Genomic and Transcriptomic Characterization of Pre-operative Chemotherapy Response in Patients with Osteosarcoma

doi:10.21203/rs.3.rs-3140734/v1

Download PDF

Article

Genomic and Transcriptomic Characterization of Pre-operative Chemotherapy Response in Patients with Osteosarcoma

https://doi.org/10.21203/rs.3.rs-3140734/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 27 Nov, 2023

Read the published version in Scientific Reports →

You are reading this latest preprint version

Osteosarcoma is a heterogeneous disease with regard to its chemotherapy response and clinical outcomes. This study aims to investigate the genomic and transcriptomic characteristics related to pre-operative chemotherapy response. Samples from 25 osteosarcoma patients were collected to perform both whole exome and transcriptome sequencing. Osteosarcoma had the relatively high level of chromosomal copy number variant (CNV) burden. Chemotherapy responders showed the higher chromosomal CNV burden than non-responders (p = 0.0775). The percentage of COSMIC signature 3, associated with homologous recombination repair deficiency, was higher in responders (56%) than in non-responders (45%). Transcriptomic analysis suggested that 11 genes were significantly up-regulated in responders and 18 genes were up-regulated in non-responders. Both GSEA and KEGG enrichment analysis indicted that four pathways related to cardiomyopathy were up-regulated in responders, while neuroactive ligand−receptor interaction was up-regulated in non-responders. Finally, a random forest-based classifier was developed to classify chemotherapy response using 23 differentially expressed genes, with the area under the curve of 0.843 (95% CI, 0.654-1.000). Osteosarcoma had the heterogeneous mutational profile with frequent occurrence of CNVs. Transcriptomic analysis identified several signaling pathways associated with chemotherapy responsiveness to osteosarcoma. Transcriptomic characteristics classifier provides a potential research direction for predicting the chemotherapy response.

Health sciences/Oncology/Cancer/Bone cancer

Health sciences/Medical research/Biomarkers/Predictive markers

Health sciences/Medical research/Translational research

Osteosarcoma is the most common primary malignant tumor of bone. Although it is a rare disease in all ages, primary osteosarcoma commonly occurs in children and adolescents aged 10–24 years. According to the statistical analyses of 5 016 osteosarcoma patients from the Surveillance, Epidemiology, and End Results program from 1975 to 2017¹, the age-adjusted incidence of osteosarcoma was 3.3 / 1 000 000 in all age groups and 7.2 / 1 000 000 in population aged 10–24 years. Pre-operative (neoadjuvant) chemotherapy followed by definite surgery and postoperative (adjuvant) chemotherapy is the mainstay of therapy for osteosarcoma, significantly improving the survival and prognosis in approximately two-thirds of patients with localized disease².

Despite the remarkable efficacy of chemotherapy, there is a lack of uniform and objective criteria for evaluating the chemotherapy response in osteosarcoma patients. How to evaluate the effectiveness of chemotherapy, especially pre-operative chemotherapy, is crucial for the adjustment of treatment plans and patient prognosis assessment. Currently, the commonly applied evaluation method to determine the chemotherapy response and tumor prognosis is based on the tumor cell necrosis rate^3,4. However, the evaluation and application of tumor cell necrosis rate have certain limitations: the procedure of tumor necrosis rate analysis is complex; heterogeneous tumors are difficult to carry out; the evaluation time required is long, resulting in a late outcome which is not conducive to subsequent treatment. Huvos⁵ first proposed a histological evaluation method for osteosarcoma necrosis rate in the 1970s, which has been used for more than 40 years. However, this method cannot be applied to evaluate the chemotherapy response before surgery, and it does not comprehensively consider clinical and imaging aspects. Therefore, it is necessary to understand the molecular characteristics related to chemotherapy response and find a new effective evaluation method for chemotherapy response in osteosarcoma.

The rapid development and widespread application of high throughput sequencing techniques promote the understanding of tumor development and progression from the molecular level, and thus change the clinical treatment modes and survival outcomes of patients in various of cancers, especially in non-small cell lung cancer and colorectal cancer^6,7. In addition to supporting treatment decisions for cancer patients, next-generation sequencing has also been utilized in multiple clinical settings, including detecting tumor susceptibility genes, early diagnosis and screening, evaluating patient prognosis and monitoring minimal residual diseases and tumor progression^8–11. This study aims to investigate the genomic and transcriptomic characteristics of osteosarcoma using DNA- and RNA-based next-generation sequencing techniques, find clinical and molecular factors related to chemotherapy response and try to construct a machine learning-based classifier to predict chemotherapy response.

