|Home | About | Journals | Submit | Contact Us | Français|
We previously demonstrated that select cytokine gene polymorphisms in interleukin (IL)-8 are a significant predictor for pain and analgesia in patients with lung cancer. This study explores the role of thirteen potentially functional polymorphisms in cytokine genes including IL-1β, IL-6, IL-8, IL-10, IL-18, tumor necrosis factor (TNF-α), and nuclear factor kappa-B subunit 1 (NFkappaB1) in pain severity in patients with pancreatic cancer. We evaluated a series opatients with histologically-confirmed adenocarcinoma of the pancreas (n=484) who had completed a self-administered survey of pain prior to initiating any cancer treatment. DNA (n=156) available for a subset of white patients was assayed and assessed for association with pain severity. Results showed that 26% (128/484) reported experiencing severe pain (score of > 7 on a 0–10 scale). Severe pain varied by stage of disease (odds ratio [OR] Stage II=4.02, 95% confidence interval (CI)=1.07, 15.07; Stage III=5.02, 95% CI=1.28, 19.61; Stage IV=6.90, 95% CI=1.96, 24.29), ethnicity (OR non-Hispanic blacks=3.67; 95% CI=1.44, 9.38), reports of depressed mood (OR=1.94; 95% CI=1.09, 3.43), and female sex (OR=1.78; 95% CI=1.04, 3.05). Controlling for these covariates, IL8-251T/A (OR=2.43, 95% CI=1.3, 4.7, P<0.009) significantly predicted severe pain in a subset of white patients. When we adjusted for reported analgesic use, we found that IL8-251T/A persisted as a predictor for severe pain, with carriers of TT and AT genotypes having more than a threefold risk (OR=3.23, 95% CI=1.4, 4.7) for severe pain relative to the AA genotypes. We provide preliminary evidence of the role of IL-8 in the severity of pain in pancreatic cancer patients. Additional studies are needed in larger cohorts of patients.
Pancreatic cancer is the fourth leading cause of adult cancer-related deaths, with many patients presenting with severe and often debilitating symptoms. An understanding of the extent to which genetic variability plays a role in severe pain may prove useful in identifying patients at high risk for pain and, importantly, may help in developing individualized therapy for patients who will benefit most from symptom intervention.
There is compelling evidence that cytokines play an important role in the pathogenesis of cancer-related pain. Cytokines released during inflammation or tissue damage (as in the cancer process) modify the activity of nociceptors contributing to pain hypersensitivity. Pro-inflammatory cytokines released by activated glial cells in response to inflammation or other cytokines (as produced by cancer) have been shown to cause hyperexcitability in pain transmission neurons. The resulting enhanced release of substance P and excitatory amino acids from presynpatic terminals also produces an exaggerated pain response.1–6 Activation of glia in CNS regions due to cancer is also linked to the production of sickness responses. Chemokines may also produce enhanced sensitivity to pain via direct actions on chemokine receptors expressed by nociceptive neurons.7
Single nucleotide polymorphisms (SNPs) in cytokine genes have been shown to alter their expression or function and thus may serve as a stable marker for the risk of cancer-related pain. For example, the –251T/A SNP in the promoter of the interleukin-8 (IL8) gene, shown to be associated with altered gene expression, is associated with pain severity in lung cancer patients8. However, previous genetic association studies have selected only one or at most a few polymorphisms of putative genes for cancer-related pain. In this study, we assessed the role of thirteen polymorphisms in cytokines genes including IL-1β, IL-6, IL-8, IL-10, IL-18, Tumor necrosis factor (TNF-α), and nuclear factor kappa-B subunit 1 (NFkappaB1) as correlates of pain severity in patients with adenocarcinoma of the pancreas.
Patients with newly diagnosed, previously untreated and histologically confirmed adenocarcinoma of the pancreas who registered at the University of Texas M.D. Anderson Cancer Center (MDACC) for cancer treatment during the period 06/2002-07/2006 comprised the study sample (Figure 1). This study was approved by the Institutional Review Board at M. D. Anderson Cancer Center.
Patients completed a standardized self- administered questionnaire at patient registration, as part of their initial work-up. The questionnaire included items on basic demographic information, history of comorbid conditions and other clinical variables.
