Diagnostic biopsy-to-treatment interval and risk of recurrence in localized pancreatic ductal adenocarcinoma: every week counts

Introduction

Pancreatic ductal adenocarcinoma (PDAC) is an aggressive malignancy representing the third leading cause of cancer mortality in the US, with projections suggesting it will become the second leading cause by 2030 (1). PDAC accounts for approximately 90% of all pancreatic neoplasms. Across all stages, median overall survival ranges from 4 to 8 months, with a 5-year survival rate of 13% (2). More than half of patients present with metastatic disease at diagnosis, and no validated population-wide screening strategy currently exists (3). Recent advances in surgical technique, systemic therapy, and multidisciplinary care have improved 5-year overall survival from 4% to 13% over the past three decades (4).

Treatment of localized PDAC is multimodal and may include combinations of chemotherapy, surgical resection, and radiation therapy (RT), tailored to disease stage and patient performance status (5). Neoadjuvant or adjuvant systemic regimens, including FOLFIRINOX and gemcitabine plus nab-paclitaxel, are central components of curative-intent therapy (6). For patients with unresectable disease, conventional or hypofractionated RT may be used to facilitate downstaging or as definitive treatment (5). Despite these advances, most patients with initially localized disease ultimately develop local recurrence or distant metastasis, underscoring the need to identify and optimize all modifiable prognostic variables (2).

One underexplored modifiable variable is the interval between diagnostic biopsy and initiation of treatment. In several other aggressive malignancies, delays in treatment initiation have been associated with inferior oncologic outcomes (7-9). Whether a similar relationship exists in localized PDAC has not been well characterized. The present study was designed to evaluate the association between the biopsy-to-treatment interval and the risk of recurrence or distant metastasis in patients with localized PDAC. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0444/rc).

Methods

Study design and patient population

This retrospective cohort study was conducted at Northwell Health, a large tertiary care academic medical center. The study was approved by the institutional review board (IRB) of Northwell Health (No. 24-0285) and individual consent for this retrospective analysis was waived. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. Ninety-one consecutive patients diagnosed with localized PDAC between 2016 and 2024 who received curative-intent treatment were included. Eligible treatments included chemotherapy, RT, and/or surgical resection. Patients were excluded if metastatic disease was present at the time of diagnostic biopsy or if no curative-intent treatment was administered. Pre-eligibility exclusions (patients with localized disease who were screened but did not enter the analytic cohort) could not be enumerated retrospectively from existing databases; this is acknowledged as a limitation.

Data collection and outcome definition

Clinical, demographic, and treatment-related data were extracted from electronic medical records. The primary exposure was the interval (in days) between diagnostic biopsy and initiation of any curative-intent treatment, analyzed as a continuous variable. The primary outcome was time from treatment initiation to recurrence or distant metastasis, defined as any locoregional or distant tumor event identified by imaging, histopathology, or documentation in the multidisciplinary tumor board record. For surgical patients (R0/R1/R2 resection), locoregional recurrence was defined as any new radiographic, histopathologic, or biopsy-confirmed evidence of tumor within the pancreatic bed, peripancreatic lymph node basins, or adjacent operative field. For non-surgical patients (chemotherapy alone or chemoradiation without resection), an event was recorded when there was unequivocal radiographic progression of the primary tumor or peripancreatic disease beyond the treated volume, or when new locoregional disease developed, per multidisciplinary tumor board adjudication. A documented prior radiographic response or stabilization was not required as a precondition for event capture. Distant metastasis was defined for both groups as any new lesion outside the locoregional field documented by imaging, histopathology, or tumor-board record. The following tumor- and patient-level prognostic variables were not consistently captured for the analytic cohort and were therefore not available for analysis: resectability category (resectable, borderline resectable, locally advanced, unresectable), tumor grade, surgical margin status, lymph node involvement, baseline CA 19-9, tumor size, and specific chemotherapy regimen.

Statistical analysis

Descriptive statistics were calculated to characterize the cohort and summarize treatment timelines. Continuous variables were reported as medians with interquartile ranges (IQR); categorical variables were reported as frequencies and percentages.

Time to development of recurrence or distant metastasis was analyzed using standard survival methods. Patients who had not experienced the endpoint at last follow-up were censored at that date. Kaplan-Meier survival curves were constructed for categorical grouping variables and compared using the log-rank test. Follow-up time was summarized using the reverse Kaplan-Meier method, in which the occurrence of the primary endpoint was treated as a censoring event and the end of follow-up was treated as the event of interest; this approach provides an unbiased nonparametric estimate of the distribution of potential follow-up time and appropriately accounts for informative truncation due to early events. The median follow-up and corresponding 95% confidence interval (CI) were derived from the reverse Kaplan-Meier estimator.

