Systemic Treatment for Metastatic or Recurrent Disease




RR (%)

PFS (months)

OS (months)


Cisplatin-based combinations

CF vs. C

Phase II


35 vs. 19

TTP: 27 vs. 18 weeks

2 years: 18 vs. 9 %


DCF vs. CF

Phase III

TTP: 5.6 vs. 3.7 m

2 years: 18 vs. 9 %



Phase II






Docet/oxal (DO) vs. DOF vs. DOX

Randomized phase II


23 vs. 47 vs. 26

4.5 vs. 7.7 vs. 5.6

9 vs. 13.6 vs. 11.3



Phase II






ECF vs. EOF vs. ECX vs. EOX

Phase III non-inferiority
9.9 vs. 9.3 vs. 9.9 vs. 11.2


CF vs. CX

Phase III non-inferiority


29 vs. 41

5.0 vs. 5.6

9.3 vs. 10.5
CF vs. FLO

Phase III
3.9 vs. 5.8

8.8 vs. 10.7


CF vs. IF

Phase III


26 vs. 32

TTP: 4.2 vs. 5

8.7 vs. 9


IF vs. cis/iri

Randomized phase II


42 vs. 32

TTP: 6.5 vs. 4.2

10.7 vs. 6.9


CF vs. cis/S-1

Phase III
7.9 vs. 8.6



Phase II



TTP: 7.0




Phase II

TTP: 6.6



Taxane-based combinations

Docet/iri vs. DF

Randomized phase II


38 vs. 33

4.2 vs. 4.4

8.6 vs. 9.4


DF vs. PF

Randomized phase II


33 vs. 42

FFS: 4.2 vs. 3.6

9.3 vs. 9.9


Irinotecan-based combinations

5-FU vs. CF vs. IF

Randomized phase II


13 vs. 27 vs. 40

3.2 vs. 4.9 vs. 6.9

6.8 vs. 9.5 vs. 11.3


Iri/S-1 vs. S-1

Phase III
42 vs. 27

TTF: 4.5 vs. 3.6

12.8 vs. 10.8


Anthracycline-based combinations

CF vs. 5-FU vs. FAM

Phase III
51 vs. 26 vs. 25
37 vs. 31 vs. 29 weeks



Phase II





Phase III


45 vs. 21
8.9 vs. 5.7


ECX vs. CX

Randomized phase II


38 vs. 37

6.4 vs. 6.5

Etoposide-based combinations


Phase II




20 %


ELF vs. CF vs. FAMTX

Phase III


9 vs. 20 vs. 12
7.2 vs. 7.2 vs. 6.7


C cisplatin; CF cisplatin and 5-FU; Cis/Iri cisplatin and irinotecan; Cis/Pacli cisplatin and paclitaxel; DCF docetaxel, cisplatin, and 5-FU; Docet/Iri docetaxel and irinotecan; DF docetaxel and 5-FU; DOF docetaxel, oxaliplatin, and 5-FU; DOX docetaxel, oxaliplatin, and capecitabine; EAP etoposide, doxorubicin, and cisplatin; ELF etoposide, leucovorin, and 5-FU; EOF epirubicin, oxaliplatin, and 5-FU; ECX epirubicin, cisplatin, and capecitabine; FAM 5-FU, doxorubicin, and mitomycin; FAMTX 5-FU, doxorubicin, and methotrexate; FLO 5-FU and oxaliplatin; IF irinotecan and 5-FU; PF paclitaxel and 5-FU

