Real-world feasibility and safety of neoadjuvant FOLFIRINOX in pancreatic ductal adenocarcinoma: a retrospective single-center study Authors: Soetkin D’Haens, Antoon Billiet, Halit Topal, Baki Topal, Filip Van Herpe, Jeroen Dekervel Published: 2026 ABSTRACT BackgroundNeoadjuvant chemotherapy in non-metastatic pancreatic ductal adenocarcinoma (PDAC) is increasingly being studied and used in clinical practice. However, this strategy is associated with significant toxicity, biopsy-related and biliary complications and possible treatment delay. We aimed to assess the feasible number of cycles of neoadjuvant (modified) FOLFIRINOX and the real-world toxicity, to inform an optimal interventional trial design.MethodsIn this descriptive monocentric study, we identified patients with borderline resectable or locally advanced PDAC, treated with upfront (m)FOLFIRINOX from 2020 to 2023. We analyzed patient and disease characteristics, dosing of therapy, occurrence of serious adverse events (SAEs) and reasons for discontinuation, and compared this data with available evidence in the literature.ResultsEighty-one patients were included. Median age was 65, most patients had an ECOG performance status of 0. 62.5% of tumors were borderline resectable, the others were locally advanced. 49 patients (60.5%) received mFOLFIRINOX, 32 (39.5%) received FOLFIRINOX and the median number of cycles administered was six. Most common reasons for discontinuation were surgery (n=45, 55.6%) and persistently inoperable or progressive disease (n=25, 30.9%), with toxicity accounting for a much smaller proportion (n=5, 6.2%). Global dose intensities gradually decreased over time. Dose reduction was necessary in 43 patients (53%), and 37 patients (45.7%) needed inpatient care. However, supportive care measures allowed most patients to overcome toxicity and finish the planned treatment cycles.ConclusionsWith adequate supportive care and dose modifications, the majority of patients were able to complete the assigned chemotherapy cycles. Eight cycles of mFOLFIRINOX appear feasible for most patients and may serve as an optimal regimen for evaluation in future interventional trials. Highlight box Key findings • In this real-world, single-center cohort of patients with non-metastatic pancreatic ductal adenocarcinoma (PDAC), neoadjuvant (modified) FOLFIRINOX was feasible, with the majority of patients able to complete a predefined number of treatment cycles. • Despite frequent toxicity and hospitalizations, treatment discontinuation due to toxicity was uncommon, and relative dose intensity remained high. What is known and what is new? • Neoadjuvant FOLFIRINOX improves outcomes in selected patients with non-metastatic PDAC, but is associated with substantial toxicity, raising concerns about real-world tolerability. • This study provides real-world evidence that neoadjuvant (m)FOLFIRINOX can be delivered with acceptable tolerability in routine clinical practice, with low rates of toxicity-driven treatment discontinuation. What is the implication, and what should change now? • These findings support the use of neoadjuvant (m)FOLFIRINOX in carefully selected patients with borderline resectable and locally advanced PDAC. • Emphasis should be placed on proactive supportive care and individualized dose adjustments rather than avoiding intensive neoadjuvant chemotherapy due to toxicity concerns. Introduction Pancreatic ductal adenocarcinoma (PDAC) is an aggressive neoplasm. It is responsible for almost five percent of cancer-related deaths, and thereby the sixth leading cause of cancer-related deaths worldwide (1). With increasing mortality, it is expected to become the second leading cause of cancer-related mortality in Western countries by 2030 (2). Radical resection offers the only chance of cure. However, due to its aggressive tumor biology and asymptomatic nature (3), most patients with PDAC are diagnosed with locally advanced or metastatic disease with surgery only feasible in 15–20% of cases at the time of diagnosis (4). Despite curative resection, the majority of patients develop postoperative recurrence, with a considerable proportion having disseminated disease (5). Early systemic spread and the insufficient effect of systemic therapies make systemic disease control challenging (3). The prognosis remains unsatisfactory with a 5-year overall survival of