Optimal Management of Resectable Pancreatic Head Cancer in the Elderly Patient: Does Neoadjuvant Therapy Offer a Survival Benefit?

Abstract

Introduction:

Neoadjuvant therapy (NAT) is a growing strategy for patients with resectable pancreatic cancer (PDAC). Elderly patients are at increased risk of treatment withdrawal due to functional decline, and the benefit of NAT in this cohort remains to be studied. The objective of this study was to compare outcomes of elderly patients with resectable head PDAC who underwent NAT or a surgery first (SF) approach.

Methods:

All patients 75 years and older with radiographically resectable (NCCN criteria) PDAC who underwent pancreaticoduodenectomy (PD) at a single institution from 2008–2017 were analyzed. Baseline characteristics and perioperative outcomes were compared between SF and NAT cohorts. Recurrence free (RFS) and overall survival (OS) were analyzed by treatment strategy.

Results:

158 patients were identified: SF=90 (57%) and NAT=68 (43%). Patients in the SF cohort were older (80 versus 78 years; p=0.01), but there were no differences in preoperative comorbidities or frailty indices. SF patients had a trend toward higher rates of major complications (38% vs 24%, p=0.06) with higher comprehensive complication index totals (CCI 20.9 vs 20, p=0.03). There were similar rates of adjuvant therapy. NAT was associated with significantly longer OS (24.6 vs 17.6 mos, p=0.01) in both the intent to treat and the resected cohorts. On MVA, NAT remained an independent predictor of OS (HR 0.60, p=0.02).

Conclusion:

NAT is safe and effective for elderly patients with PDAC. This study suggests NAT to be associated with fewer complications after surgery, equal rates of adjuvant therapy receipt, and increased overall survival over a surgery first approach.

INTRODUCTION

With increasing incidence and overall aging of the population, pancreatic ductal adenocarcinoma (PDAC) in the elderly presents a looming health problem^{1, 2}. Despite improvement in survival for other malignancies, pancreatic cancer remains a dismal diagnosis with an overall 5-year survival that has only recently reached 10%³. While much of this mortality burden is due to the large portion of patients who present with advanced disease at diagnosis, outcomes still lag even for patients who undergo resection.

Although surgery presents the only chance for cure for the 15–20% of patients with resectable PDAC on presentation, surgery alone is insufficient therapy as the majority of resected patients will experience early disease recurrence and poor survival⁴. Systemic chemotherapy, with 5-fluorouracil or gemcitabine-based regimens, significantly prolongs survival following pancreatic resection and forms the cornerstone of optimal pancreatic cancer care^5–9. Despite this, less than 40% of patients above the age of 70 go on to complete adjuvant chemotherapy following resection, leading to diminished survival^10–13.

Compared to a surgery first (SF) approach, neoadjuvant therapy (NAT) offers the advantage of increasing systemic therapy completion rates and allows for monitoring of disease response or progression, but outcomes in elderly patients undergoing NAT are understudied^{14, 15}. Additionally, since elderly patients are at increased risk for treatment dropout in the absence of disease progression, any benefit of NAT in this vulnerable population must be clearly demonstrated prior to changing practice¹⁶.

Due to the paucity of data on the optimal strategy for treating elderly patients with localized PDAC, the aim of the current study was to analyze a large institutional dataset of elderly PDAC patients (≥75 years old) and compare short and long-term outcomes of a NAT versus SF approach. Our hypothesis was that elderly patients treated with NAT would have higher rates of adverse perioperative outcomes (complications, discharge to skilled nursing facility) but improved overall survival compared to a surgery first approach.

METHODS

Study Design and Population

A retrospective study was performed for all resectable PDAC patients age 75 years and older who underwent PD either as SF or following NAT at our institution between 2008 and 2017. Patients were excluded if they had borderline or locally advanced disease at time of presentation as determined by NCCN radiographic criteria¹⁷. This study was approved by the Institutional Review Board. The decision to purse SF or NAT at our institution is mostly undertaken by a pancreatic cancer multidisciplinary tumor board. Although the decision to use FOLFIRINOX or gemcitabine /abraxane is made on a case-by-case basis, the latter is the preferred regimen for patients > 75 at our institution due to its favorable safety profile.

