Which Cases Are Found and Missed by Neuroblastoma Screening at 1 Year? Results From the 1992 to 1995 Study in Three Federal States of Germany

Abstract

PURPOSE: Neuroblastoma screening during the first half-year of life is associated with a two- to three-fold overdiagnosis. Because regression processes seem to be confined to infancy, we investigated whether screening at 1 year would be associated with fewer overdiagnoses, and we investigated the characteristics of thus-detected and not-detected patients. PATIENTS AND METHODS: Thin-layer chromatography was used for semiquantitative assessment of urine samples dried on filter paper and obtained when patients were 10 to 14 months old (sample 1) and 17 to 19 months old (sample 2). Abnormal results were reanalyzed quantitatively from the same specimen by high-performance liquid chromatography and/or gas chromatography–mass spectrometry. RESULTS: A total of 200,054 children of the German federal states Lower Saxony, Northern Rhine-Westphalia, and Bremen were screened from May 1992 to April 1995. Of 229,078 investigated samples (100%), 228,245 (99.6%) were first, 657 (0.3%) were second, and 176 (0.08%) were third urine specimens. The compliance rate was 27.8%, but it continued to increase throughout the study period and in the last year it was 43.3%. The second screening offered at 18 months was accepted by only 12.1% (24,259) of the children. Thirty children underwent clinical examination, and nine asymptomatic neuroblastoma cases were detected (stage 1, n = 4; stage 2, n = 2; stage 3, n = 2; stage 4, n = 1; detection rate, 1:22,228). The results of 21 tests were false-positive. Ten children with false-negative test results presented 8 to 35 months later with neuroblastoma (stage 1 tumor, n = 1; stage 2, n = 1; stage 3, n = 1; stage 4, n = 7; five of nine tumors were N-myc–amplified tumors). Three children were nonsecretors at the time of diagnosis. Fifty-two patients were “missed” (not screened), and 37 children developed neuroblastoma before the age of screening (early cases). During the same period, a total of 23.6 cases per million children within the screening area and 24.0 cases per million children outside the screening area were diagnosed as neuroblastoma cases (not significant [NS]). In prescreening times in the area of the later screening states, 20.7 cases per million children were found (NS). CONCLUSION: Screening at 1 year of age demonstrated a lower detection rate than earlier screening programs and did not produce a “halo effect.” The good prognostic features of early-detected cases and the poor characteristics of not-detected-but-late-presenting cases corresponded to those of the related age groups.