Efficacy and tolerability of fluorouracil, leucovorin, oxaliplatin and docetaxel (FLOT) in unselected patients with advanced gastric and gastroesophageal cancer: does age really matter?

Abstract

Purpose

Fluorouracil, leucovorin, oxaliplatin and docetaxel (FLOT) regimen has shown strong efficacy as perioperative therapy for patients with locally advanced gastric (GC) and gastroesophageal (AEG) carcinoma. In the palliative situation, FLOT is recommended only for young fit patients. Data of efficacy and tolerability of FLOT in elderly patients are scarce and controversial. Thus, this study aimed to provide real-life experience of elderly patients with GC and AEG treated with FLOT as first-line palliative chemotherapy.

Methods

Patients with advanced or metastatic GC or AEG and treated with FLOT as first-line palliative therapy between 2010 and 2021 were analyzed. Patients were grouped into < 65 years old (n = 35) and ≥ 65 years old (n = 22) groups. Overall survival (OS), progression-free survival (PFS), feasibility and toxicity were analyzed.

Results

The median OS was 10.4 months with no significant difference between both groups (HR 0.86; 95% CI 0.48, 1.57; p = 0.632). The ECOG performance status showed powerful influence on OS in the subgroup analysis with median OS of 12.3 months for ECOG = 0 compared to 5.0 months for ECOG ≥ 1 (p = 0.015) as well as in multivariate analysis (HR 2.62; 95% CI 1.36, 5.04; p = 0.004).

Conclusion

In the present study the ECOG performance status showed a stronger prognostic value than patient age in FLOT as first- line therapy in a real-life cohort with advanced and metastatic GC and AEG. The performance status should therefore be considered in the therapeutic decision making of elderly patients with GC and AEG.

Introduction

Gastric cancer (GC) is the fifth most common cancer worldwide causing the fourth highest cancer-associated mortality (Van Cutsem et al. 2016). Furthermore, the rising prevalence of obesity and gastroesophageal reflux disease with consecutive Barrett´s esophagus has resulted in a 2.5-fold increase in the incidence of gastroesophageal adenocarcinoma (AEG) with stabilizing trends in the last decades (Arnold et al. 2017; Buas and Vaughan 2013). In both tumor entities, nonspecific symptoms often result in a late diagnosis. Accordingly, 60% of cases present advanced stages and 30% metastatic disease by the time of diagnosis due to early tumor dissemination (Wagner et al. 2017). Additionally, 40% of patients initially treated with curative intention will relapse (Wagner et al. 2017).

While patients with curative therapy intention show a median overall survival of 50 months, survival through palliative chemotherapy shows actual benefits compared to best supportive care (3–5 months), although it only reaches 7.2–10.3 months (Al-Batran et al. 2019; Merchant et al. 2021; Rosenberg et al. 2019; Zhou et al. 2020).

The beneficial effect of palliative chemotherapy in advanced GC and AEG has been shown in multiple trials. Current trends aiming for personalized medicine recommend novel targeted therapy regimens in combination with SC, such as trastuzumab for HER2-positive carcinoma or, most recently, nivolumab for PD-L1-positive tumors (Gong et al. 2016; Bang et al. 2010; Janjigian et al. 2021).

For the most frequent HER2- or PD-L1-negative carcinomas, platin/5-FU-based chemotherapy is the most often applied first-line treatment. In case of medical contraindications, 5-FU/irinotecan-based chemotherapy can be used (Janjigian et al. 2021; Wagner et al. 2017). The benefit of adding docetaxel for treating GC and AEG has been demonstrated in the neoadjuvant situation. High rates of R0 resection and extended survival were shown with either perioperative or neoadjuvant systemic chemotherapy (SC) FLOT regimen (5-FU/LV plus oxaliplatin + docetaxel) (Cunningham et al. 2006; Ronellenfitsch et al. 2013; Schuhmacher et al. 2010; Ychou et al. 2011). Since publication of the FLOT4 trial, the FLOT regimen has become the new perioperative therapy standard for patients with GC and AEG and curative therapy intention (Al-Batran et al. 2019).

In addition, docetaxel also showed a benefit in first-line therapy for advanced GC and AEG as part of the DCF regimen (docetaxel, cisplatin, and fluorouracil), with a median OS of 12 months (Van Cutsem et al. 2006). After observing significantly higher hematological toxic events under DCF, the FLOT regimen (docetaxel, oxaliplatin, leucovorin and 5FU) was shown to provide an alternative option for SC in palliative intention with a better toxicity profile. A phase II trial has shown median progression-free survival (PFS) and OS periods of 5.2 and 11.1 months, respectively, in patients with previously untreated metastatic disease. Although the comparison of tolerability has not been statistically confirmed, the FLOT regimen seemed to induce less hematological toxic effects compared to DCF with neutropenia rates of 48% vs. 82% (Al-Batran et al. 2008). Current guidelines recommend FLOT as first-line therapy only for relatively young fit patients and strong tumor remission desire. Since the FLOT-4 landmark trial emphasized once again the highly active FLOT regimen in neoadjuvant treatment, the efficacy of palliative FLOT in a real-world cohort is of major interest (Al-Batran et al. 2019).

