1 Detection and monitoring of circulating tumor DNA in adrenocortical carcinoma
2 Garinet Simon1, Nectoux Juliette1,2, Neou Mario1, Pasmant Eric2,8, Jouinot Anne1,3, 3
4 Sibony Mathilde4, Orhant Lucie2, Pipoli da Fonseca Juliana7, Perlemoine Karine1,
5 Bricaire Léopoldine6, Groussin Lionel1,6,9, Soubrane Olivier9, Dousset Bertrand5, Libe
6 Rossella1,6,9, Letourneur Franck7, Bertherat Jerome 1,6,10, Assié Guillaume 1,6,10
7 8 1Institut Cochin INSERM U1016/UMR8104 and CNRS UMR-S8104, Paris, France
2Laboratory of Genetics and Molecular Biology, Hôpital Cochin, Assistance Publique
9 10 - Hôpitaux de Paris, France
11 3Department of Medical Oncology, Hôpital Cochin, Assistance Publique - Hôpitaux
12 de Paris, France
13 4Department of Pathology, Hôpital Cochin, Assistance Publique - Hôpitaux de Paris,
14 France
5Department of Digestive and Endocrine Surgery, Assistance Publique - Hôpitaux de
15 16 Paris, France
17 6Department of Endocrinology, Cochin Hospital, Assistance Publique - Hôpitaux de
18 Paris, France
19 7Institut Cochin GENOMI’C platform, Paris, France
20 8INSERM UMR745, Biological and Pharmaceutical Sciences University, Université
21 Paris Descartes, Sorbonne Paris Cité, Paris
22 9Department of Hepato-Pancreato-Biliary Surgery, Hôpital Beaujon, Assistance
23 Publique - Hôpitaux de Paris, France
24 10Reference Center for Rare Adrenal Diseases, Reference Center for Rare Adrenal
25 Cancer Network COMETE, Hôpital Cochin, AssistancePublique - Hôpitaux de Paris,
26 Paris, France
27 Correspondance : Pr Guillaume Assié
28 Département “Endocrinologie, Métabolisme et Diabètes”
29 Bâtiment Faculté, 3ème étage 30 24 rue du faubourg Saint Jacques
31
75014 Paris-France
32 Tel : +33 1 44 41 23 93
Email : guillaume.assie@aphp.fr
Number of words: 1432 (except, acknowledgments, legends and references)
33 34 35 36 37 38 39 40
Keywords: ctDNA, biomarker, adrenocortical carcinoma, NGS, digital droplet PCR
41
42 43 44 45
46 47 48
2
49 50
Dear editor,
Adrenocortical carcinomas (ACC) are rare but aggressive tumors. Currently, the 51 52 single curative approach is complete surgery. Prognosis, response to treatment and recurrence remain unpredictable, which stresses the need for new biomarkers. (Else 53 et al. 2014). Recent exome sequencing approaches of tumors identified in 60% of 54 ACC recurrent somatic mutations in 20 genes (Assié et al. 2014a, b; Juhlin et al. 55 2015; Zheng et al. 2016). Somatic mutations can be used as surrogate biomarkers 56 for detecting circulating tumor DNA (ctDNA) in blood. This ctDNA corresponds to 57 fragments of DNA released directly by tumor cells into the blood stream among the 58 circulating cell-free DNA (ccfDNA). Discrimination of ctDNA from ccfDNA of non- 59 60 tumoral origin is based on the detection of somatic mutations, specific of cancer cells. The amount of ccfDNA and the detection of ctDNA largely depend on tumor type and disease stage (Bettegowda et al. 2014). In ACC, ctDNA detection has been 61 62 recently reported in one patient (Creemers et al.). However, beyond this proof of concept, the proportion of ACC patients with detectable ctDNA is not established. 63 64 The aim of this study was to assess to which extent ctDNA can be detected in ACC patients, using two highly sensitive techniques: deep NGS and droplet digital PCR (ddPCR). We also assessed the evolution of ctDNA during the course of the disease. 65 66 67
68 We prospectively and randomly included 11 patients with ACC. For four patients 69 70 blood was sampled before primary surgery, for three patients at the time of a small relapse or metastasis occurring <2 years after primary surgery, and for four patients in the setting of a rapidly growing metastatic disease. Median age was 48 (range 31 71 72 73
to 81). Six patients had glucocorticoids hypersecretion. Two patients had androgen hypersecretion, while two patients presented no hypersecretion (Table 1). All
74 patients were informed of the project, and signed a written consent for the genetic 75 76
study of the tumor. A prior agreement from the local ethics committee was obtained, under the COMETE -TACTIC framework.
