cancers

Article Rescue of Non-Informative Circulating Tumor DNA to Monitor the Mutational Landscape in NSCLC

Stefanie Mayer 1, Gerlinde Schmidtke-Schrezenmeier 2, Christian Buske 3, Frank G. Rücker 4, Thomas F.E. Barth 1, Peter Möller 1 and Ralf Marienfeld 1,* 1 Institute of Pathology, University Medical Center Ulm, 89070 Ulm, Germany; [email protected] (S.M.); [email protected] (T.F.E.B.); [email protected] (P.M.) 2 Department of Internal Medicine II, University Medical Center Ulm, 89070 Ulm, Germany; [email protected] 3 Institute of Experimental Tumor Research, University Medical Center Ulm, 89070 Ulm, Germany; [email protected] 4 Department of Internal Medicine III, University Medical Center Ulm, 89070 Ulm, Germany; [email protected] * Correspondence: [email protected]; Tel.: +49-731-500-56303



Received: 11 June 2020; Accepted: 14 July 2020; Published: 16 July 2020


Abstract: In non-small cell lung cancer (NSCLC) the usage of plasma-derived circulating tumor DNA (ctDNA) have come into focus to obtain a comprehensive genetic profile of a given lung cancer. Despite the usage of specific sampling tubes, archived plasma samples as well as inappropriately treated blood samples still cause a loss of information due to cell lysis and contamination with cellular DNA. Our aim was to establish a reliable protocol to rescue ctDNA from such non-informative samples to monitor the mutational landscape in NSCLC. As a proof-of-concept study we used archived plasma samples derived from whole blood EDTA samples of 51 patients suffering from NSCLC. Analysis of the isolated plasma DNA determined only a small fraction of ctDNA in a range of 90–250 bp. By applying a specific purification procedure, we were able to increase the informative ctDNA content and improve in a cohort of 42 patients the detection of driver mutations from 32% to 79% of the mutations found in tissue biopsies. Thus, we present here an easy to perform, time and cost effective procedure to rescue non-informative ctDNA samples, which is sufficient to detect oncogenic mutations in NGS approaches and is therefore a valuable technical improvement for laboratories handling liquid biopsy samples.

Keywords: liquid biopsy; cell-free DNA; NGS; precision medicine; targeted therapy

1. Introduction Lung cancer is the leading cause in cancer related deaths throughout the world, with non-small cell lung cancer (NSCLC) being the most common subtype [1,2]. However, despite improvements in diagnosis and treatment, the 5-year survival rate remains dismal, due to late diagnosis, metastases and poor responsiveness to chemotherapy [3]. Therefore, early detection methods as well as personalized treatment using small molecule inhibitors on the basis of genetic profiling of a tumor sample are of high importance. During the last years several oncogenic mutations and genetic rearrangements have been discovered in NSCLC, with EGFR, KRAS, ALK, MET, BRAF, PIK3CA, ROS1, HER2 and RET being the most abundant alterations in NSCLC adenocarcinoma [4]. Some of these oncogenic mutations, e.g., EGFR L858R, can be targeted with specific tyrosine kinase inhibitors [5]. However, tumor heterogeneity is common in NSCLC samples, and mutational profiles vary between different metastatic sites or even between different subclones in one site leading to an intra-individual

Cancers 2020, 12, 1917; doi:10.3390/cancers12071917 www.mdpi.com/journal/cancers Cancers 2020, 12, 1917 2 of 12 heterogeneity in their genetic profile [6]. Thus, profiling a single tissue biopsy snapshot might be insufficient to unravel the tumor heterogeneity, potentially leading to false negative results [7]. For instance, a driver mutation located in an underrepresented subclone might be missed thus excluding a therapeutic option leading to treatment failure [1]. As the analysis of multiple biopsies appears to determine the complete tumor mutational landscape of NSCLC is not feasible, the usage of circulating tumor DNA isolated from blood plasma came into focus. Sozzi et al. demonstrated that plasma DNA concentration is increased in cancer patients compared to healthy donors, suggesting that an increase in the plasma DNA amount is an early event in lung carcinogenesis [8]. Likewise, in breast cancer higher plasma DNA levels were observed, despite inter-individual variations [9]. Hence, analysis of circulating tumor DNA (ctDNA) as a non-invasive tool is extensively studied. However, ctDNA is only defined by its molecular size of 90–250 bp and includes a minor amount of small fragment cell-free DNA from normal cells, but consists of a majority of small fragment cell-free DNA from tumor cells [10]. This technique is thought to provide an insight into the tumor heterogeneity, allowing one to monitor the mutational evolution during treatment, and offering a chance for molecular genotyping in case sufficient tissue material is not available. Therefore, ctDNA is thought to be an alternative surrogate for molecular analysis in cancer patients, requiring only a small amount of blood [11]. For instance, ctDNA from late stage NSCLC patients is used to test for additional EGFR mutations leading to resistance against EGFR tyrosine kinase inhibitors [12]. Of note, there are several limitations for the use of plasma samples. It is recommended to perform plasma isolation soon after blood withdrawal, while inappropriate handling and storage leads to non-informative liquid biopsy samples due to leukocyte lysis and therefore contamination with genomic DNA [13–15], thereby leading to non-suitable therapeutic options and treatment failure of those inappropriately handled samples [16]. Further, the usage of different blood collection tubes and centrifugation protocols for plasma isolation might cause a contamination of the plasma samples with high molecular genomic DNA to variable degrees [17–20]. However, how to use those non-appropriate handled liquid biopsy samples for further NGS approaches and how to rescue ctDNA from those samples remains a matter of debate. As the contamination with genomic DNA is a main issue for daily routine lab work in molecular pathology as well as for multi-center studies including archived plasma samples, our aim was to establish a protocol to rescue non-informative DNA samples derived from inappropriately handled liquid biopsy samples and evaluate the feasibility of NGS approaches.

2. Results

2.1. Differences in Sample Handling Leading to Contamination with Cellular DNA To analyze the applicability of archived plasma samples for mutational testing, a total of 51 patients suffering from NSCLC of the histological subtype adenocarcinoma and squamous cell carcinoma were included in a proof-of-concept study. Baseline characteristics of this patient cohort are summarized in Table1. As blood sampling and plasma preparation was performed before liquid biopsy analysis was part of the routine laboratory setup, the regulations regarding processing, shipping and storage time of the blood samples prior to plasma preparation was less well defined. Hence, duration to plasma preparation after blood withdrawal was variable spanning from a few hours to several days at room temperature. For the majority of the samples plasma isolation was performed 1–2 days after blood was drawn (49%, n = 25), whereas for eleven patients (21%) plasma was isolated within 24 h and for one patient plasma isolation was performed 7 days after blood withdrawal (Figure1A). Cancers 2020, 12, 1917 3 of 12

Table 1. Characteristics of the cohort.

