Journal of Clinical Medicine

Article Impact of New Chemotherapy Regimens on the Treatment Landscape and Survival of Locally Advanced and Metastatic Patients

Markus Kieler 1 , Matthias Unseld 1, Daniela Bianconi 1, Martin Schindl 2 , Gabriela V. Kornek 1, Werner Scheithauer 1 and Gerald W. Prager 1,* 1 Department of Medicine I, Division of Oncology, Comprehensive Cancer Center Vienna, Medical University of Vienna, Währinger Gürtel 18-20, 1090 Vienna, Austria; [email protected] (M.K.); [email protected] (M.U.); [email protected] (D.B.); [email protected] (G.V.K.); [email protected] (W.S.) 2 Department of Surgery, Division of General Surgery, Pancreatic Cancer Unit, Comprehensive Cancer Center Vienna, Medical University of Vienna, Währinger Gürtel 18-20, 1090 Vienna, Austria; [email protected] * Correspondence: [email protected]; Tel.: +43-1-40400-44500

 Received: 19 February 2020; Accepted: 26 February 2020; Published: 28 February 2020 

Abstract: Background: New chemotherapy regimens for the treatment of metastatic pancreatic cancer have changed the therapy paradigm. We aimed to assess their impact on the treatment landscape and clinical outcome at our academic institution. Methods: In this single institutional posthoc registry analysis, we assessed characteristics and survival rates from all patients with locally advanced and metastatic pancreatic cancer who started a systemic treatment between 01/2011 and 12/2017. Survival analyses were performed by Kaplan-Meier and Cox proportional hazards model. Results: A total of 301 patients started a systemic treatment in the observation period. In the first line treatment, we observed a shift from the four different main regimens (/nab-paclitaxel, modified FOLFIRINOX, gemcitabine/ +/ erlotinib or gemcitabine alone) to gemcitabine/nab-paclitaxel and modified − FOLFIRINOX that add up to more than 80% of administered first line treatments in each of the time cohorts (2011–2013 vs. 2014–2017). The rate for first line modified FOLFIRINOX treatment was balanced between the two groups (19% and 15%). Median overall survival differed significantly between the two time cohorts (8.89 versus 11.9 months, p = 0.035). Survival rates for different first to second line treatment sequences (modified FOLFIRINOX to gemcitabine/nab-paclitaxel, gemcitabine/nab-paclitaxel to fluoropyrimidines plus nanoliposomal , or gemcitabine/nab-paclitaxel to fluoropyrimidines plus oxaliplatin) were not significantly different and median overall survival ranged from 14.27 to 15.64 months. Conclusion: Our study provides real-world evidence for the effectiveness of the new chemotherapy regimens and underscores the importance of the choice of the front-line regimen when considering different sequencing strategies.

Keywords: pancreatic ductal adenocarcinoma; treatment landscape; treatment patterns; treatment sequences; new chemotherapy regimens; nanoliposomal irinotecan; nab-paclitaxel; FOLFIRINOX

1. Introduction Pancreatic ductal adenocarcinoma (PAC) is projected to become the second leading cause of cancer deaths in the United States, as well as Europe in the next decade. Unlike most other members of the top five cancers with the highest number of cancer deaths, such as colorectal, breast, lung and prostate cancer, its incidence is increasing [1,2]. One reason is that the majority of patients is diagnosed at a metastatic or locally advanced disease stage, due to lack of specific symptoms in early stages and yet

J. Clin. Med. 2020, 9, 648; doi:10.3390/jcm9030648 www.mdpi.com/journal/jcm J. Clin. Med. 2020, 9, 648 2 of 15 no effective screening program [3]. In addition, high relapse rates after resection with curative intent are observed. Thus, the five-year survival rate of pancreatic cancer ranges between~5 to 10% with no raising numbers within the last 20 years [4]. Moreover, aggressive tumor biology and unresponsiveness to most of the currently clinically tested molecularly targeted therapies, as well as immunotherapies, which have revolutionized anti-cancer treatment in many other cancers, contribute to this devastating prognosis [5,6]. Systemic chemotherapy still remains the standard of care for patients with metastatic PAC. In the past, fluorouracil (5-FU) has been the most commonly used until in 1996, gemcitabine (Gem) was approved as it was shown to modestly improve survival and relieve disease-related symptoms [7]. After the introduction of Gem, many different Gem-based combinations have been clinically tested with the majority of trials not demonstrating a meaningful clinical benefit. However, two meta-analyses found that combinations with platins or capecitabine improved the survival of patients with unresectable PAC and good performance status [8,9]. Of note, in 2006, a prospective randomized clinical phase III trial demonstrated that the addition of erlotinib to Gem marginally extended the lives of patients affected from inoperable PAC by only two weeks. So far, this is the only targeted therapy approved for the treatment of advanced PAC [10]. More recently in 2011, the PRODIGE4/ACCORD11 (Partenarait de Recherche en Oncologie Digestive/Actions Concertées dans les Cancers Colorectaux et Digestifs) study evaluated the combination of 5-FU, leucovorin (LV), irinotecan, and oxaliplatin (FOLFIRINOX) versus Gem monotherapy and reported a median overall survival (mOS) of 11.1 months against 6.8 months in the control arm (p < 0.001) [11]. Two years later, the MPACT trials also demonstrated superior survival with the combination of Gem plus nab-paclitaxel (Gem/NabP) over Gem alone with an mOS of 8.5 versus 6.7 months (p < 0.0001) [12]. The latest approved therapy for the treatment of unresectable PAC is nanoliposomal irinotecan with 5-FU/LV (FP/Nal-IRI) according to the positive results of the NAPOLI-1 trial, which evaluated this regimen in patients that progressed under a Gem based therapy [13]. Another clinically tested second line treatment option after front-line therapy with Gem is fluoropyrimidines in combination with oxaliplatin, although there is conflicting evidence concerning the efficacy of the different regimens [14,15]. With these recent developments in the clinical management of patients with advanced PAC, our aim was to determine the impact of new therapeutic options on the treatment landscape and clinical outcome in a cohort of patients with advanced or metastatic PAC, who have started their treatment at the Comprehensive Cancer Center of the Medical University of Vienna during 2011 and 2017.

2. Experimental Section

2.1. Study Design This is a single-center, retrospective, observational registry study, including patients with histologically or cytologically proven non-resectable PAC which was either locally advanced or metastasized and who have started a systemic treatment at the Medical University of Vienna between 01/2011 and 12/2017. The decision for the systemic treatment was always at the discretion of the treating physician. Furthermore, all patients in this study were discussed within multidisciplinary board sessions. If a patient was progressing to frontline chemotherapy and maintained sufficient performance status to undergo a second or later line chemotherapy, potential treatment options, which were available at that time, were discussed with the patient during the informed consent discussion. The decision for second line chemotherapy was not influenced by industry-funded trials, as there were no ongoing second line chemotherapy trials during the whole observation period. The selection of the optimal treatment options was based on the at that time available national and international treatment guidelines. For the comparison of the main study cohorts, the time of first administration of systemic chemotherapy was retrieved. The electronic medical history was queried for patient demographics, performance status, date of diagnosis, date of advanced disease, diagnosis and carbohydrate antigen 19-9 (CA19-9) level at baseline, treatment details and overall survival. ECOG (Eastern Cooperative J. Clin. Med. 2020, 9, 648 3 of 15

Oncology Group) performance status was derived, if not stated explicitly, from the medical history, including comorbidities and overall assessment of the treating physician. Recurrent PAC after resection of curative intent was stated as stage IV disease. Date of disease progression on treatment and date of death were recorded. The here presented data analysis received prior approval by the ethical committee of the Medical University of Vienna (EK.No. 1806/2017) and was performed according to Helsinki criteria of good scientific practice.

