- Email: [email protected]
Genomic testing and precision medicine — What does this mean for gynecologic oncology?
Gynecologic Oncology 140 (2016) 3–5
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
State of the Science
Genomic testing and precision medicine — What does this mean for gynecologic oncology?☆
1. Introduction Anti-cancer therapy based on pre-selection of cancers and patients through genomically characterized and identified molecular biomarkers has seen dramatic improvements in the treatment for breast cancer, lung cancer, certain leukemias and other malignancies. For gynecologic cancers, precision medicine through genomic testing has advanced the identification of high risk individuals as well as the treatment of specific types of ovarian cancer, i.e. germline BRCA mutated (gBRCAm), but has yet to significantly affect the standard treatment of BRCA wild type (BRCAwt) ovarian cancer, cervical or endometrial cancer. Nonetheless, the genomic characterization of gynecologic malignancies has provided the foundation for improved understanding of pathogenesis of gynecologic cancers, treatment advances, and outcome prediction for women with gynecologic cancers and is certain to have profound impact in the future. 2. Genomic testing in gynecologic cancers Currently, genomic testing in cancer typically involves next generation sequencing (NGS), also called massively parallel sequencing, which allows for characterization of cancer tissue or germline samples at the genomic, transcriptomic and epigenetic levels and gives data on both DNA and RNA for mutations, copy number alterations (CNAs), and somatic rearrangements. The Cancer Genome Atlas (TCGA) projects have undertaken genomic analyses of both high grade serous ovarian cancer (HGSC) as well as endometrial cancer [1,2], and the genomic characteristics of cervical cancer have been independently investigated by several groups [3,4]. Table 1 summarizes the genomic landscape based on NGS of the different histologic subtypes of ovarian, endometrial and cervical cancers. 2.1. Mutations Within ovarian cancers, there is heterogeneity in genomic profiles, although certain alterations appear to be more common in certain histological subtypes. HGSC is characterized by overall low mutational burden, mutations in TP53, high CNAs, alterations in homologous
☆ Disclosures: Ursula Matulonis: consultant for Astrazeneca, Tesaro, and Genentech/ Roche. Joyce Liu: consultant for Astrazeneca and Genentech/Roche and conducted clinical research on trials sponsored by AstraZeneca, Genentech/Roche, Boston Biomedical, Atara Biotherapeutics, and Merrimack Pharmaceuticals. Panagiotis A. Konstantinopoulos: Advisory Board for Vertex Inc.
http://dx.doi.org/10.1016/j.ygyno.2015.12.001 0090-8258/© 2015 Elsevier Inc. All rights reserved.
recombination (HR) and aberrations in certain pathways such as PI3 Kinase, RAS, Notch, FOXM1, and RB1 signaling/cell cycle control [1]. TCGA importantly found that approximately 50% of HGCS have alterations in DNA repair and HR deficiency (HRD), as alterations in HR genes such as BRCA1 or BRCA2, both germline (gBRCAm) and tumor (tBRCAm) have now been linked to both improved prognosis as well as response to poly (ADP ribose) polymerase (PARP) inhibitors [5–8]. The discovery of BRCA1/2 mutation as a biomarker for response to PARP inhibitors represents one of the most important steps in personalized treatment for ovarian cancer [6–8]. Of note, other histologies have also demonstrated alterations in HRD; nonserous ovarian cancers were found to have similar rates of mutations in HR genes as compared to HGSC, and these HRD-driven cancers included clear cell, endometrioid, and carcinosarcoma histologies [9]; in fact, high grade endometrioid cancers are thought to share genomic characteristics with HGSC and are often classified together with HGSC. Clear cell cancers display mutations in ARID1A, PIK3CA, PTEN as well as genomic complexity [10]. Mucinous cancers as well as low grade serous cancers (LGSC) both display KRAS mutations, which may be related to their relative chemotherapy insensitivity [11,12]. Small cell cancers of the ovary, which are rare but of particular importance due their occurrence predominantly in younger women and high lethality, are associated with SMARCA4 somatic or germline mutations [13]. Sex cord stromal tumors also have unique mutations present; FOXL2 mutations have been described in granulosa cell tumors, and DICER1 mutations in Sertoli-Leydig cell tumors [14,15]. In endometrial cancer, TCGA described 4 separate classifications — POLE ultramutated, microsatellite instability (MSI) hypermutated, copy-number low, and copy-number high [2]. Uterine papillary serous carcinomas (UPSC) shared genomic characteristics with HGSC and triple negative breast cancer (TNBC) with presence of TP53 mutations, low mutational load, and high degree of CNAs [2]. The majority of endometrioid cancers demonstrated few CNAs or TP53 mutations, but instead had mutations in PTEN, CTNNB1, PIK3CA, ARID1A, ARID5B, and KRAS [2]. In cervical cancers, one study observed KRAS mutations in adenocarcinomas but not in squamous cell cancers [3]. Additionally, the presence of a PIK3CA mutation was associated with a poorer outcome [3]. Besides PIK3CA mutations, cervical squamous cell cancers were found in a second study to possess somatic mutations in MAPK1 (8%), inactivating mutations in HLA-B (9%), and mutations in EP300 (16%), FBXW7 (15%), NFE2L2 (4%), TP53 (5%) and ERBB2 (6%) [4]; only 1 squamous cell cancer out of 79 was found to have a KRAS mutation corroborating the findings of Wright et al. [3,4]. Of note, although vulvar cancer has not undergone extensive genomic analysis, EGFR mutations have been detected and may be associated with poorer prognosis; EGFR inhibitors have demonstrated some activity in single institution phase II studies [16].
