Comparison of adjuvant gemcitabine and capecitabine with gemcitabine monotherapy in patients with resected pancreatic cancer (ESPAC-4): a multicentre, open-label, randomised, phase 3 trial
Neoptolemos JP et al., Lancet. 2017 Mar 11;389(10073):1011-1024
In the adjuvant setting, gemcitabine is usually favoured over 5-fluorouracil with folinic acid (5FU/FA) due to a more tolerable safety profile. In an open-label phase III ESPAC-3 trial, patients with pancreatic cancer were randomized to receive 5FU/FA or gemcitabine after surgical resection (JAMA 2010; 304: 1073–81). The trial failed to show superiority of gemcitabine over 5FU/FA (OS 23 mos. vs. 23.6 mos.; p=0.39). Researchers conducted a trial to determine whether the addition of capecitabine to gemcitabine could improve survival in the adjuvant setting.
- Patients: histologically or cytologically-proven ductal adenocarcinoma of the pancreas after complete resection (R0 or R1), no evidence of metastasis, PS ≤ 2
- Study design: phase III, 2 group open-label RCT
- Experimental: Gemcitabine 1,000 mg/m2 IV weekly (3 out of 4 wks.) x 6 cycles Capecitabine 1,660 mg/m2 PO x 21 days then 7 days rest x 6 cycles
- Control: Gemcitabine 1,000 mg/m2 IV q3 out of 4 wks. x 6 cycles
- Primary: OS
- Secondary: 2 and 5-yr. OS, RFS, safety, QOL
- 732 patients randomized to gem/cap (n=365) and gem (n=367)
- Baseline patient characteristics: 57% male, median age 65 yrs., 55% PS 2, 41% never smokers, 74% concurrent conditions, 74% non-diabetic, CA19-9 preop 150.5 KU/L, CA19-9 postop 18.7 KU/L, CRP preop 8 mg/L, postop CRP 5 mg/L, 40% R0, 60% R1, 50% moderately differentiated, 39% poorly differentiated, 80% lymph node positive, maximum tumor size 30 mm, 89% stage III, 51% Whipple, 34% pylorus-preserving resection, 14% venous resection, 15% radical resection, 78% standard resection, 72% cholecystectomy, 48% local invasion, 71% without postop complications
Table 1. Efficacy Outcomes
- Tumor recurrence (66% overall): 65% (gem-cap) vs. 66% (gem)
- Subgroup analysis:
- Factors significantly associated with survival on univariate analysis: smoking, preop and postop CA19-9, preop CRP, resection margin status, tumour grade, lymph nodes status, maximum tumour size, tumour stage, venous resection, local invasion
- Factors significantly associated with survival on multivariate analysis: resection margin status, postop CA19-9, tumour grade, lymph node status, maximum tumour size
- Median postop CA19-9: 17.7 KU/L in R0 vs. 20 KU/L in R1 groups (p=0.11)
Table 2. Survival by postop CA19-9 cutoff score
- Treatment summary
- Subsequent treatment: 33% vs. 39% Gem/cap group: chemo (31%) chemoRT (4%), surgery (3%) Gem group: chemo (32%), chemoRT (4%), surgery (5%)
- Completed 6 cycles of treatment: 54% vs. 65%
- Median dose intensity (planned protocol): 78% vs. 83%
- Treatment discontinuation before 6 cycles due to toxicity: 46% vs. 35%
- Most common grade 1-2 adverse events with gem-cap: fatigue (64%), anemia (58%), diarrhea (41%)
- Most common grade 3-4 adverse events with gem: fatigue (64%), anemia (56%), neutropenia (49%)
- QOL - No significant effect in QOL in both arms
Table 3. Adverse events
- At a median follow-up of 43.2 mos., adjuvant treatment with capecitabine combined with gemcitabine resulted in significant improvement in OS by 2.5 mos. (absolute difference) and 5 year survival was improved from 16.3% (gemcitabine alone) to 28.8% (gemcitabine + capecitabine). Early separation of OS curves was observed and greater benefit was derived with R0 resection margins. However, RFS did not differ between the treatment arms.
- Combination treatment increased toxicity as expected, particularly grade 3-4 neutropenia (38%) and diarrhea (5%). Most adverse events were low grade. Slightly more than half of the patients in each arm completed the planned 6 cycles of chemotherapy and up to 45% discontinued treatment due to toxicity. Overall, 36% went on to receive subsequent therapy.
- Postop CA19-9 concentration was a significant predictor of survival in this trial. Patients with lower postop CA19-9 levels derived the greatest benefit in OS with treatment. Similar findings were observed in the ESPAC-3(v1) trial (Br J Cancer 2009;100: 246–50).
- The combination of gemcitabine and capecitabine in the adjuvant setting has been shown in this study to provide a 5-year survival of 28.8%. This is higher than any 5-year survival statistics reported for 5-fluorouracil (21.1% in ESPAC-1, 15.9% in ESPAC3(v2)) and for gemcitabine alone (22.5% CONKO-001, 17.5% in ESPAC3(v2)).
