Deletion of serum lectin-reactive alpha-fetoprotein by acyclic retinoid: a potent biomarker in the chemoprevention of second primary hepatoma

H Moriwaki,I Yasuda,Y Shiratori,T Uematsu,M Okuno,Y Muto

Abstract:A goal of cancer chemoprevention is the deletion of latent premalignant or malignant clones before they expand to a clinically detectable tumor. However, such clonal deletion has not been demonstrated in clinical studies. We have evaluated serum levels of lectin-reactive alpha-fetoprotein (AFP-L3), which suggests the presence of latent hepatoma cells, in a randomized controlled trial that used acyclic retinoid to prevent second primary hepatomas in patients who had received treatments that cured initial hepatomas. The trial involved 21 patients in each acyclic retinoid (600 mg daily) and placebo group and consisted of a 12-month period of drug administration and a subsequent follow-up period. Serum AFP-L3 was determined at entry and at the end of the 12-month treatment period using lectin-affinity electrophoresis and antibody-affinity blotting. Although neither treatment affected serum levels of total AFP, acyclic retinoid significantly reduced AFP-L3 levels after a 12-month administration (P < 0.01). Acyclic retinoid not only deleted AFP-L3 from patients who had been positive for AFP-L3 at entry but also prevented the appearance of AFP-L3 in patients who had been negative at entry (P < 0.01). In contrast, placebo significantly raised the incidence of AFP-L3-positive patients after a 12-month administration from that at entry (P < 0.05). Patients positive for AFP-L3 after a 12-month treatment had a significantly higher risk of second primary hepatomas in the subsequent follow-up period (P = 0.03). Acyclic retinoid may have deleted a clone of latent hepatoma cells producing AFP-L3 and thereby inhibited second primary hepatomas. Serum AFP-L3 may be a useful intermediate biomarker in the chemoprevention of second primary hepatomas by acyclic retinoid.