The use of FLOT in elderly patients should be carefully evaluated. Data of efficacy and tolerability of FLOT in this setting are scarce and controversial. In a randomized trial, Al-Batran et al. investigated the efficacy of the FLOT regimen in comparison to the less-toxic FLO regimen in elderly patients (≥ 65 years) with advanced gastroesophageal cancer. While the results indicated no significant difference in median OS between the two patient groups (OS 17.3 and 14.5 months), FLOT induced a substantial increase in toxicity. On the other hand, FLOT improved the response and the PFS rates in the subgroups of patients with locally advanced cancer aged between 65 and 70 years (Al-Batran et al. 2013).

In the present study, we analyzed real-life data of patients with advanced GC or AEG treated with FLOT as first-line therapy in palliative intention. The efficacy and tolerability of FLOT in older patients (≥ 65 years) was analyzed and compared to the younger patients (< 65 years) in our cohort of patients. Furthermore, patient and treatment characteristics associated with better tolerability and prolonged overall survival were identified and further analyzed.

Methods

Patients and eligibility criteria

All patients with newly diagnosed or recurrent GC or AEG, treated at the University Hospital of Bonn, Germany, between 2010 and 2021 were evaluated for inclusion into this study (Fig. 1). Patients were included if the following inclusion criteria were fulfilled: (1) pathologically confirmed malignancy; (2) unresectable disease due to metastatic disease (e.g., distant organs, malignant ascites); (3) no prior palliative systemic chemotherapy (SC) treatment before FLOT administration; (4) patient was eligible for first-line therapy with FLOT and received at least one cycle of FLOT therapy; (5) no other active malignancy than GC or AEG. Additionally, patients with cancer relapse after curative treatment and adjuvant or neoadjuvant chemotherapy, completed at least 6 months prior to start of palliative first-line therapy with FLOT, were included, provided they fulfilled all other inclusion criteria.

Fig. 1.

Flowchart of patients

Treatment

Therapy decision was performed in our weekly multidisciplinary tumor boards attended by representatives from the departments of visceral surgery, oncology, gastroenterology, radiology, radiotherapy and pathology. Patients were considered inoperable because of metastatic stage at diagnosis (cM1) or due to poor performance status (severe comorbidities) or when patient refused the operation. Systemic CT with the FLOT regimen was offered to patients with sufficient performance status and high remission pressure. All patients included in this study received at least one cycle of oxaliplatin (85 mg/m²) administrated as 2-h intravenous infusion, docetaxel (50 mg/m²) administrated as 1-h infusion, followed by leucovorin (200 mg/m²) and 5-FU (2600 mg/m²) infusion over 24 h. These therapy cycles were repeated every 2 weeks. SC was stopped due to toxicity, death, patient’s wish and radiological or clinical disease progression. Antiemetic or other supportive medicines were given at the discretion of the treating physicians. Dose reduction, delay or interruption of FLOT was performed due to toxicity or according to patient’s wish to avoid high-grade toxicity and to maintain the patient’s quality of life. For patients with recurrent tumor disease, prior surgical and neoadjuvant/adjuvant therapies are shown in Table 1. Additional treatments others than FLOT are also recorded in Table 2.

Table 1.

