Chemotherapy of Advanced Small-bowel Adenocarcinoma: A Multicenter Study
A total of 154 patients with SBA were screened in the 13 centers. We excluded patients with completely resected disease (n = 44), patients who had not received chemotherapy, who were not evaluated for the tumor response (n = 11), or who had received a chemotherapy regimen other those specified for this study (n = 6). Therefore, the study population consisted of 93 patients (male, 53%; median age, 56 years; duodenal primary, 59%) with advanced SBA (metastatic stage, 86%) treated with one of the four pre-specified chemotherapy regimens (Table 1). Seventy-seven patients (83%) had previously received curative (n = 26) or palliative/exploratory (n = 51) surgery. Eighteen patients had received adjuvant chemotherapy (including two patients treated with adjuvant chemoradiotherapy).
There were no significant differences in patient characteristics across the chemotherapy regimen subgroups, except for the baseline CA 19-9 level (P = 0.047).
A total of 870 chemotherapy cycles were administered (median, 8; range, 2–37). Median follow-up was 26.7 months (95% CI 20.1–39.9 months). Eighteen patients were lost to follow-up, six before disease progression.
All the patients but one were assessable for toxicity. Severe toxicity was significantly more frequent in patients treated by LV5FU2-cisplatin compared with other subgroups of patients (P = 0.001). Severe adverse events were observed in 0 (0%), 22 (46%), 7 (39%) and 12 (75%) patients treated by LV5FU2, FOLFOX, FOLFIRI and LV5FU2-cisplatin, respectively. The main toxicity was hematologic. Neutropenia grade 3–4 was observed in 12 (25%), 4 (22%) and 6 (37%) patients treated by FOLFOX, FOLFIRI and LV5FU2-cisplatin, respectively. Neurotoxicity was observed only in patients treated by platinum-based chemotherapy. In the LV5FU2-cisplatin subgroup, two patients developed grade 3 sensory neurotoxicity. In the FOLFOX subgroup, nine patients discontinued oxaliplatin because of grade 3 neurotoxicity (n = 4), grade 2 neurotoxicity (n = 3), grade 2 allergic reactions (n = 1) or toxic death (n = 1).
The tumor response was assessable in 68 patients with measurable disease (Table 2). There were no complete responses and 18 partial responses. There was no significant difference in the ORR across the four chemotherapy regimen subgroups (P = 0.18).
Median PFS was 6.6 months (95% CI 5.0–7.7) overall, and, respectively, 7.7 months (95% CI 2.1 to not done), 6.9 months (95% CI 5.0–9.9), 6.0 months (95% CI 4.9–8.1) and 4.8 months (95% CI 2.2–8.1) in the LV5FU2, FOLFOX, FOLFIRI and LV5FU2-cisplatin subgroups (P = 0.16) (Figure 1A).
(Enlarge Image)
Figure 1.
Kaplan–Meier survival analysis of PFS and OS. PFS (A) and OS (B) according to the chemotherapy regimen in the entire population. PFS (C) and OS (D) according to the chemotherapy regimens among patients treated with platinum-based chemotherapy. PFS, progression-free survival; OS, overall survival.
Among the four potentially predictive variables selected by univariate analysis (predisposing conditions, PS, the CEA level and the chemotherapy regimen), two were independently associated with significantly longer PFS, namely PS (0–1 versus 2: HR, 3.26; 95% CI 1.75–6.09) and the baseline CEA level (normal versus increased: HR, 2.34; 95% CI 1.38–3.96) (Table 3).
Median OS was 15.1 months (95% CI 11.6–19.0) overall, and, respectively, 13.5 months (95% CI 4.1–34.4), 17.8 months (95% CI 14.2–24.2), 10.6 months (95% CI 8.1–28.3) and 9.3 months (95% CI 4.9–17.8) in the LV5FU2, FOLFOX, FOLFIRI and LV5FU2-cisplatin subgroups (P = 0.25) (Figure 1B).
Among the eight potentially predictive variables selected by univariate analysis (age, PS, prior surgery, disease stage, liver metastasis, primary site, CEA and CA 19-9 levels), three were independently associated with significantly longer OS, namely PS (0–1 versus 2: HR, 11.00; 95% CI 4.22–28.72), baseline CEA (normal versus increased: HR, 2.85; 95% CI 1.19–6.81) and baseline CA 19-9 (normal versus increased: HR, 2.26; 95% CI 1.06–4.78) (Table 3).
There was no significant difference in patient characteristics between the FOLFOX and LV5FU2-cisplatin subgroups (Table 1).