The REAL-2 study was a large randomized phase III trial that demonstrated that oxaliplatin was non-inferior to cisplatin and capecitabine was non-inferior to 5-FU [11] in patients with metastatic gastroesophageal cancer. In a two-by-two design, the trial randomized 1002 patients to receive epirubicin (E) plus CF (ECF), E plus oxaliplatin and 5-FU (EOF), E plus cisplatin and capecitabine (ECX), or E plus oxaliplatin and capecitabine (EOX). The median OS in the ECF, EOF, ECX, and EOX cohorts were 9.9, 9.3, 9.9, and 11.2 months, respectively. The non-inferiority of capecitabine to 5-FU was also confirmed in the ML17032 trial. ML17032 was a randomized non-inferiority open-label trial to compare capecitabine (1000 mg/m2 twice daily, days 1–14) and cisplatin 80 mg/m2 on day 1 (XP) with CF (5-FU 800 mg/m2/day days 1–5) administered every 3 weeks in 316 patients with metastatic gastric cancer [29]. RR (41 % versus 29 %) and median OS (10.5 months versus 9.3 months) were superior in the XP arm. Interestingly, median PFS was similar in both groups (XP = 5.6 months versus CF = 5.0 months). Of note, concerns have been raised about the generalizability of the results of the study since the median age of enrolled subjects (56 years; range 22–74) was significantly younger than most patients diagnosed with metastatic gastric cancer [30]. Nevertheless, these results are in line with other studies that suggest that capecitabine and 5-FU have equivalent efficacy in patients with gastroesophageal cancers. A meta-analysis of patients enrolled in the REAL-2 and ML17032 studies suggested that OS may be superior in 654 patients who received capecitabine-based regimens compared with 664 patients treated with 5-FU-based regimens [31]. There was, however, no significant difference in PFS between the two groups, a finding that makes interpretation of the OS results more difficult. It is worth noting that concerns have been raised about the safety and efficacy of capecitabine in patients who have undergone a gastrectomy [32], but these concerns do not seem to have been validated in retrospective and prospective studies [3335]. Similarly, a study by the Arbeitsgemeinschaft Internistische Onkologie (AIO) confirmed that oxaliplatin can be a reasonable alternative to cisplatin in the first-line setting. The phase III study compared 5-FU 2600 mg/m2 via a 24-h infusion, leucovorin 200 mg/m2 and oxaliplatin 85 mg/m2 (FLO), or 5-FU 2000 mg/m2 via 24-h infusion, leucovorin 200 mg/m2 and cisplatin 50 mg/m2 (CF). Both regimens were administered every 2 weeks. There was a trend toward improved median PFS with FLO versus CF (5.8 versus 3.9 months; p = 0.77) while OS was not statistically different (10.7 versus 8.8 months). FLO was better tolerated than CF. It was associated with significantly less anemia (54 % versus 72 %), nausea (53 % versus 70 %), vomiting (31 % versus 52 %), alopecia (22 % versus 39 %), fatigue (19 % versus 34 %), renal toxicity (11 % versus 34 %), thromboembolic events (0.9 % versus 7.8 %), and serious adverse events related to the treatment (9 % versus 19 %). FLO resulted, however, in increased incidence of peripheral neuropathy (63 % versus 22 %). In patients older than 65 years (N = 94), treatment with FLO was associated with significantly improved RR (41.3 % versus 16.7 %; p = 0.12), TTF (5.4 versus 2.3 months; p < 0.001), PFS (6.0 versus 3.1 months; p = 0.029), and OS (13.9 versus 7.2 months) compared to CF. Since the median age of diagnosis of gastric cancer in the United States and Europe is greater than 65 [2, 30], these results are important and suggest that the combination of 5-FU and oxaliplatin is better tolerated than CF and is associated with equal if not better efficacy.

Several studies have investigated the substitution of cisplatin with irinotecan [13, 14]. Dank et al. conducted a phase III study to compare irinotecan 80 mg/m2, leucovorin 500 mg/m2, and 5-FU 2000 mg/m2 via 22-h infusion (IF) for 6 out of 7 weeks versus cisplatin 100 mg/m2 and 5-FU 1000 mg/m2 via 24-h infusion day 1–5, every 4 weeks (CF) in 333 patients with adenocarcinoma of the stomach and GE junction. Compared to CF, TTF was improved with IF (4 versus 3.4 months; p = 0.018), while TTP (5 versus 4.2 months; p = 0.088), OS (9 versus 8.7 months), and RR (31.8 versus 25.8 %) were not significantly different. IF seemed to be better tolerated than CF and was associated with fewer toxicity-related deaths (0.6 % versus 3 %) and better rates of discontinuation for toxicity (10.0 % versus 21.5 %), severe neutropenia, thrombocytopenia, and stomatitis, but not diarrhea. The final statistical analysis indicated that IF did not provide a significant TTP or OS superiority over CF, and the results of non-inferiority of IF were borderline. These results were less impressive than those reported with the comparison of oxaliplatin versus cisplatin [12]. They, however, suggest that irinotecan has activity in metastatic esophagogastric cancer and may be considered in select situations.

