Correlation of Clinical Stage and Performance Status With Quality of Life in Patients Seen in a Pancreas Multidisciplinary Clinic

Although several measures of QoL have been observed in all patients, certain measures seem to correlate specifically with worse disease status. Routine assessment of QoL is suggested to guide treatment decisions.

Abstract

Introduction:

The objectives of this study were to evaluate quality of life (QoL) in patients presenting to the Johns Hopkins Pancreas Multidisciplinary Clinic (PMDC), and to examine associations between disease status, performance status, and QoL in order to identify patient subgroups that are most at risk for reduced QoL.

Patients and Methods:

Data from 77 patients were evaluated. At initial presentation, disease and performance status were assessed, as well as QoL, which was obtained with the European Organisation for Research and Treatment of Cancer QLQ-PAN26 questionnaire. Statistical analyses examined associations between QoL, disease status, and performance status.

Results:

Digestive symptoms (P < .003) significantly differed by pancreatic disease status (resectable, resected, locally advanced, and metastatic). Patients with a worse performance status, defined as Eastern Cooperative Oncology Group ≥ 1, were more likely to report symptomatic pancreatic pain (P = .001), digestive symptoms (P = .017), cachexia (P = .004), and ascites (P < .001) compared with patients with a performance status of 0. The majority (92%) of patients reported a significant fear of future health problems, regardless of disease status or performance status.

Conclusion:

Although several measures of QoL have been observed in all patients, certain measures appear to correlate specifically with worse disease status. Therefore, routine assessment of QoL is suggested in order to guide treatment decisions. Further investigation on optimizing the use of QoL measures and patient-reported outcomes to better tailor management is warranted.

Introduction

Worldwide, approximately 230,000 cases of pancreatic cancer are diagnosed annually, with an overall 5-year survival of less than 5%.^1–3 Patients with pancreatic cancer often suffer from intrusive symptoms and adverse effects related to tumor progression and treatments such as chemotherapy and radiation. Given the high mortality rate of this malignancy, it is particularly important to optimize treatment for these patients and to select appropriate interventions that reduce problematic symptoms and treatment adverse effects while maximizing their remaining quality of life (QoL).

Several recent studies have attempted to incorporate self-reported QoL measures as outcomes for patients with pancreatic cancer,^4–6 hepatobiliary cancers,⁷ and chronic pancreatitis. Broadly speaking, QoL measures generally assess patients' functional status, symptoms, or supportive care needs. Although the increasing use of QoL measures in these studies is promising, these tools tend to remain underused, and significant gaps remain in the current QoL literature in pancreatic cancer.⁸ For instance, prior studies have generally examined QoL as an outcome of specific treatment modalities such as chemotherapy, surgery, and/or radiation therapy.^4,6,8,9 Although these studies are helpful in quantifying longitudinal changes in QoL over time related to specific treatment modalities, they do not allow for comparisons of QoL in patients with pancreatic cancer based on disease characteristics at the time of presentation (eg, disease status, performance status). Such comparisons could allow clinicians to identify and manage subgroups of patients at greater risk for reduced QoL based on easily identifiable patient characteristics.

One forum for utilizing patient QoL measures is in the context of cancer multidisciplinary clinics (MDCs), which involve input of clinicians with a range of specialties. The use of MDCs is growing because these clinics leverage expertise from multiple fields to ensure the most appropriate course of treatment, and because they are patient centered in that they prioritize patient and family needs and preferences. We have previously reported that a pancreas MDC can increase the accuracy of initial staging and improve survival for select patients with presumed pancreatic cancer.¹⁰ Moreover, a patient's participation in an MDC may reduce the wait between diagnosis and treatment and therefore lower patient anxiety while also increasing patient exposure to a range of support services and improving enrollment onto clinical trials.¹⁰ In the context of an MDC, it is possible to compare patients across a range of presenting clinical characteristics at diagnosis, thus allowing for identification of subgroups of patients at a higher risk for symptom distress and reduced QoL.

The primary purpose of this cross-sectional cohort study was to evaluate how QoL scores change based on clinical stage at presentation to the Johns Hopkins Pancreas Multidisciplinary Clinic (PMDC). A secondary purpose was to examine associations between QoL as they relate to self-reported symptoms, clinical characteristics, and performance status. Understanding how QoL relates to disease status could give us further insight into characteristics seen in at-risk groups, which may help health care teams tailor management and address areas of need.