77 Tumor and leukocytes DNA were extracted for each patient. Next-generation 78 sequencing (NGS) workflow based on a custom AmpliSeq™ (Thermofisher, Villebon,
79
80
81
82
83
84 85 86
France) panel was designed for sequencing the 20 genes known to be frequently mutated in ACCs, using a Ion Torrent PGM™ Sequencer (Life Technologies, Carlsbad, California). Seventeen somatic mutations were detected in eight patients, (Table 1). TP53 (5 hits) and CTNNB1 (3 hits) were the most affected genes, followed by NF1 (2 hits) and single-hit mutations in TERT, RPL22, ATRX, MED12 and MEN1. Four patients had more than one somatic mutation, ranging from 2 to 6. Tumor cellularity, clonality and heterozygocity status was assessed with our R algorithm TARGOMICS (Garinet et al. 2017, data not shown). Three patients presented no mutations in the genes studied.
87 88
89
For the eight patients with at least one somatic mutation identified in the tumor, ccfDNA was extracted from plasma collected either before surgery, or at the time of relapse, metastases or follow-up obtained after double centrifugation of 20ml of blood (BCT Cell-Free DNA™ Blood Collection Tube). Concentrations were variable,
90
91
from 3 to 422 ng/ml of plasma (Table 1). For each patient, tumor mutations were searched in ccfDNA either by deep sequencing or by ddPCR. A library with a unique
92 93
94 amplicon harboring the mutation was subsequently prepared for each patient relative to the mutation identified in the tumor DNA and sequenced with an expected depth of 100000X.
95 96 97
Four mutations were detected by NGS, in two patients, 1 and 7 (Table 1). Notably,
98 these patients displayed massively metastatic diseases, with a rapid evolution. Allelic
ratios in tumors indicated that these mutations were present in all tumor cells. Concentrations of ccfDNA were high -17 and 422 ng/ml -. For patient 1, the three somatic mutations - in RB1, TP53 and CTNNB1 genes- were found in ccfDNA. The allelic ratios were close -11.7%, 16.6% and 14.3% respectively -. For patient 7, the somatic mutation in CTNNB1 was found, with an allelic ratio of 13.8%. For both patients, mutations were also detected with ddPCR. Allelic ratios -15.2% and 15.3% for patient 1 and patient 7 respectively- were comparable to those obtained by NGS.
In contrast with these two patients, no ctDNA mutation was detected for 6 out of 8 patients, neither by deep NGS, nor by ddPCR (Table 1). Several parameters may impact detection of ctDNA. First the tumor burden and progression may have a role. Indeed tumor volume was small and slowly/not progressing for three negative 110 patients -patients 2, 3, 5- and intermediate for one -patient 8 -. Yet, two of the six 111 negative patients -patients 4 and 6- also presented with a large and progressing tumor. Tumor burden is not fully associated with ctDNA detection. The clonality of mutations may also have an impact. Indeed no ctDNA could be detected for the two patients with subclonal mutations -patients 2 and 5 -. One tumor -patient 5- was of borderline malignancy (Weiss score of 2), and it remains to be checked whether ctDNA can be detected in benign adrenal tumor.