2.2. Statistics Descriptive statistics were calculated as mean, median or percentages as appropriate. Statistical comparison of categorical variables was calculated with Fisher’s exact test and comparison of metric variables with the unpaired t test. Overall survival (OS) from first, second or third treatment line was calculated from the time of the first administration of the respective treatment line to death. OS was depicted by Kaplan Meier plots. For group comparisons, the Log-Rank test (LRt) and the Gehan-Breslow-Wilcoxon test (GBWt) (in the case of non-constant hazard ratios) were used, respectively. Multivariate survival analysis was performed with the Cox proportional hazard model to evaluate the treatment effect with adjustment for stratification factors. In the case of missing variables, the entire data of that individual was deleted from the multivariate analysis. For the covariates ECOG performance status and number of metastatic sites, categories were combined because of the low number of events. A p-value of <0.05 was considered as statistically significant, and the term “significant” was used to indicate that the computed p-values were below this cut-off value. The assessment of respectively constant and non-constant hazard ratios was based on a graphical approach. No formal sample size calculation has been performed because of the retrospective character of the study. Statistical analysis of Kaplan-Meier plots, Log-Rank test and the Gehan-Breslow-Wilcoxon test was performed with GraphPad Prism Software Prism 7 for Windows, Version 7.03, February 20, 2017 (GraphPad Software, La Jolla, CA, USA). Multivariate survival analysis (Cox proportional hazard model) was performed with R for Windows, Version 3.6.2, as well as RStudio for Windows, Version 1.2.5033.

3. Results

3.1. Patient Characteristics Between 2011 and 2017, a total of 301 patients started a systemic treatment, due to recently diagnosed metastatic or unresectable locally advanced PAC. The total cohort was split up into two nearly equally sized cohorts, cohort A (2011–2013, n = 132) and cohort B (2014–2017, n = 169), according to the time point of initiating systemic chemotherapy for this disease. Table1 lists patient and tumor characteristics of the two main cohorts.

Table 1. Patient Characteristics of patients and tumors of the two main study cohorts.

Cohort A Cohort B p-Value 2011–2013 (n = 132) 2014–2017 (n = 169) Median age at diagnosis of advanced 66.3 (57.3–72.2) 66.0 (58.3–73.3) ns disease (median, range) Gender (%) Female 53 (41) 90 (53) * Male 79 (59) 79 (47) Disease stage (%) Locally advanced (unresectable) 34 (26) 43 (25) ns Metastatic 98 (74) 126 (75) Prior surgical resection (%) 23 (17) 32 (19) ns J. Clin. Med. 2020, 9, 648 4 of 15

Table 1. Cont.

Cohort A Cohort B p-Value 2011–2013 (n = 132) 2014–2017 (n = 169) 505.35 686.40 Median CA 19-9 levels in kU/l (range) ns (68.05–2261.25) (88.25–3771.00) CA 19-9 levels (%) Within normal range 20 (15) 19 (11) ns Above normal range 101 (77) 132 (78) n/a 11 (8) 18 (11) ns Site of metastatic disease (%) Liver 73 (55) 84 (50) ns Lung 15 (11) 21 (12) ns Peritoneum 10 (8) 30 (18) * Other 11 (8) 7 (4) ns Number of metastatic sites (%) 0 34 (26) 43 (25) ns 1 80 (60) 102 (60) ns 2 14 (11) 18 (11) ns 3 4 (3) 6 (4) ns ≥ ECOG Performance Status (%) 0 67 (51) 100 (59) ns 1 52 (39) 61 (36) ns 2 8 (6) 5 (3) ns n/a 5 (4) 3 (2) ns Abbreviations: CA 19-9, Carbohydrate antigen 19-9; ECOG, Eastern Cooperative Oncology Group; ns, not significant (p > 0.05).

The median age in both groups was nearly identical (66.3 (interquartile range (IQR)) 57.3–72.2) versus 66.0 years (IQR 58.3–73.3)). In cohort A there were 53 (41%) female and 79 (59%) male patients, while in cohort B there were 90 female (53%) and 79 (47%) male patients. About one fourth in each cohort presented with locally advanced (unresectable) disease (26% and 25%), while the majority had metastatic disease (74% and 75%) at time of initiation with systemic treatment. In cohort A 23 (17%) and in cohort B 32 (19%) patients had a prior surgical resection. The median CA 19-9 level was 505.35 kU/l (68.05–2261.25 kU/l) in cohort A and 686.40 kU/l (88.25–3771.00 kU/l) in cohort B. While a minor proportion in each of the cohorts had CA 19-9 levels within the normal range (15% and 11%), most of the patients had increased CA 19-9 levels at the start of their systemic therapy (77% and 78%). In 8% and 11%, the levels for CA 19-9 were not available. In both cohorts the predominant site of metastatic disease was liver (55% and 50%), followed by lung (11% and 12%), peritoneum (8% and 18%) and other anatomical locations (8% and 4%). Relative numbers for metastatic sites for patients were equivalent. There were 60% with one metastatic site, as well as 11% with two in each cohort, while there were 3% and 4% of patients who had three different organs affected by metastatic spread. Most of the patients presented with good ECOG performance status (90% and 95% with ECOG 0-1) and 6%, as well as 3% had an ECOG performance status of two.

3.2. Treatment Landscape For the complete overview of administered therapies subdivided by the two cohorts, we refer to Figure1 and Table2. J. Clin. Med. 2020, 9, 648 5 of 15 J. Clin. Med. 2020, 9, 648 5 of 15