only 8–10% across all disease stages (1). Based on anatomical criteria and vascular involvement, non-metastatic PDAC is commonly classified into resectable, borderline resectable, and locally advanced disease. Resectable PDAC is characterized by the absence of arterial involvement and limited or no venous involvement, allowing for upfront surgical resection with a high likelihood of achieving negative margins. Borderline resectable PDAC involves limited contact with major mesenteric vessels, where resection is technically feasible but associated with a higher risk of positive margins. Locally advanced PDAC is defined by extensive arterial and/or venous involvement precluding immediate surgical resection. These distinctions are clinically relevant, as they influence treatment strategy, prognosis, and the potential role of neoadjuvant therapy (6). According to recent European Society for Medical Oncology (ESMO) Clinical Practice Guidelines, upfront surgery followed by adjuvant chemotherapy, preferably modified FOLFIRINOX, remains the standard approach for clearly resectable tumors. Nevertheless, even after curative-intent resection and adjuvant therapy, recurrence rates remain high, reflecting early systemic dissemination in a substantial proportion of patients. In contrast, for patients with borderline resectable and locally advanced PDAC, neoadjuvant treatment strategies have gained increasing acceptance. In these settings, neoadjuvant chemotherapy (NAC) aims to improve systemic disease control, increase the likelihood of margin-negative resection, and allow biological selection by identifying patients who develop early disease progression and are unlikely to benefit from surgery (7). Multiagent regimens such as FOLFIRINOX have demonstrated promising activity in borderline resectable and locally advanced PDAC, with meaningful rates of downstaging and secondary resection, although no major improvements in survival have been shown so far (5). In recent years, interest has also grown in extending neoadjuvant strategies to selected patients with resectable PDAC, as illustrated by trials such as PREOPANC and ESPAC-5 (8,9). However, neoadjuvant FOLFIRINOX is associated with substantial toxicity, including gastrointestinal side effects, peripheral neuropathy, and myelosuppression, which may lead to a decline in performance status. In addition, the need for pre-treatment biopsy confirmation and biliary drainage may introduce delays and procedure-related complications (5,10). The feasibility and safety of neoadjuvant FOLFIRINOX remain a matter of debate. The recently published NORPACT-1 trial, which compared neoadjuvant FOLFIRINOX with upfront surgery in resectable PDAC, did not demonstrate a survival benefit and reported high rates of serious adverse events (SAEs) and treatment discontinuation: 57% of patients in the neoadjuvant group had at least one grade 3 or worse adverse event, only 52% of patients was able to complete all four cycles of NAC and dose reduction during neoadjuvant FOLFIRINOX was necessary in 54% of patients (11). These findings underscore the importance of real-world data to better understand treatment tolerability, dose intensity, and the achievable duration of neoadjuvant therapy in routine clinical practice. In this descriptive retrospective monocentric study, we analyzed real-world data from patients with borderline resectable or locally advanced PDAC treated with upfront neoadjuvant (modified) FOLFIRINOX from 01-01-2020 until 31-06-2023. Our primary objective was to descriptively assess the real-world feasibility of neoadjuvant (m)FOLFIRINOX, defined as the number of cycles that could be delivered prior to treatment discontinuation, and to characterize real-world toxicity and dose intensity, with the aim of informing the design of future interventional trials. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0041/rc). Methods Study design and patients This descriptive, single-center retrospective study included all patients with newly diagnosed non-metastatic PDAC, discussed at the gastrointestinal oncology multidisciplinary team meeting at University Hospitals Leuven between 01-01-2020 and 30-06-2023. Patients were planned for neoadjuvant FOLFIRINOX or modified FOLFIRINOX. Patients with metastatic disease at diagnosis, as well as patients with upfront resectable PDAC were excluded. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee Research UZ / KU Leuven (No. S69000). Given the retrospective nature of the study, the requirement for informed consent was waived by the ethics committee. NAC Eligible patients were started on FOLFIRINOX (oxaliplatin 85 mg/m2 irinotecan 180 mg/m2, leucovorin 400 mg/m2 and 5-fluorouracil (5-FU) 400 mg/m2 administered by intravenous bolus followed by 2,400 mg/m2 given as a 46-hour continuous infusion) or modified FOLFIRINOX (oxaliplatin 85 mg/m2, irinotecan 150 mg/m2, leucovorin 400 mg/m2 and 5-FU 2,400 mg/m2 given as a 46-hour continuous infusion). At the time of treatment, modified FOLFIRINOX was increasingly used in routine clinical practice, although its use and definition were not yet fully standardized. Choice of either regimen was mostly at the discretion of the treating oncologist based on patient characteristics. Treatment was administered per standard protocol biweekly in cycles of four, after which oncological re-evaluation occurred to decide on continuation of treatment or proceeding to surgery. The duration of NAC was not strictly predefined. While treatment duration of 3–4 months is commonly reported, chemotherapy was continued beyond this period in selected patients in the absence of disease progression and based on tolerability, reflecting a total neoadjuvant treatment approach. Dose adjustments and prescription of supportive care were at the discretion of the treating oncologist. Primary growth factor support prophylaxis was used in 100% of patients, according to institutional practice. Data collection A data query was performed from the electronic health record. The data were extracted from patient records and collected in a study specific secured Excel file. The analyzed patient and disease characteristics included date of birth, gender, localization and staging of PDAC (using TNM staging system), local extension according to the NCCN staging system (upfront resectable, borderline resectable or locally advanced) (12), ECOG performance status scale, CA19-9 and bilirubin levels at the start of chemotherapy, relevant comorbidities (e.g., diabetes mellitus, cardiovascular disease, chronic obstructive pulmonary disease) and definitive pathology staging. Furthermore, dosing of NAC, occurrence of SAEs, reasons for hospitalization, need for dose reduction (component and dose) and reason for stopping neoadjuvant treatment (preplanned, toxicity, disease progression, death) were assessed. SAEs were graded using the Common Terminology Criteria for Adverse Events (CTCAE)-version 4.0.3 (13). According to institutional practice, all patients were screened for DPYD deficiency prior to starting chemotherapy, and the dose of 5-FU was adjusted according to the identified variant. UGT1A1 deficiency was only assessed in case of severe gastrointestinal toxicity. These screening practices are relevant as they can significantly impact the toxicity profile observed during treatment. These descriptive data were gathered and compared with available evidence in literature. Statistical analysis Statistical analysis included descriptive analyses only: percentage of patients receiving 1,2,3,… up to 12 cycles of (m)FOLFIRINOX, percentage of patients receiving dose reductions and thereby the relative dose intensity, percentage of patients experiencing serious adverse reactions leading to hospitalization or death. Feasibility was assessed descriptively by evaluating the number of administered cycles prior to treatment discontinuation, reasons for treatment discontinuation, relative dose intensity, and the need for dose modifications or hospitalization. No predefined feasibility threshold was applied, in line with the exploratory nature of this real-world study. Given the descriptive nature of this study, a formal power calculation was not esteemed necessary. Missing data were handled using a complete-case approach. Missing values were not imputed and were excluded from the calculation of medians, percentages, and other descriptive statistics. For each analysis, only available data were included. Results Patient characteristics Patient demographics and disease characteristics at baseline were summarized in Table 1. A total of 81 patients were included in the study, with an equal distribution between male and female patients (51.9% vs. 48.1%). The median age was 65 years [interquartile