As a result of this and other trials, irinotecan was evaluated in the first- and second-line setting for patients with advanced esophagogastric cancer. Pozzo et al. [14] conducted a randomized phase II study that compared irinotecan 80 mg/m2, leucovorin 500 mg/m2, and 5-FU 2000 mg/m2 via 22-h infusion weekly for 6 out of 7 weeks versus irinotecan 200 mg/m2 and cisplatin 60 mg/m2 given every 3 weeks in 115 patients with GEJ and gastric cancer. Median TTP (6.5 versus 4.2 months; p < 0.0001), OS (10.7 versus 6.9 months; p = 0.0018), and overall RR (42.4 % versus 32.1 %) were superior in the 5-FU-containing arm. Additionally, grade 3/4 toxicity was more pronounced in the cisplatin/irinotecan arm (65.7 % versus 27 %). As expected, however, diarrhea was more common in the 5-FU/irinotecan arm (27 % versus 18 %). Taken together, these results suggest that irinotecan has encouraging activity in the first-line setting in esophagogastric cancer, but the combination of 5-FU and irinotecan may be preferred over the combination of irinotecan with cisplatin.

S-1 was also investigated in combination with cisplatin in the first-line setting of patients with advanced gastric and gastroesophageal cancer. The FLAGS trial [15] was a multinational phase III study that compared cisplatin 100 mg/m2 and 5-FU 1000 mg/m2/day for 120 h versus cisplatin 75 mg/m2 and S-1 25 mg/m2 twice daily for 21 days in 1053 patients. Both regimens were repeated every 28 days. Although cisplatin/S-1 was found to be associated with a favorable adverse effect profile (grade 3/4 neutropenia [32.3 % versus 63.6 %], complicated neutropenia [5.0 % versus 14.4 %], stomatitis [1.3 % versus 13.6 %], and treatment-related deaths [2.5 % versus 4.9 %; p < 0.05]), it did not prolong OS (8.6 versus 7.9 months; HR 0.92; 95 % CI 0.80–1.05; p = 0.20) compared with cisplatin/5-FU.

The combination of cisplatin and paclitaxel has been investigated in several phase II trials [16, 3639], most of which included predominantly patients with advanced esophageal cancer. Kornek et al. [16] conducted a phase II study of paclitaxel 160 mg/2 and cisplatin 60 mg/m2, both given every 2 weeks in 45 patients with metastatic gastroesophageal cancer. Of note, the study allowed for granulocyte colony-stimulating factor and erythropoietin to be administered for neutropenia and anemia, respectively. The confirmed RR was 44 %. The median TTP and median OS were 7 and 11.2 months, respectively. Grade 4 neutropenia occurred in 4 (11 %) of patients. Non-hematologic adverse reactions included grade 3 peripheral neuropathy in 6 patients (13 %) and anaphylaxis in 2 patients, and one patient each experienced grade 3 emesis, diarrhea, and infection, respectively. These results are intriguing but should be interpreted with caution given the small size of the trial and the potential for selection bias. Other phase II trials with modifications of this regimen in metastatic esophageal cancer raised concerns about hematological toxicity especially neutropenia [38]. The combination of carboplatin and paclitaxel seems to be better tolerated [40] but has not been systematically evaluated in a randomized fashion in advanced esophageal cancer.

