Transcript LUNG CANCER SCREENING Pitfalls, Promise & Practice

LUNG CANCER SCREENING
Albert A. Rizzo, MD FACP FCCP
Section Chief Pulmonary/ Critical Care Medicine
Christiana Care Health System, Newark De
&
Immediate Past-Chair
National Board of Directors
American Lung Association
September 2012
Disclosures
•
•
•
•
•
Pfizer – Speakers’ Bureau and Research
Novartis – Speakers’ Bureau and Research
Boehringer Ingelheim – Researach
CSL Behring – Consulting
Intermune - Research
Some initial facts
• In the year 2012, ACS estimates 260,000 new cases of
lung cancer and 160,000 lung cancer related deaths
• Worldwide in 2008, 1.4 million lung cancer deaths
• 75% of patients with lung cancer present with
symptoms due to advanced local or metastatic disease
not amenable to cure
• 5 yr survival of 16%
What is Screening?
• Screening for cancer means testing for cancer
before there are any symptoms.
• Screening for some types of cancer has
reduced deaths by early detection and
treatment. Now there is a test that can reduce
death from lung cancer through early
detection.
• The best way to prevent lung cancer is to
never smoke or stop smoking now. If your
patients are still smoking, talk to them
about ways you can help them quit
smoking.
Reasons That Screening in Lung Cancer
would be effective
High morbidity and mortality
Significant prevalence (0.5-2.2%)
Identifiable risk factors (allow targeting)
Lengthy pre-clinical phase
Therapy is more effective in early stage disease
A Guiding Principle of Lung Cancer
Screening
• Make sure that the benefits of treating the
smaller number of patients who will receive a
lung cancer diagnosis through screening
outweigh the harms that could occur from
screening a large population of healthy
individuals.
Ideal Screening Test
High Sensitivity
High Specificity
Safe and Acceptable
Low Cost
Decrease mortality and/or improve quality of life
Outcomes to be assessed
Cancer detection rates
Stage at detection
Survival
Disease-specific mortality
Overall mortality
Potential Harms of Screening
Detection of benign abnormalities
Radiation risks
Prolonged follow-up
Overdiagnosis
Lung Cancer Screening Trials
Understanding terms…
Survival =
# alive with lung cancer
# with lung cancer
Mortality =
# lung cancer deaths
# screened
The problem with survival
Lead time bias:
• In screening, survival is confounded by lead-time
− Earlier detection increases survival even if death is not
delayed
CT DX
Sx DX
Lead time
Cancer
D
Survival
Survival
The problem with survival
Length time bias:
Screening tends to detect more slowly growing cancers
Aggressive cancer
1- 2 of 8
Short biological life span
3 - 4 of 8
Indolent cancer
Longer biological life span
Time
Screen
Growth Rate of Lung Cancer
Winer-Muram. Radiology 2002;223(3):798-805.
