Patterns of Self-Reported Symptoms in Pancreatic Cancer Patients Receiving Chemoradiation

Abstract

Pancreatic cancer is a lethal disease characterized by multiple disease-related symptoms. Chemoradiation therapy is a standard of treatment for locally advanced pancreatic cancer. Although shown to prolong survival, there is little information about treatment-related symptoms or the palliative benefits of chemoradiation. We assessed symptoms of patients with locally advanced pancreatic cancer receiving chemoradiation to determine the prevalence, and co-occurrence, of symptoms and to identify the extent to which symptoms interfered with function. Forty-eight patients were treated with chemoradiation on a Phase I protocol. Patients received radiotherapy (50.4 Gy in 28 fractions), capecitabine (median dose 825 mg/m² twice daily), and bevacizumab (2.5–10 mg/kg). Symptom severity and its interference with function were prospectively assessed (at presentation, during, and after chemoradiation) in 43 consenting patients using the M.D. Anderson Symptom Inventory. Results showed that 95% of patients reported at least one of the 13 symptoms assessed at presentation. The most commonly reported symptoms of moderate to severe (≥5 on a 0–10 scale) intensity at presentation were lack of appetite (24%), pain (19%), fatigue (19%), and sleep disturbance (10%). We observed an increase in patients reporting moderate to severe fatigue, nausea, and sleep disturbance during chemoradiation. McNemar tests for paired binary observations showed the proportion of patients reporting moderate to severe symptoms significantly (P < 0.001) decreased after chemoradiation at 94 days follow-up (lack of appetite = 7%, pain = 7%, fatigue = 13%, sleep disturbance = 7%). This study demonstrates the feasibility and usefulness of symptom assessment in chemoradiation protocols. Future studies with larger cohorts are needed to further characterize multiple symptoms associated with chemoradiation.

Introduction

Pancreatic cancer is one of the most lethal human cancers, representing the fourth leading cause of cancer-related deaths in men and the fifth leading cause of cancer deaths in women in the United States. Pancreatic cancer is refractory to standard treatments: the one-year survival rate is only 24% and the five-year survival rate is about 4%. Even for those patients with resectable disease, the five-year survival rate is only 17%.¹ It is estimated that in 2006, about 33,730 people in the United States will be found to have pancreatic cancer and about 32,300 will die of the disease.²

In the United States, the treatment and management of pancreatic cancer is based on the extent of disease at presentation, categorized as localized and potentially surgically resectable, locally advanced and surgically unresectable, or metastatic.³ Because pancreatic cancer disseminates to distant sites early in its natural history, most patients present at advanced stages of the disease. Chemoradiation is a standard treatment option for many patients presenting with localized disease.^4,5 For patients with locally advanced disease, randomized trials have demonstrated improvement in median survival with chemoradiation compared with either chemotherapy or radiation therapy alone.^6,7

In a disease with such a dismal prognosis, symptom management and quality of life are paramount considerations. Patients with pancreatic cancer usually present with symptoms such as weight loss, fatigue, and pain. The purpose of this study was to assess patterns of self-reported symptoms in patients with locally advanced pancreatic cancer before, during, and after chemoradiation. Using a prospectively administered multisymptom assessment tool, we determined the prevalence, persistence, and co-occurrence of symptoms and identified the extent to which symptoms interfered with function.

Patients and Methods

The study sample included 48 patients with locally advanced, inoperable pancreatic adenocarcinoma who were receiving chemoradiation therapy. After obtaining informed consent in conjunction with a Phase I clinical trial, we prospectively assessed symptoms before treatment, weekly during treatment, and during the first clinical assessment after chemoradiation. Radiotherapy was administered at a dose of 50.4 Gy in 28 fractions delivered to the primary tumor and peripancreatic adenopathy identified by CT scan over 5.5 weeks using a conformal technique as described previously.⁶ Elective nodal treatment was not used. Bevacizumab was administered two weeks before radiotherapy, every two weeks during radiotherapy (12 patients each at 2.5, 5.0, 7.5, and 10 mg/kg), and after radiotherapy until disease progression. Capecitabine was administered with radiotherapy on Days 14–52 at 650 mg/m² for 12 patients and at 825 mg/m² twice daily for 36 patients. Dose adjustment of capecitabine was mandated for Grade 2 nonhematologic toxicity and occurred in 43% of patients, mostly for gastrointestinal toxicity.