Baseline parameters

Parameter, units reference interval	All (n = 57)		< 65 years (n = 35)		≥ 65 years (n = 22)		P value
	N	%	N	%	N	%
Age
< 65	35	61.4	35	100	0	0
≥ 65	22	38.6	0	0	22	100
Sex							0.353
Male	40	70.2	23	65.7	17	77.73
Female	17	29.8	12	34.3	5	22.7
Tumor type							0.399
Gastric cancer	35	61.4	23	65.7	12	54.5
AEG	22	38.6	12	34.3	10	45.5
Disease stage							0.889
Locally advanced	8	14.0	5	14.3	3	13.6
Metastatic	45	78.9	28	80.0	17	77.3
Locally recurrent	4	7.0	2	5.7	2	9.1
Signet-ring cell carcinoma	22	38.6	15	42.9	7	31.8	0.130
Localization of metastasis				0.0		0.0
Liver	25	43.9	12	34.3	13	59.1	0.066
Lung	11	19.3	6	17.1	5	22.7	0.733
Lymph nodes	33	57.9	18	51.4	15	68.2	0.212
Skeleton	13	22.8	10	28.6	3	13.6	0.191
Peritoneum	24	42.1	18	51.4	6	27.3	0.072
Brain	4	7.1	3	8.6	1	4.5	0.709
Ovary	3	5.3	3	8.6	0	0.0	0.276
Other	2	3.5	1	2.9	1	4.5	1.000
Lauren classification							0.232
Intestinal type	11	19.3	5	14.3	6	27.3
Diffuse type	20	35.1	15	42.9	5	22.7
Mixed type	4	7.0	3	8.6	1	4.5
Indeterminable	22	38.6	12	34.3	10	45.5
HER-2	9	15.8	6	17.1	3	13.6	1.000
PD-1	4	7.0	4	11.4	0	0.0	0.237
Microsatellite instability	2	3.5	2	5.7	0	0.0	1.000
ECOG							0.441
0	34	59.6	21	60.0	13	59.1
1	14	24.6	9	25.7	5	22.7
2	1	1.8	0	0.0	1	4.5
3	0	0	0	0.0	0	0.0
Unknown	8	14	5	14.3	3	13.6
Biochemical condition
Hemoglobin (mg/dl)	12.6 (10.9; 14.9)		12.7 (11.3; 14.9)		12.2 (9.8; 15.0)		0.263
Age by diagnosis	58.0 (47.4; 65.9)		51.7 (43.5; 57.9)		69.0 (65.6; 73.5)		< 0.001
CRP (mg/dl)	11.3 (2.7; 25.1)		9.8 (0.8; 22.8)		12.7 (3.8; 37.9)		0.679
Creatinine (mg/dl)	0.81 (0.72; 0.93)		0.8 (0.7; 0.87)		0.84 (0.74; 0.95)		0.088
LDH (U/l)	231 (181; 361)		239.5 (187; 363.5)		212 (176.5; 360)		0.490
CA19-9 (kU/l)	53,5 (5.2; 346.5)		10.9 (4.1; 254.8)		561.7 (61.3; 2012.5)		0.013
CEA (µg/l)	1.9 (1.2; 4.5)		1.7 (0.8; 3.6)		2.6 (1.4; 26.5)		0.110

Bold values indicate statistically significance (p<0.05)

CA19-9 carbohydrate antigen 19-9, CEA carcinoembryonic antigen, CRP C-reactive protein, ECOG Eastern Cooperative Oncology Group performance status, HER-2 human epidermal growth factor receptor 2, LDH lactate dehydrogenase, PD-1 programmed cell death protein 1

Table 2.

Therapy characteristics

Characteristics	All (n = 57)		< 65 years (n = 35)		≥ 65 years (n = 22)		P value
	N	%	N	%	N	%
Initially reduced dose	8	14.0	7	20.0	1	4.5	0.117
Dose reduction during therapy	30	52.6	19	54.3	11	50.0	0.752
Concomitant trastuzumab	8	14.0	5	14.3	3	13.6	0.444
Reason for therapy discontinuation							0.182
Toxicity	7	12.3	2	5.7	5	22.7
PD	23	40.4	16	45.7	7	31.8
dead	4	7.0	3	8.6	1	4.5
Patient request	3	5.3	2	5.7	1	4.5
Poor general condition	10	17.5	4	11.4	6	27.3
Other	10	17.5	8	22.9	2	9.1
Second/third line therapy	28	49.1	17	48.6	11	50.0	0.741
Trastuzumab	1	3.57	1	5.9	0	0.0
Nivolumab	1	3.57	1	5.9	0	0.0
Ramucirumab + paclitaxel	13	46.43	10	58.8	3	27.3
FOLFIRI	7	25.00	0	0.0	7	63.6
EOX	1	3.57	1	5.9	0	0.0
Other second-line therapies	5	17.86	4	23.5	1	9.1
Duration of FLOT chemotherapy in weeks	16.1 (10.9; 25.9)		15.7 (8.7; 26)		17 (12.1; 26.2)		0.854
FLOT chemotherapy cycles	8 (5; 10)		8 (5; 11)		8 (5; 9.3)		0.503

EOX epirubicin/oxaliplatin/capecitabine, FOLFIRI leucovorin/fluoruracil/irinotecan, PD progressive disease

Data collection and study design

This is a single institution retrospective analysis of patients with inoperable advanced or metastatic GC and AEG who were treated with FLOT as first-line therapy in a daily practice setting.

To evaluate the relationship between patient age and the efficacy of FLOT we allocated all patients according to their age at the time of therapy start into one of two groups (< 65 years vs. ≥ 65 years).

Clinical baseline parameters were recorded prior to therapy start (Table 1). Primary endpoint was overall survival (OS) defined as the time between first administration of SC with FLOT up to death or end of observation period in July 2021. When lost to follow-up, patients were censored at the date of their last visit. Secondary endpoints were progression-free survival (PFS), defined as the period between start of SC and disease progression or death, therapy response and the toxicity rates under SC. Therapy response was evaluated every 2–3 months with magnetic resonance imaging or computed tomography scan. Responses to therapy were defined as overall response rate (ORR), by summing up patients who showed complete or partial response to therapy. Disease control rate (DCR) encompassed patients who had complete response, partial response or stable disease.

