Pulmonary Metastases Exhibit Epigenetic Clonality: Implications for Precision Cancer Therapy
Emily S. Reardon, MD, Julie A. Hong, MS, David M. Straughan, MD, Saïd C. Azoury, MD, Mary Zhang, MS, and David S. Schrump, MD, MBA
Thoracic Epigenetics Laboratory, Thoracic and GI Oncology Branch, Center for Cancer Research, National Cancer Institute, Bethesda, Maryland
Background. Development of effective cancer thera- pies may be limited by intratumoral heterogeneity, which facilitates outgrowth and organ-specific dis- semination of treatment resistant clones. At present, limited information is available regarding epigenetic landscapes of pulmonary metastases. This study was undertaken to characterize epigenetic signatures of pulmonary metastases and to identify potential thera- peutic targets.
Methods. RNA and DNA were extracted from 65 pulmonary metastases resected from 12 patients (5 with sarcoma, 7 with adrenocortical carcinoma). Quantitative reverse transcription polymerase chain reaction tech- niques were used to evaluate expression levels of cancer- testis (CT) genes (NY-ESO-1, MAGE-A3, MAGE-A9, MAGE-A12, GAGE1, CT-45, SSX-1, and SSX-2), tumor suppressor (TS) genes (p16 and RASSF1A), and genes encoding epigenetic modifiers (DNMT1, DNMT3A, DNMT3B, EZH2, EED, and SUZ12), aberrantly expressed in human malignant diseases. Pyrosequencing tech- niques were used to quantitate DNA methylation levels in LINE1, NBL2, and D4Z4 repetitive sequences and
promoter methylation status of differentially regulated genes. Results of these analyses were compared with a standardized panel of normal lung tissues.
Results. Pulmonary metastases exhibited histologically related and patient-specific global DNA demethylation. Significant interpatient heterogeneity of gene expression was observed even among patients with similar tumor histologic features. Epigenetic signatures appeared consistent among metastases from the same patient, irrespective of the time of resection (synchronous/meta- chronous) or the anatomic location. EZH2, EED, and SUZ12 (core components of Polycomb repressive complex-2 [PRC-2]) were upregulated in the majority of metastases.
Conclusions. Pulmonary metastases exhibit patient- specific epigenetic clonality, which may be exploited for precision therapies targeting aberrant CT or TS gene expression. PRC-2 may be a shared target for epigenetic therapy of pulmonary metastases.
(Ann Thorac Surg 2015; :- ) @ 2015 by The Society of Thoracic Surgeons
A dvances in next-generation sequencing and prote- omics have revealed that phenotypic heterogeneity coincides with distinct and complex gene expression as well as kinome signatures of cancer cells and stromal elements in primary malignant neoplasms of various histologic types [1, 2]. Within these neoplasms are sub- populations of cancer stem cells (also referred to as tumor initiating cells), which through genetic and epigenetic mechanisms exhibit increased tumor-initiating capacity, self-renewal potential, and resistance to chemotherapy as well as radiation therapy [3, 4]. At present, intratumoral heterogeneity, clonal selection, and dissemination of treatment-resistant clones pose significant challenges for the development of efficacious precision cancer therapies.
Approximately one third of all patients dying of extrathoracic malignant diseases, including sarcomas,
gastrointestinal, or genitourinary cancers, develop me- tastases to lungs, pleura, or mediastinum; approximately 20% of these patients, particularly those with osteosar- comas or soft tissue sarcomas, have pulmonary disease as the sole site of distant metastases [5]. As systemic therapies have become more refined, pulmonary meta- stasectomy has emerged as an integral component of multimodality therapy for cancers of diverse histologic types, with encouraging results in properly selected pa- tients [6-9]. Unfortunately, most patients undergoing pulmonary metastasectomy ultimately develop inoper- able, refractory disease, possibly secondary to the emer- gence of cancer stem cells exhibiting multidrug resistance [10-12].
Alterations in chromatin structure during malignant transformation recapitulate epigenomic states in normal stem cells [13]. Global DNA demethylation results in
Accepted for publication May 15, 2015.
Presented at the Fifty-first Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 24-28, 2015.
Address correspondence to Dr Schrump, Thoracic and GI Oncology Branch, CCR/NCI Bldg 10, Rm 4-3942, 10 Center Drive, MSC 1201, Bethesda, MD 20892-1201; e-mail: david.schrump@nih.gov.
The Appendix can be viewed in the online version of this article [http://dx.doi.org/10.1016/j.athoracsur.2015. 05.089] on http://www.annalsthoracicsurgery.org.
de-repression of endogenous retroviruses, loss of imprinting, and activation of cancer-testis (CT) genes, which normally exhibit stage-specific expression during germ cell development in testes or placenta. In the context of genome-wide DNA demethylation, tumor suppressor (TS) genes such as p16 and RASSF1A are silenced by site-specific DNA hypermethylation or Poly- comb repressive complexes (PRCs).
Limited information is available regarding genetic and epigenetic landscapes of pulmonary metastases. The following study was undertaken to characterize epige- netic signatures in pulmonary metastases in an attempt to develop novel precision therapies for these neoplasms.
