Gemcitabine/nab-paclitaxel plus S-1 combination compared with gemcitabine/nab-paclitaxel in advanced pancreatic ductal adenocarcinoma: a retrospective study

Abstract

Background:

Pancreatic adenocarcinoma is one of the leading causes of cancer-related mortality. The median overall survival (OS) of patients with advanced pancreatic ductal adenocarcinoma (PDAC) is less than 1 year. The recent phase III NAPOLI-3 study showed that triplet chemotherapy improved OS compared to the commonly used gemcitabine/nab-paclitaxel (GA) regimen. S-1 is widely used in Asia for the treatment of PDAC, and a combination strategy with S-1 has shown encouraging results in previous studies.

Objectives:

We aimed to determine whether the addition of S-1 to GA offers more benefits than GA alone as a first-line treatment for advanced PDAC.

Design:

This was a retrospective study of patients with advanced PDAC treated at the Taichung Veterans General Hospital, Taiwan. All patients received first-line GA-based therapy, including GA alone or GA in combination with S-1 (GAS).

Methods:

Retrospectively reviewed clinical characteristics and outcomes, including the objective response rate (ORR), progression-free survival (PFS), and OS, were compared between the two groups. Some patients underwent genetic testing using the FoundationOne CDx (Foundation Medicine). The effects of the treatment on major genetic alterations were also analyzed.

Results:

Between January 1, 2022, and February 10, 2025, 89 patients were enrolled. Among these, 45 and 44 underwent GA and GAS, respectively. The median follow-up time was 10.2 months. ORR was significantly higher in patients who received GAS than in those who did not (48% vs 18%, p = 0.001). GAS showed both significant improvement of PFS (10.0 vs 5.2 months; p = 0.004) and OS (not reached vs 8.5 months; p = 0.016) compared with GA. Multivariate analysis identified only the treatment regimen as an independent prognostic factor for OS. GAS also improved the survival of patients harboring major genetic alterations. The Grade ⩾3 hematological adverse events were more in GAS, according to the CTCAE, version 5.0. No treatment-related mortality was found.

Conclusion:

The S-1 could be combined with GA as another potential triplet combination treatment option for advanced PDAC.

Plain language summary

Adding S-1 to standard chemotherapy with gemcitabine and nab-paclitaxel improved survival for people with advanced pancreatic cancer

Pancreatic cancer is one of the most deadly cancers, and most patients with advanced disease live less than a year. Standard treatment often includes two drugs: gemcitabine and nab-paclitaxel. A newer drug called S-1 is also used in Asia, and adding it to this standard treatment may improve outcomes. We looked back at medical records of 89 patients with advanced pancreatic cancer treated at Taichung Veterans General Hospital between 2022 and 2025. Patients received either gemcitabine plus nab-paclitaxel (GA) or the same treatment with S-1 added (GAS). The results showed that patients who received GAS had better tumor response, longer time before the cancer worsened, and lived longer compared to those who received GA. These benefits were also seen in patients with common genetic changes in their tumors. The GAS treatment caused more side effects affecting blood counts, but no deaths were linked to treatment. Overall, adding S-1 to standard chemotherapy may be a promising option for people with advanced pancreatic cancer.

Keywords

combination chemotherapy S-1 advanced pancreatic cancer prognostic biomarker

Introduction

Metastatic pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related mortality worldwide, with a 5-year relative survival rate of only 13.3%.¹ Despite recent advancements in chemotherapy, treatment outcomes remain unsatisfactory, and there is a persistent need for more effective first-line regimens.

Gemcitabine monotherapy, initially established in 1997, has demonstrated modest survival benefits over 5-fluorouracil (5-FU) and has become a standard of care.² Subsequent studies identified FOLFIRINOX (irinotecan, oxaliplatin, 5-FU, and leucovorin) as a superior alternative in selected patients with good performance status and normal bilirubin levels.³ In addition, the MPACT trial showed that combining gemcitabine with nab-paclitaxel (GA) significantly improved survival compared with gemcitabine alone.⁴ Recently, the phase III NAPOLI-3 trial demonstrated that NALIRIFOX (liposomal irinotecan, oxaliplatin, 5-FU, and leucovorin) significantly prolonged both progression-free survival (PFS) and overall survival (OS) compared to the GA regimen,⁵ further emphasizing the potential of multiagent chemotherapy in improving outcomes.