• Median DT 181 days
• 22% DT >= 465 days
• 94% >= 1 yr grow 0.5-3.0
cm
Overdiagnosis Bias (Pseudodisease)
Death
Other causes
No screen
Screen
Autopsy
Dx
CT Dx
Screening detects cancer (pseudodisease) that would remain
subclinical before death from other causes
Effects of Overdiagnosis
•Falsely increases sensitivity of test
•Falsely increases PPV of test
•Falsely increases incidence
•Falsely improves stage distribution
•Falsely improves case survival
•Does not decrease pop mortality
Volunteer Bias
May not represent general population
More concerned they have increased risk
More than usually health conscious (lower risk)
Screening WILL prolong survival.. regardless of mortality
benefit
Dx
Average
Test A
N =100
Years
0
2
4
6
8
10
Sensitive
Test B
N =100
Test A
Prevalence: 3/100 = 3%
5 Yr Survival: 2 living / 4 with known cancer = 50%
10 Yr Survival: 1 living / 4 with lung cancer = 25%
Mortality at 10 yrs: 3 deaths / 100 screened = 3% mortality
Test B
Baseline Prevalence: 10/100 = 10%
5 Yr Survival: 8 living / 10 with lung cancer = 80%
10 Yr Survival: 7 living / 10 with lung cancer = 70%
Mortality at 10 yrs: 3 deaths / 100 screened = 3% mortality
= aggressive cancer
= average cancer
= non-significant cancer
= screen diagnosed cancer
= cancer DEATH
Screening WILL prolong survival.. regardless of mortality
benefit
Dx
Average
Test A
N =100
Years
0
2
4
6
8
10
Sensitive
Test B
N =100
Test A
Prevalence: 3/100 = 3%
5 Yr Survival: 2 living / 4 with known cancer = 50%
10 Yr Survival: 1 living / 4 with lung cancer = 25%
Mortality at 10 yrs: 3 deaths / 100 screened = 3% mortality
Test B
Baseline Prevalence: 10/100 = 10%
5 Yr Survival: 8 living / 10 with lung cancer = 80%
10 Yr Survival: 7 living / 10 with lung cancer = 70%
Mortality at 10 yrs: 3 deaths / 100 screened = 3% mortality
= aggressive cancer
= average cancer
= non-significant cancer
= screen diagnosed cancer
= cancer DEATH
CT-detected Lung Cancers will Have ↑
survival
• Lead-time: advance time of Dx even absent a
delay in death
• Screening selects for biologically favorable
cancers (length bias)
• Some proportion = overdiagnosis (overtreatment)
− Biologically benign behavior
− Death from competing cause
• True ↑ in longevity indeterminate
History of lung cancer screening
Early lung cancer screening with sputum and CXRs in 1970’s
Radiology
2011
Limitations of early lung cancer
screening trials
• Studies each had
different designs.
Combined analysis
limited
• Small sample sizes
• Primarily addressed the
benefit of sputum
cytology, not CXR
• Some degree of
contamination (control
arm receives the
screening intervention.
CXR lung cancer screening
• CXR RCT screening trials: Mayo | MSK | Johns Hopkins | Czech studies
−
−
−
−
−
Studied some combination of CXR and sputum cytology
Observed increases in lung cancer detection rate over controls
Improved survival of screen-detected lung cancers
No mortality benefit
Long term follow-up: endpoints unchanged
Mayo Lung Project Incidence
Screening
Control
4618
4593
Lung cancers detected
206
160
Resectable cancers, %
46
32
5-year survival, %
33
15
122
115
No of patients
Lung cancer deaths
Early Low Dose CT lung cancer screening Trials
• Early low dose CT screening: Single arm
studies
− ELCAP: Henschke CI et al. Lancet 1999; 354:99-105.
− Japanese Studies
• Sobue et al. J Clin Oncol. 2002;20:911-920
• Sone et al. British J Cancer 2001; 84:25-32
• Nawa et al. Chest. 2002;122:15-20
− Mayo Study. Swensen SJ et al. Radiology 2005; 235:259265
Early low dose CT screening: Single arm studies
N
[+] Screens
Total
Lung Ca
Stage I
NSCLC
ELCAP CT Prevalence
1000
233 (23%)
27 (2.7%)
23 (85%)
ELCAP CXR Prevalence
1000
68 (7%) *
7(0.3%)
4 (57%)
Incidence Year 1-2
1184
2.5%
7
(2 interval)
5
US Trial
* 50% were false shadows ; positive screens are actually 3.5%
•
•
•
Trial design: Single arm—subjects received both annual CXR and CT screening
Eligibility: Asymptomatic | > 60 yrs | > 10 pack yrs
Trial endpoints: The frequencies of
•
•
•
•
Nodule detection
Nodules representing malignancy
Malignant nodules that are curable
ELCAP DID NOT ADDRESS MORTALITY, BUT DEMOSTRATED THE IMPORTANCE
OF CT OVER CXR IN LUNG CANCER DETECTION
Henschke CI et al. Lancet 1999; 354:99-105
Henschke CI et al. Cancer 2001; 92:153-159
Early low dose CT screening: CONCLUSIONS
• CT detection rates
− CT has ~6-fold higher nodule detection rate than
CXR
− CT has ~3-fold higher cancer detection rate than CXR
− Benign nodules = majority of detected nodules (> 90%)
• CT has 5-fold increase in resectable lung cancers
• No proven benefit in lung cancer mortality
True stage shift which requires not only an increase in earlystage disease but a concomitant decrease in late-stage
disease when compared with a control population.
