Futility in adenosquamous pancreatic cancer

Abstract

Background

Adenosquamous carcinoma of the pancreas (ASCP) is a rare and aggressive subtype of pancreatic cancer with a dismal prognosis. Futility in ASCP has been inadequately studied. The aim is to assess the incidence of futility in ASCP cases within a European cohort.

Methods

Retrospective, multicenter European study including all consecutive patients who underwent surgery for ASCP between 2010 and 2024. Inclusion criteria: patients operated for ASCP during the study period. Exclusion criteria: patients without a confirmed pathological diagnosis of ASCP, those who did not undergo surgery, or had extra-pancreatic disease. A pancreatectomy was considered futile if a patient died from postoperative complications within 90 days, or if cancer-related mortality or recurrence occurred within 6 months of the operation.

Results

194 patients from 29 hospitals in 11 European countries were studied. Surgeries included 125 pancreaticoduodenectomies, 59 left pancreatectomies, and 10 total pancreatectomies. Major complications were observed in 25.3% of patients. Postoperative mortality was 5.7%. The rate of futility was 47.9%. Eleven patients (11,8%) died from postoperative complications, 69 patients (74,2%) had recurrence, and 13 patients (14%) died within 6 months due to cancer. In the multivariate analysis, a positive retroperitoneal margin, lymphatic invasion, and not receiving chemotherapy were associated with futility.

Conclusions

The futility observed in ASCP is notably high and related to a positive retroperitoneal margin, lymphatic invasion, and not receiving adjuvant chemotherapy. Defining futile patients with an international consensus definition is crucial and has significant implications for shared decision-making and care optimization.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12957-026-04351-1.

Background

Adenosquamous carcinoma of the pancreas (ASCP) is a rare and aggressive subtype characterized by both glandular and squamous differentiation, with squamous components constituting at least 30% of the tumor. ASCP accounts for approximately 0.5% to 4% of pancreatic malignancies [1–10]. This type of cancer exhibits poor differentiation, is larger, and has more positive lymph nodes compared to pancreatic adenocarcinoma (PDAC) [1, 10–13].

ASCP has a notably inferior prognosis and short survival rates compared to PDAC and other pancreatic tumors [1–3, 5, 7, 9, 11]. Studies indicate that the disease-specific two-year survival rate is approximately 30%, while the five-year survival rate is only 7%, with a median survival time ranging from 7 to 11 months [2, 3, 5, 8, 12]. However, some research suggests that the overall survival rates for patients with ASCP and PDAC may be similar [2, 3, 5, 8, 12].

Surgical management tools typically focus on positive outcomes [14]. However, the measurement of adverse postoperative outcomes is less frequently studied, likely due to these findings’ negative implications for surgeons [15]. One of the most interesting negative tools is futility, which refers to patients for whom surgery does not yield the expected results and could be considered ineffective, contributing significantly to increased healthcare costs. Avoiding surgery in futile patients would not only reduce costs but also alleviate the considerable emotional distress caused by surgical ineffectiveness for both patients and their families [14–18].

Defining what constitutes futile or ineffective surgical treatment is a significant challenge, as this concept involves multiple factors. In the case of cancer patients, futile surgery might be understood as procedures that provide no real benefit, which can occur for several reasons: the intended procedures cannot be completed due to unforeseen findings during surgery, major postoperative complications lead to the patient’s death, or if there is an early recurrence of the disease that results in a very short survival rate. These surgeries fail to deliver the expected benefits, resulting in a low survival rate. In short, these are surgeries that have not provided the expected benefits [15, 18–19].

Futility in pancreatic surgery has been inadequately studied. Only four manuscripts addressed this issue. The studies identified factors such as age over 79, G3 differentiation, American Society of Anesthesiologists (ASA) score, CA 19 − 9 levels, and tumor size associated with surgical futility [18].

Given the lack of futility studies in ASCP, a tumor type known for its poor survival rates and potentially higher futility rates than those observed in PDAC, we propose a study to assess the incidence of futility in ASCP cases within a European cohort. This research aims to enhance our understanding and provide clinical recommendations to improve the management of ASCP patients.

Methods

Patients and study design

This retrospective, multicenter European study included all consecutive patients who underwent elective pancreatic surgery for ASCP at participating centers between January 1, 2010, and June 30, 2024. All patients were discussed at a multidisciplinary pancreatic tumor board meeting before pancreatic resection. The STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines were followed [20].

