Prophylactic Cranial Irradiation for Patients With Locally Advanced Non–Small-Cell Lung Cancer

This review by Dr. Gore emphasizesthe significance of theproblem of brain metastases inpatients with locally advanced non-small-cell lung cancer (NSCLC). Thearticle should prompt medical and radiationoncologists to consider enrollingpatients in the ambitious study ofprophylactic cranial irradiation (PCI)led by the Radiation Therapy OncologyGroup (RTOG L-0214). Statisticsfrom the ongoing RTOG study arecomplicated, but essentially, the researchersare looking for a 20% increasein median survival for patientsreceiving PCI. This would make theimpact of PCI in NSCLC comparableto that observed in limited small-celllung cancer (SCLC).The meta-analysis of PCI in limitedSCLC found a 5% absolute increasein 3-year survival for patientsreceiving PCI-ie, a 3-year survivalof 15.3% for the control arm comparedto 20.7% for the PCI arm.[1]RTOG L-0214 requires over 1,000patients be accrued in order to detectthe 20% increase in median survival.Although other centers around theworld are interested in participatingin the study, the lack of validation ofthe neurocognitive function tests innon-English-speaking patients restrictsthe study to North America.Dr. Gore, a principal investigatorfor this study, fought long and hard toopen RTOG L-0214. It has been endorsedby all of the major cooperativecancer groups, and is open throughthe Clinical Trials Support Unit. Thestudy was activated in September2002. It is possible that centers arestill putting the protocol through theirreview boards, but accrual to this studyhas not yet met expectations. Possiblereasons for the low accrual could includethe following explanations.Physicians may think the questionof PCI in NSCLC has alreadybeen answered.
Phase III evidence will be criticalin determining the value of PCI inNSCLC. Dr. Gore goes through thethree previously published random-

ized studies in some detail.[2-4] Theserandomized studies are also summarizedin a table in Dr. Gore's article,but their message loses some impactas the table also includes phase II studies. Table 1 above is another attemptto show differences betweenthe ongoing RTOG study and theseolder phase III studies.First, it is clear that the earlier studiescontained a relatively small numberof patients. The statistical backgroundin the manuscripts does notindicate how the numbers of patientsto be accrued was determined. Butclearly the studies were not poweredto detect an achievable survival difference.Another important differenceis that, in the early series, patientswere randomized much earlier in thecourse of their disease (usually afterdiagnosis, prior to any treatment).It is quite probable that RTOGL-0214 will accrue a much differentpopulation of patients than the previousrandomized studies.[2-4] TheRTOG study mandates a normal brainmagnetic resonance imaging scan priorto study entry, whereas the earlierstudies had only computed tomographyscans or nuclear brain scans. Thecurrent study accrues patients onlyafter definitive treatment, randomizingthem to PCI or observation within16 weeks of completing all therapy.Therefore, there is a significant selectionfactor favoring good performancepatients with a response todefinitive treatment and no evidenceof distant metastases documented followingdefinitive treatment. Thedrawback of the late accrual and PCIdelivery is that many patients willalready have relapsed in the brain. Arecent retrospective review of patientswith stage III NSCLC determined that46.5% of brain metastases presentedwithin 16 weeks of completion oftherapy.[5]Physicians may not be convincedthat PCI is safe.
Dr. Gore states that it "took severaldecades for PCI to be accepted as asafe and effective method of managingcentral nervous system (CNS) micrometastasesin patients withsmall-cell lung cancer." But PCI isnot given to all limited SCLC patientsfollowing a complete response to chemotherapy.[6] A national survey ofrandomly selected institutions in theUnited States found that during theyears 1998-1999, PCI was given toonly 23% of patients treated for limited-stage SCLC. What cannot be determinedfrom such a survey is whether PCI was recommended.Cmelak et al surveyed 9,176 oncologistsin 1997 to determine if theyrecommended PCI to their patientswith SCLC.[7] The survey also askedthe oncologists if they believed thatPCI improved survival or quality oflife. Only 13% of oncologists returnedthe survey, but the results (summarizedin Table 2) are nonetheless provocative.Cmelak et al concluded thatmost oncologists recommend PCI inlimited-stage SCLC despite the factthat many do not believe that it leadsto an increase in survival or quality oflife. Perhaps patients pick up on thisambivalence.Are physicians concerned aboutneurotoxicity following PCI? Probably,but they might also be minimizingthe impact of the subsequentdevelopment of brain metastases onneurocognitive function. Our abilityto control brain metastases once theyare diagnosed is not impressive. TheRTOG reported the results of the Mini-Mental Status Exam (MMSE) beforeand after two different regimens ofexternal-beam irradiation for patientswith brain metastases.[8] Prior to anytherapy, it was noted that fewer than20% of patients had no neurologicsymptoms and were fully active athome or work without assistance. Despitetreatment, only 35% to 45% ofpatients had documented improvement2 to 3 months later, with a similarpercentage of patients showing adecline in the MMSE. Needless tosay, progression of brain metastaseswas associated with a subsequent declinein MMSE scores.Neurocognitive function was alsocarefully evaluated in a prospectiverandomized study testing the benefitof motexafin gadolinium in additionto standard cranial irradiation in patientswith brain metastases.[9] Approximatelyhalf of the patients enrolledin this study had NSCLC.Neurocognitive function was foundto be impaired to some extent prior toany therapy in over 90% of patients.Although neurocognitive functionand CNS recurrence rates are importantclinical objectives, the NationalCancer Institute mandated that survivalbe the primary end point for RTOGL-0214. However, the investigators aremaking a concerted effort to assess theneuropsychological impact of brainmetastases and the use of PCI. MMSE,Hopkins Verbal Learning Test, andActivities of Daily Living Scale datawill be collected at study entry and atspecified follow-up intervals.Patients may be refusingto participate.
Of the possible reasons for lowaccrual to RTOG L-0214, patient refusalto participate will be the mostdifficult to determine. Patients maybe concerned about the impact of PCIon their subsequent neurocognitivefunction and quality of life. In mypractice, an increasing number of patientswith diagnosed brain metastasesare refusing whole-brain irradiation.Many patients specifically request stereotacticradiosurgery only. Thismight reflect their own research intothe efficacy and toxicity of wholebrainirradiation, or it might reflectpriming by referring physicians. If wecan't convince patients to have wholebrainirradiation in the setting of establishedbrain metastases, can weconvince them to have it in the preventivesetting?In summary, the concept of PCIfor NSCLC is not new. Older, outdatedstudies did not demonstrate anysurvival benefit. But RTOG L-0214,the ongoing phase III study of PCI, has a solid rationale and is sufficientlydifferent from the older studies towarrant strong support. Whether a sufficientnumber of patients can be accruedremains to be seen.