N LS T
National Lung
Screening Trial
National Cancer Institute
NLST design and time posts
NL S T
• RCT
• 1:1 randomization to CT or CXR
• Launched in 08-2002 across ~ 33 sites
National Lung
Screening Trial
Final Analysis
National Cancer
Institute
Interim analyses
CT
53,476
HighRisk
Subjects
2002
CXR
T0
03
04
T1
05
06
07
Follow up
T2
08
09
10
Standardized Eligibility
 Males | Females
N LS T
National Lung
Screening Trial
National Cancer Institute
 55-74 Yrs
 Asymptomatic
 Current or former smokers ≥ 30 pack
yrs
 Former smokers have quit within ≤ 15
yrs
 No prior lung cancer
 No cancer within past 5 yrs
 No chest CT w/in prior 18 months
Endpoints
• Lung cancer specific mortality
− 90% power to detect a 20% mortality reduction with
LDCT
− Compliance: 85% CT | 80% CXR
− Contamination: 5% CT | 10% CXR
• All cause mortality
• Medical resource utilization
• ACRIN secondary aims:
− Short | long term effects on smoking habits and beliefs
− Cost-effectiveness
− Specimen biorepository for molecular biomarkers of risk | early Dz
NLST Research Team. Radiology 2011; 258(1):243-
True and false positive screens
Low Dose Helical CT
Screening Result
Round 1
N (%)
Round 2
N (%)
CXR
Round 3
N (%)
Round 1
N (%)
Round 2
N (%)
Round 3
N (%)
Total Positives
7,193 (100)
6,902 (100)
4,054 (100)
2,387 (100)
1,482 (100)
1,175 (100)
Lung cancer
No lung cancer
270 (4)
6,923 (96)
168 (2)
6,734 (98)
211 (5)
3,843 (95)
136 (6)
2,251 (94)
65 (4)
1,417 (96)
78 (7)
1,097 (93)
• 649 CT-detected lung cancers | 279 CXR-detected lung cancers
• 370 excess cancers by CT screening: 2.3 fold increase with CT
• Most positive screens did not have lung cancer
Screen positivity rate by screening round & arm
Low dose helical CT
CXR
Number
screened
Number
positive
%
Positive
Number
screened
Number
positive
% Positive
Screen 1 (T0)
26,314
7,193
27.3
26,049
2,387
9.2
Screen 2 (T1)
24,718
6,902
27.9
24,097
1,482
6.2
Screen 3 (T2)
24,104
4,054
16.8*
23,353
1,175
5.0*
All screens
75,136
18,149
24.2
73,499
5,044
6.9
Positive screen: Nodule ≥ 4 mm or other findings potentially related to lung cancer.
* Abnormality stable for 3 rounds could be called negative according to protocol.