Patients rated their pain “during the past week” on an 11-point numeric scale, (0= no pain and 10= pain as bad as you can imagine).9 We also assessed depressed mood using the item “During the past 4 weeks, have you felt downhearted and blue?”10–13 Rating of symptoms was reported at time of presentation at M. D. Anderson, and prior to any cancer treatment. Clinical data, including stage of disease, were abstracted from patients’ medical records.
Charts were reviewed for information on opioid dose by a supportive care specialist (S.Y.). Because of the different types of opioids prescribed, we translated the daily opioid dose to morphine equivalent daily dose (MEDD). We used the conversion factors shown in Table 1 and calculated the total dose of opioids.
The University of Texas M. D. Anderson Cancer Center is a member of The Texas Medical Center Genetics (TexGen) Consortium, supported by philanthropic funds. This longitudinal multi-institutional initiative collects and stores clinical data and specimens from patients to evaluate genetic predictors of outcome and response to therapy. Patients are approached for the study and blood samples are obtained at the same time as the patient’s diagnostic blood work. Blood samples are placed on ice immediately upon collection and stored at −80°C, formalin-fixed, etc. The majority of samples are processed within the same working day in the TEXGEN laboratory. For this study, archived DNA was requested from the TEXGEN repository.
One hundred fifty-six white patients had DNA available for genotyping analyses. We genotyped polymorphisms in cytokine genes including IL-1β, IL-6, IL-8, IL-10, IL-18, TNF-α, and NFkappaB1, chosen on the basis of their known biological processes, reports in the literature, the type of SNPs, i.e., nonsynonymous, 3′UTR, 5′UTR, promoter regions; prediction of functionality, and minor allele frequency >0.05%.14 SNP genotype was determined in multiplex using Applied Biosystems SNPlex genotyping technology.
Descriptive statistics were used to summarize patient characteristics. The Kolmogorov-Smirnov Z test was used to assess normality distribution for pain severity.
Comparisons across groups were performed using Pearson’s Chi-square, Kruskall-Wallis or analysis of variance (ANOVA). Multiple logistic regression analysis was performed to explore the significant correlates of severe pain. We used the National Comprehensive Cancer Network cut-off score for severe pain;15 a score > 7/10 is considered as a pain emergency and treatment is initiated with short-acting opioids. Variable selection for potential covariates (clinical, epidemiological, and genetic correlates) was conducted by using a step-wise selection method with entry testing based on the significance of the score statistic. We also assessed if an observed association between genetic factors and pain severity may be due to differences in the use of analgesics.
The study sample included 484 newly diagnosed patients with adenocarcinoma of the pancreas. The mean age of diagnosis was 62 years (standard deviation [SD]=10.5). There were more male (58.5%) and non-Hispanic whites (82%). Almost half of the patients presented with advanced stage of disease (43% Stage IV). Sixty-six percent of the tumors were found to be located in the head of the pancreas. Fifty-one percent reported a co-morbid condition, with hypertension being the most prevalent comorbidity (42%), followed by diabetes (25%).
Ninety percent of the sample reported some level of pain. The mean level of pain was 4.7 (SD=3); the median was 5 and the mode was 3. The Kolmogorov-Smirnov Z test showed that pain severity was not normally distributed for white patients (Z=2.501; P=0.0001). Using > 7 as a cut-off for severe pain, we found 26% of the sample reported experiencing severe pain.
Table 2 summarizes the distribution of severe pain with respect to clinical and epidemiologic variables. More women (52%) reported severe pain than men (48%; P<0.005). African Americans (12%) were overrepresented in the severe pain group versus 4% in the moderate pain group (P=0.007). We also observed that the proportion of severe pain differed by stage of disease (P<0.007), suggesting the need to control for these variables in the multivariable analyses. There was no difference in pain experienced by tumor location within the pancreas, nor by smoking status or comorbidity. Predictably, those reporting a depressed mood were significantly more likely (P=0.001) to report severe pain. For the multivariable model, we found that stage of disease (odds ratio [OR] Stage II=4.02, 95% confidence interval [CI]=1.07, 15.07; Stage III=5.02, 95% CI=1.28, 19.61; Stage IV=6.90, 95% CI=1.96, 24.29), ethnicity (OR non-Hispanic blacks=3.67, 95% CI=1.44, 9.38), reports of depressed mood (OR=1.94; 95% CI=1.09, 3.43), and female sex (OR=1.78, 95% CI=1.04, 3.05) remained as significant predictors of pain severity.