Material and Methods
Patient Tissues and RNA/DNA Extraction
Sixty-five pulmonary metastases resected from 12 pa- tients were included in this study. All patients were treated at the National Cancer Institute (NCI) on institu- tional review board-approved clinical protocols, and they met eligibility criteria for standard of care pulmonary metastasectomy. Resected metastases were immediately placed on wet ice and were dissected away from adjacent normal lung tissues. Representative frozen sections were obtained to verify histologic features and tumor cell viability. Individual metastases were then frozen at -80℃ in an anonymized tissue bank in the Thoracic Epigenetics Laboratory, Center for Cancer Research, NCI with patient identification, time of resection and anatomic location tracked using Labmatrix (BioFortis, Columbia, MD). Three pieces from each metastasis were analyzed to perform all experiments in triplicate and to examine the extent of heterogeneity within these nodules. A total of 195 metastasectomy samples were analyzed. RNA and DNA were simultaneously extracted from these speci- mens by using the All Prep DNA/RNA Mini Kit (Qiagen, Valencia, CA). RNA and DNA from normal lung tissues from nonsmokers who were undergoing lung cancer re- sections at the NCI, human testis (RNA and DNA ob- tained from Life Technologies, Carlsbad, CA and from BioChain, Newark, CA), LP9 cell line (Coriell Bio- repository, Camden, NJ), and A549 cell line (ATCC, Manassas, VA) were used as controls.
Real-Time Reverse Transcription Polymerase Chain Reaction Analysis
Complementary DNA sequences were made using the Reverse Transcription Kit (Bio-Rad, Hercules, CA). Quantitative reverse transcription polymerase chain re- action (qRT-PCR) analyses of expression levels of genes encoding CT antigens (CTAs), tumor suppressors, DNA methyltransferases (DNMTs), and Polycomb group pro- teins were performed as described [14], by using primers listed in Supplementary Table 1.
Pyrosequencing Analysis
Bisulfite modification of DNA was performed using pro- tocols and reagents in the EpiTect Bisulfite Kit (Qiagen).
Pyrosequencing analyses of DNA repetitive elements and promoter regions of NY-ESO-1, MAGE-A3, RASSF1A, and p16 were performed as described [14, 15], by using primers listed in Supplementary Table 1.
Statistical Analysis
Standard error of the mean (SEM) is indicated by bars on each figure and was calculated using GraphPad Prism, version 6.0 (GraphPad Software, San Diego, CA). All ex- periments were done with at a minimum of triplicate samples, and p values were calculated using two-tailed t tests.
Results
Patients’ Demographics
A total of 65 metastases were resected from 12 patients, including 5 patients with sarcomas (2 chondrosarcomas and 3 synovial sarcomas) and 7 patients with adrenocor- tical carcinomas (ACCs) (Table 1). The median age of all patients was 46 years (range, 17 to 66 years). The median age of the patients with sarcomas was 35 years (range, 17 to 62 years), and the median age of the patients with ACCs was 53 years (range, 32 to 66 years). Ten patients developed pulmonary metastases, with an average disease-free interval of 19 months after definitive treat- ment of their primary tumors. Two patients had pulmo- nary metastases at the time of diagnosis. All patients had control of extrathoracic disease when they were referred for pulmonary metastasectomy. Nine patients received chemotherapy as part of multidisciplinary management of either their primary disease or subsequent metastases. Two patients received radiation to their primary sites of disease; none received radiation to their metastases. One patient received a brief course of a histone-deacetylase inhibitor before arrival at the NCI. Two patients with synovial sarcomas underwent adoptive transfer of autol- ogous T cells genetically engineered to recognize NY- ESO-1 after a nonmyeloablative preparative regimen with subsequent disease progression in the lungs. All metastasectomy procedures were deemed complete re- sections. No postoperative deaths and no major compli- cations occurred. Seven patients ultimately died of disease progression, 3 patients are alive with disease, and 2 patients currently have no evidence of disease.
