The impact of distress at diagnosis in patients with pancreatic cancer undergoing pancreatectomy

Abstract

Background

Cancer-related distress is associated with low quality of life and oncologic outcomes in cancer patients. At present, there are limited data regarding the clinical implications of distress in patients with pancreatic cancer. The present study aimed to investigate the association between distress at diagnosis and the surgical outcomes of patients with curative-intent surgery for pancreatic cancer.

Methods

Since 2014, distress thermometer (DT) surveys have been distributed to all patients with presumed cancer in the outpatient clinic of Samsung Medical Center. We retrospectively reviewed the clinicopathological data of patients who underwent curative-intent surgery for pancreatic cancer between 2014 and 2021. The survival of the patients according to DT score was analyzed using Kaplan-Meier graph and z-test. Risk factor analysis was performed to identify the impact of distress on postoperative complications.

Results

Among 1,050 patients with pancreatectomy, 130 patients responded to a DT survey. Thirty-three (25.4%), 67 (51.5%), and 30 (23.1%) patients presented with mild, moderate, and severe distress, respectively. In the stage II group, patients with moderate distress showed better survival compared to those with mild or severe distress. Higher body mass index (p = 0.043) and severe distress at diagnosis (p = 0.034) were found to be independent risk factors for major complications.

Conclusion

More than 70% of the patients had moderate to severe distress at diagnosis. Distress was associated with increased risk of major complications after pancreatectomy. Further research is needed to explore the potential effect of distress on outcomes of patients with pancreatic cancer.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12885-025-14365-9.

Background

Distress is common in patients diagnosed with cancer. The National Comprehensive Cancer Network (NCCN) defined cancer-related distress as ‘a multi-determined unpleasant emotional experience that may interfere with the ability to cope with cancer’ [1]. One study reported that up to 60% of cancer patients presented with psychological distress [2]. Several studies have suggested that distress is associated with pain, quality of life, and even oncologic outcome in cancer patients [3–7]. Some authors proposed that stress-induced immunosuppression and an exaggerated inflammatory response could be potential mechanisms linking distress to surgical and oncologic outcomes [8].

Despite the lethality and treatment burden of pancreatic cancer, there are highly limited data regarding psychological distress in patients with pancreatic cancer. Clark et al. reported that patients with pancreatic cancer were more vulnerable to distress than patients with other cancers [9]. In 2022, Chung et al. discussed the potential positive impact of distress management in patients undergoing palliative treatment for pancreatic cancer [10].

Pancreatectomy is the only curative treatment for pancreatic cancer, but there has yet to be any study reporting on the prevalence and impact of distress in patients undergoing curative-intent surgery. In the present study, we aimed to explore the clinical significance of psychological distress in terms of surgical and oncological outcomes in patients undergoing pancreatectomy for pancreatic cancer.

Materials and methods

Patient database

The Korean version of the distress thermometer (DT) and problem list (Fig. 1) has been distributed to all patients who are examined for newly-diagnosed cancer in the outpatient clinic of Samsung Medical Center since 2014. The survey consists of a 11-point numerical scale indicating the severity of distress (0 = no distress, 10 = extreme distress) and the list of concerns that such distress might originate from. The distress thermometer was distributed via mobile text message on the day of the first outpatient visit when cancer was suspected, often before treatment decision-making. Patients were grouped according to the severity of distress: no or mild (DT 0 ~ 3), moderate (DT 4 ~ 7), and severe (DT 8 ~ 10) groups.

Fig. 1.

The Korean version of the distress thermometer survey distributed to patients in the outpatient clinic of Samsung Medical Center

In total, 1,050 patients underwent curative-intent surgery for pancreatic cancer between 2014 and 2021. Among them, 130 patients responded to the DT survey preoperatively. Demographic characteristics and peri-operative laboratory data including tumor markers were reviewed retrospectively. The final pathology report provided the size and differentiation of the tumor, lymph node (LN) status, and R status defined by 1-mm rule from the Royal College of Pathologists [11]. The staging was based on the 8th edition of the AJCC staging system [12].

This study was approved by the Institutional Review Boards of Samsung Medical Center (SMC 2023-06-144). The Institutional Review Boards of Samsung Medical Center waived the need for written in-formed consent from the participants since the research involved no more than minimal risk to subjects, and there was no reason to assume rejection of agreement.

