Share this article on:

Added Value of Whole-Exome and Transcriptome Sequencing for Clinical Molecular Screenings of Advanced Cancer Patients With Solid Tumors

Koeppel, Florence, PhD*; Bobard, Alexandre, PhD; Lefebvre, Céline, PhD; Pedrero, Marion, PhD; Deloger, Marc, PhD§; Boursin, Yannick, MSc§; Richon, Catherine*; Chen-Min-Tao, Romy, MSc§; Robert, Guillaume, MSc§; Meurice, Guillaume, PhD§; Rouleau, Etienne, PharmD, PhD*∥; Michiels, Stefan, PhD; Massard, Christophe, MD, PhD#**; Scoazec, Jean-Yves, MD, PhD; Solary, Eric, MD, PhD††‡‡; Soria, Jean-Charles, MD, PhD‡#**; André, Fabrice, MD, PhD‡#**; Lacroix, Ludovic, PharmD, PhD*∥

doi: 10.1097/PPO.0000000000000322
Original Article

Comprehensive genomic profiling using high-throughput sequencing brings a wealth of information, and its place in the clinical setting has been increasingly prominent. This review emphasizes the utility of whole-exome sequencing (WES) and transcriptome sequencing (RNAseq) in patient care and clinical research, based on published reports as well as our experience with the MOSCATO-01 (MOlecular Screening for CAncer Treatment Optimization) molecular triage trial at Gustave Roussy Cancer Center. In this trial, all contributive samples of patients with advanced solid tumors were analyzed prospectively with targeted gene sequencing (TGS) and comparative genomic hybridization. In addition, 92 consecutive metastatic patients with contributive biopsies were sequenced for WES and RNAseq and compared with TGS and comparative genomic hybridization. Whole-exome sequencing allowed the reporting of additional variants in relevant genes in 38% of patients. Mutation detection sensitivity of WES was 95% compared with TGS. Additional information derived from WES and RNAseq could influence clinical decision, including fusion transcripts, expression levels, allele-specific expression, alternate transcripts, RNA-based pathogen diagnostic, tumor mutation load, mutational signatures, expression signatures, HLA genotyping, and neoepitope prediction. The current challenge is to be able to process the large-scale data from these comprehensive genome-wide technologies in an efficient way.

From the *Genomic Core–Molecular Biopathology Unit and Biological Resource Center, AMMICA, INSERM US23/CNRS;

SIRIC-SOCRATE 2.0/IHU Cancer–Molecular Medicine in Oncology, Gustave Roussy;

INSERM U981, Gustave Roussy, Université Paris XI;

§Bioinformatic Core, AMMICA, INSERM US23/CNRS UMS3655, Gustave Roussy;

Department of Medical Biology and Pathology, Gustave Roussy;

Biostatistics and Epidemiology Unit, Gustave Roussy, Paris-Saclay University, CESP, INSERM, Paris-Sud University Faculty of Medicine

#Drug Development Department (DITEP), Gustave Roussy, and Paris-Sud University Faculty of Medicine

**Department of Medical Oncology, Gustave Roussy;

††INSERM U1170, Gustave Roussy, Paris-Sud University Faculty of Medicine, Paris-Saclay University; and

‡‡Department of Hematology, Gustave Roussy, Villejuif, France.

Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. This work was supported by grant INCa-DGOS-INSERM-6043.

Reprints: Florence Koeppel, PhD, Institut de Cancérologie Gustave Roussy, 114 rue Edouard Vaillant, 94805 Villejuif cedex, France. E-mail:

Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (

Recent advances in next-generation sequencing (NGS) technologies have opened a new era for the management of cancer and provide now an almost exhaustive view of molecular alterations in cancer. Although comprehensive molecular portraits are constantly fueling the concept of precision medicine, identifying reliable molecular markers of diseases, disease states, and drug response has proven to be challenging. Because precision medicine is highly dependent on technology, an inventory of strengths and drawbacks of each technology is needed to set up an adequate workflow to orient patients toward molecular targeted agents.1,2

Targeted gene sequencing (TGS) offers cost-effective, streamline sequencing of a specific panel of genes to identify single-nucleotide variations (SNVs) and small insertion and deletions (indels). Panels are therefore specific to defined applications (e.g., specific disease or pathway) and cover limited numbers of genes. A recurring question is how many genes between 1 and 24,000 should be included in a TGS panel to be efficient? In the case of the MOSCATO-01 (MOlecular Screening for CAncer Treatment Optimization) study, multiple solid tumor types at advanced stages were considered, and a hotspot panel with 74 genes was selected.

Comparative genomic hybridization (CGH) arrays and single-nucleotide polymorphism (SNP) arrays are whole-genome screening techniques based on DNA microarrays, enabling detection of copy number alterations (CNAs). Comparative genomic hybridization arrays are using immobilized large genomic inserts hybridized against differentially labeled test and reference samples, whereas SNP arrays are based on immobilized oligonucleotides sampling for SNPs in the genome with second readout for the actual copy number of the investigated genomic loci in a single detection color.