Inclusion and exclusion criteria

The inclusion criteria encompassed all patients who underwent surgery for ASCP during the study period. The exclusion criteria consisted of patients without a confirmed pathological diagnosis of ASCP, those who did not undergo surgery, and patients with extra-pancreatic disease.

Definitions

The Charlson Comorbidity Index [21] and the ASA classification [22] summarized patients’ comorbidities. Intraoperative complications were categorized according to Satava’s classification [23]. Postoperative complications were assessed and classified using the Clavien-Dindo classification of surgical complications and the Comprehensive Complication Index (CCI^®) [24, 25]. Major complications were defined as Clavien-Dindo grade IIIa or higher. Resection margins, including transection and circumferential margins, were classified as R0 (distance margin to tumor ≥ 1 mm), R1 (distance margin to tumor < 1 mm), or R2 (macroscopically positive margin) according to the Royal College of Pathologists definition [26]. Tumors were staged using the TNM classification [27].

Complications, readmissions, and mortality were recorded up to 90 days postoperatively. Specific complications such as post-pancreatectomy hemorrhage, biliary leakage, delayed gastric emptying, and pancreatic fistula were categorized according to the definitions of the International Study Group of Pancreatic Surgery (ISGPS) [28–31].

A pancreatectomy intended to be curative was considered futile if a patient died from postoperative complications within 90 days, or if cancer-related mortality or recurrence occurred within 6 months of the operation [18].

Data collection

Each participating center appointed a dedicated contact person responsible for communicating with the study coordinator. This person received login codes and passwords for the online electronic case report form (eCRF) environment, specifically REDCap^® (Research Electronic Data Capture, Vanderbilt University, Nashville, Tennessee, USA). Each data collector was provided a separate login account, and the chief study coordinators could monitor all activity.

Ethics

Approval was obtained from the Aragon Research Ethics Committee (CEICA, C.I. PI24/012, 24/01/2024). Due to the aggressiveness of the pathology studied, its poor prognosis, and the retrospective nature of the study, the ethics committee did not consider informed consent necessary. All data were collected anonymously, without patient identifiers. Participating centers were instructed to link the patient’s local medical record numbers to an anonymous study patient ID. Informed consent was waived as there was no associated risk. All actions were documented by Good Clinical Practice (GCP) guidelines.

Statistical analysis

Data were analyzed using R software for MacOS version 5.2.4 (The R Foundation for Statistical Computing, 2004–2025). Statistical tests employed included Student’s t-test, Mann-Whitney U test, Chi-square test, and Fisher’s exact test as appropriate. Categorical data were expressed as frequencies and percentages, while continuous data were presented as either means and standard deviations or medians and interquartile ranges, depending on the distribution. Subgroup analyses were performed to compare characteristics and treatment outcomes, using the Chi-square test, the Mann-Whitney U test, and the Kruskal-Wallis test as appropriate. We evaluated the association between the variables and futile surgery by the univariate logistic regression analysis, the cut-off points were set at p values less than 0.05, using the significant variables identified by the univariate analysis, we then performed a multiple logistic regression analysis to predict futile surgery.

Results

Baseline data

194 patients from 29 hospitals in 11 European countries were included. 57.2% (111) were male. The most common ASA score was ASA II (58.5%). The median Charlson Comorbidity Index was 5 (Interquartile Range [IQR] 3–6). Additionally, 80% of the patients had an ECOG performance score of 0–1, and the median Body Mass Index (BMI) was 24.2 (Table 1).

Table 1.