Global Methylation Status of Pulmonary Metastases
In light of observations that cancers exhibit global DNA demethylation, pyrosequencing experiments were un- dertaken to examine DNA methylation in Long Inter- spersed Element-1 (LINE-1), D4Z4, and NBL2 repetitive DNA sequences. Results of these experiments are sum- marized in Figure 1. Overall, pulmonary metastases exhibited hypomethylation of these sequences relative to normal lung tissues (Fig 1A). In addition, histologically related changes in DNA methylation within these repetitive elements were apparent. As a group, ACC metastases exhibited more hypomethylation of LINE-1 and NBL2 compared with sarcoma metastases, whereas
| Patient (P) | Diagnosis | Sex | Diagnosis Age at | Stage at Diagnosis® | Pulmonary Metastases at Diagnosis (Y/N) | Location of Primary Tumor | Treatment of Primary Tumor (S/C/R) | Disease-Free Interval (mo) | Treatment of Subsequent Pulmonary Metastases (S/C/ACT/R) | Number of Thoracotomies | Current Status |
|---|---|---|---|---|---|---|---|---|---|---|---|
| P1 | Chondrosarcoma | M | 17 | IIB | N | Extremity | S | 22 | S | 2 | DOD |
| P2 | Chondrosarcoma | F | 62 | III | N | Extremity | S | 30 | S | 2 | NED |
| P3 | Synovial sarcoma | M | 29 | IV | Y | Chest wall | C/S | 3 | C/ACT/S | 3 | DOD |
| P4 | Synovial sarcoma | F | 32 | III | N | Extremity | S/R | 36 | S | 4 | DOD |
| P5 | Synovial sarcoma | M | 36 | III | N | Extremity | R/S | 9 | C/S | 2 | DOD |
| P6 | ACC | F | 54 | IV | Y | Adrenal | C/S | 9 | C/S | 5 | DOD |
| P7 | ACC | F | 59 | II | N | Adrenal | S | 47 | C/S | 3 | DOD |
| P8 | ACC | F | 59 | IV | N | Adrenal | S/C | 1 | C/S | 3 | AWD |
| P9 | ACC | F | 44 | IV | N | Adrenal | S/C | 4 | C/S | 3 | DOD |
| P10 | ACC | F | 32 | II | N | Adrenal | S/C | 21 | C/S | 2 | DOD |
| P11 | ACC | F | 66 | II | N | Adrenal | S | 10 | S/C | 2 | AWD |
| P12 | ACC | M | 56 | II | N | Adrenal | S | 32 | C/S | 2 | AWD |
a Data from American Joint Committee on Cancer. AJCC cancer staging manual, 7th ed. New York: Springer, 2010.
ACT = adoptive cell transfer (ACT) with genetically modified T-cell receptors that recognize cancer-testis gene, NY-ESO-1;
disease; NED = no evidence of disease; R = radiation; S = surgery.
AWD = alive with disease;
C = chemotherapy; DOD = died of
sarcoma metastases had more hypomethylation of D4Z4 relative to ACC metastases (Fig 1B). Interpatient hetero- geneity was also observed even among tumors of similar histologic types. The extent of DNA hypomethylation in the repetitive elements appeared to be quite uniform across metastases from any given patient irrespective of the anatomic location or the time of resection (Figs 1C and 1D).
Cancer-Testis, Tumor Suppressor, DNA Methyltransferase, and Polycomb Gene Signatures in Pulmonary Metastases
The qRT-PCR experiments were next performed to examine expression levels of genes encoding CTAs and tumor suppressors, frequently de-repressed or silenced by epigenetic mechanisms in human cancers, DNMT1, DNMT3A, DNMT3B, and Polycomb group proteins, EZH2, EED, and SUZ12. Raw mRNA copy numbers were then converted to heat maps to examine patterns of gene expression among the various metastases more clearly. Each horizontal bar in Figure 2 corresponds to a single qRT-PCR run for each metastasis; because the metastases were trisected, three separate analyses were performed for each resected tumor. As a whole, GAGE1, SSX-1, SSX- 2, and CT-45 CT genes were repressed in the majority of tumors despite activation of other CT genes. NY-ESO-1 expression was seen only in synovial sarcoma metasta- ses. MAGE-A3, MAGE-A9, and MAGE-A12 were differ- entially expressed, with upregulation most frequently observed in the sarcoma metastases. Overall, pulmonary metastases exhibited low expression of p16; RASSF1A also appeared to be commonly downregulated in ACC metastases. DNMT1, DNMT3A, and DNMT3B expression levels were elevated in the majority of pulmonary me- tastases. Furthermore, EZH2, EED, and SUZ12 were also upregulated in pulmonary metastases. This initial anal- ysis demonstrated impressively similar signatures among triplicate samples from the same metastases and among metastases from the same patient regardless of anatomic location or time of resection (Fig 3). This latter point was clearly evident when average gene expression levels for all metastases in each patient were compared (Fig 4).
Correlation of Promoter DNA Methylation Status With Cancer-Testis and Tumor Suppressor Gene Expression in Pulmonary Metastases
Additional experiments were performed to confirm that activation of CT genes and repression of TS genes coin- cided with promoter DNA methylation status. Repre- sentative results pertaining to NY-ESO-1, MAGE-A3, RASSF1A, and p16 are depicted in Figure 5. Metastases with low levels of NY-ESO-1 and MAGE-A3 expression had higher levels of DNA methylation in the promoters of these CT genes, whereas metastases with high expression of NY-ESO-1 and MAGE-A3 exhibited hypomethylation of the respective promoters (Figs 5A and 5B). Metastases with low-level RASSF1A expression exhibited hyper- methylation of the RASSF1A promoter, whereas metas- tases with high RASSF1A expression had lower levels of RASSF1A promoter methylation (Fig 5C). These
ARTICLE IN PRESS
A.
LINE-1
NBL2
D4Z4
80-
100-
80-
mean methylation %
mean methylation %
mean methylation %
60-
80-
60-
…
60-
40
40-
40-
20
20
20-
0
0
0
Normal
Tumors
Normal
Tumors
Normal
Tumors
B.