Whole-exome sequencing (WES) is the sequencing of the entire set of exons in the human coding genome (approximately 200,000 exons), after exome enrichment of genomic DNA. It is widely used to identify somatic driver mutations as well as CNA within gene-coding regions in various cancer types. It is also used for discovery of germline mutations underlying Mendelian disorders, as well as diagnosis. Indeed, gene-coding regions harbor more than 85% of mutations in monogenic diseases.3 It is noteworthy that both SNV/indels and CNA are also detected by WES.

An emerging trend in cancer molecular profiling is transcriptome sequencing (RNAseq). It is the sequencing of all protein-coding and noncoding transcripts (RNAs) in a given sample. Before sequencing, the coding component may be extracted from the RNA sample (poly(A) selection), or ribosomal RNA may be removed (ribodepletion). While the major advantage of RNAseq for clinical care is the detection of fusion transcripts, it is also relevant for the investigation of expression levels and expression signatures, alternative splicing patterns, and allele-specific expression when combined with DNA sequencing. The main characteristics of these technologies are summarized in Table 1.



This review addresses the current use of WES and RNAseq as well as their potential for clinical decision making in terms of scope, quality, and turnover time. To what extent is WES/RNAseq useful for the clinic? What is the concordance of the molecular profiling between the different approaches? What information can comprehensive genomic profiling add, compared with targeted sequencing approaches (TGS), currently and in the future? To our knowledge, few data are available comparing alterations identified with WES/RNAseq to TGS/CGH in the context of clinical practice in oncology, which is why we are including previously unpublished data collected as part of the MOSCATO-01 precision medicine triage trial.4 This is a follow-up on our previous report of the Gustave Roussy Cancer Center experience on genomics in the clinic.1

Currently, there are 135 genes with a known potential drug or diagnosis implication, based on the TARGET database v3 (archive. What is at stake here is the full description of cancer genomic alterations. This article offers a return on experience on the TGS/CGH and WES/RNAseq technologies used side by side.

Back to Top | Article Outline


The choice of the sequencing technique (Table 1) depends on several factors, including the nature of the starting material: solid biopsy (primary site, metastasis, lymph node), liquid biopsy (normal DNA, cell-free circulating DNA extracted from plasma, circulating tumor cells, urine, feces, saliva, ascites), the tissue preservation method (fresh frozen or formalin-fixed samples), and the type of genomic analysis that is needed (detection of SNV and multinucleotide variants, indels, CNA and fusion transcripts, global analysis such as tumor mutation load [TML], mutation signatures, or expression signatures; Fig. 1).



An increasing number of publications are related to “whole-exome sequencing” (Fig. 2), reflecting the still growing interest in using WES compared with other sequencing strategies. However, implementing WES in the clinical setting is facing several challenges, such as analytic performance (sensitivity, specificity, robustness, etc.), turnaround time, cost, and constraints on starting material (quantity, preservation method). In addition, the management of such large-scale data, associated to complexity of the bioinformatics analysis, and the complex interpretation of results,5 the traceability of the whole workflow, and release of comprehensive report for clinicians are numerous important points to be solved and controlled. Yet, expected benefits trigger several teams to take this challenge through dedicated precision medicine projects. Table 2 lists examples of molecular triage trials that include WES as part of the prospective molecular profiling.





Large panels represent a compromise between small panels and comprehensive sequencing. Clinical benefits and findings were recently reported.9,15,16 The results of the MSKCC-IMPACT program were evaluated based on the first 10,000 patients who had a molecular portrait.17 Nearly 37% had at least 1 actionable mutation, and 11% of patients were able to participate in clinical trials of treatments that directly targeted the genetic alterations in their tumors. Large panels may cover regions with most of the clinically relevant known variants, as well as address known fusions and amplifications. Yet they do not identify unexpected alterations. Also, there is yet no consensus on the list of genes to screen. For example, the FoundationOne 315 gene panel and the MSKCC-IMPACT 410 gene panel have only 248 genes in common (80%). Also, the WES approach ensures coverage of all genes analyzed in various panels.

The ongoing ASCO Targeted Agent and Profiling Utilization Registry and the NCI Molecular Analysis for Therapy Choice (MATCH) trials18 demonstrate the potential translatability of comprehensive molecular sequencing, despite major remaining challenges on interpreting the results, integrating additional information and matching molecular portraits to the best therapeutic strategy.