S-1, an oral fluoropyrimidine derivative widely used in Asia, has demonstrated efficacy in both gastrointestinal and non-gastrointestinal cancers. S-1 consists of tegafur, a prodrug of 5-FU, combined with two biochemical modulators of 5-FU: 5-chloro-2,4-dihydroxypyridine (gimeracil or CDHP), a competitive inhibitor of dihydropyrimidine dehydrogenase, and oteracil potassium, which inhibits the phosphorylation of 5-FU.⁶ In pancreatic cancer, S-1 demonstrated better tolerability and non-inferiority to gemcitabine in a large phase III trial.⁷ Although the combination of gemcitabine and S-1 was not superior to gemcitabine monotherapy, meta-analysis suggested a significantly improved objective response rate (ORR).⁸ Furthermore, phase II trials conducted in Taiwan evaluating triplet combinations with gemcitabine, nab-paclitaxel, S-1, and leucovorin (GASL) have shown a promising ORR of up to 53.5%, with an acceptable safety profile.⁹

Given the strong evidence supporting multiagent chemotherapy strategies, our study aimed to investigate whether the GAS regimen (gemcitabine, nab-paclitaxel, and S-1) improves survival outcomes compared to the GA regimen (gemcitabine and nab-paclitaxel) as a first-line treatment for patients with advanced PDAC in real-world conditions.

Methods

Patients

This study enrolled patients with either metastatic or locally advanced PDAC, including those with de novo advanced disease and those with recurrence after curative surgery between January 1, 2022, and February 10, 2025. Patients who received neoadjuvant or adjuvant therapy during the early disease stages were also included. Patients with confirmed recurrence during adjuvant therapy or within 6 months of the completion of adjuvant therapy were excluded. All patients received at least one cycle of GA or GAS as the first-line treatment for advanced-stage disease.

Clinical data were retrospectively collected, including age, sex, disease status (metastatic or locally advanced), Eastern Cooperative Oncology Group (ECOG) performance status, presence of peritoneal or liver metastasis, primary tumor location, and baseline levels of carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) prior to the initiation of systemic chemotherapy. Subsequent treatment after progression to the GA or GAS regimens and adverse events were recorded.

Treatment and assessments

The GAS regimen consisted of gemcitabine 800 mg/m², nab-paclitaxel 125 mg/m², and S-1 administered at 2-week intervals. The S-1 dosage was given at 30 mg for a body-surface area (BSA) <1.25 m², 40 mg for a BSA between 1.25 and 1.5 m², and 50 mg for a BSA ⩾1.5 m², given twice daily for 7 days. The GA regimen consisted of gemcitabine 1000 mg/m² and nab-paclitaxel 125 mg/m², administered either every 2 weeks or on days 1 and 8 of a 3-week cycle. Dose adjustments at treatment initiation and the choice of either regimen were based on the treating physicians. A conversion surgery was also allowed if the patient responded well to systemic treatment. Conversion surgery was defined as surgery performed with curative intent in patients who initially present with advanced-stage disease but become eligible for resection after demonstrating a favorable response to frontline treatment.

Treatment response was assessed using computed tomography or magnetic resonance imaging, according to the Response Evaluation Criteria in Solid Tumors version 1.1. The ORR was defined as the proportion of patients who achieved a partial or complete response. PFS signifies the duration from the initiation of chemotherapy to imaging-defined progression or death. OS was defined as the time from the initiation of chemotherapy to death from any cause. Treatment-related adverse events were graded according to the CTCAE, version 5.0.

Molecular testing

Next-generation sequencing was performed on 55 patients using tissue-based panel testing by FoundationOne CDx (Foundation Medicine, Inc., Cambridge, MA, USA). Common genetic alterations in PDAC, including mutations in KRAS, TP53, CDKN2A, SMAD4, and ARID1A, were analyzed. These molecular profiles were incorporated into the subgroup analyses of OS.

Statistical analyses

Patient characteristics are summarized using means for normally distributed continuous variables and medians with interquartile ranges for continuous variables. Categorical variables are reported as frequencies and percentages. Categorical variables were compared using the Chi-squared test or Fisher’s exact test, as appropriate. Continuous variables were compared using nonparametric tests such as the Mann–Whitney U test.

Survival analysis was conducted using the Kaplan–Meier method, and differences in survival curves were assessed using the log-rank test. Univariate and multivariate Cox proportional hazards models were used to identify the factors associated with OS in patients with advanced PDAC. Variables with a p-value <0.1 in the univariate analysis were included in the multivariate model. All statistical tests were two-sided, and a p-value <0.05 was considered statistically significant.

This study was approved by the Institutional Review Board of Taichung Veterans General Hospital (approval number: CE25226B). The requirement for informed consent was waived because of the retrospective nature of the study. The reporting of this retrospective cohort study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.¹⁰ The completed STROBE checklist is provided as Table S1.