Univariate and multivariable Cox proportional hazards (PH) models were used to estimate the association between the biopsy-to-treatment interval and the risk of recurrence or distant metastasis. The multivariable model included five covariates: the biopsy-to-treatment interval (per 7 days), age (per 5 years), sex (female vs. male), RT (yes/no), and surgery (yes/no); radiation and surgery were entered as separate independent binary covariates rather than as a single composite treatment variable. An interaction term between radiation and surgery was evaluated but excluded from the final model due to non-significance. The PH assumption was evaluated for all covariates using scaled Schoenfeld residual tests (individual and global) and visual inspection of Schoenfeld residual plots; no meaningful violations were identified. Complete data were available for all variables included in the analysis (0% missing); no patients were excluded for missing covariate or outcome data. No formal a priori sample size calculation was performed; with 61 events and 5 covariates, the events-per-variable (EPV) ratio was approximately 12.2, meeting the conventional 10-EPV guideline. Results are reported as unadjusted and adjusted hazard ratios (HR) with 95% Wald CIs. Statistical significance was defined as P<0.05. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Patient and treatment characteristics

During the study period, 91 consecutive patients meeting the inclusion criteria were identified and analyzed; no patients were excluded for missing data. Baseline patient and treatment characteristics are summarized in Table 1. Of the 91 patients, 39 (42.9%) were male, and 52 (57.1%) were female. Forty-five patients (49.5%) were known to have died during the follow-up period. Treatment modalities received were as follows: chemotherapy alone in 4 patients (4.4%), chemotherapy plus radiation in 7 (7.7%), chemotherapy plus radiation plus surgery in 33 (36.3%), and chemotherapy plus surgery in 47 (51.6%). RT was omitted in 51 patients (56.0%), and surgery was omitted in 11 patients (12.1%). The median biopsy-to-treatment interval was 27 days (IQR: 21–38; range: 9–87) among patients without recurrence or distant metastasis, and 29 days (IQR: 24–44; range: 13–199) among patients who developed recurrence or distant metastasis.

Recurrence and metastasis

A total of 61 patients (67.0%) developed recurrence or distant metastasis during follow-up. The median time to recurrence or metastasis was 21.7 months (95% CI: 17.9–24.9) (Figure 1). The median follow-up time across all 91 patients, estimated using the reverse Kaplan-Meier method, was 45.1 months (95% CI: 30.0–66.4). There was no statistically significant difference in time to recurrence by sex; median time was 19.2 months in females versus 24.6 months in males (log-rank P=0.15) (Figure 2).

Figure 1 Kaplan-Meier curve showing the time from initiation of curative-intent treatment to recurrence or distant metastasis among 91 patients with localized pancreatic ductal adenocarcinoma. The median time to recurrence or distant metastasis was 21.7 months (95% CI: 17.9–24.9). Tick marks indicate censored observations. Patients at risk are shown below the time axis. CI, confidence interval.

Figure 2 Kaplan-Meier curves of time to recurrence or distant metastasis stratified by sex. Median time to recurrence was 19.2 months (95% CI: 12.7–24.0) in females and 24.6 months (95% CI: 21.6–NE) in males. The difference between groups was not statistically significant by log-rank test (P=0.15). Tick marks indicate censored observations. Patients at risk by sex group are shown below the time axis. CI, confidence interval; HR, hazard ratio; NE, not estimable.

Association between biopsy-to-treatment interval and outcomes

Univariate and multivariable Cox regression results are presented in Table 2. Longer time from biopsy to treatment initiation was significantly associated with increased risk of recurrence or distant metastasis. On univariate analysis, each additional 7-day delay corresponded to a 7% higher hazard (HR, 1.07; 95% CI: 1.01–1.14; P=0.02). This association strengthened after multivariable adjustment for age, sex, radiation, and surgery: each additional 7-day delay was associated with a 9% increase in hazard [adjusted HR (aHR), 1.09; 95% CI: 1.02–1.16; P=0.01]. Age and sex were not independently associated with the primary outcome (Table 2). Surgery was associated with a 1.85-fold increased hazard of recurrence on univariate analysis (HR, 1.85; 95% CI: 0.88–3.93; P=0.11) and reached statistical significance on multivariable analysis (aHR, 2.54; 95% CI: 1.12–5.76; P=0.03), consistent with the established role of surgical resection in curative-intent PDAC treatment. Radiation was not significantly associated with the outcome on either univariate or multivariable analysis (Table 2).