In France, 28 molecular genetics cores are providing TGS-based testing for care. This program enables all cancer patients to receive free testing for biomarkers, such as KRAS, EGFR, KIT, and BRAF mutations. Funding for these laboratories comes from the French Ministry of Health. In 2015, 75,000 patients benefited from molecular predictive tests, among which were 15,000 NGS tests. The number of NGS platforms settled in France increased from 1 to 44 between 2014 and 2017 (gen&tiss communication). In 2016, the French government launched the “France Genomic Medicine 2025,” a plan aiming to develop very high-throughput genetic analysis in standard care. By 2020, 235,000 whole exomes or whole genomes are planned to be sequenced each year for clinical oncology and rare genetic diseases.19

Gustave Roussy has been performing multiplex gene mutation screening since 2013, delivering results on a panel of 74 genes in an average time frame of 10 days from sample reception to reporting. Whole-exome sequencing and RNAseq analyses have been implemented in Gustave Roussy since 2015 through a locally based core operated by a privately held company—IntegraGen—with an average time frame of 20 days from the isolation of nucleic acids to the release of bioinformatics analysis. In the MOSCATO-01 trial, biopsy samples are analyzed prospectively with TGS, CGH, and RNAseq. In addition, WES is performed for selected conditions and when the patient is a candidate for specific trials including immunotherapy.

Back to Top | Article Outline


Key Assets and Drawbacks of WES, TGS, and CGH

Whole-exome sequencing by definition covers a much larger set of genes than TGS, yet sequencing depth is typically much lower because of cost considerations. Targeted gene sequencing therefore presents a more favorable signal-to-noise ratio and is relatively straightforward to analyze. The increase in the number of genes with large-panel TGS (≥300 genes) represents a compromise to achieve high coverage rates on a limited number of regions of interest. Yet, it is not optimal for regions with repeats, and issues with design may lead to low coverage areas. In addition, panels have to continually adapt to new biomarkers for innovative targeted therapy discoveries and do not allow for the collection of retrospective information for such novelties. Moreover, in most panel designs, intron-exon junctions are not always fully covered, and therefore some structural variants are not detected.

Back to Top | Article Outline

Identification of Single-Nucleotide Variants and Small Insertion-Deletions

Detection of somatic driver mutations that occur during the development of a tumor is currently the primary application of NGS in oncology.20 There is a general agreement that TGS is the standard method when several genes must be tested in the same patient. In order to be used for clinical practice, WES should at least present equivalent analytical validity to TGS for main recurrent driver mutations.21

To our knowledge, few data are available in the literature comparing alterations identified with WES to TGS in the context of clinical practice in oncology, yet comparative results are concordant.22 In the MOSCATO-01 trial, biopsy DNA samples were analyzed independently by TGS and WES. Variants derived from both technologies were validated by an expert geneticist and compared. Of 132 TGS-validated variants, 126 were also found in WES; therefore, WES sensitivity was 95% compared with TGS. Variants missing in WES corresponded to constitutional variants classified as variants with unknown pathogenicity or variants with allele frequencies of less than 12% in low coverage depth area. Median depth was on average 83×/131× in normal/tumor WES, compared with 831× in TGS (Supplemental Digital Content 1,, for material and methods). Allele frequencies determined by WES and TGS were correlated (correlation coefficient, 0.906; Fig. 3A).



Despite the lower sequencing depth and the lower sensitivity linked to a limited coverage depth on the relevant genes, WES is able not only to substitute for TGS but also to enlarge SNV detection by identifying more rare variants of cancer-related genes. Indeed, sequencing studies have shown that, in addition to the high prevalence of recurrent drivers, most tumor types are also characterized by a “long tail” of gene alterations in a small proportion (<1%) of patients,23 and some of these rare genomic alterations may impact the clinical management of patients.

In the MOSCATO-01 trial, a list of 271 cancer-relevant genes was set, based on van Allen et al.24 These listed genes were given high priority in the WES variant calling, and pathogenic variants or variants suspected to be pathogenic in these genes were considered relevant variants and included in the clinical report. Whole-exome sequencing allowed the reporting of additional variants in relevant genes in 38% of patients (35 of 92 patients, including 27 patients for whom TGS had already identified relevant variants and 8 patients for whom TGS had not; Fig. 4A). Among patients with no relevant variant on TGS, WES identified 1 or more relevant variants in 29% of cases (8 of 28 patients), similarly to a previous report.25 As expected, the prevalence of unknown variants was higher in “WES only” variants (Fig. 4B).



Back to Top | Article Outline

Identification of CNAs

Comparative genomic hybridization and WES efficiency in identifying CNAs have been reported to be equivalent, despite slight differences.26–29 Whole-exome sequencing offers single-nucleotide resolution and absolute counts of read numbers and therefore can provide more sensitive and accurate CNA results. Moreover, direct sequencing enables substantial increases in discoveries of smaller structural variation events.30 Figure 5 shows concordance of copy number profiles derived from CGH and WES.



Back to Top | Article Outline

Starting Material

Whole-exome sequencing requires typically 5 to 10 times more DNA input than TGS (Table 1), but recent technical advances have allowed analysis with amounts compatible with most biopsy samples. When biopsy is not available or not contributive because of low tumor cell content, molecular profiling can be performed using circulating free DNA (cfDNA) isolated from liquid biopsy.31 The latter approach can circumvent all the drawbacks associated with on-purpose biopsy (medical risk, organizational constraints, and additional costs). Also, because tissue biopsies are localized, they are not representative of the tumor heterogeneity. Circulating free DNA enables a more integrated view of the tumor profile, with mutations potentially originating from multiple clones and multiple sites (primitive and/or metastatic), even though all sites may not be represented equally in cfDNA, with necrotic sites releasing more DNA in particular. However, the quantity of tumor-derived DNA in cfDNA is low and variable, depending on the tumor type and stage, leading to a lower sensitivity than biopsy profiling. Feasibility of both TGS and WES has been demonstrated on cfDNA.32–34 Yet, RNA remains extremely difficult to isolate from plasma because of low quantity and short half-life, and there is no report yet of RNAseq using plasma samples.