Results

Patients’ characteristics

A total of 89 patients with histologically confirmed advanced PDAC were retrospectively reviewed, with a median follow-up time of 10.2 months. Among these, 45 received GA and 44 received GAS combination therapy. Most baseline characteristics were comparable between the two groups, except for ECOG performance status (Table 1); a higher proportion of patients in the GA group were classified as ECOG performance status ⩾2 (p = 0.022). No statistically significant differences were observed in age, sex, tumor location, metastatic condition, and baseline CEA and CA19-9 in advanced disease status or major genetic alterations.

Table 1.

Baseline characteristics.

Clinical variables	GAn = 45(%)		GASn = 44(%)		p-Value
Age					0.909
<65	21	(46.7)	20	(45.5)
⩾65	24	(53.3)	24	(54.5)
Sex					0.167
Male	30	(66.7)	23	(52.3)
Female	15	(33.3)	21	(47.7)
Disease status					0.652^a
Locally advanced	1	(2.3)	6	(9.3)
Metastatic	44	(97.7)	38	(90.7)
ECOG performance					0.022*
0	22	(48.9)	30	(68.2)
1	14	(31.1)	13	(29.5)
>2	9	(20.0)	1	(2.3)
Peritoneal carcinomatosis					0.317
No	25	(55.6)	29	(65.9)
Yes	20	(44.4)	15	(34.1)
Liver metastasis					0.924
No	20	(44.4)	20	(45.5)
Yes	25	(55.6)	24	(65.9)
Metastatic sites (number)					0.828^a
0	1	(2.3)	2	(4.5)
1	14	(31.1)	16	(36.5)
2	19	(42.2)	18	(40.9)
⩾3	11	(24.4)	8	(18.1)
Tumor location^b					0.666
Head	21	(48.8)	23	(53.5)
Body/tail	22	(51.2)	20	(46.5)
Baseline CEA^c
<5 ng/mL	19	(43.2)	17	(38.6)	0.665
⩾5 ng/mL	25	(56.8)	27	(61.4)
Baseline CA19-9^c
<34 ng/mL	5	(11.4)	8	(18.2)	0.367
⩾34 ng/mL	39	(88.6)	36	(81.8)
Genetic alteration^d
KRAS	21	(95.5)	29	(87.9)	0.638
TP53	15	(68.2)	26	(78.8)	0.376
CDKN2A	9	(40.9)	20	(60.6)	0.152
SMAD4	8	(36.4)	8	(24.2)	0.332
ARID1A	3	(13.6)	8	(24.2)	0.495

Chi-square test. Independent samples t test. Mann–Whitney U test.

Fisher’s exact test.

Two patients in GA and one patient in GAS with unknown primary of pancreatic cancer.

One patient in the GA group had missing data for both CEA and CA19-9.

Only 55 patients in our cohort had next-generation sequencing.

p < 0.05.

CA19-9, carbohydrate antigen 19-9; CEA, carcinoembryonic antigen; ECOG, Eastern Cooperative Oncology Group; GA, gemcitabine/nab-paclitaxel; GAS, GA in combination with S-1.

Treatment outcomes

The ORR was significantly higher in the GAS group than in the GA group (48% vs 18%, p = 0.001), with a greater number of patients achieving partial response (PR) in the GAS group (n = 21) than in the GA group (n = 8). No complete response was observed in either group. Disease control rates (DCR) were 84% and 51% in the GAS and GA groups, respectively (Figure 1). Regarding survival, PFS was significantly longer in the GAS group, with approximately double the median PFS compared to the GA group. The median PFS was 10.0 months (95% confidence interval (CI): 7.5–12.4) in the GAS group versus 5.2 months (95% CI: 3.1–7.3) in the GA group (hazard ratio (HR): 0.49; 95% CI: 0.30–0.81; p = 0.004; Figure 2(a)). Importantly, a significant difference in OS was observed in the GAS group. The median OS was 8.5 months (95% CI: 6.6–11.8) in the GA group, whereas the median OS was not reached in the GAS group (95% CI: 21.6–31.5), with an HR of 0.46 (95% CI: 0.24–0.88; p = 0.016; Figure 2(b)). Owing to this promising response, 12 patients in the GAS group underwent conversion surgery, and their clinical characteristics are summarized in Table S2. To account for the potential impact of conversion surgery on survival outcomes, an additional survival analysis was conducted after excluding 12 patients who underwent surgery. Although the difference in PFS was not statistically significant, there was a trend in favor of the GAS group with PFS of 8.2 months in the GAS group and 5.2 months in the GA group (HR: 0.62; 95% CI: 0.37–1.05; p = 0.073; Figure S1(A)). The significant difference in OS remained, with a median OS that was not reached in the GAS group and 8.5 months in the GA group (HR: 0.48; 95% CI: 0.24–0.88; p = 0.043; Figure S1(B)).