range (IQR), 57 to 71 years], and most patients had an ECOG performance status of 0 (n=55, 67.9%), with the remainder having a status of 1 (n=26, 32.1%). Common comorbidities included cardiovascular disease and diabetes mellitus, each present in 16 patients (19.8%), while COPD was rare (n=1, 1.2%). The median BMI was 24 kg/m2 (IQR, 21–28 kg/m2), median CA19-9 level was 130 U/mL (IQR, 22–675 U/mL), and median bilirubin at the start of the chemotherapy was 0.59 mg/dL (IQR, 0.39–1.2 mg/dL). Five patients (6.2%) were DPYD heterozygous, with the remainder being wild type (n=76, 93.8%). Only two patients were UGT1A1 heterozygous (2.5%), but most of the patients were not tested (n=74, 91.4%). Tumors were most frequently located in the head of the pancreas (n=50, 61.7%), followed by the body (n=25, 30.9%) and tail (n=6, 7.4%). In terms of cancer staging, 50 patients (61.7%) had borderline resectable PDAC, while 31 patients (38.3%) had locally advanced disease. Regarding treatment, 49 patients (60.5%) received mFOLFIRINOX, and 32 (39.5%) received FOLFIRINOX. Table 1Patient demographics and disease characteristicsVariableData, n=81 Sex Male39 (48.1) Female42 (51.9)Age, years65 [57–71]ECOG 055 (67.9) 126 (32.1)Comorbidity Cardiovascular16 (19.8) COPD1 (1.2) Diabetes mellitus16 (19.8)BMI, kg/m224 [21–28]CA19-9, U/mL130 [22–675]Bilirubin, mg/dL0.59 [0.39–1.2]DPYD Wild type76 (93.8) Heterozygous5 (6.2)UGT1A1 Not known/tested74 (91.4) Wild type5 (6.2) Heterozygous2 (2.5)Primary tumor location Head50 (61.7) Body25 (30.9) Tail6 (7.4)Type of pancreatic cancer Borderline resectable50 (61.7) Locally advanced31 (38.3)Therapy received mFOLFIRINOX49 (60.5) FOLFIRINOX32 (39.5)Data are presented as n (%) or median [IQR]. Percentages might not sum to 100 as a result of rounding. BMI, body mass index; CA19-9, carbohydrate antigen 19-9; COPD, chronic obstructive pulmonary disease; CTCAE, Common Terminology Criteria for Adverse Events; ECOG, Eastern Cooperative Oncology Group; IQR, interquartile range. Patient demographics and disease characteristics Data are presented as n (%) or median [IQR]. Percentages might not sum to 100 as a result of rounding. BMI, body mass index; CA19-9, carbohydrate antigen 19-9; COPD, chronic obstructive pulmonary disease; CTCAE, Common Terminology Criteria for Adverse Events; ECOG, Eastern Cooperative Oncology Group; IQR, interquartile range. NAC-treatment Consecutive cycles of (m)FOLFIRINOX and reasons for treatment discontinuation were summarized in Figure 1 and Table 2. A total of 81 patients initiated neoadjuvant therapy, with the full cohort completing the first three cycles. After cycle three, one patient discontinued due to toxicity and one patient had progressive disease. 79 patients (97.5%) were able to continue up to four cycles and a planned restaging, whereafter a more substantial drop occurred, as 27 patients (33.3%) discontinued therapy. Most of them stopped due to planned surgery (n=15, 18.5%) or progressive or persistently inoperable disease after restaging (n=10, 12.3%), only two patients (2.5%) had to stop therapy due to toxicity. Since guidelines advise restaging every two months/four cycles, another significant discontinuation occurred before cycle nine, with 23 patients (28.4%) stopping therapy (n=17 or 21% for surgery, n=6 or 7.4% for progression). From cycle nine onward, discontinuations were exclusively due to progressive or persistently inoperable disease. By the end of cycle 12, only six patients (7.4%) remained on therapy. The median number of cycles of chemotherapy administered was six. Flowchart of consecutive cycles of (m)FOLFIRINOX and reasons for stopping. Table 2Neoadjuvant (m)FOLFIRINOX treatment-related variablesVariableData, n=81Number of (m)FOLFIRINOX cycles 32 (2.5) 427 (33.3) 58 (9.9) 67 (8.6) 71 (1.2) 823 (28.4) 91 (1.2) 103 (3.7) 113 (3.7) 126 (7.4)Median number of cycles6Reasons for stopping (m)FOLFIRINOX Toxicity5 (6.2) Surgery45 (55.5)Persistently inoperable/progressive disease25 (30.9)Data are presented as n (%) unless otherwise stated. Percentages might not sum to 100 as a result of rounding. Neoadjuvant (m)FOLFIRINOX treatment-related variables Data are presented as n (%) unless otherwise stated. Percentages might not sum to 100 as a result of rounding. Overall, the most common reasons for discontinuation were surgical resection (n=45, 55.5%) and disease