Assessment of circulating biomarkers for detection of lung cancer in a high-risk cohort

Abstract

BACKGROUND:

There is an urgent need for early detection of lung cancer. Screening with low-dose computed tomography (LDCT) is now implemented in the US. Supplementary use of a lung cancer biomarker with high specificity is desirable.

OBJECTIVE:

To assess the diagnostic properties of a biomarker panel consisting of cytokeratin 19 fragment (CYFRA 21-1), carcinoembryonic antigen (CEA) and cancer antigen 125 (CA125).

METHODS:

A cohort of 250 high-risk patients was investigated on suspicion of lung cancer. Ahead of diagnostic work-up, blood samples taken. Cross-validated prediction models were computed to assess lung cancer detection properties.

RESULTS:

In total 32% (79/250) of patients were diagnosed with lung cancer. Area under the curve (AUC) for the three biomarkers was of 0.795, with sensitivity/specificity of 57%/93% and negative predictive value of 83%. When combining the biomarkers with US screening criteria, the AUC was 0.809, while applying only US screening criteria on the cohort, yielded an AUC of 0.62. The ability of the biomarkers to detect stage I-II lung cancer was substantially lower; AUC 0.54.

CONCLUSIONS:

In a high-risk cohort, the detection properties of the three biomarkers were acceptable compared to current LDCT screening criteria. However, the ability to detect early stage lung cancer was low.

Keywords

Lung cancer screening biomarkers cytokeratin 19 fragment (CYFRA 21-1)carcinoembryonic antigen (CEA)cancer antigen 125 (CA125)

1. Introduction

Lung cancer remains the leading cause of cancer-related death worldwide [1]. The global incidence of lung cancer continues to increase due to aging, world population growth and smoking [2]. Lung cancer stage at diagnosis has a high impact on prognosis [3] and significant efforts to increase the detection of early disease and reduce mortality have been undertaken [4].

Following the results of the US National Lung Screening Trial (NLST) [5], the United States Preventive Services Task Force (USPSTF) has recommended annual low-dose computed tomography (LDCT) for people at high risk of lung cancer [6]. In 2021 the USPSTF changed the screening criteria and now recommends annual LDCT in adults aged 50 to 80 years who have a 20 pack-year smoking history and currently smoke or have quit smoking within the past 15 years [7]. European lung cancer screening trials have also been conducted [8, 9, 10, 11], and several European countries are currently in the process of implementing LDCT lung cancer screening [12].

Although LDCT screening proved beneficial in the NLST with reduced overall mortality, the largest European screening trial (NELSON) [10] has only found reduced lung cancer-specific mortality. Furthermore, issues such as overdiagnosis, false positives, nodule management and complications cannot be ignored [13]. Therefore, it has been suggested to add a blood biomarker panel to either screening criteria or as part of the screening result to improve discrimination between malignant or benign indeterminate nodules [14].

Currently, several types of cancer biomarkers are being investigated [15]. The Scottish ECLS trial [16] used a positive autoantibody as part of the inclusion criteria of a two-year biannually LDCT screening trial program. The intervention resulted in a significant diagnostic stage shift. However, the control group received usual clinical care, and thus, the incremental contribution of the autoantibody test to LDCT screening cannot be assessed. Other potential lung cancer biomarkers that are yet to be tested in conjunction with LDCT screening include several types of protein panels and circulating tumor DNA (ctDNA) [15].

In this study, the potential utilization of a 3-biomarker panel consisting of cytokeratin 19 fragment (CYFRA 21-1), carcinoembryonic antigen (CEA) and cancer antigen 125 (CA125) was tested in a high-risk cohort. These three biomarkers have been tested previously in conjunction with surfactant protein B [19], showing promises results regarding optimization of inclusion criteria in LDCT. However, surfactant protein B is not commercially available and, hence, the study aimed to assess the diagnostic properties without surfactant protein B. The primary aim of the study was to assess the ability of the biomarker panel to detect lung cancer at any stage. The secondary aims included assessing the detection of stage I-II lung cancer using the biomarker panel, evaluating the performance of the US LDCT screening criteria if applied on the cohort and the performance of a combination of the biomarker panel with current US LDCT screening criteria or tobacco pack years.