Figure 1.

Treatment response between GA and GAS. The ORR was significantly higher in the GAS group than in the GA group (48% vs 18%), with more PRs (n = 21 in GAS vs n = 8 in GA). Disease control rates were 84% with GAS and 51% with GA. No complete responses occurred in either group.

Figure 2.

Kaplan–Meier survival curves comparing GA and GAS regimens. (a) The median PFS was 10.0 months (95% CI: 7.5–12.4) in the GAS group versus 5.2 months (95% CI: 3.1–7.3) in the GA group (HR: 0.49; 95% CI: 0.30–0.81; p = 0.004). (b) The median OS was 8.5 months (95% CI: 6.6–11.8) in the GA group, whereas the median OS was not reached in the GAS group (95% CI: 21.6–31.5; HR: 0.46; 95% CI: 0.24–0.88; p = 0.016).

We also conducted univariate and multivariate analyses to identify the factors associated with OS. In the univariate analysis, ECOG performance status ⩾2, presence of liver metastasis, and elevated CEA levels were significantly associated with worse OS. However, in the multivariate analyses, only the treatment regimen remained statistically significant (HR: 0.42; 95% CI: 0.20–0.86; p = 0.018; Table 2).

Table 2.

Summary of univariate and multivariate analyses of OS and PFS.

Patient characteristics	OS				PFS
	Univariate		Multivariate		Univariate		Multivariate
	HR (95% CI)	p-Value	HR (95% CI)	p-Value	HR (95% CI)	p-Value	HR (95% CI)	p-Value
Treatment
GA	Reference		Reference		Reference		Reference
GAS	0.46(0.24–0.88)	0.019*	0.42(0.20–0.86)	0.018*	0.49(0.30–0.80)	0.004**	0.52(0.29–0.94)	0.029*
Age, years
<65-year-old	Reference				Reference
⩾65-year-old	0.90(0.48–1.68)	0.734			0.85(0.53–1.38)	0.515
Sex
Female	Reference				Reference
Male	0.89(0.47–1.66)	0.708			0.96(0.59–1.56)	0.862
ECOG performance
0	Reference				Reference		Reference
1	1.57(0.78–3.13)	0.204			1.48(0.88–2.50)	0.143	1.63(0.95–2.78)	0.075
⩾2	3.80(1.56–9.26)	0.003**			2.66(1.21–5.82)	0.015*	1.53(0.66–3.53)	0.317
Tumor location
Head	Reference				Reference
Body/tail	0.88(0.47–1.68)	0.709			0.85(0.52–1.40)	0.529
Peritoneal carcinomatosis
No	Reference				Reference
Yes	0.60(0.29–1.22)	0.160			0.79(0.48–1.33)	0.378
Liver metastasis
No	Reference		Reference		Reference
Yes	1.94(1.00–3.76)	0.049*	1.74(0.84–3.59)	0.137	1.48(0.91–2.41)	0.112
Baseline CEA (ng/mL)
⩽5	Reference		Reference		Reference
>5	2.43(1.18–5.01)	0.016*	1.87(0.88–4.01)	0.105	1.62(0.98–2.68)	0.062
Baseline CA19-9 (ng/mL)
⩽34	Reference				Reference
>34	1.82(0.71–4.71)	0.215			1.64(0.78–3.44)	0.190
Previous surgery^a
No	Reference				Reference
Yes	0.27(0.06–1.16)	0.079			0.64(0.39–1.43)	0.281

Cox regression analyses of factors associated with OS. Variables with p < 0.1 in the univariate analysis were entered into the multivariate model.

Previous surgery was defined as surgery performed with curative intent. Seven patients in the GA group and two patients in the GAS group underwent such surgery.

p < 0.05. **p < 0.01.

CA19-9, carbohydrate antigen 19-9; CEA, carcinoembryonic antigen; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; GA, gemcitabine/nab-paclitaxel; GAS, GA in combination with S-1; HR, hazard ratio; OS, overall survival.

Subgroup analysis of OS in GAS combination

The analyses revealed statistically significant improvements in OS among patients aged <65 years, both male and female patients, those with an ECOG performance status of 1, and those with liver metastases or a single metastatic site. Additional subgroups showing significant benefits included patients with tumors located in the pancreatic body or tail, those who did not undergo prior curative surgery, and those with elevated baseline CEA or CA19-9 levels. In terms of molecular characteristics, none had mutations in BRCA1, BRCA2, or PALB2, so we focused on analyzing the five most common alterations (KRAS, TP53, CDKN2A, SMAD4, and ARID1A). Statistically significant OS benefits from GAS were observed in patients harboring KRAS, CDKN2A, and ARID1A mutations (Figure 3).

