Screening for Lung Cancer

What the PCP Needs to Know

Eric Rich, MDSt. Luke’s Idaho Pulmonary

Associates

Outline Lung cancer impact

Background: Where have we been? CXR and sputum screening

Low dose CT/NLST: Where are we going?

Pro’s and con’s of screening

Cost effectiveness and comparison to other screening tests

Future directions

Lung Cancer

Lung Cancer Leading cause of cancer deaths 2011 estimates

221,130 new cases will be diagnosed (Breast 230,480)

>150,000 deaths due to lung cancer (Breast 39,970)

Survival 1-year relative survival 43% 5-year survival

All stages combined 16% Localized disease 53% (only 15% detected are early

stage)

NIH Funding 2010

Funding does not reflect epidemiology

Cancer $5,823 (Millions) Breast $763 Prostate $331 Lung $201 Lymphoma $195

Principles of early disease detection

J Thorac Dis 2013;5(S5)

Philadelphia Pulmonary Neoplasm Research Project 1951

Veterans Administration study 1958-1961 South London Lung Cancer Study 1955-1963

Nonrandomized, uncontrolled studies North London Cancer Study 1960’s.

Randomized to screening vs. no screening Kaiser Foundation Health Plan screening trial 1964-

1979 Annual CXR, spiro, questionnaire. Controlled.

Mayo Lung Project (MLP) Randomized to sputum cytology and CXR every 4

months vs. advised to seek annual screening No difference in mortality (median 20.5 years follow-

up)

Screening: CXR and Sputum

Cytology

No benefit of screening on mortality

Small studies, majority of follow-up less than 10 years

Self-selection bias, lead-time bias, overdiagnosis bias

Screening: CXR and Sputum

Cytology

Early Lung Cancer Action Project (ELCAP). 31,567 at risk patients 1993-2005 screened with low-dose CT

(LDCT), and 27,456 repeat screenings 7-18 months after previous screening. Smokers and nonsmokers, exposures, aged 40-86 Evaluated outcome of stage I cancers diagnosed by screening CT

Prevelance study: Initial CT: 21% positive tests, 13% required additional work-up Follow-up CT: 5% positive

484 lung cancers 412 (85%) stage I

• Estimated 10 year survival 88%, regardless of treatment

• Survival rate 92% if surgical resection in 1 month

“In a population at risk for lung cancer, such screening could prevent some 80% of deaths from lung cancer.”

Conclusion: Annual spiral CT screening can detect lung cancer

that is curable

Concerns Raised

Not randomized

No control group

Lacked unbiased outcome measure

Did not address harms from screening

Survival always increased by early detection Deaths are not delayed

Arch Intern Med. 2007 Nov 26;167(21):2289-95.

•“Study Sees Gain on Lung Cancer” 2006•“Cigarette Company Paid for Lung Cancer Study”• Disclosure line reported no conflicts of

interest

•“Review Casts More Doubts on a Lung Cancer Study” • 90% of consent forms cannot be located

•Authors received royalties from GE, makes CT scanners (2007)•Stock ownership and consulting for company that makes biopsy needles for lung cancer diagnosis

•Inventors on 27 patents and applications: Lung cancer (2006) screening technology embedded in I-ELCAP protocol

Contradictory results in different publications using same data

Statistically unlikely results Percentage of “Early” stage diagnoses unlikely

Outlier results Stage at screen detection inconsistent with other

studies Were patients with advanced disease excluded or

reclassified? Lung cancer survival is not consistent with other

data. Raises question of poor capture of lung cancer death.

J Natl Cancer Inst. 2011 Jul 6;103(13):1002-6.

“I believe that in the case of ELCAP data, the findings cannot be properly interpreted without understanding the basis of some of the conflicting, improbable, and outlier findings.”