FigureFigure 1. 1.Treatment Treatment patterns. patterns. Treatment Treatment regimens regimens from firstfrom to thirdfirst lineto third of locally line advancedof locallyor advanced metastatic or patientsmetastatic who startedpatients a systemicwho started treatment a betweensystemic 2011–2013 treatment and 2014–2017.between 2011–2013 Abbreviations: and Gem2014–2017./NabP, gemcitabineAbbreviations:/nab-paclitaxel; Gem/NabP, FOLFIRINOX, gemcitabine/nab-pac 5-FU, leucovorinlitaxel; (LV),FOLFIRINOX, irinotecan, and5-FU, oxaliplatin; leucovorin Gem /(LV),Ox, gemcitabineirinotecan, /andoxaliplatin; oxaliplatin; FP/Ox, Gem/Ox, fluoropyrimidines gemcitabine/ox/oxaliplatin,aliplatin; FPFP/Ox,/Iri, fluoropyrimidines fluoropyrimidines/oxaliplatin,/irinotecan; FPFP/Iri,/Nal-IRI, fluoropyrimidines/irinotec fluoropyrimidines/nanoliposomalan; FP/Nal-IRI, irinotecan; fluoropyrimidines/nanoliposo Cap, capecitabine; CRTx,mal chemoradiation; irinotecan; Cap, MMC,capecitabine; mitomycin CRTx, C. chemoradiation; MMC, mitomycin C. Table 2. Administered chemotherapy regimens. Table 2. Administered chemotherapy regimens. First Line p Second Line p Third Line p First Line p Second Line p Third Line p Chemotherapy 2011–2013 2014–2017 2011–2013 2014–2017 2011–2013 2014–2017 Chemotherapy 2011–2013 2014–2017 2011–2013 2014–2017 2011–2013 2014–2017 116 Gem/NabP 34 (26%) 116 **** 18 (21%) 13 (13%) ns 10 (26%) 10 (17%) ns Gem/NabP 34 (26%) (69%) **** 18 (21%) 13 (13%) ns 10 (26%) 10 (17%) ns (69%) mFOLFIRINOXmFOLFIRINOX 25 25 (19%) (19%) 26 26 (15%) (15%) ns ns 4 (5%) 4 (5%) 5 (5%) 5 (5%) ns ns 2 (5%) 2 (5%) 4 (7%) 4 (7%) ns ns Gem/OxGem/Ox 16 16 (12%) (12%) 5 5(3%) (3%) ** ** 5 (6%) 5 (6%) 2 (2%) 2 (2%) ns ns 0 (0%) 0 (0%) 4 (7%) 4 (7%) ns ns Gem/Ox/Gem/Ox / 1616 (12%) (12%) 0 0(0%) (0%) **** **** 3 (3%) 3 (3%) 5 (5%) 5 (5%) nsns 1 (3%)1 (3%) 0 (0%)0 (0%) ns ns ErlotinibErlotinib GemGem alone alone 18 18 (14%) (14%) 14 14 (8%) (8%) ns ns 6 (7%) 6 (7%) 2 (2%) 2 (2%) ns ns 2 (5%) 2 (5%) 3 (5%) 3 (5%) ns ns Gem/Erlotinib 4 (3%) 0 (0%) * 9 (10%) 1 (1%) ** 2 (5%) 2 (3%) ns Gem/Erlotinib 4 (3%) 0 (0%) * 9 (10%) 1 (1%) ** 2 (5%) 2 (3%) ns FP/Ox 1 (1%) 2 (1%) ns 12 (14%) 19 (18%) ns 5 (13%) 16 (28%) ns FP/IriFP/Ox 1 1(1%) (1%) 2 0(1%) (0%) ns ns 12 (14%) 1 (1%) 19 (18%) 2 (2%) ns ns 5 (13%) 8 (21%) 16 (28%) 6 (10%) ns ns FP/Nal-IRIFP/Iri 0 1(0%) (1%) 0 1(0%) (1%) ns ns 1 (1%) 0 (0%) 2 (2%) 31 (30%) ns **** 8 (21%) 0 (0%) 6 (10%) 14 (24%) ns **** FPCap/Nal-IRI 0 0(0%) (0%) 1 1(1%) (1%) ns ns 0 (0%) 4 (5%) 31 (30%) 5 (5%) **** ns 0 (0%) 1 (3%) 14 (24%) 2 (3%) **** ns Gem/CapCap 0 0(0%) (0%) 1 1(1%) (1%) ns ns 4 (5%) 2 (2%) 5 (5%) 0 (0%) ns ns 1 (3%) 0 (0%) 2 (3%) 2 (3%) ns ns Ox 1 (1%) 0 (0%) ns 0 (0%) 0 (0%) ns 1 (3%) 0 (0%) ns Gem/Cap 0 (0%) 1 (1%) ns 2 (2%) 0 (0%) ns 0 (0%) 2 (3%) ns CRTx 15 (11%) 3 (2%) *** 16 (19%) 19 (19%) ns 2 (5%) 0 (0%) ns MMCOx 0 1(0%) (1%) 0 0(0%) (0%) ns ns 0 (0%) 1 (1%) 0 (0%) 0 (0%) ns ns 1 (3%) 1 (3%) 0 (0%) 0 (0%) ns ns Local therapyCRTx 151 (1%) (11%) 3 0(2%) (0%) *** ns 16 (19%) 2 (2%) 19 (19%) 0 (0%) ns ns 2 (5%) 2 (5%) 0 (0%) 0 (0%) ns ns Gem/Cap/ MMC0 0(0%) (0%) 0 0(0%) (0%) ns ns 1 (1%) 3 (3%) 0 (0%) 0 (0%) ns ns 1 (3%) 1 (3%) 0 (0%) 0 (0%) ns ns Erlotinib Local therapy 1 (1%) 0 (0%) ns 2 (2%) 0 (0%) ns 2 (5%) 0 (0%) ns No therapy 0 (0%) 0 (0%) ns 46 (35%) 65 (38%) ns 94 (71%) 106 (63%) ns Gem/Cap/ 0 (0%) 0 (0%) ns 3 (3%) 0 (0%) ns 1 (3%) 0 (0%) ns ErlotinibAbbreviations: Gem/NabP, gemcitabine/nab-paclitaxael; mFOLFIRINOX, modified 5-FU, leucovorin, Noirinotecan, therapy and 0 (0%)oxaliplatin; 0 (0%)Gem/Ox, nsgemcitabine/ox 46 (35%)aliplatin; 65 (38%) FP/Ox, ns fluoropyrimidines/oxaliplatin, 94 (71%) 106 (63%) ns FP/Iri, fluoropyrimidines/irinotecan; FP/Nal-IRI, fluoropyrimidines/nanoliposomal irinotecan; Cap, Abbreviations: Gem/NabP, gemcitabine/nab-paclitaxael; mFOLFIRINOX, modified 5-FU, leucovorin, irinotecan, andcapecitabine; oxaliplatin; Gem CRTx,/Ox, chemoradiation; gemcitabine/oxaliplatin; MMC, FP /mitomycinOx, fluoropyrimidines C, p, p-value;/oxaliplatin, ns, not FP /significantIri, fluoropyrimidines (p > 0.05)./ irinotecan;Asterisks FP indicate/Nal-IRI, p fluoropyrimidines-value, * 0.01 to 0.05,/nanoliposomal ** 0.001 to irinotecan;0.01, *** 0.0001 Cap, capecitabine;to 0.001, ****< CRTx, 0.0001 chemoradiation;. MMC, mitomycin C, p, p-value; ns, not significant (p > 0.05). Asterisks indicate p-value, * 0.01 to 0.05, ** 0.001 to 0.01, *** 0.0001 to 0.001, **** < 0.0001.

J. Clin. Med. 2020, 9, 648 6 of 15 J. Clin. Med. 2020, 9, 648 6 of 15

InIn first first line line systemic systemic therapy, therapy, Gem /Gem/NabPNabP ranks inrank boths in cohorts both oncohorts top, however, on top, relativehowever, frequencies relative difrequenciesffer substantially differ (26%substantially in cohort (26% A versus in cohort 69% inA cohortversus B). 69% Modified in cohort FOLFIRINOX B). Modified (mFOLFIRINOX, FOLFIRINOX no(mFOLFIRINOX, bolus 5-FU) was no the bolus second 5-FU) commonly was the second used regimen commonly in the used two regimen cohorts in (19% the vs.two 15%) cohorts followed (19% byvs. Gem 15%) alonefollowed (14% by versus Gem alone 8%). (14% In cohort versus A, 8%). further In cohort commonly A, further used commonly regimens used were regimens gemcitabine were andgemcitabine oxaliplatin and (Gem oxaliplatin/Ox) (12%), (Gem/Ox) oxaliplatin (12%), plus oxalip erlotiniblatin (Gem plus/ Oxerlotinib/Erlotinib) (Gem/Ox/Erlotinib) (12%), chemoradiation (12%), (CRTx)chemoradiation (11%) and (CRTx) other therapies(11%) and (6%, other see therapies Table2), (6%, whereas, see Table in cohort 2), whereas, B further in treatmentscohort B further were subsumedtreatments to were other subsumed therapies, to due other to theirtherapies, low frequencies due to their (8%, low seefrequencies Table2). (8%, see Table 2). TheThe proportion proportion of of patients patients which which startedstarted aa secondsecond lineline therapytherapy waswas 65%65% ((nn == 86) in cohort A and 62%62% (n(n= = 104)104) inin cohortcohort BB (see(see FigureFigure1 ).1). In In descending descending orderorder of of relative relative frequencies, frequencies, administeredadministered therapiestherapies in in cohort cohort AA werewere GemGem/NabP/NabP (21%),(21%), CRTx (19%), fluoropyrimidines fluoropyrimidines and oxaliplatin oxaliplatin (FP/Ox) (FP/Ox) (14%),(14%), gemcitabine gemcitabine/erlotinib/erlotinib (Gem(Gem/Erlotinib)/Erlotinib) (10%),(10%), GemGem alonealone (7%),(7%), GemGem/Ox/Ox (6%), mFOLFIRINOX (5%),(5%), capecitabinecapecitabine (Cap)(Cap) (5%)(5%) andand otherother therapiestherapies (14%).(14%). For cohortcohort B thesethese werewere nanoliposomalnanoliposomal irinotecanirinotecan andand fluoropyrimidinesfluoropyrimidines (FP(FP/Nal-IRI)/Nal-IRI) (30%), CRTx (18%), FPFP/Ox/Ox (18%), Gem/NabP Gem/NabP (13%), mFOLFIRINOXmFOLFIRINOX (5%),(5%), GemGem/Ox/Erlotinib/Ox/Erlotinib (5%),(5%), CapCap (5%)(5%) andand otherother therapiestherapies (7%).(7%). RatesRates forfor administeredadministered thirdthird lineline therapiestherapies werewere 29%29% ((nn == 38) in cohortcohort A andand 37%37% ((n == 63) in cohortcohort B. B. In In cohort cohort A theseA these therapies therapies were were Gem /Gem/NabPNabP (26%), (26%), fluoropyrimidines fluoropyrimidines and irinotecan and irinotecan (FP/Iri) (21%),(FP/Iri) FP (21%),/Ox (13%), FP/Ox Gem (13%),/Erlotinib Gem/Erlotinib (5%), mFOLFIRINOX (5%), mFOLFIRINOX (5%), Gem (5%), alone Gem (5%), alone CRTx (5%), (5%), CRTx local (5%), to thelocal site to of the the site tumor of the directed tumor directed therapies therapies (5%) and (5 other%) and therapies other therapies (13%). In (13% cohort). In B cohort third lineB third therapy line consistedtherapy consisted of FP/Ox of (25%), FP/Ox FP (25%),/Nal-IRI FP/Nal-IRI (22%), Gem(22%/),NabP Gem/NabP (16%), (16%), FP/Iri (10%),FP/Iri (10%), mFOLFIRINOX mFOLFIRINOX (6%), Gem(6%),/Ox Gem/Ox (6%), Gem (6%), alone Gem (5%)alone and (5%) other and therapiesother therapies (10%). (10%).