One hundred fifty-six white patients had DNA available for genotyping analyses. We limited our analyses to white patients to avoid the potential confounding from population stratification. Patients with DNA did not differ from the study population in demographics, stage of disease, and other variables that might suggest a selection bias. (Table 3 shows the distribution of each genotype by pain severity levels.) We found that IL8-251T/A (TT=18.2%; TA=33.3%; AA=48.5%; P<0.026) and IL1 B-31T/C genotypes (TT=30.3; TC+CC= 69.7%; P<0.044) were significantly correlated with severe pain on univariate analysis.
We further assessed the extent to which polymorphisms in IL8 and clinical and epidemiological factors influence pain severity and found that the polymorphism in IL8 was the most important predictor of severe pain (OR=4.82; 95% CI=1.48, 15.74), followed by sex (women: OR=3.82; 95% CI=1.02, 14.28).
Because analgesic use may confound the observed relationship between IL8-251T/A and pain severity, we reviewed charts for patients’ report of current medications (same day that pain was assessed). We found that the mean MEDD was 28.71mg/24h (SD=97.91). We did not observe any significant association between MEDD and severe pain (non-severe= 24.4 versus severe=45.61; P>0.26). Table 4 shows that the significant association between IL8-251T/A and severe pain persisted even after adjusting for MEDD, sex, age, stage of disease, and comorbidities.
We studied several SNPs and a false association could arise due to multiple comparisons. Therefore, we calculated the false positive report probability (FPRP) for the IL8-251T/A association that was found to be significant. The FPRP is the probability that the significant finding is false. FPRP calculations depend on the observed P-value for the association, prior probability that the association between the genetic variant and the disease is real and the statistical power of the test.16 For our analyses, we assumed a range of prior probabilities 0.05 to 0.10. These are considered moderate range of prior probabilities. For the statistical power calculations, we used odds ratios of 1.4, 3.23, and 7.7, representing the lower confidence interval, point estimate, and upper confidence interval of our observed association between IL8-251T/A and pain severity (Table 4), respectively. The noteworthiness of an association is defined as having FPRP value below 0.2. Table 5 provides the noteworthiness of our significant association under different prior probabilities. From this table, we see that our observed association is noteworthy for prior probabilities that are greater than 0.07 for our observed OR=3.23.
We found that a functional polymorphism in the IL8 gene is a significant correlate of pain severity, independent of stage or any of the clinical or demographic variables. This is consistent with our previous study17 in lung cancer patients, which showed IL8-251T/A as a significant correlate of pain severity. Importantly, we found that the association persisted even after adjusting for reported analgesic use. Taken together, these studies provide further support of the biological importance of polymorphisms in IL8 as a proximate and stable marker for pain severity in patients with cancer.