4
State of the Science
Table 1 Genomic characteristics of gynecologic cancers and their histologic subtypes (references [1–4,9–15,19,20]). Cancer (by anatomic site)
Ovarian
Endometrial
Histologic subtype
Mutations
Copy number alterations
High grade serous or endometrioid
TP53, BRCA1, BRCA2, CDK12
Low grade serous Clear cell Mucinous Small cell Sex cord stromal
KRAS, BRAF PIK3CA, ARID1A, PTEN KRAS SMARCA4 FOXL2 (granulosa), DICER1 (Sertoli–Leydig)
High grade serous
TP53, PIK3CA, FBXW7, PPP2R1A
Endometrioid
PTEN, PIK3CA, PIK3R1, KRAS, CTNNB1, FGFR2, POLE, MMR genes
Clear cell
PIK3CA, ARID1A
Adenocarcinoma
PIK3CA, KRAS, ELF3, CBFB
Squamous cell carcinoma
PIK3CA, MAPK1, HLA-B, EP300, FBXW7, NFE2L2, TP53, ERBB2
Cervical
CCNE1 (amp), MYC (amp), MECOM (amp), EMSY (amp), PIK3CA (amp), KRAS (amp), PTEN (del), RB1 (del), NF1 (del) 9p (loss), CDKN2A/2B (del) MET (amp), ERBB2 (amp) ERBB2 (amp)
Promoter methylation BRCA1, RAD51C TMS1/ASC
CCNE1 (amp), MYC (amp), ERBB2 (amp)
ERBB2 (amp) MYC (amp), PIK3CA (amp), SOX2 (amp), ERBB2 (amp), MCL1 (amp) MYC (amp), GLI2 (amp), ERBB2 (amp), BIRC3 (amp), YAP1 (amp), TP63 (amp), LRP1B (del)
Abbreviations: amp = amplication del = deletion
2.2. Copy number alterations CNAs can be assessed via NGS, and both HGSC and UPSC are characterized by significant CNAs. Findings of interest include the association of cyclin E and c-myc CNAs with poorer prognosis, although these have not been established in prospective trials [17]. Of note, ERBB2 amplification has been reported in clear cell and mucinous ovarian cancers as well as in UPSC, leading to speculation that ErbB2-directed therapies would be of interest in these cancers [10]. At present, however, these findings have not yet been successfully translated to directed care in gynecologic cancers. In cervical cancer, CNAs, including focal amplifications at 17q12 (encompassing ERBB2) and 8q24 (encompassing MYC), were observed in both squamous cell and adenocarcinomas of the cervix [4], although the clinical implications of these CNAs are not yet known. Additionally, Ojesina et al. found that integration sites for Human Papilloma Virus (HPV) occurred closer to amplified regions than would be expected by chance alone, supporting the hypothesis that HPV integration may trigger genome amplification [4].
2.3. Epigenetic silencing/methylation Epigenetic anomalies and silencing present in ovarian as well as endometrial cancer include abnormal DNA methylation, atypical histone modifications, and dysregulated microRNA expression leading to altered gene-expression [1,2]. For example, TCGA identified BRCA1 promoter methylation as present in approximately 11.5% of HGSC cases, although this did not appear to confer a survival benefit like that seen with tBRCAm or gBRCAm [1]. At this time, methylation information on gynecologic cancers should be considered for research based discovery and should not be used in routine clinical care. Genomic interrogation of gynecologic cancers is a powerful tool for development of new therapeutic agents, for molecular subclassification of gynecologic cancers, and in some instances, for prognosis. However, at this time, due to the lack of incorporation of NGS into prospective studies that have tested agents that have benefit in any gynecologic cancer, NGS cannot be routinely recommended for gynecologic cancers with the exception of translational research and discovery work as part of a clinical study.
3. Identification of specific genomic alterations of interest in gynecologic cancers While incorporation of NGS into clinical decision-making for gynecologic cancers remains premature, specific genomic alterations have proven to be of significant interest, such as the use of PARP inhibitors in gBRCAm ovarian cancer and the phase III testing of MEK inhibitors for LGSC. Below we discuss some genomic alterations of current interest and exploration in gynecologic cancers. 3.1. Identification of HRD-related ovarian cancers PARP inhibitors have demonstrated the most striking activity in ovarian cancers that possess either tBRCAm or gBRCAm [6–8], and currently, BRCA1/2 testing is the only FDA-approved genomic mechanism to identify patients for whom PARP inhibitors might be prescribed. However, because up to 50% of HGSC have alterations in HR genes, with BRCA1/2 mutations comprising about 20% of HGSC (14% gBRCAm and 6% tBRCAm) [1], multiple assays are now being tested to identify ovarian cancers that have HRD characteristics but do not have a deleterious BRCA1/2 mutation. Myriad HRD testing, which is being employed by ongoing niraparib studies, provides an HRD score based on 3 components: loss of heterozygosity (LOH), telemeric allelic imbalance, and large scale state transitions. Clovis Oncology is separately developing an HRD test in its trials of rucaparib which is comprised of 2 elements: tumor BRCA1/2 status and high or low genomic LOH status established through NGS [18]. Cancers in this test are classified as one of 3 groups: BRCAm, BRCAwt and LOH high, and BRCAwt and LOH low, with the latter group having demonstrated the lowest responsiveness to rucaparib. One of the criticisms of these assays that use LOH is that they are insensitive to reversion of HRD, which may occur upon development of resistance to platinum after patients have been exposed to multiple previous regimens. When reversion of HRD to HR proficiency occurs, the cumulative genomic defects as the result of the original HRD do not reverse; therefore, these assays still interpret these HR-proficient tumors as HRD, thereby limiting their accuracy of detecting PARP inhibitor response. One way to overcome this problem is by development of dynamic, functional biomarkers of HRD, whereby the HR pathway is mechanistically evaluated directly on tumor specimens; however such assays preclude the use of formalin-fixed paraffin-embedded specimens and are technically complex, limiting their reproducibility.
