PLOS ONE RESEARCH ARTICLE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Almudena Cotes Sanchı́s1, Javier Gallego ID2*, Raquel Hernandez3, Virginia Arrazubi4, Ana Custodio5, Juana Marı́a Cano6, Gema Aguado7, Ismael Macias8, Carlos Lopez9, Flora López10, Laura Visa11, Marcelo Garrido12, Nieves Martı́nez Lago13, Ana Fernández Montes14, Marı́a Luisa Limón15, Aitor Azkárate ID16, Paola Pimentel17, Pablo Reguera18, Avinash Ramchandani19, Juan Diego Cacho20, Alfonso Martı́n Carnicero21, a1111111111 Mónica Granja22, Marta Martı́n Richard23, Carolina Hernández Pérez24, Alicia Hurtado25, a1111111111 Olbia Serra26, Elvira Buxo27, Rosario Vidal Tocino ID28, Paula Jimenez-Fonseca29, a1111111111 Alberto Carmona-Bayonas30 a1111111111 1 Medical Oncology Department, Hospital General Universitario de Elda, Alicante, Spain, 2 Medical a1111111111 Oncology Department, Hospital General Universitario de Elche, Elche, Spain, 3 Medical Oncology Department, Hospital Universitario de Canarias, Tenerife, 4 Medical Oncology Department, Complejo Hospitalario de Navarra, Pamplona, Spain, 5 Medical Oncology Department, Hospital Universitario La Paz, CIBERONC CB16/12/00398, Madrid, Spain, 6 Medical Oncology Department, Hospital General Universitario de Ciudad Real, Ciudad Real, Spain, 7 Medical Oncology Department, Hospital Universitario Gregorio Marañón, Madrid, Spain, 8 Medical Oncology Department, Hospital Universitario Parc Tauli, Sabadell, Spain, OPEN ACCESS 9 Medical Oncology Department, Hospital Universitario Marqués de Valdecilla, IDIVAL, Santander, Spain, Citation: Cotes Sanchı́s A, Gallego J, Hernandez R, 10 Medical Oncology Department, Hospital Universitario Doce de Octubre, Madrid, Spain, 11 Medical Arrazubi V, Custodio A, Cano JM, et al. (2020) Oncology Department, Hospital Universitario El Mar, Barcelona, Spain, 12 Medical Oncology Department, Second-line treatment in advanced gastric cancer: Pontificia Universidad Católica de Chile, Santiago de Chile, Chile, 13 Medical Oncology Department, Complejo Hospitalario Universitario de A Coruña, Coruña, Spain, 14 Medical Oncology Department, Data from the Spanish AGAMENON registry. PLoS Complejo Hospitalario de Orense, Orense, Spain, 15 Medical Oncology Department, Hospital Universitario ONE 15(7): e0235848. https://doi.org/10.1371/ Virgen del Rocı́o, Sevilla, Spain, 16 Medical Oncology Department, Hospital Universitario Son Espases, journal.pone.0235848 Mallorca, Spain, 17 Medical Oncology Department, Hospital Santa Lucı́a, Cartagena, Spain, 18 Medical Editor: Hakan Buyukhatipoglu, Harran Universitesi, Oncology Department, Hospital Universitario Ramón y Cajal, Madrid, Spain, 19 Medical Oncology Department, Hospital Universitario Insular de Gran Canaria, Las Palmas de Gran Canaria, Spain, 20 Medical UNITED STATES Oncology Department, Hospital Universitario Marqués de Valdecilla, IDIVAL, Santander, Spain, 21 Medical Received: April 6, 2020 Oncology Department, Hospital San Pedro, Logroño, Spain, 22 Medical Oncology Department, Hospital Universitario Clı́nico San Carlos Madrid, Spain, 23 Medical Oncology Department, Hospital Universitario Accepted: June 24, 2020 Santa Creu i Sant Pau, Barcelona, Spain, 24 Medical Oncology Department, Hospital Universitario Nuestra Señora de la Candelaria, Tenerife, Spain, 25 Medical Oncology Department, Hospital Universitario Published: July 31, 2020 Fundación Alcorcón, Madrid, Spain, 26 Medical Oncology Department, Catalan Institute of Oncology, Peer Review History: PLOS recognizes the L’Hospitalet, Spain, 27 Medical Oncology Department, Hospital Universitario Vall d’Hebron, Barcelona, benefits of transparency in the peer review Spain, 28 Medical Oncology Department, Complejo Asistencial Universitario de Salamanca, IBSAL, Salamanca, Spain, 29 Medical Oncology Department, Hospital Universitario Central de Asturias, ISPA, process; therefore, we enable the publication of Oviedo, Spain, 30 Hematology and Medical Oncology Department, Hospital Universitario Morales Meseguer, all of the content of peer review and author University of Murcia, IMIB, Murcia, Spain responses alongside final, published articles. The editorial history of this article is available here: * j.gallegoplazas@gmail.com https://doi.org/10.1371/journal.pone.0235848 Copyright: © 2020 Cotes Sanchı́s et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which Abstract permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Background Data Availability Statement: All relevant data are Second-line treatments boost overall survival in advanced gastric cancer (AGC). However, within the paper and its Supporting Information files. Additional data are accessible in case of need there is a paucity of information as to patterns of use and the results achieved in actual clini- in the Agamenon registry. cal practice. PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 1 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Funding: The authors received no specific funding Materials and methods for this study. The study population comprised patients with AGC in the AGAMENON registry who had Competing interests: Dr. Javier Gallego declares received second-line. The objective was to describe the pattern of second-line therapies advisory role for Amgen, Bayer, BMS, Ipsen, Lilly, Merck, Roche, Servier, and travel grants from administered, progression-free survival following second-line (PFS-2), and post-progres- Novartis, Amgen. No other authors have sion survival since first-line (PPS). competing interests to declare. This does not alter our adherence to PLOS ONE policies on sharing data and materials. Results 2311 cases with 2066 progression events since first-line (89.3%) were recorded; 245 (10.6%) patients died during first-line treatment and 1326/2066 (64.1%) received a second- line. Median PFS-2 and PPS were 3.1 (95% CI, 2.9–3.3) and 5.8 months (5.5–6.3), respec- tively. The most widely used strategies were monoCT (56.9%), polyCT (15.0%), ramuciru- mab+CT (12.6%), platinum-reintroduction (8.3%), trastuzumab+CT (6.1%), and ramucirumab (1.1%). PFS-2/PPS medians gradually increased in monoCT, 2.6/5.1 months; polyCT 3.4/6.3 months; ramucirumab+CT, 4.1/6.5 months; platinum-reintroduction, 4.2/6.7 months, and for the HER2+ subgroup in particular, trastuzumab+CT, 5.2/11.7 months. Cor- relation between PFS since first-line and OS was moderate in the series as a whole (Ken- dall’s τ = 0.613), lower in those subjects who received second-line (Kendall’s τ = 0.539), especially with ramucirumab+CT (Kendall’s τ = 0.413). Conclusion This analysis reveals the diversity in second-line treatment for AGC, highlighting the effec- tiveness of paclitaxel-ramucirumab and, for a selected subgroup of patients, platinum rein- troduction; both strategies endorsed by recent clinical guidelines. Introduction Advanced gastric cancer (AGC) is the third leading cause of cancer death worldwide [1]. Che- motherapy (CT) is capable of improving overall survival (OS) and quality of life for individuals with AGC compared to best supportive care (BSC) [2]. In first line, platin-fluoropyrimidine schedules are the most widely recommended option [3], whereas the standard of care is the combination of trastuzumab and cisplatin-fluoropyrimidine for tumors that amplify or over- express human epidermal growth factor receptor-2 (HER2+) [4]. The benefit of first-line is limited; up to 25%-30% display progression at their first evaluation of response [5] and median progression-free survival (PFS) is 4–7 months [2], with approximately 50% of patients in suit- able conditions to receive second-line treatment after progression since first-line [6, 7]. Numerous drugs have proven activity in second-line for AGC [8, 9]. Thus, a small random- ized trial (NCT00144378) confirmed for the first time that the use of irinotecan vs BSC in sec- ond line discreetly prolonged OS [6]. In the COUGAR-2 study, docetaxel incremented OS versus BSC and likewise demonstrated a benefit in quality of life [10]. Both drugs again improved OS compared to BSC in a phase III trial [7], while the WJOG-4007 study detected no differences between them or between paclitaxel and irinotecan [9]. More recently, the use of ramucirumab plus paclitaxel vs paclitaxel in second line was seen to increase OS in all sub- groups in the RAINBOW trial [11]. For its part, the REGARD study corroborated a gain in OS with ramucirumab vs BSC [12]. Both studies with ramucirumab were bolstered by favorable PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 2 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry quality of life analyses, as well as real-world data [13–15]. This positions ramucirumab as the recommended second-line strategy, whether in combination or monotherapy [16]. There are minimal data concerning how the use of the various alternatives available for second-line treat- ment has evolved, in addition to their efficacy in actual clinical practice [17]. Moreover, pembrolizumab has demonstrated efficacy in a second line study of carcinoma of the esophagus and of the gastroesophageal junction, in the pre-specified subgroup of PDL1-CPS�10 [18], while efficacy in second-line was unproven for advanced gastric or gas- tro- gastroesophageal junction adenocarcinoma in the KEYNOTE-061 phase III study [19]. Treatment in second and successive lines for HER2+ tumors does not currently differ from the rest, given the absence of evidence in favor of anti-HER2 therapy [20, 21]. Nevertheless, these tumors are molecularly dissimilar. Based on retrospective analysis, certain individuals who do not receive first-line treat- ment until progression might profit from reintroducing platin-fluoropyrimidine doublets, when the treatment-free interval exceeds three months [22]. This subgroup of patients is excluded from most recent second-line clinical trials for AGC [11, 12], and most updated clinical guidelines consider reintroduction of the first-line to be an appropriate alternative [16]. Likewise, treatment options with proven efficacy exist in various third-line scenarios [23– 25]. This availability of options beyond first line makes survival susceptible to the outcomes associated with successive lines of treatment, which could have implications when designing clinical trials. We must therefore revisit the value of intermediate endpoints, such as PFS, as surrogates for OS [26–30]. Against this backdrop, we have conducted this study to evaluate patterns of use and out- comes related to each type of strategy in second line and the surrogate function of PFS in an AGC registry (AGAMENON). Material and methods Patients and design The patient population assessed derive from the Spanish AGAMENON registry that enlists the collaboration of 34 Spanish hospitals and one center in Chile and recruits consecutive cases of unresectable or metastatic, locally advanced adenocarcinoma of the stomach, gastroesophageal junction, or distal esophagus [31–39]. Eligibility criteria are: patients with AGC, aged >18 years, who received first-line treatment with polyCT routinely administered in clinical practice (two- or three-agent schedule, with or without platin) [40], and experienced tumor progression or died during first-line treatment. Those cases that had completed neoadjuvant or adjuvant treatment before 6 months were excluded. Data are managed through a website (http://www.agamenonstudy.com/) consisting of fil- ters and a query-generating system to guarantee reliability and control missing and inconsis- tent data, as well as errors. Telephone and on-line monitoring (PJF) further guarantee data quality. Objectives The primary objective of this study was to describe the pattern of second-line therapies admin- istered from 2008 onward and the associated outcomes. The secondary objective was to assess the correlation between PFS and OS over time, in terms of clinical-pathological variables and use of second lines. PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 3 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Variables Post-progression survival (PPS) and PFS-2 were defined as the time between initiation of sec- ond-line and all-cause mortality or progression, respectively, censuring those event-free indi- viduals at the time of the last follow up. OS and PFS-1 were defined as the interval between commencement of first-line treatment and death for any cause or progression, respectively, censoring at last follow up. Second-line strategies were categorized as: (1) platinum-reintroduction, defined as provid- ing a second-line platin-based schedule to individuals who had received platin in first-line with no evidence of progression when plantin was stopped; (2) maintenance of trastuzumab post-progression, consisting of changing the backbone of CT in progression to first-line with- out discontinuing trastuzumab; (3) regimens containing ramucirumab included use of pacli- taxel, irinotecan, or