Patients and Methods

Patients

Between January 2007 and March 2009, data were prospectively collected on 110 patients at initial presentation to the Johns Hopkins PMDC.¹⁰ Institutional review board approval was obtained before chart review. Patients with significant missing demographic data, functional status, or questionnaire data were excluded, leaving data from 77 patients available for final analysis. Details related to the clinical setting and collection of patient information at Johns Hopkins PMDC have been previously described.¹⁰ In short, standardized pathologic, radiologic, laboratory, and clinical data on each patient are collected, evaluated, and discussed in the PMDC setting.¹⁰ Although the majority of patients evaluated at PMDC have pancreatic adenocarcinoma, patients with pancreas tumors of other histologic variants are also evaluated. Pancreas disease status was classified as (1) benign/nonadenocarcinoma, including benign masses, intraductal papillary mucinous neoplasms (IPMN), and neuroendocrine tumors; (2) resectable pancreatic adenocarcinoma before resection; (3) pancreatic adenocarcinoma in patients seeking adjuvant therapy recommendations after curative resections; (4) locally advanced pancreatic adenocarcinoma, including both borderline resectable and unresectable disease; and (5) metastatic pancreatic adenocarcinoma. These classifications were used because they represent clinically meaningful distinctions that typically guide treatment decisions.

Measures

QoL.

QoL was assessed at initial presentation to the Johns Hopkins PMDC using the European Organisation for Research and Treatment of Cancer (EORTC) QLQ-PAN26, a questionnaire validated in previous studies for pancreatic disease.^4,8,11 This questionnaire comprises 26 items relating to self-reported symptoms during the past week, with responses scored using a 4-point scale in which 1 = not at all and 4 = very much. Questions cover a range of symptoms including disease-related symptoms and emotional issues specific to pancreatic cancer.^8,12 The EORTC QLQ-PAN26 consists of the following items or groups of items: pancreatic pain, gastrointestinal symptoms, cachexia, hepatic symptoms, ascites, altered bowel habits, body image, adverse effects, fear of future health, ability to plan for future health, sexuality, and health care satisfaction (Data Supplement). For each patient, the score for each category was calculated using a linear transformation with standardization of the raw scores. With the exception of health care satisfaction, a high score (100%) indicated more pronounced symptoms and, in general correlated with a lower QoL. On the basis of prior research,⁸ a symptom score of 50% (3) was considered symptomatic with moderate to severe impairment of QoL. However, a high health care satisfaction score was a positive result, associated with greater patient satisfaction.

Performance status.

Performance status (PS) was assessed by clinicians using two separate metrics: Eastern Cooperative Oncology Group (ECOG) PS^13,14 and Karnofsky performance status (KPS).^13,15

Statistical Analyses

Statistical analyses were performed using STATA, version 9 (Stata, College Station, TX). Symptomatic QLQ-PAN26 scores were assessed for association with pancreatic disease status and PS. PS was considered symptomatic if ECOG PS was ≥ 1 (restricted in strenuous work) or KPS was ≤ 80% (normal activity with effort, some signs/symptoms of disease). Baseline characteristics were compared by disease status using analysis of variance for continuous variables and χ² tests for dichotomous variables to determine statistical significance, defined as P < .05. For QoL, χ² tests were used to compare the proportion of symptomatic respondents to pancreas disease status and PS.

Results

Baseline characteristics are shown in Table 1. Patients were classified by disease status as follows: benign lesions/nonadenocarcinomas (15.5%), resectable (17.0%), locally advanced (35.0%), metastatic (15.5%), and resected (17.0%). Mean ECOG PS and KPS scores were 0.5 and 91.6, respectively. Performance status and tumor size significantly differed across disease status groups (tumor size, P = .045; ECOG > 1, P = .009; KPS < 80%, P = .041). There were no significant differences for age, sex, or carbohydrate antigen 19-9 by disease status.

Table 1.