Discussion

This retrospective single-center cohort study identified an association between a longer interval from diagnostic biopsy to initiation of curative-intent treatment and an increased hazard of recurrence and distant metastasis in patients with localized PDAC. To our knowledge, this is among the first studies to specifically examine the biopsy-to-treatment interval—as opposed to imaging-to-surgery or adjuvant treatment timing—as a distinct exposure in this disease context. We emphasize at the outset that the observational design and the absence of several established prognostic covariates require that this finding be interpreted cautiously and as hypothesis-generating.

The statistical findings merit careful interpretation. The aHR of 1.09 per 7-day delay (95% CI: 1.02–1.16; P=0.01) indicates that each additional week of delay was associated with a 9% increase in the hazard of recurrence or metastasis after accounting for age, sex, radiation, and surgery. The association persisted and modestly strengthened on multivariable analysis relative to the univariate estimate (HR, 1.07). Given the aggressive biology of PDAC and propensity for early micrometastatic dissemination, even modest delays in treatment initiation could plausibly allow subclinical disease progression that translates into measurable differences in clinical outcomes; however, an equally plausible explanation is that patients with more complex or advanced disease accrue both longer pre-treatment intervals (for biliary drainage, nutritional optimization, staging workup, or multidisciplinary coordination) and worse outcomes for reasons independent of the delay itself. Our data cannot distinguish between these explanations.

Prior studies examining treatment timing in PDAC have yielded mixed results. A systematic review of 10 studies enrolling 181,344 patients found that prolonged wait times were associated with decreased overall survival in 3 studies, improved survival in 2, and no impact in 5 (10). A systematic review and meta-analysis by Sugumar et al. (11) found that delays of 3 to 5 weeks between resection and adjuvant chemotherapy were associated with a significant increase in the hazard of death (pooled HR, 1.86) and in disease-free survival events, although no significant hazard increase was observed at 6–8- or 9–12-week delay windows; the authors’ overall conclusion was that the evidence regarding the optimal time-to-adjuvant interval remained inconclusive. A study using the US National Cancer Database found no survival benefit associated with earlier surgical resection within 12 weeks of diagnosis in stage I–II PDAC (12), and a single-center study by Jacobsen et al. similarly found no impact of time from imaging to surgery on survival outcomes (13). The heterogeneity across these studies likely reflects differences in the timing endpoints examined (imaging vs. biopsy, surgery vs. any treatment), patient populations, and outcome measures. Our study contributes a distinct perspective by focusing on the biopsy-confirmed diagnosis as the starting point, which may better capture clinically actionable delays in the multidisciplinary workup and treatment planning process. Preoperative biliary drainage and the related timing considerations have also been examined in the context of periampullary surgery (14).

The independent prognostic significance of surgery in our multivariable model (aHR, 2.54; P=0.03) is consistent with established evidence supporting the centrality of surgical resection in curative-intent therapy for localized PDAC and provides internal validity to our analytical approach. Notably, age and sex were not independently associated with the primary outcome in this cohort, suggesting that the biopsy-to-treatment interval is associated with recurrence risk independently of these patient-level characteristics, although unmeasured confounders may operate through paths these variables do not capture.

Strengths and limitations

Strengths of this study include the use of a consecutive single-institution cohort with uniform multidisciplinary care, a clinically actionable exposure definition starting at biopsy-confirmed diagnosis, complete covariate data for the analytic cohort, and adherence to the PH assumption verified by Schoenfeld residual testing.

Several limitations must be prominently acknowledged. First, and most importantly, the analysis did not capture several established independent prognostic variables in PDAC: resectability category (resectable, borderline resectable, locally advanced, unresectable), tumor grade, surgical margin status (R0/R1/R2), lymph node involvement, baseline CA 19-9 level, tumor size, and specific chemotherapy regimen. Because patients with more complex disease typically require longer pre-treatment workups (biliary drainage, nutritional optimization, vascular reconstruction planning), the observed association between the biopsy-to-treatment interval and recurrence risk is at substantial risk of confounding by indication. The reasons for individual delays—whether attributable to patient factors, system-level barriers, or deliberate clinical strategy—were not captured in a structured form and could not be retrospectively adjudicated. The causal interpretation of our findings is therefore limited. Second, the cohort is restricted to patients who initiated curative-intent treatment; patients with localized disease who were screened but did not enter the analytic cohort could not be enumerated retrospectively, which constitutes a form of survivor-inclusion bias. Third, the sample size of 91 patients (61 events) yields an EPV ratio near the conventional 10-EPV threshold for the multivariable Cox model; while this satisfies the standard guideline, the precision of effect estimates is limited and overfitting cannot be entirely excluded. Fourth, the single-institution, retrospective design limits generalizability and introduces the potential for selection bias. Fifth, no formal a priori sample size or power calculation was performed. Prospective multi-institutional studies with larger cohorts, granular capture of resectability and tumor-biology covariates, and pre-specified handling of delay attribution are needed to confirm and extend these findings.