Clinicopathologic characteristics and evaluation of chemotherapy response

The clinicopathologic features of 25 patients with pathologic diagnoses of osteosarcoma were summarized in Supplementary Table S2. The median age at diagnosis was 15 years (range 8–32 years) and 75% patients were male. The most common site of osteosarcoma in this study was femur (60%), followed by tibia (20%), humerus (12%), fibula (4%) and multifocal (4%). Majority of the patients were conventional osteosarcoma (88%) and SSS stage IIB (92%). The tumor volume ranged from 24.633 to 14705.856 cm³ with the median of 371.57 cm³. Pre-operative chemotherapy and surgery were conducted in 84% and 92% of patients, respectively. Pathological evaluation suggested that the tumor necrosis rate was higher than 90% in 36% of patients. Increased ossification, clear margin, decreased blood supply and decreased circumference after chemotherapy were observed in 52%, 36%, 40% and 36%, respectively.

To evaluate the chemotherapy response, we designed an integrated rating scale based on the pathological and imaging findings (Supplementary Table S3). A patient was considered as a chemotherapy responder when the sum of scores for all indicators was greater than the upper quartile. As a result, among 21 patients who received pre-operative chemotherapy, 9 patients were classified as chemotherapy responders and 12 patients were non-responders. Tumor necrosis rate (p < 0.0001), change of margin (p = 0.0158) and circumference (p = 0.0071) were significantly related to the results of integrated evaluation (Table 1). Two patients without tumor necrosis rate information were evaluated as one responder and one non-responder according to the results of other indicators.

Table 1

Comparison of single indicators and integrated evaluation of chemotherapy response
	Responder (N = 9)	Non-responder (N = 12)	P value
Tumor necrosis rate-No.(%)			< 0.0001
>90%	8 (88.9%)	0 (0%)
<90%	0 (0%)	11 (91.7%)
NA	1 (11.1%)	1 (8.3%)
Ossification-No.(%)			0.4878
Increase	5 (55.6%)	8 (66.7%)
No change	4 (44.4%)	3 (25.0%)
Decrease	0 (0%)	1 (8.3%)
Margin-No.(%)			0.0158
Clear	7 (77.8%)	2 (16.7%)
No change	2 (22.2%)	7 (58.3%)
Unclear	0 (0%)	3 (25.0%)
Blood supply-No.(%)			0.2194
Decrease	6 (66.7%)	4 (33.3%)
No change	3 (33.3%)	6 (50.0%)
Increase	0 (0%)	2 (16.7%)
Circumference-No.(%)			0.0071
Decrease	7 (77.8%)	2 (16.7%)
No change	1 (11.1%)	2 (16.7%)
Increase	0 (0%)	7 (58.3%)
NA	1 (11.1%)	1 (8.3%)
Abbreviations: NA, not available.

Genomic landscape of osteosarcoma

A total of 1570 somatic mutations were detected in 25 patients with osteosarcoma, including 1488 single nucleotide variants and small insertions and deletions, 72 CNVs and 7 structural variants. The median number of mutations was 49 (range 2–215). MYC (8q24.21), NCOR1 (7p12-11.2), PHOX2B (4p13) and TP53 (17p13.1) were the top 4 commonly mutated genes, and mutated in 32%, 28%, 28% and 24% of osteosarcomas, respectively (Fig. 1A). Osteosarcoma had the extremely low level of TMB (median 0.47 muts/Mb, range 0.04–1.62 muts/Mb) and relatively high level of chromosomal CNV burden (median: 5, range 0–15). A total of 25 significantly amplified and 5 deleted regions were identified with the q value < 0.1 (Fig. 1B). COSMIC mutational signature analysis showed that osteosarcoma was mainly associated with signature 3 (55%, failure of DNA double-strand break-repair by homologous recombination), 22 (14%, aristolochic acid exposure) and 20 (13%, defective DNA mismatch repair) (Fig. 1C). To examine the relationship between copy number and expression level of CNVs, Spearman correlation analysis was performed and the result indicated that copy number was positively correlated with Z-score, with the Spearman coefficient of 0.5568 (p < 0.0001) (Fig. 1D).