2. Materials and methods

2.1 Inclusion and exclusion

At the Department of Internal Medicine, Lillebaelt Hospital Vejle, Denmark, a total of 250 patients were consecutively included in the study. They were all referred from their family doctor on suspicion for lung cancer, aged 18 years or above and able to participate in diagnostic procedures. Patients with a previous cancer diagnosis (except non-melanoma skin cancer) were excluded from the study. Lung cancer diagnosis and staging were obtained using the IASLC 8 ${}^{\text{th}}$ edition [17]. Patients where lung cancer was ruled out after diagnostic work-up, served as controls.

2.2 Biomarker analysis

Prior to diagnostic work-up, blood samples were taken in BD Vacutainer ${}^{\text{\textregistered}}$ Clot Activator Tubes (BD Biosciences, San Jose, CA, USA). After 30 minutes, the samples were centrifuged at 2000xg for 10 minutes and serum was collected and stored at $-$ 80 ${}^{\circ}$ C until use. The concentration of CA125, CEA, and CYFRA 21-1 was assessed in the serum using the Human Circulating Cancer Biomarker Magnetic Bead Panel I (Merck, Darmstadt, Germany) on a Bio-Plex 200 ${}^{\text{\textregistered}}$ analyzer (Bio-Rad Laboratories, Inc., Hercules, CA, USA) following the manufacturer’s recommendations. After thawing, the samples were mixed, centrifuged and diluted 1:6 using the Serum Matrix reagent. The three types of beads were prepared by sonication and vortexing after which 150 uL of each were mixed and the solution was brought to a final volume of 3 mL using the Bead Diluent reagent. After preparing plates according to instructions, 25 uL of the diluted sample and 25 uL of the bead solution was added to the wells and incubated overnight at 4 ${}^{\circ}$ C. Plates were washed three times in the Wash Buffer Reagent and added 25 uL Detection Antibodies. After one hour of incubation, 25 uL Streptavidin-Phycoerythrin was added followed by 30 minutes of incubation. Plates were washed 3 times in Wash Buffer, added 100 uL Sheath Fluid reagent and analyzed. Four samples, two controls included in the kit and two serum samples from lung cancer patients, all served as quality controls and were analyzed in either duplicates or triplicates in each run. Intra assay variation for CA125, CEA and CYFRA 21-1 were 15%, 13% and 13% respectively, and the intra assay variations were 12%, 14% and 15%, respectively.

2.3 Statistics

Normally distributed values are presented as means and standard deviation, while non-normally distributed values are described as median and IQR. Categorical values are presented as frequencies and percentages. Cross-validated logistic regression prediction models were computed using K-fold cross-validation technique (5-fold) in R, Caret package [18]. Results are presented as receiver operating characteristics (ROC) curves with area under the curve (AUC) and confidence intervals (CI). $P$ values $<$ 0.05 are considered statistically significant. All statistical analyses were calculated using R (R Core Team, 2014).

2.4 Ethical considerations

The study was observational and diagnostic work-up or treatment of a potential lung cancer were not influenced by participation in the study. The study was approved by the Danish Data Protection Agency (18/33058) and the Regional Committee on Health Research Ethics in Southern Denmark (S-20180052). All participants gave informed consent to participate. The study conforms with The Code of Ethics of the World Medical Association (Declaration of Helsinki), printed in the British Medical Journal (18 July 1964).