3.3.3.3. ClinicalClinical OutcomeOutcome Kaplan-MeierKaplan-Meier curvescurves forfor thethe twotwo cohortscohorts areare depicteddepicted inin FigureFigure2 2..

FigureFigure 2. 2.Survival Survival accordingaccording toto treatmenttreatment lines.lines. Kaplan-Meier curves of patientspatients from initiation of the firstfirst ( A(A),), second second ( B(B)) and and thirdthird ((CC)) lineline treatmenttreatment accordingaccording toto thethe timetime periodsperiods when the firstfirst systemic treatmenttreatment was was administered. administered.

InIn cohort cohort A, A, 132 132 events events occurredoccurred (100%)(100%) andand inin cohortcohort BB 157157 (93%).(93%). The mOS from the beginning ofof first first line line treatment treatment waswas 8.898.89 monthsmonths inin cohortcohort AA andand statistically significantlysignificantly longer in cohort B by 3.013.01 months months (mOS (mOS 11.9 11.9 months; months;p p= =0.035 0.035 (LRt), (LRt), hazard hazard ratio ratio (HR) (HR) 0.77, 0.77, confidence confidence interval interval (CI) (CI) 0.61 0.61 to 0.98).to 0.98). The The proportions proportions of patientsof patients who who were were alive alive at 12, at 1812, and 18 and 24 months 24 months after after initiation initiation of first of first line therapyline therapy were 37.1%were 37.1% versus versus 49.1%, 49.1%, 16.7% versus16.7% 28.7%versus and 28.7% 8.3% and versus 8.3% 17.3%. versus Multivariate 17.3%. Multivariate analysis confirmedanalysis confirmed that the time that whenthe time the when first treatment the first trea linetment was initiatedline was isinitiated a prognostic is a prognostic factor for factor survival for (HRsurvival 0.77, (HR CI 0.60-0.99, 0.77, CI 0.60-0.99,p = 0.043) p (see = 0.043) Figure (see3). Figure 3).

J. Clin. Med. 2020, 9, 648 7 of 15 J. Clin. Med. 2020, 9, 648 7 of 15

Figure 3. Forest plot for the adjusted survival analysis for all patients. Cox proportional hazard model Figureto analyze 3. Forest the eff plotect offor multiple the adjusted variables survival on the analysis survival for outcome all patients. of all Cox patients. proportional Abbreviations: hazard model ECOG, toEastern analyze Cooperative the effect Oncologyof multiple Group variables performance on the survival scale. N outcome= 266 (patients of all patients. with missing Abbreviations: values of ECOG,variables Eastern were excludedCooperative from Oncology the analysis). Group Asterisks performance indicate scale.p-value, N = *266 0.01 (patients to 0.05, ** with 0.001 missing to 0.01, values*** 0.0001 of variables to 0.001. were # indicates excluded number. from the analysis). Asterisks indicate p-value, * 0.01 to 0.05, ** 0.001 to 0.01, *** 0.0001 to 0.001. # indicates number. From beginning of second line therapy, mOS was 7.14 months versus 9.39 months (p = 0.010 (GBWt),From HR beginning 0.77, CI 0.57of second to 1.04) line and therapy, from the mO beginningS was 7.14 of third months line therapy,versus 9.39 mOS months was 5.89 (p months= 0.010 (GBWt),versus 6.15 HR months 0.77, CI ( p0.57= 0.414 to 1.04) (LRt), and HR from 0.89, the CI beginning 0.59 to 1.34). of third Gem line/NabP therapy, and mFOLFIRINOX mOS was 5.89 weremonths the versusmost commonly 6.15 months administered (p = 0.414 (LRt), first line HR therapies 0.89, CI in0.59 both to cohorts1.34). Gem/NabP and accounted and mFOLFIRINOX for 67% of all first were line thetherapies most commonly from 2011 administered to 2017. When first comparing line therapies these firstin both line cohorts regimens and between accounted the for two 67% time of cohorts,all first line therapies from 2011 to 2017. When comparing these first line regimens between the two time cohorts, mOS did not differ significantly for first line Gem/NabP (8.23 months versus 9.82 months, p

J. Clin. Med. 2020, 9, 648 8 of 15

J. Clin. Med. 2020, 9, 648 8 of 15 J. Clin. Med. 2020, 9, 648 8 of 15 mOS did not differ significantly for first line Gem/NabP (8.23 months versus 9.82 months, p = 0.260 = 0.260 (LRt), HR 0.75, CI 0.49 to 1.14), mFOLFIRINOX (10.38 months versus 16.14 months,p p = 0.409 (LRt),= 0.260 HR (LRt), 0.75, HR CI 0.75, 0.49 toCI 1.14),0.49 to mFOLFIRINOX 1.14), mFOLFIRINOX (10.38 months (10.38 months versus versus 16.14 months, 16.14 months,= 0.409 p = (LRt),0.409 HR (LRt), HR 0.84, CI 0.48 to 1.49) and Gem/NabP or mFOLFIRINOX (9.62 months versus 10.61 months, 0.84,(LRt), CI HR 0.48 0.84, to 1.49)CI 0.48 and to 1.49) Gem and/NabP Gem/NabP or mFOLFIRINOX or mFOLFIRINOX (9.62 months (9.62 mont versushs versus 10.61 months,10.61 months,p = 0.244 p = 0.244 (LRt), HR 0.79, CI 0.57 to 1.09) (see Figure 4). (LRt),p = 0.244 HR (LRt), 0.79, CIHR 0.57 0.79, to CI 1.09) 0.57 (seeto 1.09) Figure (see4 ).Figure 4).

Figure 4. Survival between most popular first line regimens between the two time periods. Kaplan- FigureFigure 4.4. SurvivalSurvival betweenbetween mostmost popularpopular firstfirst line line regimens regimens between between the the two two time time periods. periods. Kaplan-Meier Kaplan- Meier curves of patients who started a first line therapy with gemcitabine/nab-paclitaxel (A), curvesMeier ofcurves patients of patients who started who a firststarted line a therapy first line with therapy gemcitabine with /gemcitabine/nab-paclitaxelnab-paclitaxel (A), mFOLFIRINOX (A), mFOLFIRINOX (B) or with either one of these regimens (C) according to the year when the first (mFOLFIRINOXB) or with either (B one) or of with these either regimens one of (C )these according regimens to the (C year) according when the to firstthe year systemic when treatment the first was systemic treatment was administered. Horizontal grid lines indicate the survival proportion of 25% administered.systemic treatment Horizontal was administered. grid lines indicateHorizontal the grid survival lines indicate proportion the ofsurvival 25% and proportion intersections of 25% with and intersections with vertical grid lines indicate the survival time for the respective Kaplan-Meier verticaland intersections grid lines with indicate vertical the survivalgrid lines time indicate for the the respective survival time Kaplan-Meier for the respective curves atKaplan-Meier this point. curves at this point. curves at this point. In neither time cohort nor in the whole cohort from 2011 to 2017, patients treated with these two In neither time cohort nor in the whole cohort from 2011 to 2017, patients treated with these two first lineIn neither regimes time di cohortffered nor substantially in the whole in cohort survival from when 2011 comparedto 2017, patients to all treated other therapieswith these (mOS two in first line regimes differed substantially in survival when compared to all other therapies (mOS in first line regimes differed substantially in survival when compared to all other therapies (mOS in cohortcohort A 9.62 versus versus 8.46 8.46 months, months, p p= =0.6110.611 (LRt), (LRt), HR HR 0.96, 0.96, CI CI0.68 0.68 to 1.35; to 1.35; in cohort in cohort B 10.61 B 10.61 versus versus cohort A 9.62 versus 8.46 months, p = 0.611 (LRt), HR 0.96, CI 0.68 to 1.35; in cohort B 10.61 versus 15.3715.37 months, pp == 0.6410.641 (LRt), (LRt), HR HR 0.98, 0.98, CI CI0.64 0.64 to 1.51; to 1.51; in the in thetotal total cohort cohort 10.18 10.18 versus versus 11.01 11.01months, months, p 15.37 months, p = 0.641 (LRt), HR 0.98, CI 0.64 to 1.51; in the total cohort 10.18 versus 11.01 months, p p= =0.3560.356 (LRt), (LRt), HR HR 0.88, 0.88, CI CI0.69 0.69 to 1.13) to 1.13) (see (see Figure Figure 5). 5). = 0.356 (LRt), HR 0.88, CI 0.69 to 1.13) (see Figure 5).

