Relationship of inflammatory markers and pain in patients with head andneck cancer prior to anticancer therapy
Abstract
Introduction
Clinical and epidemiological studies suggest an association between chronic inflammationand certain types of cancer (1), including cancerof the head and neck (HNC) (2). The tumormicroenvironment is rich in cytokines and other inflammatory mediators that influenceimmunosuppression, cancer cell growth, tissue remodeling, and angiogenesis (3). Cancer and immune cells produce and secretemediators that activate and sensitize primary afferent nociceptors, resulting in pain(4).
Pain is a common symptom in cancer patients, including those with HNC (5,6).Inflammation is widely recognized as both a cause of pain and a possible therapeutictarget for treating pain (7). Inflammatorymediators such as proinflammatory cytokines and C-reactive protein (CRP) can modulatenociception (8) and contribute to theamplification and persistence of pain in cancer patients (4,9).
CRP is an acute-phase protein and a marker of inflammation. The synthesis of CRP inhepatocytes may be regulated by proinflammatory cytokines such as interleukin-1 (IL-1),IL-6, and tumor necrosis factor-alpha (TNF-α). CRP and proinflammatory cytokines arepresent at increased levels in various malignancies (10). An association between elevated serum CRP, TNF-α and IL-6 has beendemonstrated in HNC patients (11-13), and these inflammatory mediators may be asource of pain around the tumor site (5).However, there have not been any published research studies associating pain with CRPand TNF-α levels in HNC patients prior to anticancer therapy, a time when cliniciansmight employ therapeutics to prevent or reduce the occurrence of persistent pain. Thus,the purpose of this study was to evaluate the association between serum CRP, TNF-αlevels, and perceived pain in HNC patients prior to anticancer therapy.
Material and Methods
Patients
This study was approved by the Institutional Ethics Committee for human subjects(UFES #99.242/2012). We evaluated 127 patients with primary head and neck squamouscell carcinoma who had undergone medical examinations at the Santa Rita de CassiaHospital, AFECC, Vitória, ES, Brazil, and 9 healthy volunteers as controls. Patientswere excluded if they had already been treated for cancer, had recurrent malignantdisease, were unable to speak Portuguese, had impaired functional status thatprevented them from answering the questionnaires, or were younger than 18 years ofage. Clinical data (gender, age, tumor location, and tumor stage) were obtained frommedical records. Self-reported “average pain” in the past 24 h, as assessed using theBrief Pain Inventory (BPI) (14), was used toallocate the cancer patients to “pain” or “no pain” groups.
Inflammatory marker assays
Blood samples were collected and centrifuged at 4°C and 1000 g for15 min. The sera were immediately frozen in liquid nitrogen for storage at -80°Cuntil analysis of CRP and TNF-α concentrations.
The serum concentrations of CRP were determined using a Dimension¯ RxLMax¯ autoanalyzer (Siemens Healthcare Diagnostics, Germany) using theCRP extended range (RCRP) Flex¯ reagent cartridge (SiemensDimension¯, USA), which is based on the particle-enhanced turbidimetricimmunoassay (PETIA) technique. The results are reported in mg/L.
Serum TNF-α concentrations were determined using standard ELISA protocols accordingto the kit manufacturer's specifications (Human TNF-α UltraSensitive kits,Invitrogen-BioSource International, USA). The results are reported in pg/mL.
Statistical analysis
Data are reported as means±SD. Mean values were compared using one-way analysis ofvariance (ANOVA) followed by the Fisher least significant difference posthoc comparison. Spearman correlations were used to determine associationsbetween self-reported pain and inflammatory markers. P<0.05 was considered to bestatistically significant. The statistical analyses were performed using SPSSStatistics for Windows, Version 17.0 (SPSS Inc., USA).
Results
The main features of our series of 127 HNC patients are summarized in Table 1. The mean age of the patients was 57.9years, and most were males. The most frequent primary tumor site was the oral cavity,the majority of the patients had advanced cancers (stages T3 and T4), and 59.1% reportedthe presence of pain.