other cytotoxics in combination with ramucirumab; (4) ramucirumab in monotherapy; additionally, patients could receive (5) monoCT or (6) polyCT. Statistics Survival functions were Kaplan-Meier estimates. Correlation between PFS and OS was quanti- fied by Kendall’s τ associated with Clayton’s copula models for bivariate survival data. Sensitiv- ity to second line was evaluated using multivariable binary logistic regression (covariates were HER2 status, histological subtype, signet ring cells, hepatic tumor burden, number of meta- static sites, and interval of time since withdrawal of platin in first-line). Continuous variables were analyzed by means of restricted cubic splines. Treatment effect was appraised using a Cox multivariable proportional hazards model. No data-driven criteria were used for the model specification. The covariates for the multivariable model were chosen by theoretical considerations, as recommended in the literature [41]. Thus, ECOG PS (<2, �2), Lauren’s his- topathological subtype (intestinal, diffuse), number of metastatic sites (�2, >2), liver tumor burden (�50, >50%), HER2 status (negative, positive), PFS-1, and best response to first-line (complete or partial response, stable disease, progressive disease) were used as confounding factors. Restricted cubic splines were used to model the non-linear effect of PFS-1. Given that it is an observational, fixed sample size study, inferences should be interpreted in accordance with the magnitude of the CI with a descriptive purpose (hypothesis generator). Analyses were performed with the R v3.1.6 software package, with rms and Copula.surv libraries [42, 43]. Ethics statement This study has Compliance with Ethical Standards. This study was approved in November, 4 th 2014 by the Ethics, Research and Investigation Committee in Hospital Morales Meseguer, Mur- cia, Spain. The Research Ethics Committee from Morales Meseguer General University Hospital first, and then all the rest of Autonomous Communities and participating hospitals approved the study. The Spanish Agency of Medicines and Medical Devices categorized this study as a post- marketing, prospective follow-up study. In every alive prospective or retrospective registered patient, written informed consent was obtained in order to be included in the study. Participants who were not alive at data collection had previously provided written informed consent to use their medical records for the purposes of research. This was carried out according to the require- ments stated in the international guidelines regarding carrying out epidemiological studies and put forth in the International Guidelines for Ethical Review of Epidemiological Studies (Council for the International Organizations of Medical Sciences–CIOMS-, Geneva, 1991), as well as the Declaration of Helsinki (Seoul revision, October, 2008). This document defines the principles that must be scrupulously respected by any and all personas involved in the research. The treat- ment, communication, and conveyance of the personal data of all participants was adapted to PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 4 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry the Organic Law 3/2018, dated December 5, regarding the Protection of Personal Data requiring approval by a Clinical Research Ethics Board (CREB). Results Patients and second-line treatments At the time of analysis, 2311 cases had been recorded that met eligibility criteria, 2066 progres- sion events since first-line (89.3%) and 2103 deaths (90.9%). Of the latter, 245 (10.6%) died during first-line. Median PFS-1 was 5.6 months (95% CI, 5.5–5.9), while median OS was 10.2 months (95% CI, 9.8–10.7). Second-line therapy was given to 1326/2066 (64.1%); 366 (17.7%) received three lines, and 98 (4.7%), four or more. Baseline characteristics are summarized in Table 1. The most common strate- gies were: monoCT 56.9% (n = 755), polyCT 15.0% (n = 199), ramucirumab+CT 12.6% (n = 167), platinum-reintroduction 8.3% (n = 110), trastuzumab-continuing schedules 6.1% (n = 81), and ramucirumab monotherapy 1.1% (n = 14). S1 Table displays characteristics per strategy used. In subjects treated with platinum-reintroduction, the reason for discontinuing platinum in first-line before progression was: having completed the number of cycles established by their center’s protocol (71.8%), toxicity (18.2%), patient request (2.7%), and other reasons (7.3%). Of the participants who received trastuzumab in second-line, 14/81 (17.3%) had not received it in first-line. Trastuzumab was withheld from those 14 patients in first-line because their HER2 status was unavailable (7 cases); due to cardiac comorbidity (n = 3), or oncologist’s decision (n = 4). S2 Table shows the data of use of these strategies by HER2 status. Fig 1 illustrates the usage trend of these strategies over time, revealing that the only one with an upward trend is the incorporation of ramucirumab from 2012 onward. Response rate to second lines The response rate to second lines was 12.7% (n = 168); 28.5% had stable disease (n = 378) and the disease control rate (response or stable disease) was 41.2%. Progression occurred in 55.1% (n = 731) and information regarding response was unavailable for 3.7% (n = 49) of the cases. Fig 2 illustrates response rates by second-line strategy and HER2 status. For descriptive pur- poses, the probability of response to second-line has been represented depending on histopath- ological subtype, prior response to first-line, HER2 status and platin-free interval (S1–S3 Figs; S3 Table). The underlying model suggests differences according to these features. For instance, in diffuse tumors not responding previously to platin, the odds of achieving response to ramu- cirumab+CT vs monoCT increased as a function of platin-free interval: odds ratio (OR) 1.53 (95% CI, 0.69–3.72) at one month; OR 2.22 (95% CI, 1.30–3.81) at three months, and OR 2.90 (95% CI, 1.41–5.97) at six months. Plots with the probability of response for HER2+ and HER2-negative tumors can be seen in S2 and S3 Figs, respectively. Survival endpoints in second lines At the time of analysis, 93.7% had suffered a progression event and 86.2% died after second- line. Median PFS-2 and PPS were 3.1 (95% CI, 2.9–3.3) and 5.8 months (95% CI, 5.5–6.3), respectively. Fig 3 presents survival for both endpoints. Survival endpoints for each treatment group are laid out in Table 2. The