IL-8 is located on 4q13-q21. The IL8-251T/A has been shown to affect IL-8 levels, with the A-allele associated with increased IL-8 production.18 As a major mediator of the inflammatory response, IL8 serves as a chemical signal that attracts neutrophils. Elevated IL8 levels have been demonstrated in patients with chronic pain conditions such as chronic back pain19 and post-herpetic neuralgia. Among pancreatic cancer patients, higher IL-8 levels were associated with poor performance status and weight loss.20 The association of IL8 in the pathogenesis of pain is hypothesized to result from their direct involvement in nociceptive signal transduction. Recent investigations suggest that dorsal root ganglion neurons are actively involved in chemokine signaling and that chemokines act as messengers between peripheral immune cells and sensory afferent neurons at inflamed sites. In animal studies, Oh et al.7 observed that isolated dorsal root ganglion neurons express the chemokine receptors and further, they observed that chemokines (IL8) also produced allodynia after injection into the rat paw. In an earlier study, Cunha et al.21 measured the hyperalgesic effect of IL8 in a rat paw pressure test and found that IL-8 evoked dose-dependent hyperalgesia that was blocked by specific anti-IL-8 serum suggesting that IL-8 causes hyperalgesia. It was also observed that intraperitoneal administration of a specific antiserum against IL-8 partly blocked the nociceptive response in mice.22 Chemokines also are suggested to produce enhanced sensitivity to pain via direct actions on chemokine receptors expressed by nociceptive neurons.23
Variants in cytokine genes, including IL-1β, IL-6, IL-10, IL-18, TNF-α, and NFkappaB1 have been shown to be significant correlates of pain in other studies.24–27 In univariate analyses, we found a significant association for polymorphisms in IL-1β and a borderline association of IL10 with pain but these associations did not persist in the multivariable analyses. IL-1 participates in activating T cells and inducing the expression of adhesion molecules and has been implicated in pain response.24,25 IL10 inhibits cytokine production by macrophages and explain accessory functions of macrophages during T cell activation and is considered to play a significant role in pain therapy.27,28 Variants in IL6 and TNF-α were not significantly associated with pain in this analysis. However, with our small sample, we had low power to detect these associations.
Consistent with other studies, we found a high prevalence of severe pain at presentation in patients with adenocarcinoma of the pancreas. We found that women and minorities had higher risk for severe pain and that the risk persisted even after controlling for stage of disease, suggesting the greater need for symptom surveillance, treatment and control for these subgroups. It is well established that stage III and IV patients have a higher incidence of severe pain because unresectable tumors invade nerves. We also confirmed the significance of depressed mood as a covariate of pain severity, consistent with other clinical and population-based studies of pain, suggesting that depressed mood should always be assessed and treated in order to improve upon pain and the general well-being of cancer patients.
Among the strengths of this study is that newly diagnosed patients with adenocarcinoma of the pancreas with no previous cancer treatment were assessed for pain, thus providing a representative view of the prevalence of pain that is not treatment-related in pancreatic cancer patients.
There are limitations to our study. It should be noted that we do not have information on symptoms such insomnia/sleep disturbance, lack of appetite, delirium etc. all of which may have an impact on pain severity. It could also be argued that because we assessed 13 SNPS and given the small sample, the association we observed may be false positive findings (observed P=0.009 versus with Bonferroni correction P=0.003). However, in conducting the FPRP analyses, we found that our observed association is noteworthy.
Furthermore, the results of these studies should be replicated. Chanock and colleagues29 have noted that although researchers use a priori knowledge of polymorphisms and gene functions in the candidate gene approach these studies have yielded informative but conflicting results. Prospective studies with larger studies are needed to further elucidate causal relationships between cytokines and pain.
Advances in molecular technology now allow us to assess genetic variations that are stable markers of symptom severity, prognosis and outcome.30 Studies that elucidate the independent contributions of the disease process and host factors involved in symptom outcomes have high clinical significance. In the future, we anticipate that genotyping could become an integral component of an individualized treatment program for patients with pancreatic cancer. Because genetic polymorphisms are stable markers and easily and reliably assayed, understanding the extent to which genetic variability plays a role on symptoms may prove useful in identifying cancer patients at high-risk for symptom development and importantly, could help in understanding patients who might benefit most from symptom intervention, and ultimately in developing personalized and more effective symptom therapies.
This work was supported by a Public Health Service grant CA109043 and CA128069 from the National Cancer Institute. Patient consent and acquisition of blood specimens were supported by the Various Donor Fund for Pancreatic Cancer Research and National Institutes of Health grant CA101936 (SPORE in Pancreatic Cancer), The Texas Medical Center Genetics (TEXGEN) Consortium, and the Center for Clinical and Translational Sciences (CCTS) of the University of Texas Health Science Center at Houston.
We thank Erika Thompson, MS, Co-Director, DNA Analysis Core Facility; Dr. Luis Vence, Co-Director, Immune Monitoring Core Laboratory; and Dr. Kelly Merriman and Elizabeth Thompson, MS, Department of Epidemiology, The University of Texas M. D. Anderson Cancer Center.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.