Surgical and oncological outcomes

Postoperative length of stay and complications were reviewed based on medical records. Postoperative complications were graded using Clavien– Dindo (C-D) classification [13]. Complications of C-D grade III or higher were defined as major complications. Complications were assessed during the in-hospital stay and within 30 days after surgery. After discharge, patients were followed up with contrast-enhanced computed tomography (CT) and tumor markers assessed every three months by surgeons or medical oncologists. When a patient presented with any suspicious lesions at CT or elevated tumor markers, additional positron emission tomography (PET) scan or biopsy was performed to confirm the recurrence of cancer.

Recurrence-free survival (RFS) was calculated as the time interval from the date of operation to the date when recurrence was confirmed. Overall survival (OS) was calculated as the time from the date of operation to death from any causes. The last follow-up of survival data was updated in December 2023.

Statistical analysis

Binary logistic regression analysis was performed to identify risk factors for major complications. Variables that were found to have p-values < 0.1 in univariable analysis were included in multivariable analysis. Odds ratios (ORs) were reported with 95% confidence intervals (CIs).

Kaplan-Meier survival graphs were plotted to compare RFS and OS among DT severity groups. Z-test was performed to determine the statistical difference of survival outcomes at postoperative three and five years among patient groups. To evaluate prognostic factors associated with survival, a Cox proportional hazards regression analysis was performed. Variables with p < 0.1 in univariate analysis were included in the multivariable model. Hazard ratios (HRs) with 95% CIs were reported. All statistical analyses were performed using IBM SPSS version 26 (SPSS Inc., Chicago, IL, USA) and SAS version 9.4 (SAS Institute Inc, Cary, NC, USA).

Results

Patient characteristics

Tables 1 and 2 list the demographic and clinicopathological data of the patients. The mean age at operation was 64.0 years, and there were 72 (55.3%) male patients. Twenty-two (16.9%) patients received neoadjuvant treatment. There were 78 (60.0%) patients with pancreaticoduodenectomy, 44 (33.8%) with left-sided pancreatectomy, and 8 (6.2%) with total pancreatectomy. In terms of the pathology report, 47 (36.2%) patients had stage I disease, 60 (46.1%) had stage II disease, and 23 (17.7%) had stage III disease. The R0 resection rate was 82.3%. Ninety-three (71.5%) patients underwent adjuvant treatment.

Table 1.

Patient demographics and operative outcomes (n = 130)

Variables	N (%) or mean (± SD)
Demographic and preoperative data
Age at operation (years)	64.0 (± 8.9)
Sex
Male	72 (55.3%)
Female	58 (44.7%)
BMI (kg/m2)	23.4 (± 2.8)
ASA score ≥ III	20 (15.4%)
Underlying DM	43 (33.1%)
Preop. Symptomsa	37 (28.5%)
Preop. CA 19 − 9 (U/mL)	520.2 (± 1036.7)
Preop. CA 19 − 9 > 37 U/mL	86 (66.2%)
Preop. biliary drainage	58 (44.6%)
Neoadjuvant treatment	22 (16.9%)
Distress at diagnosis	122 (93.8%)
Median DT score	5 (± 2.7)
Surgical outcomes
Operation types
Pancreaticoduodenectomy	78 (60.0%)
Left-sided pancreatectomy	44 (33.8%)
Total pancreatectomy	8 (6.2%)
Minimally invasive surgery	10 (7.7%)
Combined vascular resection	37 (28.5%)
Estimated blood loss (mL)	382.2 (± 297.8)
Complications	57 (43.8%)
Clavien-Dindo grade ≥ III	26 (20.0%)
Hospital stay (days)	11.3 (± 6.0)

^aPreoperative symptoms included any of the following; abdominal pain, jaundice and weight loss of more than 5% of usual body weight over 6 to 12 months

BMI body mass index, ASA the American Society of Anesthesiologists, DM Diabetes Mellitus, Preop. Preoperative, CA 19 − 9 carbohydrate antigen 19 − 9, DT distress thermometer

Table 2.