highest median PFS-2 and PPS were obtained with platinum-reintroduction: 4.2 (95% CI, 3.3–5.0) and 6.7 months (95% CI, 5.5–10.2) and with ramucirumab+CT: 4.1 (95% CI, 3.4–5.2) and 6.5 months (95% CI, 5.5–8.7), respectively. In the case of HER2+ tumors, trastuzumab-containing regi- mens achieved a median PFS-2 of 4.8 months (95% CI, 3.6–5.7) and PPS of 10.5 months (95% PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 5 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Table 1. Characteristics at the time of diagnosis. Variables Total, n (%), Patients receiving 2nd-line, n (%), n = 2311 n = 1326 Age, median (range) 64 (20–89) 63 (20–86) Sex, female 672 (29.1) 370 (27.9) Lauren subtype Diffuse 745 (32.2) 409 (30.8) Intestinal 991 (42.8) 589 (44.4) Mixed 107 (4.6) 61 (4.6) Not Available 468 (20.2) 267 (20.1) Signet ring cells 657 (28.4) 354 (26.7) HER2-positive 502 (21.7) 318 (23.9) ECOG-PS basal 0 533 (23.2) 361 (27.2) 1 1457 (63.0) 849 (64.0) �2 321 (12.8) 114 (8.8) Tumor stage at diagnosis, locally advanced 134 (18.1) 74 (5.5) unresectable Histological grade 1 225 (9.7) 152 (11.5) 2 628 (27.2) 361 (27.2) 3 933 (40.4) 525 (39.6) Not available 525 (22.7) 288 (21.7) First-line treatment Anthracycline-based 464 (20.1) 285 (21.5) Cisplatin-based doublet 472 (20.4) 302 (22.8) Docetaxel-based 276 (11.9) 154 (11.6) Irinotecan-based 43 (1.9) 25 (1.9) Oxaliplatin-based 911 (39.4) 498 (37.6) Other 145 (6.3) 62 (4.7) Metastases sites Ascites 545 (23.6) 289 (21.8) Peritoneal 1011 (43.7) 559 (42.2) Bone 235 (10.2) 112 (8.4) Lung 308 (13.3) 185 (14.0) Liver 876 (37.9) 522 (39.4) Burden of liver disease >50% 453 (19.6) 247 (18.6) Number of metastases >2 629 (27.2) 332 (25.0) Primary tumor site Esophagus 183 (7.9) 113 (8.5) GEJ 306 (13.2) 166 (12.5) Stomach 1822 (78.8) 1046 (79.0) PFS-1 5.6 (5.4–5.9) 6.8 (6.5–7.1) Best response to first-line Complete response 22 (1.0) 18 (1.4) Partial response 661 (28.6) 465 (35.1) Stable disease 1028 (44.5) 570 (43.0) Progression disease 600 (26.0) 273 (20.6) Abbreviations: ECOG-PS, Eastern Cooperative Group Performance Status; GEJ, gastroesophageal junction. https://doi.org/10.1371/journal.pone.0235848.t001 PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 6 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Fig 1. Time trends in the use of second-line schedules based on HER2 status. Abbreviations: polyCT, polychemotherapy; monoCT, monochemotherapy; Ram+CT, ramucirumab+chemotherapy. https://doi.org/10.1371/journal.pone.0235848.g001 CI, 5.5–12.1). In a sensitivity analysis, after excluding 14 subjects without first-line trastuzu- mab, the remaining patients obtained a similar median PFS-2/PPS, 4.80 (CI 95%, 3.45–5.75) and 10.8 months (CI 95%, 7.1–14.6) respectively. MonoCT yielded the worst results with median PFS-2 of 2.6 months (95% CI, 2.4–2.7) and PPS of 5.1 months (95% CI, 4.6–5.7). In the multivariable Cox model, taking monoCT as reference, ramucirumab+CT (HR 0.62; 95% CI, 0.51–0.74), platinum-reintroduction (HR 0.76; 95% CI, 0.61–0.94), polyCT (HR 0.81; 95% CI, 0.69–0.96), and trastuzumab+CT (HR 0.58, 95% CI; 0.44–0.77, in HER2+) were associated with better PFS-2. The data as per HER2 subtype are detailed in Table 2 and Fig 4. Correlation of PFS & OS with each treatment strategy The correlation between PFS-1 and OS is moderate in the complete series (n = 2311, Kendall’s τ = 0.613), lower in individuals who received a second-line (Kendall’s τ = 0.539). The possibility of having effective second lines available dilutes the surrogate value of PFS-1, principally in individuals who receive CT-ramucirumab. Correlations for each treatment strategy are as follows: ramuciru- mab+CT (Kendall’s τ = 0.413), polyCT (Kendall’s τ = 0.503), monoCT (Kendall’s τ = 0.539), trastu- zumab+CT (Kendall’s τ = 0.566), and platinum-reintroduction (Kendall’s τ = 0.585) (S4 Table). Discussion Within the context of AGC, second-line therapy has been proven to enhance OS compared to BSC to a statistically significant extent [44]. In a meta-analysis of 10 clinical trials, polyCT was PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 7 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Fig 2. Response rates according to HER2 and strategy. Abbreviations: Pl reint, platinum reintroduction; poly-CT, polychemotherapy; P, paclitaxel; Ram, ramucirumab; CT, chemotherapy; Trastu, trastuzumab; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; NA, not available. � Paclitaxel was the cytotoxic used in all patients with HER-negative tumors who received ramucirumab+chemotherapy, whereas paclitaxel and other cytotoxics were associated with ramucirumab in HER+ tumors. https://doi.org/10.1371/journal.pone.0235848.g002 more effective than monoCT [45], while a network meta-analysis suggests that the combina- tion of paclitaxel+ramucirumab is most likely to be the best schedule available to date [46]. However, data with reference to real-world use of second lines (without the usual clinical trial selection biases) are scant. Moreover, there is a paucity of information about the strategies cli- nicians apply pragmatically, such as platinum-reintroduction or using trastuzumab beyond progression. To investigate these aspects, we evaluated the use of second-line in the 64.1% of the AGC registry patients who received it, a percentage similar to that observed by other authors [6, 7]. The individuals who received second-line tended to be those who had benefitted most from first-line, with longer PFS-1. Our data corroborate that polyCT and CT+ramucirumab is superior to monoCT in daily practice [45, 46]. Bearing in mind the safety profile of each strategy in indirect comparisons, and the available scientific evidence, this would endorse the established role of ramucirumab+- paclitaxel as the current standard of second-line treatment in AGC. The AGAMENON data endorse this consideration, by revealing a trend toward increased use of ramucirumab, alone and in combination, compared to the remaining second-line strategies, which are declining. Furthermore, the study indicates that histopathological subtype, therapy administered, time since platin withdrawal, and better response to first-line might be among the factors associated with response. In particular, chemosensitivity to second-line are continuously and non- PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 8 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Fig 3. Survival curves for PFS-2 and