3-fold increase in positive screens in CT arm.
lung cancers diagnosed in NLST
Screen Result and Time Period
CT (%)
Total T0-T2 Screen-detected lung cancers
Total Screen [-] lung cancers T0-T2
Total NO screen lung cancers
Total lung cancers in arm
Total Lung Cancers NLST
892 NO screen cancers include:
649 (61.2%)
44
CXR (%)
279 (29.6%)
(4.2%)
137 (14.6%)
367 (34.6%)
525 (55.8%)
1060 (100.0%)
941 (100%)
2001
never screened (N = 35) | due for screen (N = 55)
post-screen time period (N = 802)
stage distribution for lung cancers by screen status
CT Screens (1040 cancers)
CXR Screens (929 cancers)
Numbers reflect only lung cancers of known stage
Cumulative number of lung cancer deaths
cumulative lung cancer deaths by time from
randomization
500
400
CXR
LDCT
300
200
100
0 0
1
2
3
4
Years post randomization
5
6
7
diagnostic follow-up of positive screens
Category
Total positives
LDCT
T0 (%)
T1 (%)
CXR
T2 (%)
7191 (100%)
6901 (100%)
Confirmed lung cancer
270 (3.8%)
168 (2.4%)
211 (5.2%)
Participants with complete
diagnostic F/U
7049 (98%)
6740 (98%)
3913 (97%)
Total (%)
T0 (%)
T1 (%)
T2 (%)
Total (%)
4054 (100%) 18,146 (100%) 2387 (100%) 1482 (100%) 1174 (100%) 5043 (100%)
649 (3.6%)
136 (5.7%)
65 (4.4%)
78 (6.6%)
279 (5.5%)
17,702 (98%) 2348 (98%)
1456 (98%)
1149 (98%)
4953 (98%)
Clinical procedure
72.2%
47.3%
55.0%
58.9%
60.2%
49.7%
57.3%
56.4%
Imaging Exam
81.1%
37.4%
51.3%
57.9%
85.6%
66.5%
78.9%
78.4%
CXR
18.2%
9.1%
16.6%
14.4%
36.9%
26.2%
31.8%
32.6%
Chest CT
73.1%
30.4%
41.1%
49.8%
65.8%
51.2%
62.0%
60.6%
PET or PET-CT
10.3%
5.2%
10.0%
8.3%
7.6%
7.2%
9.8%
8.0%
Invasive Procedures
T0 (%)
T1 (%)
T2 (%)
Total (%)
T0 (%)
T1 (%)
T2 (%)
Total (%)
Percutaneous FNA/Core
2.2%
1.1%
2.4%
1.8%
3.5%
2.5%
4.5%
3.5%
Bronchoscopy
4.6%
2.6%
4.8%
3.8%
4.6%
3.8%
5.4%
4.5%
Surgical procedure(s)
4.2%
2.9%
5.6%
4.0%
5.2%
3.5%
5.8%
4.8%
Mediastinoscopy
0.9%
0.5%
0.6%
0.7%
0.9%
0.8%
1.7%
1.1%
Thoracoscopy
1.2%
0.8%
2.5%
1.3%
0.9%
0.8%
1.7%
1.1%
Thoracotomy
2.8%
2.2%
4.2%
2.9%
4.1%
3.0%
3.8%
3.7%
screening-related complications from invasive
procedures
CT
lung cancer
N
Positive screens 649
%
100
CT
CXR
CXR
NO cancer lung cancer No cancer
N
%
17,053 100
N
%
N
%
279
100 4,674 100
Major complication
75
11.6
12
0.1
24
8.6
4
< 0.1
Death 60 days after any procedure
10
1.5
11
< 0.1
11
3.9
3
< 0.1
Death 60 days after invasive procedure
10
1.5
6
< 0.1
10
3.8
0
0
NLST Summary
• CT-detects more lung cancers than CXR x 2.3 folds
• True stage shift observed in CT arm
• 20% lung cancer mortality reduction CT vs. CXR
− Absolute risk reduction = 0.4% (AR CT= 1.3% | CXR = 1.7%)
• Few major complications
• NNS (Number needed to screen) : 320
− NNS (Breast Cancer): US: 238, NZ: 781
• NCI_2012 and J med Screen, 2001;8(3):114-5
• Need for diagnostic algorithm to decrease false positives
Take Away Points from NSLT
• Population was younger than the general
population
• Population was fairly well educated
• Population included more former smokers
• Operative mortality was low (1%)
Rational for Lung Cancer Screening
Kaplan–Meier Estimates of Overall Survival (Panel A) and the Time to Progression of Disease (Panel B) in the
Study Patients, According to the Assigned Treatment.