Pathology and oncologic outcomes of the patients (n = 130)

Variables	N (%) or mean (± SD)
Pathology
AJCC 8th stage
I	47 (36.2%)
II	60 (46.1%)
III	23 (17.7%)
Tumor size (cm)	2.9 (± 1.1)
LN metastasis	77 (59.2%)
Tumor differentiation
Well differentiated	2 (1.5%)
Moderately differentiated	95 (73.1%)
Poorly or undifferentiated	30 (23.1%)
R0 resection	107 (82.3%)
Oncologic outcomes
Adjuvant treatment	93 (71.5%)
Median recurrence-free survival	12.4 months
5-year recurrence-free survival rate	20.5%
Median overall survival	28.0 months
5-year overall survival rate	24.5%

AJCC the American Joint Committee on Cancer, LN lymph node

Median RFS was 12.4 months with a 5-year RFS rate of 20.5%. Median OS was 28.0 months with a 5-year OS rate of 24.5%.

Results of distress thermometer questionnaire

Among all patients, 122 (93.8%) had preoperative distress (Table 1). The median DT score was 5. In terms of severity, 33 (25.4%) patients had no or mild distress, 67 (51.5%) had moderate distress, and 30 (23.1%) had severe distress (Fig. 2A). For each of the five domains—practical, social, emotional, physical, and spiritual/religious—patients were considered to have domain-specific distress if they endorsed at least one item within that category on the problem list. Most of the patients responded that they had emotional (83.8%) and physical (85.3%) distress (Fig. 2B).

Fig. 2.

The distribution chart of distress thermometer score (A) and the list of concerns (B)

Risk factors for surgical complications

There were 57 (43.8%) patients with postoperative complications (Table 1) and there were 26 (20.0%) patients who suffered from major complications. The mean length of hospital stay was 11.3 days.

Table 3 presents the risk factor analysis for major complications. Body mass index (BMI), neoadjuvant treatment, and distress at diagnosis were included in the multivariable analysis. BMI (OR: 1.187, 95% CI: 1.015–1.387, p = 0.031) and distress at diagnosis (OR: 1.219, 95% CI: 1.024–1.451, p = 0.026) were statistically significant risk factors for increasing major postoperative complications. In particular, patients with severe distress had an approximately four-fold increased risk of major complications compared to those in other DT groups (OR: 3.891, 95% CI: 1.044–14.502, p = 0.043).

Table 3.

Univariate and multivariate analyses of risk factors for major complications

Variables	Univariable p	OR	95% CI	Multivariable p
Age at operation	0.820
Sex, male (ref. female)	0.860
BMI	0.043	1.187	1.015–1.387	0.031
ASA score, III-IV (ref. I-II)	0.859
Underlying DM	0.457
Preop. elevated CA 19 − 9	0.756
Preop. biliary drainage	0.538
Neoadjuvant treatment	0.079	0.145	0.017–1.215	0.075
Distress at diagnosis (ref. no distress)	0.034	1.219	1.024–1.451	0.026
Severe distress (ref. mild to moderate)		3.891	1.044–14.502	0.043
Pancreaticoduodenectomy (ref. others)	0.960
Minimally invasive surgery	0.424
Combined vascular resection	0.498
Estimated blood loss	0.606
Pathologic stage, III (ref. I-II)	0.561

OR odds ratio, CI confidence interval, BMI body mass index, ASA the American Society of Anesthesiologists, DM Diabetes Mellitus, Preop. Preoperative, CA 19 − 9 carbohydrate antigen 19 − 9

Survival analysis

The median follow-up duration for survival analysis was 52 months. Kaplan-Meier survival graphs were plotted to show RFS and OS of all patients stratified according to DT severity (Fig. 3). The median RFS was 15.5, 13.5, and 8.2 months in no or mild, moderate, and severe DT group, respectively. In terms of OS, the median survival was 27.8, 31.2, and 16.3 months. However, there was no statistically significant difference between the groups in either RFS (p = 0.253) or OS (p = 0.244).

Fig. 3.