3. Results

3.1 Lung cancer diagnosis

Table 1
Participants’ characteristics

	Lung cancer	Control
$n$	79	171
Sex (male/female)	40/39	88/83
Age	69 (9)	65 (12)
Pack years	35 [19–48]	15 [0–40]
Ever smoker	72 (91%)	120 (70%) ${}^{\scriptscriptstyle\#}$
Never smoker	7 (9%)	49 (29%) ${}^{\scriptscriptstyle\#}$
Stage I-II/III-IV	29 (37%)/50 (63%)

Presented as frequencies (%), mean (standard deviation) or median [IQR]. ${}^{\scriptscriptstyle\#}$ Smoking status was missing for two subjects.

Of the 250 included patients, 79 (32%) were diagnosed with lung cancer. Background information on patients is given in Table 1. Age and sex distribution did not differ substantially between groups while lung cancer patients more often were smokers compared to controls ( $p<$ 0.05). Among lung cancer patients, 37% were diagnosed in stage I-II vs. 63% in stage III-IV.

3.2 Biomarker prediction model

Median concentrations of the markers in lung cancer patients vs. controls were: CA125: 10 kU/L vs. 6 kU/L; CEA 1837: $\mu$ g/L vs. 601 $\mu$ g/L; CYFRA 21-1: 1341 ng/L vs. 0.1 ng/L.

The lung cancer detection abilities of the three potential biomarkers individually and a cross-validated biomarker model containing all three markers are presented as ROC curves in Fig. 1. The AUC values of the three biomarkers are presented in Table 2.

Table 2
Biomarker detection abilities

Biomarker	AUC	95% CI	Odds ratio per 1000 unit increase	95% CI
CA125	0.685	[0.609–0.761]	1.961	0.006–743.4
CYFRA 21-1	0.686	[0.616–0.756]	1.143	1.046–1.273
CEA	0.752	[0.680–0.824]	1.555	1.271–1.996

AUC: Area under the curve; 95% CI: 95% confidence interval; CA125: Cancer antigen 125; CYFRA 21-1: Cytokeratin 19 fragment; CEA: carcinoembryonic antigen.

Figure 1.

AUC of ROC curves: CA125: 0.685, CI 0.609–0.761 (red); CYFRA 21-1: 0.686, CI 0.616–0.756 (green); CEA: 0.752, CI 0.680–0.824 (gray); biomarker panel: 0.795, CI 0.729–0.861 (blue).

The cross-validated biomarker prediction model yielded a sensitivity and specificity of 57% and 93%, respectively. The negative predictive value was 83%.

3.3 Prediction models combining clinical parameters and biomarkers

Table 3
Prediction models detecting stage I-II or III-IV lung cancer

Stage I-II

Stage III-IV

Biomarker model

0.54 [0.44-0.63]

0.88 [0.83–0.94]

Biomarker

+

tobacco

pack years model

0.66 [0.55-0.76]

0.86 [0.79–0.93]

Biomarker

+

USPSTF

screening criteria

0.67 [0.58–0.76]

0.87 [0.81–0.93]

USPSTF screening criteria

0.62 [0.51–0.72]

0.61 [0.53–0.69]

Figure 2.

AUC of ROC curves: USPSTF LDCT screening criteria: 0.624, CI: 0.550–697 (red); combination of biomarkers and tobacco pack years: 0.796, CI: 0.731–0.862 (blue); combination of biomarkers and USPSTF LDCT screening criteria: 0.809, CI: 0.750–0.868 (grey).

If applying the current USPSTF LDCT screening criteria on the cohort, 28 patients (35%) with lung cancer would not be eligible for screening. Figure 2 depicts the ROC curves of cross-validated prediction models containing clinical parameters with/without a combination of the three measured biomarkers. The current USPSTF LDCT screening criteria on the cohort had an AUC of 0.624 (CI: 0.550–697). When combining the three biomarkers and tobacco pack years the AUC was 0.796 (CI: 0.731–0.862). Moreover, for the combination of the biomarkers and the current USPSTF LDCT screening criteria the AUC was 0.809 (CI: 0.750–0.868).