Secondary events were assessed by using the Common Terminology Criteria for Adverse Events (CTCAE v.5.0) and recorded for grades three to five. Therapy response under SC was evaluated by computer tomography or magnetic resonance imaging every 3 months according to the Response Evaluation Criteria in Solid Tumors (RECIST 1.1).

This retrospective study was reviewed and approved by the Ethics Committee of the Medical Faculty of the University of Bonn (No. 341/17). Written informed consent before therapy with FLOT was obtained from all patients. Datasets were anonymized in compliance with national and international patient’s privacy policies.

Statistical analysis

Normal distribution of continuous variables was tested with the Shapiro–Wilk test. Continuous variables are expressed as medians and first and third quartiles, differences were assessed using Student’s unpaired t test or non-parametric Mann–Whitney test, as appropriate. Categorical variables are shown as absolute frequencies and percentages. The Pearson’s Chi-squared test or Fisher´s exact test was used to compare them. Estimated survivals were compared by log-rank test and displayed in Kaplan–Meier curves and furthermore presented as medians with 95% confidence interval (95% CI). Univariate and multivariate analyses were performed using Cox regression forward conditional models. If parameters showed p values ≤ 0.05 in univariate analysis, they were included in multivariate analysis. Two-tailored p values ≤ 0.05 were considered statistically significant. SPSS version 22 (IBM Corporation, Armonk, NY, USA) was used for statistical analysis.

Results

Patient characteristics and therapy

Baseline characteristics of all patients are listed in Table 1. Fifty-seven patients were included in the final analysis and allocated to either the < 65 years group (n = 35) or the ≥ 65 years group (n = 22), depending on their age at start of SC. The median age of all patients was 58 (range) years and 70.2% were male. The median age of the younger group was 51.7 years (range 31.8–63.4 years) and the median age of the older group was 69.9 years (range 65.1–76.9 years). Most patients were included at a metastatic stage of disease (78.9%) with 61.4% of patients having GC and 38.6% having AEG. All carcinomas were histologically proven, 38.6% were signet-ring cell type, while the leading subtype according to Lauren´s criteria was the diffuse type (35.1%). The immunohistochemistry showed positive results for human epidermal growth factor receptor 2 (HER2), programmed death-ligand 1 (PD-L1) and microsatellite instability (MSI) in 24.6%, 7.0% and 3.5% of cases, respectively. Most metastases were lymphatic (57.9%), hepatic (43.9%) and peritoneal (42.1%). The ECOG performance status was 0 in 35.1% and 1 in 40.4%. Between the patients of the ≥ 65 years group and the patients of the < 65 years groups, only CA19-9 levels at baseline eshowed a statistically significant difference. The older group had higher CA19-9 values (561.7 vs. 10.9; p = 0.013).

Therapy data are displayed in Table 2. All patients were treated with the FLOT regimen and the median number of therapy cycles was eight (1–23), with no difference between both groups. The most common reason for FLOT discontinuation was disease progression (40.4%) with similar distribution in both groups. Median treatment duration was 16.1 (range) weeks (Q1 10.9; Q3 25.9) without statistical difference between both groups. A dose adjustment during FLOT treatment was required by 52.6% of patients. Second-line chemotherapy was administered in 49.1% of cases, with a larger proportion of patients receiving ramucirumab and paclitaxel (46.4%) or FOLFIRI (5-FU/FA plus irinotecan) (25.0%).

Efficacy

The median follow-up period was 40.6 months (95% CI 36.7; 44.6). The median OS as the primary end point was 10.4 months (95% CI 8.1, 12.6) in the < 65 years group and 10.4 months (95% CI 7.3, 13.6) in the ≥ 65 years group (HR 0.86; 95% CI 0.48, 1.57; p = 0.632) (Fig. 3a). The cumulative overall survival was 10.4 months (95% CI 8.5, 12.3), respectively (Fig. 2a). Likewise, regarding both cancer entities separately, there was no difference in OS between the two age groups (Fig. 2c, d). The overall estimated PFS was 7.8 months (95% CI 6.9, 8.7) for all patients (Fig. 2 b) with 7.5 months (95% CI 5.1, 9.9) in the younger group and 8.0 months (95% CI 4.7, 11.3) in the older group (HR 0.69; 95% CI 0.40, 1.20; log-rank p = 0.191) (Fig. 3c, d). Therapy response showed no statistical difference between both groups (Table 3). One patient (1.8%) experienced complete response and 23 patients (40.4%) experienced partial response leading to an overall response rate of 42.1% (p = 0.577). Additionally, 20 patients (35.1%) had stable disease contributing to a control rate of 77.2%. Dose adjustments were necessary in 30 cases (52.6%) but had no statistical influence on OS (p = 0.366) or PFS (p = 0.132).

Fig. 3.

Kaplan–Meier survival analysis, with log-rank p. a Overall survival: <65-≥-ears. b Overall survival: ECOG=0 vs. ECOG=1/2. c Progression-free survival: <65-≥-years. d Progression-free survival: ECOG=0 vs. ECOG=1/2

Fig. 2.