State of the Science
Similarly, the development of a companion diagnostic for olaparib has focused on either gBRCAm in the United States or gBRCAm and tBRCAm in Europe, and olaparib's current registration studies have focused on both gBRCAm or tBRCAm for eligibility. It is important to point out that while deleterious mutations in the BRCA1/2 genes suggest a higher likelihood of response to a PARP inhibitor, it is not currently possible to prospectively predict the responsiveness of a given cancer to a PARP inhibitor even when a BRCA1/2 mutation is present. Many factors play a role in this and likely include other mutational events and perhaps where in the BRCA gene the mutation lies. Other mutated genes that may play a significant role in ovarian cancer development as well as sensitivity to platinum and PARP inhibitor agents include Fanconi Anemia (FA) genes (mainly FANCN (PALB2), FANCA, FANCI, FANCJ (BRIP1), FANCL, and FANCC), core HR RAD genes (such as RAD50, RAD51, RAD51B, RAD51C, and RAD54L), and DNA damage response genes involved in HR (such as ATM, ATR, CHEK1, and CHEK2) [19,20]. 3.2. TP53 TP53 is ubiquitously altered in HGSC through mutation; this could represent a universal target for HGSC but has to date been elusive [1]. However, currently several agents that target aberrant p53 (mainly cell cycle checkpoint inhibitors such as Wee1, Chk1 or ATR inhibitors) are in clinical trials [21]. 3.3. PI3K and RAS/RAF/MEK pathways Abnormalities in the PI3K and RAS/RAF/MEK pathways have been identified in ovarian cancer and endometrial cancer, but have not yielded consistent prognostic nor therapeutic information [1,2]. Mutations in PIK3CA were seen in both adenocarcinomas and squamous cell cancers of the cervix with the observation that PIK3CA mutations were associated with shorter survival [3]. Given the low response rate to single agent PI3K pathway inhibitors, combined PI3K pathway and MEK inhibitors have been studied in both endometrial and cervical cancer, but this combination's potential has been limited by toxicities. 3.4. LGSC and MEK inhibitors Though BRAF mutations are less common in ovarian cancers, MEK inhibitors appear to have significant activity in LGSC [22] and are currently being tested in phase III clinical trials. Interestingly, in a Phase II trial of the MEK inhibitor selumetinib in LGSC, the presence of KRAS or BRAF mutation did not appear to correlate with drug activity [22]; thus, the exact mechanism of action of MEK inhibitor efficacy in LGSC remains unclear. 3.5. Predictive biomarkers for immune-oncology (IO) agents Whether NGS techniques can identify certain cancers amenable to IO agents and strategies is still unknown. However, the presence of POLE mutations or MSI in endometrial cancer may identify a group of cancers that should undergo further study with these agents [2]. Whether presence of BRCA1/2 mutation or HRD may render an ovarian cancer more sensitive to IO agents is also currently being studied. 4. Conclusions Although our ability to apply genomic findings to clinical practice is currently still limited, it is a very exciting time in the development of personalized medicine for gynecologic malignancies. As further genomic characterization continues in the gynecologic cancers, more successes in the field of personalized medicine are expected. Successes such as the approval of olaparib in gBRCAm ovarian cancer and the observation of PARP inhibitor activity in BRCAwt ovarian cancers that have other HRD biomarkers demonstrate the potential of precision medicine. It
5
will be critical that future studies of targeted therapies incorporate tissue collection and genomic characterization to capitalize on this potential.
