Breast cancer (BC) is the most common female cancer and CTC is a predictive marker of poor survival and metastatic relapse ^[1][58]. The detection rate of CTCs correlates with the number of metastatic sites, and BC patients with brain metastasis may have the highest CTC counts ^[2][59].

The hormone status of BC, such as expression of the estrogen receptor (ER) or progesterone receptor (PR), indicates the feasibility of ER-targeted endocrine therapy ^[3][60]. However, no correlation was found between total CTC number and/or ER expression status as determined by immunocytostaining and the intensity of ER staining in primary tumors ^[4][20]. Only 81.3% of patients were positive for ER expression in CTCs, while ER-negative CTCs were also found in ER-positive patients, delineating the genetic inconsistencies between CTC counts. ER status in CTCs might have predictive power with regard to response and resistance to endocrine therapy and may thus help in the choice of better treatment options ^[4][20]. One study performed Sanger sequencing on CTC WGAs (MALBAC), which resulted in the identification of the ESR1-Y537S variant known to produce a constitutively active receptor and ESR1-T570I (a novel mutation) in exon 8 ^[5][25]. This study found ESR1-Y537S heterozygously and homozygously in single CTCs and confirmed mutations in matched cell-free DNA (cfDNA) in one patient. Interestingly, in another patient, heterozygote ESR1-T570I and homozygote ESR1-Y537S were found in a single CTC, but ESR1-T570I could not be detected in matched cfDNA ^[5][25]. Thus, using two entities extractable from a blood biopsy, CTCs and cfDNA biomarkers may complement each other and enhance the chance of finding disease-related variants. However, in another study that screened for exon 4, 6 and 8 ESR1 mutations after WGA (Picoplex, MALBAC), none was found in individual CTCs ^[4][20].

The PI3K/AKT/mTOR pathway (Phosphoinositide 3-kinase/protein kinase B/mammalian target of rapamycin) regulates cell growth, survival, and angiogenesis. Upregulated activity has been linked to oncogenesis and is a major therapeutic target ^[6][61]. In BC, mutations in PIK3CA are found in about 40% of ER-positive cancers and have been implicated in resistance to HER2-based therapies ^[7][62]. Pharmacologic targeting of PIK3Ca in HR (hormone receptor) +/HER2-metastatic BC offers significant benefits to patients with endocrine therapy resistance ^[8][44]. Several single CTC-based studies ^{[8][9][10][11][12]}[19,26,27,44,63] were conducted to study mutations in the PIK3CA gene. Heterogenous expression of PIK3CA mutations among CTCs and matched primary tumors, and even among CTCs from the same patient, was observed. Individual PIK3CA mutations found in Ampli1-amplified CTCs included E542K and H1047R ^[8][44], as well as E542K, E545K and H1047R, as was determined in a second study ^[10][26]. Another study found PIK3CA mutations (E542K, E545K, H1047R, H1047L and M1043V) in exon 9 and 20 in at least one CTC in 36.4% of the patients tested ^[13][64]; similar data were reported in other studies ^[11][12][27,63] (Table 12). 

Prostate cancer (PC) is the most common cancer type diagnosed in men; eventually, it develops into castrate-resistant prostate cancer (CRPC) following standard of care androgen deprivation therapy (ADT). Commonly altered genes during CRPC progression include AR (androgen receptor), ERG (ETS-related gene), c-MET (tyrosine-protein kinase MET), PTEN (phosphatase and tension homology deleted on chromosome 10) and PI3K/AKT signaling pathway genes. AR alterations in CTCs, especially AR amplification and expression of splice variant AR-V7, predict poor treatment outcomes for ADT ^{[20][21][38][39]}[71,72,82,83]. ERG amplification of CTCs is also informative for treatment selection and might contribute to resistance to taxane therapy ^[21][72].

WGA-based single-CTC analysis found significant numbers of shared mutations in PTEN, GRM8 and TP53 among PC CTCs, particularly if they were of epithelial phenotype. Some recurrent mutations found in CTCs correlated with matched metastatic tissue. Interestingly, sequencing multiple CTCs did not significantly change the number of mutations found ^[19][6]. This may indicate that heterogeneity is less of an issue, as these mutations may be shared by most CTCs and are likely early events in cancer formation. Both epithelial and non-epithelial CTCs showed CTC-exclusive alterations affecting invasion, DNA repair mechanism, cancer-driver, and cytoskeleton genes ^[19][6]. The shared mutations between matched tissue and CTCs might provide insights into the metastatic spread of cancer and the origins of CTCs, as it is assumed that more mutations are acquired during cancer progression and spread.