Conclusions

In this single-institution retrospective cohort of patients with localized PDAC, a longer interval from diagnostic biopsy to initiation of curative-intent treatment was associated with an increased hazard of recurrence and distant metastasis on multivariable analysis. These findings suggest a potentially important relationship between time to treatment initiation and oncologic outcomes but should be interpreted cautiously given the retrospective single-center design, the limited sample size, the absence of key prognostic covariates including resectability status, and the possibility of unmeasured confounding (including confounding by indication). Validation in larger, prospective, multi-institutional cohorts that capture these covariates is warranted before any inference about optimal treatment timing is made.

Acknowledgments

The authors thank Noah Siebel for his assistance with data review.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0444/rc

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0444/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0444/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the institutional review board of Northwell Health (No. 24-0285) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bekkali NLH, Oppong KW. Pancreatic ductal adenocarcinoma epidemiology and risk assessment: Could we prevent? Possibility for an early diagnosis. Endosc Ultrasound 2017;6:S58-S61.
2. Stoop TF, Javed AA, Oba A, et al. Pancreatic cancer. Lancet 2025;405:1182-202. [Crossref] [PubMed]
3. Saba H, Goggins M. Familial Pancreatic Cancer. Gastroenterol Clin North Am 2022;51:561-75. [Crossref] [PubMed]
4. Mannucci A, Goel A. Advances in pancreatic cancer early diagnosis, prevention, and treatment: The past, the present, and the future. CA Cancer J Clin 2026;76:e70035. [Crossref] [PubMed]
5. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Pancreatic Adenocarcinoma. Available online: https://www.nccn.org/professionals/physician_gls/pdf/pancreatic.pdf. Accessed March 29, 2026.
6. Damm M, Efremov L, Birnbach B, et al. Efficacy and Safety of Neoadjuvant Gemcitabine Plus Nab-Paclitaxel in Borderline Resectable and Locally Advanced Pancreatic Cancer-A Systematic Review and Meta-Analysis. Cancers (Basel) 2021;13:4326. [Crossref] [PubMed]
7. Yuan M, Behrami E, Pannullo S, et al. The Relationship Between Tumor Volume and Timing of Post-resection Stereotactic Radiosurgery to Maximize Local Control: A Critical Review. Cureus 2019;11:e5762. [Crossref] [PubMed]
8. Roth O'Brien DA, Kaye SM, Poppas PJ, et al. Time to administration of stereotactic radiosurgery to the cavity after surgery for brain metastases: a real-world analysis. J Neurosurg 2021;135:1695-705. [Crossref] [PubMed]
9. Petrusan AM, Feier CVI, Muntean C, et al. The Hidden Cost of Delay: Post-Pandemic Evolution of Advanced Ovarian Cancer Profiles. Medicina (Kaunas) 2026;62:598. [Crossref] [PubMed]
10. Frountzas M, Liatsou E, Schizas D, et al. The impact of surgery delay on survival of resectable pancreatic cancer: A systematic review of observational studies. Surg Oncol 2022;45:101855. [Crossref] [PubMed]
11. Sugumar K, Hue JJ, De La Serna S, et al. The importance of time-to-adjuvant treatment on survival with pancreatic cancer: A systematic review and meta-analysis. Cancer Rep (Hoboken) 2021;4:e1390. [Crossref] [PubMed]
12. Swords DS, Zhang C, Presson AP, et al. Association of time-to-surgery with outcomes in clinical stage I-II pancreatic adenocarcinoma treated with upfront surgery. Surgery 2018;163:753-60. [Crossref] [PubMed]
13. Jacobsen A, Hobbs M, Merkel S, et al. Time to Surgery Does Not Affect Overall or Disease-Free Survival of Patients with Primary Resectable PDAC. J Clin Med 2022;11:4433. [Crossref] [PubMed]
14. Maatouk M, Kbir GH, Ben Dhaou A, et al. Pancreatic surgery after preoperative biliary drainage in periampullary cancers: does timing matter? A systematic review and meta-analysis. HPB (Oxford) 2025;27:10-20.