References [1] The Cancer Genome Atlas Research Network, Integrated genomic analysis of ovarian cancer, Nature 474 (2011) 609–615. [2] Cancer Genome Atlas Research Network, C. Kandoth, N. Schultz, A.D. Cherniack, R. Akbani, Y. Liu, H. Shen, et al., Integrated genomic characterization of endometrial carcinoma, Nature 497 (2013) 67–73. [3] A.A. Wright, B.E. Howitt, A.P. Myers, S.E. Dahlberg, E. Palescandolo, P. Van Hummelen, et al., Oncogenic mutations in cervical cancer: genomic differences between adenocarcinomas and squamous cell carcinomas of the cervix, Cancer 119 (2013) 3776–3783. [4] A.I. Ojesina, L. Lichtenstein, S.S. Freeman, C.S. Pedamallu, I. Imaz-Rosshandler, T.J. Pugh, et al., Landscape of genomic alterations in cervical carcinomas, Nature 506 (2014) 371–375. [5] K.L. Bolton, G. Chenevix-Trench, C. Goh, S. Sadetzki, S.J. Ramus, B.Y. Karlan, et al., Association between BRCA1 and BRCA2 mutations and survival in women with invasive epithelial ovarian cancer, JAMA 307 (2012) 382–390. [6] J.F. Liu, P.A. Konstantinopoulos, U.A. Matulonis, PARP inhibitors in ovarian cancer: current status and future promise, Gynecol. Oncol. 133 (2014) 362–369. [7] C.L. Scott, E.M. Swisher, S.H. Kaufmann, Poly (ADP-ribose) polymerase inhibitors: recent advances and future development, J. Clin. Oncol. 33 (2015) 1397–1406. [8] C.C. Gunderson, K.N. Moore, PARP inhibition in ovarian cancer: state of the science, Gynecol. Oncol. 136 (2015) 8–10. [9] K.P. Pennington, T. Walsh, M.I. Harrell, M.K. Lee, M.H. Rendi, A. Thornton, et al., Germline and somatic mutations in homologous recombination genes predict platinum response and survival in ovarian, fallopian tube, and peritoneal carcinomas, Clin. Cancer Res. 20 (2014) 764–775. [10] D.S.P. Tan, R.E. Miller, S.B. Kaye, New perspectives on molecular targeted therapy in ovarian clear cell carcinoma, Br. J. Cancer 108 (2013) 1553–1559. [11] G.L. Ryland, S.M. Hunter, M.A. Doyle, F. Caramia, J. Li, S.M. Rowley, M. Christie, et al., Mutational landscape of mucinous ovarian carcinoma and its neoplastic precursors, Genome Med 7 (2015) 87 in press. [12] S.M. Hunter, M.S. Anglesio, G.L. Ryland, R. Sharma, Y.E. Chiew, S.M. Rowley, et al., Molecular profiling of low grade serous ovarian tumours identifies novel candidate driver genes, Oncotarget 6 (2015) in press. [13] P. Jelinic, J.J. Mueller, N. Olvera, F. Dao, S.N. Scott, R. Shah, et al., Recurrent SMARCA4 mutations in small cell carcinoma of the ovary, Nat. Genet. 46 (2014) 424–426. [14] S.P. Shah, M. Kobel, J. Senz, R.D. Morin, A.C. Blaise, K.C. Wiegand, et al., Mutation of FOXL2 in granulosa-cell tumors of the ovary, N. Engl. J. Med. 360 (2009) 2719–2729. [15] A. Heravi-Moussavi, M.S. Angelesio, S.-W. Grace Cheng, J. Senz, W. Yang, L. Prentice, et al., Recurrent somatic DICER1 mutations in nonepithelial ovarian cancers, N. Engl. J. Med. 366 (2012) 234–242. [16] N.S. Horowitz, A.B. Olawaiye, D.R. Borger, W.B. Growdon, C.N. Krasner, U.A. Matulonis, et al., Phase II trial of erlotinib in women with squamous cell carcinoma of the vulva, Gynecol. Oncol. 127 (2012) 141–146. [17] A.M. Patch, E.L. Christie, D. Etemadmoghadam, D.W. Garsed, J. George, S. Fereday, et al., Whole-genome characterization of chemoresistant ovarian cancer, Nature 521 (2015) 489–494. [18] E.M. Swisher, I.A. McNeish, R.L. Coleman, J. Brenton, S.H. Kaufmann, A.R. Allen, et al., ARIEL 2/3: An integrated clinical trial program to assess activity of rucaparib in ovarian cancer and to identify tumor molecular characteristics predictive of response, J. Clin. Oncol. 32 (5s) (2014) (suppl; abstr TPS5619). [19] H. Song, E. Dicks, S.J. Ramus, J.P. Tyrer, M.P. Intermaggio, J. Hayward, et al., Contribution of germline mutations in the RAD51B, RAD51C, and RAD51D genes to ovarian cancer in the population, J. Clin. Oncol. 33 (2015) 2901–2907. [20] S.J. Ramus, H. Song, E. Dicks, J.P. Tyrer, A.N. Rosenthal, M.P. Intermaggio, et al., Germline mutations in the BRIP1, BARD1, PALB2, and NBN genes in women with ovarian cancer, J. Natl. Cancer Inst. 107 (2015) in press. [21] M.J. Duffy, N.C. Synnott, P.M. McGowan, J. Crown, D. O'Connor, Gallagher WM.p53 as a target for the treatment of cancer, Cancer Treat. Rev. 40 (2014) 1153–1160. [22] J. Farley, W.E. Brady, V. Vathipadiekal, H.A. Lankes, R. Coleman, M.A. Morgan, et al., Selumetanib in women with recurrent low-grade serous carcinoma of the ovary or peritoneum: an open-label, single-arm, phase 2 study, Lancet Oncol. 14 (2013) 134–140.