Back to Top | Article Outline

Turnaround Time and Cost

Turnaround time is critical for therapeutic decision making. The maximum acceptable delay between biopsy and clinical report editing was set to be 21 days. This time frame was reached at Gustave Roussy and other sites after optimization.35 Yet, an improvement of the turnaround time would be beneficial because it is still significantly longer than immunohistochemistry or targeted sequencing. Also, informatics tools and knowledge databases are strongly needed to help the clinical annotation of variants and speed up the reporting process, in order to facilitate interpretation of results, decrease the time spent by expert geneticists, and therefore decrease the overall delay to clinical decision. Overall, WES can substitute for TGS/CGH for identifying mutations and CNAs in a time frame that is considered acceptable for diagnosed cancer patients.

A key aspect to consider is the significantly higher cost of very high-throughput sequencing compared with TGS and CGH analysis. It includes both sequencing costs and indirect costs such as bioinformatics analysis, which may be automatized, but mostly data storage and structuration, despite the sharp decline in recent years. Moreover, the higher cost of WES is also related to the requirement for normal tissue analysis for somatic variant calling; nevertheless, enlargement of TGS panel to several hundred genes could lead to the same requirement and therefore increase TGS cost.

Overall, WES is not cost-effective compared with TGS when considering only hotspot mutations screening. Yet, when considering all the information derived from WES and comparing WES to stepwise testing, WES is effective regarding cost, turnover time, and amount of starting material. Whole-exome sequencing is of particular value in the following cases: when no actionable mutations are found in guidelines-recommended testing36 (although the added value of this may depend on the histotype25), for a disease that is refractory to standard lines of therapy,4 for rare tumor types,37,38 for tumors of unknown primary origin,39 and also for patients who are suspected to respond positively to immunotherapies or to combinations of targeted therapies and immunotherapies.

Back to Top | Article Outline


Fusion Transcripts

The most valuable application of RNAseq for the clinic today is fusion transcripts identification. Fusions are relevant for diagnosis (e.g., EWSR1/FLI1 in Ewing sarcoma,40 DNAJB1/PRKACA in fibrolamellar hepatocellular carcinoma41), theranostics (e.g., FGFR3/TACC3 in glioblastoma,42 CD74/ROS1 in lung adenocarcinoma43), and prognosis (e.g., TMPRSS2/ERG in prostate cancer44).

RNAseq allows a 1-shot screening of all possible fusions including fusion with theranostic impact, not limited to 1 histological type such as ALK, RET, and FGFR fusion.45 Among the 92 patients, 26 patients had no fusion identified by bioinformatics analysis, whereas 66 patients harbored a total of 144 fusions. Thirty of these fusions had a high Oncofuse46 prediction score of greater than 0.8 (27 patients). Among all the fusion transcripts identified from the RNAseq, 15 fusions were already described in the literature (e.g., PAX7/FOXO1 in alveolar rhabdomyosarcoma47); the others were not yet known (e.g., HBS1L/RAD51B in nasopharynx carcinoma). For the latter ones, a functional characterization is needed to address their potential role as cancer drivers. It is also noteworthy that fusion proteins are usually more easily druggable than mutated proteins.

Back to Top | Article Outline

TML and Mutational Signature

Tumor mutation load results from a global analysis of WES and is valuable to the clinical practice. Tumor mutation load has also been derived from large-panel TGS using dedicated algorithms.48 A range of TML values was obtained using WES data from MOSCATO-01 patients (Supplemental Digital Content 2,, displaying both intertumor and intratumor type heterogeneity. Immune checkpoint inhibitors (anti-cytotoxic T-Lymphocyte Associated Protein 4, anti-programmed cell death-1, and anti–programmed cell death ligand-1) improve survival in certain cancers (melanoma, lung cancer, bladder cancer, etc.), and several studies have suggested that high tumor mutational burden constitutes a valuable biomarker to identify patients who would benefit from those therapies.49–51 Tumor heterogeneity may also predict emergence of subclones at relapse.50,52

In addition, mutational signatures are derived not only from WGS53 but also WES data. In the MOSCATO-01 WES subset of data, the APOBEC signature inferred from WES was correlated with a higher expression of APOBEC3B based on RNAseq expression levels (correlation coefficient, 0.44; P = 1.7e−5; Supplemental Digital Content 3, The tobacco signature was correlated with the smoking status (Wilcoxon test P = 0.02). Unlike TGS, WES is able to bring information about the mutational status of the tumor, thus giving an insight into the dominant mutational processes, even though this information does not currently have theranostic value, and more research is needed on this topic.