Patients 51 age 63.3 9.7 ± Cancers 2020, 12, x N (%) 3 of 12 male 32 (63) female>2 19 (37)12 (23) n/a N (%) 4 (8) adenocarcinomaUICC7 stage 45 (88)N (%) squamous cell carcinomaIA 6 (12) 2 (4) metastatic sitesIIA N (%) 1 (2) 0IIB 9 (18) 1 (2) 1IIIA 16 (31) 4 (8) 2IIIB 10 (20) 4 (8) >2IV 12 (23)38 (74) n/a 4 (8) n/a 1 (2) UICC7 stage N (%) As blood sampling and plasma preparation was performed before liquid biopsy analysis was IA 2 (4) part of the routine laboratory setup, theIIA regulations regarding 1 (2) processing, shipping and storage time of the blood samples prior to plasma preparationIIB was less 1 (2) well defined. Hence, duration to plasma preparation after blood withdrawal wasIIIA variable spanning 4 (8) from a few hours to several days at room temperature. For the majority of the samplesIIIB plasma isolation 4 (8) was performed 1–2 days after blood IV 38 (74) was drawn (49%, n = 25), whereas for elevenn/a patients (21 1%) (2) plasma was isolated within 24 h and for one patient plasma isolation was performed 7 days after blood withdrawal (Figure 1A).

A B

FigureFigure 1. Delay 1. Delay in plasma in plasma separation. separation. Fifty-one Fifty- patientsone patients were were enrolled enrolled in the in study the study of which of which plasma plasma preparationpreparation was performedwas performed at di ffaterent different time pointstime points after bloodafter blood withdrawal. withdrawal. For all For samples all samples plasma plasma

DNADNA was isolated.was isolated. (A) Storage(A) Storage time time of whole of whole blood blood in K2 inEDTA K2EDTA tubes tubes at room at room temperature temperature before before plasmaplasma isolation. isolation. (B) Contamination (B) Contamination of ctDNA of (arrow) ctDNA with (arrow) high molecular with high genomic molecular DNA (prohibition genomic DNA sign)(prohibition was analyzed sign) using was an analyzed Agilent Bio-Analyzer. using an Agilent Bio-Analyzer.

HighHigh inter-individual inter-individual variations variations in plasma in plasma DNA DNA concentration concentration were were observed observed ranging ranging from from non-detectablenon-detectable plasma plasma DNA DNA to 100 to ng100/µ ng/µL withL w aith mean a mean concentration concentration of 9.52 of ng9.52/µ Lng/µ independentL independent of of storagestorage time before time b plasmaefore plasma isolation isolation (Figure2 ). (Figure However, 2). However, mean plasma mean DNA plasma concentration DNA concentration of samples of storedsamples 4 days stored at room 4 days temperature at room beforetemperature plasma before isolation plasma was isolation significant was higher significant compared higher to compared those storedto 2those days stored or less 2 days at room or less temperature at room temperature before plasma before has plasma been isolatedhas been (isolatedp < 0.0001, (p < Figure0.0001,2 Figure). These2). data These underline data underline the impact the ofimpact delay of in delay sample in processingsample processing on plasma on DNAplasma content. DNA content.

Cancers 2020, 12, 1917 4 of 12 Cancers 2020, 12, x 4 of 12

Cancers 2020, 12, x 4 of 12

Figure 2. 2. RelevanceRelevance of of time time delay delay for fortotal total plasma plasma DNA DNA concentration. concentration. Plasma Plasma DNA concentrations DNA concentrations of plasma samples isolated at different timepoints after venipuncture. Samples#9, #19, #21 and #27 were of plasma samples isolated at different timepoints after venipuncture. Samples#9, #19, #21 and #27 notwere processed not processed due due to a to lack a lack of materialof material (ND (ND= = notnot documented).documented). Figure 2. Relevance of time delay for total plasma DNA concentration. Plasma DNA concentrations To monitorof plasma the the samples size size distribution isolated distribution at different of the of timepoints the plasma plasma DNAafter DNAven andipuncture. therefore and thereforeSamples#9, define the#19, define degree#21 and the #27 of degree of contaminationcontaminationwere with not processed high molecular due to a lackcellular cellular of material DNA, DNA, ( NDwe we =analyzed analyzednot documented) the the fragment fragment. size size of of the the extracted extracted plasma DNAplasma using DNA a using Bio-Analyzer a Bio-Analyzer 2100. 2100. Circulating Circulating tumor tumor DNA DNA (ctDNA) (ctDNA) has a a fragment fragment size size of of90– 90–250bp, while250bp high-molecular, whileTo high monitor-molecular cellular the size cellular DNA,distribution DNA, probably probably of the from plasmafrom lysed ly s DNAimmuneed immune and cells, therefore cells, are are high definehigh molecular molecular the degree fragments of contamination with high molecular cellular DNA, we analyzed the fragment size of the extracted clusteringfragments atclustering the length at the of length1000 bpof ≥ [100021]. Fragmentbp [21]. Fragment size analysis size analysis suggested suggested that only that aonly minor a fraction minorplasma fraction DNA of the using isolated a Bio≥ -plasmaAnalyzer DNA 2100. was Circulating in the range tumor expec DNAted (ctDNA) for ctDNA has a(Figure fragment 1B). size All of 90– of the isolated plasma DNA was in the range expected for ctDNA (Figure1B). All samples contained samples250bp contained, while longerhigh-molecular DNA fragments cellular ranging DNA, probably from 14% from to 99% lys withed immune a mean cells,of 69%, are suggesting high molecular longera contaminationfragments DNA fragments clustering with high ranging at molecularthe length from of cellular14% ≥1000 to 99% DNA.bp [21 with ]. We Fragment a subdivided mean size of 69%, analysis our suggesting samples suggested into a contaminationthat three only a withsubgroups highminor— molecular fraction(1) plasma of cellular theisolation isolated DNA. within plasma We24 h subdividedDNA after bloodwas in withdrawal, ourthe range samples expec (2) intoplasmatedthree for isolation ctDNA subgroups—(1) between(Figure 1B 1 ). plasmaAll isolationand 2 samplesdays within after contained blood 24 h afterwithdrawal longer blood DNA and withdrawal, fragments (3) plasma ranging (2) isolation plasma from at 14% isolationleast to 3 99% days between with after a meanblood 1 and ofwithdrawal. 6 29%, days suggesting after blood withdrawalThe percentagea contamination and of (3) isolated plasma with ctDNA isolation high representing molecular at least cellular3a daysfragment after DNA. region blood We of withdrawal. subdivided 90–250 bp was our The decreased samples percentage intowithof three isolated ctDNAprolongedsubgroups representing time —between(1) plasma a fragmentblood isolation withdrawal region within ofand 24 90–250 hplasma after bpblood isolation, was withdrawal, decreased with a mean(2) with plasma of prolonged 42 isolation% ctDNA timebetween for between 1 bloodplasmaand withdrawal samples 2 days handled after and blood the plasma samewithdrawal d isolation,ay after and blood with(3) plasmawithdrawal a mean isolation of and 42% 23at % ctDNAleast ctDNA 3 days forfor aftersamples plasma blood for samples withdrawal.which handled theplasma sameThe isolation daypercentage after was blood of performed isolated withdrawal ctDNA at least representing and3 days 23% after ctDNA a bloodfragment for withdrawal samplesregion of (Figu for90–250 whichre 3Abp –was plasmaC). Thus,decreased isolation the with was prolonged time between blood withdrawal and plasma isolation, with a mean of 42% ctDNA for performedcontent of longer at least DNA 3 daysfragments after was blood slightly withdrawal increased (Figurefrom 58%3A–C). to 77% Thus, when plasma the content isolation of longerwas DNA performedplasma 3 samplesdays after handled venipuncture the same d compareday after blood to thosewithdrawal isolated and within 23% ctDNA 24 h for(Figure samples 3A –forC). which fragments was slightly increased from 58% to 77% when plasma isolation was performed 3 days However,plasma despite isolation a larger was variation performed between at least sampl 3 dayses withinafter blood the same withdrawal subgroup (Figu in regardre 3A– Cof). their Thus, the afterctDNA venipuncturecontent content, of longerwe observed compared DNA fragments a significant to those was isolateddifference slightly withinincreas betweened 24 thesefrom h (Figure 58two% togroups3 A–C).77% when (p However, = 0.042). plasma despiteisolation awas larger variationperformed between 3 samplesdays after within venipuncture the same compared subgroup to in regardthose isolated of their within ctDNA 24 content, h (Figure we observed 3A–C). a significantA However, difference despite between a largerB thesevariation two between groups sampl (p = 0.042).es withinC the same subgroup in regard of their ctDNA content, we observed a significant difference between these two groups (p = 0.042). A B C