Kaplan–Meier survival analysis. a Overall survival: < 65 years vs. ≥ 65 years. b Overall survival ECOG = 0 vs. ECOG = 1/2. c Progression-free survival: < 65 years vs. ≥ 65 years. d Progression-free survival ECOG = 0 vs. ECOG = 1/2

Table 3.

Response rates

	All (n = 57)		< 65 years (n = 35)		≥ 65 years (n = 22)		P value
	N	%	N	%	N	%
Best response
CR	1	1.8	1	2.9	0	0.0	1.000
PR	23	40.4	15	42.9	8	38.1	0.784
SD	20	35.1	13	37.1	7	33.3	0.953
PD	13	22.9	7	20.0	6	28.6	0.524
ORR	24	42.1	16	45.7	8	38.1	0.577
DCR	44	77.2	29	80.0	15	71.4	0.313

CR complete response, DCR disease control rate, MR marginal response, ORR overall response rate, PR partial response, PD progressive disease, SD stable disease

p value refers to Chi-squared test for < 65 years and ≥ 65 years

Univariate and multivariate analysis

The univariate cox regression analysis showed an influence on survival for ECOG = 0 compared to ECOG ≥ 1 (HR: 2.62; 95% CI 1.36, 5.04; p = 0.004) and for hemoglobin (HR: 0.87; 95% CI 0.77, 0.99; p = 0.034). Both parameters remained statistically significant independent predictors of survival in multivariate analysis with only marginally changed HR (ECOG = 0 vs. ECOG ≥ 1: HR 2.69, 95% CI 1.05, 6.90; p = 0.040; hemoglobin: HR 0.81, 95% CI 0.68, 0.96, p = 0.015) (Table 4).

Table 4.

Univariate and multivariate time‐to‐event analysis of baseline and therapy characteristics

Parameters	P value	HR	HR 95% CI
			Under	Upper
Univariate analysis
Age grouped	0.633	0.864	0.475	1.573
ECOG = vs 1/2	0.004	2.616	1.358	5.040
Locally advanced	0.229	Reference
Locally recurrent	0.836	0.918	0.407	2.068
Metastatic	0.104	0.267	0.054	1.313
Gastric cancer vs AEG	0.924	0.971	0.529	1.780
Lauren-intestinal type	0.844	Reference
Lauren-diffuse type	0.866	0.933	0.416	2.091
Lauren-mixed type	0.563	0.705	0.215	2.310
Signet-ring cell carcinoma	0.668	1.401	0.301	6.513
Initial tumor resection	0.646	0.867	0.472	1.595
Hemoglobin	0.034	0.873	0.770	0.990
LDH	0.839	1.000	0.999	1.001
CA19-9	0.627	1.000	1.000	1.000
Liver metastases	0.479	1.233	0.690	2.203
Peritoneal metastases	0.597	1.172	0.651	2.111
DCR	0.150	0.647	0.358	1.170
ORR	0.071	0.552	0.289	1.053
Second-line SC	0.238	0.705	0.394	1.260
Multivariate analysis
ECOG 0 vs 1/2	0.040	2.685	1.045	6.899
Hemoglobin	0.015	0.807	0.679	0.959

CA19-9 carbohydrate antigen 19-9, DCR disease control rate, ECOG Eastern Cooperative Oncology Group performance status, LDH lactate dehydrogenase, ORR objective response rate

In categorial parameters, HR refers to the parameter value marked as reference

In compliance with Cox regression analysis, the subgroup analysis for overall survival showed a statistically significant survival benefit for patients with ECOG = 0 compared to ECOG ≥ 1. Patients with the best performance status (ECOG = 0) showed a median overall survival of 12.3 months (95% CI 8.7, 15.8), while the others (ECOG ≥ 1) survived for a median 5.0 months (95% CI 0.7, 9.3) (HR 2.62, 95% CI 1.36, 5.04, log-rank p = 0.003) (Fig. 3B). Additionally, the ECOG = 0 group had a longer non-statistically significant PFS (8.2 months; 95% CI 7.0, 9.4 vs. 4.6 months; 95% CI 2.3, 7.0) (Fig. 3D).

Toxicity

The frequency of adverse events is listed in Table 5. The most common grade 3 or higher adverse events were neutropenia (40%), anemia (22%) and fatigue (16%) with even distribution in both groups. Less common adverse events, such as leukopenia (10%), pain (12%) or fever (14%), occurred with the same frequency in both groups. In the < 65 years group, one patient died of an atypical pneumonia during a neutropenia episode. Apart from this case, no treatment-related mortality has been observed.

Table 5.

