Page 78 - Read Online

Shi et al. J Cancer Metastasis Treat 2018;4:47 I http://dx.doi.org/10.20517/2394-4722.2018.32 Page 11 of 19
[55]
are also present in the primary tumor and may be found in the metastatic lesions , suggesting that single-
cell analysis on CTCs is an effective option to non-invasively monitor cancer progression and predict meta-
static risk. Last but not the least, single-cell analysis facilitates researchers to dissect tumor heterogeneity at a
much higher resolution than before. For example, the degree of karyotypic anomalies in human cancer is as-
sociated with tumor progression and therapeutic response to cancer treatment [102] . However, current karyo-
typic analysis methods rely on a small fraction of dividing mitotic subpopulations in the sample and do not
provide in-depth information on copy number variations (CNV) [102,103] . Single-cell whole genome sequencing
offers a significant advantage over traditional methods in analyzing karyotypic anomalies and CNVs at a
much higher resolution.

Understanding tumor evolution
Tumor evolution is a dynamic process and describes the emergence of cancer cell subpopulations under
environmental pressure. As the tumor grows, each generation of cells acquire novel somatic mutations that
provide cells with survival advantages thereby determining the overall fitness of the clonal population [104] .
Waves of clonal expansion and contraction driven by changes in the tumor microenvironment govern the
life cycle of a tumor. Single-cell sequencing can potentially identify low abundance clones carrying driver
mutations, which can be further leveraged to refine therapeutic strategies. Although low abundance driver
mutations are possible to detect by deep exome sequencing, the fraction of cells carrying the mutation, or
the zygosity of the change (relevant for loss of function mutations in tumor suppressor genes) are hard to es-
timate without single cell sequencing. A computational approach to map single-cell mutational profile from
exome sequencing was successfully used to chart the chronological acquisition of mutations and create a
phylogenic map of tumor evolution in both glioblastoma multiforme and secondary acute myeloid leukemia
(AML) [105,106] . A similar analysis in breast cancer identified three clonal populations in the primary tumor
[74]
of which only one clone was present in the metastatic lesion . This observation supports the hypothesis
that rare clones present in the primary tumor harbor genetic signatures of metastasis even before they have
spread and colonized distant sites [74,107,108] . In a follow-up breast cancer study, aneuploidy rearrangements
were shown to occur early in tumor evolution, which remained highly stable as the tumor grew, whereas,
[77]
point mutations generated clonal diversity . A similar pattern is observed in lymphoblastic leukemia pa-
tients where recurrent translocations appear earlier than structural nucleotide variants [109] . This suggests that
large structural alterations offer selective advantage early during tumor growth followed by accumulation of
mutations producing clonal diversity. This is supported by the finding that subclonal populations arise more
frequently in tumors with high mutational burdens such as bladder and colon cancer, but not in tumors with
low mutational burden such as renal cell carcinoma [76,110,111] . A clonal progression of multiple mutations was
mapped in hematopoietic stem cells of AML patients, suggesting the clonal evolution of AML genomes from
founder mutations [112] . An interesting finding from single-cell analysis is that phenotypic diversity fails to re-
capitulate genotypic diversity detected in subclones strongly implicating that a large proportion of genotypic
variation may lack functional consequences, appearing and disappearing without contributing to tumor
evolution [113] .

Disease diagnosis and therapeutic stratification of patients
Modern cancer treatment relies heavily on accurate molecular and immuno/histopathological tissue analysis
of needle biopsies or surgically resected tissues for diagnosis. Tumor heterogeneity often confounds accuracy
of disease diagnostics by subsampling a subset of tumor cells that may not represent the whole tumor. This
calls for obtaining multiregional and longitudinal samples to guide therapeutic intervention, which is often
not routine. High-resolution single-cell analysis of tumor samples or CTCs can aid in refining diagnostic
parameters and patient stratification.

In a single-cell sequencing study of CTCs from metastatic lung cancer, patients who share the same subtype
of lung cancer displayed similar patterns of copy number variations in their CTCs, providing a potential