Figure 3. 3. DelayDelay in inplasma plasma preparation preparation leads leads to increased to increased plasma plasma DNA contamination. DNA contamination. CtDNA content CtDNA content and content of genomic DNA from plasma DNA samples isolated (A) at the same day, (B) at 1–2 days and contentFigure of3. Delay genomic in plasma DNA preparation from plasma leads DNA to increased samples plasma isolated DNA (A )contamination. at the same day, CtDNA (B) atcontent 1–2 days and (C) at ≥3 days after venipuncture. and (Cand) at content3 days of aftergenomic venipuncture. DNA from plasma DNA samples isolated (A) at the same day, (B) at 1–2 days ≥ and (C) at ≥3 days after venipuncture. 2.2.2.2. Size Selection Selection Based Based Purification Purification of Ct ofDNA CtDNA from fromArchived Archived Historic Historic PlasmaPlasma Samples Samples 2.2. Size Selection Based Purification of CtDNA from Archived Historic Plasma Samples The high degree of contaminating high-molecular DNA might affect the mutational profiling of the ctDNAThe probably high degree resulting of contaminating in false negative high-molecular sequencing DNA results. might In affect order the to mutational rescue the profiling contaminated of plasmathe DNA ctDNA samples, probably we resulting established in false a size-selection negative sequencing protocol results.for the purificationIn order to rescueof the ctDNA the (Figure4).

Cancers 2020, 12, x 5 of 12 Cancers 2020, 12, x 5 of 12 Cancerscontaminated2020, 12, 1917 plasma DNA samples, we established a size-selection protocol for the purification of5 of 12 thecontaminated ctDNA (Fi gureplasma 4). DNA samples, we established a size-selection protocol for the purification of the ctDNA (Figure 4).

Figure 4. Purification procedure for the isolation of ctDNA from contaminated plasma DNA samples. ContaminatedFigure 4. Purification plasma DNA procedure was mixed for the with isolation AMPure of ctDNA XP magnetic from contaminated beads, incubated plasma and DNA high samples. molecular DNAContaminatedFigure bound 4. toPurification the plasma magnetic procedure DNA beads was for were mixedthe isolation isolated. with AMPureof ctDNA The pellet XPfrom magnetic contaminated containing beads, the plasma incubated high DNA molecular and samples. high DNA was washedmolecularContaminated twice DNA with plasma bound 80% to DNA ethanol,the magnetic was air-driedmixed beads with andwere AMPure theisolated. DNA XP The was magnetic pellet eluted containing beads, with 20 incubated theµL h RNaseigh molecular and free high water. MagneticDNAmolecular beads was washedDNA were bound added twice to with tothe the magnetic80% supernatant ethanol, beads air were- (0.6dried isolated. volumeand the TheDNA of supernatant),pellet was containingeluted with incubated, the 20 h µighL RNase molecular pelleted free on a water.DNA was Magnetic washed beads twice werewith 80% added ethanol, to the air supernatant-dried× and the (0. 6DNA× volume was eluted of supernatant), with 20 µL RNase incubated, free magnetic rack. Washing, drying and elution of the ctDNA was done similarly. pelletedwater. Magnetic on a magnetic beads rack. were Washing, added todrying the supernatantand elution of (0 .the6× volumectDNA was of supernatant),done similarly. incubated, Withpelleted this protocol on a magnetic we selectively rack. Washing, purified drying DNA and elution fragments of the with ctDNA a sizewas ofdone 90–250 similarly. bp from the Agilent With this protocol we selectively purified DNA fragments with a size of 90–250 bp from the DNAAgilent 1000With DNA DNA this ladder 1000protocol DNA used we ladder as selectively control used (Figure aspurified control5A). DNA (Figure To explorefragments 5A). To whether with explore a size the whether purificationof 90– 250the bppurification procedurefrom the is able toprocedureAgilent rescue DNA non-informative is able 1000 to DNA rescue ladder non plasma- informativeused DNA as con sample plasmatrol (Figure forDNA a 5 further A).sample To explore analysisfor a further whether of known analysis the driverpurification of known mutants, we processeddriverprocedure mutants, 51 is archived able we to processed rescue liquid non biopsy51-informative archived samples liquid plasma using biopsy DNA the samples purification sample using for a procedure thefurther purification analysis and determined procedureof known its efficacyanddriver by determin anmutants, analysised itswe withefficacy processed the by Agilent an51 analysisarchived Bio-Analyzer with liquid the biopsy Agilent 2100. samples Bio-Analyzer using 2100.the purification procedure andA determined its efficacy by an analysis with the Agilent Bio-BAnalyzer 2100. A B