LINE-1
NBL2
D4Z4
**
*
80
100-
80
mean methylation %
mean methylation %
mean methylation %
60
80
60-
…
60
40-
…
40-
40
20
20-
20-
…
0
0
0
Normal
Sarcoma
ACC
Normal
Sarcoma
ACC
Normal
Sarcoma
ACC
C.
D.
80
80
LINE-1 mean methylation %
D4Z4 mean methylation %
60
60
40
40
20
20
0
0
m-c
u-c
NL
P1
P3
P6
P7
P9
P10
m-c
u-c
NL
P1
P3
P6
P7
P9
P10
Sarcoma
ACC
Sarcoma
ACC
relationships were not evident for p16 (Fig 5D), possibly because of the limited number of CpG methylation sites interrogated with pyrosequencing or silencing by Poly- comb proteins without DNA methylation.
DNA Methyltransferase and Polycomb Gene Expression in Pulmonary Metastases
Because human malignant diseases of diverse histologic types exhibit aberrant expression of DNMTs and Polycomb
CTG
TSG
DNMT
PRC-2
Chondrosarcoma
Synovial Sarcoma
Adrenocortical Carcinoma
Normal Lung Positive Control
copy # / actin x e4
| NY-ESO-1 | MAGE-A3 MAGE-A9 MAGE-A12 -A12 GA GAGE1 | SSX-1 | SSX-2 | CT-45 | p16 | RASSF1A | DNMT1 | DNMT3A | DNMT3B | EZH2 | EED | SUZ12 | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| < 100 | 100 - | 500 - | 1,000 - | 5,000 - | 10,000 - 15,000 |
|---|---|---|---|---|---|
| 500 | 1,000 | 5,000 | 10,000 | 15,000 |
Fig 2. Cancer-testis (CT), tumor suppressor (TS), DNA methyltransferase (DNMT), and Polycomb gene signatures in pulmo- nary metastases. Each row corresponds to a single quantitative reverse tran- scription polymerase chain reaction run for each metastasis; three separate analyses were performed for each resected tumor. Each column corresponds to a single CT, TS, DNMT, or Polycomb gene. Gene expression levels were similar among triplicates from the same metastases and among metastases from the same patient irrespective of anatomic location or time of resec- tion. Polycomb repressive complex-2 (PRC-2) was also upregulated in the majority of pulmonary metastases. Results were compared with RNA extracted from normal lung tissue and positive controls. (CTG = human testis; DNMT/PRC-2 = A549, lung cancer cell line; TSG = LP9, normal meso- thelial cell line.)
proteins [16, 17], additional analyses were performed to examine expression of genes encoding DNMT1 (mainte- nance DNMT), DNMT3A, and DNMT3B (de novo DNMTs), as well as EZH2, EED, and SUZ12 (core compo- nents of PRC-2). Results of this analysis are depicted in Figures 2 and 3. In metastases from 8 of the 12 patients (75%), average DNMT1 and DNMT3B mRNA copy numbers were not dramatically different from those observed in normal lung. In contrast, in 12 of 12 patients (100%), DNMT3A levels were higher in metastases than in normal lung. These observations were, however, not statistically significant, possibly because of the small sample size.
The majority of pulmonary metastases exhibited over- expression of one or more Polycomb genes. Because overexpression of any of these genes would increase the function of PRC-2, an equation was devised to reflect overall PRC-2 activity in the metastases (Fig 6). Briefly, a
PRC-2 metastasis score was calculated by summation of the ratios of mRNA copy numbers for EZH2, EED, and SUZ12 in individual normal lung specimens (n = 4) or average mRNA copy numbers for these genes in 65 in- dividual metastases from 12 patients relative to the mean mRNA copy number for these genes in 4 normal lung specimens, divided by 3. Overall, the PRC-2 score was elevated in pulmonary metastases relative to normal lung (Fig 6A; p < 0.05). In 8 of the 12 patients (67%), including 4 patients with sarcomas and 4 patients with ACCs, Poly- comb gene expression was markedly higher in metastases (n = 34) than in normal lung. This finding was reflected in a significantly higher PRC-2 metastasis score in these patients relative to the 4 patients with Polycomb gene expression levels approximating those of normal lung (Fig 6B; mean PRC-2 score, 13.03 ± 1.5 vs 4.58 ± 0.75; p = 0.004).