Similarly, the combination of cisplatin and docetaxel has been evaluated in phase II trials. Roth et al. [17] conducted a phase II trial to evaluate the outcomes of docetaxel 85 mg/m2 and cisplatin 75 mg/m2 administered every 3 weeks. Dose escalation of docetaxel to 100 mg/m2 was performed in 5 patients and was discontinued for excessive toxicity. The median TTP and OS were 6.6 and 9 months, respectively. RR was 56 % (95 % CI 41–71 %). Hematological toxicity was, however, pronounced. Grade ≥3 toxicities were neutropenia 81 %, anemia 32 %, and thrombocytopenia 4 %. The study included 9 episodes of febrile neutropenia in 8 patients, 2 of them with docetaxel at 100 mg/m2.

Non-hematological toxicity included alopecia 36 %, fatigue 9 %, mucositis 9 %, diarrhea 6 %, nausea/vomiting 4 %, neurologic 2 %, and one episode of anaphylaxis precluding treatment administration. Other phase II trials evaluated similar regimens with lower doses of docetaxel and showed reasonable efficacy and better toxicity profiles [4143]. Taken together, these results indicate that the taxanes demonstrated encouraging activity in esophageal cancer, but their combination with cisplatin may be associated with excessive toxicity.

Platinum agents continue to play a pivotal role in the first-line treatment of patients with esophageal cancer. The bulk of the evidence suggests that oxaliplatin is at least as effective and significantly better tolerated than cisplatin and therefore may be considered the preferred platinum agent in the first-line setting in combination with 5-FU. Carboplatin has not been as rigorously investigated and therefore there is not sufficient data to recommend for or against it.

Taxane-Based Combinations

Taxanes have been evaluated in various combinations in the first-line setting in metastatic gastroesophageal cancer. As detailed above, taxanes have been combined with platinum agents in regimens such as DCF [7], cisplatin/docetaxel [17, 4143], and cisplatin/paclitaxel [16, 3639]. Additionally, taxanes have been evaluated with other chemotherapeutic agents such as irinotecan and 5-FU. A randomized phase II study evaluated a three-weekly regimen of docetaxel 60 mg/m2 and irinotecan 250 mg/m2 (DI) versus docetaxel 85 mg/m2 and 5-FU 750 mg/m2/day as a continuous infusion days 1–5 (DF) in 85 patients with advanced esophagogastric cancer [18]. The ORR and TTP for DI were 37.5 % and 4.2 months, respectively. The ORR and TTP for DF were 33.3 % and 4.4 months, respectively. The median OS was 8.6 months (95 % CI 6.1–12.2) with DI and 9.4 months (95 % CI 7.7–11.0) with DF, which was not statistically significantly different. It is important to note, however, that the ORR was the primary endpoint of the trial. Grade 3–4 neutropenia, febrile neutropenia, and diarrhea were more frequent in the DI arm as compared with the DF arm (83.3 % versus 69.8 %, 40.5 % versus 18.6 %, and 42.9 % versus 16.3 %, respectively). Docetaxel may, however, be associated with greater toxicity than paclitaxel. A randomized phase II study evaluated the combination of 5-FU with docetaxel (DF) or paclitaxel (PF) in the first-line setting in advanced gastric cancer [19]. Seventy-seven patients received three-weekly cycles of paclitaxel 175 mg/m2 or docetaxel 75 mg/m2 in combination with 5-FU 500 mg/m2 continuous infusion on days 1–5. Dose reduction was required more frequently with DF versus PF (19 % versus 9 %, respectively). Grade 3 or 4 toxicity was more common with DF versus PF (85 % versus 68 %), but this difference was not statistically significant (p = 0.09). PF and DF were associated with similar response rates (42 % versus 33 %, respectively; p = 0.53), failure-free (3.6 versus 4.2 months; p = 0.92), and OS (9.9 versus 9.3 months; p = 0.42). These results suggest that paclitaxel and docetaxel have comparable efficacy but paclitaxel may be associated with a more favorable side effect profile. Taken together, these data suggest that combinations of taxanes with other chemotherapeutic options are active in first-line setting in esophagogastric cancer. Other combinations, however, such as 5-FU and oxaliplatin [12], may be associated with more favorable side effect profiles and may therefore be better options in this setting.