Figure 5. CtDNA levels in rescued plasma DNA samples. (A) Using the size selection protocol small FigurefragmentsFigure 5. CtDNA 5. C intDNA a levels range levels in of rescued in90 rescued–250 bp plasma plasma from DNAthe DNA Agilentsamples. samples. high ( (AsensitivityA)) Using Using the theDNA size size ladderselection selection were protocol protocolselectively small small fragmentsseparated.fragments in a (in rangeB )a Successful range of 90–250of 90purification–250 bp bp from from from the the all Agilent Agilent 51 plasma highhigh samples sensitivity sensitivity was DNA independent DNA ladder ladder were of werestorage selectively selectively time separated.beforeseparated. (plasmaB) Successful(B) isolation.Successful purification purification from from all all 5151 plasmaplasma samples samples was was independent independent of storage of storage time time beforebefore plasma plasma isolation. isolation. The plasma DNA samples were significantly purified up to 99% ctDNA with a mean of 77% ThectDNA plasmaThe compared plasma DNA DNA to samples31% samples ctDNA were wereof the significantly significantlynon-purified purifiedpurifiedsamples (upp up < to0.0001, to 99% 99% FigurectDNA ctDNAs 5withB and with a 6mean). a Moreover, mean of 77 of% 77% ctDNApurificationctDNA compared compared was to 31%independent to 31% ctDNA ctDNA of theof the the storage non-purified non-puri timefied before samplessamples plasma ( (pp < < isolation0.0001,0.0001, Figure (Figure Figuress 5 B55 B)andB and and 6). the6Moreover,). Moreover,degree ofpurification contamination was independent with higher ofmolecular the storage DNA time before before si zeplasma selection isolation procedure (Figure (F 5igureB) and 6, theTable degree S1). purification was independent of the storage time before plasma isolation (Figure5B) and the degree Forof contamination example, plasma with preparation higher molecular of sample DNA #43 before was si doneze selection 7 days afterprocedure venipuncture (Figure 6and, Table ctDNA S1). of contaminationcontentFor example, was 4% withplasma prior higher purification. preparation molecular However, of sample DNA after #43 before wassize donesizeselection selection 7 days ctDNA after procedure content venipuncture was (Figure 94% and (Figure6 ,ctDNA Table 6). S1). For example,Outcontent of 51 was plasma samples 4% prior preparation only purification. for four of samples sample However, we #43 wer wasaftere unable donesize selectionto 7 daysreach aftera ctDNA purity venipuncture ofcontent 50%, while wasand 94% for ctDNA47 (Figure samples content6). was 4%weOut prior reachedof 51purification. samples purity only of 58%for However, four and samples higher after we size(Figure wer selectione unable6, Table ctDNAto S1reach).In contenta addition purity of was, purification50%, 94% while (Figure for efficacy 476 samples). Out was of 51 samplesindependentwe only reached for four purityof the samples UICC7 of 58% westage, and were higheras unablewe obtained (Figure to reach 6purification, Table a purity S1). Inof of 60% addition50%, for while patients, purification for 47with samples NSCLC efficacy we stage was reached purityindependent of 58% and of higher the UICC7 (Figure stage,6, Table as we S1). obtained In addition, purification purification of 60% for e ffi patientscacy was with independent NSCLC stage of the UICC7 stage, as we obtained purification of 60% for patients with NSCLC stage IA, 78% for patients with NSCLC stage IIA/B, 87% for patients with NSCLC stage IIIA/B and 76% for patients with NSCLC stage IV. Cancers 2020, 12, x 6 of 12

CancersIA, 202078%, 12for, 1917 patients with NSCLC stage IIA/B, 87% for patients with NSCLC stage IIIA/B and 76%6 for of 12 patients with NSCLC stage IV.

FigureFigure 6. Improved6. Improved sensitivity sensitivity of of oncogenic oncogenic gene gene mutationsmutations in rescued plasma plasma DNA DNA samples. samples. CtDNA CtDNA contentcontent pre pre and and post post purification purification as well as well as duration as duration to plasma to plasma preparation preparation was shown was shown for each for sample. each DNAsample. was purifiedDNA was up purified to 44–99% up to with 44– a99% mean with of a 77% mean ctDNA. of 77% ComparisonctDNA. Comparison of resequencing of resequencing analyses usinganalyses tumor using tissue tumor samples, tissue non-purified samples, non plasma-purified DNA plasma samples DNA samples (pre purified (pre purifi ctDNA)ed ctDNA and purified) and DNApurified samples DNA (post samples purified (post ctDNA) purified ofctDNA 42 patients) of 42 patients with non-small with non-small cell lung cell lung cancer cancer (NSCLC) (NSCLC with) an eight-genewith an eight panel-gene including panel includingEGFR, KRAS,EGFR, IDH1, KRAS, IDH2, IDH1, PDGFRA, IDH2, PDGFRA, NRAS, BRAFNRAS,and BRAFKIT and. Samples KIT. withSamples oncogenic with mutations oncogenic inmutationsKRAS (middle in KRAS part) (middle or EGFR part) (loweror EGFR part) (lower are part) labeled are lab green,eled allgreen, samples all withoutsamples driver without mutations driver weremutations labeled were red labeled (ND = rednot (ND documented; = not documented; N/A = not N/A available). = not available).

In caseIn case the the purification purification was was incomplete,incomplete, this this process process can can be berepeated repeated as often as often as needed as needed until the until thebest best purity purity with with a minimal a minimal loss loss of ctDNA of ctDNA is ach isieved. achieved. Moreover, Moreover, to avoid to a avoidloss of athe loss ctDNA of the fraction ctDNA fractionduring during the purification the purification step, step,the high the highmolecular molecular DNA DNAfraction fraction was also was eluted also eluted and monitored and monitored for forremaining remaining ctDNA. ctDNA. In In this this case, case, we we did did a a second second purification purification procedure procedure using using the the eluate eluate from from the the beadsbeads to avoid to avoid loss loss of ctDNA. of ctDNA. In total, In tota meanl, mean reduction reduction between between pre-purified pre-purified and post-purified and post-purified ctDNA wasctDNA 11 ng/ mLwas plasma,11 ng/m rangingL plasma, from ranging 0 to 123from ng 0/ mLto 123 plasma ng/m (TableL plasma S2). (T Afterable S2 the). After purification the purification procedure procedure there was no significant difference between samples of different subgroups (Figure 5B). there was no significant difference between samples of different subgroups (Figure5B).