PPS (n = 1326). Abbreviations: PFS-2, progression-free survival to second-line of treatment; PPS, post-progression survival. https://doi.org/10.1371/journal.pone.0235848.g003 linearly related to the platin-free period. PFS-1 is a known and consistent predictive factor dur- ing second-line therapy [44, 47]. Diffuse tumors are more refractory to second-line treatment than the intestinal subtype, although this depends on the interaction with the platin-free inter- val. Thus, even in adverse scenarios, such as treatment-resistant diffuse tumors, the probability of response is twofold in those exposed to ramucirumab+CT vs monoCT, indicating that treat- ment choice is key to achieving benefit. Based on retrospective analysis, reintroduction of the same drug combination should be contemplated for patients in whom first-line treatment was discontinued and time to progres- sion exceeded three months, provided that any toxicity issues have been resolved [22] and as recommended in the most recently updated guidelines [16]. In this registry, the reintroduction of firs-line platin-based therapy (10% of Her2- patients) was associated with the highest disease control rate and median PPS. These results are comparable to those of the study by Okines PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 9 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Table 2. Survival endpoints based on the strategy for HER+ and HER-negative tumors. Variables n/events Median PFS-2, months (95% CI) n/events Median PPS, months (95% CI) All Mono-CT 755/723 2.6 (2.4–2.7) 755/677 5.1 (4.6–5.7) Poly-CT 199/194 3.4 (2.7–3.9) 199/172 6.3 (5.6–7.2) Ram-CT 167/139 4.1 (3.4–5.2) 167/104 6.5 (5.5–8.7) Plat reintroduction 110/104 4.2 (3.3–5.0) 110/99 6.7 (5.5–10.2) HER2-negative Ram 14/11 2.8 (1.8-NA) 14/10 5.0 (3.0-NA) Mono-CT 592/566 2.6 (2.4–2.8) 592/527 4.9 (4.3–5.4) Poly-CT 177/172 3.4 (2.7–4.8) 177/157 6.2 (5.5–7.1) Ram-CT 125/104 3.8 (3.3–5.1) 125/84 6.5 (5.1–9.4) Plat reintroduction 100/95 4.1 (3.2–4.8) 100/91 6.6 (5.4–9.8) HER2-positive Mono-CT 163/157 2.7 (2.4–3.2) 163/150 6.7 (5.2-NA) Poly-CT 22/22 3.0 (2.5–5.7) 22/20 8.6 (5.0–14.9) Ram-CT 42/35 4.7 (3.2–6.3) 42/30 7.3 (5.5–12.1) CT + Trastuzumab 81/72 4.8 (3.6–5.7) 81/66 10.5 (5.5–12.1) Plat reintroduction 10/9 5.2 (3.1-NA) 10/8 11.7 (7.3–13.3) Abbreviations: Ram, ramucirumab; Plat, platinum; CT, chemotherapy; PFS-2, progression-free survival to second-line of treatment; PPS, post-progression survival. https://doi.org/10.1371/journal.pone.0235848.t002 et al that revealed that the reintroduction of platin was associated with median PFS-2 and PPS of 3.9 and 6.6 months, respectively [22], depending on prior chemosensitivity to platin, platin- free interval, and histological subtype. Therefore, given that platin is sometimes discontinued due to cumulative toxicity, proceeding with fluoropyrimidine until progression [48], platin reintroduction might be an especially useful option in intestinal tumors, sensitive to platin in first-line, with a prolonged platin-free interval and in the absence of residual toxicity. Another strategy arising in this real world evidence analysis is that of using trastuzumab fol- lowing progression, which in this registry accounts for 25.5% of HER+ tumors, although evi- dence for trastuzumab in second-line treatment of AGC is lacking [49, 50]. Likewise, our data corroborate the favorable prognostic effect of HER2+ status that is maintained beyond first- line [34]. Still, the reader must be mindful of the current lack of positive results in clinical trials that have assessed anti-HER2 therapy in second-line [20, 21], as well as the confirmed benefit of ramucirumab in cases in which trastuzumab was administered in first-line [51]. Finally, we have examined the surrogate function of PFS-1 within the context of the avail- ability of treatment strategies after first-line. OS has traditionally been the gold-standard end- point in clinical trials of first-line therapy for AGC; nonetheless, PFS continues to be routine in various randomized AGC studies [26–28]. The advantages of PFS include shortened study duration, smaller sample size, and the absence of interference of post-progression therapies. Overall, the use of intermediate endpoints calls for statistical proof of the validity of the surro- gate, as well as the justification of the clinical value that delaying progression has for the patient’s quality of life [52]. Our data reveal that the correlation between PFS-1 and OS is mod- erate in actual practice, with a magnitude slightly lower than that reported in the literature [53]. In fact, the possibility of having effective second lines available dilutes the surrogate value of PFS-1, principally in individuals who receive CT-ramucirumab. Bearing in mind the grad- ual increase in the use of ramucirumab in our series, this would call into question the appro- priateness of substituting OS for PFS-1 in studies of first line in AGC. PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 10 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Fig 4. Survival functions since initiation of second-line by HER2 status and treatment strategy. Abbreviations: PFS-2, progression-free survival to second-line of treatment; PPS, post-progression survival; CT, chemotherapy; Pacli, paclitaxel. https://doi.org/10.1371/journal.pone.0235848.g004 There are several limitations implicit in observational studies, such as this one, in which the criteria that mediate in the decisions regarding second lines depend on the evolution of the disease that are not initially present and, as such, are difficult to capture in a registry of this kind. In addition, patients in this registry received first-line treatment with polyCT, excluding more fragile patients who were only candidates for monoCT. Nevertheless, survival endpoints and baseline characteristics are adequately typified through regular reviews and updating of the information. PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 11 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Conclusion In short, our study provides the largest real world practice data set regarding the use of second lines for AGC, backing up the scientific evidence derived from previous clinical trials and smaller retrospective analyses. Our analysis reveals the diversity in second-line treatment for AGC, highlighting the effectiveness of paclitaxel-ramucirumab and, for a selected subgroup of