It remains challenging to conclude whether ctDNA detection was negative because of a limited sensitivity. Indeed, one theoretical limitation of the sensitivity remains the initial quantity of DNA in the test sample. In our study, quantities of ccfDNA were low (<10ng/ml) for three patients -patients 2, 3 and 5-, and the number of DNA copies analyzed in ddPCR remained <300 for patients 2, 3 and 8. The minimum detectable ratio is thus lowered, and this is a detection limit de facto. However, it might be 123 possible that some ACC do not release any ctDNA. This is the case for some tumor
99 100 101 102 103 104 105 106 107 108 109
112 113 114 115 116 117 118 119 120 121 122
124 types, like glioma or renal cell carcinoma for which ctDNA is detected in less than 50% of patients, with currently no clear explanation (Bettegowda et al. 2014). Further studies will be needed to test this hypothesis. Compared to the study of Creemers et al., who used a NGS panel, our study used two of the currently most sensitive molecular biology technologies, including deep NGS and ddPCR. However, though being limited, our series suggests at least that patients with the most aggressive forms of the disease do secrete ctDNA. This subgroup of patients with aggressive disease is certainly the subgroup in which such a tool as ctDNA may have the most obvious applications, both in terms of prognosis and in terms of disease follow-up.
Amounts of ctDNA were monitored during follow-up and our data suggest that quantity of ctDNA parallels disease evolution (Fig. 1). Patient 1 had a pancreatic recurrence and no hepatic lesion at the time of first blood sampling. After 6 months, both ctDNA absolute quantity and allelic ratios increased -from 2.7 to 9.9 x 103 copies of DNA in a 10 ml sample, and from 13.5 to 27% respectively - (Fig. 1A). This increase paralleled disease progression, with the appearance of liver metastases and growth of the pancreatic lesion (Fig. 1C and D). Patient 7 had small pulmonary and liver metastases at the time of first blood sampling. After one month, ctDNA absolute quantity remained similar, while allelic ratio decreased - from 56 to 48 x 103 copies of DNA in a 10 mL sample, and from 13.8% to 8.9% of ctDNA respectively -. Two months later, both ctDNA absolute quantity and allelic ratio increased importantly, reaching 264 x 103 copies and 31.6% of ctDNA. This rapid increase paralleled disease progression, with a rapid and massive growth of metastases (Fig.1 E, F and G), leading to patient’s death one month later. For both patients, quantities of non-tumoral ccfDNA were also at a higher level than observed
142 143 144 145 146 147 148
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141
in healthy controls, and increased strongly when patient 7 was in terminal phase. This finding has been observed in other studies, and could be explained by an overall inflammation and excessive cell death releasing more ccfDNA (Zhang et al. 2017). Therefore in clinical routine, ctDNA should be monitored with both the ratio relative to ccfDNA, and the absolute quantity. Monitoring ctDNA could help to monitor the response to treatment and disease progression in ACC patients. Further studies are needed, to compare ctDNA with imaging techniques for this purpose, both in terms of sensitivity.
In conclusion, ctDNA is detectable in a subset of ACC patients. When detected, ctDNA can be accurately quantified and seems to follow tumor dynamics. However, ctDNA could not be detected in several patients, including some with large tumor burden, and despite the use of highly sensitive technologies. Thus, despite some promising value, it is not currently possible to foresee the exact place of ctDNA in ACC management. Its applications as a potential biomarker remain to be determined on a larger cohort, with a longitudinal monitoring.
149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174
175 Declaration of interest
The authors report no conflict of interest in this work
Fundings:
We thank Brou de Lauriere Dotation Funds, Carigest-promex Charitable Foundation, La Ligue Contre le Cancer -Carte d’identité des Tumeurs program-, ENSAT FP7 grant 259735, Comete-TACTIC grant PRT-K 14079 and INCA_DGOS 8663, and Departement Hospitalo- Universitaire (DHU) AUTHORS for their financial support.