Median PFS was 6.8 months (95% CI 5.0–8.1) among all patients treated with platinum-based chemotherapy. Median PFS was significantly shorter in the LV5FU2-cisplatin group than in the FOLFOX group (4.8 versus 6.9 months; P = 0.02) (Figure 1C). Five potentially predictive variables were selected by univariate analysis (threshold, 20%): predisposing conditions (P = 0.14), PS (P = 0.0001), primary site (P = 0.16), CEA level (P = 0.03) and chemotherapy regimen (P = 0.03). Among these variables, two were independently associated with significantly longer PFS in multivariate analysis, namely PS (0–1 versus 2: HR, 6.84; 95% CI 2.86–16.34; P = 0.02) and the chemotherapy regimen (FOLFOX versus LV5FU2-cisplatin: HR, 2.25; 95% CI 1.12–4.50; P < 0.0001).
Median OS was 15.5 months (95% CI 13.7–20.4) overall, and was significantly shorter in the LV5FU2-cisplatin group than in the FOLFOX group (9.3 versus 17.8 months; P = 0.04) (Figure 1D). Five potentially predictive variables were selected by univariate analysis (threshold, 20%): age (P = 0.15), PS (P < 0.0001), prior surgery (P = 0.051), CEA level (P = 0.09) and chemotherapy regimen (P = 0.04). Among these variables, four were independently associated with significantly longer OS in multivariate analysis, namely age (≤65 versus >65 years: HR, 2.64; 95% CI 1.06–6.62; P = 0.04), PS (0–1 versus 2: HR, 14.20; 95% CI 4.65–43.11; P < 0.0001), the chemotherapy regimen (FOLFOX versus LV5FU2-cisplatin: HR, 2.75; 95% CI 1.18–6.41; P = 0.02) and prior surgery (none versus curative: HR, 0.21; 95% CI 0.06–0.73; P =0.049).
Among the patients who received first-line LV5FU2, FOLFOX, FOLFIRI or LV5FU2-cisplatin, respectively, 5 (50%), 26 (54%), 10 (53%) and 12 (75%) patients received a second line of chemotherapy. Most patients who received first-line platinum-based chemotherapy subsequently received the FOLFIRI regimen (n = 29; 76%), while patients who first received FOLFIRI usually received the FOLFOX regimen as second-line therapy (n = 5; 50%).
Results
Patient Characteristics
A total of 154 patients with SBA were screened in the 13 centers. We excluded patients with completely resected disease (n = 44), patients who had not received chemotherapy, who were not evaluated for the tumor response (n = 11), or who had received a chemotherapy regimen other those specified for this study (n = 6). Therefore, the study population consisted of 93 patients (male, 53%; median age, 56 years; duodenal primary, 59%) with advanced SBA (metastatic stage, 86%) treated with one of the four pre-specified chemotherapy regimens (Table 1). Seventy-seven patients (83%) had previously received curative (n = 26) or palliative/exploratory (n = 51) surgery. Eighteen patients had received adjuvant chemotherapy (including two patients treated with adjuvant chemoradiotherapy).
There were no significant differences in patient characteristics across the chemotherapy regimen subgroups, except for the baseline CA 19-9 level (P = 0.047).
A total of 870 chemotherapy cycles were administered (median, 8; range, 2–37). Median follow-up was 26.7 months (95% CI 20.1–39.9 months). Eighteen patients were lost to follow-up, six before disease progression.
toxicity
All the patients but one were assessable for toxicity. Severe toxicity was significantly more frequent in patients treated by LV5FU2-cisplatin compared with other subgroups of patients (P = 0.001). Severe adverse events were observed in 0 (0%), 22 (46%), 7 (39%) and 12 (75%) patients treated by LV5FU2, FOLFOX, FOLFIRI and LV5FU2-cisplatin, respectively. The main toxicity was hematologic. Neutropenia grade 3–4 was observed in 12 (25%), 4 (22%) and 6 (37%) patients treated by FOLFOX, FOLFIRI and LV5FU2-cisplatin, respectively. Neurotoxicity was observed only in patients treated by platinum-based chemotherapy. In the LV5FU2-cisplatin subgroup, two patients developed grade 3 sensory neurotoxicity. In the FOLFOX subgroup, nine patients discontinued oxaliplatin because of grade 3 neurotoxicity (n = 4), grade 2 neurotoxicity (n = 3), grade 2 allergic reactions (n = 1) or toxic death (n = 1).
Tumor Response
The tumor response was assessable in 68 patients with measurable disease (Table 2). There were no complete responses and 18 partial responses. There was no significant difference in the ORR across the four chemotherapy regimen subgroups (P = 0.18).
