTIVE: MicroRNAs (miRNAs) are promising biomarkers in ovarian cancer. Their kinetics during treatment might be useful for monitoring disease burden, and guiding treatments in patients treated with peri-operative chemotherapy and interval debulking surgery (IDS). METHODS: Serial blood samples of patients enrolled in the randomized phase II CHIVA trial, comparing first line carboplatin-paclitaxel +/- nintedanib (NCT01583322) and IDS, were investigated to assess the kinetics of 11 relevant miRNAs. Their prognostic/predictive values regarding the likelihood of complete IDS, and the patient survival, were assessed and compared to those of CA125 kinetics. The selection of the miRNAs (miR-15b-5p, miR-16-5p, miR-20a-5p, miR-21-5p, miR-93-5p, miR-122-5p, miR-150-5p, miR-195-5p, miR-200b-3p, miR-148b-5p and miR-34a-5p) was based on the expression levels found with a large explorative panel, and on the literature data. RESULTS: 756 serial blood samples from 119 patients were analyzed for a total of 8172 miRNA assays, and 1299 CA125 values. The longitudinal kinetics of the miRNA expressions were highly inconsistent, and were not related to CA125 dynamics. The miRNA changes during neoadjuvant treatment were not found associated with RECIST tumor response or IDS outcomes. Decreases of miR-34a-5p and miR-93-5p were associated with PFS benefit (p = .009) and OS benefits (p < .001), respectively, using univariate tests. CONCLUSIONS: The longitudinal kinetics of miRNA expressions during neoadjuvant treatment in ovarian cancer patients were inconsistent, and were not found to be associated with tumor burden changes. Although some prognostic value could be discussed, no predictive value regarding tumor responses or IDS quality could be identified.

GINECO Prospective Non-interventional PROSPECTYON Study: Trabectedin Plus Pegylated Liposomal Doxorubicin for Platinum-sensitive Recurrent Ovarian Cancer.

BACKGROUND: Trabectedin and pegylated liposomal doxorubicin (PLD) is an effective combination therapy for platinum-sensitive recurrent ovarian cancer (ROC), particularly for disease relapsing within 6-12 months of platinum therapy. The non-interventional PROSPECTYON study evaluated trabectedin/PLD in French clinical practice. PATIENTS AND METHODS: Patients with ROC after at least one platinum-based regimen received 1.1 mg/m2 trabectedin plus 30 mg/m2 PLD every 3 weeks. Efficacy and safety were evaluated in subgroups according to platinum-free interval [6-12 versus ≥12 months (partially or fully platinum sensitive, respectively)]. RESULTS: Recurrent disease was partially platinum-sensitive in 58 patients and fully sensitive in 33 patients treated between July 2014 and June 2016. Patients in both subgroups received a median of six cycles of trabectedin and PLD. The most common grade 3 or more toxicities were haematological. Median progression-free survival was 6 months for both subgroups. CONCLUSION: Trabectedin/PLD is a valuable treatment option for partially or fully platinum-sensitive ROC.

Assessment of the prognostic role of a 94-single nucleotide polymorphisms risk score in early breast cancer in the SIGNAL/PHARE prospective cohort: no correlation with clinico-pathological characteristics and outcomes.

BACKGROUND: Genome-wide association studies (GWAS) have to date identified 94 genetic variants (single nucleotide polymorphisms (SNPs)) associated with risk of developing breast cancer. A score based on the combined effect of the 94 risk alleles can be calculated to measure the global risk of breast cancer. We aimed to test the hypothesis that the 94-SNP-based risk score is associated with clinico-pathological characteristics, breast cancer subtypes and outcomes in early breast cancer. METHODS: A 94-SNP risk score was calculated in 8703 patients in the PHARE and SIGNAL prospective case cohorts. This score is the total number of inherited risk alleles based on 94 selected SNPs. Clinical data and outcomes were prospectively registered. Genotyping was obtained from a GWAS. RESULTS: The median 94-SNP risk score in 8703 patients with early breast cancer was 77.5 (range: 58.1-97.6). The risk score was not associated with usual prognostic and predictive factors (age; tumor, node, metastasis (TNM) status; Scarff-Bloom-Richardson grade; inflammatory features; estrogen receptor status; progesterone receptor status; human epidermal growth factor receptor 2 (HER2) status) and did not correlate with breast cancer subtypes. The 94-SNP risk score did not predict outcomes represented by overall survival or disease-free survival. CONCLUSIONS: In a prospective case cohort of 8703 patients, a risk score based on 94 SNPs was not associated with breast cancer characteristics, cancer subtypes, or patients' outcomes. If we hypothesize that prognosis and subtypes of breast cancer are determined by constitutional genetic factors, our results suggest that a score based on breast cancer risk-associated SNPs is not associated with prognosis. TRIAL REGISTRATION: PHARE cohort: NCT00381901 , Sept. 26, 2006 - SIGNAL cohort: INCa RECF1098, Jan. 28, 2009.