This study was approved by the Institutional Review Board at M.D. Anderson Cancer Center.

Assessment of Symptoms and Functional Impairment

Self-reported symptoms and symptoms’ interference with function were assessed using the M.D. Anderson Symptom Inventory (MDASI) before treatment, weekly during treatment, and during the first clinical assessment after chemoradiation. The MDASI is a brief measure of the severity and impact of cancer-related symptoms.⁸ The “core” MDASI is a 19-item scale consisting of 13 symptoms frequently reported by cancer patients: pain, fatigue, nausea, disturbed sleep, distress (emotional), shortness of breath, lack of appetite, drowsiness, dry mouth, sadness, vomiting, difficulty with remembering, and numbness or tingling. The MDASI also contains six items that describe the extent to which these symptoms have interfered with different aspects of the patient’s life during the past 24 hours: general activity, mood, walking ability, normal work (including both work outside the home and housework), relations with other people, and enjoyment of life. Both the symptom and interference items are rated on an 11-point scale, with 0 being “does not interfere” and 10 being “completely interferes.” Based on previous studies on pain and fatigue,^9,10 we provisionally categorized symptoms as mild (ratings of 0–4), moderate (ratings of 5–6), or severe (ratings of 7–10). The development and validation process for the MDASI have been reported elsewhere.^8,11

Patients were asked to complete the MDASI at baseline before starting chemoradiation, weekly during chemoradiation, and monthly after the completion of chemoradiation. The first follow-up after treatment was coincident with the restaging CT scan (around day 94 from initiation of therapy). Patients’ hospital records were reviewed for age, sex, tumor location, histology, and comorbid conditions.

Statistical Analyses

Descriptive statistics were used to characterize the study population. We determined the proportion of patients reporting moderate to severe (score of ≥5 on a 0–10 scale) symptoms before, during, and after chemoradiation. The difference between the proportions for the most prevalent symptoms before and after chemoradiation was tested with the McNemar test for paired binary observations.

We used hierarchical cluster analysis to identify co-occurring/coexisting self-reported symptoms. Clusters or groupings of symptoms were formed using the average linkage (centroid method) between symptom items, and the distances between symptom items were calculated using squared Euclidian distances.¹²

We used mixed model methods to assess to what extent symptoms are associated with reported interference with function. Mixed models take into consideration the correlation between symptoms collected from the same subject.¹³ We assumed an autoregressive of order 1 correlation structure in the model, in which symptoms assessed closer in time are assumed to have a larger correlation compared with those farther apart in time. We applied the Aikaike Information Criterion to determine the final model.¹³ Symptoms, comorbid medical conditions, and demographic variables were used as potential predictors. Further variable selection in the model was conducted by using backward elimination. With the goal of having the most parsimonious model, only variables with P < 0.05 were included in the final model. We limited the post-treatment analysis to the first follow-up clinic visit.

A significance level of 5% (two-sided) was used for all the analyses. Because of the multiplicity of tests conducted, this significance level may result in false positive results, but this was considered acceptable for an exploratory study.¹⁴ Analyses were conducted using SPSS¹² and SAS.¹³

Results

The study sample included 48 patients with locally advanced, inoperable pancreatic adenocarcinoma receiving chemoradiation therapy. Patient characteristics are shown in Table 1. The median age was 58 years (range 41–80 years). Fifty-two percent were men, and 92% were Caucasian. Sixty-nine percent of the tumors were located in the head/uncinate process of the pancreas, 13% located in the body of pancreas, 8% in the neck of pancreas, and 10% in multiple locations. Ninety-four percent of the tumors were unresectable because of abutment or encasement of the celiac axis or the superior mesenteric artery. Seventy-four percent of the patients reported comorbid conditions (having a history of any of the following diseases: diabetes, heart, lung, liver, kidney, stroke).