progression (n=25, 30.9%). Among these last group, 14 patients (17.3%) had locally advanced disease and 11 (13.6%) had borderline resectable disease at baseline. There is no clear signal that extended treatment duration beyond eight cycles increased the risk of progression. Only five patients (6.2%) had to discontinue therapy due to intolerable toxicity. Subgroups were too small to identify a difference in drop out due to toxicity between patients treated with FOLFIRINOX vs. mFOLFIRINOX (n=2 vs. n=3). Nevertheless, regarding future interventional trials, a regimen with four months (eight cycles) of NAC appears feasible for the vast majority of patients. Figure 2 shows the relative dose intensity per cycle per cytotoxic agent, calculated as the ratio of the actual dose administered to the standard dose during treatment. As expected, dose intensities started high in early cycles but gradually decreased over time. Irinotecan showed the steepest decline, oxaliplatin maintained the highest relative intensity across cycles. This trend likely reflects a cumulative toxicity over the course of therapy, and a decreasing tolerability due to physical deconditioning from neoadjuvant treatment or disease progression. Nevertheless, dose intensity for the total cohort never fell below 79%. Relative dose intensity per cycle per product and comparison between surgery and non-surgery group. 5-FU, 5-fluorouracil. When we compare the mean dose intensity of patients who were able to undergo surgery after NAC with the non-surgery group, no difference was seen in 5-FU (mean of 90% in both groups) or oxaliplatin (94% vs. 96%), and a 9% difference in advantage of the surgery group was seen in irinotecan (89% vs. 80%). Adverse events and dose reduction Table 3 summarizes the reasons for hospitalization due to AEs during NAC. In total, 37 patients (45.7%) needed inpatient treatment, reflecting a substantial rate of treatment-related complications. The main reason for hospitalization was gastrointestinal toxicity, with anorexia (n=9, 11.1%), diarrhea (n=7, 8.6%) and nausea (n=2, 2.5%) being the most common. 4.9% of all patients (n=4) had a documented enteritis on imaging. Furthermore, a relevant number of infectious complications were seen: five patients were treated for cholangitis due to stent dysfunction (6.2%), a smaller portion of patients for pneumonitis (n=2, 2.5%), febrile neutropenia (n=1, 1.2%) and urosepsis (n=1, 1.2%). One patient of 45 years old, ECOG 0, died during NAC due to a severe viral pneumonitis with hypoxic respiratory failure. Table 3Reasons for hospitalization during NACHospitalizations during NACData (%), n=37 (45.7%) Anorexia (CTCAE 3)9 (11.1)Diarrhea (CTCAE 3)7 (8.6)Nausea (CTCAE 3)2 (2.5)Ileus/gastric outlet obstruction (CTCAE 3)2 (2.5)Enteritis CTCAE 21 (1.2) CTCAE 32 (2.5) CTCAE 41 (1.2)Pneumonitis CTCAE 31 (1.2) CTCAE 5†1 (1.2)Cholangitis (stent dysfunction) (CTCAE 3)5 (6.2)Febrile neutropenia (CTCAE 3)1 (1.2)Infectious syndrome (CTCAE 3)1 (1.2)Chills (CTCAE 2)1 (1.2)Urosepsis (CTCAE 3)1 (1.2)Renal insufficiency (CTCAE 3)1 (1.2)Hyperglycemia (CTCAE 3)1 (1.2)†, death due to severe viral pneumonitis with hypoxic respiratory failure. CTCAE, Common Terminology Criteria for Adverse Events; NAC, neoadjuvant chemotherapy. Reasons for hospitalization during NAC †, death due to severe viral pneumonitis with hypoxic respiratory failure. CTCAE, Common Terminology Criteria for Adverse Events; NAC, neoadjuvant chemotherapy. Table 4 shows the reasons for dose reduction. Dose reduction was necessary in 43 patients (53.1%), preceded by treatment delay in 18 patients (22.2%). Again, gastrointestinal toxicity was the most common reason for treatment delay and dose reduction, with diarrhea (n=11, 13.6%), anorexia (n=6, 7.4%) and nausea (n=3, 3.7%) being the most important. Furthermore, hematological toxicity led to dose reductions in 13.6 % of patients (n=11), including thrombocytopenia (n=6, 7.4%), neutropenia (n=4, 4.9%), and anemia (n=1, 1.2%). Neurotoxicity, in the form of polyneuropathy, accounted for 6.2% of dose modifications (n=5). Additionally, general weakness was reported as a reason for dose reduction in 8.6% of patients. Table 4Reasons for dose reductionReasons for dose reductionData (%), n=43 (53.1%) Gastrointestinal toxicity Diarrhea11 (13.6) Anorexia6 (7.4) Nausea3 (3.7)Hematological toxicity Thrombocytopenia6 (7.4) Neutropenia4 (4.9) Anemia1 (1.2)Neurotoxicity Polyneuropathy5 (6.2)General weakness7 (8.6) Reasons for dose reduction Discussion The objective of this study was to descriptively assess the real-world feasibility of neoadjuvant (m)FOLFIRINOX, defined as the number of cycles that could be delivered prior to treatment discontinuation, and to characterize real-world toxicity and dose intensity, with the aim of informing the design of future interventional trials. NAC offers potential benefits, such as controlling early systemic spread, reducing the risk of positive margins after resection, giving an effective dose of chemotherapy prior to surgical intervention and associated complications, and selecting patients with favorable biology by excluding patients who develop metastases on treatment (10). Nevertheless, in previous studies, FOLFIRINOX is only considered suitable for patients with an excellent performance status and few relevant comorbidities, due to high grades of toxicity. In the NORPACT-1, only 52% of patients initiating neoadjuvant FOLFIRINOX completed all four cycles, the other 48% was not able to due to toxicity, decline in performance status, or disease progression (11). In this descriptive retrospective study, we analyzed 81 patients who initiated neoadjuvant therapy with (m)FOLFIRINOX at a single academic hospital. The majority of patients discontinued NAC after a planned restaging at four or eight cycles. The median number of cycles of chemotherapy administered was 6. While neoadjuvant treatment durations of 3–4 months are commonly reported, the optimal number of cycles remains undefined. In our cohort, some patients received extended treatment durations, reflecting a total neoadjuvant approach aiming to maximize systemic disease control. This variability highlights the lack of consensus regarding treatment duration in routine clinical practice. Main reasons for discontinuation of therapy were planned surgery (n=45, 55.6%), followed by persistently inoperable or progressive disease (n=25, 30.9%). Only five patients (6.2%) had to stop because of intolerable toxicity. Progression during neoadjuvant therapy was observed in both borderline resectable and locally advanced disease, with a higher absolute number in locally advanced patients, as expected given their more advanced disease biology. There is no clear signal that extended treatment duration beyond eight cycles increased the risk of progression. These findings support the importance of careful restaging and individualized treatment duration. Dose intensities were calculated and started high in early cycles, but gradually decreased over time, likely reflecting a cumulative toxicity over the course of therapy, and a decreasing tolerability due to physical deconditioning from neoadjuvant treatment or disease progression. Irinotecan showed the steepest decline, oxaliplatin maintained the highest relative intensity across cycles. The observed decline in irinotecan relative dose intensity may partly be explained by variability in starting dose between standard and modified FOLFIRINOX regimens (150–180 mg/m2), as well as the tendency to preferentially reduce irinotecan in response to toxicity in clinical practice. SAEs with need for an inpatient admission were observed in 37 patients (45.7%). Gastrointestinal toxicity was the main reason for hospitalization, followed by infectious problems such as cholangitis due to stent dysfunction. Nevertheless, the majority of patients were successfully managed through this toxicity with supportive therapy and were able to complete the remaining NAC-cycles as planned. One patient died due to complications. Dose reduction was necessary in 43 patients (53.1%), preceded by treatment delay in 18 patients (22.2%). Again, gastrointestinal toxicity was the most common reason for treatment delay and dose reduction, followed by hematological toxicity, neurotoxicity in the form of polyneuropathy and general weakness. A neoadjuvant strategy with FOLFIRINOX is known to be associated with toxicity and treatment-related complications, including biliary problems. The NORPACT-1 trial encountered challenges in regimen completion, with only 52% of patients able to complete all four cycles, and 57% of overall adverse events. The use of full-dose neoadjuvant FOLFIRINOX, and not a modified regimen, might have contributed to increased