Comparative characteristics between futile and non-futile groups

	Total 194	No futility 101 (52.1%)	Futility 93 (47.9%)	P value
BASELINE CHARACTERISTICS
Gender, n (%)				0.527
Male	111 (57.2)	60 (59.4)	51 (54.8)
Female	83 (42.8)	41 (40.6)	42 (45.2)
Age, years (IQR)	69 (62–75)	69 (63–74)	70 (60–78)	0.476
Ethnicity, n (%)
Asian  Caucasian  Latino  Other	6 (3.1) 183 (94.3) 2 (1.0) 3 (1.5)	5 (4.9) 92 (91.1) 1 (1.0) 3 (3.0)	1 (1.1) 91 (97.8) 1 (1.1) 0 (0.0)	0.080
ASA score, n (%)
I  II  III  IV	14 (7.3) 113 (58.5) 62 (32.1) 4 (2.1)	4 (4.0) 69 (68.3) 27 (26.7) 1 (1.0)	10 (10.9) 44 (47.8) 35 (38.0) 3 (3.3)	0.019
Surgical history, n (%)
No  Yes	126 (65.3) 67 (34.7)	65 (64.4) 36 (35.6)	61 (66.3) 31 (33.7)	0.776
Liver surgery, n (%)	1 (0.5)	1 (1.0)	0 (0.0)	0.252
Pancreatic surgery, n (%)	1 (0.5)	1 (1.0)	0 (0.0)	0.252
Cholecystectomy, n (%)	12 (6.2)	7 (6.9)	5 (5.4)	0.653
Other supramesocolic surgery, n (%)	13 (6.7)	6 (5.9)	7 (7.6)	0.659
Inframesocolic surgery, n (%)	52 (26.8)	27 (26.5)	25 (27.2)	0.981
Body mass index, kg/m2 (IQR)	24.2 (22–28.4)	25.5 (22.9–29.1)	23.4 (21.7–26.6)	0.021
ECOG, n (%)
0  1  2  3	74 (40.4) 70 (38.3) 29 (15.8) 10 (5.5)	37 (38.5) 38 (39.6) 17 (17.7) 4 (4.2)	37(42.5) 32 (36.8) 12 (13.8) 6 (6.9)	0.719
Charlson Comorbidity Index, median (IQR)	5 (3–6)	5 (3–6)	5 (3–6)	0.975
PREOPERATIVE DATA
Abdominal pain, n (%)
Yes   No	104 (53.6) 90 (46.4)	53 (52.5) 48 (47.5)	51 (54.8) 42 (45.2)	0.742
Jaundice, n (%)
Yes  No	84 (43.3) 110 (56.7)	48 (47.5) 53 (52.5)	57 (61.3) 36 (38.7)	0.216
Asymptomatic, n (%)
Yes  No	16 (8.2) 178 (91.8)	9 (8.9) 92 (91.1)	7 (7.5) 86(92.5)	0.726
Acute cholangitis, n (%)
Yes  No	5 (2.6) 189 (97.4)	1 (1.0) 100 (99)	4 (4.3) 89 (95.7)	0.196
Weight loss, n (%)
Yes  No	62 (32.0) 132 (68.0)	37 (36.6) 64 (63.4)	25 (26.9) 68 (73.1)	0.146
Bilirubin, mg/dL (IQR)	0.8 (0.4–4.4)	0.8 (0.4–4)	0.9 (0.4–4.7)	0.898
Ca 19.9, UI/l (IQR)	71 (19–276)	64 (22–169)	73,5 (17.6–386)	0.372
Hemoglobin, gr/dL (IQR)	12.5 (11.6–13.5)	12.7 (11.8–13.7)	12.3 (10.9–13.3)	0.064
AST, UI/l (IQR)	30 (20–79)	30.5 (20–86.5)	30 (21–70)	0.901
ALT, UI/l (IQR)	36 (19–99)	40.2 (20–103)	32 (18–93)	0.269
Location of the tumor, n (%)
Head  Neck  Body  Tail  Body-tail	125 (64.8) 7 (3.6) 25 (13.0) 36 (18.6) 8 (4.1)	71 (70.3) 3 (3.0) 14 (13.9) 13 (12.9) 0 (0.0)	54 (58.7) 4 (4.3) 11 (12.0) 23 (16.3) 8 (8.7)	0.009
Neoadjuvant chemotherapy, n (%)
Yes  No	44 (22.7) 150 (77.3)	21 (20.8) 80 (79.2)	23 (24.7) 70 (75.3)	0.513
INTRAOPERATIVE DATA
Minimal invasive approach, n (%)
Yes   No	17 (8.8) 177 (91.2)	8 (7.9) 93 (92.1)	9 (9.7) 84 (90.3)	0.666
Laparoscopic Robotic Hybrid	12 (6.2) 3 (1.5) 2 (1.0)	6 (5.9) 1 (1.0) 1 (1.0)	6 (6.5) 2 (2.2) 1 (1.1)	0.883 0.510 0.953
Conversion, n (%)	8 (4.4)	3 (3.2)	5 (5.9)	0.375
Satava, n (%)
1  2  3  4	177 (91.2) 6 (3.1) 10 (5.2) 1 (0.5)	96 (95.0) 2 (2.0) 3 (3.0) 0 (0.0)	80 (87.0) 4 (4.3) 7 (7.6) 1 (1.1)	0.191
Bleeding, n (%)	1 (0.5)	1 (1.0)	0 (0.0)	0.252
Type of resection, n (%)
Pancreatoduodenectomy  Left pancreatectomy  Total pancreatectomy	125 (64.4) 59 (30.4) 10 (5.2)	72 (71.3) 26 (25.7) 3 (3.0)	53 (57.0) 33 (35.5) 7 (7.5)	0.080
POSTOPERATIVE DATA
Length of stay, days (IQR)	13 (9–20)	13 (10–18)	13 (9–23)	0.546