•Longitudinal analysis 3246 current or former smokers•Annual CT scans, comprehensive evaluation and treatment of detected nodules•Compared predicted to observed number of new cases, lung cancer resections, advanced lung cancer cases, and deaths from lung cancer•144 cases of lung cancer compared to 44.5 expected cases (RR 3.2)•109 lung resections compared to 10.9 expected (RR 10)•No decline in number of advanced lung cancer or deaths from lung cancer“Screening for lung cancer with LDCT may increase the rate of lung cancer diagnosis and treatment, but may not meaningfully reduce the risk of advanced lung cancer or death from lung cancer.”

LDCT detects more nodule and lung cancers, including early stage cancers, than CXR.

No randomized, controlled trial has shown that this leads to a mortality benefit

Therefore, in 2002 the NCI funded the NLST.

Methods

Randomized comparison of screening with LDCT vs. screening with CXR.

33 participating centers in the USA. 10 Lung Screening Study sites (LSS) 23 American College of Radiology Imaging

Network (ACRIN)

Participants Eligible participants

Age 55-74 at randomization History of cigarette smoking of at least 30

pack-years If former smoker, quit within last 15 years

Exclusion criteria Diagnosis of lung cancer CT of the chest in prior 18 months Hemoptysis Unexplained weight loss > 6.8 Kg in previous year

August 2002-April 2004 patients enrolled

Screening from August 2002 through September 2007

Followed for events through December 31, 2009

Participating Sites

ACRIN centers collected additional data for cost-effectiveness, quality of life, and smoking cessation.

15 ACRIN centers collected serial blood, sputum, and urine specimens.

Lung-cancer and other tissue specimens were obtained at ACRIN and LSS centers and used to construct tissue microarrays.

Screening Three screenings (T0, T1, T2) at 1-year intervals.

Screenings outside NLST estimated by questionaires to 500 participants annually

Low-dose CT exposure average 0.6-1.5 mSv Average effective dose with diagnostic CT 8 mSv.

LDCT Noncalcified nodule >4mm labeled as “suspicious”

for lung cancer. Adenopathy or effusion could be classified as

positive At T2 stable abnormalities could be classified as

minor Guidelines for nodule follow-up were provided by

radiologists, but not mandated.

Data Collection Medical records

Diagnostic procedures and complications for patients with positive screening tests and in those with lung cancer diagnosis

Pathology and tumor-staging reports Operative procedures and initial treatments

Vital Status Questionnaire annually (LSS) or semiannually (ACRIN) Lost to follow-up patients names and SSN submitted to

National Death Index for probable vital status End-point verification team Death caused by lung cancer and a death resulting from

treatment of lung cancer counted as lung-cancer deaths

Intention-to-screen analysis Estimated 90% power to detect 21% decrease in

mortality Secondary analyses compared rate of death from any

cause and incidence of lung cancer

•Vital status known for 97% of patients•Median duration of follow-up 6.5 years•Maximum duration 7.4 years

•Adherence to screening high• 95% LDCT• 93% CXR

Trial stopped early by DSMB on October 20, 2010. Definitive result reached for primary end point.

•At least one positive result during screening• LDCT 39.1%• CXR 16%

•Clinically significant abnormality other than suspicious for lung cancer• LDCT: 7.5%• CXR: 2.1%

What happens to patients with positive

screening tests? First round of screening

90% of positives led to a diagnostic evaluation

Screening rounds combined LDCT: 24.2% positive results, 23.3% false

positive 96.4% false positives

CXR: 6.9% positive results, 6.5% false positive 94.5% false positives

Complications

Adverse Events All patients: LDCT 1.4%, CXR 1.6% Major complication from procedure

0.06% in those without lung cancer confirmation

11.6% in those with lung cancer confirmation

16 patients died within 60 days after invasive procedure

Lung Cancer

LDCT: 1060 lung cancers (645 per 100,000 person-years) 649 diagnosed after positive screening test 44 diagnosed after negative screening test 367 after missing screening or after screening

complete

CXR: 941 lung cancers (572 per 100,000 person-years) 279 diagnosed after positive screening test 137 after negative screening test 525 after missing screening or after screening

complete

Rate ratio 1.13 (95% [CI] 1.03-1.23)