Figure 3.

Subgroup analysis of OS in patients treated with GAS. Significant OS benefits were observed in several clinical and molecular subgroups, including younger age, ECOG 1, limited metastatic burden, elevated tumor markers, and mutations in KRAS, CDKN2A, and ARID1A.

Adverse events

In the GAS group, the most frequently reported nonhematological treatment-related adverse events were fatigue (59.1%), nausea/vomiting (56.8%), and elevated liver function (40.9%). The overall nonhematological toxicity profile was similar in the GA group, except for renal function impairment, which was more commonly observed in GA-treated patients (26.6% vs 6.8% in the GAS group), but the incidence of severe renal impairment (grade ⩾3) was comparable between the two groups (2.3% vs 2.2%). Anemia was the most frequent hematological toxicity in both groups, occurring in 68.8% of patients in the GA group and 59.1% of patients in the GAS group. Grade ⩾3 neutropenia was reported more often in the GAS group, whereas severe anemia and thrombocytopenia were more common in the GA group. No treatment-related fatalities occurred in either group (Table 3).

Table 3.

Adverse events.

Side effects	GAN = 45(%)				GASN = 44(%)
Side effects	Any grade		Grade 3 or 4		Any grade		Grade 3 or 4
Fatigue	27	(60.0)	3	(6.6)	26	(59.1)	0
Nausea/vomiting	17	(37.7)	0		25	(56.8)	0
Neutropenia	12	(26.6)	3	(6.6)	10	(22.7)	5	(11.3)
Anemia	31	(68.8)	6	(13.3)	26	(59.1)	1	(2.2)
Thrombocytopenia	11	(24.4)	6	(13.3)	7	(15.9)	3	(6.8)
Liver function impairment	22	(48.8)	1	(2.3)	18	(40.9)	3	(6.8)
Renal function impairment	12	(26.6)	1	(2.3)	3	(6.8)	1	(2.2)

Patients who met the eligibility criteria were included. Shown are adverse events of any cause that occurred in at least 10% of the patients in both groups. No grade 5 events were recorded.

GA, gemcitabine/nab-paclitaxel; GAS, GA in combination with S-1.

Discussion

To our knowledge, this is the first retrospective study to demonstrate that GAS combination therapy offers a survival benefit over the standard GA regimen in patients with advanced PDAC. The GA regimen was previously established as the standard of care following the MPACT trial,⁴ which showed that adding nab-paclitaxel to gemcitabine significantly improved OS compared to gemcitabine alone. GA remained the preferred first-line option until the NALIRIFOX regimen demonstrated significantly prolonged PFS and OS compared with GA in the NAPOLI-3 study.⁵ However, for patients with an ECOG performance status of 2, the GA regimen remains the first treatment of choice.¹¹

Before the NAPOLI-3 trial by Wainberg et al., several multiagent combinations were investigated to improve survival outcomes in patients with metastatic PDAC. Earlier regimens incorporating gemcitabine, capecitabine, or cisplatin have yielded only modest improvements in survival,¹² and the addition of targeted therapies to gemcitabine has been largely disappointing.¹³ S-1 is an oral fluoropyrimidine derivative that has been extensively studied for the treatment of advanced PDAC, particularly in East Asian populations. A phase II trial evaluating S-1 monotherapy in patients with gemcitabine-refractory metastatic PDAC reported an ORR of 15%,¹⁴ indicating a modest efficacy in a second-line setting. In the first-line setting, the GEST trial assessed S-1 as monotherapy and in combination with gemcitabine (GS). While the GS regimen did not demonstrate a statistically significant improvement in OS (HR: 0.88; 97.5% CI: 0.71–1.08; p = 0.15), it met the non-inferiority criterion compared to gemcitabine monotherapy (HR: 0.96; 97.5% CI: 0.78–1.18; p < 0.001 for non-inferiority). Notably, the GS arm achieved a higher ORR of 29.3% compared to gemcitabine alone, and its efficacy was greater in older adult patients with locally advanced PC.¹⁵ Based on this evidence, GS or S-1 monotherapy is listed as the treatment of choice in Taiwan’s pancreatic cancer treatment recommendation.¹⁶ Based on the NAPOLI-3 and PRODIGE studies, the triplet combination showed good response and survival in metastatic PDAC. Therefore, in the TCOG T5221 study (NCT05026905), a phase II trial conducted in Taiwan, the researchers tried to compare different GA-based triplet regimens in metastatic PDAC. In the preliminary results of this study, the GASL regimen (gemcitabine, nab-paclitaxel, S-1, and leucovorin) demonstrated a promising ORR of 53.3%, which was significantly higher than the 17.4% observed with GA plus oxaliplatin.⁹ Notably, the ORR of 23% reported in the historical MPACT trial for the GA regimen. These results highlight the potential of the GAS regimen to achieve superior tumor control in advanced PDAC.