Figure 5. Survival with most popular first line regimens versus all other therapies within the two time FigureFigure 5. SurvivalSurvival with with most most popular popular firs firstt line line regimens regimens versus versus all allothe otherr therapies therapies within within the two the twotime time periods. Kaplan-Meier curves of patients who started a first line therapy with either one of the two periods.periods. Kaplan-Meier curves curves of of patients patients who who star startedted a first a first line line therapy therapy with with either either one oneof the of two the two regimens gemcitabine/nab-paclitaxel and mFOLFIRINOX compared to all other therapies between regimensregimens gemcitabinegemcitabine/nab-paclitaxel/nab-paclitaxel and and mFOLFI mFOLFIRINOXRINOX compared compared to all to allother other therapies therapies between between 2011–2013 (A), 2014–2017 (B), as well as for the whole time period (C). 2011–20132011–2013 ( A), 2014–2017 2014–2017 ( (BB),), as as well well as as for for the the whole whole time time period period (C). (C ).

The ratio of patients who started a second line therapy was similar in both cohorts, however, the TheThe ratio of patients patients who who started started a asecond second line line th therapyerapy was was similar similar in both in both cohorts, cohorts, however, however, the the frequency of used regimens, unlike to first line regimens, was different. When excluding CRTx, which frequencyfrequency ofof used regimens, regimens, unlike unlike to to first first line line re regimens,gimens, was was different. different. When When excluding excluding CRTx, CRTx, which which was applied in both cohorts at comparable rates (19% versus 18% in second line), the survival from waswas appliedapplied inin bothboth cohortscohorts atat comparablecomparable ratesrates (19%(19% versusversus 18%18% inin secondsecond line),line), thethe survivalsurvival fromfrom first first line and second line of patients treated with the most commonly used second line regimens linefirst and line second and second line of line patients of patients treated treated with thewith most the commonlymost commonly used second used second line regimens line regimens (cohort A: (cohort A: 21% Gem/NabP, 14% Gem/(Cap)/Erlotinib, 14% FP/Ox; cohort B: 30% FP/Nal-IRI, 18% 21%(cohort Gem A:/NabP, 21% Gem/NabP, 14% Gem/(Cap) 14%/ Erlotinib,Gem/(Cap)/Erlotinib, 14% FP/Ox; 14% cohort FP/Ox; B: 30% cohort FP/Nal-IRI, B: 30% 18%FP/Nal-IRI, FP/Ox) 18% differed FP/Ox) differed significantly (mOS from first line 11.41 versus 15.21 months, p = 0.012 (LRt), HR 0.64, significantlyFP/Ox) differed (mOS significantly from first (mOS line 11.41from versusfirst line 15.21 11.41 months, versus 15.21p = 0.012 months, (LRt), p = HR0.012 0.64, (LRt), CI HR 0.41 0.64, to 0.99; CI 0.41 to 0.99; mOS from second line 5.72 versus 7.47 months, p = 0.021 (GBWt), HR 0.68, CI 0.44 to mOSCI 0.41 from to 0.99; second mOS line from 5.72 second versus line 7.47 5.72 months, versusp = 7.470.021 months, (GBWt), p = HR 0.021 0.68, (GBWt), CI 0.44 HR to 1.05)0.68, (seeCI 0.44 Figure to 6). 1.05) (see Figure 6). 1.05) (see Figure 6).

J. Clin. Med. 2020, 9, 648 9 of 15 J. Clin. Med. 2020, 9, 648 9 of 15

FigureFigure 6.6. SurvivalSurvival according according to to first first line and second line treatment for most popular regimens. Kaplan-MeierKaplan-Meier curves curves of of patients patients who who were were treated treated with withone of one the most of the popular most popularsecond line second treatment line treatmentregimes, calculated regimes, from calculated the initiation from the of the initiation first (A of) and the second first (A (B)) and treatment second line (B )according treatment to line the accordingyear when tothe the first year systemic when tr theeatment first systemic was administered. treatment wasThe most administered. popular treatment The most lines popular were treatmentgemcitabine/nab-paclitaxel, lines were gemcitabine gemcitabine/(capecitabine)/erlotinib,/nab-paclitaxel, gemcitabine/(capecitabine) as/erlotinib, well as well asas fluoropyrimidinesfluoropyrimidines/oxaliplatin/oxaliplatin for for the cohortthe cohort 2011–2013 2011–2013 and fluoropyrimidines and fluoropyrimidines/nanoliposomal/nanoliposomal irinotecan, asirinotecan, well as fluoropyrimidines as well as fluoropyrimidi/oxaliplatinnes/oxaliplatin for the cohort for 2014–2017. the cohort 2014–2017. Multivariate analysis (Cox proportional hazard model) with potentially influencing variables was Multivariate analysis (Cox proportional hazard model) with potentially influencing variables performed for these two cohorts. The difference in survival between patients treated with the most was performed for these two cohorts. The difference in survival between patients treated with the commonly used second line regimens (2011–2013 vs. 2014–2017) remained statistically significant (HR most commonly used second line regimens (2011–2013 vs. 2014–2017) remained statistically 0.56, CI 0.34–0.94, p = 0.029) (see Figure7). significant (HR 0.56, CI 0.34–0.94, p = 0.029) (see Figure 7). Among the total cohort, three different first to second line treatment sequences were analyzed. The survival curves for the sequences, mFOLFIRINOX followed by Gem/NabP (mOS 15.64 months), Gem/NabP followed by FP/Nal-IRI (mOS 14.27 months), Gem/NabP followed by FP/Ox (mOS 13.62 months), did not differ significantly from each other (p = 0.692) (see Figure8).

J. Clin. Med. 2020, 9, 648 10 of 15 J. Clin. Med. 2020, 9, 648 10 of 15

Figure 7. Forest plot for the adjusted survival analysis from patients treated with the most commonly Figureused second 7. Forest line plot regimens for the of adjusted both time survival cohorts. analysis Cox proportional from patients hazard treated model with to the analyze most thecommonly effect of usedmultiple second variables line regimens on the survival of both outcome time cohorts. of patients Cox proportional who have been hazard treated model with to the analyze most commonly the effect ofused multiple second variables line treatment on the regimens. survival Abbreviations:outcome of patients ECOG, who Eastern have Cooperative been treated Oncology with the Group most commonlyperformance used scale. secondN = 82 (patientsline treatment with missing regimens values. Abbreviations: of variables were ECOG excluded, Eastern from theCooperative analysis). OncologyAsterisks indicateGroup performancep-value, * 0.01 scale. to 0.05. N = #82 indicates (patients number. with missing values of variables were excluded from the analysis). Asterisks indicate p-value, * 0.01 to 0.05. # indicates number.

Among the total cohort, three different first to second line treatment sequences were analyzed. The survival curves for the sequences, mFOLFIRINOX followed by Gem/NabP (mOS 15.64 months), Gem/NabP followed by FP/Nal-IRI (mOS 14.27 months), Gem/NabP followed by FP/Ox (mOS 13.62 months), did not differ significantly from each other (p = 0.692) (see Figure 8).

J. Clin. Med. 2020, 9, 648 11 of 15 J. Clin. Med. 2020, 9, 648 11 of 15

Figure 8. SurvivalSurvival with with different different first first to to second second line line treatment treatment sequences. sequences. Kaplan-Meier Kaplan-Meier curves curves of ofpatients patients who who were were treated treated with with one one of ofthe the three three treatment treatment sequences sequences from from first first to to second second line: mFOLFIRINOX followed by gemcitabinegemcitabine/nab-paclitaxel/nab-paclitaxel (red line), gemgem/nab-paclitaxel/nab-paclitaxel followed by fluoropyrimidinesfluoropyrimidines/nanoliposomal/nanoliposomal irinotecan (blue lin line)e) and gemcitabinegemcitabine/nab-paclitaxel/nab-paclitaxel followed by fluoropyrimidinesfluoropyrimidines/oxaliplatin/oxaliplatin (green(green line).line).

