Screening and Early Diagnosis in Lung Cancer.ppt

Screening and early diagnosis in lung cancer

ESMO Stockholm 2008

[email protected]

Epidemiology of lung cancer

• 1912 - 374 cases worldwide (Adler I. Primary malignant growths of the lungs and bronchi. New York: Longmans, Green, and Company; 1912)

• 20th century– 100 million tobacco-related deaths

• 21st century– expected 1 billion deaths; 1/3 due to lung cancer!!

Smoking and smoking cessation

• Smoking is and will remain the leading cause of cancer death (Jemal A et al. Ca Cancer J Clin 2002; 52: 23-47)

• Smoking is increasing among young people and women (Tickle JJ et al. Favorite movie stars, their tobacco use in contemporary movies, and its association with adolescent smoking. Tob Control 2001; 10: 16-22)

Cessation:• Reduces death rate to non-cancer causes immediately (Barone-

Adesi F et al. Short-term effects of Italian smoking regulation on rates of hospital admission for acute myocardial infarction. Eur Heart J 2006; 27:2468.)

• Very limited effect on lung cancer death rate(Enstrom JE et al. Smoking cessation and mortality trends among 118,000 Californians, 1960-1997. Epidemiology 1999, 100: 500-512)

Lung cancer classification based on:

• TNM-staging (revision due early 2009)• histology (SCLC versus NSCLC)• further histological description: squamous,

adeno, large cell• location in thorax: central versus peripheral

Lung cancer outcome• Strongly related to size and stage: small tumors

do better!

0%

20%

40%

60%

80%

100%

0 5 10 15Years After Enrollment

Deaths / N MST 5-YearcT1, <=2cm 217/ 423 68 53%cT1, >2-3 cm 248/ 445 52 47%cT2, <=5 cm 787/ 1345 43 43%cT2, >5-7 cm 272/ 411 30 36%cT2, > 7 cm 127/ 173 17 26%cT3 354/ 486 19 29%

After stereotactic radiotherapy (Lagerwaard FJ et al. Int J Radiation Oncology Biol Phys 2008; 70: 685-692)

After surgery (Rami-Porta R et al. Journal of Thoracic Oncology. 2007; 2:593-602)

How to improve the outcome of lung cancer?

• Improve systemic therapy (may take decades)

• detect tumors at a curable stage

►try to find lung cancer at an earlier stage: screen the high risk group

However……..

• This is only effective if the outcome of the screen-detected tumors is identical to the outcome of the clinically found tumors…..

How can we screen?• Investigation of patient tissue or cells: sputum cytology ►mainly larger airways; squamous, SCLC

• Radiological imaging: chest X-ray; spiral CT►mainly lung parenchyma; adenoca, large cell

• Endoscopic imaging (bronchoscopy)►only larger airways; squamous, SCLC

• Metabolic imaging (FDG-PET)►size limit > 8 mm

Chest X-ray screening improved outcome for detected lung cancer,

Marcus P et al. Lung Cancer Mortality in the Mayo Lung Project:Impact of Extended Follow-up J Natl Cancer Inst 2000; 92: 1308-1316

Especially in resected early stage but …….

Marcus P et al. Lung Cancer Mortality in the Mayo Lung Project:Impact of Extended Follow-up J Natl Cancer Inst 2000; 92: 1308-1316

Marcus P et al. Lung Cancer Mortality in the Mayo Lung Project:Impact of Extended Follow-up J Natl Cancer Inst 2000; 92: 1308-1316

Chest X-ray studies: no effect on survival

Number of lung cancer deaths

In the example shown, the diagnosis of disease is made earlier in the screened group, resulting in an apparent increase in survival time (lead-time bias), although the time of death is the same in both groups.

Patz EF et al, NEJM 2000; 343: 1627

Potential biases in screening studies: Lead-time bias

The probability of detecting disease is related to the growth rate of the tumor. Aggressive, rapidly growing tumors have a short potential screening period (the interval between possible detection and the occurrence of symptoms). Thus, unless the screening test is repeated frequently, patients with aggressive tumors are more likely to present with symptoms. More slowly growing tumors have a longer potential screening period and are more likely to be detected when they are asymptomatic. As a result, a higher proportion of indolent tumors is found in the screened group, causing an apparent improvement in survival.

Potential biases in screening studies: Length-time bias

Overdiagnosis bias is an extreme form of length-time bias. The detection of very indolent tumors in the screened group produces apparent increases in the number of cases of lung cancer (three in the screened group in the figure and one in the control group) and in survival (two of three patients in the screened group were treated and died of natural causes, without evidence of disease [66 percent survival], and the one patient in the control group did not survive [0 percent survival]), with no effect on mortality (one death from lung cancer in each group). Two patients in the control group died with undiagnosed lung cancer that did not affect their natural life span.

