Classification of Cancer Pain Syndromes

Chronic pain occurs in about one-third of all cancerpatients and in about three-quarters of those with advanced disease.[1] A majorfactor in the undertreatment of cancer pain is inadequate pain assessment.[2]Pain assessment provides the basis for inferred pathophysiology that directsdiagnostic evaluation and treatment decisions. Pain syndrome identificationplays an important role in this process—much of clinical medicine is based onpattern recognition of symptoms and signs, leading to a specific diagnosis andtherapeutic strategy.

Provocative maneuvers that reproduce pain are especially valuable. Forexample, a positive femoral stretch test or reverse straight-leg-raisingmaneuver can reproduce back pain and lead the clinician to investigate furtherfor a retroperitoneal mass.

Classification of Pain Syndromes

In this issue of ONCOLOGY, Drs. Caraceni and Weinstein provide acomprehensive review of the classification of cancer pain syndromes. Althoughthey describe several classification schemes, syndromic classification includingunderlying pathophysiology probably provides the most useful clinicalinformation.

Foley’s seminal work in cancer pain syndrome identification is still usefultoday.[3] For instance, impending epidural spinal cord compression can berecognized by localized spinal pain followed by radicular pain. The inferredpathophysiology is somatic pain due to vertebral body metastasis, followed byneuropathic pain due to infiltration of the adjacent nerve roots. Recognitionand appropriate treatment of this syndrome can prevent a neurologic catastrophe.

Similarly, postmastectomy pain usually presents in a highly stereotypicalmanner. The pain is typical of neuropathic pain with a burning, stabbingcomponent; it is localized to the axilla, upper inner arm, and chest wall, andusually persists beyond the normal healing time of 3 months. Recognition of thispain syndrome as treatment- related pain rather than due to cancer progressionprovides valuable prognostic information.

A recent retrospective cohort study of postmastectomy pain syndromechallenged the traditional view that this is an uncommon condition, finding aprevalence rate of 29%.[4] Crystallization of pain syndromes as distinctclinical entities provides a basis for large-scale epidemiologic studies tofurther define the prevalence of disease- and treatment-related conditions.However, this is complicated by the observation that patients often present withmultiple pain syndromes that are due to both disease progression and treatment.

Validity and Reliability of Classification Systems

Although several classification systems for cancer pain syndromes have beenproposed, few have been tested for validity and reliability. Validity andreliability testing are essential features of any classification system. Theyallow reproducible predictable diagnoses so that different centers know thatthey are recognizing the same entities. Beyond the clinical application of asyndromic classification, such a system could be a valuable research tool forassessing prognosis and treatment outcome.

The Edmonton Staging System for cancer pain tested some components ofvalidity in a prospective multicenter trial.[5] The study showed that incidentpain and neuropathic pain are important predictors of intense discomfort that isless amenable to pharmacologic intervention. As Drs. Caraceni and Weinsteinpoint out, further validation of cancer staging is necessary to correlate painintensity to objective measures of tumor activity.

Perhaps the most useful application of a validated classification of cancerpain syndromes is to perform large-scale observational studies of treatmentoutcome to compensate for the dearth of high-quality randomized controlledtrials in cancer pain research. Most patients with cancer pain and progressivedisease are poor candidates for controlled clinical trials, especially trialsthat require long-term monitoring of repetitive dosing. For example, thetreatment strategies for neuropathic cancer pain are extrapolated almostentirely from nonmalignant neuropathic pain models.[6] Randomized controlledtrials are still limited in the information they provide because they reflectefficacy—ie, outcome of a defined population in a randomized controlled trialof appropriate size—rather than effectiveness, or real world outcome inordinary or average settings.[7]

Conclusions

Multicenter observational studies of treatment outcome using a patientregistry and a national or international database modeled on valid, reliablecancer pain syndromes would significantly enhance cancer pain research andtreatment. This could provide definitive outcome data in a wide variety ofcancer pain syndromes—data that are not presently available.

References:

1. Bonica JJ: Treatment of cancer pain: Current status and future needs, inFields HL, Dubner R, Cervero F (eds): Advances in Pain Research and Therapy, vol9, pp 589-616. New York, Raven Press, 1985.

2. Portenoy RK: Cancer pain: Pathophysiology and syndromes. Lancet339:1026-1031, 1992.

3. Foley KM: Pain syndromes in patients with cancer, in Bonica JJ,Ventafridda V (eds): Advances in Pain Research and Therapy, vol 2,pp 59-75. New York, Raven Press, 1979.

4. Smith WCS, Bourne D, Squair J, et al: A retrospective cohort study ofpost-mastectomy pain syndrome. Pain 83:91-95, 1999.

5. Bruera E, Schoeller T, Wenk R, et al: A prospective multicentre assessmentof the Edmonton staging system for cancer pain. J Pain Symptom Manage10:348-355, 1995.

6. Moulin DE: Neuropathic cancer pain: Syndromes and clinical controversies,in Portenoy RK, Bruera E (eds): Supportive Care Medicine, pp 7-29. New York,Oxford, 1998.

7. Lohr KN, Eleazer K, Mauskopf J: Review: Health policy issues andapplications for evidence-based medicine and clinical practice guidelines.Health Policy 46:1-19, 1998.