4. Discussion 4. Discussion New therapeutic agents for metastatic PAC have recently expanded treatment options for this New therapeutic agents for metastatic PAC have recently expanded treatment options for this disease. Our aim was to assess the impact of these novel chemotherapy regimens on the treatment disease. Our aim was to assess the impact of these novel chemotherapy regimens on the treatment landscape and clinical outcome in a cohort of patients, which falls in the period of their introduction to landscape and clinical outcome in a cohort of patients, which falls in the period of their introduction the clinic. These treatment options include Gem/NabP, FOLFIRINOX, FP/Nal-IRI and FP/Ox. In concert to the clinic. These treatment options include Gem/NabP, FOLFIRINOX, FP/Nal-IRI and FP/Ox. In withconcert the with timeline the timeline of clinical of introduction clinical introduction of these regimens, of these regimens, we observed we a observed major increase a major of Gemincrease/NabP of whenGem/NabP comparing when firstcomparing line regimens first line in regimens the period in from the period 2011 to from 2013 2011 against to 2013 2014 against to 2017 2014 from to 26% 2017 to p 69%from ( 26%< 0.0001), to 69% while (p < 0.0001), the rates while for administration the rates for ofadministration mFOLFIRINOX of mFOLFIRINOX as a front-line regimen as a front-line differed p notregimen substantially differed (19%not substa versusntially 15%, (19%= 0.4415). versus This 15%, increase p = 0.4415). of Gem This/NabP increase in the of first Gem/NabP line setting in wasthe mainlyfirst line accounted setting was to amainly reduction accounted of theother to a redu chemotherapyction of the regimens other chemotherapy Gem/Ox, Gem regimens/Ox/Erlotinib Gem/Ox, and CRTxGem/Ox/Erlotinib with less evidence and CRTx as eff withective less treatment evidence options. as effective treatment options. There have been two other studies from an Amer Americanican and a Japanese group assessing treatment patterns overover recent recent years years where where modern modern first linefirst chemotherapies line chemotherapies were introduced were introduced and which and showed which a similarshowed trend a similar towards trend increased towards usage increased of Gem /usageNabP asof aGem/NabP preferred first as linea preferred regimen rangingfirst line from regimen ~40% toranging 57% [ 16from,17]. ~40% In contrast, to 57% another[16,17]. In multi-institutional contrast, another large multi-institutional study from Canada large study showed from a tendency Canada towardshowed increased a tendency administration toward increased of FOLFIRINOX administrati ason first of line FOLFIRINOX treatment (60%) as first [18 ].line In countries,treatment where(60%) S-1[18]. is In in countries, clinical use where for PAC, S-1 changes is in clinical in frontline use fo treatmentr PAC, changes patterns in havefrontline also beentreatment observed. patterns However, have thealso use been of observed. Gem/NabP However, or FOLFIRINOX the use of does Gem/NabP not exceed or FOLFIRINOX 25% for each treatment does not exceed option, 25% according for each to twotreatment recently option, published according studies to from two Japan recently and Taiwanpublished [19 ,studies20]. Referring from Japan to our and results Taiwan regarding [19,20]. the secondReferring and to third our lineresults setting, regarding Gem/NabP the second was the and most third commonly line setting, used regimenGem/NabP in thewas earlier the most time periodcommonly (21% used and regimen 26%), while, in the if consideringearlier time theperiod high (21% rate and of this 26%), regimen while, in if theconsidering frontline settingthe high of rate the laterof this period, regimen it was in the subsequently frontline setting replaced of the by later FP/Nal-IRI period, (30% it was and subsequently 22%). To sum replaced up, when by FP/Nal-IRI comparing the(30% treatment and 22%). landscape To sum ofup, the when two timecomparing periods, the we treatment observed alandscape clear increase of the of two Gem time/NabP, periods, a regimen we withobserved high a evidence clear increase as an e ffofective Gem/NabP, front-line a therapyregimen with with the high later evidence line treatment as an optioneffective FP /front-lineNal-IRI. therapyMost with importantly, the later line the treatment median survival option FP/Nal-IRI. of patients, who started their systemic treatment in the laterMost time period,importantly, improved the median when compared survival of to patien the earlierts, who one started (mOS 11.9their versus systemic 8.89, treatmentp = 0.035) in with the survivallater time rates period, at 12, improved 18 and 24 when months compared after initiation to the earlier of first one line (mOS therapy 11.9 of versus 37.1% 8.89, versus p = 49.1%, 0.035) 16.7% with versussurvival 28.7% rates andat 12, 8.3% 18 and versus 24 months 17.3%. after This initiation survival of benefit first line was therapy still evident of 37.1% when versus comparing 49.1%, 16.7% the survivalversus 28.7% rates and from 8.3% the beginningversus 17.3%. of second This survival line of thebenefit patients was whostill startedevident theirwhen first comparing line therapy the betweensurvival rates 2011 andfrom 2013 the (mOSbeginning 7.14 versusof second 9.39 line months, of thep =patients0.010), who but not started from their the beginning first line therapy of third line.between We 2011 analyzed and 2013 if the (mOS increased 7.14 versus ratio of9.39 first months, line treatment p = 0.010), options but not withfrom higherthe beginning evidence of asthird an eline.ffective We therapyanalyzed in if the the later increased time period ratio could of first be responsibleline treatment for theoptions improvement with higher in survival, evidence but as didan effective therapy in the later time period could be responsible for the improvement in survival, but did not observe a difference when comparing Gem/NabP or mFOLFIRINOX, as well as both regimens

