Lag times between lymphoproliferative disorder and clinical diagnosis of chronic lymphocytic leukemia: a prospective analysis using plasma soluble CD23.
Kaaks R., Sookthai D., Łuczyńska A., Oakes CC., Becker S., Johnson T., Johansson A., Melin B., Sjöberg K., Trichopoulos D., Trichopoulou A., Lagiou P., Mattiello A., Tumino R., Masala G., Agnoli C., Boeing H., Aleksandrova K., Brennan P., Franceschi S., Roulland S., Casabonne D., de Sanjose S., Sánchez MJ., Huerta JM., Ardanaz E., Sala N., Overvad K., Tjønneland A., Halkjær J., Weiderpass E., Bueno-de-Mesquita HB., Vermeulen R., Peeters PH., Vineis P., Kelly RS., Khaw KT., Travis RC., Key TJ., Riboli E., Nieters A.
BACKGROUND: Chronic lymphocytic leukemia (CLL) is a chronic disease that often progresses slowly from a precursor stage, monoclonal B-cell lymphocytosis (MBL), and that can remain undiagnosed for a long time. METHODS: Within the European Prospective Investigation into Cancer cohort, we measured prediagnostic plasma sCD23 for 179 individuals who eventually were diagnosed with CLL and an equal number of matched control subjects who remained free of cancer. RESULTS: In a very large proportion of CLL patients' plasma sCD23 was clearly elevated 7 or more years before diagnosis. Considering sCD23 as a disease predictor, the area under the ROC curve (AUROC) was 0.95 [95% confidence interval (CI), 0.90-1.00] for CLL diagnosed within 0.1 to 2.7 years after blood measurement, 0.90 (95% CI, 0.86-0.95) for diagnosis within 2.8 to 7.3 years, and 0.76 (95% CI, 0.65-0.86) for CLL diagnosed between 7.4 and 12.5 years. Even at a 7.4-year and longer time interval, elevated plasma sCD23 could predict a later clinical diagnosis of CLL with 100% specificity at >45% sensitivity. CONCLUSIONS: Our findings provide unique documentation for the very long latency times during which measurable B-cell lymphoproliferative disorder exists before the clinical manifestation of CLL. IMPACT: Our findings have relevance for the interpretation of prospective epidemiologic studies on the causes of CLL in terms of reverse causation bias. The lag times indicate a time frame within which an early detection of CLL would be theoretically possible. Cancer Epidemiol Biomarkers Prev; 24(3); 538-45. ©2014 AACR.