Potential biases in screening studies: Overdiagnosis bias

Other problems with screening studies for lung cancer

• What type of tumors will be found: by CT-screening only parenchymal lesions• If there is no reduction in detection of late stages it will never effect survival

• Crossoverof patients due to availability of screening instruments outside the study

• Sticky-diagnosis bias. Because the target cancer in a screening trial is more likely to be diagnosed in the screened group than in the control group, deaths are more likely to be attributed to the target cancer in the screened group.

• Slippery-linkage bias. Death due to a procedure related to the study but not perse in a case with the target cancer (for instance benign tumor), is often not scored appropriate

Baseline Annual repeat

Total scans 13,122 10,245Number of Tumors 112 55

Stage I (%) 55-85 60-100Stage III,IV (%) 3-36 0-36Rate of Cancer (%) 0.4-2.7 0.07-1.1Mean diameter (mm) 14-21 10-16Invasive procedures for benign lesions (%) 4-22 14-15Interval cancers (%) NA 0-22

Low dose spiral-CTScreening results from 5 large

observational studies

ALCA: anti lung cancer association, JapanELCAP: Early Lung Cancer Action Project, USAMayo Clinic Study, USALung Cancer Screening study, Milan, ItalyHitachi Health Center, Japan

Mulshine J, NEJM, 2005

I-ELCAP results

• 31,567 baseline and 27,456 repeat scans 7-18 months after baseline

• 484 reviewed LC cancer cases detected

• All cases: 10-yr survival 80%• Stage I resected cases (85%): 10-yr survival

92%

• Stage I untreated cases (n=8): dead within 5-yrs

Henschke C, NEJM 2006

I-ELCAP results

Henschke C, NEJM 2006

I-ELCAP results

• Estimated 10-yr survival data• Median FU 3.3 yr, small numbers at risk after 5 yrs

(n=90) and 10 yrs (n=2)• Of all resected stage I LC cases (n=375) survival

curves of only those resected < 1 months presented (n=302).

• No lung cancer specific mortality data

Comments on ELCAP

• Length-time bias?• Overdiagnosis?• Reduction in later stages at annual

screen?• Is the population found comparable to the

normal lung cancer population?

I-ELCAP: length-time bias?

• Of 35 stage I LC cases detected at baseline in ELCAP-I and ELCAP-II 9 had > DT 400 days (26%)

• So far DT’s of all baseline cases have not been reported

• But 87% either invasive growth or DT < 400 days; according to authors overdiagnosis uncommon

• Does invasive growth exclude overdiagnosis ?Henschke C et al. J Clin Imag 2006;30:11-15.

Does invasive growth exclude overdiagnosis ?

• In retrospective study of 149 stage I LC cases 2 CT scans were made

• Median interval 130 days (range 25-2493 days)

* Jennings SG et al, Radiology 2006;241:554-563

Distribution of Doubling times

* Jennings SG et al, Radiology 2006;241:554-563

Tumor Doubling time and LC survival

• Median % volume change +54% (range -81%-10.759%)

• 48/149 (32%) proven invasive LC cases showed no growth or DT > 1 year

• Survival lower in fast growing tumors (p=0.002) also for non-adenocarcinomas/BAC (p<0.05).

• Elderly people had more slow or non-growing tumors (p = 0.047)

* Jennings SG et al, Radiology 2006;241:554-563

I-ELCAP: reduction in number of late stages?

More LN metastases: • baseline 7%• Annual repeat 11%

Different histological distribution:• baseline: adeno/BAC 76%, SCC 13%• annual repeat: 48% adeno/BAC, SCC 22%

I-ELCAP results

7% 11%

less frequent LN metastases

Henschke C et al. Arch Intern Med 2006;166:321-325

Nodule consistency and frequency of LN metastases

Henschke C et al. Arch Intern Med 2006;166:321-325

More baseline lung cancer cases found by CT screening in females

• Henschke, Miettinen (2004): higher risk (2.7) of detecting lung cancer in females than in males at the baseline CT screen

• Controlled for age and pack-years of smoking

• Confounders discussed and found unlikely to contribute (underreporting of smoking in females)

Henschke and Miettinen, Lung Cancer 2004

Effect of Gender in Mayo CT trial

• VDT of 48 cancers detected during 5 annual CT’s

• Males 38% (n=18) , Females 62% (n=30)

Lindell et al Radiology 2007

Male Female

Mean VDT (d) 234±447 688 ±1321

Median VDT (d) 92 217

Possibly Overdiagnosed cancers especially in females

• Overdiagnosed cancers are those with a VDT > 400 days

• Mayo Clinic study: 23% had VDT > 400 days; for all histological subtypes VDT was longer in females