Baseline Patient Characteristics

Characteristic	Benign		Resectable		Resected		Locally Advanced		Metastatic		P
	No.	%	No.	%	No.	%	No.	%	No.	%
No. of patients	12	15.5	13	17	13	17	27	35	12	15.5
Age, years											.182
Mean	61.1		67.3		63.3		68.7		64.0
SD	12.7		9.1		8.3		10.2		9.2
Male sex	5	41.7	4	30.8	8	61.5	15	55.6	6	5.0	.514
Tumor size, cm*											.045
Mean	2.3		2.6		2.5		3.7		3.2
SD	1.4		1.4		0.0		1.3		1.5
CA 19-9 , U/mL†											.072
Mean	23.4		426.8		129.4		491.9		2,069.2
SD	18.8		724.3		256.5		818.6		4,130.0
ECOG PS											.009
Mean	0.1		0.5		0.3		0.9		0.6
SD	0.3		0.7		0.5		0.8		0.5
KPS, %
Mean	98.3		92.3		96.2		85.6		92.5
SD	3.9		14.2		6.5		18.3		7.5		.041

Abbreviations: CA 19-9, carbohydrate antigen 19-9; ECOG PS, Eastern Cooperative Oncology Group performance status; KPS, Karnofsky performance status; SD, standard deviation.

^*Patients evaluable for tumor size: benign (n = 12), resectable (n = 12), locally advanced (n = 26), metastatic (n = 8), resected (n = 2).

^†CA 19-9: patients evaluable: benign (n = 7), resectable (n = 12), locally advanced (n = 26), metastatic (n = 9), resected (n = 10).

Table 2 demonstrates the proportion of symptomatic patients by QoL across disease status groups. Symptomatic scores related to digestive symptoms, including indigestion, flatulence, and diet limitations, significantly differed by disease status (P = .003). In a separate analysis including patients with benign lesions, there was difference in the proportion of patients with a limited ability to plan for the future according to disease status (P = .229). A similar but nonsignificant trend was seen for cachexia (P = .259), ascites (P = .364), indigestion (P = .107), and sexuality (P = .060). More than half (55%) of the patients reported symptomatic sexual problems. Sexual problems were most common among patients after tumor resection (55.6%) and were fairly uncommon among patients in other disease categories (≤ 20%). Irrespective of disease status, the single item that consistently scored high was a fear of future health problems.

Table 2.

Percentage of Patients With Symptomatic Scores Based on Pancreas Disease Status and Nonadvanced Versus Advanced Stage of Disease

QoL Item(s)	Disease Status						Nonadvanced Versus Advanced Stage of Disease
	n	Resectable (17%)	Resected (17%)	Locally Advanced (35%)	Metastatic (16%)	P	n	Nonadvanced* (49%)	Advanced† (51%)	P
Pancreatic pain	65	30.8	15.4	40.7	33.3	.457	77	21.1	38.5	.095
Digestive symptoms	64	7.7	30.8	66.7	54.6	.003	76	13.2	63.2	< .001
Altered bowel habits	52	16.7	7.7	25.9	16.7	.567	75	13.9	23.1	.308
Hepatic symptoms	65	15.4	7.7	14.8	25.0	.694	77	7.9	18.0	.189
Cachexia	65	23.1	7.7	37.0	25.0	.259	77	10.5	33.3	.016
Ascites	64	15.4	15.4	37.0	27.3	.364	76	10.5	34.2	.013
Indigestion	64	7.7	7.7	33.3	36.4	.107	76	7.9	34.2	.005
Body image	63	15.4	16.7	19.2	25.0	.933	75	10.8	21.1	.226
Fear of future health problems	59	83.3	66.7	69.2	58.3	.610	71	72.7	65.8	.528
Limited ability to plan for future	54	40.0	12.5	53.9	45.5	.229	66	17.2	51.4	.004
Health care satisfaction	55	66.7	88.9	79.2	90.0	.610	67	66.7	66.7	.520
Sexuality	30	12.5	55.6	20.0	0.0	.060	42	26.9	12.5	.269

NOTE. Numbers in each column represent the proportion of patients that have both characteristics of pancreatic disease status and QoL symptomatic score. Bold font indicates significance at P < .05.

Abbreviation: QoL, quality of life.

^*Nonadvanced includes benign, resectable, and resected pancreatic cancer.