Progression-free Survival
Median PFS was 6.6 months (95% CI 5.0–7.7) overall, and, respectively, 7.7 months (95% CI 2.1 to not done), 6.9 months (95% CI 5.0–9.9), 6.0 months (95% CI 4.9–8.1) and 4.8 months (95% CI 2.2–8.1) in the LV5FU2, FOLFOX, FOLFIRI and LV5FU2-cisplatin subgroups (P = 0.16) (Figure 1A).
(Enlarge Image)
Figure 1.
Kaplan–Meier survival analysis of PFS and OS. PFS (A) and OS (B) according to the chemotherapy regimen in the entire population. PFS (C) and OS (D) according to the chemotherapy regimens among patients treated with platinum-based chemotherapy. PFS, progression-free survival; OS, overall survival.
Among the four potentially predictive variables selected by univariate analysis (predisposing conditions, PS, the CEA level and the chemotherapy regimen), two were independently associated with significantly longer PFS, namely PS (0–1 versus 2: HR, 3.26; 95% CI 1.75–6.09) and the baseline CEA level (normal versus increased: HR, 2.34; 95% CI 1.38–3.96) (Table 3).
Overall Survival
Median OS was 15.1 months (95% CI 11.6–19.0) overall, and, respectively, 13.5 months (95% CI 4.1–34.4), 17.8 months (95% CI 14.2–24.2), 10.6 months (95% CI 8.1–28.3) and 9.3 months (95% CI 4.9–17.8) in the LV5FU2, FOLFOX, FOLFIRI and LV5FU2-cisplatin subgroups (P = 0.25) (Figure 1B).
Among the eight potentially predictive variables selected by univariate analysis (age, PS, prior surgery, disease stage, liver metastasis, primary site, CEA and CA 19-9 levels), three were independently associated with significantly longer OS, namely PS (0–1 versus 2: HR, 11.00; 95% CI 4.22–28.72), baseline CEA (normal versus increased: HR, 2.85; 95% CI 1.19–6.81) and baseline CA 19-9 (normal versus increased: HR, 2.26; 95% CI 1.06–4.78) (Table 3).
PFS and OS after Platinum-based Chemotherapy
There was no significant difference in patient characteristics between the FOLFOX and LV5FU2-cisplatin subgroups (Table 1).
Median PFS was 6.8 months (95% CI 5.0–8.1) among all patients treated with platinum-based chemotherapy. Median PFS was significantly shorter in the LV5FU2-cisplatin group than in the FOLFOX group (4.8 versus 6.9 months; P = 0.02) (Figure 1C). Five potentially predictive variables were selected by univariate analysis (threshold, 20%): predisposing conditions (P = 0.14), PS (P = 0.0001), primary site (P = 0.16), CEA level (P = 0.03) and chemotherapy regimen (P = 0.03). Among these variables, two were independently associated with significantly longer PFS in multivariate analysis, namely PS (0–1 versus 2: HR, 6.84; 95% CI 2.86–16.34; P = 0.02) and the chemotherapy regimen (FOLFOX versus LV5FU2-cisplatin: HR, 2.25; 95% CI 1.12–4.50; P < 0.0001).
Median OS was 15.5 months (95% CI 13.7–20.4) overall, and was significantly shorter in the LV5FU2-cisplatin group than in the FOLFOX group (9.3 versus 17.8 months; P = 0.04) (Figure 1D). Five potentially predictive variables were selected by univariate analysis (threshold, 20%): age (P = 0.15), PS (P < 0.0001), prior surgery (P = 0.051), CEA level (P = 0.09) and chemotherapy regimen (P = 0.04). Among these variables, four were independently associated with significantly longer OS in multivariate analysis, namely age (≤65 versus >65 years: HR, 2.64; 95% CI 1.06–6.62; P = 0.04), PS (0–1 versus 2: HR, 14.20; 95% CI 4.65–43.11; P < 0.0001), the chemotherapy regimen (FOLFOX versus LV5FU2-cisplatin: HR, 2.75; 95% CI 1.18–6.41; P = 0.02) and prior surgery (none versus curative: HR, 0.21; 95% CI 0.06–0.73; P =0.049).
Second-line Chemotherapy
Among the patients who received first-line LV5FU2, FOLFOX, FOLFIRI or LV5FU2-cisplatin, respectively, 5 (50%), 26 (54%), 10 (53%) and 12 (75%) patients received a second line of chemotherapy. Most patients who received first-line platinum-based chemotherapy subsequently received the FOLFIRI regimen (n = 29; 76%), while patients who first received FOLFIRI usually received the FOLFOX regimen as second-line therapy (n = 5; 50%).
SHARE
![Breast Cancer Screening (PDQ®): Screening - Health Professional Information [NCI]-Special Population](https://cdnus.oss-us-west-1.aliyuncs.com/libs/images/NewsJpg/cate_6/039_215x160.jpg)