References
Assié G, Jouinot A & Bertherat J 2014a The “omics” of adrenocortical tumours for personalized medicine. Nature Reviews. Endocrinology 10 215-228. (doi:10.1038/nrendo.2013.272)
Assié G, Letouze E, Fassnacht M, Jouinot A, Luscap W, Barreau O, Omeiri H, Rodriguez S, Perlemoine K, René-Corail F et al. 2014b Integrated genomic characterization of adrenocortical carcinoma. Nature Genetics 46 607-612. (doi:10.1038/ng.2953)
Bettegowda C, Sausen M, Leary RJ, Kinde I, Wang Y, Agrawal N, Bartlett BR, Wang H, Luber B, Alani RM et al. 2014 Detection of circulating tumor DNA in early- and late-stage human malignancies. Science Translational Medicine 6 224ra24. (doi:10.1126/scitranslmed.3007094)
Creemers SG, Korpershoek E, Atmodimedjo PN, Dinjens WNM, Koetsveld V, M P, Feelders RA & Hofland LJ Identification of mutations in Cell-free Circulating Tumor DNA in Adrenocortical Carcinoma: a Case Series. The Journal of Clinical Endocrinology & Metabolism. (doi:10.1210/jc.2017-00174)
Else T, Kim AC, Sabolch A, Raymond VM, Kandathil A, Caoili EM, Jolly S, Miller BS, Giordano TJ & Hammer GD 2014 Adrenocortical carcinoma. Endocrine Reviews 35 282-326. (doi:10.1210/er.2013-1029)
Garinet S, Néou M, de La Villeon B, Faillot S, Sakat J, Da Fonseca JP, Jouinot A, Le Tourneau C, Kamal M, Luscap-Rondof W et al. 2017 Calling Chromosome Alterations, DNA Methylation Statuses, and Mutations in Tumors by Simple Targeted Next-Generation Sequencing: A Solution for Transferring Integrated Pangenomic Studies into Routine Practice? The Journal of Molecular Diagnostics: JMD 19 776-787. (doi:10.1016/j.jmoldx.2017.06.005)
176 177 178 179 180 181 182 183
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208
Juhlin CC, Goh G, Healy JM, Fonseca AL, Scholl UI, Stenman A, Kunstman JW, Brown TC, Overton JD, Mane SM et al. 2015 Whole-exome sequencing characterizes the landscape of somatic mutations and copy number alterations in adrenocortical carcinoma. The Journal of Clinical Endocrinology and Metabolism 100 E493-502. (doi:10.1210/jc.2014-3282)
Untergasser A, Cutcutache I, Koressaar T, Ye J, Faircloth BC, Remm M & Rozen SG 2012 Primer3 — new capabilities and interfaces. Nucleic Acids Research 40 e115. (doi:10.1093/nar/gks596)
Zhang W, Xia W, Lv Z, Ni C, Xin Y & Yang L 2017 Liquid Biopsy for Cancer: Circulating Tumor Cells, Circulating Free DNA or Exosomes? Cellular Physiology and Biochemistry: International Journal of Experimental Cellular Physiology, Biochemistry, and Pharmacology 41 755-768. (doi:10.1159/000458736)
Zheng S, Cherniack AD, Dewal N, Moffitt RA, Danilova L, Murray BA, Lerario AM, Else T, Knijnenburg TA, Ciriello G et al. 2016 Comprehensive Pan-Genomic Characterization of Adrenocortical Carcinoma. Cancer Cell 29 723-736. (doi:10.1016/j.ccell.2016.04.002)
209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
1 Figure legend:
2 Figure 1: Monitoring of ctDNA and tumor evolution in patients with positive detection.
A, B: relative and absolute concentration of ctDNA among ccfDNA in patients 1 and 3
4 7 respectively. C, D: Growth of pancreatic and hepatic lesions at CT-scan in patient 5 1. E, F, G: Increase of pulmonary lesions at CT-scan in patient 7.