•Stage I at Diagnosis• LDCT: 50%• CXR: 31.1%

•Stage III or IV at Diagnosis• LDCT: 32.9%• CXR: 40.9%

Mortality Lung-Cancer specific mortality

LDCT: 356 deaths in 144,103 person years 247 per 100,000 person years

CXR: 443 deaths in 143,368 person years 309 per 100,000 person years

Relative rate reduction: 20.0% (95% [CI] 6.8 to 26.7, P=0.004)

Number needed to screen to prevent one death 320 people

“Screening with the use of low-dose CT reduces mortality from lung cancer.” 20% relative reduction.

Rate of positive results was higher with LDCT Three times higher (24% positive)

High rate of false positives with LDCT 96.4% of all positive results

Higher rate of early stage detection 50% stage I

Major complications are rare for evaluation of positive result Surgical resection mortality 1% (4% general US)

Pro’s of Study Randomized controlled trial

Intention-to-screen analysis

Appropriate control group, no concern for systematic differences in two groups Internal Validity

Standardized reading of CTs

Extremely high adherence to screening (>90%)

Minimal loss to follow-up (3%, 4%)

Collection of data for further study Tissue, blood, sputum Cost-effectiveness Quality of life

First study to show a mortality benefit in lung cancer screening

Concerns Overdiagnosis

Detection of cancers that never would have become symptomatic.

Follow populations over time for excess cancer diagnoses in screened group. Mayo study suggests ~ 15 years of follow-up.

USPSTF estimates 10-12%

Would the same results be obtained in clinical practice? (external validity) Radiologists were trained, read large volume of CTs CT scanners now more advanced “Healthy volunteer effect”, as patients were younger

and more educated

Large variance in risk (age, smoking, family history) Who is really at highest risk Does incorporation of sputum, peripheral blood profiles,

etc increase benefit?

•Risk of radiation?•Effects on quality of life, anxiety?•Cost?•Role for smoking cessation?•Who should really be screened?•How does this compare to other accepted screening tools?•Can we do this on a local level? Infrastructure? Who’s in charge?

Downstream effects

Positive scans and incidental findings require clinical and radiologic follow-up 50% increase visits per participant (1 visit per

participant)

Risk of radiation:

LDCT 0.6-1.5 mSv (Mammography 0.7 mSv, background radiation in US 2.4 mSv per year)

Lifetime excess risk 0.85% (50 yo female smoker, 25 scans)

2-5 excess deaths per 10,000 screened 15-100 lives saved per 10,000 screened

Quality of Life 3 studies

HRQoL, anxiety, lung-cancer specific distress 2500 screening participants

Transient negative psychological effects with indeterminate or positive result

Effects subsided, no difference at 12-24 months

Improved with minimizing the waiting time for results

Smoking Cessation

Is screening a teachable moment?

ELCAP study: 23% of active smokers quit after baseline CT scan! Background population quit rate: 4%

NLST collected data on smoking cessation with full results pending Subgroup of 430 NLST participants at 1 year post-study No change overall in risk perceptions 9.7% quit, and 6.6% relapsed at 1 year

Changes risk perceptions among current/former smokers?

Opportunity for captive audience, improved motivation?

Smoking cessation counseling and treatment likely improves mortality benefits above screening alone

Cancer. 2013 Apr 1;119(7):1306-13.J Thorac Dis

2013;5(S5)524-539

Modeling study: LDCT screening annually over 15 years

NLST criteria

Compared with/without smoking cessation Light smoking cessation intervention

Behavioral treatment Intensive cessation intervention

Behavioral and pharmacologic treatment

Lung cancer screening is cost effective from a commercial payer perspective

Becomes even more cost effective when linked with smoking cessation interventions

PLoS One 2013 Aug 7;8(8)

Relative risk reduction for death 15-20% Age 40: Absolute mortality benefit 4 per

10,000 (at 10.7 years) Age 50: Absolute mortality benefit 5 per

1,000. False positives

50% over 10 years 25% with biopsies

NCI, USPSTF

Mammography is socially and economically accepted.