Table 3 shows the performance of cross-validated prediction models in the detection of stage I-II and stage III-IV lung cancer. The models containing the measured biomarkers are remarkably better at detecting stage III-IV lung cancer than stage I-II lung cancer. However, when applying USPSTF LDCT screening criteria, the detection of stage I-II and stage III-IV lung cancer are equally low with AUC values of 0.62 and 0.61, respectively.

4. Discussion

This study evaluated the use of the three biomarkers CYFRA 21-1, CEA and CA125 in detection of lung cancer in a high-risk cohort with a lung cancer prevalence of 32%. Using the three serum biomarkers in a cross-validated prediction model, the AUC of the ROC curve was 0.795 and the negative predictive value (NPV) 83%. However, if the biomarker model was to detect stage I-II lung cancer, the AUC of the biomarker prediction model was only 0.54. A combination of CYFRA 21-1, CEA and CA125 with current USPSTF LDCT screening criteria yielded an AUC of 0.809, a value quite similar to what was obtained using the biomarkers alone.

The diagnostic properties of CYFRA 21-1, CEA and CA125 have previously been investigated in case-control studies in conjunction with other biomarkers. With the addition of surfactant protein B (Pro-SFTPB) and smoking history, a validation study yielded an AUC of 0.83 [19]. The present study aimed at testing the diagnostic properties of the three commercially available biomarkers CYFRA 21-1, CEA and CA125 in a different setting. As observed, this study yields quite similar results (AUC 0.81) using only three biomarkers in combination with age and smoking history.

When assessing the individual properties of serum biomarkers in other populations with lung cancer and healthy controls, AUC values of CEA and CA125 have provided quite similar results [20, 21]. In contrast, the diagnostic properties of CYFRA 21-1 have both been described as substantially better with an AUC between 0.85–0.87 [20] and worse with an AUC of 0.55–0.68 [21]. These differences may very well arise from differences in the study populations in terms of smoking history, study design, pathological lung cancer type and lung cancer stage distribution. In this context, the strength of the current study is the prospective cohort design, resulting in a lung cancer stage distribution as seen in clinical practice.

Using the biomarker prediction model, this study found a remarkable difference in detection properties of stage I-II lung cancer vs. stage III-IV lung cancer (AUC 0.54 vs. 0.88). This is in accordance with previous studies, although they have been using other statistical measures. Molina et al. found that at least one in five biomarkers (CEA, CA125, CYFRA 21-1, squamous cell carcinoma antigen, neuron-specific enalase (NSE)) was abnormally high in 87% of patients with locoregional disease, while this was the case in 100% of patients with metastases [22]. For this reason, biomarkers such as these have been proposed as a tool for distant metastasis detection in lung cancer; either using CYFRA 21-1 as a single biomarker [23] or a biomarker panel consisting of CEA, CA125, CA199, CA153, CYFRA 21-1 and NSE [24]. However, in countries where positron emission tomography is standard in lung cancer staging, the incremental contribution of such a panel is doubtful. Naturally, this also means that the usefulness of these biomarkers in lung cancer screening is limited, since the obvious goal of lung cancer screening programs is detection of early stage resectable lung cancer. Compared to previous studies, the advantages of the current study are the cross-validated prediction model analyses of stage distribution and the prospective cohort design giving rise to clinically realistic outcomes.

Currently, an LDCT screening program is implemented in the US. The individual LDCT scans have a high sensitivity and specificity, and the negative predictive value of the NLST was 99.9% [5]. However, the inclusion criteria for lung cancer screening mean, that a large portion of patients with lung cancer will not be offered screening [25]. In the study cohort, 35% of lung cancer patients would not have been eligible for lung cancer screening, and the corresponding AUC value – if inclusion criteria were to predict lung cancer, is poor at 0.62. Furthermore, the screening program yields a high number of false positive CT scans with indeterminate nodules [26], resulting in invasive diagnostic procedures which are associated with potential risks. The addition of a highly specific biomarker panel with a corresponding high NPV to inclusion criteria of age and smoking history, would be highly desirable. This would potentially minimize the number of false positive and true negative LDCT scans. However, the current panel of CYFRA 21-1, CEA and CA125 with an NPV of 83% and inferior detection abilities of stage I-II lung cancer does not seem to provide this advantage.