Recurrence-free survival (A) and overall survival (B) according to distress thermometer severity in all patients (n = 130)

Z-tests were performed to compare the survival rates at postoperative three and five years in patients with stage II pancreatic cancer (n = 60, Fig. 4). Figure 4A shows a comparison of RFS between DT severity groups. Patients with a moderate DT score showed better RFS than those with no or mild (3-year rate: 50.0% vs. 17.9%, p = 0.036; 5-year rate: 42.9% vs. 8.9%, p = 0.017) or severe (3-year rate: 50.0% vs. 26.5%, p = 0.143; 5-year rate: 42.9% vs. 13.2%, p = 0.064) DT scores. Comparing OS within stage II patients, the 3-year survival rate of the moderate DT group was superior to that of the no or mild DT group (52.2% vs. 20.3%, p = 0.028). Otherwise, there was no statistically significant difference.

Fig. 4.

Z-test of recurrence-free survival (A) and overall survival (B) according to distress thermometer severity in patients with stage II pancreatic cancer (n = 60)

A Cox regression analysis was conducted to identify independent prognostic factors for OS (Table 4). In the multivariable analysis, combined vascular resection (HR 1.599, 95% CI: 1.036–2.468, p = 0.034), pathologic stage (III) (HR 1.288, 95% CI: 1.172–2.330, p = 0.043), and pathologic LN metastasis (HR 1.426, 95% CI: 1.117–2.217, p = 0.013) were significantly associated with worse OS. However, distress at diagnosis and severe distress were not found to be statistically significant predictors.

Table 4.

Univariate and multivariate Cox regression analyses of risk factors for overall survival

Variables	Univariable p	HR	95% CI	Multivariable p
Age at operation	0.117
Sex, male (ref. female)	0.237
BMI	0.280
ASA score, III-IV (ref. I-II)	0.719
Underlying DM	0.571
Preop. elevated CA 19 − 9	0.710
Preop. biliary drainage	0.744
Neoadjuvant treatment	0.345
Distress at diagnosis (ref. no distress)	0.295
Severe distress (ref. mild to moderate)	0.112
Pancreaticoduodenectomy (ref. others)	0.734
Minimally invasive surgery	0.349
Combined vascular resection	0.027	1.599	1.036–2.468	0.034
Estimated blood loss	0.056	1.000	1.000–1.001	0.341
Major complications	0.191
Pathologic stage, III (ref. I-II)	0.042	1.288	1.172–2.330	0.043
Tumor size	0.075	1.172	0.978–1.404	0.086
Pathologic lymph node metastasis	0.015	1.426	1.117–2.217	0.013
Tumor differentiation, WD (ref. MD/PD)	0.525
R0 resection	0.127
Adjuvant treatment	0.103

HR hazard ratio, CI confidence interval, BMI body mass index, ASA the American Society of Anesthesiologists, DM Diabetes Mellitus, Preop. Preoperative, CA 19 − 9 carbohydrate antigen 19 − 9, WD well-differentiated, MD moderately differentiated, PD poorly differentiated

Discussion

Distress is common in patients with cancer and it can affect the quality of life of patients during cancer treatment. However, there has been very limited research examining the relationship between distress and prognosis of pancreatic cancer. The present study aimed to investigate the impact of distress on surgical and oncological outcomes in patients with pancreatic cancer undergoing curative surgery. The results showed that a high level of distress increased the risk of major postoperative complications and that the survival was different between DT severity groups.

The clinical utility of the DT has been well established in oncology settings. The original English version of DT is an 11-point visual analog scale (0 = no distress, 10 = extreme distress) assessing the level of distress experienced over the past week. It is accompanied by a 44-item problem list that identifies potential sources of distress across physical, emotional, social, practical, and spiritual or religious concerns (Supplementary Fig. 1) [1]. As recommended by the NCCN, the DT serves as a validated, multidomain screening tool for distress in cancer patients and has been widely adopted for routine psychosocial assessment [14, 15]. In this study, we used the Korean version of the DT, which has been consistently utilized in prior research conducted at our institution [7]. Based on its practicality and familiarity in our clinical workflow, it was deemed suitable for retrospective analysis. However, future studies may consider exploring and validating alternative distress screening tools to better capture the multifaceted nature of psychosocial distress in Korean patients with cancer.