Figure 1 reports serum CRP and TNF-α levels inuntreated HNC patients with pain, with no pain, and controls. The median serum CRPlevels were higher in patients with pain (n=35, P<0.01) than in those notexperiencing pain (n=54, P<0.01), or in controls (n=9, P<0.01). The median serumTNF-α levels were higher in patients with pain (n=32, P<0.05) than in those notexperiencing pain (n=22, P<0.05), or in controls (n=9, P<0.05).
Spearman correlations (Table 2) revealedpositive associations between pain, CRP (P<0.05) and tumor stage (P<0.01). TNF-αlevels were positively correlated with pain, but the degree of association was notsignificant.
Discussion
In this cohort of HNC patients prior to anticancer therapy, the levels of TNF-α and CRPin those reporting pain were higher than the levels observed in patients without pain,as well as in control subjects. The presence of pain was positively and significantlyassociated with both CRP level and tumor stage. Normally, cancer pain is classified intothree categories: pain caused by tumor growth, pain caused by treatment, and painunrelated to cancer (15). We excluded pain causedby treatment because the evaluation of our patients was performed before any type ofcancer treatment. We also excluded pain unrelated to cancer because all patientsreported pain related to the lesion. Tumor growth may cause pain by compressing andinvading surrounding tissues, including muscle, bone, and peripheral nerves.Additionally, the head and neck site has a rich blood supply and a large number ofnerves that may affect tumor growth and pain (15,16).
Studies suggest that inflammation caused by tumor-induced mediators such as CRP andcytokines (4,9) may be a potential cause of cancer-related pain. Although the exactmolecular mechanism by which cytokines influence pain is not fully elucidated, studiessuggest that cytokines released during inflammation and tissue damage, as in the case ofcancer, modify the activity of nociceptors and thus contribute to hyperalgesia (17).
Studies have shown an increase in preoperative CRP levels in patients with squamous cellcarcinoma of the head and neck (13,18) and reveal that TNF-α can stimulate theproduction of CRP in hepatocytes (19). However,no studies have explored the relationship between the inflammatory activity of thesemediators and pain in HNC patients prior to cancer treatment. A significant, positivecorrelation between cancer pain and CRP has been demonstrated (9), but not in HNC patients. Recently, in breast cancer, higherlevels of CRP were detected in women with pain than in women without pain, prior toanticancer therapy (20). Furthermore, otherstudies have found a relationship between higher CRP level and worse overall outcome inpatients with oral squamous cell carcinoma and HNC (13,18). Here, we show for the firsttime that the presence of pain in patients with HNC is positively associated with CRPlevel.
In this study, TNF-α levels were higher in patients with pain, but the correlation wasnot significant. However, TNF-α stimulates immune cells to produce nociceptive agentsthat interact with primary afferent nociceptors in the cancer microenvironment (4), and there is evidence in the literatureregarding the involvement of TNF-α in tumor progression (11,21) and in the development of painin some cancers (4,22).
Clinical studies in HNC show increased levels of IL-1β, IL-6, IL-10, and TNF-α, anddecreased levels of IL-12 (11,12,23). Amore recent study found high levels of cytokines, including TNF-α, in the saliva of HNCpatients treated with radiotherapy (24), but thatresult was not related to pain, and the study was conducted during administration ofantineoplastic therapy. In another study, no significant differences were found in thelevels of TNF-α in the serum and saliva of oral cancer patients compared with controls,but high levels of IL-1β and IL-6 were detected (25).
Furthermore, TNF-α has been associated with other cancer symptoms such as fatigue,insomnia, and cognitive impairment (26,27). Given that treatment with cisplatin, paclitaxelor radiation increases levels of proinflammatory cytokines such as TNF-α and IL-1β(28-30), clinicians might employ interventions to reduce the symptoms prior toanticancer therapy.