In our cohort, the GAS regimen was associated with a significantly longer PFS of 10.0 months (p = 0.004), aligning with a previous phase II randomized study that reported a PFS of 8.6 months (95% CI: 7.4 months–not reached). The OS for the entire cohort was not reached at the end of the observation period. With a median follow-up of 10.2 months, the GA group demonstrated a median OS of 8.5 months—comparable to the historical outcomes, whereas the median OS for the GAS group remained unreached. Furthermore, the TCOG T5221 study evaluating the GASL regimen reported encouraging OS outcomes ranging from 17.1 to 40.7 months, with all patients alive at the time of the data cutoff. Collectively, these findings indicate the efficacy of the GAS regimen in the treatment of advanced PDAC. In addition, 12 patients in the GAS group underwent conversion surgery during the frontline treatment, whereas none of the patients in the GA group underwent surgery. This discrepancy may be attributable to the higher ORR observed with the GAS regimen. Considering that conversion surgery may be a major confounding factor for OS, we conducted an additional analysis that excluded these patients. Even after excluding patients who underwent surgery, the GAS group continued to demonstrate a significant survival benefit (p = 0.043). The GAS regimen has also been studied in Japan and China for different treatment schedules.^17–19 The studies were phase I, designed to determine the maximum tolerated dose of the GAS regimen. They all infused GA on days 1 and 8 of a 3-week cycle with 14 days of consecutive S-1. Chang et al. found nab-paclitaxel 125 mg/m², gemcitabine 1000 mg/m², and S-1 80 mg/day (BSA >1.5 m²) have acceptable toxicity with ORR of 36.8% and DCR of 94.7%. The median PFS was 5.3 months, and the median OS was 10.3 months. On the other hand, Sai et al. found nab-paclitaxel 90 mg/m², gemcitabine 600 mg/m², and S-1 50/70/80 mg/day (BSA <1.25 m²/1.25–1.50 m²/>1.5 m²) is tolerable. In total, 58% of patients achieved PR, and the median PFS was 7.6 months. A phase II study evaluating the efficacy of neoadjuvant GAS (nab-paclitaxel 125 mg/m², gemcitabine 1000 mg/m² on day 1, and S-1 60–100 mg/day on days 1–7 during a 14-day cycle) in borderline resectable PDAC, the response and DCRs were 43% and 96%, respectively, with 96% of the patients successfully receiving curative surgery. Interestingly, this dose is identical to our real-world cohort.²⁰ Although the dosing schedules are different, these findings suggest that GAS may provide good treatment efficacy in advanced PDAC, both in patients deemed ineligible for curative resection and those with the potential to undergo surgery.

In the GAS subgroup, OS was further analyzed across various clinical and molecular subgroups. Patients younger than 65 years, regardless of sex, those with an ECOG performance status of 1, and individuals with liver metastases or a single metastatic site showed improved outcomes. Additional subgroups demonstrating significant benefits included patients with tumors located in the pancreatic body or tail, those with de novo metastatic disease, and those with elevated baseline CEA or CA19-9 levels. Notably, many of these characteristics are traditionally associated with more advanced disease, suggesting that the GAS regimen may confer survival benefits, even in higher-risk populations. In addition, statistically significant OS benefits of GAS were observed in patients with KRAS, CDKN2A, and ARID1A mutations. These genetic alterations are commonly found in advanced PDAC, suggesting that GAS may be useful in patients harboring these alterations.

Most patients in our cohort received a liposomal irinotecan-based regimen as second-line therapy. Notably, a substantial proportion of the patients in the GA group (64.4%) were unable to receive any subsequent treatment following disease progression. In comparison, only 34.1% of the patients in the GAS group did not receive further therapy after first-line treatment (Table S3). Subsequent treatment of metastatic PDAC that fails after first-line treatment is important. In a recent GENERATE trial in Japan,²¹ both the GA and mFOLFIRINOX arms demonstrated an OS of more than 1 year, emphasizing the importance of subsequent treatment because approximately 70% of patients in this trial received subsequent treatment, compared to previous studies, such as NAPOLI-3, in which only 50% of patients received subsequent treatment.⁵ In a retrospective study in Korea, subsequent FOLFIRINOX therapy resulted in a median survival of 9.7 months.²² In our previous report, we used NALIRIFOX as a second-line treatment, which provided 6.9 months of survival after failure of gemcitabine-based treatment.²³ In this study, the GAS regimen provided a better response, and more patients received subsequent treatment while facing disease progression, possibly contributing to prolonged survival.