4.1 Limitations

The cohort in this study, is formed solely on the suspicion of lung cancer by the family doctor. Hence, the baseline status regarding smoking history in lung cancer patients and controls are not equal. Furthermore, the results cannot be compared to LDCT screening trials since this is a cross-sectional prediction study, as compared to a 3-year clinical trial (NLST). Furthermore, the prevalence of lung cancer in the current study is 32%, while lung cancer incidence in NLST was around 600 cases per 100,000 person-years (NLST). Also, strict comparisons of biomarker panel prediction models and screening criteria are not possible but performed here only to display the short-comings of LDCT screening criteria.

5. Conclusions

In conclusion, in this high-risk cohort, the combination of CYFRA 21-1, CEA and CA125 is able to predict lung cancer with a sensitivity of 57% and specificity of 93%. However, the biomarker panel is best for detection of late-stage lung cancer, making it less suitable for screening purposes. While the individual CT scans in LDCT screenings programs have excellent negative predictive values, the screening inclusion criteria still leaves out a large portion of lung cancer patients; in the current cohort 35% of patients would not be eligible for screening. Hence, there is still room for improvement and a biomarker panel with a high negative predictive value could potentially be used a screening tool before LDCT. Based on the results of the current study, the combination of CYFRA 21-1, CEA and CA125 does not hold the sufficient diagnostic properties for this purpose.

Footnotes

Acknowledgments

The authors would like to thank Kristina Kock Hansen for identifying potential participants and administrating the database, Anne-Mette Gintberg, Marianne Mikkelsen, Rikke Maria Iversen and Marianne Kammer for enrolling the participants. Karin Larsen for administrating funding and ethics board approval, Nilosa Ushanthan, Sara Egsgaard and Camilla Davidsen for performing the laboratory analyses and Sandra Esperanza Rubio-Rask for entering data into the database. This work was supported by grants from Region of Southern Denmark [grant number: J.nr. 20/14276, Efond 481]; Gangstedfonden [grant number A35818], and Lillebaelt Hospital Research Foundation. The sponsors had no role in study design, in the collection, analysis and interpretation of data, in the writing of the report, and in the decision to submit the article for publication.

Author contributions

Conception: Morten Borg, Torben Frøstrup Hansen, Sara W.C. Wen, Anders Jakobsen, Rikke Fredslund Andersen, Line Nederby.

Interpretation or analysis of data: Morten Borg, Ole Hilberg, Ulla Møller Weinreich.

Preparation of the manuscript: Morten Borg.

Revision for important intellectual content: Morten Borg, Torben Frøstrup Hansen, Sara W.C. Wen, Anders Jakobsen, Rikke Fredslund Andersen, Ole Hilberg, Ulla Møller Weinreich, Line Nederby.

Supervision: Ole Hilberg, Ulla Møller Weinreich, Line Nederby.

References

Mattiuzzi

and Lippi

, Current Cancer Epidemiology, J Epidemiol Glob Health 9 (2019), 217–222.

Global Burden of Disease Cancer Collaboration Fitzmaurice

Akinyemiju

T.F.

Al Lami

F.H.

Alam

Alizadeh-Navaei

Allen

Alsharif

Alvis-Guzman

Amini

Anderson

B.O.

Aremu

Artaman

Asgedom

S.W.

Assadi

Atey

T.M.

Avila-Burgos

Awasthi

Ba Saleem

H.O.

Barac

Bennett

J.R.

Bensenor

I.M.

Bhakta

Brenner

Cahuana-Hurtado

Castaneda-Orjuela

C.A.

Catala-Lopez

Choi

J.J.

Christopher

D.J.

Chung

S.C.

Curado

M.P.

Dandona

das Neves

Dey

Dharmaratne

S.D.

Doku

D.T.

Driscoll

T.R.