Back to Top | Article Outline

Analysis of Germline Alterations

Use of a normal sample is strongly recommended for WES.24 Normal sample sequencing allows distinguishing somatic variants from germline variants by performing a differential analysis. It decreases the complexity of interpretation with identification of rare germline polymorphisms, which are not filtered out using population frequency criteria. Sequencing normal samples also enables the reporting of incidental findings, assuming the patient agreed to it. Beyond care applications, WES offers major advantages for research and discovery applications. Whole-exome sequencing allows the creation of extremely rich databases, available for translational research, constitution of cohorts based on molecular traits, and also the development of new analysis pipelines to derive more complex and robust information from the sequencing data.

It is important to also consider the impact of normal tissue analysis in terms of legal aspects related to germline variant screening. When WES is performed to identify somatic aberrations, it often includes simultaneous analysis of germline DNA from a matched normal sample as a control and as such could lead to incidental detection of genetic susceptibilities. The American College of Medical Genetics and Genomics has recommended analysis of 56 specific genes associated with hereditary syndromes, including cancer. The consensus reached is that reporting of incidental and secondary findings would likely have medical benefits for the patients undergoing clinical NGS testing as well as their families.54 These considerations emphasize the need to train clinicians prescribing genomic tests in order to correctly inform patients about the possibilities and consequences of incidental genetic findings.

Back to Top | Article Outline

Expression Levels

RNAseq-derived expression levels provide precious identification of highly expressed oncogenic targets. Beyond that, it also provides a measure of oncogenic pathway activation, as it offers an increased dynamic range of expression compared with RNA microarrays.55 In addition, RNAseq may be used to detect transcript isoforms and splice variants (e.g., exon 14 splice site in MET 56) and decipher the immune contexture based on expression signatures.57,58 RNAseq could also be used as a surrogate for immunohistochemistry in case of screen failure or technical challenges.59

Indeed in MOSCATO-01, gene expression evaluated by RNAseq was compared with immunohistochemistry staining for HER2 and ESR1 in 7 breast cancer samples and was found to be fully concordant (Supplemental Digital Content 4,

Another clinically relevant input of RNAseq is to independently confirm expression of driver variants identified in DNA analysis.60 This is possible only for genes with significant expression levels. In our data, 75% of curated TGS variants were found expressed in RNAseq (Fig. 3B). When considering all mutations obtained from targeted sequencing, some had very different allele frequencies between WES and RNAseq, indicating allelic imbalance (Fig. 6). In addition, the integrated analysis of WES and RNAseq can document a homozygous expression of deleterious variants or correlate overexpression with amplification or gene fusion.



Back to Top | Article Outline

HLA Typing and Neoepitope Prediction

Whole-exome sequencing– and RNAseq-derived information could assess the immunogenicity of a tumor and may be able to identify patients that would benefit from immunotherapeutic interventions. However, the relevance of the predicted neoepitopes still needs to be evaluated using biological models.61

In order to benchmark HLA genotyping algorithms, 10 MOSCATO samples were selected based on their WES-derived TML (5 samples with a high TML and 5 samples with a low TML). The concordance between the criterion standard method and the algorithmic HLA genotyping was found to be 100% with all 3 algorithms tested for the second-digit HLA type and, respectively, 98.3%, 100%, and 95% for Optitype, HLA-VBSeq, and Phlat for the fourth digit based on using the normal control DNA sequences from WES. Based on these results, Optitype was used to determine major histocompatibility complex class I alleles for the 92 samples with WES. A list of putative neoantigens could be predicted from WES for each patient. Neoantigens were then shortlisted based on their RNAseq-derived expression levels.

Back to Top | Article Outline

Viral Infection Diagnostic

RNAseq is also applicable for molecular detection of several pathogens as viral infection. In particular, human papillomavirus (HPV) has been identified as a causative agent in a subset of oropharyngeal squamous cell carcinoma62 and is a marker of good prognosis.63

Human papillomavirus status can be derived from RNAseq data. In MOSCATO-01, genotypes of HPV16, HPV18, and HPV33 were quantified from RNAseq and from DNA-based polymerase chain reaction (PCR) in all 9 head and neck cancer patients available in our WES subcohort. RNAseq and PCR results were concordant (Supplemental Digital Content 5,, even though the PCR assay was based on DNA and not RNA. HPV16 was found in 2 patients, whereas the other types were not found. Concordance was verified on a larger set of 34 patients with head and neck cancer (data not shown), among which 7 were positive for HPV16 with both techniques and all were negative for HPV18 and HPV33.