Is this a model to follow in lung cancer?

Assessed variation in efficacy of NLST participant data according to 5 year risk of lung cancer Efficacy False positive rate Number of lung cancer deaths prevented

Results Number of lung cancer deaths per 10,000 person

years increased across higher risk quartiles Decreased false-positives in those at higher risk The 60% at highest risk accounted for 88% of

prevented lung cancer deaths!

Higher risk quartiles account for most prevented deaths

Number needed to screen improves dramatically by risk classification

Empirical evidence for targeting highest risk population for LDCT

Potential to narrow the population needing to be screened, limit cost, and limit false positive results

Principles of early disease detection

J Thorac Dis 2013;5(S5)

ACCP, ASCO, ATS, NCCN, ACS Recommendations similar to NLST criteria ACS recommends strongly in screening at an

organized program/center with experience in LDCT

Screening Implementation

More than just providing a CT scan

Commitment to infrastructure Invitation and recruitment High quality control to track nodules over

time Primary care, pulmonary, thoracic surgery,

oncology Quality improvement Risk stratification Diagnosis and treatment Participant support and follow-up

Conclusion Low-dose CT screening appears to reduce risk of death

from lung cancer Similar to mammography?

There is a high rate of false positives, requiring further follow-up. The risk of complications from further evaluation is low.

The risk of cumulative radiation exposure is likely low

Which populations are at highest risk, and what other epidemiologic, radiographic, and biochemical/genetic data can be incorporated to improve screening?

Cost-effectiveness studies and long term follow-up to identify extent of overdiagnosis is required

More than just a CT scan, and the potential public health impact is

large and requires a multidisciplinary approach to patient

care.

Selected ReferencesNational Lung Screening Trial Research Team, Aberle DR, Adams AM, Berg CD, Black WC, Clapp JD, et al.

Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011 Aug 4;365(5):395–409.

Moyer VA. Screening for Lung Cancer: U.S. Preventive Services Task Force Recommendation Statement. Ann. Intern. Med. 2013 Dec 31.

Henschke CI, Yankelevitz DF, Libby DM, et al. Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med 2006;355:1763-71.

Bach PB, Jett JR, Pastorino, et al. Computed tomography screening and lung cancer outcomes. JAMA 2007;297:953-61.

Kovalchik SA, Tammemagi M, Berg CD, Caporaso NE, Riley TL, Korch M, et al. Targeting of low-dose CT screening according to the risk of lung-cancer death. N Engl J Med. 2013 Jul 18;369(3):245–54.

Villanti AC, Jiang Y, Abrams DB, Pyenson BS. A Cost-Utility Analysis of Lung Cancer Screening and the Additional Benefits of Incorporating Smoking Cessation Interventions. Gorlova OY, editor. PLoS ONE. 2013 Aug 7;8(8):e71379.

Marshall HM, Bowman RV, Yang IA, Fong KM, Berg CD. Screening for lung cancer with low-dose computed tomography: a review of current status. J Thorac Dis. 2013 Oct;5(Suppl 5):S524–39.

Bach PB. Inconsistencies in findings from the early lung cancer action project studies of lung cancer screening. J Natl Cancer Inst 2011;103:1002-6.

Welch HG, Woloshin S, Schwartz LM, et al. Overstating the evidence for lung cancer screening: the International early lung cancer action program (I-ELCAP) study. Arch Intern Med 2007;167:2289-95.

Marcus PM, Bergstralh EJ, Zweig MH, et al. Extended lung cancer incidence folow-up in the Mayo Lung Project and overdiagnosis. J Natl Cancer Inst 2006;98:748-56.

Swensen SJ, Jett JR, Hartman TE, et al. Lung cancer screening with CT: Mayo Clinic experience. Radiology 2003;226:756-61.