2.3.2.3. Sample Sample Purification Purification Rescues Rescues Non-Informative Non-Informative Plasma Plasma DNADNA Samples for Mutational Mutational Analysis Analysis In order to determine the impact of the purification procedure on the usability of the plasma In order to determine the impact of the purification procedure on the usability of the plasma DNA samples for a mutational analysis, we subjected 42 purified ctDNA samples as well as their DNA samples for a mutational analysis, we subjected 42 purified ctDNA samples as well as their non-purified counterparts of which sequencing data from the corresponding tissue material was non-purified counterparts of which sequencing data from the corresponding tissue material was available to a targeted resequencing approach with an eight gene panel including BRAF, EGFR, IDH1, available to a targeted resequencing approach with an eight gene panel including BRAF, EGFR, IDH1, IDH2, KIT, KRAS, NRAS and PDGFRA used for molecular diagnostics in our routine laboratory work. IDH2, KIT, KRAS, NRAS PDGFRA The results were comparedand with theused mutation for molecular status obtained diagnostics with in the our corresponding routine laboratory 42 tissue work. Thesamples. results were We also compared included with the the two mutation purified status ctDNA obtained samples with, which the correspondingdid not reach the 42 tissue50% ctDNA samples. Wepurity also included (#8 and the#10), two while purified the non ctDNA-purified samples, samples which #1 didand not#30 reachwere theexcluded 50% ctDNA due to purity insufficient (#8 and #10),DNA while amounts. the non-purified samples #1 and #30 were excluded due to insufficient DNA amounts. AnalysisAnalysis of the of tissue the tissue biopsy biopsy samples samples revealed revealed 23 samples 23 samples without without detectable detectable oncogenic oncogenic mutation, fivemutation, samples five with samples an oncogenic with anEGFR oncogenicmutation EGFR and mutation 14 samples and 14 with samples an oncogenic with an oncogenicKRAS mutation. KRAS Nonemutation. of the oncogenicNone of theEGFR oncogenicmutations EGFR and mutations only six and of theonly 14 six oncogenic of the 14 KRASoncogenicmutations KRAS mutations detected in thedetected tissue biopsies in the weretissue observed biopsies inwere the observed non-purified in the ctDNA non-purified samples ctDNA (Figure samples6). By contrast, (Figure analysis6). By of thecontrast, 42 purified analysis ctDNA of the samples 42 purifi revealeded ctDNA 14 samples samples harboring revealed oncogenic 14 samples mutations, harboring with oncogenic oncogenic KRASmutations,mutations with being oncogenic the most KRAS prominent mutations ( beingn = 10), the followed most prominent by oncogenic (n = 10),EGFR followedmutations by oncogenic (n = 4). Of theEGFR total mutations 17 samples (n = 4). with Of anthe oncogenic total 17 samplesKRAS withmutation, an oncogenic only in KRAS six samples mutation, (35%) only the in mutation six samples was detectable(35%) the in mutation all three approacheswas detectable (tissue in all biopsy, three approaches non-purified (tissue ctDNA biopsy, and non purified-purified ctDNA). ctDNA For and two samplespurified (12%) ctDNA).KRAS Formutations two samples were observed (12%) KRAS intissue mutations biopsy were and observed purified ctDNA, in tissue whereas biopsy andin six samplespurified (35%) ctDNA,KRAS whereasmutation in was six detectedsamples only (35%) in KRAS the corresponding mutation was tissue detecte biopsy.d only Further, in the we observedcorresponding six samples tissue with biopsy. an oncogenic Further, we mutation observed in sixEGFR samples, of which with an three oncogenic samples mutation (50%) showed in EGFR the, of which three samples (50%) showed the same mutation in tissue biopsy and purified ctDNA. same mutation in tissue biopsy and purified ctDNA. Moreover, four oncogenic mutations in EGFR Moreover, four oncogenic mutations in EGFR (n = 1) or KRAS (n = 3) were exclusively identified in (n = 1) or KRAS (n = 3) were exclusively identified in purified ctDNA, while no oncogenic mutations were seen only in non-purified ctDNA or in a tissue biopsy and non-purified sample. When comparing the results obtained from the tissue biopsy with those of our non-purified ctDNA samples, we obtained Cancers 2020, 12, x 7 of 12

Cancerspurified2020 ,ctDNA,12, 1917 while no oncogenic mutations were seen only in non-purified ctDNA or in a 7tissue of 12 biopsy and non-purified sample. When comparing the results obtained from the tissue biopsy with those of our non-purified ctDNA samples, we obtained an overall concordance of 70%, while for the anpurified overall ctDNA concordance samples of 70%,an overall while forconcordance the purified of ctDNA71% (Figure samples 5A) an was overall achieved. concordance Regarding of 71% the (Figuremutation5A) specific was achieved. concordance, Regarding we theobserved mutation an increase specific concordance,from 32% to we79% observed after purification. an increase In from most 32%of the to 79%cases after for purification.which we detected In most the of theoncogenic cases for mutations which we in detected all three the samples, oncogenic the mutationspurification in allled threeto a samples,distinct increase the purification in the variant led to a allele distinct frequencies increase in (VAFs), the variant except allele for frequencies samples #13 (VAFs), and #34. except For forthose samples cases #13 with and a VAF #34. Forof 1% those in the cases purified with a samples, VAF of 1% the in mutation the purified in the samples, non-purified the mutation sample in thewas non-purifiednot detected sample (Figure was 7), notleading detected to a (Figuresensitivity7), leading of 58% to for a sensitivitythe purified of 58%or 33% for thefor the purified non- orpurified 33% forctDNA the non-purified samples. Collectively, ctDNA samples. our results Collectively, show that our our results purification show that protocol our purification is a helpful protocoltool to rescue is a helpfulinappropriately tool to rescue handled inappropriately ctDNA with handled lower quality. ctDNA With with lowermaterial quality. costs Withbelow material €10 per costs samples, below as €only10 per magnetic samples, beads, as only ethanol magnetic and beads,Agilent ethanol Bio-Analyzer and Agilent chip are Bio-Analyzer required, the chip purification are required, is also the a purificationcost-effective is alsomethod a cost-e to rescueffective ctDNA method from to rescue blood ctDNA samples from with blood clinical samples relevance. with clinical relevance.

FigureFigure 7. 7.Decreased Decreased allele allele frequency frequency in in non-purified non-purified ctDNA ctDNA samples. samples. Heatmap Heatmap showing showing the the variant variant alleleallele frequency frequency (%) (%) of detected of detected oncogenic oncogenic mutations mutations in purified in and purified non-purified and non ctDNA-purified in comparison ctDNA in tocomparison the detected to mutations the detected using mutations tissue biopsies using tissue (*estimated biopsies from (*estimated Sanger sequencing). from Sanger sequencing). 3. Discussion 3. Discussion Tumor heterogeneity leads to variations in the mutational spectrum at different tumor sites [6]. Tumor heterogeneity leads to variations in the mutational spectrum at different tumor sites [6]. While tissue biopsy being the gold standard in tumor genotyping, inaccessible tumor sites, While tissue biopsy being the gold standard in tumor genotyping, inaccessible tumor sites, highly highly metastatic tumors as well as tissue biopsies with insufficient tumor cell content due to metastatic tumors as well as tissue biopsies with insufficient tumor cell content due to fine needle fine needle biopsies display major problems in daily routine [22]. For those cases molecular genotyping biopsies display major problems in daily routine [22]. For those cases molecular genotyping may not may not be feasible at all [23]. Therefore, analysis of ctDNA derived from liquid biopsies is thought to be feasible at all [23]. Therefore, analysis of ctDNA derived from liquid biopsies is thought to unravel unravel the aggregation of all mutations at a metastatic tumor, thus, being an alternative to conventional the aggregation of all mutations at a metastatic tumor, thus, being an alternative to conventional biopsies [24]. However, several studies had demonstrated that liquid biopsy processing has a high biopsies [24]. However, several studies had demonstrated that liquid biopsy processing has a high impact on further approaches [13,25,26]. For blood stored in tubes containing EDTA longer than impact on further approaches [13,25,26]. For blood stored in tubes containing EDTA longer than 24 h 24 h prior plasma isolation, contamination of ctDNA with cellular DNA from dying immune cells prior plasma isolation, contamination of ctDNA with cellular DNA from dying immune cells was was observed [13]. As this is one main issue in archived blood samples with generally improper observed [13]. As this is one main issue in archived blood samples with generally improper handling, handling, improvement of these methods are mandatory to rescue these inappropriately handled improvement of these methods are mandatory to rescue these inappropriately handled samples. The samples. The impact of the improper handling of blood and plasma samples is also evident in impact of the improper handling of blood and plasma samples is also evident in our proof-of-concept our proof-of-concept study. For instance, overall plasma DNA concentrations from liquid biopsies study. For instance, overall plasma DNA concentrations from liquid biopsies stored for four days stored for four days harbor distinctively more DNA than those stored for one or two days (Figure2). harbor distinctively more DNA than those stored for one or two days (Figure 2). Furthermore, also Furthermore, also the percentage of contaminating longer DNA fragments increased with the delay in the percentage of contaminating longer DNA fragments increased with the delay in plasma plasma preparation (Figure3). This is in line with a study from Parpart-Li et al., which demonstrated preparation (Figure 3). This is in line with a study from Parpart-Li et al., which demonstrated that that storage temperature plays an important role in ctDNA quality, hence, storage temperature at 4 C storage temperature plays an important role in ctDNA quality, hence, storage temperature at ◦4°C delays contamination with cellular DNA of up to 3 days [27]. Further, Wong et al. observed not only delays contamination with cellular DNA of up to 3 days [27]. Further, Wong et al. observed not only the processing time but also shipping and storage temperature as well as physical shock to have an the processing time but also shipping and storage temperature as well as physical shock to have an effect on cell integrity in blood samples used for a prenatal diagnosis [28]. Other studies suggested effect on cell integrity in blood samples used for a prenatal diagnosis [28]. Other studies suggested that utilization of Cell-Free DNA BCT tubes (Streck, La Vista, Nebraska) or PaxGene tubes (Qiagen, that utilization of Cell-Free DNA BCT tubes (Streck, La Vista, Nebraska) or PaxGene tubes (Qiagen, Hilden, Germany) significantly decrease contamination of the ctDNA with cellular DNA [13,20] by Hilden, Germany) significantly decrease contamination of the ctDNA with cellular DNA [13,20] by preventing cell damage due to cell-preserving reagents. However, even with the usage of such specific preventing cell damage due to cell-preserving reagents. However, even with the usage of such sample tubes, lysis of peripheral blood lymphocytes (PBLs) is not completely abolished [21]. The data specific sample tubes, lysis of peripheral blood lymphocytes (PBLs) is not completely abolished [21]. presented here clearly show that our purification procedure is capable of rescuing such contaminated The data presented here clearly show that our purification procedure is capable of rescuing such plasma DNA samples. For instance, we enriched ctDNA from 31% to a mean purity of 77%, which

