Table 4.
Selective recent studies investigating the predictive value of CTCs in NSCLC patients treated with TKI. Abbreviations: NSCLC, non-small cell lung cancer; CTC, circulating cancer cell; TKI, tyrosine kinase inhibitor; ISET, isolation by size of tumour cells; CK, cytokeratin; EMT, epithelial–mesenchymal transition; VIM, vimentin; PD-L1, programmed cell death ligand 1; PFS, progression-free survival; PFS 12 m: one-year progression-free survival; RR, response rate; EGFR, epidermal growth factor receptor; ALK, anaplastic lymphoma kinase; OS, overall survival; CNG, copy number gain; pts, patients.
| Study | Analysis Time-Point | Patients (n) |
Type of Targeted Treatment | Detection Method | Predictive Relevance |
|---|---|---|---|---|---|
| Ntzifa et al. [104] | At baseline, after 1 month, and at PD | 30 | Osimentinib | Parsortix ISET |
There was a strong positive correlation of VIM expression with PIM-1 expression at baseline and increased PD-L1 expression levels at PD. The high prevalence of VIM-positive CTCs suggests a dynamic role of EMT during osimertinib treatment |
| Kallergi et al. [105] | At baseline, post-cycle 1, and at the end of treatment | 47 | Osimertinib | ISET | The decrease in CTCs occurring early during osimertinib treatment is predictive of better outcome |
| Pantazaka et al. [106] | At baseline, post-cycle 1, and at the end of treatment | 42 | Osimertinib | ISET | Significant correlation between PD-L1 and pS6 phenotypes at all time points was found. Survival analysis revealed that CK + pS6+ and CKlowpS6+ phenotypes after 1st cycle were related to significantly decreased PFS12m (p = 0.003) and PFS (p = 0.021), respectively |
| Isobe et al. [107] | At baseline | 30 | Osimertinib | ClearCell FX | RR to osimertinib was worse in pts with high than in those with low BIM-γ mRNA expression (26.6% vs. 73.3%, respectively; p = 0.011). PFS did not significantly differ between groups (p = 0.13) |
| Tamminga et al. [92] | At baseline | 86 | TKI therapy (n = 34) | CellSearch | Pts with CTCs vs. pts without CTCs: RR: TKI: 25% vs. 73% (p = 0.02) PFS: 3.3 mo. vs. 8 mo., p = 0.01 OS: 5.2 mo. vs. 12.1 mo., p = 0.03 |
| Jiang et al. [108] | At baseline, 1 month after treatment, and every 2 months | 232 | EGFR-TKIs | CytoploRare | Pts with baseline low CTCs vs. high CTCs: PFS: 412 vs. 267 days; HR = 0.48; p < 0.001 OS: 836 vs. 583 days; HR = 0.52; p = 0.002 Pts with EGFR19delmut and low CTCs in comparison with high CTCs have prolonged PFS (HR = 0.51, p = 0.014) and OS (HR = 0.52, p = 0.036) Pts with EGFR L858R) and low CTCs in comparison with high CTCs have prolonged PFS (HR = 0.5, p = 0.023) and OS (HR = 0.43, p = 0.0007) |
| Pailler et al. [103] | At baseline and at an early time-point (2 months) | 39 | Crizotinib | ISET | The dynamic change of CTC with ALK-CNG was the strongest factor associated with PFS (14 mo. for pts with decreased ALK-CNG vs. 6.1 mo. for pts with stable or increased ALK-CNG, p = 0.025). No correlation with OS |
| Pailler et al. [61] | At PD | 17 | Crizotinib Lorlatinib |
ISET CellSearch RosetteSep |
Multiple mutations in various genes (EGFR, KRAS, BRAF genes and TP53 pathways) in ALK-independent pathways were predominantly identified in CTCs of crizotinib-resistant patients. In one lorlatinib-resistant patient, 2 single CTCs out of 12 harboured ALK compound mutations. CTC-1 harboured the ALKG1202R/F1174C compound mutation, virtually similar to ALKG1202R/F1174L present in the corresponding tumour biopsy. CTC-10 harboured a second ALKG1202R/T1151M compound mutation not detected in the tumour biopsy |