J. Clin. Med. 2020, 9, 648 12 of 15 not observe a difference when comparing Gem/NabP or mFOLFIRINOX, as well as both regimens to all other therapies among the two time cohorts and the total cohort (see Figure5). There was also no significant survival difference between patients who started Gem/NabP or mFOLFIRINOX, as well as with either one of these two regimens among the two-time cohorts (see Figure4). However, when comparing later time points of these survival curves, it was apparent that the survival curves separated at a time point where most patients already started a second line chemotherapy (compare intersection of indicated grid lines with survival curves from Figure4C). In detail, at the time point of mOS of the respective time cohort, 90% and 84% of the patients who started a systemic therapy between 2011 to 2013 and 2014 to 2017 and also received a second line treatment already had begun this second or later treatment line. Next, we compared the survival of patients who received a second line treatment with the most commonly used regimens (cohort A: 21% Gem/NabP, 14% Gem/(Cap)/Erlotinib, 14% FP/Ox; cohort B: 30% FP/Nal-IRI, 18% FP/Ox) and observed that indeed there was a difference in survival time from first and second line treatment between the time cohorts (mOS from first line 11.41 versus 15.21 months, p = 0.012; mOS from second line 5.72 versus 7.47 months, p = 0.021). This survival benefit remained significant in the multivariate model when comparing patients who have been treated with the most commonly used second line regimens of the respective time cohorts (HR 0.56, CI 0.34–0.94, p = 0.029). These results suggest, that a change in treatment patterns, in particular that Gem/NabP moved from the second and third line to the first line setting in the later cohort and that FP/Nal-IRI was introduced as an effective second line regimen, was responsible for the observed survival benefit at our institution. Our hypothesis is strengthened by other reports. A recent study by Kordes M et al. reported that the sequence of gemcitabine-based therapy after 5-FU/oxaliplatin with or without irinotecan was associated with poorer outcomes than a 5-FU-based treatment after a gemcitabine-based doublet therapy [21]. Another study by Caparello and colleagues reports for the subgroup of patients who received Gem/NabP after the failure of first line FOLFIRINOX disappointing results in terms of both activity (RR: 7%; DCR: 23%) and survival (median PFS: 1.95 months; median OS: 5.4 months) [22]. Likewise, also other studies report only modest activity of second line Gem/NabP with median overall survival rates of 5.69 months and 5.75 months [23,24]. In descending order after Gem/NabP another frequently used second line therapy option in the earlier cohort was Gem/Erlotinib. In light of missing data concerning the effectiveness of Gem/Erlotinib as a second line treatment regimen and considering the only very modest improvement by the addition of erlotinib to gemcitabine reported in the literature, we believe that this regimen should not be considered as an effective second line treatment option [10,25]. In contrast to that, the most commonly used second line regimens in the earlier cohort are proven effective options as second line therapies, which has been shown particularly for FP/Nal-IRI [13]. This is also reflected by current treatment guidelines which mainly suggest FP/Nal-IRI after the failure of a gemcitabine-based first line chemotherapy. There are limited data concerning the impact of new chemotherapy regimens on the outcome of patients with advanced or metastatic PAC in Caucasians, highlighting the main findings of our study. We are aware of another study conducted in a Taiwanese cohort, which also observed improved survival over a seven-year time period up to 2016 [20]. However, the authors associated this development to the increased use of S-1, which is not in clinical use in non-Asian countries, and thus, makes the results difficult to compare. In line with the increased survival of patients treated with modern chemotherapy regimens, we also noted that the ratio of patients who received a third line therapy increased from 29% to 37% over time. These numbers also show a positive trend towards a prolonged treatment period when compared to other studies, which evaluated historical treatment patterns from 2005 to 2015 in Europe, as well as the US and reported rates between 3 to 22% for third line treatment [16,26,27]. This opens up the discussion for the optimal treatment sequences. The mOS seen across the analyzed first to second line treatment sequences was encouraging, ranging from 13.62 to 15.64 months. Given that FOLFIRINOX is associated with a higher toxicity rate, it is noteworthy that the survival of patients who also received a second line therapy was comparable and did not differ significantly J. Clin. Med. 2020, 9, 648 13 of 15 from patients, who had not been treated with mFOLFIRINOX in the first line. There are limiting data concerning the effectiveness of different treatment sequences, but a recently published study, which evaluated the use of Gem/NabP followed by mFOLFIRINOX or vice versa in 48 patients did not report any difference in survival between the two groups (mOS 13.7 and 13.8 months, p = 0.9) [28]. Supporting these and our findings, also a study by Glassman and colleagues reported promising survival rates for the sequence Gem/NabP followed by FP/Nal-IRI (mOS 23.0 months). They did not find a significant difference to the sequence FOLFIRINOX followed by Gem/NabP and FP/Nal-IRI (mOS 25.5 months) [29]. Furthermore, in the previously mentioned study by Kordes et al. patients who received FOLFIRINOX had a shorter median OS (9.9 months, 95% CI; 8.1–11.7) than previously reported [21]. In that respect and in light of no head-to-head prospective comparison trials, there is still no evidence that FOLFIRINOX in the palliative setting is associated with increased survival compared to Gem/NabP. This triplet-therapy, however, is regarded more toxic when compared to doublet-regimens, which should be considered in particular if palliation is the primary goal of the treatment. Our study has several limitations, including being conducted as a single-institution analysis, the relatively small sample size of the different sequence subgroups, the inclusion of patients only when they were presented to the multidisciplinary tumor board, its retrospective character, as well as missing data for adverse events and toxicity. Furthermore, we cannot rule out a possible bias resulting from potential earlier switching to second line chemotherapy in the earlier time cohort, due to closer monitoring of disease activity. However, we have no reason to believe that this has occurred because standard diagnostic and therapeutic work-up, including intervals between reassessment of treatment responses have not changed during the observational period. Moreover, provision of better supportive therapy, which would allow more patients to complete systemic treatment for advanced-stage PAC could have impacted the results of our study. In this respect, there might also be additional influencing factors like better handling of dose adjustments when encountering toxicities. These potential influencing factors are, however, not very likely to explain the improved survival between our cohorts, as individual components of the different chemotherapies like platinum salts, fluoropyrimidines, irinotecan, gemcitabine, as well as erlotinib and nab-paclitaxel are classic antineoplastic drugs which were already in use long before the observation period of this study started. Thus, it is not very likely that our clinical management concerning toxicities and dose adjustments has changed during 2011 and 2017. Emerging promising predictive biomarkers like mismatch-repair deficiency (or MSI-high) for checkpoint blockade or BRCA mutations for PARP inhibitors and platinum salts are currently entering the clinic, but have not been used for any patient in the current study [30,31]. With the year 2018, our cancer center started to implement testing for mismatch-repair deficiency and BRCA mutations of tumors from PAC patients. Although, only a small subset of patients is affected by these conditions, it is an important step towards personalized treatment of PAC patients, which will hopefully improve the prognosis and add new effective treatment options.

5. Conclusions This study describes the changes in the treatment landscape in the clinical management of patients with unresectable PAC and furthermore underscores the effectiveness of the new chemotherapy regimens, which hopefully invigorates the discussion about optimal treatment sequencing, as well as the choice of the optimal frontline regimen. It is encouraging that survival rates have increased at our institution. This positive trend is most likely explained by the introduction of novel treatment options. Furthermore, the high rate of patients that started a third line therapy (37%) is stimulating. Additional analyses of sequential treatment in a sufficiently large study population are warranted.

Author Contributions: Conceptualization, M.K. and G.W.P.; data curation, M.K., M.U., M.S., G.V.K., W.S. and G.W.P.; formal analysis, M.K., M.U. and G.W.P.; investigation, M.K. and G.W.P.; methodology, M.K., M.U., D.B. and G.W.P.; supervision, G.W.P.; visualization, M.K.; writing—original draft, M.K. and G.W.P.; writing—review J. Clin. Med. 2020, 9, 648 14 of 15 and editing, M.U., D.B., M.S., G.V.K. and W.S. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Acknowledgments: We want to thank for the biostatistical support, which was provided by ASOKLIF (Arbeitsgruppe zur Systemoptimierung klinischer Forschungsprojekte). Conflicts of Interest: G.W.P.: Personal financial interests: Merck Serono, Roche, Amgen, Sanofi, Lilly, Servier, Taiho, Bayer, Halozyme, BMS, Celgene, Pierre Fabre, Shire, Institutional financial interests—clinical trials: Celgene, Array, Servier, Bayer, BostonBiomedical, Merck, BMS; All other authors declare no conflict of interest.

References

1. Rahib, L.; Smith, B.D.; Aizenberg, R.; Rosenzweig, A.B.; Fleshman, J.M.; Matrisian, L. Projecting cancer incidence and deaths to 2030: The unexpected burden of thyroid, liver, and cancers in the United States. Cancer Res. 2014, 74, 2913–2921. [CrossRef] 2. Malvezzi, M.; Bertuccio, P.; Rosso, T.; Rota, M.; Levi, F.; La Vecchia, C.; Negri, E. European cancer mortality predictions for the year 2015: Does lung cancer have the highest death rate in EU women? Ann. Oncol. 2015, 26, 779–786. [CrossRef] 3. Singhi, A.D.; Koay, E.J.; Chari, S.T.; Maitra, A. Early Detection of Pancreatic Cancer: Opportunities and Challenges. Gastroenterology 2019, 156, 2024–2040. [CrossRef] 4. Kleeff, J.; Korc, M.; Apte, M.; La Vecchia, C.; Johnson, C.D.; Biankin, A.V.; Neale, R.E.; Tempero, M.; Tuveson, D.A.; Hruban, R.H.; et al. Pancreatic cancer. Nat. Rev. Dis. Primers 2016, 2, 16022. [CrossRef] 5. Balachandran, V.P.; Beatty, G.L.; Dougan, S.K. Broadening the Impact of Immunotherapy to Pancreatic Cancer: Challenges and Opportunities. Gastroenterology 2019, 156, 2056–2072. [CrossRef] 6. Neesse, A.; Bauer, C.A.; Öhlund, D.; Lauth, M.; Buchholz, M.; Michl, P.; Tuveson, D.A.; Gress, T. Stromal biology and therapy in pancreatic cancer: Ready for clinical translation? Gut 2019, 68, 159–171. [CrossRef] 7. Burris, A.H.; Moore, M.J.; Andersen, J.; Green, M.R.; Rothenberg, M.L.; Modiano, M.R.; Cripps, M.C.; Portenoy, R.K.; Storniolo, A.M.; Tarassoff, P.; 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, 2403–2413. [CrossRef] 8. Heinemann, V.; Böck, S.; Hinke, A.; Labianca, R.; Louvet, C. Meta-analysis of randomized trials: Evaluation of benefit from gemcitabine-based combination chemotherapy applied in advanced pancreatic cancer. BMC Cancer 2008, 8, 82. [CrossRef] 9. Sultana, A.; Smith, C.T.; Cunningham, D.; Starling, N.; Neoptolemos, J.P.; Ghaneh, P. Meta-analyses of chemotherapy for locally advanced and metastatic pancreatic cancer. J. Clin. Oncol. 2007, 25, 2607–2615. [CrossRef] 10. Moore, M.J.; Goldstein, D.; Hamm, J.; Figer, A.; Hecht, J.R.; Gallinger, S.; Au, H.J.; Murawa, P.; Walde, D.; Wolff, R.A.; et al. 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, 1960–1966. [CrossRef] 11. Conroy, T.; Desseigne, F.; Ychou, M.; Bouche, O.; Guimbaud, R.; Bécouarn, Y.; Adenis, A.; Raoul, J.-L.; Gourgou-Bourgade, S.; De La Fouchardiere, C.; et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N. Engl. J. Med. 2011, 364, 1817–1825. [CrossRef] 12. Von Hoff, D.D.; Ervin, T.; Arena, F.P.; Chiorean, E.G.; Infante, J.; Moore, M.; Seay, T.; Tjulandin, S.A.; Ma, W.W.; Saleh, M.N.; et al. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N. Engl. J. Med. 2013, 369, 1691–1703. [CrossRef] 13. Wang-Gillam, A.; Li, C.-P.; Bodoky, G.; Dean, A.; Shan, Y.-S.; Jameson, G.; Macarulla, T.; Lee, K.-H.; Cunningham, D.; Blanc, J.F.; et al. Nanoliposomal irinotecan with fluorouracil and folinic acid in metastatic pancreatic cancer after previous gemcitabine-based therapy (NAPOLI-1): A global, randomised, open-label, phase 3 trial. Lancet 2016, 387, 545–557. [CrossRef] 14. Oettle, H.; Riess, H.; Stieler, J.M.; Heil, G.; Schwaner, I.; Seraphin, J.; Görner, M.; Mölle, M.; Greten, T.F.; Lakner, V.; et al. Second-line oxaliplatin, folinic acid, and fluorouracil versus folinic acid and fluorouracil alone for gemcitabine-refractory pancreatic cancer: Outcomes from the CONKO-003 trial. J. Clin. Oncol. 2014, 32, 2423–2429. [CrossRef] J. Clin. Med. 2020, 9, 648 15 of 15