Schiller JH et al. N Engl J Med 2002;346:92-98.
FY 2010 Federal Research Dollars
Per Death
Federal spending from the
combined FY2010 research
dollars of the National Cancer
Institute, Department of
Defense and Centers for
Disease Control and
Prevention.
85% of all lung cancers are
linked to cigarette use.
Effects of stopping smoking
at various ages on the
cumulative risk (%) of death
from lung cancer up to age
75, at death rates for men in
UK in 1990. Nonsmoker
rates were taken from US
prospective study of
mortality
Peto R, BMJ, 2000
Cost of Lung Cancer Care
• Annual Cost of Lung Cancer treatment in US
– $10 billion
• Estimated Annual treatment cost
– $21,000 per patient
– $47,000 per patient for those who live more than
one year
Heathcare Mang Sci 1999, J Clin Oncol 1997
Actuarial Analysis Shows That Offering Lung
Cancer Screening As An Insurance Benefit Would
Save Lives At Relatively Low Cost
HEALTH AFFAIRS 31,NO. 4 (2012): 770–779
Cancer
Type
Screening Test
Cost/life-year
saved (Original
Study)
Study
date
Cost/life-year
saved (2012
dollers)
Cervical
Pap Smear
33,000
2000
50,162 - 75,181
Colorectal Colonoscopy
11,900
1999
18,705 - 28,958
Breast
mammography
18,800
1997
31,309 – 51,274
Lung
LDCT (Baseline)
18,862
2012
18,862
LDCT (Lowest Cost
Scenario)
11,708
2012
11,708
LDCT (Highest Cost
Scenario)
26,016
2012
26,016
NCCN Guidelines: Lung Cancer Screening
October 26, 2011
•
•
•
NCCN. Org
•
Lung cancer screening with CT should be
part of a program of care and should not
be performed in isolation as a free
standing test.
The risks and benefits of lung cancer
screening should be discussed with the
individual before doing a screening LDCT
scan.
It is recommended that institutions
performing lung cancer screening use a
multidisciplinary approach that may
include specialties such as radiology,
pulmonary medicine, thoracic oncology,
and thoracic surgery.
Management of downstream testing and
follow-up of small nodules are imperative
and may require establishment of
administrative processes to ensure the
adequacy of follow-up.
Typical Organ Radiation Doses from Various Radiologic Studies.
Brenner DJ, Hall EJ. N Engl J Med 2007;357:2277-2284.
Ionizing Radiation
•
•
•
•
Linear No Threshold Model
Roentgens (R) vs Grays (Gy) vs Sieverts (Sv)
3 mSv – annual background dose
1000 mSv – 4-5% risk of fatal cancer
– Myeloma, leukemia, lung, thyroid, breast
• CXR 0.1 mSv, CT chest 8 mSv, LDCT chest 1.5 mSv
Institute of Medicine
• Use of CT scanning increased three fold since
1993
• Top environmental causes of breast cancer
– CT scanning
– Postmenopausal Hormone Therapy
• 30% of all CT scans “medically unnecessary”
• Justification, Optimization, Limitation
Estimated Dependence of Lifetime Radiation-Induced Risk of Cancer on Age at Exposure for Two of the Most
Common Radiogenic Cancers.
Brenner DJ, Hall EJ. N Engl J Med 2007;357:2277-2284.