Constitutional variants are not associated with HER2-positive breast cancer: results from the SIGNAL/PHARE clinical cohort.

Human epidermal growth factor receptor 2-positive breast cancer is a subtype of interest regarding its outcome and the impressive impact of human epidermal growth factor receptor 2 targeted therapy. Constitutional variants may be involved in the aetiology of human epidermal growth factor receptor 2-positive breast cancer, and we propose a case-case study to test the hypothesis that single nucleotide polymorphisms may be associated with human epidermal growth factor receptor 2 status. A Genome-Wide Association Study was used in a cohort of 9836 patients from the SIGNAL/PHARE study (NCT00381901-RECF1098). The main goal was to identify variants specifically related to human epidermal growth factor receptor 2-positive breast cancer. A two-staged genotyping strategy was carried out to cover as large a proportion of the genome as possible. All subjects were genotyped using the Illumina HumanCore Exome chip set. Principal Components Analysis and k-means were then used to characterize the ancestry of the participants. A random sample of subjects from the main "European" cluster was genotyped with the Omni5 chip set. These data were then used to impute missing genotypes from the remaining subjects genotyped only using the HumanCore Exome array. From the 9836 patients, a total of 8703 cases including 3230 patients with human epidermal growth factor receptor 2-positive breast cancer were analyzed. Despite having 80% power to detect an odds ratio of 1.23 in this population, no variant achieved genome-wide significance for association with the occurrence of human epidermal growth factor receptor 2-positive breast cancer vs. any other subtype of breast tumour. Our study was unable to identify constitutional polymorphisms that are strongly associated with human epidermal growth factor receptor 2-positive status among breast cancer patients.

GWAS in the SIGNAL/PHARE clinical cohort restricts the association between the FGFR2 locus and estrogen receptor status to HER2-negative breast cancer patients.

Genetic polymorphisms are associated with breast cancer risk. Clinical and epidemiological observations suggest that clinical characteristics of breast cancer, such as estrogen receptor or HER2 status, are also influenced by hereditary factors. To identify genetic variants associated with pathological characteristics of breast cancer patients, a Genome Wide Association Study was performed in a cohort of 9365 women from the French nationwide SIGNAL/PHARE studies (NCT00381901/RECF1098). Strong association between the FGFR2 locus and ER status of breast cancer patients was observed (ER-positive n=6211, ER-negative n=2516; rs3135718 OR=1.34 p=5.46×10-12). This association was limited to patients with HER2-negative tumors (ER-positive n=4267, ER-negative n=1185; rs3135724 OR=1.85 p=1.16×10-11). The FGFR2 locus is known to be associated with breast cancer risk. This study provides sound evidence for an association between variants in the FGFR2 locus and ER status among breast cancer patients, particularly among patients with HER2-negative disease. This refinement of the association between FGFR2 variants and ER-status to HER2-negative disease provides novel insight to potential biological and clinical influence of genetic polymorphisms on breast tumors.

Telomere length is a prognostic biomarker in elderly advanced ovarian cancer patients: a multicenter GINECO study.

PURPOSE: Age induces a progressive decline in functional reserve and impacts cancer treatments. Telomere attrition leads to tissue senescence. We tested the hypothesis that telomere length (TL) could predict patient vulnerability and outcome with cancer treatment. PATIENTS AND METHODS: An ancillary study in the Elderly Women GINECO Trial 3 was performed to evaluate the impact of geriatric covariates on survival in elderly advanced ovarian cancer patients receiving six cycles of carboplatin. TL was estimated from peripheral blood at inclusion using standard procedures. RESULTS: TL (in base pairs) was estimated for 109/111 patients (median 6.1 kb; range [4.5-8.3 kb]). With a cut-off of 5.77 kb, TL discriminated two patient groups, long telomere (LT) and short telomeres (ST), with significantly different treatment completion rates of 0.80 (95% CI [0.71-0.89]) and 0.59 (95% CI [0.41-0.76]), respectively (odds ratio [OR]=2.8, p=0.02). ST patients were at higher risk of serious adverse events (SAE, OR=2.7; p=0.02) and had more unplanned hospital admissions (OR=2.1; p=0.08). After adjustment on FIGO stage, TL shorter than 6 kb was a risk factor of premature death (HR=1.57; p=0.06). CONCLUSION: This exploratory study identifies TL as predictive factor of decreased treatment completion, SAE risk, unplanned hospital admissions and OS after adjustment on FIGO stage.