One of the significant findings of the present study is the association between preoperative distress and surgical complications. Many previous studies have reported that postoperative complications negatively affect psychological outcomes and well-being of patients [16], but the effect of psychological distress on the incidence of complications has been rarely investigated. A review article in 2016 investigated the impact of depression on surgical complications [17]. The authors suggested that the immune suppression in depressive disorders might expose patients to higher risks of postoperative complications. Another study showed that patients with anxiety and depression had an increased risk of 30-day surgical mortality [8]. Preoperative emotional status may also affect postoperative pain. As an example, De Cosmo et al. identified that patients with preoperative anxiety and depression had more severe pain with increased use of analgesics [18].

Despite these potential associations, there remains no consensus on the optimal cut-off value of the DT in predicting postoperative outcomes. Although the NCCN Guidelines recommend a DT score of ≥ 4 as a threshold for clinically significant distress [1], prior studies have reported varying cut-points depending on cancer type, clinical setting, and outcome of interest. For example, Hong et al. evaluated DT performance in patients with nasopharyngeal cancer and concluded that no potential cut-off score demonstrated acceptable sensitivity [19]. In a more recent study by Abu-Odah et al., a DT score of ≥ 6 was found to be an acceptable and effective threshold for screening distress in patients with advanced-stage cancer in palliative care settings [20]. In our study, DT stratification was exploratory and based on internal distribution patterns. We acknowledge this as a limitation and emphasize the need for prospective validation to determine an appropriate threshold in patients undergoing surgery for pancreatic cancer.

The important aspect of preoperative distress is that distress is preoperatively manageable by psychological interventions. A systemic review by Lanini et al. argued that perioperative psychologic intervention may improve surgical outcomes by modulating patients’ stress response of patients [21]. Particularly for patients undergoing major abdominal surgery, it was reported that psychological management including relaxation techniques and cognitive-behavioral therapies might have a positive effect on surgical outcomes [22]. In the present study, all the patients underwent pancreatectomies, which involve a high risk of postoperative complications such as hemorrhage, postoperative pancreatic fistula, and intra-abdominal infection. Aside from the high probability of post-pancreatectomy complications, preoperative distress was an independent risk factor for major complications. These results show the potential for a future study including preoperative psychological intervention to improve surgical outcomes after pancreatectomies by reducing distress of patients. Based on the findings of this study, our institution has recently implemented a system in which physicians receive an automatic notification recommending psychiatric consultation whenever a patient reports severe distress on the DT survey. This aims to facilitate timely psychosocial support and improve perioperative care. We are also planning to conduct a further examination of the impact of distress in patients undergoing pancreaticoduodenectomy for other periampullary cancers.

In terms of cancer survival, previous studies have discussed the potential negative effect of distress in breast, gynecologic, and digestive cancers [6, 7, 23]. A recent study from our institution investigated the impact of distress at diagnosis on oncologic outcomes in patients with resectable colon cancer [7]. Patients with severe distress showed worse RFS than those with low distress. The authors explained the distress-recurrence association by the activation of stress-response system and reduced immunity. Even beyond the direct biologic link between distress and cancer prognosis, distress can indirectly affect oncologic outcomes by altering compliance among patients. In one study including 500 oncology patients, the authors identified that psychological distress was related to delayed time to treatment [24]. In pancreatic cancer, patients showed higher distress than patients with other cancer diagnoses [9], and more than 40% of patients experienced depression after diagnosis [25]. Distress in pancreatic cancer patients can come from the diagnosis itself or from pre-existing depression or anxiety. Davis et al. examined the frequency of depression in pancreatic cancer patients and the association between depression and treatment compliance [26]. The study included 856 patients with pancreatic cancer, and 246 (31.1%) of them underwent pancreatectomy. Among all patients, 4.6% had depression or anxiety preceding diagnosis, and these patients were less likely to receive chemotherapy. Pre-existing depression or anxiety was also associated with decreased OS in patients with metastatic pancreatic cancer. In the present study, we compared post-pancreatectomy oncologic outcomes according to DT severity. Patients with severe DT scores showed worse survival than those with moderate DT scores, but the statistical significance was somewhat uncertain. Also, survival outcomes among patients with mild and severe distress were unexpectedly similar. Cox regression analysis to adjusting for potential confounders did not identify distress at diagnosis as an independent prognostic factor for survival. This could be due to either the small number of the study cohort or the well-known devastating nature of pancreatic cancer. Nonetheless, the results from many previous studies as well as the present study suggest the existence of a potential association between distress and cancer prognosis; a prospective study with a larger cohort is necessary to further explore this.