15. Gill, S.; Ko, Y.-J.; Cripps, C.; Beaudoin, A.; Dhesy-Thind, S.; Zulfiqar, M.; Zalewski, P.; Do, T.; Cano, P.; Lam, W.Y.H.; et al. PANCREOX: A Randomized Phase III Study of /Leucovorin With or Without Oxaliplatin for Second-Line Advanced Pancreatic Cancer in Patients Who Have Received Gemcitabine-Based Chemotherapy. J. Clin. Oncol. 2016, 34, 3914–3920. [CrossRef] 16. Abrams, T.A.; Meyer, G.; Meyerhardt, J.A.; Wolpin, B.M.; Schrag, D.; Fuchs, C.S. Patterns of Chemotherapy Use in a U.S.-Based Cohort of Patients With Metastatic Pancreatic Cancer. Oncologist 2017, 22, 925–933. [CrossRef] 17. Nagai, K.; Terashima, T.; Yamashita, T.; Kido, H.; Hayashi, K.; Sunagozaka, H.; Nakai, R.; Mizuno, H.; Hayakawa, H.; Nomura, Y.; et al. Dynamical Changes of Treatment Patterns and Outcomes of Unresectable Pancreatic Cancer Patients in Real-Life Practice. J. Clin. Oncol. 2019, 37, 407. [CrossRef] 18. Cheung, W.Y.; Zhang, H.; Tang, P.A.; Spratlin, J.; Lee-Ying, R.M.; Goodwin, R.A.; Meyers, B.M.; Armstrong, D.; Ramjeesingh, R.; Vickers, M.M.; et al. A real world multicenter study of first (1L) and second (2L) line treatment patterns and outcomes in advanced pancreatic cancer (APC). J. Clin. Oncol. 2018, 36, 476. [CrossRef] 19. Terashima, T.; Yamashita, T.; Sakai, A.; Ohta, H.; Hinoue, Y.; Toya, D.; Kawai, H.; Yonejima, M.; Urabe, T.; Noda, Y.; et al. Treatment Patterns and Outcomes of Unresectable Pancreatic Cancer Patients in Real-Life Practice: A Region-Wide Analysis. Jpn. J. Clin. Oncol. 2018, 48, 966–973. [CrossRef] 20. Chou, W.-C.; Chen, Y.-Y.; Hung, C.-Y.; Chen, J.-S.; Lu, C.-H.; Chang, P.-H. Evolution of the Chemotherapeutic Landscape and Survival Outcome in Patients With Metastatic Pancreatic Cancer: A Four-Institute Cohort Study in Taiwan, 2010-2016. Cancer Manag. Res. 2019, 11, 2119–2127. [CrossRef] 21. Kordes, M.; Yu, J.; Malgerud, O.; Liljefors, M.G.; Löhr, J.M. Survival Benefits of Chemotherapy for Patients With Advanced Pancreatic Cancer in A Clinical Real-World Cohort. Cancers 2019, 11, 1326. [CrossRef] 22. Caparello, C.; Vivaldi, C.; Fornaro, L.; Musettini, G.; Pasquini, G.; Catanese, S.; Masi, G.; Lencioni, M.; Falcone, A.; Vasile, E. Second-Line Therapy for Advanced Pancreatic Cancer: Evaluation of Prognostic Factors and Review of Current Literature. Futur. Oncol. 2016, 12, 901–908. [CrossRef] 23. Zhang, H.; Kellett, C.; Lambert, P.; Kim, C.A. Efficacy and Tolerability of Second-Line Nab-Paclitaxel and Gemcitabine After Failure of First-Line FOLFIRINOX for Advanced Pancreas Cancer: A Single-Institution Experience. Clin. Color. Cancer 2018, 17, e451–e456. [CrossRef] 24. Zhang, Y.; Hochster, H.S.; Stein, S.; Lacy, J. Gemcitabine Plus Nab-Paclitaxel for Advanced Pancreatic Cancer After First-Line FOLFIRINOX: Single Institution Retrospective Review of Efficacy and Toxicity. Exp. Hematol. Oncol. 2015, 4, 29. [CrossRef] 25. Shin, S.; Park, C.M.; Kwon, H.; Lee, K.-H. Erlotinib Plus Gemcitabine Versus Gemcitabine for Pancreatic Cancer: Real-World Analysis of Korean National Database. BMC Cancer 2016, 16, 443. [CrossRef] 26. Smyth, E.N.; Bapat, B.; Ball, D.E.; André, T.; Kaye, J.A. Metastatic Pancreatic Adenocarcinoma Treatment Patterns, Health Care Resource Use, and Outcomes in France and the United Kingdom Between 2009 and 2012: A Retrospective Study. Clin. Ther. 2015, 37, 1301–1316. [CrossRef] 27. Seal, B.; Xia, F.; Rietschel, P.; Germino, R.; Asche, C.; D’Souza, A.; Morland, K.; Eaddy, M. Treatment Patterns And Cost Of Care For Patients With Pancreatic Cancer. Value Heal. 2014, 17, A132. [CrossRef] 28. Vogl, U.; Andalibi, H.; Klaus, A.; Vormittag, L.; Schima, W.; Heinrich, B.; Kafka, A.; Winkler, T.; Öhler, L. Nab-Paclitaxel and Gemcitabine or FOLFIRINOX As First-Line Treatment in Patients With Unresectable Adenocarcinoma of the Pancreas: Does Sequence Matter? BMC Cancer 2019, 19, 28. [CrossRef] 29. Glassman, D.C.; Palmaira, R.L.; Covington, C.M.; Desai, A.M.; Ku, G.Y.; Li, J.; Harding, J.J.; Varghese, A.M.; O’Reilly, E.; Yu, K.H. Nanoliposomal Irinotecan With Fluorouracil for the Treatment of Advanced Pancreatic Cancer, a Single Institution Experience. BMC Cancer 2018, 18, 693. [CrossRef] 30. Hu, Z.I.; Shia, J.; Stadler, Z.K.; Varghese, A.M.; Capanu, M.; Salo-Mullen, E.; Lowery, M.A.; Diaz, L.A.; Mandelker, D.; Yu, K.H.; et al. Evaluating Mismatch Repair Deficiency in Pancreatic Adenocarcinoma: Challenges and Recommendations. Clin. Cancer Res. 2018, 24, 1326–1336. [CrossRef] 31. Golan, T.; Hammel, P.; Reni, M.; Van Cutsem, E.; Macarulla, T.; Hall, M.J.; Park, J.-O.; Hochhauser, D.; Arnold, D.; Oh, D.Y.; et al. Maintenance Olaparib for Germline BRCA-Mutated Metastatic Pancreatic Cancer. N. Engl. J. Med. 2019, 381, 317–327. [CrossRef][PubMed]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).