Variables, Definitions and Outcomes

Demographic, disease-related and treatment characteristics, and perioperative variables were collected for all patients. Demographic variables included age, sex, body mass index (BMI), smoking history, albumin level, and hemoglobin level. Since functional status and frailty are important variables in this elderly cohort, three comorbidity/frailty scores were used: age adjusted Charlson Deyo Comorbidity Index score (AA-CCI)¹⁸, ASA physical status, and modified frailty index score (mFI-5)^{19, 20}. The latter index is a validated tool used to assess patient frailty and predict postoperative complications. It ranges from zero to five points and incorporates five risk factors: hypertension (HTN) requiring medication, history of congestive heart failure (CHF), history of chronic obstructive pulmonary disease (COPD), impaired functional health status (either partially or totally dependent on assistance for activities of daily living), and diabetes mellitus (DM) requiring either a non-insulin agent (NIDDM) or insulin therapy (IDDM)^{19, 20}. The mFI-5 is calculated for each patient by counting the number of risk factors present.

Disease and treatment characteristics included CA 19–9 level at diagnosis, tumor size at diagnosis by CT and or EUS, neoadjuvant regimen type (5-fluorouracil versus gemcitabine based), and number of neoadjuvant cycles. For the purposes of this study, one cycle of chemotherapy is defined as one month of treatment (either two treatments of 5-fluorouracil based chemotherapy on day 1 and 14, or gemcitabine-based triplet or doublets). Resection variables included pathologic stage (AJCC 8^th edition), presence of lymphovascular (LVI) or perineural invasion (PNI), and margin status (AJCC 8^th edition). Post-operative treatment variables included receipt and number of adjuvant chemotherapy (AT) cycles, total number of chemotherapy cycles received (sum of NAT+AT), site of recurrence, and receipt of salvage chemotherapy for recurrence.

Peri-operative factors were surgical approach (open, robotic assisted), vascular resection, use of an enhanced recovery pathway after surgery (ERAS), operative length, intraoperative blood loss (EBL), length of stay, in hospital comprehensive complication index score (CCI)²¹, major complications (Clavien Dindo grade III or higher)²², clinically relevant post-operative pancreatic fistula (CR-POPF) as per ISGPS criteria²³, delayed gastric emptying (DGE)²⁴, reoperation, discharge disposition (home, skilled nursing facility (SNF), or long term acute care hospital (LTAC)), discharge on tube feeds, discharge on total parenteral nutrition (TPN), readmission within 30 days of discharge, and 30 and 90-day mortality.

Statistical Methods

Descriptive statistics were used to compare patient characteristics between cohorts. Continuous data were reported as median with interquartile range (IQR) and compared using Wilcoxon rank sum test. Categorical data were reported as frequencies and percentages and compared using chi-square of Fisher exact test as appropriate. Overall survival (OS) was defined as time from diagnosis to death and recurrence free survival (RFS) as time between surgery and recurrence or death. Post-recurrence survival was defined as time from disease recurrence to death. Kaplan Meier analysis was used to compare survival between treatment cohorts and significance assessed by log rank test. To examine the impact of NAT on OS and RFS, hazard ratios (HR) for NAT were examined using Cox proportional hazard models. Multivariate Cox proportional hazards models examined the impact of NAT a priori adjusting for age, frailty CA 19–9, margin status, AJCC 8th stage, receipt of adjuvant chemotherapy, and other significant variables on univariate analysis. Since a significant number of elderly patients may start NAT but fail to reach resection either due to disease progression or performance status decline, we also performed a sensitivity analysis for overall survival by treatment strategy. This intention to treat analysis included those NAT patients that were started on therapy but dropped out either due to disease progression or functional decline. Only variables at a 0.05 level of statistical significance were incorporated in the final multivariate regression analysis. An alpha cut off of 0.05 was used for all significance tests. The data were analyzed using STATA 15 (StataCorp LP, College Station, TX, USA).

RESULTS

Patient Demographics and Treatment Characteristics

A total of 158 elderly patients underwent PD for resectable pancreatic cancer during the study period: SF=90 (57%) and NAT=68 (43%). Baseline characteristics are displayed in Table 1. The SF cohort was slightly older at time of diagnosis (80 vs 78 years, p=0.01), however, there were no significant differences in preoperative comorbidity burden as assessed by Charlson-Deyo comorbidity index, physical status as assessed by ASA, or frailty as detected by modified Frailty Index. Both cohorts had similar BMI, albumin, and hemoglobin levels. Additionally, CA 19–9 levels and clinical tumor size were similar on presentation.

