AbstractThe aim of the study was to investigate the effect of 2^nd-line pazopanib on the different CTCs subpopulations in SCLC patients and evaluate the clinical relevance of their changes. Different CTCs subpopulations were evaluated before pazopanib initiation (n = 56 patients), after one-cycle (n = 35) and on disease progression (n = 45) by CellSearch and double immunofluorescence using anti-CKs and anti-Ki67, anti-M30 or anti-Vimentin antibodies. Before treatment, CTCs were detected in 50% of patients by CellSearch whereas 53.4%, 15.5% and 74.1% patients had CK⁺/Ki67⁺, CK⁺/M30⁺ and CK⁺/Vim⁺ CTCs, respectively. One pazopanib cycle significantly decreased the number of CTCs as detected by CellSearch (p = 0.043) as well as the number of CK⁺/Ki67⁺ (p < 0.001), CK⁺/M30⁺ (p = 0^.015) and CK⁺/Vim⁺ (p < 0^.001) cells. On disease progression, both the incidence and CTC numbers were significantly increased (CellSearch, p = 0.027; CK⁺/Ki67⁺, p < 0.001; CK^+/M30⁺, p = 0.001 and CK⁺/Vim⁺, p < 0^.001)^. In multivariate analysis, the detection of CK⁺/Vim⁺ CTCs after one treatment cycle (HR: 7.9, 95% CI: 2.9–21.8; p < 0.001) and CTCs number on disease progression, as assessed by CellSearch, (HR: 2.0, 95% CI: 1.0–6.0; p = 0.005) were emerged as independent factors associated with decreased OS. In conclusion, pazopanib can eliminate different CTC subpopulations in patients with relapsed SCLC. The analysis of CTCs could be used as a dynamic biomarker of treatment efficacy.