Disclosures:

The author(s) have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.

References:

1.

Auperin A, Arriagada R, Pignon J-P, et al:Prophylactic cranial irradiation for patients withsmall-cell lung cancer in complete remission.N Engl J Med 341:476-484, 1999.

2.

Cox JD, Stanley K, Petrovich Z, et al:Cranial irradiation in cancer of the lung of allcell types. JAMA 245:460-472, 1981.

3.

Russell AH, Pajak TE, Selim HM, et al:Prophylactic cranial irradiation for lung cancerpatients at high risk for development of cerebralmetastasis: Results of a prospective randomizedtrial conducted by the RadiationTherapy Oncology Group. Int J Radiat OncolBiol Phys 21:637-643, 1991.

4.

Umsawasdi T, Valdivieso M, Chen TT, etal: Role of elective brain irradiation during combinedchemoradiotherapy for limited diseasenon-small cell lung cancer. J Neuro-Oncology2:253-259, 1984.

5.

Gaspar LE, Chansky K, Albain K, et al:Time from treatment to subsequent diagnosisof brain metastases in stage III non-small celllung cancer (NSCLC): A retrospective reviewby the Southwest Oncology Group (SWOG).Proc Am Soc Clin Oncol. In press.

6.

Movsas B, Moughan J, Komaki R, et al:Radiotherapy Patterns of Care Study in LungCarcinoma. Int J Radiat Oncol Biol Phys 54(2suppl):101, 2002.

7.

Cmelak AJ, Choy H, Shyr Y, et al: Nationalsurvey on prophylactic cranial irradiation:differences in practice patterns betweenmedical and radiation oncologists. Int J RadiatOncol Biol Phys 44:157-162, 1999.

8.

Regine WF, Scott C, Murray K, et al:Neurocognitive outcome in brain metastasespatients treated with accelerated-fractionationvs. accelerated-hyperfractionated radiotherapy:an analysis from Radiation Therapy OncologyGroup Study 91-04. Int J Radiat Oncol BiolPhys 51:711-717, 2001.

9.

Mehta PM, Rodrigus P, Terhaard CHJ, etal. Survival and neurologic outcomes in a randomizedtrial of motexafin gadolinium andwhole brain radiation therapy in brain metastases.J Clin Oncol. In press.