Relationship between genomic features and chemotherapy response

To explore the relationship between genomic characteristics and chemotherapy response, fisher’s exact test was used to find the differentially mutated genes between responders and non-responders with significance, and the result was negative, which may be explained by the low mutation rate in osteosarcoma (data not shown). Additionally, no difference in TMB between two groups was observed (median TMB for responders vs. non-responders: 0.5111 vs. 0.3778 muts/Mb, p = 0.3013) (Fig. 2A). Notably, the chromosomal CNV burden of chemotherapy responders appeared to be higher than that of non-responders with the median of 7 and 3.5, respectively, but the difference was not statistically significant (p = 0.0775) (Fig. 2B). The composition of COSMIC mutational signatures was also compared between chemotherapy responders and non-responders (Fig. 2C). Mutational signatures 1, 3, 6, 17, 20, 22 and 28 were present in both groups. Apparently, the percentage of signature 3, which was associated with failure of DNA double-strand break-repair by homologous recombination, was higher in chemotherapy responders (56%) than in non-responders (45%), which suggested that signature 3 may be related to chemotherapy response. By contrast, the percentage of two signatures of unknown etiology (signature 17 and 28) was higher in non-responders (Non-responders vs. Responders: 12% vs. 4% for signature 17; 6% vs. 1% for signature 28).

Differentially expressed genes (DEGs) and pathway enrichment analysis

The raw expression counts for all genes were executed for the differentially expressed gene analysis between chemotherapy responders and non-responders. The thresholds for adjusted p (padj) and |log2FC| were set as < 0.05 and ≥ 1, respectively. As a result, the volcano plot showed that 11 genes were significantly up-regulated in responders and 18 genes were significantly up-regulated in non-responders with the adjusted p value smaller than 0.05 (Fig. 3A). GSEA enrichment analysis was performed to identify the up-regulated pathways in chemotherapy responders and non-responders. Four pathways related to cardiomyopathy were up-regulated in responders (Fig. 3B), while neuroactive ligand − receptor interaction and olfactory transduction pathways were up-regulated in non-responders (Fig. 3C).

Due to the relative small number of DEGs, genes with p < 0.05 and |log2FC| ≥ 1 were included to perform conventional KEGG pathway enrichment analysis and PPI analysis. The significant KEGG pathways (adjusted p < 0.05) of 309 significantly expressed genes in responders included calcium signaling pathway, motor proteins, cGMP − PKG signaling pathway, circadian entrainment, arrhythmogenic right ventricular cardiomyopathy, as well as 4 pathways identified by GSEA enrichment analysis (Fig. 4A). In non-responders, neuroactive ligand-receptor interaction pathway was the only significant KEGG pathway with the adjusted p value smaller than 0.05 (Fig. 4B). Genes involved in these pathways were summarized in Supplementary Table S4. The PPI analysis was performed using STRING database and visualized by Cytoscape (Supplementary Fig. 1). A total of 709 nodes and 1888 edges were included in the network. ACTN2, TTN and MYH6 were considered as potential hub genes with the node degree score of 45, 43 and 39, respectively. ACTN2, TTN and MYH6 were overexpressed in chemotherapy responders with the |log2FC| of 1.8487, 3.6915, 2.1881 and p value of 0.0103, 0.0001, 0.0200, respectively (data not shown). In addition, all of them were involved in pathways enriched in chemotherapy responders (Supplementary Table S4).

Development of random forest classifier to predict chemotherapy response

To discriminate chemotherapy response, we developed a random forest-based classifier using 23 DEGs. The importance of genes in the classifier was illustrated in Supplementary Fig. 2A, which showed that UST, CTNNA2 and DAPL1 was the top 3 important genes. Among 21 patients with prior chemotherapy treatment and RNA sequencing, the classifier classified 12 non-responders into 11 non-responders and 1 responder, and classified 9 responders into 6 responders and 3 non-responder, with the out-of-bag estimate of error rate of 19.05%. The classifier had the area under the curve of 0.843 (95% confidence interval, 0.654-1.000) to classify chemotherapy response (Supplementary Fig. 2B).