^†Advanced includes unresectable and metastatic pancreatic cancer.

The proportion of symptomatic patients by QoL across various levels of PS is found in Table 3. Patients with lower performance status measured by both ECOG PS (≥ 1) and KPS (≤ 80%) had worse QoL scores on the following PAN26 items: symptomatic pancreatic pain (ECOG PS, P = .001; KPS, P = .014), digestive symptoms (ECOG PS, P = .017; KPS, P = .026), cachexia (ECOG PS, P = .004; KPS, P = .005), and ascites (ECOG PS, P < .001; KPS, P = .006). A lower KPS (≤ 80%) alone was associated with worse body image (P = .026), a limited ability to plan for the future (P = .016), and problems with sexuality (P = .025).

Table 3.

Percentage of Patients With Symptomatic Score by Performance Status

QoL Measure(s)	ECOG PS				KPS
	n	0 (%)	≥ 1 (%)	P	n	> 80% (%)	80% (%)	P
Pancreatic pain	77	14.3	48.6	.001	77	23.8	57.1	.014
Digestive symptoms	76	26.2	52.9	.017	76	32.3	64.3	.026
Altered bowel habits	75	15.0	22.9	.384	75	14.8	35.7	.069
Hepatic symptoms	77	11.9	14.3	.757	77	14.3	7.1	.472
Cachexia	77	9.5	37.1	.004	77	15.9	50.0	.005
Ascites	76	7.1	41.2	< .001	76	16.1	50.0	.006
Indigestion	76	16.7	26.5	.297	76	19.4	28.6	.445
Body image	75	12.5	20.0	.377	75	11.5	35.7	.026
Fear of future health problems	71	65.8	72.7	.528	71	69.5	66.7	.847
Limited ability to plan for future	66	26.5	46.9	.085	66	29.6	66.7	.016
Healthcare satisfaction	67	73.0	83.3	.312	67	78.2	75.0	.811
Sexuality	42	22.2	20.0	.866	42	16.2	60.0	.025

NOTE. The numbers in each column represent the proportion of patients that have both characteristics of performance status and QoL symptomatic score. Bold font indicates significance at P < .05.

Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance status; KPS, Karnofsky performance status.

Discussion

The overarching objective of this cross-sectional cohort study was to examine associations between self-reported QoL (QLQ-PAN26) and symptom reports with disease and PS from patients presenting to the Johns Hopkins PMDC, in an effort to identify patient subgroups most at risk for reduced QoL. We hypothesized that patients with more advanced disease status and worse PS would score higher (ie, worse) on the PAN26 questionnaire. Our findings largely supported this hypothesis, with an exception to fear of future health problems. Interestingly, the majority of the sample (69%) reported moderate to severe fear for future health problems, and this did not differ by disease status or PS. In patients with benign pancreatic masses, fear of future health problems might stem from inadequate understanding of their diagnosis or prognosis. Although all patients should receive empathetic support and reassurance by the physician as standard of care, patients with excessive worry may benefit from more intensive interventions such as counseling and/or antidepressants.¹⁶

There was a statistically statistical difference seen in PS, digestive symptoms and limited ability to plan for the future between patients with different disease statuses. Lower PS was also statistically significantly associated with pancreatic pain, cachexia, and ascites. The consistency of these findings suggests that having advanced disease correlates with symptoms that likely contribute to a poor PS. Having a poor PS has been shown to be correlated with decreased survival and limits a patient's ability to receive treatment.¹⁷ Therefore, our results suggest that more aggressive symptom management may result in improved PFS and thus better outcomes for patients with pancreatic cancer. Administration of the QLQ-PAN26 at diagnosis may be more effective at alerting caregivers of symptoms and/or concerns that may directly improve progression-free survival and subsequent outcomes.