patients, platinum reintroduction; both strategies endorsed by recent clinical guidelines. Addi- tionally, it disputes the role of PFS as a surrogate for OS with the progressive incorporation of more efficacious strategies in successive lines of treatment. Supporting information S1 Table. Characteristics at the time of diagnosis by second-line strategy. (DOCX) S2 Table. Chemotherapy regimens used based on HER2 status. (DOCX) S3 Table. Response rate depending on HER2 status and treatment strategy. (DOCX) S4 Table. Correlation between progression-free survival to first-line of treatment and over- all survival. (DOCX) S1 Fig. Probability of response to second-line depending on platinum-free interval, response to first-line, and subtype (all patients). (TIFF) S2 Fig. Probability of response to second-line depending on platinum-free interval, response to first-line, subtype and second-line strategy (HER2-negative). (TIFF) S3 Fig. Probability of response to second-line depending on platinum-free interval, response to first-line, and second-line strategy (HER2-positive subset). (TIFF) Acknowledgments We thank Priscilla Chase Duran for editing the manuscript, Natalia Cateriano, Miguel Vaquero, and IRICOM S.A. for supporting the registry website. We are indebted to all patients, as well as to AGAMENON centres and investigators who particpated in this research and made it possible. Disclaimers iii. AGAMENON registry is part of the Evaluation of Results and Clinical Practice Section included in the Spanish Society of Medical Oncology (SEOM). Author Contributions Conceptualization: Almudena Cotes Sanchı́s, Javier Gallego, Paula Jimenez-Fonseca, Alberto Carmona-Bayonas. PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 12 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry Data curation: Almudena Cotes Sanchı́s, Javier Gallego, Raquel Hernandez, Virginia Arra- zubi, Ana Custodio, Juana Marı́a Cano, Gema Aguado, Ismael Macias, Carlos Lopez, Flora López, Laura Visa, Marcelo Garrido, Nieves Martı́nez Lago, Ana Fernández Montes, Marı́a Luisa Limón, Aitor Azkárate, Paola Pimentel, Pablo Reguera, Avinash Ramchandani, Juan Diego Cacho, Alfonso Martı́n Carnicero, Mónica Granja, Marta Martı́n Richard, Carolina Hernández Pérez, Alicia Hurtado, Olbia Serra, Elvira Buxo, Rosario Vidal Tocino, Paula Jimenez-Fonseca, Alberto Carmona-Bayonas. Formal analysis: Alberto Carmona-Bayonas. Investigation: Javier Gallego, Paula Jimenez-Fonseca, Alberto Carmona-Bayonas. Methodology: Almudena Cotes Sanchı́s, Javier Gallego, Paula Jimenez-Fonseca, Alberto Car- mona-Bayonas. Project administration: Paula Jimenez-Fonseca. Supervision: Almudena Cotes Sanchı́s, Javier Gallego, Paula Jimenez-Fonseca, Alberto Car- mona-Bayonas. Validation: Javier Gallego, Paula Jimenez-Fonseca, Alberto Carmona-Bayonas. Writing – original draft: Javier Gallego, Paula Jimenez-Fonseca, Alberto Carmona-Bayonas. Writing – review & editing: Almudena Cotes Sanchı́s, Javier Gallego, Paula Jimenez-Fonseca, Alberto Carmona-Bayonas. References 1. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, et al. Cancer incidence and mortal- ity worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int. J. cancer. 2015; 136: E359–86. https://doi.org/10.1002/ijc.29210 PMID: 25220842 2. Wagner AD, Syn NLX, Moehler M, Grothe W, Yong WP, Tai B, et al. Chemotherapy for advanced gas- tric cancer. Cochrane Database Syst Rev. 2017; 8:CD004064. https://doi.org/10.1002/14651858. CD004064.pub4 PMID: 28850174 3. Smyth EC, Verheij M, Allum W, Cunningham D, Cervantes A, Arnold D. Gastric cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann. Oncol. 2016; 27:v38–49. https://doi. org/10.1093/annonc/mdw350 PMID: 27664260 4. Bang YJ, Van Cutsem E, Feyereislova A, Chung HC, Shen L, Sawaki A, et al. Trastuzumab in combina- tion with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): A phase 3, open-label, randomised controlled trial. Lancet. 2010; 376:687–97. https://doi.org/10.1016/S0140-6736(10)61121-X PMID: 20728210 5. Yamada Y, Boku N, Mizusawa J, Iwasa S, Kadowaki S, Nakayama N, et al. Docetaxel Plus Cisplatin and S-1 versus Cisplatin and S-1 in Patients with Advanced Gastric Cancer (JCOG1013): An Open- Label, Randomised, Phase 3 Trial. Lancet Gastroenterol Hepatol. 2019; 4:501. https://doi.org/10.1016/ S2468-1253(19)30083-4 PMID: 31101534 6. Thuss-Patience PC, Kretzschmar A, Bichev D, Deist T, Hinke A, Breithaupt K, et al. Survival advantage for irinotecan versus best supportive care as second-line chemotherapy in gastric cancer–a randomised phase III study of the Arbeitsgemeinschaft Internistische Onkologie (AIO). Eur. J. Cancer. 2011; 47:2306–14. https://doi.org/10.1016/j.ejca.2011.06.002 PMID: 21742485 7. Kang JH, Lee S Il, Lim DH, Park KW, Oh SY, Kwon HC. Salvage chemotherapy for pretreated gastric cancer: a randomized phase III trial comparing chemotherapy plus best supportive care with best sup- portive care alone. J. Clin. Oncol. 2012; 30:1513–8. https://doi.org/10.1200/JCO.2011.39.4585 PMID: 22412140 8. Lee K, Maeng CH, Kim T, Zang DY, Kim YH, Hwang IG, et al. A Phase III Study to Compare the Efficacy and Safety of Paclitaxel Versus Irinotecan in Patients with Metastatic or Recurrent Gastric Cancer Who Failed in First-line Therapy (KCSG ST10-01). Oncologist. 2019; 24:18–e24. https://doi.org/10.1634/ theoncologist.2018-0142 PMID: 30126861 9. Hironaka S, Ueda S, Yasui H, Nishina T, Tsuda M, Tsumura T, et al. Randomized, open-label, phase III study comparing irinotecan with paclitaxel in patients with advanced gastric cancer without severe PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 13 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry peritoneal metastasis after failure of prior combination chemotherapy using fluoropyrimidine plus plati- num: WJOG 4007 tria. J Clin Oncol. 2013; 31:4438–44. https://doi.org/10.1200/JCO.2012.48.5805 PMID: 24190112 10. Ford HER, Marshall A, Bridgewater JA, Janowitz T, Coxon FY, Wadsley J, et al. Docetaxel versus active symptom control for refractory oesophagogastric adenocarcinoma (COUGAR-02): an open- label, phase 3 randomised controlled trial. Lancet Oncol. 2014; 15:78–86. https://doi.org/10.1016/ S1470-2045(13)70549-7 PMID: 24332238 11. Wilke H, Muro K, Van Cutsem E, Oh S-C, Bodoky G, Shimada Y, et al. Ramucirumab plus paclitaxel versus placebo plus paclitaxel in patients with previously treated advanced gastric or gastro-oesopha- geal junction adenocarcinoma (RAINBOW): a double-blind, randomised phase 3 trial. Lancet Oncol. 