Adverse events

Parameter	All (n = 50)	< 65-years (n = 30)	≥ 65 years (n = 20)	P value
Hematological toxic effects
Neutropenia	20 (40)	11 (36.7)	9 (45.0)	0.621
Anemia	11 (22)	8 (26.7)	3 (15.0)	0.489
Leukopenia	5 (10)	3 (10.0)	2 (10.0)	1.000
Thrombopenia	2 (4)	1 (3.3)	1 (5.0)	1.000
Non-hematological toxic effects
Icterus	2 (4)	2 (6.7)	0 (0)	0.510
ALT—elevation	2 (4)	2 (6.7)	0 (0)	0.510
AST—elevation	2 (4)	2 (6.7)	0 (0)	0.510
Fatigue	8 (16)	4 (13.3)	4 (20.0)	0.697
Nausea	3 (6)	1 (3.3)	2 (10.0)	0.556
Diarrhea	1 (2)	1 (3.3)	0 (0)	1.000
Vomiting	1 (2)	0 (0)	1 (5.0)	0.4000
Dysphagia	3 (6)	3 (10.0)	0 (0)	0.265
Mucositis	1 (2)	0 (0)	1 (5.0)	0.400
Constipation	0 (0)	0 (0)	0 (0)	1.000
Weight loss	2 (4)	1 (3.3)	1 (5.0)	1.000
Pain	6 (12)	4 (13.3)	2 (10.0)	1.000
Edema	1 (2)	1 (3.3)	0 (0)	1.000
Thromboembolic event	0 (0)	0 (0)	0 (0)	n.a
Infection	2 (4)	2 (6.7)	0 (0)	0.504
Fever	7 (14)	0 (0)	0 (0)	n.a
Polyneuropathy	2 (4)	1 (3.3)	1 (5.0)	1.000

ALT alanine transaminase, AST aspartate

p value refers to Chi-squared test for < 65 years and ≥ 65 years

Discussion

The present study aimed to investigate patient outcome with the FLOT chemotherapy regimen in advanced and metastatic GC and AEG in a real-life unselected cohort of patients. In particular, impact of age and other factors, such as the ECOG scale, were analyzed to define further parameters for decision making regarding first-line therapy with FLOT.

Since the FLOT-4 trial established FLOT as the new therapy standard in curative intended perioperative setting in Europe and North America, the efficacy of FLOT in palliative setting has been controversial and standard first-line therapy in HER2-negative and PD-L1-negative patients with a platin/5-FU doublet might be insufficient for young and fit patients (Al-Batran et al. 2019; Epistola and Chao 2020; Jimenez-Fonseca et al. 2021). Additionally, real-world data outside of prospective trials concerning survival and toxicity is scarce, especially in older patients. Therefore, we retrospectively analyzed data of 57 patients treated with FLOT as first-line chemotherapy in palliative intention and additionally investigated the clinical outcome of patients > 65 years of age by the time of therapy start.

Our entire cohort of patients had a median OS of 10.4 months and a median PFS of 7.8 months, which is in line with results of earlier randomized trials with OS of 10.0–13.0 months and PFS of 5.2–6.3 months PFS (Al-Batran et al. 2008; Al-Batran et al. 2017; Anter and Abdel-Latif 2013; Knödler et al. 2018). The retrospective trial of Gürler et al. (2022) described a slightly longer OS (13.3 months) and a clearly longer PFS (10.9 months). The authors argue that the high rate of prophylactic Granulocyte-Colony Stimulating Factor application (G-CSF) in the FLOT arm might have contributed to less hematological adverse events and consecutively to later dose reduction or therapy cessation, generating longer PFS (Gürler et al. 2022). Earlier studies with taxane-based triplet therapies analyzing OS and PFS in patients with GC and AEG showed similar results to our trial in a palliative therapy setting. The V325 trial found an OS of 9.2 months for DCF and Wang et al. showed an OS of 10.2 months in their modified DCF arm (Van Cutsem et al. 2006; Wang et al. 2016). A phase II trial reported an OS of 10.3 months for DOF (docetaxel, oxaliplatin, fluoruracil) in patients with advanced GC and AEG (Rosenberg et al. 2019). Additionally, epirubicin-based triplet therapies in historical trials demonstrated lower OS of 8.9–9.4 months in advanced and metastatic GC and AEG (Ross et al. 2002; Webb et al. 1997).

