To the Editor: I would like to clarify the selection processleading to the inclusion of the patients in the previous reporton survival in the trial of computed tomographic (CT) screeningfor lung cancer (Oct. 26, 2006, issue).1 The process at 37 ofthe 38 participating sites was as follows. People interestedin participating in the trial were first interviewed and wereadministered a precoded questionnaire by the study staff. Thedata from the questionnaires were entered, and data entry waschecked, according to the standard protocol used at each center.The data were then transmitted electronically to the coordinatingcenter. There, eligibility was assessed by means of a computeralgorithm, and those who did not meet the eligibility criteriawere excluded (Figure 1). At the 38th site, the questionnairewas not administered, and therefore all the needed data werenot recorded before enrollment.
Figure 1. Supplement to the Previously Published Figure 1, Showing the Numbers of Patients Enrolled and Evaluated at the 38th Site as Compared with the Numbers at the Other 37 Sites.
The "31,567 Asymptomatic participants [who] underwent baseline screening" in the previously published Figure 1 comprises the "30,449 [participants who] were enrolled and underwent baseline CT" at 37 sites plus the "1130 [participants who] were enrolled at the 38th site," minus the "12 [participants who] were excluded" (30,449+1130–12=31,567). CT denotes computed tomography.
For participants who received a diagnosis of lung cancer afterbaseline screening, study records were reviewed by the steeringcommittee at one of its twice-yearly meetings. This review includedconfirmation of eligibility for the study — that is, asymptomaticstatus at the time of enrollment. If the symptoms resultingin exclusion had been present at enrollment but had not beenrecorded, the participant was excluded post hoc. This was thecase for three patients among the 37 sites. Other features relevantto overall survival or survival itself were not used as a basisfor exclusion. At the 38th site, at which symptoms at enrollmenthad not been documented, eight participants were excluded onthe basis of preexisting disqualifying symptoms, and one wasexcluded because pathological confirmation of lung cancer wasnot received by the coordinating center (Figure 1). The symptomaticstatus of participants who did not receive a diagnosis of lungcancer was not reviewed, since this information had no bearingon the research question addressed by the study.
Except for the 12 patients excluded after enrollment (the 3from the 37 sites at which the screening questionnaire had beenadministered and the 9 from the 38th site, at which the questionnairewas not administered), no patients were excluded from the studyafter they had been enrolled on the basis of the computer algorithm.
Inclusion of these 12 patients changes the 10-year survivalrate for patients with lung cancer from the 80% (95% confidenceinterval [CI], 74 to 85) reported previously for 484 patientsto 81% (95% CI, 75 to 86) for 496 patients. The other reportedfindings are not changed.
The article also reported that eight patients with clinicalstage I lung cancer remained untreated and died within 5 yearsafter diagnosis. However, only three had a pathological diagnosisof stage I lung cancer. Another four had stage I disease confirmedon CT, but further workup was delayed despite repeated promptings,and pathological diagnosis was made only after the cancer hadprogressed to stage IV. The remaining patient had a solitarynodule on baseline CT that grew at a rate consistent with primarylung cancer, refused biopsy and treatment, and died of lungcancer 6 months after the last CT showing lung cancer. Thus,all eight patients died from lung cancer within 5 years aftertheir actual or potential diagnosis during stage I.
Since, however, pathological diagnosis of lung cancer was requiredby the International Early Lung Cancer Action Project (I-ELCAP)investigators, I should have classified four of the eight patientsas having stage IV lung cancer and the remaining patient whohad not received a pathological diagnosis during stage I ashaving an interim diagnosis. The remaining 483 patients receivedan antemortem pathological diagnosis of their lung cancer. Thus,the correct number of patients who were untreated and had adiagnosis of stage I lung cancer is 3, not 8, and the totalnumber of patients who had clinical stage I lung cancer is 407,not 412 (Figure 1).
These corrections increased the 10-year Kaplan–Meier survivalrate for clinical stage I lung cancer from 88% to 90%. The overallKaplan–Meier survival rate remained the same, since allpatients with any stage of lung cancer were included in thatanalysis.
Claudia I. Henschke, M.D., Ph.D. Weill Medical College of Cornell University New York, NY 10021 chensch{at}med.cornell.edu
This letter (10.1056/NEJMc086327) was published at www.nejm.orgon July 30, 2008.
References
The International Early Lung Cancer Action Program Investigators. Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med 2006;355:1763-1771. [Erratum, N Engl J Med 2008;358:1862, 1875.] [Free Full Text]
Previous