Dubey

Ebrahimi

Edessa

El-Khatib

Endries

A.Y.

Fischer

Force

L.M.

Foreman

K.J.

Gebrehiwot

S.W.

Gopalani

S.V.

Grosso

Gupta

Gyawali

Hamadeh

R.R.

Hamidi

Harvey

Hassen

H.Y.

Hay

R.J.

Hay

S.I.

Heibati

Hiluf

M.K.

Horita

Hosgood

H.D.

Ilesanmi

O.S.

Innos

Islami

Jakovljevic

M.B.

Johnson

S.C.

Jonas

J.B.

Kasaeian

Kassa

T.D.

Khader

Y.S.

Khan

E.A.

Khan

Khang

Y.H.

Khosravi

M.H.

Khubchandani

Kopec

J.A.

Kumar

G.A.

Kutz

Lad

D.P.

Lafranconi

Lan

Legesse

Leigh

Linn

Lunevicius

Majeed

Malekzadeh

Malta

D.C.

Mantovani

L.G.

McMahon

B.J.

Meier

Melaku

Y.A.

Melku

Memiah

Mendoza

Meretoja

T.J.

Mezgebe

H.B.

Miller

T.R.

Mohammed

Mokdad

A.H.

Moosazadeh

Moraga

Mousavi

S.M.

Nangia

Nguyen

C.T.

Nong

V.M.

Ogbo

F.A.

Olagunju

A.T.

Park

E.K.

Patel

Pereira

D.M.

Pishgar

Postma

M.J.

Pourmalek

Qorbani

Rafay

Rawaf

D.L.

Roshandel

Safiri

Salimzadeh

Sanabria

J.R.

Santric Milicevic

M.M.

Sartorius

Satpathy

Sepanlou

S.G.

Shackelford

K.A.

Shaikh

M.A.

Sharif-Alhoseini

She

Shin

M.J.

Shiue

Shrime

M.G.

Sinke

A.H.

Sisay

Sligar

Sufiyan

M.B.

Sykes

B.L.

Tabares-Seisdedos

Tessema

G.A.

Topor-Madry

Tran

T.T.

Tran

B.X.

Ukwaja

K.N.

Vlassov

V.V.

Vollset

S.E.

Weiderpass

Williams

H.C.

Yimer

N.B.

Yonemoto

Younis

M.Z.

Murray

C.J.L.

and Naghavi

, Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-Years for 29 Cancer Groups, 1990 to 2016: A Systematic Analysis for the Global Burden of Disease Study, JAMA Oncol 4 (2018), 1553–1568.

Woodard

G.A.

Jones

K.D.

and Jablons

D.M.

, Lung Cancer Staging and Prognosis, Cancer Treat Res 170 (2016), 47–75.

Toumazis

Bastani

Han

S.S.

and Plevritis

S.K.

, Risk-Based lung cancer screening: A systematic review, Lung Cancer 147 (2020), 154–186.

National Lung Screening Trial Research Team Aberle

D.R.

Adams

A.M.

Berg

C.D.

Black

W.C.

Clapp

J.D.

Fagerstrom

R.M.

Gareen

I.F.

Gatsonis

Marcus

P.M.

and Sicks

J.D.

, Reduced lung-cancer mortality with low-dose computed tomographic screening, N Engl J Med 365 (2011), 395–409.

Moyer

V.A.

and U.S. Preventive Services Task Force, Screening for lung cancer: U.S. Preventive Services Task Force recommendation statement, Ann Intern Med 160 (2014), 330–338.

USPSTF, Lung Cancer: Screening, in: U.S.P.S.T.F, ed., 2021.

Becker

Motsch

Trotter

Heussel

C.P.

Dienemann

Schnabel

P.A.

Kauczor

H.U.

Maldonado

S.G.

Miller

A.B.

Kaaks

and Delorme

, Lung cancer mortality reduction by LDCT screening-Results from the randomized German LUSI trial, Int J Cancer 146 (2020), 1503–1513.