In animal models, the administration of the proinflammatory cytokines TNF-α and IL-1βinduces pain behavior (31), and treatment withanti-inflammatory cytokines or proinflammatory cytokine inhibitors has been shown toreduce pain (32,33). Human studies show a possible pathogenic role of cytokines in pain.Samples of different fluids (blood and cerebrospinal fluid) and tissue (skin and nerve)indicate that an imbalance between proinflammatory and anti-inflammatory cytokines is apotential factor for pain (34). Elevated IL-2 andTNF-α mRNA levels have been found in patients with painful neuropathy (35), while mRNA levels for the anti-inflammatorycytokines IL-4 and IL-10 were reduced (36).Additionally, some cytokines released by monocytes and macrophages, such as IL-6, IL-1β,and TNF-α, activate hepatocytes and induce the hepatic acute phase response, leading tothe synthesis of various serum proteins, such as CRP (5,37).
In this study, the association of pain with CRP and TNF-α may have been attenuated bypharmacotherapeutic agents administered according to the World Health Organization'sanalgesic ladder for cancer pain management, as TNF-α levels can be influenced bycertain medications. However, none of our patients were treated with anticytokinetherapy, such as cytokine receptor antibodies, cytokine signaling inhibitors, TNF-αblockers, or thalidomide.
In conclusion, this is the first study to report a positive association betweenperceived pain and CRP in HNC patients at diagnosis, prior to any course of anticancertherapy. Further research is needed to confirm these findings and to determine thefunctional effects of elevated CRP and TNF-α on pain modulation. Overall, the currentfindings suggest an important association between pain and inflammatory processes in HNCpatients, with potential implications for future treatment strategies.
Footnotes
Acknowledgments
The authors would like to thank the medical team of Santa Rita de Cassia Hospital,AFECC, Vitória, ES, Brazil, for their assistance with data collection. Researchsupported by CAPES, INCT-IF/REBRAFVIME, FAPES, and CNPq.
References
- 1. Cancer-related inflammationNature2008454436444[PubMed][Google Scholar]
- 2. Role of cytokines in head and neck squamous cellcarcinomaExpert Rev Anticancer Ther2006611951203[PubMed][Google Scholar]
- 3. Cytokines and their relationship to the symptoms andoutcome of cancerNat Rev Cancer20088887899[PubMed][Google Scholar]
- 4. Mechanism of cancer painMol Interv201010164178[PubMed][Google Scholar]
- 5. Pain may predict poor prognosis in patients with oralsquamous cell carcinomaOral Surg Oral Med Oral Pathol Oral Radiol Endod2011111587592[PubMed][Google Scholar]
- 6. Quality of life and oral function following radiotherapyfor head and neck cancerHead Neck199921111[PubMed][Google Scholar]
- 7. Cytokines, inflammation, and painInt Anesthesiol Clin2007452737[PubMed][Google Scholar]
- 8. Beyond neurons: evidence that immune and glial cellscontribute to pathological pain statesPhysiol Rev2002829811011[PubMed][Google Scholar]
- 9. Cancer pain and its relationship to systemicinflammation: an exploratory studyPain2011152460463[PubMed][Google Scholar]
- 10. C-reactive protein levels: a prognostic marker forpatients with head and neck cancer?Head Neck Oncol2010221[PubMed][Google Scholar]
- 11. Serum levels of IL-1b, IL-6, TNF-α, sTNF-RI and CRP inpatients with oral cavity cancerPathol Oncol Res19973126129[PubMed][Google Scholar]
- 12. Aberrant cytokine expression in serum of patients withadenoid cystic carcinoma and squamous cell carcinoma of the head andneckHead Neck200729472478[PubMed][Google Scholar]
- 13. Prognostic impact of pretherapeutic laboratory values inhead and neck cancer patientsJ Cancer Res Clin Oncol2013139171178[PubMed][Google Scholar]