Table 1.

Characteristics of the Patient Population

Characteristics	Number of Patients (n = 48)
Age (years)
Median	58
Range	41–80
Sex
Male	25 (52%)
Female	23 (48%)
Race/ethnicity
White, Anglo	44 (92%)
Blacks	2 (4%)
Asian/Pacific Islander	1 (2%)
Hispanic origin	1 (2%)
Location of tumor
Head or uncinate process	33 (69%)
Neck	4 (8%)
Body	6 (13%)
Multiple locations	5 (10%)
Karnofsky performance status
100	14 (29%)
90	32 (67%)
80	2 (4%)
Comorbidity
Present	32 (74%)
Absent	11 (26%)

The MDASI were all completed during clinic visits. Five patients refused to complete the MDASI forms, resulting to a total sample of 43 patients with symptom data. Patients who refused to complete the MDASI forms did not differ from the study population in demographics, stage of disease, and other variables that might suggest a selection bias. Thirty-seven patients completed the MDASI at baseline (before initiation of chemoradiation), for a response rate of 86% (37/43). Weekly symptom assessment was also conducted during the entire course of treatment, with an overall response rate (number of completed weekly forms to the number of forms expected) of 72%. After treatment, patients were asked to complete the MDASI every month when they visited the clinic. Thirty patients completed the MDASI on their first follow-up, during their CT restaging. We used each patient’s first post-treatment assessment in this analysis. We did not observe statistically significant differences in survival and disease progression between those with a higher proportion of missing data and those with complete data (analyses not shown).

Prevalence of Moderate to Severe Symptoms

Overall, 95% of the patients reported at least one of the 13 symptoms assessed at initial presentation. Fig. 1 shows the mean scores for these symptoms. The highest mean score was observed for fatigue, followed by lack of appetite and pain. Fig. 2 shows that the most commonly reported symptoms at moderate to severe intensity over the course of the study (initial presentation, during treatment, and after treatment) were fatigue, lack of appetite, pain, nausea, and sleep disturbance. The proportion of patients reporting symptoms at moderate to severe intensity at initial presentation include the following: lack of appetite (24%), pain (19%), fatigue (19%), sleep disturbance (10%), and nausea (4%). Patients reporting moderate to severe pain decreased during treatment (15%), and increased for fatigue (25%), nausea (13%), and sleep disturbance (12%). There was no change for lack of appetite (23%). After treatment, the proportion of patients reporting moderate to severe symptoms decreased relative to baseline (7% for lack of appetite, 7% for pain, 13% for fatigue, 3% for nausea, and 7% for sleep disturbance). McNemar test for paired binary observations before and after chemoradiation was statistically significant for these symptoms (P < 0.05).

Fig. 1.

Mean score of symptoms assessed before chemoradiation (Note: all symptom items were assessed using an 11-point (0 = none; 10 = as bad as you can imagine) numeric rating scale).

Fig. 2.

Prevalence of moderate to severe symptoms (≥5, on a 0–10 scale) before, during, and after chemoradiotherapy (n = 43) (Note: all symptom items were statistically significant (P < 0.001) for McNemar test for paired binary response before and after chemoradiation (n = 27). No statistically significant (P > 0.05) differences for survival and disease progression for patients with complete data versus those with missing data Prevalence during treatment was computed as any report of >5 symptom severity during the chemoradiation period).

We also assessed factors associated with symptom severity and found that the presence of a comorbid medical condition was a significant factor for symptom severity (P < 0.01). However, there was no significant variation by sex, age, or type of histology, possibly due to the small sample (analyses not shown). We also assessed for associations between symptom improvement and radiographic response criteria (CT scan) and found no significant association between these two variables (analyses not shown).