12/16/10
6/14/12
CTG
TSG
DNMT
PRC-
2
NY-ESO-1 MAGE-A3 MAGE-A9 MAGE-A12 GAGE1
SSX-1
SSX-2
CT-45
p16
RASSF1A DNMT1 DNMT3A DNMT3B
EZH2
EED
SUZ12
12/16/10
[
6/14/12
copy # / actin x e4
< 150
> 150
| Patient (P)/ # Metastases | CTG | TSG | DNMT | PRC-2 | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| NY-ESO-1 | MAGE-A3 MAGE-A9 MAGE-A12 | GAGE1 | SSX-1 | SSX-2 | CT-45 | p16 | RASSF1A | DNMT1 | DNMT3A | DNMT3B | EZH2 | EED | SUZ12 | ||||
| P1 4 | |||||||||||||||||
| Chondrosarcoma | P2 3 | ||||||||||||||||
| P3 3 | |||||||||||||||||
| Synovial | P4 4 | ||||||||||||||||
| Sarcoma | P5 3 | ||||||||||||||||
| P6 13 | |||||||||||||||||
| P7 9 | |||||||||||||||||
| P8 4 | |||||||||||||||||
| Adrenocortical | P9 3 | ||||||||||||||||
| Carcinoma | P10 5 | ||||||||||||||||
| P11 3 | |||||||||||||||||
| P12 11 | |||||||||||||||||
| Normal Lung Positive Control | |||||||||||||||||
copy # / actin x e4
| < 100 | 100 - | 500 - | 1,000 - | 5,000 - | 10,000 - | > 15,000 |
| 500 | 1,000 | 5,000 | 10,000 | 15,000 |
Fig 4. Interpatient heterogeneity in pulmonary metastases. Average cancer-testis (CT), tumor suppressor (TS), DNA methyltransferase (DNMT), and Polycomb gene expression levels for all metastases in each patient (P1 to P12). Interpatient variability was observed even among patients with similar tumor histologic types. (CTG = human testis; DNMT/PRC-2 = A549, lung cancer cell line; TSG = LP9, normal mesothelial cell line.)
A.
NY-ESO-1
B.
MAGE-A3
100000-
100000-
copy # / actin x e4
80000-
80000
60000-
copy # / actin x e4
60000-
40000-
40000
20000
20000
50
50
40-
40
30-
30-
20-
20-
10-
10-
0
0
NL Ts
P1
P3
NL Ts
P6
P10
mean methylation %
80-
mean methylation %
100-
60
80-
60
40
40-
20
20-
0
0
m-c u-c NL Ts
P1
P3
m-c u-c NL Ts
P6
P10
C.
RASSF1A
D.
p16
50000
10000
copy # / actin x e4
40000-
8000-
30000-
copy # / actin x e4
20000
6000-
10000
4000-
2000
500
10
400-
8-
300-
6-
200-
4-
100
2-
0
0
NL CC
P9
P6
NL CC
P10
P9
mean methylation %
30
mean methylation %
40
30
20-
20
10-
10-
0
0
m-c u-c NL CC
P9
P6
m-c u-c NL CC
P10
P9
Comment
During recent years, the genetic and epigenetic complexity of human cancers has been illuminated [18, 19]. Cancer stem cells give rise to progeny with varying differentiation capacities. Genetic instability and epigenetic plasticity in these tumor-initiating cells. together with selective pressure within the tumor
microenvironment as a manifestation of exposure to cytotoxic agents, hypoxia, reactive oxygen species, cancer-associated fibroblasts, and immune infiltrating cells, facilitate clonal evolution, acquisition of resistance to chemotherapy and radiation therapy, and tumor dissemination [3, 20-22]. Recent studies in murine models [23-25], as well as in human cancer specimens [26-27],
A.
B.
PRC-2 Score
25-
**
SUZ12
*
20-
EED
30
PRC-2 Score
EZH2
15-
Score
20
10-
10
5-
0
0
Normal
Tumors
P1
P6
P11
P12
P2
P3
P4
P5
P6
P8
P9
P10
have demonstrated that individual clones within hetero- geneous tumors have unique predilections for organ- specific metastases and that disseminated tumor cells may remain dormant for prolonged periods until they are stimulated to grow by cytokines produced within the metastatic niche [28]. These elegant studies provide clear molecular evidence of the “seed and soil” mechanisms of cancer metastases postulated by Paget more than a cen- tury ago [29].
Observations from human patients with cancer [30] and from murine tumor models [31] suggest that metastatic processes are inherently inefficient and that clonality or heterogeneity of pulmonary metastases depends to some extent on circulating tumor burden [32]. In the present study, we sought to examine epigenetic signatures of pulmonary metastases. Although metastases from a va- riety of malignant diseases were evaluated, we focused our present efforts on metastases from sarcomas and ACCs (reflecting current NCI protocol activities), to optimize patient homogeneity. Our analysis revealed patient-specific patterns of DNA demethylation and expression profiles of epigenetically regulated CT and TS genes that were remarkably consistent among metastases irrespective of the time of resection (simultaneous or sequential) or the anatomic location in the lungs. Signif- icant interpatient variability was observed even among patients with similar tumor histologic features. Overall, our findings are consistent with patient-specific epige- netic clonality of pulmonary metastases.