- 14. Pain assessment: global use of the Brief PainInventoryAnn Acad Med Singapore199423129138[PubMed][Google Scholar]
- 15. Orofacial pain in cancer: part I -mechanismsJ Dent Res200786491505[PubMed][Google Scholar]
- 16. Types and causes of pain in cancer of the head andneckCancer199270178184[PubMed][Google Scholar]
- 17. Cytokine genes and pain severity in lung cancer:exploring the influence of TNF-alpha-308 G/A IL6-174G/C andIL8-251T/ACancer Epidemiol Biomarkers Prev20071627452751[PubMed][Google Scholar]
- 18. Preoperative circulating C-reactive protein levelspredict pathological aggressiveness in oral squamous cell carcinoma: aretrospective clinical studyClin Otolaryngol201136147153[PubMed][Google Scholar]
- 19. C-reactive protein, inflammation, and innateimmunityImmunol Res200124163176[PubMed][Google Scholar]
- 20. Pain and inflammation in women with early-stage breastcancer prior to induction of chemotherapyBiol Res Nurs201315234241[PubMed][Google Scholar]
- 21. Tumor necrosis factor, cancer and anticancertherapyCytokine Growth Factor Rev2005163553[PubMed][Google Scholar]
- 22. Endogenous tumor necrosis factor alpha (TNFalpha)requires TNF receptor type 2 to generate heat hyperalgesia in a mouse cancermodelJ Neurosci20082850725081[PubMed][Google Scholar]
- 23. Investigation of interleukin 10, 12 and 18 levels inpatients with head and neck cancerJ Laryngol Otol2007121246252[PubMed][Google Scholar]
- 24. Determination of cytokine protein levels in oralsecretions in patients undergoing radiotherapy for head and neckmalignanciesRadiat Oncol2012764[PubMed][Google Scholar]
- 25. Salivary and serum interleukin 1 beta, interleukin 6 andtumor necrosis factor alpha in patients with leukoplakia and oralcancerMed Oral Patol Oral Cir Bucal201217
[PubMed][Google Scholar] - 26. Biochemical regulation of sleep and sleepbiomarkersJ Clin Sleep Med20117S38S42[PubMed][Google Scholar]
- 27. Cognitive impairment, fatigue, and cytokine levels inpatients with acute myelogenous leukemia or myelodysplasticsyndromeCancer2005104788793[PubMed][Google Scholar]
- 28. Induction of proinflammatory and chemokine genes bylipopolysaccharide and paclitaxel (Taxol) in murine and human breast cancer celllinesCytokine200115156165[PubMed][Google Scholar]
- 29. Induction of cytokines and killer cell activities bycisplatin and 5-fluorouracil in head and neck cancer patientsAnticancer Drugs200011165173[PubMed][Google Scholar]
- 30. Contribution of inflammatory cytokine release toactivation of resident peritoneal macrophages after in vivolow-dose gamma-irradiationJ Radiat Res199940253262[PubMed][Google Scholar]
- 31. Intraneural injection of interleukin-1beta and tumornecrosis factor-alpha into rat sciatic nerve at physiological doses induces signsof neuropathic painPain2005116257263[PubMed][Google Scholar]
- 32. Antinociceptive effects of interleukin-4, -10, and -13on the writhing response in mice and zymosan-induced knee joint incapacitation inratsJ Pharmacol Exp Ther2003304102108[PubMed][Google Scholar]
- 33. Etanercept reduces hyperalgesia in experimental painfulneuropathyJ Peripher Nerv Syst200166772[PubMed][Google Scholar]
- 34. Mode of action of cytokines on nociceptiveneuronsExp Brain Res20091966778[PubMed][Google Scholar]
- 35. Differential expression of cytokines in painful andpainless neuropathiesNeurology2007694249[PubMed][Google Scholar]
- 36. Reduced levels of antiinflammatory cytokines in patientswith chronic widespread painArthritis Rheum20065426562664[PubMed][Google Scholar]
- 37. Cytokine-induced sickness behaviour: mechanisms andimplicationsTrends Neurosci200225154159[PubMed][Google Scholar]