Prevalence of Multiple Symptoms/Symptom Clusters

We also determined the proportion of patients reporting multiple coexisting symptoms. Table 2 shows the total number of symptoms reported at moderate to severe intensity at presentation, during, and after chemoradiation. We observed, that at presentation, at least 24% of our patients had two or more symptoms of moderate to severe intensity. During therapy, up to 51% of our sample reported multiple symptoms that were of moderate to severe intensity; after therapy, the proportion decreased to 20% (McNemar test for paired binary observations was significant for before and after treatment).

Table 2.

Prevalence of Moderate to Severe Symptoms Before, During, and After Chemoradiation^a

	Before Chemoradiation		During Chemoradiationb		After Chemoradiaton
Number of Moderate to Severe Symptoms	n	%	n	%	n	%
0	21	58	14	34	20	66.7
1	6	16.8	5	12	4	13.3
2	1	2.8	1	2.4	3	10
3	2	5.6	6	14.6	0	0
4	2	5.6	0	0	2	6.6
5	1	2.8	2	5	0	0
6	1	2.8	4	10	1	3.3
7 or greater	2	5.6	9	22	0	0
Total	37	100	41	100	30	100

^aSignificant McNemar test (P < 0.05) for difference between proportion reporting two or more symptoms before and after chemoradiation.

^bPrevalence during treatment was computed as number of patients reporting symptoms of >5 symptom severity during the chemoradiation period.

We also determined which symptoms typically co-occurred using hierarchical cluster analyses. We found that over the course of the study (baseline, during treatment, and one month after treatment), fatigue and lack of appetite formed a distinct grouping, suggestive of a strong interrelatedness between these two symptoms.

Symptom Severity and Interference with Function

The MDASI includes six items that describe to what extent symptoms have interfered with function, including general activity, mood, walking ability, normal work (including both work outside the home and housework), relations with other people, and enjoyment of life. Using mixed model methods of analyses, we found that lack of appetite, fatigue, sleep, and pain were the symptoms that accounted for variation in these interference items (Table 3). Fatigue accounted for the largest variation across the different interference items.

Table 3.

Symptoms Significantly Associated with Interference Items

Interference Items		Predictors
Walking	Fatigue 0.2690 (P < 0.0012)	Pain 0.1632 (P < 0.017)	Lack of appetite 0.1818 (P < 0.010)
Activity	Fatigue 0.441 (P < 0.001)	Nausea 0.162 (P < 0.002)	Lack of appetite 0.178 (P < 0.0055)
Mood	Lack of appetite 0.2325 (P < 0.0001)	Sleep disturbance 0.1291 (P < 0.0194)	Nausea 0.1157 (P < 0.021)
Enjoyment	Fatigue 0.385 (P < 0.0001)	Lack of appetite 0.2140 (P < 0.0002)
Work	Fatigue 0.5179 (P < 0.001)
Relations	Pain 0.2045 (P < 0.001)

Coefficient (P-value).

Interference items were assessed using an 11-point (0 = “does not interfere” and 10 = “completely interferes”) numeric rating scale.

Discussion

This study demonstrates the feasibility and usefulness of symptom assessment in chemoradiation protocols. We had a good completion rate and were able to determine patterns of symptoms and symptom-related interference and found a decreased prevalence of symptoms at 94 days follow-up.