Paci

Puliti

Lopes Pegna

Carrozzi

Picozzi

Falaschi

Pistelli

Aquilini

Ocello

Zappa

Carozzi

F.M.

Mascalchi

and the ITALUNG Working Group, Mortality, survival and incidence rates in the ITALUNG randomised lung cancer screening trial, Thorax 72 (2017), 825–831.

10.

de Koning

H.J.

van der Aalst

C.M.

de Jong

P.A.

Scholten

E.T.

Nackaerts

Heuvelmans

M.A.

Lammers

J.J.

Weenink

Yousaf-Khan

Horeweg

van ’t Westeinde

Prokop

Mali

W.P.

Mohamed Hoesein

F.A.A.

van Ooijen

P.M.A.

Aerts

den Bakker

M.A.

Thunnissen

Verschakelen

Vliegenthart

Walter

J.E.

Ten Haaf

Groen

H.J.M.

and Oudkerk

, Reduced Lung-Cancer Mortality with Volume CT Screening in a Randomized Trial, N Engl J Med 382 (2020), 503–513.

11.

Wille

M.M.

Dirksen

Ashraf

Saghir

Bach

K.S.

Brodersen

Clementsen

P.F.

Hansen

Larsen

K.R.

Mortensen

Rasmussen

J.F.

Seersholm

Skov

B.G.

Thomsen

L.H.

Tonnesen

and Pedersen

J.H.

, Results of the Randomized Danish Lung Cancer Screening Trial with Focus on High-Risk Profiling, Am J Respir Crit Care Med 193 (2016), 542–551.

12.

van Meerbeeck

J.P.

and Franck

, Lung cancer screening in Europe: where are we in 2021? Transl Lung Cancer Res 10 (2021), 2407–2417.

13.

Brodersen

Voss

Martiny

Siersma

Barratt

and Heleno

, Overdiagnosis of lung cancer with low-dose computed tomography screening: meta-analysis of the randomised clinical trials, Breathe (Sheff) 16 (2020), 200013.

14.

Seijo

L.M.

Peled

Ajona

Boeri

Field

J.K.

Sozzi

Pio

Zulueta

J.J.

Spira

Massion

P.P.

Mazzone

P.J.

and Montuenga

L.M.

, Biomarkers in Lung Cancer Screening: Achievements, Promises, and Challenges, J Thorac Oncol 14 (2019), 343–357.

15.

Sears

C.R.

and Mazzone

P.J.

, Biomarkers in Lung Cancer, Clin Chest Med 41 (2020), 115–127.

16.

Sullivan

F.M.

Mair

F.S.

Anderson

Armory

Briggs

Chew

Dorward

Haughney

Hogarth

Kendrick

Littleford

McConnachie

McCowan

McMeekin

Patel

Rauchhaus

Ritchie

Robertson

Robles-Zurita

Sarvesvaran

Sewell

Sproule

Taylor

Tello

Treweek

Vedhara

and Schembri

, Early Diagnosis of Lung Cancer Scotland, Earlier diagnosis of lung cancer in a randomised trial of an autoantibody blood test followed by imaging, Eur Respir J 57 (2021).

17.

Goldstraw

Chansky

Crowley

Rami-Porta

Asamura

Eberhardt

W.E.

Nicholson

A.G.

Groome

Mitchell

Bolejack

and International Association for the Study of Lung Cancer Staging and Prognostic Factors Committee, Advisory Boards, and Participating Institutions, The IASLC Lung Cancer Staging Project: Proposals for Revision of the TNM Stage Groupings in the Forthcoming (Eighth) Edition of the TNM Classification for Lung Cancer, J Thorac Oncol 11 (2016), 39–51.

18.

Kuhn

, Building Predictive Models in R Using the caret Package, Journal of Statistical Software 28 (2008), 1–26.

19.

Integrative Analysis of Lung Cancer Etiology and Risk (INTEGRAL) Consortium for Early Detection of Lung Cancer Guida

Sun

Bantis

L.E.