Joyce Liu Panagiotis A. Konstantinopoulos Ursula A. Matulonis⁎ Gynecologic Oncology Program, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA 02215, United States ⁎Corresponding author at: Dana-Farber Cancer Institute, 450 Brookline Ave., Boston, MA 02215, United States. E-mail address: [email protected] (U.A. Matulonis). 18 November 2015
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
State of the Science
Genomic testing and precision medicine — What does this mean for gynecologic oncology?☆
1. Introduction Anti-cancer therapy based on pre-selection of cancers and patients through genomically characterized and identified molecular biomarkers has seen dramatic improvements in the treatment for breast cancer, lung cancer, certain leukemias and other malignancies. For gynecologic cancers, precision medicine through genomic testing has advanced the identification of high risk individuals as well as the treatment of specific types of ovarian cancer, i.e. germline BRCA mutated (gBRCAm), but has yet to significantly affect the standard treatment of BRCA wild type (BRCAwt) ovarian cancer, cervical or endometrial cancer. Nonetheless, the genomic characterization of gynecologic malignancies has provided the foundation for improved understanding of pathogenesis of gynecologic cancers, treatment advances, and outcome prediction for women with gynecologic cancers and is certain to have profound impact in the future. 2. Genomic testing in gynecologic cancers Currently, genomic testing in cancer typically involves next generation sequencing (NGS), also called massively parallel sequencing, which allows for characterization of cancer tissue or germline samples at the genomic, transcriptomic and epigenetic levels and gives data on both DNA and RNA for mutations, copy number alterations (CNAs), and somatic rearrangements. The Cancer Genome Atlas (TCGA) projects have undertaken genomic analyses of both high grade serous ovarian cancer (HGSC) as well as endometrial cancer [1,2], and the genomic characteristics of cervical cancer have been independently investigated by several groups [3,4]. Table 1 summarizes the genomic landscape based on NGS of the different histologic subtypes of ovarian, endometrial and cervical cancers. 2.1. Mutations Within ovarian cancers, there is heterogeneity in genomic profiles, although certain alterations appear to be more common in certain histological subtypes. HGSC is characterized by overall low mutational burden, mutations in TP53, high CNAs, alterations in homologous
☆ Disclosures: Ursula Matulonis: consultant for Astrazeneca, Tesaro, and Genentech/ Roche. Joyce Liu: consultant for Astrazeneca and Genentech/Roche and conducted clinical research on trials sponsored by AstraZeneca, Genentech/Roche, Boston Biomedical, Atara Biotherapeutics, and Merrimack Pharmaceuticals. Panagiotis A. Konstantinopoulos: Advisory Board for Vertex Inc.
http://dx.doi.org/10.1016/j.ygyno.2015.12.001 0090-8258/© 2015 Elsevier Inc. All rights reserved.
recombination (HR) and aberrations in certain pathways such as PI3 Kinase, RAS, Notch, FOXM1, and RB1 signaling/cell cycle control [1]. TCGA importantly found that approximately 50% of HGCS have alterations in DNA repair and HR deficiency (HRD), as alterations in HR genes such as BRCA1 or BRCA2, both germline (gBRCAm) and tumor (tBRCAm) have now been linked to both improved prognosis as well as response to poly (ADP ribose) polymerase (PARP) inhibitors [5–8]. The discovery of BRCA1/2 mutation as a biomarker for response to PARP inhibitors represents one of the most important steps in personalized treatment for ovarian cancer [6–8]. Of note, other histologies have also demonstrated alterations in HRD; nonserous ovarian cancers were found to have similar rates of mutations in HR genes as compared to HGSC, and these HRD-driven cancers included clear cell, endometrioid, and carcinosarcoma histologies [9]; in fact, high grade endometrioid cancers are thought to share genomic characteristics with HGSC and are often classified together with HGSC. Clear cell cancers display mutations in ARID1A, PIK3CA, PTEN as well as genomic complexity [10]. Mucinous cancers as well as low grade serous cancers (LGSC) both display KRAS mutations, which may be related to their relative chemotherapy insensitivity [11,12]. Small cell cancers of the ovary, which are rare but of particular importance due their occurrence predominantly in younger women and high lethality, are associated with SMARCA4 somatic or germline mutations [13]. Sex cord stromal tumors also have unique mutations present; FOXL2 mutations have been described in granulosa cell tumors, and DICER1 mutations in Sertoli-Leydig cell tumors [14,15]. In endometrial cancer, TCGA described 4 separate classifications — POLE ultramutated, microsatellite instability (MSI) hypermutated, copy-number low, and copy-number high [2]. Uterine papillary serous carcinomas (UPSC) shared genomic characteristics with HGSC and triple negative breast cancer (TNBC) with presence of TP53 mutations, low mutational load, and high degree of CNAs [2]. The majority of endometrioid cancers demonstrated few CNAs or TP53 mutations, but instead had mutations in PTEN, CTNNB1, PIK3CA, ARID1A, ARID5B, and KRAS [2]. In cervical cancers, one study observed KRAS mutations in adenocarcinomas but not in squamous cell cancers [3]. Additionally, the presence of a PIK3CA mutation was associated with a poorer outcome [3]. Besides PIK3CA mutations, cervical squamous cell cancers were found in a second study to possess somatic mutations in MAPK1 (8%), inactivating mutations in HLA-B (9%), and mutations in EP300 (16%), FBXW7 (15%), NFE2L2 (4%), TP53 (5%) and ERBB2 (6%) [4]; only 1 squamous cell cancer out of 79 was found to have a KRAS mutation corroborating the findings of Wright et al. [3,4]. Of note, although vulvar cancer has not undergone extensive genomic analysis, EGFR mutations have been detected and may be associated with poorer prognosis; EGFR inhibitors have demonstrated some activity in single institution phase II studies [16].