Irinotecan-Based Combinations

Several studies have suggested that irinotecan has encouraging activity in the first-line setting in patients with metastatic gastroesophageal cancer [13, 14, 18, 44]. As discussed above, the combination of irinotecan with platinum agents or taxanes seems to have reasonable activity but possibly more adverse events than other combinations such as 5-FU and oxaliplatin. 5-FU-irinotecan combinations seem to offer the best toxicity profile compared to other combinations such as cisplatin and irinotecan [13, 20, 41]. A randomized phase II study compared LV 200 mg/m2 (2-h infusion) followed by a bolus of 5-FU 400 mg/m2 and FU 600 mg/m2 as a 22-h continuous infusion on days 1 and 2 every 14 days (LV5FU2; arm A), LV5FU2 plus cisplatin 50 mg/m2 as 1-h infusion on day 1 or 2 (arm B), or LV5FU2 plus irinotecan 180 mg/m2 as a 2-h infusion on day 1 (arm C) in 136 patients with advanced gastroesophageal cancer [20]. The RR for arms A, B, and C were 13 % (95 % CI 3.4 % to 23.3 %), 27 % (95 % CI 14.1 % to 40.4 %), and 40 % (95 % CI 25.7 % to 54.3 %), respectively. Median PFS was 3.2 months (95 % CI 1.8–4.6 months), 4.9 months (95 % CI 3.5–6.3 months), and 6.9 months (95 % CI 5.5–8.3 months), and the OS was 6.8 months (95 % CI 2.6–11.1 months), 9.5 months (95 % CI 6.9–12.2 months), and 11.3 months (95 % CI 9.3–13.3 months) for arms A, B, and C, respectively. Treatment was discontinued as a result of toxicity in 4 %, 16 %, and 11 % of patients in arms A, B, and C, respectively. Longitudinal analysis of quality of life scores using QLQ-C30 quality of life questionnaire indicated that mean scores were higher in arm C than in arms A and B which suggests that the combination of 5-FU and irinotecan is associated with better efficacy and QOL compared to the other treatments tested [20]. Similar outcomes to the combination of irinotecan with 5-FU have been observed with the combinations of irinotecan and oral capecitabine or S-1 [21, 4547]. A randomized phase III study was conducted in Japan to compare the efficacy and safety of S-1 monotherapy versus irinotecan and S-1 [21]. Although the RR was superior with the combination of irinotecan and S-1 (41.5 % versus 26.9 %), OS (12.8 versus 10.8 months; p = 0.23) and TTF (4.5 versus 3.6 months; p = 0.157) were equivalent between the two groups. S-1 is not commercially available in the United States. Taken together, these data suggest that, compared to other irinotecan-containing combinations, the combination of 5-FU and irinotecan may offer the best balance of efficacy and tolerability. Further studies need to compare 5-FU and irinotecan with other non-irinotecan combinations such as 5-FU and oxaliplatin.