Osteosarcoma is a heterogeneous disease with regard to its histology, chemotherapy response and clinical outcomes^12,13. In this study, whole exome and transcriptome sequencing were conducted in 25 patients with osteosarcoma. A clinicopathologic integrated rating scale was developed to evaluate the chemotherapy response. In addition, clinical, genomic and transcriptomic features relevant to chemotherapy response were explored and a chemotherapy response classifier was constructed. To our knowledge, this is the first study to comprehensively investigate the clinical and molecular characteristics of chemotherapy response in osteosarcoma using multi-omics techniques, which expands our knowledge of this complex disease.

In multiple previous studies^14–16, osteosarcoma was mainly characterized by diverse variants, recurrent structural variants, high frequencies of TP53 and RB1 mutations. Similarly, our study revealed that the genetic variants in osteosarcoma were dispersedly distributed. In terms of mutation type, single nucleotide variants and small insertions and deletions were atypical events, whereas CNVs, especially amplification, were common in osteosarcoma. PHOX2B mutations were firstly identified in osteosarcoma in our study, which encodes a transcription factor participating in the development of the peripheral nervous system and is known to be related to neuroblastoma and congenital central hypoventilation syndrome¹⁷. Our study also found that osteosarcoma had the extremely low level of TMB and relatively high level of chromosomal CNV burden. In addition, it seemed that patients with high chromosomal CNV burden had a nonsignificantly better response to chemotherapy, which required to be verified in further studies. Consistent with the previous study¹⁵, mutations in osteosarcoma were mainly caused by the dysfunctional DNA damage repair in this study. Moreover, this study suggested that chemotherapy responders had more homologous recombination repair deficiency - induced mutations, which may be explained by the administration of platinum-based chemotherapy regimens in osteosarcoma and the favorable response to platinum in defective homologous recombination repair ovarian cancer¹⁸.

In this study, 11 genes were significantly up-regulated in responders and 18 genes were significantly up-regulated in non-responders. Both KEGG and GSEA enrichment analyses suggested that 4 pathways relative to cardiomyopathy were significantly enriched in responders and neuroactive ligand-receptor interaction pathway was significant in non-responders. The recommended neoadjuvant chemotherapy regimen for osteosarcoma included high-dose methotrexate, doxorubicin, cisplatin and ifoshamide. Doxorubicin can induce clinically symptomatic cardiac toxicity in 1.5%-1.7% of patients with osteosarcoma and the incidence of subclinical cardiac abnormalities may be higher¹⁹. The enrichment of cardiomyopathy-related pathways in chemotherapy responders may be related to the treatment of doxorubicin, which reminds clinicians of the occurrence of cardiac toxicity in chemotherapy responders. In a previous study, transcriptomic analysis was performed in breast cancer to investigate biomarkers related to chemotherapy sensitivity. KEGG analysis indicated that a great number of DEGs in neuroactive ligand-receptor interaction and the number of DEGs in neuroactive ligand-receptor interaction pathway subexpressed in chemotherapy resistant group was more than overexpressed genes. However, whether the entire pathway was up-regulated or down-regulated was not analyzed²⁰. In another study in breast cancer, the relationship between neuroactive ligand-receptor interaction pathway and pathological complete response to chemotherapy was not clearly elucidated²¹. One study used the paclitaxel-induced peripheral neuropathy rat model to understand the transcriptomic level of the dorsal root ganglia neurons and found that neuroactive ligand-receptor interaction was majorly involved in sensory neurons of rats with paclitaxel-induced peripheral neuropathy²². Therefore, the enrichment of neuroactive ligand-receptor interaction pathway in chemotherapy non-responder group in our study might be related to chemotherapy-induced peripheral neuropathy. PPI analysis identified ACTN2, TTN and MYH6 as hub genes resulting in chemotherapy responsiveness. ACTN2 encoded protein links the anti-parallel actin filaments, contributes to sarcomere stability and is related to cardiomyopathy²³. TTN encoded protein plays an important role in skeletal and in heart muscles²⁴. MYH6 gene provides instructions for making a protein known as the cardiac alpha (α)-myosin heavy chain, which forms part of type II myosin in cardiac muscles²⁵. All of them were overexpressed in chemotherapy responders and involved in pathways enriched in chemotherapy responders, which confirmed the pathway enrichment findings.