toxicity and reduced compliance, and may partially explain the negative results of the study. However, we did not observe any differences in drop out due to toxicity between patients treated with FOLFIRINOX vs. mFOLFIRINOX, although subgroups are too small to make definitive conclusions. Several other studies have evaluated the feasibility and tolerability of neoadjuvant FOLFIRINOX in non-metastatic PDAC, providing important context for our real-world findings. In an intent-to-treat analysis by Fong et al., 78.7% of patients completed all eight planned cycles of neoadjuvant FOLFIRINOX, while 21.4% discontinued treatment due to poor tolerability, with a median of four cycles delivered in this subgroup. Although overall toxicity was frequent (85%), grade 3–4 adverse events occurred in 42.9% of patients, and 40.2% required inpatient admission (10). Shorter neoadjuvant regimens appear to be associated with lower toxicity rates. In the ESPAC-5 trial, 79% of patients in the FOLFIRINOX arm completed all four planned cycles, with SAEs reported in 26% of patients and none exceeding grade 2, likely reflecting the limited duration of therapy (9). Data from longer neoadjuvant strategies further support the feasibility of multiagent chemotherapy in selected patients. Murphy et al. reported that 81% of patients were able to complete at least four cycles of neoadjuvant FOLFIRINOX prior to chemoradiation and surgery, with only nine percent of patients assigned to eight cycles discontinuing treatment due to toxicity (14). Similarly, in the PANACHE01-PRODIGE 48 trial in resectable PDAC, the predefined therapeutic sequence of NAC followed by surgery was achieved in 70.8% of patients treated with mFOLFIRINOX, although grade 3–4 adverse events occurred in 47% (15). Together, these findings illustrate that while neoadjuvant FOLFIRINOX is associated with substantial toxicity, the majority of patients are able to complete predefined treatment courses when adequate supportive care and dose modifications are applied. This is consistent with the feasibility, dose-intensity patterns, and low discontinuation rates due to toxicity observed in our real-world cohort. At ESMO 2024, the randomized phase II PANDAS/PRODIGE 44 trial in borderline resectable PDAC demonstrated that the addition of neoadjuvant chemoradiation to a mFOLFIRINOX backbone did not result in improved R0 resection rates or survival compared with mFOLFIRINOX alone (16). In the locally advanced setting, results from the phase III NEOPAN (PRODIGE 29–UCGI 26) trial—currently available in abstract form—suggest an improvement in progression-free survival with FOLFIRINOX compared with gemcitabine, although no significant overall survival benefit has been demonstrated (17). In addition, the Alliance A021501 trial did not show a benefit for the addition of radiotherapy in the neoadjuvant setting, highlighting ongoing uncertainties regarding the optimal treatment strategy (18). Together, these findings reinforce the role of intensive multiagent chemotherapy in advanced non-metastatic PDAC and support guideline-based approaches that prioritize multiagent systemic chemotherapy as the cornerstone of neoadjuvant treatment, without routine incorporation of radiotherapy. Tolerance of FOLFIRINOX may also differ between the neoadjuvant and adjuvant settings. In the adjuvant setting, FOLFIRINOX is administered after major pancreatic surgery, when patients may still be recovering from postoperative complications, nutritional deficiencies, and functional decline, which can negatively impact treatment tolerance. In contrast, NAC is delivered upfront in patients with a better baseline performance status, before surgical stress, potentially allowing for improved tolerance, earlier dose modifications, and proactive supportive care. However, neoadjuvant treatment is also associated with unique challenges, including biliary complications, treatment-related delays, and cumulative toxicity that may affect surgical fitness (5,10). For example, in our cohort, the number of dose reductions explicitly attributed to chemotherapy-induced peripheral neuropathy was relatively low. This observation aligns with neoadjuvant FOLFIRINOX literature in non-metastatic PDAC, where neuropathy-driven dose modifications are infrequent, likely reflecting shorter cumulative oxaliplatin exposure and proactive management