Anthracycline-Containing Combinations

One of the first multidrug combinations in advanced esophagogastric cancer was the combination of 5-FU, doxorubicin (A), and mitomycin (FAM), which achieved an RR of 40 % and a median survival of 5.5 months [48]. However, a randomized trial conducted by the NCCTG to compare FAM, FA, or 5-FU alone in 305 patients with metastatic gastric and pancreatic cancer demonstrated equivalent RR and OS for patients treated with any of the three regimens [49]. Similarly, a phase III randomized trial conducted in Korea compared 5-FU and cisplatin (CF) with 5-FU alone or FAM [22]. The RR of 5-FU and FAM were equivalent (26 % versus 25 %) and inferior to CF (51 %). However, OS was not statistically different between the three groups (37 versus 31 versus 29 weeks, respectively). Other studies suggested that the anthracycline-containing regimens were associated with impressive clinical activity. The combination of epirubicin 50 mg/m2 and cisplatin 60 mg/m2 was administered every 3 weeks for 8 cycles during a 21-week continuous intravenous infusion of 5-fluorouracil 200 mg/m2/day (ECF) and was associated with an RR of 71 % in a phase II study that included 139 patients with advanced gastroesophageal cancer [23]. The median OS was 8.2 months and the 1- and 2-year OS rates were 30 % and 10 %, respectively. Grade 3 or 4 vomiting occurred in 13 %, stomatitis in 7 %, diarrhea in 4 %, infection in 6 %, leukopenia in 21 %, and thrombocytopenia in 8 % of patients. This study was associated with impressive RR, but the OS data did not seem to be significantly better than that seen with two-drug combinations. A second trial compared ECF with 5-FU, doxorubicin, and methotrexate (FAMTX) in 274 patients with metastatic gastroesophageal cancer [24]. ECF was associated with superior RR 45 % (95 % CI 36–54 %) versus 21 % (95 % CI 13–29 %; p = .0002) with FAMTX. The median OS was 8.9 months with ECF and 5.7 months with FAMTX (p = .0009).

A meta-analysis including trials that compared CF versus CF plus an anthracycline combination suggested that the addition of anthracyclines may be associated with an improvement in OS (HR 0.77; 95 CI 0.62–0.91) [4, 50, 51]. These results, however, were based on older regimens that predominantly included 5-FU bolus infusions. It is thought that newer regimens involving high-dose infusional 5-FU such as 5-FU and oxaliplatin (FOLFOX) may be more effective than the older regimens that were included in the meta-analysis. In order to confirm this hypothesis, Yun et al. conducted a randomized phase II study to compare epirubicin, cisplatin, and capecitabine (ECX) and cisplatin and capecitabine (CX) in 91 patients with advanced gastric cancer [25]. Patients randomized to ECX and CX experienced similar RR (38 % versus 37 %, respectively) and PFS (6.4 versus 6.5 months), respectively. ECF is a commonly used regimen in the United States and Europe, but it remains a matter of debate if it offers any advantage over newer regimens such as FOLFOX.

Etoposide-Containing Regimens

Early reports using etoposide-containing combinations demonstrated encouraging results. A phase II study of etoposide, doxorubicin, and cisplatin (EAP) in 55 patients with advanced gastric cancer demonstrated an RR of 64 % and median OS of 9 months [26]. A follow-up study, however, was associated with a significantly lower RR of 33 % [52]. A subsequent study compared EAP with FAMTX and demonstrated that EAP was associated with an unimpressive RR of 20 % and a median OS of 6.1 months. Additionally, EAP was associated with 4 toxicity-related deaths [27]. ELF (etoposide, 5-FU, and leucovorin) was also evaluated with initial encouraging results that were not confirmed in subsequent studies [28, 53, 54]. A study by the EORTC compared the efficacy and safety of ELF versus CF versus FAMTX in 399 patients with advanced gastric cancer [28]. All three regimens were associated with poor RR (9 % with ELF, 20 % with CF, and 12 % with FAMTX). They were also associated with equivalent OS (7.2, 7.2, and 6.7 months, respectively). These data suggest that etoposide-containing regimens lack any meaningful activity and therefore should not be used outside of the clinical trial setting.

Second Line

Approximately 20–40 % of patients who progress on first-line chemotherapy receive second-line treatment [55]. There is no established consensus on the optimal second-line chemotherapy so far [3]. Several studies have evaluated the use of cytotoxic agents as monotherapy or in combination with variable degrees of success [3]. Since many patients receive fluoropyrimidines and platinum agents in the first line [11], investigators have evaluated the use of taxanes and irinotecan in the second-line setting [56, 57] (Table 14.2).