Whereas prior studies have examined QoL with specific treatment modalities (eg, chemotherapy), this study contributes to the literature by examining how QoL in patients with pancreatic cancer compares across patients by disease status and PS. This may help identify patients at greater risk for QoL impairments who could benefit from further assessment and more intensive treatments, which may ultimately translate into a clinical benefit for patients. For example, given that patients with more advanced disease (locally advanced and metastatic pancreatic adenocarcinoma) reported worse digestive symptoms, providers may want to consider more aggressive management that may include a proton pump inhibitor and/or pancreatic enzyme supplementation.¹⁸

Similarly, given the association between low PS and pancreatic pain, providers may consider a more aggressive approach to pain assessment and management. This may include prescribing stronger or longer acting opiates or referring patients to a pain and palliative care provider for evaluation of a celiac plexus block.^18–21

Patients with advanced cancer often have interest in maintaining sexual intimacy and report sexual concerns to be important, yet these concerns tend to be neglected by providers.²² In this study, 55% of patients evaluated were considered symptomatic in terms of lack of sexual interest or enjoyment. These findings were most common among patients with postresection status (55.6%) and less common among patients in all other disease categories (≤ 20%). It is possible that patients postresection might experience sexual problems due to surgery-related difficulties, or these patients may be more likely to notice and report these concerns compared with patients undergoing long-term treatment. Sexual difficulties may arise from physical (eg, fatigue, pain), emotional, and interpersonal factors for patients with pancreatic cancer, leading to a loss of sexual interest and intimacy.^22,23 The precise nature of sexual problems experienced by patients in this study was not examined and should be addressed in future research.

There are several aspects of the current study that limit the generalizability of its findings. The findings reported are cross-sectional associations, and thus we cannot infer causality. One of the major limitations in our study is the limited QoL data that we have reported, with up to a quarter of the data missing clinical or staging data needed to conduct the analysis. This specifically applies to answers to the questions regarding sexuality. This could have led to a bias when reporting our QoL measures and in the statistical analysis. In future studies, longitudinal data should be used to assess changes in QoL in patients with pancreatic cancer across time. Finally, without a sufficient number of patients for multivariable analyses, we were not able to adjust for potential confounders such as tumor size, age, and comorbid disease.

Results from this cross-sectional study suggest that although all patients presenting to the PMDC report impairments in QoL, patients with more advanced pancreatic cancer are, in general, more likely to report worse symptoms on the EORTC QLQ-PAN26 questionnaire. QoL measures that are most impaired, such as digestive symptoms and pancreatic pain, may be improved with basic clinical interventions such as optimizing pain management or prescribing pancreatic enzyme supplementation. Ideally, QoL questionnaires will be frequently administered, and subsequent interventions will be tailored to a patient's specific needs. However, this is not always feasible given limitations in both time and resources in most clinics. Further efforts should be made to increase the use of QoL questionnaires and to investigate how symptom-directed interventions may improve QoL in patients with pancreatic cancer.

Authors' Disclosures of Potential Conflicts of Interest

Disclosures provided by the authors are available with this article at jop.ascopubs.org.

Author Contributions

Conception and design: Shalini Moningi, Amanda J. Walker, Charles C. Hsu, Jing-Ya Wang, Joseph M. Herman

Administrative support: Timothy M. Pawlik

Collection and assembly of data: Amanda J. Walker, Charles C. Hsu

Data analysis and interpretation: Shalini Moningi, Amanda J. Walker, Charles C. Hsu, Jennifer B. Reese, Katherine Y. Fan, Lauren M. Rosati, Daniel A. Laheru, Matthew J. Weiss, Christopher L. Wolfgang, Timothy M. Pawlik, Joseph M. Herman

Manuscript writing: All authors

Final approval of manuscript: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Correlation of Clinical Stage and Performance Status With Quality of Life in Patients Seen in a Pancreas Multidisciplinary Clinic

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or jop.ascopubs.org/site/misc/ifc.xhtml.

Shalini Moningi

No relationship to disclose

Amanda J. Walker

No relationship to disclose

Charles C. Hsu

No relationship to disclose

Jennifer B. Reese

No relationship to disclose

Jing-Ya Wang

No relationship to disclose

Katherine Y. Fan

No relationship to disclose

Lauren M. Rosati

No relationship to disclose

Daniel A. Laheru

No relationship to disclose

Matthew J. Weiss

No relationship to disclose

Christopher L. Wolfgang

No relationship to disclose

Timothy M. Pawlik

No relationship to disclose

Joseph M. Herman

Honoraria: Celgene, Merrimack

Research Funding: Nucletron (Inst)

Patents, Royalties, Other Intellectual Property: Elekta Corporation