In clinics, multiple methods are used to evaluate the neoadjuvant chemotherapy response in osteosarcoma, including symptoms and signs, laboratory tests, imaging examinations and evaluation of tumor necrosis rate. Our study suggested that changes of tumor margin and circumference after chemotherapy were more related to chemotherapy response. Zeng et.al. developed a chemoresistant risk model using markers ontained from bulk RNA and single-cell RNA sequencing with the area under the curve of 0.82 in the TARGET-OS training cohort and 0.84 in the GSE33382 validation cohort²⁶. Our study used the limited internal data to develop a random forest-based classifier, and performance of the model was comparative to the previous study. Due to the inaccessible clinical information in TARGET-OS dataset and limited overlapping genes between GSE33382 dataset and our internal dataset, no suitable external dataset could be used to validate our classifier. A potential limitation of this study was the small amount number of participants, which may reduce the representativeness of certain findings and need to be further verified in large-scale studies.

In conclusion, multi-omics sequencing techniques help us better understand the molecular characteristics of osteosarcoma. Osteosarcoma is a highly heterogeneous disease with frequent occurrence of CNVs at the genomic level. Transcriptomic analysis identified several signaling pathways associated with chemotherapy responsiveness to osteosarcoma, including pathways related to cardiomyopathy and neuroactive ligand-receptor interaction. Additionally, the performance of chemotherapy response classifier was acceptable, which requires to be validated using more data before utilizing in clinics.

Patients and samples. A total of 25 patients with the definite diagnosis of osteosarcoma in Beijing Jishuitan Hospital (Beijing, China) from December 2021 to November 2022 were prospectively enrolled in this study. Clinical information including demographics, pathologic diagnoses, treatment history and imaging examinations were collected. Tumor tissue samples (including 6 fresh tissues, 18 formalin immersed tissues and 1 formalin fixed paraffin-embedded tissue) and 2 ml matched peripheral blood were collected from all participants to perform the whole exome and transcriptome sequencing. All procedures were conducted in accordance with the Declaration of Helsinki. This study was approved by the Ethics Committee of Beijing Jishuitan Hospital (Approval No. K2023013-00) and written informed consent was obtained from all participants.

Pathologic and imaging evaluation of chemotherapy response. The tumor cell necrosis rate was analyzed using the tumor tissues collected from tumor resection and separated into two groups: greater than 90% and less than 90%. The circumferences of the tumor site on the affected limb and the same site on the healthy limb were measured, and their difference was calculated before and after chemotherapy. Enhanced CT examinations on the tumor site were performed before and after chemotherapy to evaluate the degree of ossification, tumor margin and blood supply of the tumor.

DNA- and RNA-based next-generation sequencing. Whole exome and transcriptome sequencing were performed in Geneplus-Beijing (Beijing, China) as previously described^27–29. Briefly, genomic DNA and RNA were extracted from the tumor samples using FirePureTM FFPE gDNA Extraction Kit for genomic DNA and RNeasy FFPE Kit for RNA or AllPrep DNA/RNA FFPE Kit for both DNA and RNA (QIAGEN, Hilden, Germany). Genomic DNA from leukocytes was extracted using the CWE9600 Blood DNA Kit (Cwbiotech, Taizhou, China). Sequencing libraries of genomic DNA and mRNA were prepared using KAPA DNA Library Preparation Kit (Kapa Biosystems, Wilmington, MA, USA) and NEB Next Ultra™ RNA Library Prep Kit (Illumina, Inc., San Diego, CA, USA), respectively. The DNA and RNA sequencing were performed using the DNBSEQ-T7RS High-throughput Sequencing platform (MGI, Shenzhen, China). All experimental procedures followed the manufacturer's instructions. Whole transcriptome sequencing cannot be performed for one patient due to the severe degradation of RNA. The detailed quality control data of DNA and RNA sequencing was provided in Supplementary Table S1.