- In summary, gemcitabine and capecitabine should be the new standard of care for adjuvant treatment of resected pancreatic adenocarcinoma in patients who are well enough to tolerate the combination treatment.
More GI Cancer Updates
Rucaparib in relapsed, platinum-sensitive high-grade ovarian carcinoma (ARIEL2 Part 1): an international, multicentre, open-label, phase 2 trial
Swisher EM et al., Lancet Oncol. 2017 Jan;18(1):75-87.
Rucaparib is an oral poly(ADP-ribose) polymerase (PARP) inhibitor that demonstrated clinical efficacy in relapsed, platinum-sensitive, high-grade ovarian cancer (Eur J Cancer 2015; 51: S545 (abstr 2746). Investigators conducted the ARIEL2 (part 1) trial to determine molecular predictors of response to rucaparib in this subgroup of patients.
- Patients: high-grade serous or endometrioid ovarian, fallopian tube, or primary peritoneal carcinoma, received at least one previous platinum therapy, ECOG 0-1, no brain metastases
- Study design: phase II, 2-part, single-arm, open-label trial
- Intervention: Rucaparib 600 mg PO 2x/day
- Next generation sequencing was used to classify patients into three groups:
- BRCA mutation positive
- BRCA wild type, loss of heterozygosity (LOH) high
- BRCA wild type, LOH low
- Primary: PFS
- Secondary: ORR, DOR, safety, pharmacokinetics
- 206 patients enrolled; BRCA mutant (n=40), BRCA wild-type and LOH high (n=82), BRCA wild-type and LOH low (n=70), BRCA wild-type and LOH unclassified (n=12)
- 10% with confirmed BRCA germline mutation (BRCA-homologous recombination assay), 9% with somatic BRCA mutation (BRCA 1 [n=14], BRCA2 [n=5])
- 13% (21/165) with BRCA1 promoter hypermethylation, 2% with RAD51C promoter hypermethylation
- Baseline patient characteristics: median age 64, 81% epithelial ovarian cancer, 98% serous histology, 62% at least 1 previous chemotherapy, 47% 6-12 mos. progression-free interval after platinum-based treatment
Table 1: Efficacy Outcomes
Table 2: Objective response rates by homologous deficiency subgroup
- RECIST responses were significantly higher in the BRCA mutant (p<0.0001) and LOH high (p=0.0033) subgroups.
- Similarly, RECIST and CA-125 responses were better in the BRCA mutant (p<0.0001) and LOH high (p=0.0018) subgroups.
- LOH classification was highly concordant between archival and pretreatment samples (r=0.86, p<0.0001) in 117 patients.
- BRCA1 methylation was also highly concordant in 90 paired samples (p<0.0001).
- Steady-state levels of rucaparib were achieved at cycle 1 day 15, with a mean trough plasma concentration of 2026 ng/mL.
- RECIST and CA-125 responses were detected in patients with a mutation in a non-BRCA homologous recombination gene.
- RECIST responses were detected in tumors with BRCA1 methylation and RAD51C methylation.
- Most common grade 3 or higher adverse events: anemia (22%) elevated ALT/AST (12%)
- Most common serious adverse events: small intestinal obstruction (5%), disease progression (5%), anemia (4%)
Table 3: Adverse events
- Rucaparib demonstrated clinical efficacy in relapsed, platinum-sensitive, high-grade ovarian cancer. The greatest magnitude of benefit was observed in the BRCA mutant subgroup where rucaparib reduced the risk of disease progression or death by 73%.
- A similar pattern was noted with tumor responses. BRCA mutant tumors had almost 3x higher responses compared to LOH high and LOH low subgroups. Response rates were comparable between somatic and germline BRCA mutation and BRCA1 and BRCA2 mutation. Although ORR was higher with RECIST + CA-125 compared to RECIST alone, there was a consistent trend in responses according to mutation status.
- Within the BRCA wild type subgroups, LOH high and LOH low tumors had similar PFS but response rates were higher and duration of response was longer in the LOH high group. The study however was not statistically powered to compare BRCA mutant and LOH subgroups. Furthermore, the small sample size per subgroup limits the interpretation of these findings.
- Serious adverse events were reported in a quarter of the study population; no treatment-related deaths were reported. Adverse events were manageable with supportive care and dose reductions.
- With the use of homologous recombination deficiency assay, this study identified predictors of response to PARP inhibitor rucaparib in relapsed, platinum-sensitive, high-grade ovarian cancer. Study results from the ARIEL2 study are provocative and warrant further investigation in larger, adequately powered trials. An ongoing phase III trial (ARIEL3) will determine the efficacy of maintenance rucaparib after platinum-based chemotherapy in recurrent ovarian cancer and determine whether genomic LOH is a predictive biomarker for response to rucaparib (NCT01968213).
More Updates in Gynecological Cancer