Table 1:

Baseline Characteristics of Elderly (≥75 years) PDAC Patients Treated by Surgery First (SF) or Neoadjuvant Therapy (NAT) approach

Variable	Surgery First	Neoadjuvant	p Value
	n=90	n=68
Age, years	80 (77–83)	78 (76–80)	0.01
Male	39 (43%)	21 (31%)	0.11
BMI, kg/m2	25.6 (22.9–28.7)	24.7 (21.3–27.1)	0.16
Smoking History	34 (38%)	21 (31%	0.4
DM			0.9
No	55 (61%)	42 (61%)
Yes: Oral medications	21 (23%)	14 (21%)
Yes: Insulin dependent	14 (16%)	12 (18%)
AA-CCI	6 (6–7)	6 (5–7)	0.12
mFI			0.34
0	12 (13%)	11 (16%)
1	39 (43%)	28 (41%)
2+	39 (44%)	29 (43%)
ASA Status			0.99
2	2 (2%)	2 (3%)
3	75 (84%)	57 (84%)
4	13 (14%)	9 (13%)
Tumor Size (CT), cm	2.35 (1.7–3.15)	2.45 (2.15–4.0)	0.43
CA 19-9, u/mL	143 (32–422)	142 (46–472)	0.69
Albumin, g/dL	3.5 (3.0–4.0)	3.4 (3.2–3.8)	0.67
Hemoglobin, g/dL	12.4 (11–13.4)	12.2 (11–13.3)	0.64
Clinical trial participation	1 (1%)	22 (32%)	<0.001
Neoadjuvant Treatment
Gemcitabine	-	64 (94%)	-
5-Fluorouracil	-	1 (2%)	-
Both	-	3 (4%)	-
Number of Cycles	-	3 (2–4)	-
Radiation therapy	-	10 (15%)	-
Post-therapy CA 19-9	-	71.9 (31.8–200.2)	-
CA 19-9 Normalized		18 (27%)
CA 19-9 Dropped >50%		28 (41%)

Abbreviations: BMI, Body Mass Index; DM, Diabetes Mellitus; AA-CCI, Age Adjusted Charlson Comorbidity Index; mFI, modified Frailty Index; ASA status, American Society of Anesthesiologists; CA 19–9, Cancer Antigen 19–9.

All values are depicted as median (IQR) or n (%).

Among patients who underwent NAT, a majority (n=64, 94%) were treated with gemcitabine-based therapy with a median of 3 cycles (IQR 2–4). Only 1 patient (2%) received 5-fluorouracil based therapy alone while 3 patients (4%) received a combination of both gemcitabine and 5-fluorouracil based therapy. The majority of patients underwent chemotherapy alone, with 10% receiving additional radiation.

Peri-Operative Outcomes of SF and NAT Cohorts

Operative, oncologic, and postoperative outcomes are listed by treatment strategy in Table 2. There were no differences in rates of operative approach (open or robotic), vascular resection, operative time, blood loss, and conversion to open (all p>0.05). There were higher rates of ERAS management in the NAT cohort (35% vs 13%, p=0.01).

Table 2:

Perioperative Outcomes of Elderly (≥75 years) PDA Patients Treated by Surgery First (SF) or Neoadjuvant Therapy (NAT) approach