Bioinformatics analysis. After removal of terminal adaptor sequences and low-quality reads, the clean sequencing reads were aligned to the reference human genome (hg19) using BWA (version 0.7.10) and HISAT (version 2.0.4) for DNA and RNA sequencing, respectively. Genomic single nucleotide variants, small insertions and deletions, copy number variants (CNVs) and structural variants were detected using MuTect (version 1.1.4) / NChot, GATK (version 3.4–46-gbc02625) and CONTRA (version 2.0.8), respectively. Transcript assembly was performed using StringTie (version 1.2.3).

Tumor mutation burden was evaluated as the number of non-synonymous variants with the mutant allele frequency greater than 5% per megabase in the coding region. Chromosomal CNV burden represented the total level of amplifications or deletions at the chromosome level, which was calculated as previously described³⁰. Gistic 2.0 was used to detect the significantly recurrent regions with amplification or deletion. The mutational landscape was portrayed using R package ‘maftools’ (version 2.14.0). To infer the composition of known Catalogue Of Somatic Mutations In Cancer (COSMIC) mutational signatures in osteosarcoma, R package ‘yapsa’ (version 1.24.0) was performed using the COSMIC mutational signatures version 2 (https://cancer.sanger.ac.uk/signatures/signatures_v2). Differentially expressed genes were analyzed using R package ‘DESeq2’ (version 1.38.3). R package ‘clusterProfiler’ (version 4.7.1.3) was used to perform gene set enrichment analysis (GSEA) enrichment analysis and conventional Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. An online tool (STRING, https://string-db.org/) was used to analyze protein-protein interaction (PPI). The default confidence score > 0.40 was used to screen the PPI pairs. Cytoscape (version 3.9.1) was applied to visualize the PPI network. R package ‘RandomForest’ (version 4.7.1.1) was used to construct the chemotherapy response classifier and ‘pROC’ (version 1.18.0) was used to plot the receiver operator characteristic curve and calculate the area under the curve.

Statistical analysis. Z-score normalization of gene expression data was performed using the scale function in R package ‘base’ (version 4.2.2) taking the raw count matrix as an input. Spearman correlation analysis was conducted to study the correlation between copy number and Z-score of CNVs. Differences of variables between groups were assessed using Mann-Whitney test for continuous variables and Fisher’s exact test or Chi-square test for categorical variables, with P < 0.05 considered as statistically significant.

Acknowledgements: Not applicable

Authors’ Contributions: Y.Y.: data curation, writing - original draft, writing – reviewing & editing; Z.H: data curation, writing – reviewing & editing; M.Y.: formal analysis, methodology, writing – review & editing; R.C.: supervision, writing - review & editing; T.J.: writing – review & editing; Y.S.: writing – review & editing; Z.D.: writing – review & editing; H.S.: writing – review & editing; X.N.: writing – review & editing, supervision; W.L.: writing – review & editing, conceptualization, supervision.

Conflict of Interest: M.Y. and R.C. are employees of Geneplus-Beijing (Beijing, China). The remaining authors report no conflict of interest.

Ethics Statement: All procedures were conducted in accordance with the Declaration of Helsinki. All the procedures involving human subjects were conducted following the ethical guidelines approved by the ethical committee of Beijing Jishuitan Hospital (Approval No. K2023013-00). All the participants have signed the informed consent for sequencing.

Funding: This study was funded by Beijing Natural Science Foundation (grant No. L212042), Beijing Municipal Hospital Scientific Research Cultivation Programme (grant No. PX2021015), Beijing Hospitals Authority Youth Programme (grant No. QML20200403), Beijing Jishuitan Hospital Elite Young Scholar Programme (grant No. XKGG202105), Beijing JST Research Funding (grant No. ZR-202104) and National Key R&D Program of China (grant No. 2021YFC2400500).

Data availability: Data underlying the results presented in this paper may be obtained from the corresponding author upon reasonable request.