Clavien-Dindo ≥ IIIa, n (%)	49 (25.3)	24 (23.8)	25 (26.9)	0.617
Bile leak, n (%)
No  Grade A  Grade B  Grade C	183 (94.8) 4 (2.1) 3 (1.6) 3 (1.6)	94 (93.1) 3 (3.0) 1 (1.0) 3 (3.0)	89 (96.7) 1 (1.1) 2 (2.2) 0 (0.0)	0.154
Hemorrhage, n (%)
No  Grade A  Grade B  Grade C	169 (87.1) 9 (4.6) 7 (3.6) 9 (4.6)	91 (90.1) 2 (2.0) 6 (5.9) 2 (2.0)	78 (83.9) 7 (7.5) 1(2.2) 7 (7.5)	0.015
Pancreatic fistula, n (%)
No  Biochemical  Grade B  Grade C	146 (75.3) 23 (11.9) 24 (12.4) 1 (0.5)	77 (76.2) 11 (10.9) 12 (11.9) 1 (1.0)	69 (74.2) 12 (12.9) 12 (12.9) 0 (0.0)	0.673
Delayed gastric emptying, n (%)
No  Grade A  Grade B  Grade C	167 (86.5) 15 (7.8) 7 (3.6) 4 (2.1)	87 (86.1) 8 (7.9) 4 (4.0) 2 (2.0)	80 (87.0) 7 (7.6) 3 (3.3) 2 (2.2)	0.994
Intraabdominal abscess, n (%)	24 (12.4)	9 (0.9)	15 (16.3)	0.120
Cardiac arrest, n (%)	7 (3.6)	2 (2.0)	5 (5.4)	0.263
Venous thromboembolism, n (%)	2 (1.0)	1 (1.0)	1(1.1)	0.953
Stroke, n (%)	1 (0.5)	0 (0.0)	1 (1.1)	0.224
Pneumonia, n (%)	4 (2.1)	2 (2.0)	2 (2.2)	0.934
Urinary tract infection, n (%)	4 (2.1)	3 (3.0)	1 (1.1)	0.622
Reintervention, n (%) Endoscopy Interventional radiology Surgery	32 (16.5) 5 (2.6) 12 (6.2) 15 (7.7)	15 (14.9) 3 (3) 5 (5.0) 7 (6.9)	17 (18.3) 2 (2.2) 7 (7.5) 8 (8.6)	0.520 0.718 0.457 0.663
Readmission at 90 days, n (%)	31 (16.0)	14 (13.9)	17 (18.3)	0.401
HISTOLOGY DATA
Margin, n (%)
R0   R1  vR1 vascular   R2	125 (64.4) 53 (27.3) 11 (5.7) 5 (2.6)	77 (76.2) 18 (17.8) 6 (5.9) 0 (0.0)	48 (51.6) 35 (37.6) 5 (5.4) 5 (5.4)	0.001
Ck5 +, n (%)	35 (18.0)	15 (14.9)	20 (21.5)	0.229
Ck6 +, n (%)	32 (16.5)	13 (12.9)	19 (20.4)	0.156
Ck7 +, n (%)	47 (24.2)	21 (20.8)	26 (28.0)	0.245
P63 +, n (%)	25 (12.9)	12 (11.9)	13 (14.0)	0.663
P40 +, n (%)	55 (28.4)	28 (27.7)	27 (29,0)	0.840
MUC1 +, n (%)	6 (3.1)	3 (3.0)	3 (3.2)	0.918
Tumoral size, mm (IQR)	35 (25–47)	31 (20–42)	40 (30–50)	< 0.001
Lymph nodes harvested, n (%)	19 (12–28)	18 (11–26)	20 (13–33)	0.071
Vascular invasion, n (%)	125 (64.4)	53 (52.5)	72 (77.4)	< 0.001
Lymphatic invasion, n (%)	100 (51.5)	38 (37.6)	62 (66.7)	< 0.001
Perineural invasion, n (%)	152 (78.4)	73 (72.3)	79 (84.9)	0.032
No stromal invasion, n (%)	22 (11.3)	17 (16.8)	5 (5.4)	0.012
T, n (%)
T1  T2  T3  T4	22 (11.3) 64 (33.0) 99 (51.0) 9 (4.6)	16 (15.8) 38 (37.6) 44 (43.6) 3 (3.0)	6 (6.5) 26 (28.0) 55 (59.1) 6 (6.5)	0.031
Positive vascular margin, n (%)
Yes  No  Unknown	37 (19.2) 155 (80.3) 1 (0.5)	13 (12.9) 88 (87.1) 0 (0.0)	24 (26.1) 67 (72.8) 1 (1.1)	0.028
Positive retroperitoneal margin, n (%)
Yes  No  Unknown	48 (24.7) 142 (73.2) 4 (2.1)	14 (13.9) 86 (85.1) 1 (1.0)	34 (36.6) 56 (60.2) 3 (3.2)	< 0.001
POSTOPERATIVE TREATMENT
Postoperative chemotherapy, n (%)
Yes   No	128 (66.0) 66 (34.0)	78 (77.2) 23 (22.8)	50 (53.8) 43 (46.2)	0.001
Recurrence, n (%)	119 (61.3)	49 (48.5)	70 (75.3)	< 0.001
Location of recurrence, n (%)
Local  Liver metastases  Carcinomatosis  Other	49 (25.2) 79 (40.7) 34 (17.5) 15 (7.7)	25 (26.8) 25 (24.8) 9 (8.9) 7 (6.9)	21 (23.7) 54 (58.1) 25 (26.9) 8 (8.6)	0.722 < 0.001 0.001 0.790
If recurrence, type of treatment, n (%)
Chemotherapy  Radiotherapy  Surgery	83 (42.8) 13 (6.7) 2 (1.0)	36 (35.6) 7 (6.9) 2 (2.0)	47 (50.5) 6 (6.5) 0 (0.0)	0.036 0.894 0.105