Variable	Surgery First	Neoadjuvant	p Value
	n=90	n=68
Operative Features
Approach			0.79
Open	49 (55%)	34 (50%)
Robotic	40 (44%)	34 (50%)
Vascular Resection			0.98
Artery	2 (2%)	1 (1%)
Vein	16 (18%)	13 (19%)
Both	2 (2%)	2 (3%)
Converted	10 (25%)	3 (9%)	0.06
Operative length, hours	6.7 (5.4–8.1)	6.1 (5.4–7.2)	0.16
EBL, mL	375 (200–700)	250 (150–600)	0.14
ERAS Protocol	12 (30.8%)	24 (52.2%)	0.04
Pathologic Outcomes
LVI	76 (87%)	51 (77%)	0.1
PNI	85 (94%)	56 (82%)	0.02
Margin Positive	26 (29%)	24 (34%)	0.39
Stage, AJCC 8th			0.04
0	-	2 (3%)
IA	5 (5%)	13 (19%)
IB	15 (17%)	6 (9%)
IIA	-	-
IIB	37 (41%)	23 (34%)
III	32 (36%)	24 (35%)
IV	1 (1%)	-
Post-Operative Outcomes
Length of stay, days	11 (8–16)	9 (6–13)	0.01
Major Complications	34 (38%)	16 (24%)	0.06
CCI Total	20.9 (8.7–39.5)	20 (0–26.9)	0.03
CR-POPF	10 (11%)	2 (3%)	0.05
DGE	25 (28%)	17 (25%)	0.7
PSA	2 (2%)	-	0.32
Reoperation	5 (6%)	4 (6%)	0.6
Disposition			0.01
Home	40 (45%)	48 (71%)
SNF	48 (53%)	19 (28%)
LTAC	2 (2%)	1 (1%)
Discharge on tube feeds	20 (22%)	12 (18%)	0.31
Discharge on TPN	6 (7%)	2 (3%)	0.25
30 Day Readmission	41 (47%)	22 (33%)	0.09
30 Day Mortality	3 (3%)	1 (2%)	0.42
90 Day Mortality	9 (10%)	2 (2.9%)	0.08
Open	6 (12%)	2 (6%)
Robotic	3 (8%)	-
Systemic Therapy and Recurrence
Receipt of Adjuvant Therapy	43 (48%)	34 (51%)	0.71
Days from OR to chemotherapy	66 (54–84)	63 (52–82)	0.55
Total number of chemotherapy cycles (NAT+AT)	5 (3–6)	6 (6–8)	<0.001
Recurrence	49 (58%)	38 (56%)	0.76
Site of Recurrence
Local	28 (57%)	17 (45%)
Distant	21 (43%)	21 (55%)
Salvage Chemotherapy	20 (41%)	21 (55%)	0.13

Abbreviations: EBL, Estimated Blood Loss; ERAS, Enhanced Recovery After Surgery; LVI, Lymphovascular Invasion; PNI, Perineural Invasion; AJCC, American Joint Committee on Cancer; CCI total, Comprehensive Complication Index Score total; CR-POPF, Clinically Relevant Postoperative Pancreatic Fistula; DGE, delayed gastric emptying; PSA, pseudoaneurysm; SNF, Skilled Nursing Facility; LTAC, Long Term Acute Care facility; TPN, Total Parenteral Nutrition.

All values are depicted as median (IQR) or n (%).

Patients treated with NAT had improved pathologic outcomes in the resected specimen. In addition to having a greater proportion of lower stage tumors (p=0.04), NAT patients had lower rates of PNI (82% vs 94%, p=0.02) and LVI (77% vs 87%, p=0.10). Notably, there were no differences in R1 resection rates in the NAT versus SF cohorts (34% vs 29%, p=0.39).

Immediate perioperative outcomes were worse in the SF cohort, with longer hospitalizations (11 vs 9 days, p=0.01), higher overall morbidity burden (CCI total: 20.9 vs 20.0, p=0.03), and a trend towards higher rates of major complications (38% vs 24%, p=0.06). Regarding PD specific outcomes, the SF cohort had higher rates of clinically relevant post-operative pancreatic fistula (11% vs 3%, p=0.05) but similar rates of delayed gastric emptying and reoperation. There was no difference in rates of discharge on tube feeds or TPN, however SF patients had lower rates of discharge to home (45% vs 71%, p=0.01). There was no difference in rates of 30-day readmission, 30-day (3% vs 2%, p=0.42) or 90-day (10% vs 2.9% p=0.08) mortality between treatment groups.

Regarding post-operative treatment, both groups had comparable rates of adjuvant therapy receipt (48% vs 51%, p=0.71) and median time to adjuvant therapy (66 vs 63 days, p=0.55). Overall, the NAT cohort received more total cycles of chemotherapy (median 6 vs 5, p<0.001). There was no difference in rate of recurrence or site of recurrence. Following recurrence, both groups were comparable in receipt of salvage chemotherapy at (SF: 41% vs NAT: 55%, p=0.13).

Survival Outcomes for Resected Patients by Treatment Strategy

NAT patients had longer survival compared to the SF cohort (Figure 1). Median overall survival from diagnosis was 24.6 months (95% CI 16.7–29.1) for NAT versus 17.6 months (95% CI 14.2–20.3) for the SF cohort (p=0.01). There was no difference in recurrence free survival (12.6 versus 11.5 months, p=0.18) or post-recurrence survival (6.4 versus 7.0 months, p=0.70) between NAT and SF respectively.

Figure 1:

Overall survival from time of diagnosis for (A) resected patients (B) for intent to treat analysis by treatment strategy: neoadjuvant (NAT, red) vs. surgery first (SF, blue).