NLST complications (partial list)
 Major complications
–
–
–
–
–
–
–
–
–
–
–
Acute respiratory failure
Respiratory arrest
Prolonged mechanical ventilation
Cardiac arrest | MI | CHF
Hemothorax requiring tube
Empyema
Bronchopulmonary fistula
Bronchial stump leak
Injury to vital organ | vessel
Thromboembolic complications
Death
• Intermediate complications
− Respiratory distress
− Mucostasis  bronchoscopy
− Cardiac arrhythmia needing
attention
− Cardiac ischemia | ST elevation
− Pleural effusion
− Pneumothorax needing chest tube
− Fever | infection | sepsis |
pneumonia
− Transfusion for blood loss
− Steroid-induced diabetes
− Hospitalization post procedure
− Vocal cord paralysis
− Pain requiring referral to specialist
Lung Cancer Screening: American Lung Association
Recommendations
Low-dose CT screening should be recommended for those people
who meet National Lung Cancer Screening Trail criteria:
• current or former smokers, aged 55 to 74 years
a smoking history of at least one pack a day for 30 years (30
pack-years )
• no history of lung cancer
• Individuals should not receive a chest X-ray for lung cancer
screening
• Low-dose CT screening should NOT be recommended for
everyone
Should my patient be screened?
 Prior to appointment: Questions to think about
– Does my patient meet the NLST criteria or another high-risk profile that makes them
good candidates for screening?
– What are the benefits/risks for my patient if screened?
– What are my referral options? Where do I refer?
 During the average medical visit: ~12 minutes
–
–
–
–
–
Take a complete health history and determine possible co-morbidities (conduct spirometry?)
Advocate smoking cessation
Educate about symptoms of underlying lung disease
Discuss the benefits /risks and possible procedures associated with the screening process Discuss
costs of screening – health insurance reimbursement, time and personal costs
If recommending screening, discuss
• Availability – low dose CT is available and will be done, high quality machine and staff
• Low Dose CT /qualified center with expertise to follow up after test
• Follow up care through multidisciplinary approach.
 Points to Remember:
–
–
–
Do not offer a Chest X-ray as an option for screening
Refer patients only to qualified centers, those providing low-dose CT scans and a multi-disciplinary
team for follow-up
Ensure that patients know the difference between a screening test and screening process.
Screening Referrals
Refer your patient to institutions that have
experience in conducting Low Dose CT scans, as
well as, using the latest CT technology.
• Make sure that the facility uses “best practices”
for lung cancer screening
• There should be a link to an expert
multidisciplinary team that can provide followup for evaluation of nodules.
• Discussion of results and follow-up is key.
Making an Individual Decision to Get Screened:
Questions to Ask Your Doctor
• Understanding
– Are you considered high risk for lung cancer?
• NLST high risk group
• Genetics
• Smoking History – pack years
• Other exposures – occupational, etc.
– What is your current health status? Do you have co-morbidities like
COPD?
– What are my screening options?
– What does the screening process entail?
• Risks
– What are some of the complications of the diagnosis procedure?
• False Positives
• Negative results do not absolutely rule out the chance for cancer
incidence
– What is the cumulative exposure to radiation?
– What are the unknowns?
Making an Individual Decision to
Get Screened:
Questions to Ask Your Doctor, con’t
 Costs
– Does my health insurance company reimburse
for CT scans?
– How much does a CT scan cost? What about
the cost of follow up procedures?
– How will screening affect my quality of life?
(living with indeterminate diagnosis)
– How much time and personal costs will I spend
throughout the screening
Screening Flow Chart: What Happens if a Patient Chooses to Be
Screened?
Low Dose CT
Screening
Negative
Re-Screen
Risks
Indeterminate
Watchful
Follow Up
Further
Imaging
Multi-disciplinary
Approach
Smoking Cessation
Risks
Suspicious
Further Immediate and
Potentially Invasive
Procedures
Lung Cancer
Treatment Possibilities:
Surgery, Palliative Care,
Clinical Trials, Other
No Lung Cancer
Continued
Screening
Concerns: Length
of Time Intervals
Some Remaining Issues
•
•
•
•
•
Cost effectiveness
Screening frequency
Population targets
Criteria for “positive” findings
Follow up protocols to decrease false-positive
evaluations
THANK YOU
QUESTIONS?