• Hasagawa et al: 44% (27/61) VDT > 400 days

Lindell et al Radiology 2007

Hasagawa et al Br J Radiol 2000

Gender difference in tumor growth rates: females demonstrate slower

growth

387

126

559

256

0

100

200

300

400

500

600

700

CT CXR

VDT, daysMales Females

Hasegawa et al2000 Usuda et al1994

CT Screening for Lung Cancer

in Health Examination

Toru Nakagawa MD Suzushi Kusano MD

Syuichiro Hosoda MD Syuichiro Yamamoto MD

Masataka Irokawa MD

Hitachi Health Care Center, HITACHI, Ltd., JapanE-mail：：：：[email protected]

• Large population study in Japan

• Baseline detection rate 0.47%, 90% stage I• Incidence detection rate 0.09%, 100% stage I

• baseline 57/60 ( 95%) adenocarcinoma !!• Incidence 20/22 (91%) adenocarcinoma !!

• Smokers and non-smokers (females)• Lower LC incidence in Japan than in US and Europe• Slow growing adenocarcinomas

0

0.161

0.09

0.474

0.07 0.083

0

2000

4000

6000

8000

10000

12000

14000

Baseline Repeat

1st

Repeat

2nd

Repeat

3rd

Repeat

4th

Repeat

5th

00,050,10,150,20,250,30,350,40,450,5

Total of Exams. Detection Rate(%)

70％％％％ 54％％％％ 39％％％％ 24％％％％10％％％％100％％％％

Changes of the Detection Rate of Lung Cancer during Baseline and annual Repeat Screening

“ Should Spiral CT screeningfor lung cancer screening already be introduced on a large scale ?”

What would support implementation

aLC is an important diseaseaPrimary prevention is unsuccessfulaThe consequences of LC are substantialaWe have an acceptable and feasible test

(LDCT)aLC can be detected in an early stage by

LDCT

Questions to be answered:

?Has an mortality-based survival benefit been demonstrated and what is the magnitude of lead time and overdiagnosis bias

?Are the side-effects of LC screening in balance with the amount of favourable health outcomes (test/QOL/natural history)

?Will LC screening be accepted to individuals at risk and health care providers (implementation)

? Is LC screening reasonably cost-effective

Overview RCT’s on spiral CT

• NLST: USA, n= 53,000 • NELSON: Holland-Belgium, n=16,000• Italy: n= 10.000 (Pastorino, proposal)• Denmark: n=4,000, part of NELSON• ITALUNG-CT: Italy, n= 3,000• Germany/Heidelberg: n=4,000, NELSON design

National LungScreening Trial

National Cancer Institute

National LungScreening Trial

National Cancer Institute

TSLNChristine D. Berg, MDChief, Early Detection Research GroupDivision of Cancer Prevention | NCIProject Officer, LSS-NLST

Deni R. Aberle, MDCancer Imaging Program, DCTD | NCI

David Geffen School of Medicine at UCLANational PI, ACRIN-NLST

The combat against lung cancer

A Randomised Clinical Trial for

Lung Cancer Screening in high risk subjects

2004-2009

1,019

1,889

6,814

53,476

Nr screened

31/12/05

47%NoNo4-16 detector

MSCT

3

(4 units)

ITALUNG-CT

50%YesYes16-detector

MSCT

1Denmark

76%YesYes16-detector

MSCT

4NELSON

63%NoNo4-64 detector

MSCT

46NLST

Stage I,II

Double reading

Volu-

metry

CTNumber of sites

Trial

Overview ongoing RCT’s

64%/36%2,33355-69No screening

Random sample

ITALUNG-CT

55%/45%4,10450-70No screening

MediaDenmark

85%/15%15,53050-75No screening

Random sample

NELSON

59%/41%53,47655-74CXRMediaNLST

Males / Females

NoAgeControl arm

Selection procedure

Trial

Overview ongoing RCT’s

6 yrs1 yr4<10 yrs>20 PYITALUNG-CT

5 yrs1 yr5<10 yrs> 20 PYDenmark

6 yrs1 and 2 yr3<10 yrs>30 yrs >11 cig

>26 yrs >16 cig

NELSON

2 yrs1 yr3< 15 yrs> 30 PYNLST

Follow-up period

Screening interval

Screening rounds

Ex-smokers

Smoking history

Trial

Overview ongoing RCT’s

Conclusion

• LDCT screening for lung cancer is a very promising early detection tool

• Several uncertainities make it premature to advocate Spiral CT screening on a large scale

• Spiral CT screening should not be recommended outside clinical trials

Flehinger BJ et al. The effect of surgical treatment on survival from early lung cancer. Implications for screening. Chest 1992;101;1013-1018