While the DCR in the present study was 77.2% and therefore comparable with data of the randomized trials (61.0–80.8%), the retrospective trial of Gürler et al. (2022) demonstrated a DCR of 91.7% in their FLOT arm. ORR was only 42.1%, whereas prospective studies showed an ORR of 46.0–57.7% and retrospective analysis had an ORR of 63.9% (Al-Batran et al. 2008, 2017; Anter and Abdel-Latif 2013; Gürler et al. 2022; Knödler et al. 2018). Further taxane-based triplet regimens, such as like DCF and DOF, found an ORR of 33–49% and 50–73%, respectively (Liu et al. 2018; Rosenberg et al. 2019; Shah et al. 2015; Van Cutsem et al. 2006; Wang et al. 2016). Similar ORR were detected under epirubicin-based triple therapies as ECF (epirubicin, cisplatin, fluorouracil; 41–45%), ECX (epirubicin, cisplatin, capectitabine; 47%), EOF (epirubicin, oxaliplatin, fluorouracil; 42%) and EOX (epirubicin, oxaliplatin, fluorouracil; 48%) (Cunningham et al. 2008; Ross et al. 2002; Webb et al. 1997). One reason for our relatively low ORR might be the combination of initially reduced dose administration (14%) on the one hand, and the high rate of dose reduction due to adverse events and toxicity (52.6%) on the other hand. Dose reduction at the start as well as during therapy with FLOT was carried out in order to preserve the patient’s quality of life and avoid serious adverse events. In fact, this is the most crucial challenge in palliative therapy setting: the balance between quality of life by avoiding adverse events and efficient treatment by minimizing cancer-related symptoms and prolonging survival. In prospective trials, cohorts are often younger and fitter with less comorbidities than patients in a real-world setting. In the prospective trials evaluating FLOT in advanced GC and AEG, the median age was 60 years and 52 years while nearly 40% of our real-world cohort were older than 65 years with a median age of 69 years (Al-Batran et al. 2008; Anter and Abdel-Latif 2013). Due to expected changes in pharmacokinetics and pharmacodynamics, clinicians are often uncertain about the best chemotherapy regimen and dose, suspecting higher rates of toxicity. Hence, this lack of information can result in undertreatment of older patients (Hall et al. 2017; Swaminathan and Swaminathan 2015).

Looking at the median age at diagnosis of GC and AEG in Germany (72 years), there is a clear mismatch between patients included in prospective (median age 52–60 years) and retrospective (median age 58 years) trials analyzing FLOT and patients treated in a real-world setting (Al-Batran et al. 2008; Anter and Abdel-Latif 2013; Gürler et al. 2022; Knödler et al. 2018; Robert Koch-Institut 2019). In order to contribute to this current scarce data landscape, we divided our real-world cohort into two groups, generating two subgroups with a median age of 51.7 years and 69 years, respectively. Notably, there was no statistically significant difference in overall survival between the younger (10.4 months) and older group (10.4 months) (HR 0.86; 95% CI 0.48, 1.57; log-rank p = 0.632).

A pooled analysis of three randomized trials showed a similar benefit for patients with esophageal cancer, AGE and GC > 70 years for treatment with ECF, FAMTX (methotrexate, fluorouracil, doxorubicin), MCF (mitomycin C, cisplatin, fluorouracil), fluorouracil mono therapy or fluorouracil with mitomycin C compared to younger patients. The authors concluded that, compared to younger patients, older patients show similar outcomes regarding symptomatic response, tumor regression, survival and toxicities (Trumper et al. 2006).

The FLOT 65 + trial prospectively investigated therapy efficacy and toxicity of FLOT in elderly patients with advanced or metastatic GC and AEG in comparison to FLO. The median age of the FLOT group was 69 years. The ORR was higher (49% vs. 32%) and the OS (17.3 months vs. 10.4 months) was longer, while the PFS (9.0 months vs. 8.0 months) showed matchable results with our < 65 years subgroup. A subgroup analysis of the FLOT65 + trial suggested a significant therapy benefit under FLOT for locally advanced cancers compared to metastatic disease concerning OS (24.2 months vs. 7.3 months) (Al-Batran et al. 2013). Our results indicate a similar effect with an OS of 26.2 months in locally advanced cancer compared to metastatic disease (10.4 months) in the > 65 years group (data not shown). Due to the small number of patients, especially in the locally advanced group, these results should be interpreted with caution. Interestingly, within the FLOT65 + trial, (Kripp et al. 2014) analyzed the impact of FLOT on the quality of life compared to doublet therapy with FLO (Kripp et al. 2014). Although toxicity was higher in the FLOT group, no negative impact by addition of docetaxel on QOL parameters could be demonstrated. Therefore, the authors stated that especially elderly patients in need of tumor remission (tumor obstruction, downsizing) are eligible candidates for FLOT, without increased risk of deterioration of quality of life.

In accordance with these findings, our uni- and multivariate analysis did not identify patient age but ECOG performance status as an independent predictor of survival. Further, the influence was also visible in log-rank test results that showed a statistically significant prolonged OS for the ECOG = 0 group compared to the ECOG = 1/2 group. Although ECOG is known as a prognostic factor in cancer patients, there is hardly any data showing ECOG as the most important factor in predicting survival on a therapy with FLOT in patients with advanced or metastatic GC and AEG rather than age.