Because cancer cells exhibit alterations in chromatin structure that recapitulate epigenomic states in stem cells [13], it is conceivable that DNA demethylating agents can simultaneously induce growth arrest and augment immunogenicity of thoracic malignant diseases through reactivation of epigenetically silenced TS genes and upregulation of CTAs [33]. In published studies, we have demonstrated induction of NY-ESO-1 and MAGE-A3 in cancer cells of various histologic types, but not normal respiratory epithelial cells or fibroblasts, after exposure to the DNA demethylating agent decitabine (DAC); upre- gulation of the respective CTAs facilitates cytotoxic lymphocyte-mediated lysis of tumor cells [15, 34, 35]. We have also demonstrated eradication of pulmonary me- tastases in immunocompetent mice following systemic treatment with DAC and subsequent adoptive transfer of
cytotoxic lymphocytes recognizing the murine CTA, P1A [36]. Finally, in a phase I study we observed upregulation of intratumoral p16, NY-ESO-1, and MAGE-A3 in pa- tients with lung cancer after intravenous 72-hour DAC infusions [37]. Collectively, these studies provide proof of concept for the use of chromatin remodeling agents for the treatment of thoracic malignant diseases.
The present study was initiated as part of a larger effort to develop precision epigenetic therapies for patients with pulmonary metastases. If pulmonary metastases are indeed epigenetically clonal, then patients undergoing bilateral sequential metastasectomies could be ideal candidates for examining intratumoral molecular re- sponses and systemic antitumor immunity mediated by chromatin remodeling agents administered between sequential bilateral metastasectomy procedures. A pro- tocol examining the feasibility and efficacy of oral DAC and oral tetrahydrouridine (an inhibitor of cytidine deaminase that is responsible for rapid deamination of DAC in vivo) in patients undergoing bilateral pulmonary metastasectomies has recently been initiated at the NCI. Conceivably, this patient population will provide an excellent translational “platform” for the development of epigenetic cancer therapies administered alone or in conjunction with immune check point inhibitors or adoptive transfer of T cells genetically engineered to recognize CTAs.
A secondary objective of the present study was the identification of novel therapeutic targets in pulmonary metastases. Our analysis demonstrated overexpression of PRC-2, a critical mediator of stem cell pluripotency and malignancy [17, 38]. A calculated PRC-2 score, possibly reflective of overall PRC-2 activity, was significantly elevated in metastases from 8 of the 12 patients in this study. More thorough analysis of considerably more samples from patients will be required to validate these findings and to determine whether PRC-2 score correlates with predilection for disease recurrence and overall prognosis in patients with pulmonary metastases. These experiments are planned and will be included in future clinical trial design.
Potential limitations of this study include the relatively small number of patients and the lack of correlative an- alyses of the respective primary tumors. Furthermore, the relatively limited number of epigenetically regulated
genes evaluated by qRT-PCR techniques in our study may have underestimated the extent of heterogeneity in the metastases potentially detectable by more compre- hensive analyses of gene expression [39]. Another limi- tation includes the use of normal lung rather than histologically specific normal tissue controls for calcula- tion of the PRC-2 metastasis score. However, as previ- ously mentioned, this analysis was undertaken as an initial examination of epigenetic clonality of pulmonary metastases; hence we were primarily interested in pat- terns of gene expression in the metastases, rather than differences or similarities relative to normal tissues. We chose to use normal lung tissues rather than cultured epithelial cells or fibroblasts for the Polycomb gene ana- lyses to account for various stromal elements, which were admixed with the tumor cells in the metastases.
Despite the aforementioned limitations, our present findings have direct translational relevance regarding the development of treatment regimens targeting aberrant CT and TS gene expression in pulmonary metastases. Studies have been initiated in our laboratory using DNA methyl- ation arrays and next-generation sequencing techniques to analyze pulmonary metastases and their respective primary tumors comprehensively in an effort to characterize epi- genomic signatures predisposing to or corresponding with pulmonary metastases, as well as to identify epigenetic targets for precision therapy of these malignant diseases.
This work was supported by NCI Intramural grants ZIA BC 011122 (D.S.S.) and ZIA BC 011418 (D.S.S.).
References
1. Skubitz KM, Francis P, Skubitz AP, Luo X, Nilbert M. Gene expression identifies heterogeneity of metastatic propensity in high-grade soft tissue sarcomas. Cancer 2012;118:4235-43.
2. Kim MS, Zhong Y, Yachida S, et al. Heterogeneity of pancreatic cancer metastases in a single patient revealed by quantitative proteomics. Mol Cell Proteomics 2014;13:2803-11.
3. Easwaran H, Tsai HC, Baylin SB. Cancer epigenetics: tumor heterogeneity, plasticity of stem-like states, and drug resis- tance. Mol Cell 2014;54:716-27.
4. Shukla S, Meeran SM. Epigenetics of cancer stem cells: pathways and therapeutics. Biochim Biophys Acta 2014;1840: 3494-502.
5. Deo SS, Shukla NK, Khanna P, Jha D, Pandit A, Thulkar S. Pulmonary metastasectomy: review of experience at a ter- tiary cancer care center. J Cancer Res Ther 2014;10:535-9.
6. Pastorino U, Buyse M, Friedel G, et al. Long-term results of lung metastasectomy: prognostic analyses based on 5206 cases. J Thorac Cardiovasc Surg 1997;113:37-49.
7. Casiraghi M, De Pas T, Maisonneuve P, et al. A 10-year single-center experience on 708 lung metastasectomies: the evidence of the “international registry of lung metastases.” J Thorac Oncol 2011;6:1373-8.