A potentially neuroprotective role for erythropoietin with paclitaxel treatment in ovarian cancer patients: a prospective phase II GINECO trial.

PURPOSE: A prospective phase II multicenter study was performed in two steps in paclitaxel-treated ovarian cancer patients in France. A French version of the four-item Functional Assessment of Cancer Therapy/Gynecologic Oncology Group-Neurotoxicity (FACT/GOG-Ntx) questionnaire was validated. This was then used to evaluate neurotoxicity in relation to erythropoietin treatment. METHODS: Patients received standard second-line paclitaxel-based chemotherapy and erythropoietin for anemia. Neurotoxicity and hemoglobin levels were evaluated every cycle with the FACT/GOG-Ntx and NCI-CTCAE. The translated questionnaire was tested in 20 patients to confirm the translation accuracy. The final questionnaire was validated in 98 patients with internal consistency (Cronbach's coefficient) and item correlation (Pearson's r coefficient) tests. Neurotoxicity severity was analyzed according to erythropoietin intake (first three cycles versus no or late intake) and correlated with anemia. RESULTS: Patients received a median of six paclitaxel cycles (range 1-9). Neurotoxicity was validated in 484 questionnaires. Internal consistency was excellent with Cronbach's coefficients of ≥0.89 at inclusion, after 3 cycles and at study end. Inter-question correlation was high with Pearson's coefficients of 0.65-0.85. FACT/GOG-Ntx and NCI-CTCAE severity scoring was similar. Globally, the incidence of severe neurotoxicity (FACT/GOG-Ntx and NCI-CTCAE) was found significantly higher in patients with severe anemia. Of 98 evaluable patients, 31 received erythropoietin during the first three cycles. Mean hemoglobin level was significantly lower in this group from baseline to cycle 4; however, these anemic patients with early EPO intake did not develop an increase rate of severe neurotoxicity. CONCLUSIONS: The French FACT/GOG-Ntx questionnaire is a reliable and valid tool for assessing chemotherapy-induced neuropathy. This study raises the possibility that erythropoietin might play a neuroprotective role when administered with paclitaxel.

Conjunctival low-grade non-Hodgkin's lymphoma: a large single-center study of initial characteristics, natural history and prognostic factors.

To define the initial characteristics and prognostic factors of patients with conjunctival low-grade malignant lymphoma, all patients treated for low-grade lymphoma with initial conjunctival involvement were reviewed. Forty-nine cases were selected, including 45 cases with exclusive ophthalmologic conjunctival involvement. Pathologic review showed 55% of mucosa-associated lymphoid tissue type lymphoma, and 23% of lymphoplasmocytic lymphoma. Initial characteristics were median age of 62 years, nodal involvement in 17% of cases, and stage IV in 22% of patients with 10% of bone marrow involvement. With a median follow-up of 75 months, the 5-year disease-free survival (DFS) and overall survival were 65% and 83%, respectively. On multivariate analysis, nodal involvement was the only factor with a pejorative impact on DFS. Our patient cohort represents one of the largest published series defining the characteristics and prognostic factors of primary conjunctival low-grade malignant lymphoma.

Ophthalmologic and intraocular non-Hodgkin's lymphoma: a large single centre study of initial characteristics, natural history, and prognostic factors.

The aims of this study were to define the initial characteristics, natural history, and prognostic factors of patients with ophthalmologic and intraocular malignant lymphoma. All patients treated at the Institut Curie for lymphoma with ophthalmologic (orbit and/or adnexa) or intraocular involvement were retrospectively reviewed. A pathological review of all cases was performed according to the WHO classification. One hundred and forty-five patients were selected for the study. Pathological review showed 36% MALT type lymphoma, 22% lymphoplasmocytic lymphoma, and 15% diffuse large B-cell lymphoma. Ophthalmologic and ocular sites were intra-orbital in 61 cases (42%) and conjunctival in 51 cases (35%), with bilateral involvement in 10% of cases. Stage IV was found in 32% of cases, with bone marrow involvement in 12%. With a median follow-up of 90 months, the 5-year DFS and OS were 64 and 79% for low-grade NHL, and 43 and 50% for high-grade NHL. On multivariate analysis, age greater than 59 years, elevated LDH level, stage IV, high-grade histological subgroup, and presence of B-symptoms had a negative impact on OS for the overall population. In conclusion, with a median follow-up of 7.5 years, our large cohort of patients represents one of the largest published series on primary ophthalmologic and intraocular malignant lymphoma.