Our study had some limitations. First, it was a retrospective study in a single medical center with a small sample size, which introduced inherent biases in patient selection, treatment assignment, imbalance of disease status, and precise recording of side effects, so the findings in this study should be interpreted cautiously. Prospective randomized clinical trials are needed to validate the impact of different dosing strategies in combination chemotherapy. Second, we used a modified GA regimen, which may have influenced the treatment outcomes; however, our patients in the GA group showed results comparable to those of the conventional MPACT trial. Other trials using the modified GA also showed identical results,^24,25 making this modification reasonable in real-world conditions. The rationale for using a modified dosing schedule is that 41% of the patients had dose reductions of the nab-paclitaxel dose, and only 71% of all nab-paclitaxel doses were administered during the original MPACT study. Finally, the GA group included more patients with poor performance status and a higher proportion of patients who did not receive subsequent therapy, all of which may have contributed to the observed OS difference and potentially confounded the survival analysis; however, the actual difference in ORR and PFS between the groups could still indicate the potential superior effect of the GAS regimen over GA.

Conclusion

In conclusion, S-1 could be combined with GA as another potential triplet combination treatment option for advanced-stage pancreatic cancer with a comparable benefit and toxic profile to GA. Further prospective studies are warranted to confirm potential clinical benefits.

Supplemental Material

sj-docx-2-tam-10.1177_17588359261420081 – Supplemental material for Gemcitabine/nab-paclitaxel plus S-1 combination compared with gemcitabine/nab-paclitaxel in advanced pancreatic ductal adenocarcinoma: a retrospective study

Supplemental material, sj-docx-2-tam-10.1177_17588359261420081 for Gemcitabine/nab-paclitaxel plus S-1 combination compared with gemcitabine/nab-paclitaxel in advanced pancreatic ductal adenocarcinoma: a retrospective study by Kuan-Yu Tseng, Chiann-Yi Hsu, Cheng-Han Wu, Yu-Hsuan Shih and Hsin-Chen Lin in Therapeutic Advances in Medical Oncology

Supplemental Material

sj-docx-3-tam-10.1177_17588359261420081 – Supplemental material for Gemcitabine/nab-paclitaxel plus S-1 combination compared with gemcitabine/nab-paclitaxel in advanced pancreatic ductal adenocarcinoma: a retrospective study

Supplemental material, sj-docx-3-tam-10.1177_17588359261420081 for Gemcitabine/nab-paclitaxel plus S-1 combination compared with gemcitabine/nab-paclitaxel in advanced pancreatic ductal adenocarcinoma: a retrospective study by Kuan-Yu Tseng, Chiann-Yi Hsu, Cheng-Han Wu, Yu-Hsuan Shih and Hsin-Chen Lin in Therapeutic Advances in Medical Oncology

Supplemental Material

sj-jpg-1-tam-10.1177_17588359261420081 – Supplemental material for Gemcitabine/nab-paclitaxel plus S-1 combination compared with gemcitabine/nab-paclitaxel in advanced pancreatic ductal adenocarcinoma: a retrospective study

Supplemental material, sj-jpg-1-tam-10.1177_17588359261420081 for Gemcitabine/nab-paclitaxel plus S-1 combination compared with gemcitabine/nab-paclitaxel in advanced pancreatic ductal adenocarcinoma: a retrospective study by Kuan-Yu Tseng, Chiann-Yi Hsu, Cheng-Han Wu, Yu-Hsuan Shih and Hsin-Chen Lin in Therapeutic Advances in Medical Oncology

Footnotes

Acknowledgements

The authors would like to thank the participating patients and all medical staff members.

Declarations

ORCID iDs

Kuan-Yu Tseng

Yu-Hsuan Shih

Hsin-Chen Lin

Supplemental material

Supplemental material for this article is available online.

References

National Cancer Institute. SEER*stat databases: SEER November 2024 submission, https://seer.cancer.gov/data-software/documentation/seerstat/nov2024/ (2025, accessed 23 April 2025).

Burris

3rd , et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol 1997; 15(6): 2403–2413.

Conroy

, et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med 2011; 364(19): 1817–1825.

Von Hoff

, et al. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N Engl J Med 2013; 369(18): 1691–1703.

Wainberg

, et al. NALIRIFOX versus nab-paclitaxel and gemcitabine in treatment-naive patients with metastatic pancreatic ductal adenocarcinoma (NAPOLI 3): a randomised, open-label, phase 3 trial. Lancet 2023; 402(10409): 1272–1281.