Cancers 2020, 12, x 8 of 12

contaminated plasma DNA samples. For instance, we enriched ctDNA from 31% to a mean purity of Cancers77%, which2020, 12 ,could 1917 be subsequently used for next-generation sequencing approaches leading 8to of an 12 increase in the mutation specific concordance from 32% to 79% after purification. Further, when comparing the tested non-purified and purified ctDNA samples with oncogenic mutations we coulddemonstrate be subsequentlyd an increase used in for sensitivity next-generation from 33 sequencing% with non- approachespurified to leading58% with to purified an increase samples. in the mutationHowever, specific despite concordance successful enrichment from 32% toof 79%ctDNA, after not purification. all oncogenic Further, mutations when already comparing seen the in testedtissue non-purifiedbiopsy were and observed purified in ctDNA purified samples plasma with DNA oncogenic samples. mutations The reason we demonstrated for this deficiency an increase remains in sensitivityunclear, but from it might 33% be with explained non-purified by a very to 58% low with abundance purified of samples. the corresponding However, despitectDNA successfulmolecules enrichmentin the blood of or ctDNA, very sm notall all amounts oncogenic of mutations plasma DNA already samples. seen in By tissue contrast, biopsy the were additional observed four in purifiedoncogenic plasma mutations DNA samples.exclusively The seen reason with for the this purified deficiency ctDNA remains are unclear, most likely but it derived might be from explained other bytumor a very sites low in abundance the patients of the and corresponding therefore underscore ctDNA molecules the general in the potential blood or very of liquid small biopsieamountss ofin plasmaanalyzing DNA tumor samples. heterogeneity By contrast,, as one the single additional biopsy four with oncogenic a small sample mutations size exclusively cannot reveal seen the with whole the purifiedmutational ctDNA landscape are most of likely a tumor. derived Therefore, from other our tumor purification sites in the procedure patients and is able therefore to improve underscore the theoutcome general of potentialinappropriately of liquid handled biopsies plasma in analyzing samples tumor for molecular heterogeneity, genotyping as one single in a time biopsy and with cost a- smalleffective sample manner size. cannotHowever reveal, as thewe whole observed mutational a partial landscape loss of plasma of a tumor. DNA Therefore, during the our purification purification procedureprocedure, iswe able recommend to improve using the outcome the purification of inappropriately procedure only handled for cases plasma with samples less than for 50% molecular in the genotypingfragment size in of a timectDNA and (F cost-eigure ff8,ective Table manner.S2). However, as we observed a partial loss of plasma DNATaken during together, the purification we presented procedure, here we a recommenduseful procedure using theto rescue purification non-informative procedure only ctDNA for casessamples. with less than 50% in the fragment size of ctDNA (Figure8, Table S2).

Figure 8. Decision tree for the application of the plasma DNA purificationpurification protocol. For low quality ctDNA a purificationpurification procedure is recommended. A p purificationurification procedure can be repeated until the best possible result is achieved.

4. MaterialsTaken together, and Methods we presented here a useful procedure to rescue non-informative ctDNA samples.

4.4.1. Materials Patients and Methods 4.1. PatientsFrom May 2014 to July 2017, 600 patients with NSCLC were enrolled in the LuCaBiO study (Lung Cancer and Biological Outcome, NCT02613637). All patients participating were required to From May 2014 to July 2017, 600 patients with NSCLC were enrolled in the LuCaBiO study (Lung Cancer and Biological Outcome, NCT02613637). All patients participating were required to meet Cancers 2020, 12, 1917 9 of 12 the criteria: (a) the patients had to be diagnosed with NSCLC adenocarcinoma, (b) all patients accepted biopsy with sufficient tumor tissue to detect genetic mutations and (c) patients provided sufficient plasma for genetic detection. Finally, 51 patients meeting the criteria were selected for ctDNA analysis, 37% patients were female and mean age was 63.3 years. For all patients, analyses of tumor tissue were performed according to the IASLC UICC TNM (7th edition) classification. Analyses include the morphology of tumor cells, immunophenotyping and molecular genetic studies. Biobanking of patient tumor tissue was organized in the local pathology units. The study was approved by the hospital’s ethics committee (ethic code 371/13) and all patients gave signed informed consent.

4.2. Human Tissue Biopsy Human tissue biopsies used in the current study were collected and stored by the Institute of Pathology of the University Medical Centre Ulm. Tissue biopsies were fixed in formalin and further paraffin embedded and stored at room temperature until use. Pathologists assessed all samples before use. Extraction of FFPE DNA was performed using the Qiagen FFPE DNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. For each patient we used one to three tissue sections (5 µm) for DNA extraction depending on the tumor size. DNA was recovered in 25 µL of elution buffer and stored at 20 C until further use. − ◦ 4.3. Plasma Collection Whole venous blood (7.5 mL) was collected in K2EDTA tubes (Sarstedt, Nümbrecht, Germany) by peripheral blood withdrawal. Plasma separation was carried out within up to 7 days after collection. Whole blood was centrifuged for 10 min (2500 g at room temperature) and the plasma fraction was × recovered in two fresh 2 mL tubes for immediate storage at 80 C until ctDNA isolation. − ◦ 4.4. Extraction of Circulating Cell-Free DNA Circulating cell-free DNA was extracted from plasma using the QIAamp Circulating Nucleic Acid Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. For each patient we used 2 mL of plasma for ctDNA extraction and recovered ctDNA in 25 µL of elution buffer. DNA was stored at 20 C until further use. − ◦ 4.5. Quantification of FFPE and Cell-Free DNA The total amount of DNA was determined by fluorometric measurement using Qubit 3.0 Fluorometer (ThermoFisher Scientific, Waltham, MA, USA). We used 1 µL of DNA eluate gained through the QIAamp Circulating Nucleic Acid Kit (Hilden, Germany) and measured the concentration using the Qubit dsDNA HS Assay Kit (Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions. Fragment size of DNA was determined by electrophoresis using Agilent Bio-Analyzer 2100. We used 1 µL of DNA eluate and measured the fragment size distribution using the Agilent High Sensitivity Kit (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instructions.

