IP Fellows Reading List

Biomarker Tests for Lung Cancer Diagnosis

Audit of the autoantibody test, EarlyCDT®-lung, in 1600 patients: an evaluation of its performance in routine clinical practice

https://pubmed.ncbi.nlm.nih.gov/24268382/

Clinical Trial

Reference: Jett JR, Peek LJ, Fredericks L, Jewell W, Pingleton WW, Robertson JFR. Audit of the autoantibody test, EarlyCDT®-lung, in 1600 patients: an evaluation of its performance in routine clinical practice. Lung Cancer. 2014;83(1):51-55.

Background: EarlyCDT-Lung (Oncimmune) is an ELISA for 7 autoantibodies (7AAB: p53, NY-ESO-1, CAGE, GBU4-5, HuD, MAGE A4, and SOX2) in peripheral blood intended to aid in detecting early lung cancer in high-risk patients. These autoantibodies can be detected in peripheral blood up to 3-4 years before symptomatic presentation of lung cancer. At the start of this study, EarlyCDT-Lung included 6 autoantibodies (6AAB: p53, NY-ESO-1, CAGE, GBU4-5, Annexin I, and SOX2) but was changed during the study period in 11/2010 to the 7AAB panel because of improved specificity. This study evaluates outcomes 6 months after testing.

PICO:

• 1,613 patients who were deemed high-risk for lung cancer by their treating physician who ordered the EarlyCDT-Lung assay. Tests were ordered by 810 unique physicians in 720 different practices across the United States. Results were relayed back to the ordering physician who determined the subsequent follow-up plan.

• EarlyCDT-Lung testing from peripheral blood. 752 patients had the 6AAB assay (5/2009-11/2010) while 861 patients had the 7AAB assay (11/2010-8/2011).

• None

• For the 7AAB panel only because the 6AAB is no longer available
• Of all patients tested with the 7AAB panel, 10% had a positive result and 4% were diagnosed with lung cancer by 6 months. 15.7% of those who tested positive were diagnosed with lung cancer while 37.1% of those diagnosed with lung cancer tested positive
• The 7AAB panel had a specificity of 91%, sensitivity of 37%, and PPV of 16%
• A positive result on the 7AAB panel was associated with a 5.4-fold increase in the risk of a lung cancer diagnosis at 6 months
• At diagnosis for NSCLC, 57% were stage I or II, 35% were stage III or IV, and 9% had unknown staging.

Take home: In patients who are deemed to be high-risk for lung cancer by their physician, EarlyCDT-Lung may aid in further risk stratifying patients and prompt additional screening if the assay is positive. It should be noted that the assay had a low sensitivity and, thus, a negative result should not be used to exclude the possibility of lung cancer. Although authors use the Spitz model to assess lung cancer risk, the ordering clinician’s determination of risk was not standardized. Without a comparison group, the impact of EarlyCDT-Lung cannot be adequately assessed. It should also be noted that this study took place before the USPSTF’s recommendation for screening LDCTs in 12/2013 and the update in 3/2021 that effectively increases the number of patients eligible for radiographic screening.

Earlier diagnosis of lung cancer in a randomised trial of an autoantibody blood test followed by imaging

https://pubmed.ncbi.nlm.nih.gov/32732334/

Clinical Trial

Reference: Sullivan FM, Mair FS, Anderson W, et al. Earlier diagnosis of lung cancer in a randomised trial of an autoantibody blood test followed by imaging. Eur Respir J. 2021;57(1):2000670.

Background: This is a randomized controlled study evaluating whether incorporation of EarlyCDT-Lung (Oncimmune) reduces the incidence of advanced stage lung cancer (III, IV, or unspecified) at the time of diagnosis.

PICO:

• 12,208 patients in Scotland aged 50-75 at increased risk for lung cancer (i.e., smoked ≥20 pack-years or any smoking history with a first-degree relative with lung cancer)

• 6,088 patients underwent EarlyCDT-Lung testing
• If testing was positive, patients had a low-dose CT chest scans every 6 months for 2 years. If testing was negative, they continued with standard of care.

• 6,121 received standard of care. There was no routine lung cancer screening at the time of the study.

• In the intervention group, 9.8% had a positive EarlyCDT-Lung test. Of those who tested positive, 3% were diagnosed with lung cancer at 6 months compared to 0.7% in those who tested negative
• The absolute risk reduction of diagnosing advanced stage lung cancer in the intervention group was 0.3% for all study participants and 14.3% for participants who were diagnosed with lung cancer. The number needed to screen to prevent one lung cancer from being diagnosed at an advanced stage was 325
• There were nonsignificant differences in lung cancer-related and all-cause mortality between the two groups
• Lung cancer in the intervention group was diagnosed an average of 87.3 days earlier than in the control group

Take home: EarlyCDT-Lung can aid in diagnosing lung cancer at an earlier stage with a number needed to screen of 325 in order to diagnose one lung cancer at an early stage rather than an advanced stage. While mortality did not change at two years, it is presumed that there would be some long-term mortality benefit if followed for a longer duration since early lung cancers have better prognosis compared to advanced lung cancers. It is important to note that at the time of this study, routine low-dose CTs for lung cancer screening were not yet adopted in this study population. Many of the patients in this study (i.e., those ages >50-years-old who smoked ≥20 pack-years) would be eligible for a screening LDCT based on US guidelines.