IQR Interquartile range, ASA American Society of Anestesiologist, Eastern Cooperative Oncology Group, AST Aspartate aminotransferase, ALT Alanine aminotransferase

Perioperative outcomes

Intraoperative incidents were reported in only 8.8% of cases. Surgical procedures included 125 pancreaticoduodenectomies, 59 left pancreatectomies, and 10 total pancreatectomies (Table 1). A minimally invasive surgical (MIS) approach was used in 8.8% of patients. The number of lymph nodes harvested was higher in the open approach, 19 (12–29) vs. 14 (12–20) in MIS, but the difference was not statistically significant (p = 0.171) (Table 1). No differences in R status were observed by surgical approach (p = 0.309). Major complications (Clavien-Dindo grade ≥ IIIa) were observed in 25.3% of patients, and reoperation was required in 7.7% of cases. The 90-day readmission rate was 16%. Postoperative 90-day mortality was 5.7%. Eleven patients died, all due to postoperative complications. The specific pancreatic complications in the deceased patients were post-pancreatectomy hemorrhage (4 cases, 3 grade C, 1 grade A), postoperative pancreatic fistula (2 biochemical leaks, 1 grade B), DGE (1 grade B, 1 grade C), and intrabdominal abscess (1). The R0 resection rate was 64.4%, with a median tumor size of 35 mm (IQR 25–47).

Oncological outcomes

Two-thirds of patients received adjuvant chemotherapy. The most common chemotherapy regimen was FOLFIRINOX. Survival rates at 1, 3, and 5 years were 56.2%, 26.3%, and 9.8%, respectively. Meanwhile, disease-free survival rates at the same intervals were 36.6%, 16.5%, and 6.7%, respectively.

119 patients (61.3%) had disease recurrence (Table 1). The most frequent patterns were liver metastasis (40,7%), local (25,2%), and carcinomatosis (17,5%). Liver metastasis and carcinomatosis were more frequent in the futile group.