4.6. DNA Purification Procedure DNA size selection was done using Agencourt AMPure XP magnetic beads (Beckman Coulter, Brea, CA, USA) to purify ctDNA with a fragment length of 90–250 bp, according to the manufacturer’s instructions. 25 µL of plasma DNA was mixed with 1.2 AMPure XP magnetic beads and incubated × for 5 min at room temperature. After 5 min of incubation at the magnetic rack, the supernatant was convicted into a new tube. Beads containing longer fragments ( 1000 bp) were washed twice with ≥ 80% ethanol, air-dried and longer fragments were eluted with 20 µL RNase free water. Meanwhile, 0.6 AMPure XP beads were added to the supernatant and incubated for 5 min at room temperature. × After additional incubation for 5 min on the magnetic rack, the supernatant was discarded and beads Cancers 2020, 12, 1917 10 of 12 were washed twice with 80% ethanol and air-dried. Purified ctDNA was eluted with 20 µL RNase free water. The quality of purified ctDNA was checked using Agilent Bio-Analyzer High Sensitivity DNA Kit (Santa Clara, CA, USA).

4.7. Next Generation Sequencing and Data Analysis FFPE tissue DNA, non-purified and purified ctDNA were further used for amplicon sequencing. For each library up to 20 ng DNA/ctDNA were used depending on the amount of available DNA/ctDNA (min = 1.1 ng; max = 20 ng; mean =10.9 ng). The library was prepared using the Tumor Actionable Mutations Gene Read kit (Qiagen, Hilden, Germany), which allows the sequencing of BRAF, EGFR, IDH1, IDH2, KIT, KRAS, NRAS and PDGFRA gene segments. Prepared libraries were sequenced by targeted next-generation sequencing on a MiSeq platform (Illumina, San Diego, CA, USA) with paired-end 150-base pair reads (approximately 5000 coverage). Fastq files were uploaded to Qiagen × CLC Biomedical Workbench V5.2 (Hilden, Germany), samples were analyzed using a ready-to-use workflow “Identify and Add Variants” and reads were mapped against Human Genome Build 19 (hg19) as the reference. Additionally, the Integrative Genomics Viewer (IGV) was used to visualize variants.

4.8. Statistical Analysis For the calculation of the concordance, the same mutations detected in both matched FFPE tumor and ctDNA samples were classified as true positives; true negatives were identified as those where both matched tumor and ctDNA samples had no mutations; mutations identified in ctDNA, which were not found in tumor tissue DNA were classified as false positives and mutations identified in tumor tissue DNA but not in ctDNA were classified as false negatives. The overall concordance rate was defined as the ratio of the sum of the number of true positives and true negatives to the total enrolled patients. Concordance between tissue mutated samples and purified ctDNA was calculated as the ratio of the number of mutations in the tissue biopsy to purified ctDNA. Sensitivity rate was defined as the ratio of the true positives to the sum of true positives and false negatives. Significance was determined using a Student’s t test, with a p-value of <0.05 showing statistical significance.

5. Conclusions This procedure is easy to perform, time and cost effective and leads to purified ctDNA, which is sufficient to detect oncogenic mutations in NGS approaches. Moreover, the efficacy of our purification procedure is independent of the delay in plasma DNA preparation and degree of contamination with longer DNA fragments making it a valuable technical improvement in all laboratories handling liquid biopsy samples.

Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6694/12/7/1917/s1, Table S1: percentage of ctDNA content, Table S2: ctDNA content in ng/mL plasma. Author Contributions: Conceptualization, R.M. and S.M.; formal analysis, T.F.E.B., P.M.; investigation, S.M.; resources, G.S.-S., F.G.R., C.B.; data curation, G.S.-S., F.G.R., C.B.; writing—original draft preparation, S.M., R.M.; writing—review and editing, R.M., T.F.E.B. and P.M.; project administration, P.M.; funding acquisition, R.M. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by German Research Foundation (DFG), GRK 2254 HEIST. Acknowledgments: We thank Karola Dorsch and Elena Moser for excellent technical assistance. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Cancers 2020, 12, 1917 11 of 12