Assessment of plasma proteomics biomarker’s ability to distinguish benign from malignant lung nodules: results of the panoptic (Pulmonary nodule plasma proteomic classifier) trial

https://pubmed.ncbi.nlm.nih.gov/29496499/

Clinical Trial

Reference: Silvestri GA, Tanner NT, Kearney P, et al. Assessment of plasma proteomics biomarker’s ability to distinguish benign from malignant lung nodules: results of the panoptic (Pulmonary nodule plasma proteomic classifier) trial. Chest. 2018;154(3):491-500.

Background: This study is a prospective validation of a bronchial-airway gene-expression classifier. Participants were enrolled in the Airway Epithelial Gene Expression in the Diagnosis of Lung Cancer trials (AEGIS-1 and AEGIS-2).

PICO:

• 639 current or former smokers, aged 21 or older, with no history of cancer who were undergoing bronchoscopy for suspected lung cancer at 28 sites across the US, Canada, and Ireland. Participants were enrolled at two different periods as two independent prospective cohorts: AEGIS-1 and AEGIS-2. Patients were followed for 12 months.

• Brushings were obtained from a mainstem bronchus at the time of a diagnostic bronchoscopy for suspected lung cancer. Samples were evaluated in a gene expression classifier, which incorporated a patient’s age and was composed of 23 genes.

• None

• 43% of bronchoscopies were nondiagnostic. 25% of patients who were ultimately diagnosed with lung cancer had a nondiagnostic bronchoscopy. The sensitivity of bronchoscopy for detection of lung cancer was 74-76% (AEGIS-1 and AEGIS-2 groups)
• The gene-expression classifier had a sensitivity of 88-89% and specificity of 47% for identifying malignant tumors. Unlike bronchoscopy alone, these sensitivities were not significantly affected by tumor size, tumor location, cancer stage, histologic type, or presence of lymphadenopathy
• Combining bronchoscopy with the gene-expression classifier increased overall sensitivity to 96-98%. The negative likelihood ratio also improved from 0.24 to 0.06
• 11% of collected brushings were insufficient or had poor-quality RNA

Take home: In patients with a smoking history and a nondiagnostic bronchoscopy, use of a gene-expression classifier from central airway brushings taken at the time of bronchoscopy may be useful in evaluating the posttest probability of lung cancer. A negative bronchoscopy and genomic classifier with a low-to-intermediate pretest probability of lung cancer may support a more conservative approach in management. These findings are currently not applicable to never-smokers as it is thought that the genetic changes in normal airways are due to a “field of injury” from smoking.

Clinical utility of a bronchial genomic classifier in patients with suspected lung cancer

https://pubmed.ncbi.nlm.nih.gov/26896702/

Clinical Trial

Reference: Vachani A, Whitney DH, Parsons EC, et al. Clinical utility of a bronchial genomic classifier in patients with suspected lung cancer. Chest. 2016;150(1):210-218.

Background: This study evaluates the clinical impact of using a bronchial genomic classifier (Percepta GSC, Veracyte) in patients with low-to-intermediate pretest probability of lung cancer.

PICO:

• 222 patients from the AEGIS-1 and AEGIS-2 cohorts with low-to-intermediate pretest probability of lung cancer. Patients were followed for 12 months

• Theoretical application of a genomic classifier to clinical practice where a negative genomic classifier resulted in surveillance over invasive testing

• Standard of care without application of a genomic classifier

• 188 (85%) of patients with low-to-intermediate pretest probabilities of lung cancer had nondiagnostic bronchoscopies. 19% of these patients were ultimately diagnosed with lung cancer. 41% underwent 99 additional interventions, including surgical lung biopsies in 52% of these patients
• 54% of patients with benign disease had a negative classifier result. If these patients had undergone surveillance only, 46% of all invasive procedures may have been avoided in patients with benign nodules
• The genomic classifier had a false-negative rate of 11%, which all occurred in patients who had an intermediate pretest probability

Take home: In patients with a smoking history and a low-to-intermediate pretest probability for lung cancer who are undergoing a diagnostic bronchoscopy, use of a bronchial genomic classifier may reduce the number of additional invasive procedures in cases of a nondiagnostic bronchoscopy. Although a false-negative result on the classifier may delay diagnosis, authors suggest that the delay would not be significant with close surveillance.