2014; 15:1224–35. https://doi.org/10.1016/S1470-2045(14)70420-6 PMID: 25240821 12. Fuchs CS, Tomasek J, Yong CJ, Dumitru F, Passalacqua R, Goswami C, et al. Ramucirumab mono- therapy for previously treated advanced gastric or gastro-oesophageal junction adenocarcinoma (REGARD): an international, randomised, multicentre, placebo-controlled, phase 3 trial. Lancet. Else- vier; 2014; 383:31–9. https://doi.org/10.1016/S0140-6736(13)61719-5 PMID: 24094768 13. Di Bartolomeo M, Niger M, Tirino G, Petrillo A, Berenato R, Laterza MM, et al. Ramucirumab as sec- ond-line therapy in metastatic gastric Cancer: Real-World data from the RAMoss study. Target. Oncol. 2018; 13:227–34. https://doi.org/10.1007/s11523-018-0562-5 PMID: 29582224 14. Paulson AS, Hess LM, Liepa AM, Cui ZL, Aguilar KM, Clark J, et al. Ramucirumab for the treatment of patients with gastric or gastroesophageal junction cancer in community oncology practices. Gastric Cancer. Springer; 2018; 21:831–44. https://doi.org/10.1007/s10120-018-0796-z PMID: 29397460 15. Jung M, Ryu M-H, Kang M, Zang DY, Hwang IG, Lee K-W, et al. Efficacy and tolerability of ramuciru- mab monotherapy or in combination with paclitaxel in gastric cancer patients from the Expanded Access Program Cohort by the Korean Cancer Study Group (KCSG). Gastric Cancer. 2018; 21:819– 30. https://doi.org/10.1007/s10120-018-0806-1 PMID: 29427038 16. Muro K, Van Cutsem E, Narita Y, Pentheroudakis G, Baba E, Li J, et al. Pan-Asian adapted ESMO Clin- ical Practice Guidelines for the management of patients with metastatic gastric cancer: a JSMO–ESMO initiative endorsed by CSCO, KSMO, MOS, SSO and TOS. Ann. Oncol. 2018; 30:19–33. 17. Choi IS, Kim JH, Lee JH, Suh KJ, Lee JY, Kim J-W, et al. A population-based outcomes study of patients with metastatic gastric cancer receiving second-line chemotherapy: A nationwide health insur- ance database study. PLoS One. 2018; 13:e0205853. https://doi.org/10.1371/journal.pone.0205853 PMID: 30346970 18. Kojima T, Muro K, Francois E, Hsu C-H, Moriwaki T, Kim S-B, et al. Pembrolizumab versus chemother- apy as second-line therapy for advanced esophageal cancer: Phase III KEYNOTE-181 study. 2019 ASCO Annu. Meet. Chicago; 2019. 19. Shitara K, Özgüroğlu M, Bang Y-J, Di Bartolomeo M, Mandalà M, Ryu M-H, et al. Pembrolizumab ver- sus paclitaxel for previously treated, advanced gastric or gastro-oesophageal junction cancer (KEY- NOTE-061): a randomised, open-label, controlled, phase 3 trial. Lancet. 2018; 392:123–33. https://doi. org/10.1016/S0140-6736(18)31257-1 PMID: 29880231 20. Thuss-Patience PC, Shah MA, Ohtsu A, Van Cutsem E, Ajani JA, Castro H, et al. Trastuzumab emtan- sine versus taxane use for previously treated HER2-positive locally advanced or metastatic gastric or gastro-oesophageal junction adenocarcinoma (GATSBY): an international randomised, open-label, adaptive, phase 2/3 study. Lancet Oncol. 2017; 18:640–53. https://doi.org/10.1016/S1470-2045(17) 30111-0 PMID: 28343975 21. Satoh T, Xu R-H, Chung HC, Sun G-P, Doi T, Xu J-M, et al. Lapatinib plus paclitaxel versus paclitaxel alone in the second-line treatment of HER2-amplified advanced gastric cancer in Asian populations: TyTAN—a randomized, phase III study. J. Clin. Oncol. 2014; 32:2039–49. https://doi.org/10.1200/JCO. 2013.53.6136 PMID: 24868024 22. Okines AFC, Asghar U, Cunningham D, Ashley S, Ashton J, Jackson K, et al. Rechallenge with plati- num plus fluoropyrimidine+/–epirubicin in patients with oesophagogastric cancer. Oncology. 2010; 79:150–8. https://doi.org/10.1159/000322114 PMID: 21150230 23. Shitara K, Doi T, Dvorkin M, Mansoor W, Arkenau H-T, Prokharau A, et al. Trifluridine/tipiracil versus placebo in patients with heavily pretreated metastatic gastric cancer (TAGS): a randomised, double- blind, placebo-controlled, phase 3 trial. Lancet Oncol. 2018; 19:1437–48. https://doi.org/10.1016/ S1470-2045(18)30739-3 PMID: 30355453 24. Kang Y-K, Boku N, Satoh T, Ryu M-H, Chao Y, Kato K, et al. Nivolumab in patients with advanced gas- tric or gastro-oesophageal junction cancer refractory to, or intolerant of, at least two previous chemo- therapy regimens (ONO-4538-12, ATTRACTION-2): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2017; 390:2461–71. https://doi.org/10.1016/S0140-6736(17)31827-5 PMID: 28993052 PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 14 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry 25. Li J, Qin S, Xu J, Xiong J, Wu C, Bai Y, et al. Randomized, double-blind, placebo-controlled phase III trial of apatinib in patients with chemotherapy-refractory advanced or metastatic adenocarcinoma of the stomach or gastroesophageal junction. J. Clin. Oncol. 2016; 34:1448–54. https://doi.org/10.1200/JCO. 2015.63.5995 PMID: 26884585 26. Van Cutsem E, Moiseyenko VM, Tjulandin S, Majlis A, Constenla M, Boni C, et al. Phase III study of docetaxel and cisplatin plus fluorouracil compared with cisplatin and fluorouracil as first-line therapy for advanced gastric cancer: a report of the V325 Study Group. J. Clin. Oncol. 2006; 24:4991–7. https://doi. org/10.1200/JCO.2006.06.8429 PMID: 17075117 27. Yoshikawa T, Muro K, Shitara K, Oh D-Y, Kang Y-K, Chung HC, et al. Effect of first-line S-1 plus oxali- platin with or without ramucirumab followed by paclitaxel plus ramucirumab on advanced gastric cancer in East Asia: The phase 2 RAINSTORM randomized clinical trial. JAMA Netw. open. 2019; 2:e198243– e198243. https://doi.org/10.1001/jamanetworkopen.2019.8243 PMID: 31373648 28. Al-Batran S-E, Schuler MH, Zvirbule Z, Manikhas G, Lordick F, Rusyn A, et al. FAST: An international, multicenter, randomized, phase II trial of epirubicin, oxaliplatin, and capecitabine (EOX) with or without IMAB362, a first-in-class anti-CLDN18. 2 antibody, as first-line therapy in patients with advanced CLDN18. 2+ gastric and ga. 2016 ASCO Annu. Meet. Chicago; 29. Shitara K, Ikeda J, Yokota T, Takahari D, Ura T, Muro K, et al. Progression-free survival and time to pro- gression as surrogate markers of overall survival in patients with advanced gastric cancer: analysis of 36 randomized trials. Invest. New Drugs. 