6
Table 1 : Patients main clinical features
| Patients feature at the time of ccfDNA sampling | Primary tumor features | ccfDNA | Last follow up | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Nº Age | Gender | Time to surgery (months) | Primary in place | Metas- tases | Loco- regional recurrence | Tumor burden | Disease progression* | Steroid secretion | ENSAT stage | Tumor size (cm) | Weiss score | Ki67 | Mutated genes (allelic ratio) | DNA Concen- -tration (ng/ml of plasma) | NGS depth | NGS allelic ratio | ddPCR allelic ratio (positive droplets/total) | Specific death (time) |
| 1 44 | M | 21 | no | liver pancreas lung | yes | multiple large metastases | progression | none | 4 | 8 | 7 | 40% | TP53 c.611delT p.Leu204fs (93%) CTNNB/ c.134C>T pSer45Phe (92%) RB1 c.1388C>G p.Ser463* (93%) | 17.3 | 301308 155564 117413 | 11,7% 16,6% 14,3% | - 15,2% (251/1649) - | yes (28) |
| 2 38 | F | 19 | no | lung | no | micronodule | stable | A | 4 | 6 | 7 | 9% | TERT c.2657-46C>T (17%) TERT c.2677G>C p.Glu893Gln (18%) | 3 | - 121032 | - <0.1% | - 0/101 | no (47) |
| 3 48 | F | 31 | no | lung | yes | micronodules | stable | GC | 3 | 20 | 9 | 20% | RPL22 c.305T>C p. Val102A la (5%) NF1 c.4970A>Gp.Tyr1657Cys (5%) NF1 c.5843 _ 5846del p.Gln1948fs (62%) TP53 c.146G>A p.Arg49His (64%) ATRX c.2542_2545del p.Glu848fs (72%) MED12 c.5223A>T p.Pro1741Pro (17%) | 6 | - - - 105972 98586 - | - - - <0.1% <0.1% - | - - - - 0/223 - | no (36) |
| 4 55 | F | 19 | no | liver lung | no | multiple large metastases | progression | GC | 4 | 10 | 8 | 20% | NF1 c.5017G>T pGlu1673* (75%) | 31,5 | 123524 | <0.1% | 0/7658 | yes (27) |
| 5 31 | F | before | yes | no | no | small primary | none | GC | 1 | 2.5 | 1 | 5% | CTNNB/ c.134C>G p.Ser45Cys(13%) | 9,4 | 21574 | - | 0/1657 | no (17) |
| 6 53 | F | 3 | no | no | yes | multiple large nodules | progression | GC | 2 | 10 | 9 | 40% | CTNNB1 c.387A>G p.Ser129Ser (37%) | 11 | 0 (technical error) | <0.1% | 0/1199 | yes (9) |
| 7 42 | F | before | yes | lung | - | massive primitive and metastases | progression | GC | 4 | 15 | -** | 80%$ | TP53 c.76C>TpArg26Cys (71%) | 422 | 147166 | 13,8% | 15,3% (46/301) | yes (2) |
| 8 81 | F | 21 | no | lung | yes | 35mm recurrence micrometastasis | progression | A | 3 | 6 | 9 | 35% | MEN1 c.779A>G p.Glu260Gly (84%) TP53 c.229delA pArg77fs (83%) | 19 | 109369 123541 | <0.1% <0.1% | - 0/54 | no## (29) |
| 9 21 | F | before | yes | lung bone | - | large primary multiple micrometastases multiple large metastases | progression | GC MC | 4 | 10 | 8 | 35% | - (TP53 germline, 93%) | - | - | - | - | yes (6) |
| 10 53 | F | 30 | no | liver lung | no | (bone and liver) micrometastases | progression | none | 4 | 18 | 6 | 15% | - | - | - | - | - | yes (33) |
| 11 54 | M | before | yes | no | no | (lung) massive primitive | - | GC | 2 | 20 | 8 | 23% | - | - | - | - | - | yes" (0) |
*RECIST criteria
** no Weiss score could be determined since only a biopsy was performed ## palliative care only at this time
# deceased from pulmonary infection after surgery
$ determined on a lung metastases
GC : Glucocorticoïds, MC : Minerlocorticoïde, A : Androgens
A
B
60,0
ctDNA
non tumoral ccfDNA
900
ctDNA
non tumoral ccfDNA
x103 DNA copies /mL of
31,6%
50,0
x103 DNA copies /ml of
800
700
40,0
27%
plasma
600
30,0
plasma
8,9%
500
572
14,3%
26,8
400
13,8%
20,0
300
490
10,0
18.1
200
349
9,9
100
264
0,0
3.0
0
56
48
t0
m+6
to
m+1
m+2
C
D
E
F
G
999×759mm (78 × 78 DPI)