8. Kemp CD, Ripley RT, Mathur A, et al. Pulmonary resection for metastatic adrenocortical carcinoma: the National Cancer Institute experience. Ann Thorac Surg 2011;92:1195-200.
9. Pfannschmidt J, Egerer G, Bischof M, Thomas M, Dienemann H. Surgical intervention for pulmonary metas- tases. Dtsch Arztebl Int 2012;109:645-51.
10. Hong IS, Lee HY, Nam JS. Cancer stem cells: the ‘Achilles heel’ of chemo-resistant tumors. Recent Pat Anticancer Drug Discov 2015;10:2-22.
11. Cojoc M, Mabert K, Muders MH, Dubrovska A. A role for cancer stem cells in therapy resistance: cellular and molecular mechanisms. Semin Cancer Biol 2015;31: 16-27.
12. Mimeault M, Batra SK. Molecular biomarkers of cancer stem/progenitor cells associated with progression, metasta- ses, and treatment resistance of aggressive cancers. Cancer Epidemiol Biomarkers Prev 2014;23:234-54.
13. Baylin SB, Jones PA. A decade of exploring the cancer epi- genome: biological and translational implications. Nat Rev Cancer 2011;11:726-34.
14. Liu F, Killian JK, Yang M, et al. Epigenomic alterations and gene expression profiles in respiratory epithelia exposed to cigarette smoke condensate. Oncogene 2010;29: 3650-64.
15. Rao M, Chinnasamy N, Hong JA, et al. Inhibition of histone lysine methylation enhances cancer-testis antigen expression in lung cancer cells: implications for adoptive immuno- therapy of cancer. Cancer Res 2011;71:4192-204.
16. Wongtrakoongate P. Epigenetic therapy of cancer stem and progenitor cells by targeting DNA methylation machineries. World J Stem Cells 2015;7:137-48.
17. Vizan P, Beringer M, Ballare C, Di Croce L. Role of PRC2- associated factors in stem cells and disease. FEBS J 2015;282:1723-35.
18. Hoadley KA, Yau C, Wolf DM, et al. Multiplatform analysis of 12 cancer types reveals molecular classification within and across tissues of origin. Cell 2014;158:929-44.
19. Coarfa C, Pichot C, Jackson A, et al. Analysis of interactions between the epigenome and structural mutability of the genome using Genboree Workbench tools. BMC Bioinfor- matics 2014;15(Suppl 7):S2.
20. Atlasi Y, Looijenga L, Fodde R. Cancer stem cells, pluripo- tency, and cellular heterogeneity: a WNTer perspective. Curr Top Dev Biol 2014;107:373-404.
21. Seyfried TN, Huysentruyt LC. On the origin of cancer metastasis. Crit Rev Oncog 2013;18:43-73.
22. Zhang XH, Jin X, Malladi S, et al. Selection of bone metastasis seeds by mesenchymal signals in the primary tumor stroma. Cell 2013;154:1060-73.
23. Bobek V, Hoffman RM, Kolostova K. Site-specific cytomor- phology of disseminated PC-3 prostate cancer cells visual- ized in vivo with fluorescent proteins. Diagn Cytopathol 2013;41:413-7.
24. Okuda R, Sekine K, Hisamatsu D, et al. Tropism of cancer stem cells to a specific distant organ. In Vivo 2014;28:361-5.
25. Mosoyan G, Nagi C, Marukian S, et al. Multiple breast cancer cell-lines derived from a single tumor differ in their molec- ular characteristics and tumorigenic potential. PLOS One 2013;8:e55145.
26. Chu JE, Allan AL. The role of cancer stem cells in the organ tropism of breast cancer metastasis: a mechanistic balance between the “seed” and the “soil”? Int J Breast Cancer 2012;2012:209748.
27. Park HS, Kim S, Kim K, et al. Pattern of distant recurrence according to the molecular subtypes in Korean women with breast cancer. World J Surg Oncol 2012;10:4.
28. Sosa MS, Bragado P, Aguirre-Ghiso JA. Mechanisms of disseminated cancer cell dormancy: an awakening field. Nat Rev Cancer 2014;14:611-22.
29. Paget S. The distribution of secondary growths in cancer of the breast, 1898. Cancer Metastasis Rev 1989;8:98-101.
30. Tarin D, Price JE, Kettlewell MG, Souter RG, Vass AC, Crossley B. Mechanisms of human tumor metastasis studied in patients with peritoneovenous shunts. Cancer Res 1984;44: 3584-92.
31. Luzzi KJ, MacDonald IC, Schmidt EE, et al. Multistep nature of metastatic inefficiency: dormancy of solitary cells after successful extravasation and limited survival of early micrometastases. Am J Pathol 1998;153:865-73.
32. Yamamoto N, Yang M, Jiang P, et al. Determination of clonality of metastasis by cell-specific color-coded fluorescent-protein imaging. Cancer Res 2003;63:7785-90.
33. Schrump DS. Targeting epigenetic mediators of gene expression in thoracic malignancies. Biochim Biophys Acta 2012;1819:836-45.