Shirasaka

, et al. Development of a novel form of an oral 5-fluorouracil derivative (S-1) directed to the potentiation of the tumor selective cytotoxicity of 5-fluorouracil by two biochemical modulators. Anticancer Drugs 1996; 7(5): 548–557.

Ueno

, et al. Randomized phase III study of gemcitabine plus S-1, S-1 alone, or gemcitabine alone in patients with locally advanced and metastatic pancreatic cancer in Japan and Taiwan: GEST study. J Clin Oncol 2013; 31(13): 1640–1648.

Hamada

, et al. Efficacy and safety of gemcitabine plus S-1 in pancreatic cancer: a pooled analysis of individual patient data. Br J Cancer 2017; 116(12): 1544–1550.

, et al. 1528P TCOG T5221 trial: a phase II randomized study of gemcitabine and nab-paclitaxel in combination with S-1/LV (GASL) or oxaliplatin (GAP) as first-line treatment for metastatic pancreatic cancer. Ann Oncol 2024; 35(S2): S932–S933.

10.

Von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370(9596): 1453–1457.

11.

Conroy

Ducreux

; ESMO Guidelines Committee. ESMO clinical practice guideline express update on the management of metastatic pancreatic cancer. ESMO Open 2025; 10(4): 104528.

12.

Lim

, et al. Gemcitabine combined with capecitabine compared to gemcitabine with or without erlotinib as first-line chemotherapy in patients with advanced pancreatic cancer. Cancer Res Treat 2015; 47(2): 266–273.

13.

Moore

MJ.

Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer: a phase III trial of the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 2007; 25(15): 1960–1966.

14.

Morizane

, et al. A phase II study of S-1 in gemcitabine-refractory metastatic pancreatic cancer. Cancer Chemother Pharmacol 2009; 63(2): 313–319.

15.

Bai

, et al. A prospective phase II study of biweekly S-1, leucovorin, and gemcitabine in elderly patients with locally advanced or metastatic pancreatic adenocarcinoma—The Taiwan Cooperative Oncology Group T1217 study. Eur J Cancer 2022; 173: 123–132.

16.

, et al. Systemic treatments in pancreatic cancer: Taiwan pancreas society recommendation. Biomed J 2024; 47(3): 100696.

17.

Sai

, et al. Phase 1 study of gemcitabine/nab-paclitaxel/S-1 in patients with unresectable pancreatic cancer (GeNeS1S trial). Cancer Chemother Pharmacol 2021; 87(1): 65–71.

18.

Chang

, et al. A phase I study of gemcitabine/nab-paclitaxel/S-1 chemotherapy in patients with locally advanced or metastatic pancreatic ductal adenocarcinoma. Oncologist 2023; 28(7): e575–e584.

19.

Chang

, et al. Combination of gemcitabine, nab-paclitaxel, and S-1(GAS) as the first-line treatment for patients with locally advanced or advanced pancreatic ductal adenocarcinoma: study protocol for an open-label, single-arm phase I study. BMC Cancer 2021; 21: 545.

20.

Kondo

, et al. A phase II study of gemcitabine/nab-paclitaxel/S-1 combination neoadjuvant chemotherapy for patients with borderline resectable pancreatic cancer with arterial contact. Eur J Cancer 2021; 159: 215–223.

21.

Ohba

, et al. Modified fluorouracil, leucovorin, irinotecan, and oxaliplatin or S-1, irinotecan, and oxaliplatin versus nab-paclitaxel 1 gemcitabine in metastatic or recurrent pancreatic cancer (GENERATE, JCOG1611): a randomized, open-label, phase II/III trial. J Clin Oncol 2025; 43(31): 3345–3354.

22.

Park

, et al. Liposomal irinotecan plus fluorouracil/leucovorin versus FOLFIRINOX as the second-line chemotherapy for patients with metastatic pancreatic cancer: a multicenter retrospective study of the Korean Cancer Study Group (KCSG). ESMO Open 2021; 6(2): 100049.

23.

Lin

Tseng

Shih

YH.

NALIRIFOX as subsequent therapy for metastatic pancreatic cancer after gemcitabine-based chemotherapy: a real-world experience in single center. Ann Oncol 2024; 35(Suppl_4): S1450–S1504.

24.

Ahn

, et al. A modified regimen of biweekly gemcitabine and nab-paclitaxel in patients with metastatic pancreatic cancer is both tolerable and effective: a retrospective analysis. Ther Adv Med Oncol 2017; 9(2): 75–82.

25.

Kokkali

, et al. Biweekly gemcitabine/nab-paclitaxel as first-line treatment for advanced pancreatic cancer. In Vivo 2018; 32(3): 653–657.