4
State of the Science
Table 1 Genomic characteristics of gynecologic cancers and their histologic subtypes (references [1–4,9–15,19,20]). Cancer (by anatomic site)
Ovarian
Endometrial
Histologic subtype
Mutations
Copy number alterations
High grade serous or endometrioid
TP53, BRCA1, BRCA2, CDK12
Low grade serous Clear cell Mucinous Small cell Sex cord stromal
KRAS, BRAF PIK3CA, ARID1A, PTEN KRAS SMARCA4 FOXL2 (granulosa), DICER1 (Sertoli–Leydig)
High grade serous
TP53, PIK3CA, FBXW7, PPP2R1A
Endometrioid
PTEN, PIK3CA, PIK3R1, KRAS, CTNNB1, FGFR2, POLE, MMR genes
Clear cell
PIK3CA, ARID1A
Adenocarcinoma
PIK3CA, KRAS, ELF3, CBFB
Squamous cell carcinoma
PIK3CA, MAPK1, HLA-B, EP300, FBXW7, NFE2L2, TP53, ERBB2
Cervical
CCNE1 (amp), MYC (amp), MECOM (amp), EMSY (amp), PIK3CA (amp), KRAS (amp), PTEN (del), RB1 (del), NF1 (del) 9p (loss), CDKN2A/2B (del) MET (amp), ERBB2 (amp) ERBB2 (amp)
Promoter methylation BRCA1, RAD51C TMS1/ASC
CCNE1 (amp), MYC (amp), ERBB2 (amp)
ERBB2 (amp) MYC (amp), PIK3CA (amp), SOX2 (amp), ERBB2 (amp), MCL1 (amp) MYC (amp), GLI2 (amp), ERBB2 (amp), BIRC3 (amp), YAP1 (amp), TP63 (amp), LRP1B (del)
Abbreviations: amp = amplication del = deletion
2.2. Copy number alterations CNAs can be assessed via NGS, and both HGSC and UPSC are characterized by significant CNAs. Findings of interest include the association of cyclin E and c-myc CNAs with poorer prognosis, although these have not been established in prospective trials [17]. Of note, ERBB2 amplification has been reported in clear cell and mucinous ovarian cancers as well as in UPSC, leading to speculation that ErbB2-directed therapies would be of interest in these cancers [10]. At present, however, these findings have not yet been successfully translated to directed care in gynecologic cancers. In cervical cancer, CNAs, including focal amplifications at 17q12 (encompassing ERBB2) and 8q24 (encompassing MYC), were observed in both squamous cell and adenocarcinomas of the cervix [4], although the clinical implications of these CNAs are not yet known. Additionally, Ojesina et al. found that integration sites for Human Papilloma Virus (HPV) occurred closer to amplified regions than would be expected by chance alone, supporting the hypothesis that HPV integration may trigger genome amplification [4].
2.3. Epigenetic silencing/methylation Epigenetic anomalies and silencing present in ovarian as well as endometrial cancer include abnormal DNA methylation, atypical histone modifications, and dysregulated microRNA expression leading to altered gene-expression [1,2]. For example, TCGA identified BRCA1 promoter methylation as present in approximately 11.5% of HGSC cases, although this did not appear to confer a survival benefit like that seen with tBRCAm or gBRCAm [1]. At this time, methylation information on gynecologic cancers should be considered for research based discovery and should not be used in routine clinical care. Genomic interrogation of gynecologic cancers is a powerful tool for development of new therapeutic agents, for molecular subclassification of gynecologic cancers, and in some instances, for prognosis. However, at this time, due to the lack of incorporation of NGS into prospective studies that have tested agents that have benefit in any gynecologic cancer, NGS cannot be routinely recommended for gynecologic cancers with the exception of translational research and discovery work as part of a clinical study.
3. Identification of specific genomic alterations of interest in gynecologic cancers While incorporation of NGS into clinical decision-making for gynecologic cancers remains premature, specific genomic alterations have proven to be of significant interest, such as the use of PARP inhibitors in gBRCAm ovarian cancer and the phase III testing of MEK inhibitors for LGSC. Below we discuss some genomic alterations of current interest and exploration in gynecologic cancers. 3.1. Identification of HRD-related ovarian cancers PARP inhibitors have demonstrated the most striking activity in ovarian cancers that possess either tBRCAm or gBRCAm [6–8], and currently, BRCA1/2 testing is the only FDA-approved genomic mechanism to identify patients for whom PARP inhibitors might be prescribed. However, because up to 50% of HGSC have alterations in HR genes, with BRCA1/2 mutations comprising about 20% of HGSC (14% gBRCAm and 6% tBRCAm) [1], multiple assays are now being tested to identify ovarian cancers that have HRD characteristics but do not have a deleterious BRCA1/2 mutation. Myriad HRD testing, which is being employed by ongoing niraparib studies, provides an HRD score based on 3 components: loss of heterozygosity (LOH), telemeric allelic imbalance, and large scale state transitions. Clovis Oncology is separately developing an HRD test in its trials of rucaparib which is comprised of 2 elements: tumor BRCA1/2 status and high or low genomic LOH status established through NGS [18]. Cancers in this test are classified as one of 3 groups: BRCAm, BRCAwt and LOH high, and BRCAwt and LOH low, with the latter group having demonstrated the lowest responsiveness to rucaparib. One of the criticisms of these assays that use LOH is that they are insensitive to reversion of HRD, which may occur upon development of resistance to platinum after patients have been exposed to multiple previous regimens. When reversion of HRD to HR proficiency occurs, the cumulative genomic defects as the result of the original HRD do not reverse; therefore, these assays still interpret these HR-proficient tumors as HRD, thereby limiting their accuracy of detecting PARP inhibitor response. One way to overcome this problem is by development of dynamic, functional biomarkers of HRD, whereby the HR pathway is mechanistically evaluated directly on tumor specimens; however such assays preclude the use of formalin-fixed paraffin-embedded specimens and are technically complex, limiting their reproducibility.