The analysis of adverse events in the present study showed an overall frequency of any adverse event (grade 3–5 corresponding CTCAE V5.0) of 72% with no difference between the younger and older group. The most common hematological toxic effects were neutropenia (40%) and anemia (22%), which were mostly manageable. Non-hematological toxic effects were fatigue (16%), pain (12%) and fever (14%). Earlier prospective trials reported a similar rate of neutropenia (45–48%) but lower rates of anemia (4%) (Al-Batran et al. 2008; Jimenez-Fonseca et al. 2021). Interestingly, the retrospective results of Guerler et al. showed a lower frequency of neutropenia (19%) and anemia (8%), which can be partially explained by the above-mentioned high rate of prophylactic G-CSF application. Regarding other taxane-based triplets in prospective trials, there is an already known and accepted difference in safety profile between DCF and DOF. Neutropenia and anemia corresponding grade 3–5 occurred more often in the DCF therapy arms (61–82% for neutropenia, 5–65% for anemia) than in the DOF therapy arms (19–23%, 0–7%) (Liu et al. 2018; Rosenberg et al. 2019; Van Cutsem et al. 2006; Wang et al. 2016). Epirubicin-based triplets show similar rates of neutropenia (28–51%) and anemia (7–13%) (Cunningham et al. 2008; Ross et al. 2002; Webb et al. 1997). The prospective FLOT 65 + trial reported that 82% of patients had National Cancer Institute Common Toxicity Criteria (NCI-CTC) grade 3/4 adverse events and dose modifications were needed in 44%. Our subgroup of elderly patients showed similar results in terms of dose modification (50%) and rate of overall adverse events (75%), while we used the CTCAE classification system instead of NCI-CTC. Of note, FLOT patients showed a neutropenia rate of 53% and anemia rate of 11% (both grade 3/4 NCI-CTC).

Since the phase III ToGA trial confirmed the beneficial effect of additional application of trastuzumab with cisplatin and fluoropyrimidine in patients with HER2 protein overexpression as first-line therapy, future and ongoing trials are investigating the effects of further immunotherapy (Bang et al. 2010). A subgroup analysis of an observational study in Germany showed a median OS for patients with metastatic GC or AEG who were treated with FLOT + trastuzumab of 15.5 months (95% CI 9.5–20.8) (Al‐Batran et al. 2020). Eight patients of our cohort additionally received trastuzumab as first-line treatment and had a median OS of 12.3 months. For patients with advanced HER2-negative GC and AEG, the addition of the PD-1 antibody nivolumab to systemic chemotherapy showed promising results in the CheckMate 649 study. OS for the nivolumab plus chemotherapy group was 14.4 months compared to 11.1 months in the chemotherapy alone group. The applied chemotherapy regimens were a combination of cisplatin and fluoropyrimidine (Janjigian et al. 2021). Therapy experiences with PD-1 or PD-L1 checkpoint inhibitors in combination with FLOT are scarce in palliative therapy setting and should be evaluated in future studies.

Our study has some limitations. Due to the retrospective and monocentric design, a selection bias cannot be excluded, although both groups were well balanced. Also, as the number of patients is limited, subgroup analysis should be interpreted with caution. A further evaluation of GC vs. AEG with respect to the age will be also interesting. However, due to the sample size of this study, analysis of the required four groups of patients does not lead to any statistically significant results because the groups are too small. Nevertheless, we included an evaluation of GC vs AEG as a parameter (gastric cancer vs AEG) in the univariate analysis, showing no significant difference between GC and AEG as prognostic factors for survival. Inter-study comparisons should be considered as hypothesis generating and no clinical assumptions should be taken based on these data. It is a strength of our study, however, that it contributes to the scarce data field of administration of FLOT in patients with advanced and metastatic GC and AEG in daily practice. Additionally, we focused our analysis on patient age and the relevance of ECOG for outcome with FLOT in these patients. Since nearly two thirds of the patients with GC and AEG are > 65 years old, and as these patients are underrepresented in clinical studies, our results are of utmost value.

In summary, this study suggests an acceptable efficacy of the FLOT chemotherapy regimen as a first-line therapy in patients with advanced and metastatic GC and AEG from a real-life unselected cohort, including elderly patients. Furthermore, our data showed that ECOG performance status seems to have a greater prognostic value concerning OS than the age of the patients and should therefore be considered in therapeutic decision making for elderly patients with GC and AEG in high need of tumor remission. Although the toxicity was relatively high in older as well as younger patients, it was manageable.

Author contributions

Study concept, data collection and analysis, and interpretation of data were designed and performed by CM, AT and MAG-C. The first draft of the manuscript was written by CM, AT and MAG-C and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the following grants awarded to M.A. Gonzalez-Carmona: GO 1874/1-2 grant from Deutsche Forschungsgemeinschaft (DFG), BONFOR from the University of Bonn, grant number 109255 from Deutsche Krebshilfe (German Cancer Aid) and grant from the Reuthersche endowment fund of the University of Bonn.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of interest

Author MG has contributed to advisory boards for Roche, Eisai, BMS, MSD and AZ. However, these activities have no potential conflicts of interest with the manuscript. None of the other authors have any potential conflicts (financial, professional or personal) that are relevant to the manuscript.