Back to Top | Article Outline

Investigation of Resistance and Clonality

Finally, besides providing molecular triggers (actionable drivers) and immunological-related information (TML, signatures, neoepitope characterization), WES and RNAseq can bring clues to solve the major challenge of precision medicine: acquired resistance. Indeed, several mechanisms of resistance are known and may be investigated specifically, but many are still unexpected. Digging into these sequences has the potential to shed light on resistance mechanisms elicited by targeted therapies such as signaling bypass, metabolic adaptation, or apoptosis resistance, allowing the discovery of patient's specific mechanisms of resistance. Alternatively, information from WES/RNAseq at the time of diagnosis could predict the arising of resistance in the form of a genomic alteration or an aberrant mRNA expression. This last example highlights the difference between a hypothesis-driven method and a blind nonbiased approach.

Back to Top | Article Outline


The studies reviewed and the data presented here support that integrating comprehensive screening of genomic alterations for therapeutic decision making is feasible and provides an opportunity for improved clinical decision. Whole-exome sequencing and RNAseq offer a global and unbiased view of genomic abnormalities involved in cancer development and have the potential to impact clinical decision and guide treatment with matched targeted therapies available in clinical practice or clinical trials.

On the translational research side, the data collected by WES and RNAseq constitute an extremely rich database to further investigate cancer development, including clonal heterogeneity. This research would be potentiated drastically by increasing data sharing between centers. Yet, sharing genomic data is not enough; it needs to be associated with high-quality curated clinical data, such as those collected through case report forms. This process will be crucial to define the best oncogenic drivers at the individual level and the best therapeutic approaches in the future.

Back to Top | Article Outline


The authors acknowledge the clinical team, including Dr. Charles Ferté, Dr. Antoine Hollebecque, Dr. Loic Verlingue, Dr. Rastilav Bahleda, Dr. Anas Gazzah, Dr. Andrea Varga, Dr. Sophie Postel-Vinay, Dr. Yohann Loriot, Dr. Caroline Even, Dr. Thierry De Baere, Dr. Frederic Deschamps, Dr. Philippe Vielh, and Dr. Vincent Ribrag, for their implication in the enrollment of patients in MOSCATO-01 trial, and Maud Ngo-Camus, Aljosa Celebic for Clinical Data Management. The authors also acknowledge the laboratory and bioinformatics teams, including Mélanie Laporte, Isabelle Miran, Aurélie Honoré, Gladwys Faucher, Zsofia Balogh, Jonathan Sabio, Lionel Fougeat, Marie Xiberras, Leslie Girard, Catherine Lapage, and Bastien Job, for their support in the preanalytic processing of samples, storage, and tumor sample analysis (wet lab and bioinformatics) of the patients included in MOSCATO-01 trial.