References

1. Testa, U.; Castelli, G.; Pelosi, E. Lung Cancers: Molecular Characterization, Clonal Heterogeneity and Evolution, and Cancer Stem Cells. Cancers (Basel) 2018, 10, 248. [CrossRef] 2. Torre, L.A.; Bray, F.; Siegel, R.L.; Ferlay, J.; Lortet-Tieulent, J.; Jemal, A. Global cancer statistics, 2012. CA Cancer J. Clin. 2015, 65, 87–108. [CrossRef] 3. Ferlay, J.; Shin, H.R.; Bray, F.; Forman, D.; Mathers, C.; Parkin, D.M. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int. J. Cancer. 2010, 127, 2893–2917. [CrossRef] 4. Sorber, L.; Zwaenepoel, K.; Deschoolmeester, V.; Van Schil, P.E.; Van Meerbeeck, J.; Lardon, F.; Rolfo, C.; Pauwels, P. Circulating cell-free nucleic acids and platelets as a liquid biopsy in the provision of personalized therapy for lung cancerpatients. Lung Cancer 2017, 107, 100–107. [CrossRef] 5. Schmitt, M.W.; Loeb, L.A.; Salk, J.J. The influence of subclonal resistance mutations on targeted cancer therapy. Nat. Rev. Clin. Oncol. 2016, 13, 335–347. [CrossRef] 6. Jamal-Hanjani, M.; Wilson, G.A.; McGranahan, N.; Birkbak, N.J.; Watkins, T.; Veeriah, S.; Shafi, S.; Johnson, D.H.; Mitter, R.; Rosenthal, R.; et al. Tracking the Evolution of Non-Small-Cell Lung Cancer. N. Engl. J. Med. 2017, 376, 2109–2121. [CrossRef] 7. Amirouchene-Angelozzi, N.; Swanton, C.; Bardelli, A. Tumor Evolution as a Therapeutic Target. Cancer Discov. 2017, 7, 1–13. [CrossRef] 8. Sozzi, G.; Conte, D.; Leon, M.; Ciricione, R.; Roz, L.; Ratcliffe, C.; Roz, E.; Cirenei, N.; Bellomi, M.; Pelosi, G.; et al. Quantification of free circulating DNA as a diagnosticmarker in lung cancer. J. Clin. Oncol. 2003, 21, 3902–3908. [CrossRef] 9. Huang, Z.H.; Li, L.H.; Hua, D. Quantitative analysis of plasma circulating DNA at diagnosis and during follow-up of breast cancer patients. Cancer Lett. 2006, 243, 64–70. [CrossRef] 10. Underhill, H.R.; Kitzman, J.O.; Hellwig, S.; Welker, N.C.; Daza, R.; Baker, D.N.; Gligorich, K.M.; Rostomily, R.C.; Bronner, M.P.; Shendure, J. Fragment Length of Circulating Tumor DNA. PLoS Genet. 2016, 12, e1006162. [CrossRef] 11. Heitzer, E.; Ulz, P.; Geigl, J.B. Circulating tumor DNA as a liquid biopsy for cancer. Clin. Chem. 2015, 61, 112–123. [CrossRef][PubMed] 12. Park, C.K.; Cho, H.J.; Choi, Y.D.; Oh, I.J.; Kim, Y.C. A Phase II Trial of Osimertinib in the Second-Line Treatment of Non-small Cell Lung Cancer with the EGFR T790M Mutation, Detected from Circulating Tumor DNA: LiquidLung-O-Cohort 2. Cancer Res. Treat. 2019, 51, 777–787. [CrossRef] 13. Risberg, B.; Tsui, D.; Biggs, H.; Ruiz-Valdepenas Martin de Almagro, A.; Dawson, S.J.; Hodgkin, C.; Jones, L.; Parkinson, C.; Piskorz, A.; Marass, F.; et al. Effects of Collection and Processing Procedures on Plasma Circulating Cell-Free DNA from Cancer Patients. J. Mol. Diagn. 2018, 20, 883–892. [CrossRef][PubMed] 14. Page, K.; Guttery, D.S.; Zahra, N.; Primrose, L.; Elshaw, S.R.; Pringle, J.H.; Blighe, K.; Marchese, S.D.; Hills, A.; Woodley, L.; et al. Influence of plasma processing on recovery and analysis of circulating nucleic acids. PLoS ONE 2013, 8, e77963. [CrossRef] 15. Norton, S.E.; Luna, K.K.; Lechner, J.M.; Qin, J.; Fernando, M.R. A new blood collection device minimizes cellular DNA release during sample storage and shipping when compared to a standard device. J. Clin. Lab. Anal. 2013, 27, 305–311. [CrossRef][PubMed] 16. Hu, Y.; Ulrich, B.C.; Supplee, J.; Kuang, Y.; Lizotte, P.H.; Feeney, N.B.; Guibert, N.M.; Awad, M.M.; Wong, K.K.; Jänne, P.A.; et al. False-Positive Plasma Genotyping Due to Clonal Hematopoiesis. Clin. Cancer Res. 2018, 24, 4437–4443. [CrossRef] 17. Sorber, L.; Zwaenepoel, K.; Jacobs, J.; De Winne, K.; Goethals, S.; Reclusa, P.; Van Casteren, K.; Augustus, E.; Lardon, F.; Roeyen, G.; et al. Circulating Cell-Free DNA and RNA Analysis as Liquid Biopsy: Optimal Centrifugation Protocol. Cancers (Basel) 2019, 11, 458. [CrossRef] 18. Zhao, Y.; Li, Y.; Chen, P.; Li, S.; Luo, J.; Xia, H. Performance comparison of blood collection tubes as liquid biopsy storage system for minimizing cfDNA contamination from genomic DNA. J. Clin. Lab. Anal. 2019, 33, e22670. [CrossRef][PubMed] 19. Medina Diaz, I.; Nocon, A.; Mehnert, D.H.; Fredebohm, J.; Diehl, F.; Holtrup, F. Performance of Streckcf DNA Blood Collection Tubes for Liquid Biopsy Testing. PLoS ONE 2016, 11, e0166354. [CrossRef] Cancers 2020, 12, 1917 12 of 12

20. Warton, K.; Yuwono, N.L.; Cowley, M.J.; McCabe, M.J.; So, A.; Ford, C.E. Evaluation of Streck BCT and PAX gene Stabilised Blood Collection Tubes for Cell-Free Circulating DNA Studies in Plasma. Mol. Diagn. Ther. 2017, 21, 563–570. [CrossRef][PubMed] 21. Parackal, S.; Zou, D.; Day, R.; Black, M.; Guilford, P. Comparison of Roche Cell-Free DNA collection TubestoStreck Cell-Free DNA BCTs for sample stability using healthy volunteers. Pract. Lab. Med. 2019, 16, e00125. [CrossRef][PubMed] 22. Bubendorf, L.; Lantuejoul, S.; de Langen, A.J.; Thunnissen, E. Nonsmall cell lung carcinoma: Diagnostic difficulties in small biopsies and cytological specimens: Number 2 in the Series ”Pathology for the clinician” Edited by Peter Dorfmüller and Alberto Cavazza. Eur. Respir. Rev. 2017, 26, 170007. [CrossRef][PubMed] 23. Lim, C.; Tsao, M.S.; Le, L.W.; Shepherd, F.A.; Feld, R.; Burkes, R.L.; Liu, G.; Kamel-Reid, S.; Hwang, D.; Tanguay, J.; et al. Biomarker testing and time to treatment decision in patients with advanced nonsmall-cell lung cancer. Ann. Oncol. 2015, 26, 1415–1421. [CrossRef] 24. Herbreteau, G.; Vallée, A.; Charpentier, S.; Normanno, N.; Hofman, P.; Denis, M.G. Circulating free tumor DNA in non-small cell lung cancer (NSCLC): Clinical application and future perspectives. J. Thorac. Dis. 2019, 11, 113–126. [CrossRef] 25. El Messaoudi, S.; Rolet, F.; Mouliere, F.; Thierry, A.R. Circulating cell free DNA: Preanalytical considerations. Clin. Chim. Acta. 2013, 424, 222–230. [CrossRef][PubMed] 26. Rothwell, D.G.; Smith, N.; Morris, D.; Leong, H.S.; Li, Y.; Hollebecque, A.; Ayub, M.; Carter, L.; Antonello, J.; Franklin, L.; et al. Genetic profiling of tumours using both circulating free DNA and circulating tumour cells isolated from the same preserved whole blood sample. Mol. Oncol. 2016, 10, 566–574. [CrossRef] 27. Parpart-Li, S.; Bartlett, B.; Popoli, M.; Adleff, V.; Tucker, L.; Steinberg, R.; Georgiadis, A.; Phallen, J.; Brahmer, J.; Azad, N.; et al. The Effect of Preservative and Temperature on the Analysis of Circulating Tumor DNA. Clin. Cancer Res. 2017, 23, 2471–2477. [CrossRef] 28. Wong, D.; Moturi, S.; Angkachatchai, V.; Mueller, R.; DeSantis, G.; van den Boom, D.; Ehrich, M. Optimizing blood collection, transport and storage conditions for cell free DNA increases access to prenatal testing. Clin. Bio. Chem. 2013, 46, 1099–1104. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).