Cole, S., Gianferante, D. M., Zhu, B. & Mirabello, L. Osteosarcoma: a surveillance, epidemiology, and end results program-based analysis from 1975 to 2017. Cancer. 128, 2107–2118, DOI: https://doi.org/10.1002/cncr.34163 (2022).
Ritter, J. & Bielack, S. S. Osteosarcoma. Ann. Oncol. 21 Suppl 7, vii320-vii325, DOI: https://doi.org/10.1093/annonc/mdq276 (2010).
Bacci, G. et al. Prognostic factors for osteosarcoma of the extremity treated with neoadjuvant chemotherapy: 15-year experience in 789 patients treated at a single institution. Cancer. 106, 1154–1161, DOI: https://doi.org/10.1002/cncr.21724 (2006).
Bishop, M. W. et al. Assessing the prognostic significance of histologic response in osteosarcoma: a comparison of outcomes on ccg-782 and int0133-a report from the children's oncology group bone tumor committee. Pediatr. Blood Cancer. 63, 1737–1743, DOI: https://doi.org/10.1002/pbc.26034 (2016).
Huvos, A. G., Rosen, G. & Marcove, R. C. Primary osteogenic sarcoma: pathologic aspects in 20 patients after treatment with chemotherapy en bloc resection, and prosthetic bone replacement. Arch. Pathol. Lab. Med. 101, 14–18 (1977).
Miller, M. & Hanna, N. Advances in systemic therapy for non-small cell lung cancer. Bmj. 375, n2363, DOI: https://doi.org/10.1136/bmj.n2363 (2021).
Biller, L. H. & Schrag, D. Diagnosis and treatment of metastatic colorectal cancer: a review. Jama. 325, 669–685, DOI: https://doi.org/10.1001/jama.2021.0106 (2021).
Chen, M. & Zhao, H. Next-generation sequencing in liquid biopsy: cancer screening and early detection. Hum. Genomics. 13, 34, DOI: https://doi.org/10.1186/s40246-019-0220-8 (2019).
Mirabello, L. et al. Frequency of pathogenic germline variants in cancer-susceptibility genes in patients with osteosarcoma. Jama Oncol. 6, 724–734, DOI: https://doi.org/10.1001/jamaoncol.2020.0197 (2020).
Zhang, J. T. et al. Longitudinal undetectable molecular residual disease defines potentially cured population in localized non-small cell lung cancer. Cancer Discov. 12, 1690–1701, DOI: https://doi.org/10.1158/2159-8290.CD-21-1486 (2022).
Yi, Z. et al. The molecular tumor burden index as a response evaluation criterion in breast cancer. Signal Transduct. Target. Ther. 6, 251, DOI: https://doi.org/10.1038/s41392-021-00662-9 (2021).
Whelan, J. S. & Davis, L. E. Osteosarcoma, chondrosarcoma, and chordoma. J. Clin. Oncol. 36, 188–193, DOI: https://doi.org/10.1200/JCO.2017.75.1743 (2018).
Lilienthal, I. & Herold, N. Targeting molecular mechanisms underlying treatment efficacy and resistance in osteosarcoma: a review of current and future strategies. Int. J. Mol. Sci. 21, DOI: https://doi.org/10.3390/ijms21186885 (2020).
Bousquet, M. et al. Whole-exome sequencing in osteosarcoma reveals important heterogeneity of genetic alterations. Ann. Oncol. 27, 738–744, DOI: https://doi.org/10.1093/annonc/mdw009 (2016).
Jiang, Y. et al. Multi-omics analysis identifies osteosarcoma subtypes with distinct prognosis indicating stratified treatment. Nat. Commun. 13, 7207, DOI: https://doi.org/10.1038/s41467-022-34689-5 (2022).
Chen, X. et al. Recurrent somatic structural variations contribute to tumorigenesis in pediatric osteosarcoma. Cell Rep. 7, 104–112, DOI: https://doi.org/10.1016/j.celrep.2014.03.003 (2014).
Raabe, E. H. et al. Prevalence and functional consequence of phox2b mutations in neuroblastoma. Oncogene. 27, 469–476, DOI: https://doi.org/10.1038/sj.onc.1210659 (2008).
Wen, H. et al. Homologous recombination deficiency in diverse cancer types and its correlation with platinum chemotherapy efficiency in ovarian cancer. Bmc Cancer. 22, 550, DOI: https://doi.org/10.1186/s12885-022-09602-4 (2022).
Longhi, A., Ferrari, S., Bacci, G. & Specchia, S. Long-term follow-up of patients with doxorubicin-induced cardiac toxicity after chemotherapy for osteosarcoma. Anticancer. Drugs. 18, 737–744, DOI: https://doi.org/10.1097/CAD.0b013e32803d36fe (2007).