Back to Top | Article Outline


1. Lacroix L, Boichard A, André F, et al. Genomes in the clinic: the Gustave Roussy Cancer Center experience. Curr Opin Genet Dev. 2014;24:99–106.
2. Cummings CA, Peters E, Lacroix L, et al. The role of next-generation sequencing in enabling personalized oncology therapy. Clin Transl Sci. 2016;9:283–292.
3. Botstein D, Risch N. Discovering genotypes underlying human phenotypes: past successes for mendelian disease, future approaches for complex disease. Nat Genet. 2003;33(3S):228–237.
4. Massard C, Michiels S, Ferté C, et al. High-throughput genomics and clinical outcome in hard-to-treat advanced cancers: results of the MOSCATO 01 trial. Cancer Discov. 2017;7:586–595.
5. Hintzsche JD, Robinson WA, Tan AC. A survey of computational tools to analyze and interpret whole exome sequencing data. Int J Genomics. 2016;2016:7983236.
6. Facchinetti F, Loriot Y, Kuo MS, et al. Crizotinib-resistant ROS1 mutations reveal a predictive kinase inhibitor sensitivity model for ROS1- and ALK-rearranged lung cancers. Clin Cancer Res. 2016;22:5983–5991.
    7. Horak P, Klink B, Heining C, et al. Precision oncology based on omics data: the NCT Heidelberg experience. Int J Cancer. 2017;141:877–886.
      8. Mody RJ, Wu YM, Lonigro RJ, et al. Integrative clinical sequencing in the management of refractory or relapsed cancer in youth. JAMA. 2015;314:913.
        9. Cobain EF, Robinson DR, Wu YM, et al. Clinical application of comprehensive next generation sequencing in the management of metastatic cancer in adults. J Clin Oncol. 2017;35(15 suppl):101.
        10. Tuxen IV, Yde CW, Mau-Sørensen M, et al. Copenhagen prospective personalized oncology (CoPPO): genomic profiling to select patients for phase 1 trials. Ann Oncol/ESMO 2016 Congr. 2016;27:545–551.
          11. Ghazani AA, Oliver NM, St Pierre JP, et al. Assigning clinical meaning to somatic and germ-line whole-exome sequencing data in a prospective cancer precision medicine study. Genet Med. 2017;19:787–795.
            12. Van Allen EM, Wagle N, Stojanov P, et al. Whole-exome sequencing and clinical interpretation of formalin-fixed, paraffin-embedded tumor samples to guide precision cancer medicine. Nat Med. 2014;20:682–688.
              14. Beltran H, Eng K, Mosquera JM, et al. Whole-exome sequencing of metastatic cancer and biomarkers of treatment response. JAMA Oncol. 2015;1:466–474.
                15. Hyman DM, Taylor BS, Baselga J. Implementing genome-driven oncology. Cell. 2017;168:584–599.
                16. Robinson DR, Wu YM, Lonigro RJ, et al. Integrative clinical genomics of metastatic cancer. Nature. 2017;548:297–303.
                17. Zehir A, Benayed R, Shah RH, et al. Mutational landscape of metastatic cancer revealed from prospective clinical sequencing of 10,000 patients. Nat Med. 2017;23:703–713.
                18. Biankin AV, Piantadosi S, Hollingsworth SJ. Patient-centric trials for therapeutic development in precision oncology. Nature. 2015;526:361–370.
                19. Lévy Y. Genomic medicine 2025: France in the race for precision medicine. Lancet. 2016;388:2872.
                20. Vogelstein B, Papadopoulos N, Velculescu VE, et al. Cancer genome landscapes. Science. 2013;339:1546–1558.
                21. Frampton GM, Fichtenholtz A, Otto GA, et al. Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing. Nat Biotechnol. 2013;31:1023–1031.
                22. Uzilov AV, Ding W, Fink MY, et al. Development and clinical application of an integrative genomic approach to personalized cancer therapy. Genome Med. 2016;8:62.
                23. Lawrence MS, Stojanov P, Mermel CH, et al. Discovery and saturation analysis of cancer genes across 21 tumour types. Nature. 2014;505:495–501.
                24. Van Allen EM, Wagle N, Stojanov P, et al. Whole-exome sequencing and clinical interpretation of FFPE tumor samples to guide precision cancer medicine. Nat Med. 2014;20:682–688.
                25. Postel-Vinay S, Boursin Y, Massard C, et al. Seeking the driver in tumours with apparent normal molecular profile on comparative genomic hybridization and targeted gene panel sequencing: what is the added value of whole exome sequencing? Ann Oncol. 2016;27:344–352.
                26. Alkodsi A, Louhimo R, Hautaniemi S. Comparative analysis of methods for identifying somatic copy number alterations from deep sequencing data. Brief Bioinform. 2015;16:242–254.
                27. Guo Y, Sheng Q, Samuels DC, et al. Comparative study of exome copy number variation estimation tools using array comparative genomic hybridization as control. Biomed Res Int. 2013;2013:915636.
                28. Duan J, Zhang J-G, Deng H-W, et al. Comparative studies of copy number variation detection methods for next-generation sequencing technologies. PLoS One. 2013;8:e59128.
                29. Retterer K, Scuffins J, Schmidt D, et al. Assessing copy number from exome sequencing and exome array CGH based on CNV spectrum in a large clinical cohort. Genet Med. 2015;17:623–629.
                30. Wang X, Li X, Cheng Y, et al. Copy number alterations detected by whole-exome and whole-genome sequencing of esophageal adenocarcinoma. Hum Genomics. 2015;9:22.
                31. Jovelet C, Ileana E, Le Deley MC, et al. Circulating cell-free tumor DNA analysis of 50 genes by next-generation sequencing in the prospective MOSCATO trial. Clin Cancer Res. 2016;22:2960–2968.
                32. Murtaza M, Dawson S-J, Pogrebniak K, et al. Multifocal clonal evolution characterized using circulating tumour DNA in a case of metastatic breast cancer. Nat Commun. 2015;6:8760.
                33. Dietz S, Schirmer U, Mercé C, et al. Low input whole-exome sequencing to determine the representation of the tumor exome in circulating DNA of non–small cell lung cancer patients. PLoS One. 2016;11.