aCGH analysis of CTC WGA products from CRPC patients demonstrated genomic gains in >25% of CTCs. Such genomic gains were observed in AR, FOXA1, ABL1, MET, ERG, CDK12, BRD4 and ZFHX3, while common genomic losses involved PTEN, ZFHX3, PDE4DIP, RAF1 and GATA2. AR and NCOA2 amplification were found in 50% and 43.75% of CTC WGAs, respectively, while ERG amplification was found in 40% of patient CTCs. Loss of KDM6A was found in 6.25%, while KDM6A gain was found in 50% of mCRPC CTC samples. MYCN gene amplification was observed after the development of enzalutamide resistance. Similarly, PTEN gain was observed before starting enzalutamide, and PTEN loss appeared after enzalutamide treatment ^[21][72]. Another aCGH analysis of WGA CTCs found AR gain in 78% of nine patient bulk CTC samples (that is, samples combining more than a single CTC). However, AR gain in CTC WGA samples is not always found in matched tissues and may be due to previous archival tissues failing to represent tumor evolution; nevertheless, some copy number alterations, including gains and losses of chromosome 8p and 8q, are concordant between CTCs and primary tumors ^[22][73].

The detection of certain driver mutations, such as in EGFR and ALK fusion, is associated with the early stages of lung cancer, its development and drug resistance ^[25][74]. Genetic analysis of CTCs from the same patient can give overall information about deletions, fusions, insertions and SNVs in the metastatic tumor and such changes can be monitored during treatment, even in the presence of cell-to-cell heterogeneity; however, a large number of CTCs needs to be sequenced ^[29][5].

Ni. et al. observed number of mutations in different genes, such as EGFR, PIK3CA, RB1 and TP53, after exome sequencing of single-CTC WGA products. Amongst these alterations, one INDEL in the EGFR gene (K746_A750del), which is a target for tyrosine kinase inhibitors (TKIs), was found in CTCs as well as in the primary and metastatic tumors of the patients, while other mutations in PIK3CA (E545K), TP53 (T155I) and RB1 (R320*) genes were only observed in CTCs and metastatic tumors in the liver. This study also found some common CNV regions that have important roles in cancer development, such as cell proliferation, differentiation and protecting chromosomal ends from degradation. These regions include regions of gain in chromosome 8q, the c-Myc gene loci, and in chromosome 5p, the TERT gene (telomerase reverse transcriptase) loci, 17q22, 17q25.3 and 20p13. The CNV patterns of individual CTCs from the same patient were reproducible. It was also found that CNV patterns were not changed upon different drug treatments ^[29][5].

Floating tumor cells (FTCs) from the pleural fluid of lung adenocarcinoma patients were enriched and amplified. EGFR exon 19 deletion (del L747_A750), an EGFR activating mutation that makes patients eligible for EGFR inhibitor therapy, was detected in 63.2% of FTCs in one patient. In a second patient, the EML4-ALK (echinoderm microtubule associated protein-like 4–anaplastic lymphoma kinase) fusion variant, which is a novel target in a subset of non-small cell lung cancer cases, was detected in 85% of isolated FTCs. The ALK G1202R mutation, a known Alectinib-resistance mutation, was the only mutation identified throughout multiple FTC samples from another patient ^[28][13].

Colorectal cancer (CRC) is the third most commonly diagnosed cancer and second most common death-causing cancer in Australia. It is a lethal cancer with a high mortality rate due to distant metastasis. A number of driver genes are commonly identified in CRC, including mutated BRAF, KRAS, EGFR and PIK3Ca ^[9][30][40][19,76,84]. EGFR is the main therapeutic target; however, responses to EGFR inhibition are variable ^[9][19]. The key mutations found in single-cell analysis of CRC CTCs so far are KRAS, PIK3CA and EGFR mutations. Significant heterogeneous expression of KRAS, PIK3CA and EGFR was found among CTCs within the same patient and between different individuals ^[9][30][19,76]. A mutational discordance between primary tumor tissue and CTC WGAs was observed for KRAS, and remarkably different KRAS mutations in different single-CTC WGAs from the same individual patients have been observed ^[9][30][19,76]. CTCs were observed with increased EGFR expression in some patients, and EGFR gene amplification was identified in 7 out of 26 CTC WGAs for three patients ^[9][19].

Pancreatic cancer is a lethal cancer with a less than 10% 5-year survival rate. KRAS is the predominant mutated gene in pancreatic cancer, and targeting KRAS may be an attractive therapy, despite many trial failures for anti-KRAS therapies ^[41][85]. KRAS mutations have been detected in 92% of patients, with a detection rate of 27.7% in total single-CTC WGAs (REPLI-g, MDA), but not in any WGAs of control WBCs. Interestingly, at least 10 single CTCs are required to reliably detect the KRAS heterozygous allele ^[32][77], which indicates that single-cell amplification bias responsible for ADO can be reduced by sequencing at least 10 cells together. In a study on single-CTC analysis of melanoma ^[34][79], CDKN2A and PTEN deletions and amplifications of TERT, BRAF, KRAS and MDM2 were found. Moreover, new chromosomal amplifications of chromosomes 12, 17 and 19 were detected ^[34][79].