We observed that although moderate to severe pain, fatigue, lack of appetite, nausea, and sleep disturbance were prevalent at initial presentation, there was a decrease in the proportion of patients reporting these symptoms after chemoradiation therapy. Although assessed only to a limited extent, studies have explored the palliative effects of cancer therapies on disease-related symptoms. Buris et al.,¹⁵ using information on analgesic consumption and pain intensity, showed in a randomized trial that gemcitabine is effective (compared to 5-fluorouracil) in alleviating disease-related symptoms in patients with advanced pancreatic cancer. The palliative benefits of radiation have also been shown in other studies of patients with locally advanced pancreatic cancer and other gastrointestinal malignancies.^16-22 Minsky et al. found that radiation provides pain relief in 50%–85% of patients with pancreatic cancer.²³ More recently, Shinchi et al. also have shown pain relief in 80% of 10 patients with locally advanced pancreatic cancer who had chemoradiation.²⁴ Our study has extended these findings by providing preliminary evidence of the palliative benefits of chemoradiation to include symptoms other than pain, such as fatigue, lack of appetite, nausea, and sleep disturbance.

We also showed that patients who presented with symptoms (particularly fatigue, lack of appetite, and nausea) were more likely to report interference with several activities of daily living over the course of chemoradiation and even after treatment. Given the poor overall outcome and short survival of patients with pancreatic cancer, treatments should be well tolerated. The advent of conformal radiation and targeted therapies offers the promise of treatment regimens that will be well tolerated by patients. Although patients objectively appeared to tolerate this experimental regimen well, symptom data revealed that a subset experienced significant worsening of their symptoms during treatment, especially fatigue and lack of appetite. This finding highlights the need for regular symptom assessment, and for the development of effective therapy for fatigue and other symptoms.

Consistent with findings in other studies, the majority of our patients had multiple coexisting symptoms (two or more symptoms). Studies show that patients with advanced pancreatic cancer have a high incidence of tumor-related symptoms and that patients undergoing treatment (i.e., chemotherapy or radiotherapy) report significant fatigue and other symptoms during the course of treatment.^8,25-27 A study by Greenberg et al.²⁸ showed that localized radiation treatment is associated with increased fatigue and sleep requirement. In this study, we found that patients reporting fatigue also reported lack of appetite over the course of treatment.

Importantly, the use of a multisymptom assessment instrument allowed us to identify symptoms other than fatigue and to observe reduction in the proportion of patients reporting two or more symptoms of moderate to severe intensity, after chemoradiation. Among the significant implications of this finding is the need for using multisymptom assessment instrument, so that we can identify other important symptoms requiring management.

We found that the presence of other chronic comorbid medical conditions was a significant factor for the prevalence and severity of symptoms. This is not surprising because chronic diseases, such as diabetes^29,30 and cardiovascular disease,³¹ are associated with symptoms such as pain and fatigue. A large proportion of our patients presented with coexisting medical conditions; thus, symptom management can be a considerable challenge for this patient population.

To our knowledge, this is one of the first reports that has used a previously validated instrument to prospectively assess self-reported symptoms in patients with locally advanced pancreatic cancer treated with chemoradiation. The high prevalence and persistence of symptoms and their adverse effect on function highlight the importance of prospective symptom assessment.

Our study has potential limitations. Our results may have been affected by selection bias. Our treatment facility is a tertiary center and has patients not typical of other institutions (primarily white, with higher levels of socioeconomic status), and the patients in the trial also had high performance status; thus our findings may have limited generalizability. We acknowledge that the correlations identified do not reflect causality and that there may be other reasons for the observed improvement in symptoms, including powerful placebo effects related to the treatment or treatment setting. Furthermore, because we lack data on symptom therapies (e.g., pain medication or analgesic consumption), we were not able to assess whether the decrease in the prevalence of symptoms may be due to medications rather than chemoradiation. We also studied a small number of patients and, as a result, some of the reported analyses may have failed to reach statistical significance because of sample-size limitations and Type II error.

In conclusion, this study provides evidence on the importance of assessing multiple symptoms during chemoradiation. We had a good completion rate and were able to determine patterns of symptoms and symptom-related interference and found a decreased prevalence of symptoms at 94 days follow-up. Future studies with larger cohorts are needed to further characterize relief of symptoms associated with chemoradiation. As pancreatic cancer is a lethal disease associated with multiple debilitating symptoms, symptom surveillance, treatment, and control are important goals for this population.