Futility

The futility rate was 47.9%. Eleven patients (11,8% of futility cases) died from postoperative complications within 90 days; 69 patients (74,2%) had recurrence within 6 months of the operation, and 13 patients (14%) died within 6 months due to cancer. Several factors contributed to this futility, including higher ASA scores and lower BMI. Additionally, tumors in the futile group were more often located outside the head of the pancreas. Postoperative hemorrhage was also more frequent and severe in the futile group. The tumors were larger, had higher T classifications, and were more likely to have positive margins (R1) and to exhibit increased vascular, lymphatic, and perineural invasion. Receiving chemotherapy was a protector for futility.

The relationship between patients with positive margins and prior neoadjuvant chemotherapy was examined in the futility and non-futility groups. There were 37 patients with positive vascular margins. Ten (27%) received neoadjuvant chemotherapy. No statistical significance was observed between the futile and non-futile groups (p = 0.688). There were 48 patients with positive retroperitoneal margins. Thirteen (27.1%) received neoadjuvant chemotherapy. No statistical significance was observed between the futile and non-futile groups (p = 0.728).

In the univariate analysis (Table 2), factors associated with futility included tail location, ASA III-IV, margin positive, N+, vascular, lymphatic, perineural invasion, T stage, and vascular and retroperitoneal margin. In the multivariate analysis, only a positive retroperitoneal margin, lymphatic invasion, and not receiving chemotherapy were associated with futility.

Table 2.

Univariate and multivariate analysis

	Univariate			Multivariate
	OR	CI 95%	P value	OR	IC 95%	P value
Gender
Male	Ref
Female	1.21	0.68–2.13	0.521
Age	1	0.97–1.03	0.979
ASA score
I  II  III  IV	Ref 0.25 0.52 1.2	0.07–0.86 0.14–1.83 0.09–15.26	0.028 0.308 0.888	Ref 0.284 0.495 1.08	0.07–1.16 0.11–2.18 0.07–17.1	0.080 0.354 0.957
Surgical history
No  Yes	Ref 0.92	0.51–1.66	0.777
Body mass index	0.94	0.88–1.01	0.075
ECOG
0  1  2  3	Ref 0.84 0.71 1.5	0.44–1.62 0.29–1.68 0.39–5.76	0.607 0.432 0.555
Charlson Comorbidity Index	0.97	0.87–1.08	0.542
PREOPERATIVE DATA
Abdominal pain
No   Yes	Ref 1.1	0.63–1.94	0.742
Jaundice
No  Yes	Ref 0.69	0.39–1.24	0.216
Asymptomatic
No  Yes	Ref 1.2	0.43–3.37	0.727
Acute cholangitis
No  Yes	Ref 4.49	0.49–40.9	0.183
Weigth loss
No  Yes	Ref 0.64	0.35–1.17	0.147
Bilirrubin	0.98	0.94–1.02	0.292
Haemoglobin	1.00	0.99–1.02	0.617
ALT	0.99	0.99–1.01	0.260
Location
Head  Neck  Body  Tail	0.59 1.56 1.59 2.24	0.32–1.06 0.48–5.11 0.75–3.40 1.05–4.7	0.076 0.460 0.226 0.036	1.43	0.53–3.87	0.482
Neoadjuvant chemotherapy
No  Yes	Ref 1.25	0.64–2.45	0.513
INTRAOPERATIVE
Minimal invasive approach
No   Yes	Ref 1.25	0.46–3.38	0.666
Satava
1   2   3 and 4	Ref 2.4 3.2	0.43–13.44 0.82–12.46	0.319 0.094
Type of resection
Pancreatecoduodenectomy  Left pancreatectomy  Total pancreatectomy	Ref 1.72 3.17	0.92–3.22 0.78–12.83	0.087 0.106
POSTOPERATIVE
Length of stay	0.99	0.99–1.01	0.645
Clavien-Dindo ≥ IIIa
No  Yes	Ref 1.18	0.62–2.25	0.618
Bile leak
No  Yes	Ref 0.60	0.17–2.13	0.433
Hemorrhage
No  Yes	Ref 1.75	0.74–4.12	0.200
Pancreatic fistula
No  Yes	Ref 1.12	0.58–2.14	0.742
Delayed gastric emptying
No  Yes	Ref 1.01	0.45–2.28	0.981
Reintervention
No  Yes	Ref 1.28	0.6–2.74	0.521
Readmision at 90 days
No  Yes	Ref 1.39	0.64–3.01	0.403
HISTOLOGY
Margin
R0   R1-R2	Ref 3.01	1.63–5.55	< 0.001	Ref 1.23	0.37–2.57	0.970
Size	1.03	1.01–1.05	0.001	1.00	0.98–1.03	0.730
Positive lymph nodes	1.02	1.0–1.04	0.042	1.01	0.98–1.04	0.655
Vascular invasion
No  Yes	Ref 3.11	1.66–5.79	< 0.001	Ref 1.59	0.67–3.76	0.294
Lymphatic invasion
No  Yes	Ref 3.32	1.84–5.98	< 0.001	Ref 2.21	1.03–4.73	0.041
Perineural invasion
No  Yes	Ref 2.16	2.06–4.43	0.035	Ref 1.15	0.38–3.48	0.799
Stromal invasion
No  Yes	Ref. 3.56	1.26–10.08	0.017	Ref. 1.01	0.22– 5.26	0.927
T
T1  T2  T3  T4	Ref 1.83 3.33 5.33	0.63–5.28 1.20–9.23 1–28.44	0.267 0.021 0.050	Ref 1.48 2.14 1.34	0.39–5.46 0.53–8.66 0.14–12.9	0.561 0.287 0.799
Vascular margin R1
No  Yes	Ref 2.43	1.15–5.11	0.020	Ref 1.52	0.52–4.41	0.442
Retroperitoneal margin R1
Yes  No	Ref 0.34	0.17–0.65	0.001	Ref 0.39	0.15–0.98	0.046