Barron-Gallardo, C. A. et al. Transcriptomic analysis of breast cancer patients sensitive and resistant to chemotherapy: looking for overall survival and drug resistance biomarkers. Technol. Cancer Res. Treat. 21, 2091156443, DOI: https://doi.org/10.1177/15330338211068965 (2022).
Li, Y., Liu, X., Tang, H., Yang, H. & Meng, X. Rna sequencing uncovers molecular mechanisms underlying pathological complete response to chemotherapy in patients with operable breast cancer. Med. Sci. Monitor. 23, 4321–4327, DOI: https://doi.org/10.12659/msm.903272 (2017).
Sun, W. et al. Transcriptome analysis reveals dysregulation of inflammatory and neuronal function in dorsal root ganglion of paclitaxel-induced peripheral neuropathy rats. Mol. Pain. 19, 804307447, DOI: https://doi.org/10.1177/17448069221106167 (2023).
Ranta-Aho, J. et al. Mutation update for the actn2 gene. Hum. Mutat. 43, 1745–1756, DOI: https://doi.org/10.1002/humu.24470 (2022).
Chauveau, C., Rowell, J. & Ferreiro, A. A rising titan: ttn review and mutation update. Hum. Mutat. 35, 1046–1059, DOI: https://doi.org/10.1002/humu.22611 (2014).
Carniel, E. et al. Alpha-myosin heavy chain: a sarcomeric gene associated with dilated and hypertrophic phenotypes of cardiomyopathy. Circulation. 112, 54–59, DOI: https://doi.org/10.1161/CIRCULATIONAHA.104.507699 (2005).
Zeng, Z. et al. Development of a chemoresistant risk scoring model for prechemotherapy osteosarcoma using single-cell sequencing. Front. Oncol. 12, 893282, DOI: https://doi.org/10.3389/fonc.2022.893282 (2022).
Wu, H. et al. Genomic characterization reveals distinct mutation landscapes and therapeutic implications in neuroendocrine carcinomas of the gastrointestinal tract. Cancer Commun. 42, 1367–1386, DOI: https://doi.org/10.1002/cac2.12372 (2022).
Liu, Y. et al. Genomic and transcriptomic insights into the precision treatment of pulmonary enteric adenocarcinoma. Lung Cancer. 179, 107169, DOI: https://doi.org/10.1016/j.lungcan.2023.03.005 (2023).
Zhang, S. et al. Dysregulated immune and metabolic microenvironment is associated with the post-operative relapse in stage i non-small cell lung cancer. Cancers. 14, DOI: https://doi.org/10.3390/cancers14133061 (2022).
Davoli, T., Uno, H., Wooten, E. C. & Elledge, S. J. Tumor aneuploidy correlates with markers of immune evasion and with reduced response to immunotherapy. Science. 355, DOI: https://doi.org/10.1126/science.aaf8399 (2017).

Competing interest reported. Conflict of Interest: M.Y. and R.C. are employees of Geneplus-Beijing (Beijing, China). The remaining authors report no conflict of interest.

Supplementaryinformationfiles.pdf

Download PDF

Journal Publication

published 27 Nov, 2023

Read the published version in Scientific Reports →

Editorial decision: Major revision
13 Aug, 2023
Reviews received at journal
02 Aug, 2023
Reviewers agreed at journal
30 Jul, 2023
Reviewers invited by journal
30 Jul, 2023
Editor assigned by journal
30 Jul, 2023
Editor invited by journal
13 Jul, 2023
Submission checks completed at journal
13 Jul, 2023
First submitted to journal
04 Jul, 2023

You are reading this latest preprint version

Genomic and Transcriptomic Characterization of Pre-operative Chemotherapy Response in Patients with Osteosarcoma

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Clinicopathologic characteristics and evaluation of chemotherapy response

Genomic landscape of osteosarcoma

Relationship between genomic features and chemotherapy response

Differentially expressed genes (DEGs) and pathway enrichment analysis

Development of random forest classifier to predict chemotherapy response

Discussion

Materials and Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1