Muller

D.C.

Taguchi

Dhillon

Kundnani

D.L.

Patel

N.J.

Yan

Byrnes

Moons

K.G.M.

Tjonneland

Panico

Agnoli

Vineis

Palli

Bueno-de-Mesquita

Peeters

P.H.

Agudo

Huerta

J.M.

Dorronsoro

Barranco

M.R.

Ardanaz

Travis

R.C.

Byrne

K.S.

Boeing

Steffen

Kaaks

Husing

Trichopoulou

Lagiou

La Vecchia

Severi

Boutron-Ruault

M.C.

Sandanger

T.M.

Weiderpass

Nost

T.H.

Tsilidis

Riboli

Grankvist

Johansson

Goodman

G.E.

Feng

Brennan

Johansson

and Hanash

S.M.

, Assessment of Lung Cancer Risk on the Basis of a Biomarker Panel of Circulating Proteins, JAMA Oncol 4 (2018), e182078.

20.

Wang

Zhi

Yang

Tian

Zhao

Zhang

Wang

Liu

Shen

Zheng

Zhao

Yang

Xin

Stojadinovic

Avital

Lee

L.J.

Rao

and Zhang

, A novel serum based biomarker panel has complementary ability to preclude presence of early lung cancer for low dose CT (LDCT), Oncotarget 8 (2017), 45345–45355.

21.

Mazzone

P.J.

Wang

X.F.

Han

Choi

Seeley

Scherer

and Doseeva

, Evaluation of a Serum Lung Cancer Biomarker Panel, Biomark Insights 13 (2018), 1177271917751608.

22.

Molina

Filella

Auge

J.M.

Fuentes

Bover

Rifa

Moreno

Canals

Vinolas

Marquez

Barreiro

Borras

and Viladiu

, Tumor markers (CEA, CA 125, CYFRA 21-1, SCC and NSE) in patients with non-small cell lung cancer as an aid in histological diagnosis and prognosis. Comparison with the main clinical and pathological prognostic factors, Tumour Biol 24 (2003), 209–218.

23.

Cabrera-Alarcon

J.L.

Carrillo-Vico

Santotoribio

J.D.

Leon-Justel

Sanchez-Gil

Gonzalez-Castro

and Guerrero

J.M.

, CYFRA 21-1 as a tool for distant metastasis detection in lung cancer, Clin Lab 57 (2011), 1011–1014.

24.

Wang

Chu

Zeng

Wang

Sun

Chen

Wang

Zhang

and Zeng

, Development of a prediction model with serum tumor markers to assess tumor metastasis in lung cancer, Cancer Med 9 (2020), 5436–5445.

25.

Bao

Tang

Liu

Yang

and Huang

, Current lung cancer screening guidelines may miss high-risk population: a real-world study, BMC Cancer 21 (2021), 50.

26.

Walter

J.E.

Heuvelmans

M.A.

Yousaf-Khan

Dorrius

M.D.

Thunnissen

Schermann

Groen

H.J.M.

van der Aalst

C.M.

Nackaerts

Vliegenthart

de Koning

H.J.

and Oudkerk

, New Subsolid Pulmonary Nodules in Lung Cancer Screening: The NELSON Trial, J Thorac Oncol 13 (2018), 1410–1414.

Assessment of circulating biomarkers for detection of lung cancer in a high-risk cohort

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Materials and methods

2.1 Inclusion and exclusion

2.2 Biomarker analysis

2.3 Statistics

2.4 Ethical considerations

3. Results

3.1 Lung cancer diagnosis

Table 1 Participants’ characteristics

Table 2 Biomarker detection abilities

Table 3 Prediction models detecting stage I-II or III-IV lung cancer

4.1 Limitations

5. Conclusions

Footnotes

Acknowledgments

Author contributions

References

Table 1
Participants’ characteristics

Table 2
Biomarker detection abilities

Table 3
Prediction models detecting stage I-II or III-IV lung cancer