The present study has several limitations. Above all, this study is a retrospective study that included a relatively small number of patients. Since only patients who responded to DT survey and underwent curative-intent surgery were included, the results could have been influenced by selection bias. Overall, the response rate of the DT survey was 12.4% (130 out of 1,050). The low response rate can be mainly attribute to its method of distribution and a lack of awareness. The survey was basically distributed through mobile text message to patients on the day of their first outpatient clinic, and it was not compulsory. Moreover, both patients and health care providers might have regarded distress as being less important than components in treatment process of pancreatic cancer. Given these limitations, distress data for the remaining patient population could not be retrieved in a standardized manner, and this lack of universal assessment further introduces potential selection bias. This issue highlights the importance of prospective data collection using systematic and institutionally integrated distress screening protocols to ensure representative sampling and minimize bias in future studies. Another source of bias is the timing of DT measurement with respect to the date of operation. Some patients responded to the survey before getting any insight into malignancy, and others responded after they were already diagnosed with cancer and had the operation scheduled. The waiting times for surgery also differed between the patients. Particularly those receiving neoadjuvant therapy, may have experienced higher distress levels due to uncertain surgical candidacy or prolonged treatment timelines. Although resectability status was not stratified in our current analysis, we acknowledge its potential influence and plan to include it in future prospective studies using resectability criteria. In addition, an analysis of the impact of the types of concerns was omitted. It is plausible that cancer patients become distressed from symptoms of malignancy and fear, anxiety towards cancer. However, their distress can also stem from other sources such as socio-economic burdens or familial problems. Subgroup analyses based on distress severity were also limited by the small number of respondents in each category, which may limit the insights into potential contributing factors. Outcomes according to the severity and types of distress could be a potential subject of a future investigation using a large cohort of DT survey responders.

Despite these potential limitations, the present study revealed that preoperative distress could potentially affect the outcomes of patients with pancreatic cancer by interfering in the systemic stress response and increasing risks of postoperative complications. An important aspect of distress is that it can be modified preoperatively with multimodal treatment. Also, future studies including all patients with pancreatic cancer, regardless of surgical treatment, would increase the generalizability of the findings suggesting the potential link between distress and cancer prognosis. We are planning a prospective study to explore the effect of preoperative distress management in patients with pancreato-biliary cancers undergoing curative-intent treatment.

Conclusion

This study identified that preoperative psychological distress may have a potential negative impact on postoperative outcomes in patients undergoing curative-intent surgery for pancreatic cancer. These findings suggest the potential value of incorporating early psychosocial assessment and support into perioperative care. Further large-scale prospective studies are needed to validate these observations and explore effective interventions.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

BMI

Body mass index

C-D

Clavien– Dindo

CIs

Confidence intervals

CT

Computed tomography

DT

Distress thermometer

LN

Lymph node

NCCN

National Comprehensive Cancer Network

ORs

Odds ratios

OS

Overall survival

PET

Positron emission tomography

RFS

Recurrence-free survival

Author contributions

Conceptualization: Yoon SJ, Kim H, Shin SH, Heo JS, Han IWData curation: Yoon SJ, Lim SY, Jeong H, Chae H, Kim HS, Yoon SK, Kim H, Shin SH, Heo JS, Han IWFormal analysis: Yoon SJ, Lim SY, Jeong H, Chae H, Kim HS, Yoon SK, Kim H, Shin SH, Heo JS, Han IWWriting of the original draft, review, and editing: Yoon SJ, Lim SY, Jeong H, Chae H, Kim HS, Yoon SK, Kim H, Shin SH, Heo JS, Han IW.

Funding

This study was supported by Research Program of the Korean Pancreas Surgery Club for KPSC-2023-002.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The Institutional Review Boards of Samsung Medical Center (SMC 2023-06-144) approved this study. The Institutional Review Boards of Samsung Medical Center waived the need for written in-formed consent from the participants since the research involved no more than minimal risk to subjects, and there was no reason to assume rejection of agreement.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.