                34. Koeppel F, Blanchard S, Jovelet C, et al. Whole exome sequencing for determination of tumor mutation load in liquid biopsy from advanced cancer patients. PLoS One. 2017;12:e0188174.
                35. Roychowdhury S, Iyer MK, Robinson DR, et al. Personalized oncology through integrative high-throughput sequencing: a pilot study. Sci Transl Med. 2011;3:111ra121.
                36. Craig DW, O'Shaughnessy JA, Kiefer JA, et al. Genome and transcriptome sequencing in prospective metastatic triple-negative breast cancer uncovers therapeutic vulnerabilities. Mol Cancer Ther. 2013;12:104–116.
                37. Dela Cruz FS, Diolaiti D, Turk AT, et al. A case study of an integrative genomic and experimental therapeutic approach for rare tumors: identification of vulnerabilities in a pediatric poorly differentiated carcinoma. Genome Med. 2016;8:116.
                38. Menezes J, Acquadro F, Wiseman M, et al. Exome sequencing reveals novel and recurrent mutations with clinical impact in blastic plasmacytoid dendritic cell neoplasm. Leukemia. 2014;28:823–829.
                39. Gröschel S, Bommer M, Hutter B, et al. Integration of genomics and histology reveals diagnosis and effective therapy of refractory cancer of unknown primary with PDL1 amplification. Oncol Res Treat. 2016;39:183–184.
                40. Gomez NC, Davis IJ. Linking germline and somatic variation in Ewing sarcoma. Nat Genet. 2015;47:964–965.
                41. Honeyman JN, Simon EP, Robine N, et al. Detection of a recurrent DNAJB1-PRKACA chimeric transcript in fibrolamellar hepatocellular carcinoma. Science. 2014;343:1010–1014.
                42. Di Stefano AL, Fucci A, Frattini V, et al. Detection, characterization, and inhibition of FGFR-TACC fusions in IDH wild-type glioma. Clin Cancer Res. 2015;21:3307–3317.
                43. Awad MM, Katayama R, McTigue M, et al. Acquired resistance to crizotinib from a mutation in CD74-ROS1. N Engl J Med. 2013;368:2395–2401.
                44. Saramäki OR, Harjula AE, Martikainen PM, et al. TMPRSS2:ERG fusion identifies a subgroup of prostate cancers with a favorable prognosis. Clin Cancer Res. 2008;14:3395–3400.
                45. Buzyn A, Blay J-Y, Hoog-Labouret N, et al. Equal access to innovative therapies and precision cancer care. Nat Rev Clin Oncol. 2016;13:385–393.
                46. Shugay M, Ortiz de Mendíbil I, Vizmanos JL, et al. Oncofuse: a computational framework for the prediction of the oncogenic potential of gene fusions. Bioinformatics. 2013;29:2539–2546.
                47. Kashi VP, Hatley ME, Galindo RL. Probing for a deeper understanding of rhabdomyosarcoma: insights from complementary model systems. Nat Rev Cancer. 2015;15:426–439.
                48. Roszik J, Haydu LE, Hess KR, et al. Novel algorithmic approach predicts tumor mutation load and correlates with immunotherapy clinical outcomes using a defined gene mutation set. BMC Med. 2016;14:168.
                49. Rosenberg JE, Hoffman-Censits J, Powles T, et al. Atezolizumab in patients with locally advanced and metastatic urothelial carcinoma who have progressed following treatment with platinum-based chemotherapy: a single-arm, multicentre, phase 2 trial. Lancet (London, England). 2016;387:1909–1920.
                50. Rizvi NA, Hellmann MD, Snyder A, et al. Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non–small cell lung cancer. Science. 2015;348:124–128.
                51. Snyder A, Makarov V, Merghoub T, et al. Genetic basis for clinical response to CTLA-4 blockade in melanoma. N Engl J Med. 2014;2189–2199.
                52. Van Allen EM, Miao D, Schilling B, et al. Genomic correlates of response to CTLA-4 blockade in metastatic melanoma. Science (80-). 2015;350:207–211.
                53. Alexandrov LB, Nik-Zainal S, Wedge DC, et al. Signatures of mutational processes in human cancer. Nature. 2013;500:415–421.
                54. Green RC, Berg JS, Grody WW, et al. ACMG recommendations for reporting of incidental findings in clinical exome and genome sequencing. Genet Med. 2013;15:565–574.
                55. Byron SA, Van Keuren-Jensen KR, Engelthaler DM, et al. Translating RNA sequencing into clinical diagnostics: opportunities and challenges. Nat Rev Genet. 2016;17:257–271.
                56. Tong JH, Yeung SF, Chan AW, et al. MET amplification and exon 14 splice site mutation define unique molecular subgroups of non–small cell lung carcinoma with poor prognosis. Clin Cancer Res. 2016;22:3048–3056.
                57. Becht E, Giraldo NA, Germain C, et al. Immune contexture, immunoscore, and malignant cell molecular subgroups for prognostic and theranostic classifications of cancers. In: Advances in Immunology. 2016:95–190.
                58. Fridman WH, Pagès F, Sautès-Fridman C, et al. The immune contexture in human tumours: impact on clinical outcome. Nat Rev Cancer. 2012;12:298–306.
                59. Paluch BE, Glenn ST, Conroy JM, et al. Robust detection of immune transcripts in FFPE samples using targeted RNA sequencing. Oncotarget. 2017;8:3197–3205.
                60. Ku CS, Wu M, Cooper DN, et al. Exome versus transcriptome sequencing in identifying coding region variants. Expert Rev Mol Diagn. 2012;12:241–251.
                61. Strønen E, Toebes M, Kelderman S, et al. Targeting of cancer neoantigens with donor-derived T cell receptor repertoires. Science. 2016;352:1337–1341.
                62. Vokes EE, Agrawal N, Seiwert TY. HPV-associated head and neck cancer. J Natl Cancer Inst. 2015;107:1–7.
                63. Rischin D, Young RJ, Fisher R, et al. Prognostic significance of p16INK4A and human papillomavirus in patients with oropharyngeal cancer treated on TROG 02.02 phase III trial. J Clin Oncol. 2010;28:4142–4148.

                Exome; neoplasm; sequence analysis, DNA; targeted therapy

                Supplemental Digital Content

                Back to Top | Article Outline
                Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.