State of the Science
Similarly, the development of a companion diagnostic for olaparib has focused on either gBRCAm in the United States or gBRCAm and tBRCAm in Europe, and olaparib's current registration studies have focused on both gBRCAm or tBRCAm for eligibility. It is important to point out that while deleterious mutations in the BRCA1/2 genes suggest a higher likelihood of response to a PARP inhibitor, it is not currently possible to prospectively predict the responsiveness of a given cancer to a PARP inhibitor even when a BRCA1/2 mutation is present. Many factors play a role in this and likely include other mutational events and perhaps where in the BRCA gene the mutation lies. Other mutated genes that may play a significant role in ovarian cancer development as well as sensitivity to platinum and PARP inhibitor agents include Fanconi Anemia (FA) genes (mainly FANCN (PALB2), FANCA, FANCI, FANCJ (BRIP1), FANCL, and FANCC), core HR RAD genes (such as RAD50, RAD51, RAD51B, RAD51C, and RAD54L), and DNA damage response genes involved in HR (such as ATM, ATR, CHEK1, and CHEK2) [19,20]. 3.2. TP53 TP53 is ubiquitously altered in HGSC through mutation; this could represent a universal target for HGSC but has to date been elusive [1]. However, currently several agents that target aberrant p53 (mainly cell cycle checkpoint inhibitors such as Wee1, Chk1 or ATR inhibitors) are in clinical trials [21]. 3.3. PI3K and RAS/RAF/MEK pathways Abnormalities in the PI3K and RAS/RAF/MEK pathways have been identified in ovarian cancer and endometrial cancer, but have not yielded consistent prognostic nor therapeutic information [1,2]. Mutations in PIK3CA were seen in both adenocarcinomas and squamous cell cancers of the cervix with the observation that PIK3CA mutations were associated with shorter survival [3]. Given the low response rate to single agent PI3K pathway inhibitors, combined PI3K pathway and MEK inhibitors have been studied in both endometrial and cervical cancer, but this combination's potential has been limited by toxicities. 3.4. LGSC and MEK inhibitors Though BRAF mutations are less common in ovarian cancers, MEK inhibitors appear to have significant activity in LGSC [22] and are currently being tested in phase III clinical trials. Interestingly, in a Phase II trial of the MEK inhibitor selumetinib in LGSC, the presence of KRAS or BRAF mutation did not appear to correlate with drug activity [22]; thus, the exact mechanism of action of MEK inhibitor efficacy in LGSC remains unclear. 3.5. Predictive biomarkers for immune-oncology (IO) agents Whether NGS techniques can identify certain cancers amenable to IO agents and strategies is still unknown. However, the presence of POLE mutations or MSI in endometrial cancer may identify a group of cancers that should undergo further study with these agents [2]. Whether presence of BRCA1/2 mutation or HRD may render an ovarian cancer more sensitive to IO agents is also currently being studied. 4. Conclusions Although our ability to apply genomic findings to clinical practice is currently still limited, it is a very exciting time in the development of personalized medicine for gynecologic malignancies. As further genomic characterization continues in the gynecologic cancers, more successes in the field of personalized medicine are expected. Successes such as the approval of olaparib in gBRCAm ovarian cancer and the observation of PARP inhibitor activity in BRCAwt ovarian cancers that have other HRD biomarkers demonstrate the potential of precision medicine. It
5
will be critical that future studies of targeted therapies incorporate tissue collection and genomic characterization to capitalize on this potential.