Univariate analysis demonstrated that NAT, age, CA 19–9 at diagnosis, baseline albumin, margin status, tumor stage, complications, discharge disposition, and adjuvant chemotherapy were associated with survival. On multivariate Cox proportional hazards analysis, after adjusting for age, frailty, tumor stage, margin status, CA19–9 level at diagnosis, and receipt of adjuvant chemotherapy, NAT was independently associated with survival (HR: 0.60, p=0.02) along with adjuvant chemotherapy and CA 19–9 at diagnosis (Table 3).

Table 3:

Multivariate Analysis of Factors Associated with Overall Survival in Elderly (≥75 years) PDA Resected Cohort

Variable	OR	95% CI	P Value
Neoadjuvant Chemotherapy	0.60	0.40–0.91	0.02
Adjuvant Chemotherapy	0.42	0.28–0.61	<0.001
Age, years	1.02	0.96–1.08	0.49
Frailty (mFI ≥2)	1.20	0.82–1.75	0.35
CA 19-9 at Diagnosis	1.01	1.00–1.01	0.01
AJCC 8th Stage (resected specimen)			0.27
IA	ref
IB	1.13	0.51–2.53
IIA	2.04	1.01–4.10
III	2.17	1.08–4.36
IV	1.04	0.12–8.89
Positive Margin	1.40	0.94–2.07	0.10

Abbreviations: mFI, modified Frailty Index; CA 19–9, Cancer Antigen 19–9; AJCC, American Joint Committee on Cancer.

Survival Outcomes: Intention to Treat Sensitivity Analysis

Among elderly patients with resectable pancreatic head tumors treated with a NAT strategy, 6 (8.8%) patients failed to undergo resection: 3 due to functional decline in the absence of disease progression and 3 to disease progression (during NAT or at time of surgery). On intent to treat, NAT was still associated with improved survival with median OS of 23.3 months (95% CI 16.7–28.5) compared to 17.6 months for the SF cohort (95% CI 14.2–20.3) (p=0.01). On multivariate analysis, after adjusting for age, frailty, and CA 19–9 level at diagnosis, NAT trended towards significance with a HR of 0.71 (95% CI 0.49–1.03, p=0.07).

DISCUSSION

This study demonstrates that NAT may offer significant survival benefit without undue risk for older patients with resectable pancreatic head cancer. Elderly patients receiving NAT had fewer complications after PD, were equally likely to receive adjuvant therapy, and displayed improved overall survival compared to a surgery first approach.

The burden of cancer in the elderly is expected to increase markedly in the near future. By 2030, elderly patients are anticipated to account for 70% of all new cancer diagnoses²⁵. Aging of the U.S. population will be associated with a corresponding increase in pancreatic cancer cases in the coming decades². Despite this looming crisis, there is limited comparative effectiveness data driving treatment decisions in elderly patients with pancreatic cancer. Elderly patients are historically underrepresented in pancreatic cancer clinical trials²⁶. Though they account for a majority of pancreatic cancer diagnoses, patients over 65 years constitute just 36% of cancer clinical trials^{3, 27}. Thus, there is an absence of high-quality evidence to guide pancreatic cancer treatment decisions for this vulnerable population.

A small number of studies have explored outcomes for elderly patients undergoing neoadjuvant chemotherapy for pancreatic cancer. In 2014, Cooper et al reported their institutional experience with neoadjuvant chemoradiation in resectable and borderline resectable patients above the age of 70¹⁴. In their study of 153 patients who underwent NAT, only 74 (48%) patients completed treatment and underwent pancreatectomy. Major complication rates were similar (33% following PD) but discharge to home and readmission rates were lower than the current study; 91% and 16% respectively. Neoadjuvant therapy was associated with increased median overall survival among patients who underwent resection (33.8 vs 15.1 mos, p=0.001), however this advantage was lost when the intent to treat cohort was analyzed (15.1 vs 16.6, P=0.53). In their analysis, 85% of patients over 70 completed multimodal therapy with a reported survival benefit suggesting a potential role for chemoradiation in the elderly patient. Miura et al evaluated outcomes of neoadjuvant chemoradiation +/− chemotherapy within their institutional cohort of patients with borderline resectable or resectable pancreatic cancer²⁸. Comparing outcomes between elderly (≥ 75 years) and non-elderly patients, 24 of the 36 (67%) of elderly patients studied underwent resection following NAT. Within this cohort, 8% of patients had a major postoperative complication and 28% were discharged to a nursing facility. In their analysis, 13% of elderly patients were readmitted within 30 days of discharge. Median OS was reported as 27.2 months for the elderly cohort who completed all therapy (NAT and surgery). The authors concluded that a majority of elderly patients were able to complete all intended therapy and that NAT was a feasible strategy in older adults.