34. Hong JA, Kang Y, Abdullaev Z, et al. Reciprocal binding of CTCF and BORIS to the NY-ESO-1 promoter coincides with derepression of this cancer-testis gene in lung cancer cells. Cancer Res 2005;65:7763-74.
35. Schrump DS, Nguyen DM. Targeting the epigenome for the treatment and prevention of lung cancer. Semin Oncol 2005;32:488-502.
36. Guo ZS, Hong JA, Irvine KR, et al. De novo induction of a cancer/testis antigen by 5-aza-2’-deoxycytidine augments
adoptive immunotherapy in a murine tumor model. Cancer Res 2006;66:1105-13.
37. Schrump DS, Fischette MR, Nguyen DM, et al. Phase I study of decitabine-mediated gene expression in patients with cancers involving the lungs, esophagus, or pleura. Clin Cancer Res 2006;12:5777-85.
38. Kemp CD, Rao M, Xi S, et al. Polycomb repressor complex-2 is a novel target for mesothelioma therapy. Clin Cancer Res 2012;18:77-90.
39. Malchenko S, Seftor EA, Nikolsky Y, et al. Putative multi- functional signature of lung metastases in dedifferentiated chondrosarcoma. Sarcoma 2012;2012:820254.
DISCUSSION
DR P. VAN SCHIL (Antwerp, Belgium): Did you observe any changes according to the therapy given to the patient? For example, the patient given multimodality therapy as e.g. induc- tion chemotherapy followed by surgery?
DR REARDON: It is difficult to assess. All of these patients were treated as part of an NIH-funded clinical trial and received a variety of multimodality treatments; therefore, it is not possible to associate clinical outcome with one modality versus another. Moreover, that was not really the intention of our analysis.
DR VAN SCHIL: Regarding histology, did you compare the growth pattern between the primary tumor and the metastasis?
DR REARDON: No. We did not examine the primary tumors in this study. This was an initial attempt, really driven by our interest in clinical trial design, to see if pulmonary metastases demon- strate clonality comparing one patient to another. At the NCI, we have initiated a clinical trial, which will begin enrollment in the upcoming months, designed to evaluate patients with bilateral metastases. Patients will undergo resection in one hemithorax, receive 8 to 12 weeks of oral epigenetic therapy, followed by resection of the contralateral hemithorax. We will then be able to assess whether there is upregulation or inactivation of the genes that we discussed here, in particular, the cancer-testis antigens, which remain attractive immunotherapy targets. Also, at the time in which this study was conducted, we did not have access to many of the patients’ primary tumors. Certainly, if we are able to obtain the primary tumors, we will then be able to perform more comparative analyses, with the goal being to potentially identify epigenetic signatures that might predict metastasis and/or serve as biomarkers for more aggressive disease.
DR VAN SCHIL: We have a fellow looking at our series of renal cancer metastases, and in some cases the growth pattern was different between the primary tumor and the metastasis, so angiogenic versus nonangiogenic growth pattern.
DR REARDON: That is interesting. Unfortunately, the majority of the patients who present to the NCI have metastatic disease and have had definitive treatment of their primary tumors else- where. We did not have access to the primary tumors at the time of this study.
DR J. DONINGTON (New York, NY): Excellent paper. I share a similar question as Dr. Van Schil about the primaries. I think it’s pretty important. Also, in your heavily pretreated patients, had anyone been on HDAC inhibitors?
DR D. RAYMOND (Cleveland, OH): This may be a very ignorant question, but looking at the whole notion of an epigenetic signature, one of the common clinical scenarios that we run into is trying to differentiate people who have multifocal disease or metastatic disease. Could you foresee using such a signature to help us differentiate those patients who may have multifocal synchronous primaries versus metastatic disease so that we can better prognosticate and treat?
DR REARDON: That is a clever idea. The data that we have presented here include patients with metastatic sarcomas and adrenal cortical carcinomas to the lung, and therefore it is diffi- cult to extrapolate these findings to the case that you describe, in which we’d be seeking to differentiate synchronous versus metachronous non-small cell lung cancers. That being said, it may be theoretically possible to use patient-specific epigenetic and/or genetic signatures as markers of a patient’s tumor biology in time and thus, potentially differentiate between tumors of different origins. Our study cannot, unfortunately, answer that question.
DR A. CHANG (Ann Arbor, MI): It’s hard sometimes to get these patient samples, and cell line studies also have potential flaws, but were you able to go back and compare cell lines from met- astatic lesions and see if you saw the same upregulation in Pol- ycomb repressive complexes across more tissue types?
DR REARDON: Sure. We do have a cell line derived from one of the adrenocortical carcinoma patients, but none of that work has been done yet. One of the other interesting questions that we would like to address in evaluating these patients’ tumor speci- mens, and this will require obtaining the patients’ primary tu- mors, is: Is there evidence of organ-specific metastasis? Can we identify genetic/epigenetic signatures that are unique to pul- monary metastases and the tumor microenvironment within the lung and use these as diagnostic and/or therapeutic tools? It may be useful to work with tumor-derived cell lines to answer these types of questions.