OR Odds ratio, CI Confidence interval, ASA American Society of Anestesiologist, ECOG Eastern Cooperative Oncology Group, ALT Alanine aminotransferase

Discussion

In our study involving 194 ASCPs, we applied the futility criteria outlined by Crippa et al. [18]. The futility rate in our series was notably high, approaching 50%. The parameter most frequently associated with futility was the first six-month recurrence. In our multivariate analysis, the following factors were linked to the risk of futility: positive retroperitoneal margin, lymphatic invasion, and not receiving chemotherapy.

Introducing a new concept like futility involves addressing several challenges. First, not all surgeons are familiar with this concept, and there may be a lack of understanding surrounding it. Additionally, the term “futility” often has a negative connotation, which can lead to initial resistance. One of the main issues is the absence of an international consensus on which parameters should be included in the futility category. As a result, published studies may not be comparable, as they likely measure different parameters or employ varying cut-offs [18, 19, 32].

In the limited literature available, the concept of oncological surgical futility typically encompasses several parameters. These include the non-performance of previously planned surgery due to intraoperative findings, commonly referred to as aborted surgery; the presence of affected margins in pancreatic tumors, which has significant prognostic implications; postoperative mortality, usually assessed at 90 days; early recurrence, which directly influences survival; and tumor-related mortality during the first 6 to 12 months post-surgery [18, 19, 32–34].

Only four articles have been published on futility in pancreatic tumors. The factors related to futility can be grouped into patient fragility (age and ASA classification) and biological factors (tumor size, grade, neoadjuvant therapy and CA19-9 levels) [18, 32–34].

Ahmed et al. developed a machine learning model to assess futility in these tumors, using data previously gathered by their team [33, 34]. They defined futility as death occurring within the first 12 months after surgery. Key predictors of futile outcomes after pancreatic surgery included advanced age (over 79 years), tumor size greater than 4 cm, and poor differentiation of the tumor. Additionally, neoadjuvant therapy was linked to a lower risk of futility. The researchers found a futility rate of 25%, but they did not consider early recurrence when defining patients as futile.

A study by Citterio et al. examined early recurrence and its significant impact on survival. The researchers identified several risk factors for early recurrence, including a CA19-9 level greater than 100 U/ml, G3 tumors, and the absence of adjuvant chemotherapy. The early recurrence rate was 21.6%; however, the study did not report deaths occurring within the first year. Based on these findings, the researchers developed a risk score to reduce surgical futility and to create personalized treatment strategies to manage tumor progression and maximize survival in patients with PDAC [32].

Crippa et al. examined the concept of futility in patients with PDAC who underwent upfront surgery. They defined futility in the context of potentially curative upfront pancreatectomy as a situation where death occurs due to postoperative complications, PDAC-related events, or cancer recurrence within six months of the operation, and we apply this definition in our series. The study found that the futility rate was 18.9%. The researchers identified several preoperative variables associated with futile resection, including the ASA classification, serum CA 19 − 9 levels, and tumor size. They categorized patients into three groups based on their risk of futility [18].