strategies (19). By contrast, in the adjuvant setting after pancreatic resection, sensory neuropathy is a more common cause of oxaliplatin discontinuation and dose reduction, with some large cohorts reporting neuropathy-related cessation in the majority of patients receiving modified FOLFIRINOX (20). These contextual differences should be considered when interpreting comparative toxicity profiles and planning perioperative therapy. Our real-world data suggest that, despite a substantial rate of hospitalizations and dose reductions, discontinuation of neoadjuvant (m)FOLFIRINOX due to toxicity was uncommon, supporting the concept that upfront administration of intensive chemotherapy is feasible in carefully selected patients. These observations complement evidence from adjuvant trials, where completion rates of full-dose FOLFIRINOX are often limited, and highlight the need to balance treatment intensity, timing, and patient selection in perioperative strategies. Intensified regimens may contribute to increased treatment-related adverse events, as suggested by findings from recent perioperative trials such as PREOPANC-2 (21). As mFOLFIRINOX is proven to be better tolerated and associated with comparable outcomes compared to FOLFIRINOX in previous studies (22), we think it must be feasible for future clinical trials to use protocols with four months/eight cycles of neoadjuvant mFOLFIRINOX. A multidisciplinary monitoring by an experienced team is important, whereby it is necessary to prospectively identify patients who are suited to tolerate preoperative therapy, and reduce doses when necessary. This design has been adopted by the ongoing PREOPANC-3 randomized-controlled trial which investigates the superiority of eight cycles of neoadjuvant mFOLFIRINOX vs. upfront surgery for resectable PDAC (23), as well as part of a perioperative regimen in the Alliance A021806, a phase III randomized trial comparing perioperative mFOLFIRINOX and surgery vs. upfront surgery followed by adjuvant mFOLFIRINOX for resectable PDAC (24). Our study has several limitations. First, it is a descriptive retrospective study conducted at a single institution, which may introduce selection bias. However, clinical practice and patient selection criteria for neoadjuvant therapy can vary considerably between countries and institutions. In that context, our findings provide valuable insight into what can be achieved in a high-volume, experienced center with close multidisciplinary monitoring. Second, adverse events were only recorded if they required hospitalization, potentially underestimating the incidence of milder toxicities. Additionally, decisions regarding NAC dose reduction or discontinuation were made by the treating physician without predefined criteria. Fourth, both modified FOLFIRINOX and FOLFIRINOX regimens were used and no subanalysis was performed between these groups, which may result in minor differences in tolerability. Another important limitation of this study is the absence of survival analyses. This was a deliberate methodological choice. In a real-world, retrospective feasibility study, survival outcomes would be inherently and strongly biased by patient fitness, treatment tolerance, and treatment completion, with fitter patients both more likely to tolerate NAC and to experience longer survival. As a result, any association between treatment exposure, dose intensity, or number of delivered cycles and survival would be highly susceptible to confounding by indication and immortal time bias, and could easily be misinterpreted as a causal relationship. To overcome these limitations, further large-scale prospective studies are required. Despite these limitations, we believe that the findings of this study offer valuable insights and can inform the design of future interventional studies. Conclusions Based on our study results and comparing our results with evidence in literature, we conclude that, with adequate supportive care and dose modifications when necessary, the majority of patients were able to complete the assigned chemotherapy cycles. Four months (eight cycles) of mFOLFIRINOX appears feasible for the vast majority of patients and may serve as an optimal regimen for evaluation in future interventional trials. Supplementary The article’s supplementary files as Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0041/prf