Table 14.2
Clinical trials evaluating cytotoxic chemotherapy in second-line treatment of advanced esophageal/gastric cancer




RR (%)

PFS (months)

OS (months)


Docetaxel vs. BSC

Phase III

5.2 vs. 3.6 m



Phase III

7.36 m


Irinotecan vs. BSC

Phase III


4.0 vs. 2.4

Docet vs. iri vs. PEP02

Randomized phase II


16 vs. 7 vs. 14
7.7 vs. 7.8 vs. 7.3


Docet vs. iri vs. BSC

Phase III

5.2 vs. 6.5 vs. 3.8 m


Paclitaxel vs. iri

Phase III


21 vs. 1

3.6 vs. 2.3

9.5 vs. 8.4 m


BSC best supportive care, Docet docetaxel, Iri irinotecan, PFS progression-free survival, RR response rate, vs. versus


The COUGAR-02 study was a phase III multicenter trial that evaluated the use of docetaxel 75 mg/m2 intravenously every 3 weeks versus active symptom control in 168 patients with esophageal or gastric adenocarcinoma who had progressed on or within 6 months of a platinum-fluoropyrimidine combination. Median OS in the docetaxel group was 5.2 months (95 % CI 4.1–5.9) versus 3.6 months (95 % CI 3.3–4.4) in the active symptom control group (HR 0.67, 95 % CI 0.49–0.92; p = 0.01). As expected, docetaxel was associated with higher incidence of grade 3–4 neutropenia (15 % versus 0 %), infection (19 % versus 3 %), and febrile neutropenia (7 % versus 0 %). Patients receiving docetaxel experienced better symptom control and reported less pain (p = 0.0008) and less nausea and vomiting (p = 0.02) and constipation (p = 0.02). Global health-related quality of life (HRQoL) was not worsened by chemotherapy (p = 0.53). However, disease-specific HRQoL measures were in favor of docetaxel therapy in reducing dysphagia (p = 0.02) and abdominal pain (p = 0.01). It is worth noting that paclitaxel is thought to have similar efficacy to docetaxel but is associated with a more favorable adverse effect profile [19]. The beneficial effect of paclitaxel alone or in combination with biological agents was demonstrated in the REGARD trial, a phase III trial that randomized 355 patients with GEJ and gastric adenocarcinoma who have progressed on first-line platinum- or fluoropyrimidine-containing agents to paclitaxel alone or in combination with the vascular endothelial growth factor (VEGF) receptor blocker, ramucirumab [61]. Paclitaxel was administered intravenously at a dose of 80 mg/m2 on days 1, 8, and 15 of a 4-week cycle. Median PFS, TTP, and OS with paclitaxel were 2.86, 3.0, and 7.36 months, respectively. Grade ≥3 adverse events occurring in >5 % of patients on paclitaxel were neutropenia 18.8 %, febrile neutropenia 2.4 %, leukopenia 6.7 %, hypertension 2.4 %, anemia 10.3 %, fatigue 4.0 %, abdominal pain 3.3 %, and asthenia 3.3 %. In an attempt to improve survival outcomes of patients treated with second-line chemotherapy, investigators have attempted to combine taxanes with other agents such as irinotecan, capecitabine, or platinum agents or with cisplatin and fluorouracil [6266], but combination therapy with cytotoxic chemotherapeutics did not seem to offer any benefit over monotherapy.