2012; 30:1224–31. https://doi.org/10.1007/s10637-011-9648- y PMID: 21350803 30. Liu L, Yu H, Huang L, Shao F, Bai J, Lou D, et al. Progression-free survival as a surrogate endpoint for overall survival in patients with third-line or later-line chemotherapy for advanced gastric cancer. Onco. Targets. Ther. 2015; 8:921. https://doi.org/10.2147/OTT.S82365 PMID: 25960663 31. Carmona-Bayonas A, Jiménez-Fonseca P, Custodio A, Sánchez Cánovas M, Hernández R, Pericay C, et al. Anthracycline-based triplets do not improve the efficacy of platinum-fluoropyrimidine doublets in first-line treatment of advanced gastric cancer: real-world data from the AGAMENON National Cancer Registry. Gastric Cancer. 2017; 14:1379–88. 32. Carmona-Bayonas A, Jiménez-Fonseca P, Lorenzo MLS, Ramchandani A, Martı́nez EA, Custodio A, et al. On the Effect of Triplet or Doublet Chemotherapy in Advanced Gastric Cancer: Results From a National Cancer Registry. J. Natl. Compr. Cancer Netw. 2016; 14:1379–88. 33. Jiménez-Fonseca P, Carmona-Bayonas A, Lorenzo MLS, Plazas JG, Custodio A, Hernández R, et al. Prognostic significance of performing universal HER2 testing in cases of advanced gastric cancer. Gas- tric Cancer. 2016; 20:465–74. https://doi.org/10.1007/s10120-016-0639-8 PMID: 27599830 34. Custodio A, Carmona-Bayonas A, Fonseca PJ, Sánchez ML, Antonio Viudez R, Hernández JM, et al. Nomogram-based prediction of survival in patients with advanced oesophagogastric adenocarcinoma receiving first-line chemotherapy: a multicenter prospective study in the era of trastuzumab. Br. J. Can- cer. 2017; 116:1526–35. https://doi.org/10.1038/bjc.2017.122 PMID: 28463962 35. Jiménez-Fonseca P, Carmona-Bayonas A, Hernández R, Custodio A, CANO JM, Lacalle A, et al. Lau- ren subtypes of advanced gastric cancer influence survival and response to chemotherapy: Real-World Data from the AGAMENON National Cancer Registry. Br. J. Cancer. 2017; 117:775–82. https://doi.org/ 10.1038/bjc.2017.245 PMID: 28765618 36. Visa L, Jiménez-Fonseca P, Martı́nez EA, Hernández R, Custodio A, Garrido M, et al. Efficacy and safety of chemotherapy in older versus non-older patients with advanced gastric cancer: A real-world data, non-inferiority analysis. J. Geriatr. Oncol. 2017; 9:254–64. https://doi.org/10.1016/j.jgo.2017.11. 008 PMID: 29242093 37. Carmona-Bayonas A, Jiménez-Fonseca P, Echavarria I, Cánovas MS, Aguado G, Gallego J, et al. Sur- gery for metastases for esophageal-gastric cancer in the real world: data from the AGAMENON national registry. Eur. J. Surg. Oncol. 2018; 44:1191–8. https://doi.org/10.1016/j.ejso.2018.03.019 PMID: 29685755 38. Viúdez A, Carmona-Bayonas A, Gallego J, Lacalle A, Hernández R, Cano JM, et al. Optimal duration of first-line chemotherapy for advanced gastric cancer: data from the AGAMENON registry. Clin. Transl. Oncol. 2019;1–17. 39. Carmona-Bayonas A, Jiménez-Fonseca Paula et al. Multistate Models: Accurate and Dynamic Meth- ods to Improve Predictions of Thrombotic Risk in Patients with Cancer. Thromb. Haemost. 2019; 119:1849–59. https://doi.org/10.1055/s-0039-1694012 PMID: 31461750 40. NCCN. NCCN clinical practice guidelines for gastric cancer [Internet]. 2019 [cited 2019 Mar 11]. Avail- able from: https://www.nccn.org/professionals/physician_gls/pdf/gastric.pdf 41. Harrell F. Regression modeling strategies: with applications to linear models, logistic and ordinal regres- sion, and survival analysis. 2nd ed. New York: Springer; 2015. PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 15 / 16 PLOS ONE Second-line treatment in advanced gastric cancer: Data from the Spanish AGAMENON registry 42. Harrell Jr F, Frank E, Maintaner Frank E. Package ‘rms’. [Internet]. 2015 [cited 2020 Jan 1]. p. 229. Available from: http://cran.r-project.org/web/packages/rms/index.html 43. Emura T, Lin C-W, Wang W. A goodness-of-fit test for Archimedean copula models in the presence of right censoring. Comput. Stat. Data Anal. 2010; 54:3033–43. 44. Janowitz T, Thuss-Patience P, Marshall A, Kang JH, Connell C, Cook N, et al. Chemotherapy vs sup- portive care alone for relapsed gastric, gastroesophageal junction, and oesophageal adenocarcinoma: a meta-analysis of patient-level data. Br. J. Cancer. 2016; 114:381. https://doi.org/10.1038/bjc.2015. 452 PMID: 26882063 45. Zhang Y, Ma B, Huang X-T, Li Y-S, Wang Y, Liu Z-L. Doublet versus single agent as second-line treat- ment for advanced gastric cancer: a meta-analysis of 10 randomized controlled trials. Medicine (Balti- more). 2016; 95:e2792. 46. Zhu X, Ko Y-J, Berry S, Shah K, Lee E, Chan K. A Bayesian network meta-analysis on second-line sys- temic therapy in advanced gastric cancer. Gastric Cancer. 2017; 20:646–54. https://doi.org/10.1007/ s10120-016-0656-7 PMID: 27722826 47. Fanotto V, Cordio S, Pasquini G, Fontanella C, Rimassa L, Leone F, et al. Prognostic factors in 868 advanced gastric cancer patients treated with second-line chemotherapy in the real world. Gastric Can- cer. 2017; 20:825–33. https://doi.org/10.1007/s10120-016-0681-6 PMID: 28028664 48. Park SR, Kim M-J, Nam B-H, Kim CG, Lee JY, Cho S-J, et al. A randomised phase II study of continu- ous versus stop-and-go S-1 plus oxaliplatin following disease stabilisation in first-line chemotherapy in patients with metastatic gastric cancer. Eur. J. Cancer. 2017; 83:32–42. https://doi.org/10.1016/j.ejca. 2017.06.008 PMID: 28711577 49. Nevala-Plagemann C, Moser J, Gilcrease GW, Garrido-Laguna I. Survival of patients with metastatic HER2 positive gastro-oesophageal cancer treated with second-line chemotherapy plus trastuzumab or ramucirumab after progression on front-line chemotherapy plus trastuzumab. ESMO open. 2019; 4: e000539. https://doi.org/10.1136/esmoopen-2019-000539 PMID: 31423338 50. Li Q, Jiang H, Li H, Xu R, Shen L, Yu Y, et al. Efficacy of trastuzumab beyond progression in HER2 posi- tive advanced gastric cancer: a multicenter prospective observational cohort study. Oncotarget. 2016; 7:50656. https://doi.org/10.18632/oncotarget.10456 PMID: 27409420 51. De Vita F, Borg C, Farina G, Geva R, Carton I, Cuku H, et al. Ramucirumab and paclitaxel in patients with gastric cancer and prior trastuzumab: subgroup analysis from RAINBOW study. Futur. Oncol. 2019; 15:2723–31. 52. Begg CB. Justifying the Choice of Endpoints for Clinical Trials. J. Natl. Cancer Inst. 2013; 105:1594–5. https://doi.org/10.1093/jnci/djt289 PMID: 24108810 53. Paoletti X, Oba K, Bang Y-J, Bleiberg H, Boku N, Bouché O, et al. Progression-free survival as a surro- gate for overall survival in advanced/recurrent gastric cancer trials: a meta-analysis. J. Natl. Cancer Inst. 2013; 105:1667–70. https://doi.org/10.1093/jnci/djt269 PMID: 24108811 PLOS ONE | https://doi.org/10.1371/journal.pone.0235848 July 31, 2020 16 / 16