Our current study confirms and builds on these findings. To our knowledge, this is the first analysis to specifically examine the impact of NAT in a resectable-only cohort of patients with pancreatic cancer. As NAT becomes the standard of care for borderline resectable patients, its benefit is thought to extend to resectable cases, since even those patients have a dismal prognosis due to occult systemic disease at diagnosis. Another potential benefit is sparing of patients with aggressive tumor biology the morbidity of unnecessary surgery, a benefit of utmost importance for the elderly population, in whom recovery can be protracted and quality of life significantly impacted by surgery. Our study suggests NAT to be safe and beneficial in this population.

This study demonstrate that NAT is associated with lower rates of overall and pancreatectomy specific complications and decreased length of stay compared to a SF approach, as shown by other studies²⁹. While the SF cohort was slightly older, there were no significant differences in preoperative comorbidity or frailty indices between the two groups. Thus, small differences in age may not fully account for the disparate postoperative outcomes. A likely explanation maybe the higher rates of clinically relevant postoperative pancreatic fistula in the SF cohort (11% vs 3%). CR-POPF is a root cause of poor outcomes following PD, and rates of CR-POPF have been shown to be lower following the administration of NAT³⁰. Our study confirms this. Furthermore, higher rates of CR-POPF and longer hospital stay may also explain the markedly higher rates of discharge to a site other than home for SF patients in this study (55% vs 29%, p=0.01). Overall, this analysis suggests that NAT is not associated with worse perioperative outcomes following resection among elderly patients with pancreatic cancer.

In addition to being feasible and safe, this analysis suggests a survival benefit to NAT in elderly patients who complete resection. Notably, both cohorts had similar rates of adjuvant chemotherapy receipt. However, the NAT cohort received more total cycles (6 vs 5, p<0.001) over the course of their treatment. Correlative data from ESPAC 3 and other studies suggests that the receipt of 6 cycles of adjuvant therapy is associated with improved survival compared to fewer cycles following resection^{31, 32}. This may account for the longer overall survival seen in the NAT cohort. On multivariate analysis, NAT was an independent predictor of survival even after adjusting for known tumor prognostic factors. This survival advantage persisted when analyzing the intent to treat cohort, making this study the first to document the benefit of NAT in elderly patients with resectable PDA on an intent to treat basis.

There are several limitations to this analysis. First, while the dataset was prospectively collected, this was a single institution retrospective analysis with results that may not be generalizable to other centers. Second, although the cohorts only differed slightly in age, with surrogates for preoperative frailty, comorbidity, and physical status (mFI, CCI, and ASA) being similar, there were likely unmeasured differences contributing to selection bias guiding physician treatment towards NAT versus SF. Third, while patients had similar preoperative tumor size, not all patients underwent a triphasic CT scan or EUS. Additionally, the number of vascular resections observed, while similar in both cohorts (SF: 22%, NAT 23%, p= 0.98), was higher than expected for resectable tumors, and suggests radiographic under-staging. Thus, while the higher rates of early-stage disease on pathology in NAT cohort could be due to treatment effect, it is possible there are unmeasured differences in stage at diagnosis between groups that may be contributing to the observed differences in outcomes. Finally, this study primarily considered outcomes of patients who successfully underwent resection. Although the NAT cohort included patients who failed to reach resection due to disease progression or performance status decline, our institution did not historically track patients who were planned for SF but were found to be unresectable due to disease progression at time of exploration. Therefore, we could not account for SF patients who were found to be unresectable at time of surgery. However, it is likely that this cohort is small and its inclusion would only further strengthen the improved outcomes seen for NAT.

Despite these limitations, the current study is the first to examine outcomes of elderly patients with resectable pancreatic head cancer undergoing NAT versus a SF approach. In this analysis, NAT appears safe, feasible and associated with improved survival. Clinical trials are needed to confirm the benefit of neoadjuvant therapy in this vulnerable population.

Synopsis:

Neoadjuvant therapy in elderly patients (75 years and older) with resectable pancreatic cancer is safe and feasible. Patients who underwent neoadjuvant chemotherapy had fewer complications and improved survival compared to a surgery first cohort.