Since irinotecan is typically not used in the first-line setting and given its promising activity, investigators have evaluated its benefit in the second-line setting. A phase III study was conducted to compare irinotecan 250 mg/m2 intravenously every 3 weeks to be increased to 350 mg/m2 depending on toxicity versus best supportive care (BSC) in the second-line therapy of patients with advanced cancers of the GEJ and stomach [67]. The study was unfortunately closed prematurely due to poor accrual. Nevertheless, analysis of data from the 40 patients that were accrued on the study showed clinically meaningful results. The median OS was 4.0 months (95 % CI 3.6–7.5) with irinotecan and 2.4 months with BSC (HR 0.48; 95 % CI 0.25–0.92; p = 0.012). Treatment with irinotecan did not result in any objective responses but SD was observed in 53 % of patients. Improvement of tumor-related symptoms occurred in 50 % of patients treated with irinotecan and 7 % of patients treated with BSC. These results were confirmed in several phase III studies that showed that taxanes and irinotecan have comparable efficacy in the second-line setting of patients with gastroesophageal cancer. Roy et al. conducted a randomized phase II study that compared docetaxel 75 mg/m2, irinotecan 300 mg/m2, and PEP02, a liposomal formulation of irinotecan, all given every 3 weeks in the second-line therapy of 44 patients with advanced GEJ or gastric adenocarcinoma [59]. The RR was comparable in all three arms: 15.9 % (95 % CI 6.6–30.1) with docetaxel, 6.8 % (95 % CI 1.4–18.7) with irinotecan, and 13.6 % (95 % CI 5.2–27.4) with PEP02. The median OS was 7.3 months (95 % CI 3.84–9.17) in the PEP02 arm, 7.8 months (95 % CI 4.90–9.20) in the irinotecan arm, and 7.7 months (95 % CI 5.32, 12.32) in the docetaxel arm. Median PFS was similar in all the three arms [2.7 months (95 % CI 1.54–3.65) with PEP02, 2.6 months (95 % CI 1.48–4.34) with irinotecan, and 2.7 months (95 % CI 1.41–5.45) with docetaxel]. Similarly Kang et al. conducted a randomized phase III study to compare docetaxel 60 mg/m2 every 3 weeks or irinotecan 150 mg/m2 every 2 weeks or BSC in 202 patients with pretreated gastric cancer [57]. The study demonstrated that salvage chemotherapy was superior to BSC but was not powered to compare the two chemotherapy arms. The median OS increased from 3.8 months (95 % CI 3.1–4.5) in the BSC arm to 5.3 months (95 % CI 4.1–6.5) in the salvage chemotherapy arm. An exploratory analysis demonstrated that the median OS of patients treated with docetaxel (5.2 months; 95 % CI 3.8–6.6) was not statistically different than that of patients treated with irinotecan (6.5 months; 95 % CI 4.5–8.5; two-sided p = .116). A Japanese phase III study also compared biweekly irinotecan administered at a dose of 150 mg/m2 with paclitaxel at a dose of 80 mg/m2 on days 1, 8, and 15 every 4 weeks in 223 patients with advanced gastric cancer refractory to platinum and fluoropyrimidines [60]. Median OS was 9.5 months in patients treated with paclitaxel and 8.4 months in patients treated with irinotecan (HR 1.13; 95 % CI 0.86–1.49; p = .38). Median PFS was 3.6 months in the paclitaxel group and 2.3 months in the irinotecan group (HR 1.14; 95 % CI 0.88–1.49; p = .33). Response rate was 20.9 % in the paclitaxel group and 13.6 % in the irinotecan group (p = .24). Grade 3 to 4 adverse events in the paclitaxel and irinotecan groups were neutropenia (28.7 % versus 39.1 %), anemia (21.3 % versus 30.0 %), and anorexia (7.4 % versus 17.3 %), respectively. Treatment-related deaths occurred in two patients (1.8 %) in the irinotecan group. Of note, 89.8 % of patients in the paclitaxel group and 72.1 % of patients in the irinotecan group received third-line chemotherapy (p = .001). It is worth noting that two randomized trials have compared irinotecan monotherapy with the combination of irinotecan and cisplatin or irinotecan and 5-FU (FOLFIRI regimen) and showed that the combination was not associated with improved OS [68, 69]. These results suggested that single-agent therapy with irinotecan and taxanes has comparable efficacy in the second-line setting of patients with gastroesophageal cancer, and both are considered appropriate choices. The combination of either agent with additional cytotoxic chemotherapy does not appear to be associated with improved outcomes.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Nov 27, 2016 | Posted by in ONCOLOGY | Comments Off on Systemic Treatment for Metastatic or Recurrent Disease

Full access? Get Clinical Tree

Get Clinical Tree app for offline access