References [1] The Cancer Genome Atlas Research Network, Integrated genomic analysis of ovarian cancer, Nature 474 (2011) 609–615. [2] Cancer Genome Atlas Research Network, C. Kandoth, N. Schultz, A.D. Cherniack, R. Akbani, Y. Liu, H. Shen, et al., Integrated genomic characterization of endometrial carcinoma, Nature 497 (2013) 67–73. [3] A.A. Wright, B.E. Howitt, A.P. Myers, S.E. Dahlberg, E. Palescandolo, P. Van Hummelen, et al., Oncogenic mutations in cervical cancer: genomic differences between adenocarcinomas and squamous cell carcinomas of the cervix, Cancer 119 (2013) 3776–3783. [4] A.I. Ojesina, L. Lichtenstein, S.S. Freeman, C.S. Pedamallu, I. Imaz-Rosshandler, T.J. Pugh, et al., Landscape of genomic alterations in cervical carcinomas, Nature 506 (2014) 371–375. [5] K.L. Bolton, G. Chenevix-Trench, C. Goh, S. Sadetzki, S.J. Ramus, B.Y. Karlan, et al., Association between BRCA1 and BRCA2 mutations and survival in women with invasive epithelial ovarian cancer, JAMA 307 (2012) 382–390. [6] J.F. Liu, P.A. Konstantinopoulos, U.A. Matulonis, PARP inhibitors in ovarian cancer: current status and future promise, Gynecol. Oncol. 133 (2014) 362–369. [7] C.L. Scott, E.M. Swisher, S.H. Kaufmann, Poly (ADP-ribose) polymerase inhibitors: recent advances and future development, J. Clin. Oncol. 33 (2015) 1397–1406. [8] C.C. Gunderson, K.N. Moore, PARP inhibition in ovarian cancer: state of the science, Gynecol. Oncol. 136 (2015) 8–10. [9] K.P. Pennington, T. Walsh, M.I. Harrell, M.K. Lee, M.H. Rendi, A. Thornton, et al., Germline and somatic mutations in homologous recombination genes predict platinum response and survival in ovarian, fallopian tube, and peritoneal carcinomas, Clin. Cancer Res. 20 (2014) 764–775. [10] D.S.P. Tan, R.E. Miller, S.B. Kaye, New perspectives on molecular targeted therapy in ovarian clear cell carcinoma, Br. J. Cancer 108 (2013) 1553–1559. [11] G.L. Ryland, S.M. Hunter, M.A. Doyle, F. Caramia, J. Li, S.M. Rowley, M. Christie, et al., Mutational landscape of mucinous ovarian carcinoma and its neoplastic precursors, Genome Med 7 (2015) 87 in press. [12] S.M. Hunter, M.S. Anglesio, G.L. Ryland, R. Sharma, Y.E. Chiew, S.M. Rowley, et al., Molecular profiling of low grade serous ovarian tumours identifies novel candidate driver genes, Oncotarget 6 (2015) in press. [13] P. Jelinic, J.J. Mueller, N. Olvera, F. Dao, S.N. Scott, R. Shah, et al., Recurrent SMARCA4 mutations in small cell carcinoma of the ovary, Nat. Genet. 46 (2014) 424–426. [14] S.P. Shah, M. Kobel, J. Senz, R.D. Morin, A.C. Blaise, K.C. Wiegand, et al., Mutation of FOXL2 in granulosa-cell tumors of the ovary, N. Engl. J. Med. 360 (2009) 2719–2729. [15] A. Heravi-Moussavi, M.S. Angelesio, S.-W. Grace Cheng, J. Senz, W. Yang, L. Prentice, et al., Recurrent somatic DICER1 mutations in nonepithelial ovarian cancers, N. Engl. J. Med. 366 (2012) 234–242. [16] N.S. Horowitz, A.B. Olawaiye, D.R. Borger, W.B. Growdon, C.N. Krasner, U.A. Matulonis, et al., Phase II trial of erlotinib in women with squamous cell carcinoma of the vulva, Gynecol. Oncol. 127 (2012) 141–146. [17] A.M. Patch, E.L. Christie, D. Etemadmoghadam, D.W. Garsed, J. George, S. Fereday, et al., Whole-genome characterization of chemoresistant ovarian cancer, Nature 521 (2015) 489–494. [18] E.M. Swisher, I.A. McNeish, R.L. Coleman, J. Brenton, S.H. Kaufmann, A.R. Allen, et al., ARIEL 2/3: An integrated clinical trial program to assess activity of rucaparib in ovarian cancer and to identify tumor molecular characteristics predictive of response, J. Clin. Oncol. 32 (5s) (2014) (suppl; abstr TPS5619). [19] H. Song, E. Dicks, S.J. Ramus, J.P. Tyrer, M.P. Intermaggio, J. Hayward, et al., Contribution of germline mutations in the RAD51B, RAD51C, and RAD51D genes to ovarian cancer in the population, J. Clin. Oncol. 33 (2015) 2901–2907. [20] S.J. Ramus, H. Song, E. Dicks, J.P. Tyrer, A.N. Rosenthal, M.P. Intermaggio, et al., Germline mutations in the BRIP1, BARD1, PALB2, and NBN genes in women with ovarian cancer, J. Natl. Cancer Inst. 107 (2015) in press. [21] M.J. Duffy, N.C. Synnott, P.M. McGowan, J. Crown, D. O'Connor, Gallagher WM.p53 as a target for the treatment of cancer, Cancer Treat. Rev. 40 (2014) 1153–1160. [22] J. Farley, W.E. Brady, V. Vathipadiekal, H.A. Lankes, R. Coleman, M.A. Morgan, et al., Selumetanib in women with recurrent low-grade serous carcinoma of the ovary or peritoneum: an open-label, single-arm, phase 2 study, Lancet Oncol. 14 (2013) 134–140.
Joyce Liu Panagiotis A. Konstantinopoulos Ursula A. Matulonis⁎ Gynecologic Oncology Program, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA 02215, United States ⁎Corresponding author at: Dana-Farber Cancer Institute, 450 Brookline Ave., Boston, MA 02215, United States. E-mail address: [email protected] (U.A. Matulonis). 18 November 2015