Our futility rate is significantly higher than that reported in previously published studies [18, 32–34]. This is the first measurement of futility in ASCP, making comparisons to prior studies impossible because those studies focused on pancreatic adenocarcinoma, which is less aggressive than ASCP. So, this higher rate may be linked to this histological variant, with a median survival of 7 to 11 months [2, 3, 5, 8, 12], which is inferior to that of non-ASCP pancreatic cancer. Additionally, Crippa et al. [18] included only patients who underwent upfront surgery, whereas 25% of our patients received neoadjuvant therapy, indicating that they had borderline tumors. The complication rate of 25% falls within the accepted benchmark range [35, 36].

In our study, only three factors were associated with treatment futility in the multivariate analysis: positive retroperitoneal margin, lymphatic invasion, and absence of chemotherapy. In other manuscripts, factors such as ASA score, age, tumor size, CA 19 − 9 levels, tumor grade, and not receiving adjuvant therapy have been associated with futility.

It is well established that R1 resection margins and lymphatic invasion are associated with earlier recurrence in pancreatic cancer, making it logical that these factors would contribute to increased futility in affected patients. Our R0 rate was low (64%), likely due to ASCP’s aggressive nature, and R0 is crucial for improving oncologic outcomes. In the presented series, the surgical approach (open vs. MIS) does not affect the rate of R1 or the number of lymph nodes harvested, as in previous published studies, but the MIS rate is only 10%.

The only parameter identified as a futility factor in previous studies and corroborated by our data is the absence of adjuvant chemotherapy. Most patients with ASCP should benefit from adjuvant chemotherapy. However, some do not receive this treatment because of severe complications following pancreatectomy, which can leave patients too fragile to tolerate chemotherapy. This scenario increases the risk of relapse and mortality, thus establishing it as a futility factor. Our low rate of specific complications, such as DGE, decreased the risk of chemotherapy delay. The lack of specific guidelines for ASCP tumors is also a limitation, as the need for chemotherapy may differ from that of ductal adenocarcinoma.

In a recent study by Campbell et al. [37] among 168 patients resected adenosquamous carcinoma, recurrence-free survival data showed upfront surgery had significantly longer recurrence-free survival compared with neoadjuvant chemotherapy (18 months vs. 5 months, p = 0.046). Multivariable analysis showed adjuvant therapy was associated with improved overall survival (hazard ratio [HR], 0.27; P < 0.001), whereas age ≥ 65 (HR, 1.79, p = 0.030) was associated with worse overall survival.

Our study’s limitations include focusing solely on operated patients, while non-operated and non-resected patients, who could provide valuable data for survival comparisons, were not included. Additionally, there are no previous studies to compare our findings and no consensus on the parameters to consider when assessing futility. On the other hand, our strengths lie in including many patients with a rare variant, marking the first time futility has been measured in ASCP.

The observed futility in ASCP is notably high and is associated with a positive retroperitoneal margin, lymphatic invasion, and not receiving adjuvant chemotherapy. Defining futile patients with an international consensus definition is crucial and has significant implications for shared decision-making and care optimization.

Abbreviations

ASA

American Society of Anesthesiologists

ASCP

Adenosquamous carcinoma of the pancreas

BMI

Body Mass Index

CCI

Comprehensive Complication Index

eCRF

Electronic case report form

GCP

Good Clinical Practice

HR

Hazard ratio

IQR

Interquartile Range

ISPGS

International Study Group of Pancreatic Surgery

PDAC

Pancreatic adenocarcinoma

REDCap

Research Electronic Data Capture

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

Authors’ contributions

All authors met the ICMJE authorship criteria. Contributions: (I) Conception and design: Mario Serradilla-Martín, José M. Ramia. (II) Administrative support: All authors. (III) Provision of study materials or patients: All authors. (IV) Collection and assembly of data: Mario Serradilla-Martín, José M. Ramia, Celia Villodre. (V) Data analysis and interpretation: Mario Serradilla-Martín, José M. Ramia, Celia Villodre. (VI) Manuscript writing: José M. Ramia. (VII) Final approval of manuscript: All authors.

Funding

No funding.

Data availability

The database is available for review upon request.

Declarations

Ethics approval and consent to participate

Not aplicable.

Competing interests

The authors declare no competing interests.