Table 2.

Multiple biopsy next-generation sequencing dissecting tumor heterogeneity.

Patients (Samples)	Technology	Findings and Potential Clinical Implications	Ref.
88 pts (46 matched PT and MTs and 42 non-metastatic PTs)	WES	Computationally calculated tumor heterogeneity was highly variable, with 70% sub-clone consistency between PT and LM, while high heterogeneity correlated to worse outcomes	[56]
69 pts (Matched PT and MT samples)	tNGS (WGS on 4)	KRAS, NRAS, and BRAF mutations were 100% consistent and recurrent alterations were highly similar, suggesting that NGS of either PT or MT could suffice	[57]
27 pts (97 samples from PT and MTs and 68 samples from a single PT)	tNGS (100 gene panel)	Inter- and intra-tumor variability was due to CNAs, which were highly discordant between PT and MT	[18]
23 (118 MR tissue samples from matched PT and MTs)	WES	Although extensive inter- and intratumor heterogeneity was identified, matched PT and MTs were highly concordant for driver mutations, suggesting the early acquisition of aggressive alterations responsible for metastasis, while the modeof tumor evolution and sub-clonality correlated with disease stage	[58]
18 (Matched PT and LM samples)	tNGS	79.3% of SNVs in the PT were detected in the LM, while 81.7% of LM mutations were found in the PT, suggesting linear progression	[59]
18 (Matched PT and MT samples)	tNGS	While concordance was 93.5%, most tumors showed at least one discordance due to co-evolution, suggesting that sampling over therapy could be useful	[60]
17 (213 matched PT, LN and MT)	Polyguanine-repeat analysis	In 65% and 35% of cases, LN and distant metastases originated from distinct and single PT subclones respectively	[61]
14 pts (70 MR samples from PT and matched liver and/or lung MTs)	tNGS	RAS status was preserved in MTs, while emerging mutations in other genes were also identified	[62]
12 (Matched PT and MT)	WGS	15% and 19% of mutations were PT- and MT-specific respectively, while late metastasis is supported Recurrent non-coding mutations: ncRNAs RP11-594N15.3, AC010091, SNHG14, 3’ UTRs of FOXP2, DACH2, TRPM3, XKR4, ANO5, CBL, CBLB MT-specific oncotargets: FAT1, FGF1, BRCA2, KDR, and AKT2-, AKT3-, and PDGFRA-3’ UTRs	[63]
10 early CRC (53 MR samples)	MR-WES	This study supports a shift from Darwinian to neutral evolution during CRC progression	[64]
9 (75 MR PT and 2 LM samples)	MR-WES	All cancers exhibited high ITH due to neutral evolution and drug resistance was attributed to pre-existing minor subclones	[65]
6 (3-5 biopsies per patient)	MR-WES, RNAseq	Although ITH was universal, transcriptomics-guided classification could be independent of ITH	[66]
5 pts (35 MR PT and LM samples)	MR-WES of the PT and MT	Branching evolution was identified, with prevalent CNA-based ITH as a putative source of metastasis	[67]
4 (23 MR PT and MT samples)	MR-WES	Significant inter- but limited intra-patient variability was identified MTs had lower ITH than PTs, while polyclonal seeding was detected	[68]
2 (36 spatiotemporal PT and MT samples)	WES	Different modes of evolution and metastatic progression were identified, depending on the immune microenvironment of the metastatic site Distinct MTs showed different clinical, genomic and immune features An immunoediting score was developed and correlated to immune response and prognosis	[69]

Abbreviations: colorectal cancer (CRC), copy-number alteration (CNA), intra-tumor heterogeneity (ITH), liver metastasis (LM), lymph node (LN), metastatic tumor (MT), multi-regional (MR), next-generation sequencing (NGS), non-coding RNA (ncRNA), patients (pts), primary tumor (PT), RNA sequencing (RNAseq), single-nucleotide variant (SNV), targeted NGS (tNGS), whole-exome sequencing (WES), whole-genome sequencing (WGS).