Lung Cancer Screening 2009:

Lung Cancer Screening 2009: Should I Be Screened for Lung Canc= er?

Frederic W. Grannis Jr. M.D.

Long Beach CA

This page was last updated on January 8, 2009.

In October 2006, the International Early Lung Cancer Action Program, a research consortium of radiologists, pulmonologists, thoracic surgeons, pathologists, epidemiologists and other experts from medical centers around= the world reported in the New England Journal of Medicine on the long-term surv= ival of 484 cases of lung cancer detected during the course of a decade of scree= ning of more than 30,000 individual study subjects who were at high risk of lung cancer because of active or former cigarette smoking or exposure to other carcinogens. The report demonstrated that the actuarial survival of all patients detected with lung cancer was 80% at ten years. = An update of the data in 2008 shows no decrease in longer-term survival.

Why is this study important?

Experience gathered over the past thirty years indicates that we have had little impact on reducing the death toll of lung cancer by relying upon imp= rovements in treatment. Five year survi= val for Non-small cell lung cancer has improved from 12% thirty years ago to on= ly 15or 16% today. Five-year sur= vival in small cell lung cancer remains around 4%. Despite persistent optimism with t= his approach, it is now clear to me that, in the absence of a major breakthroug= h in the treatment of lung cancer, future progress in reducing deaths will depend upon prevention of lung cancer by tobacco control and smoking cessation, an= d in smokers who cannot quit and in ex-smokers it can only be accomplished by ea= rly detection of lung cancer.

Never before has a series of patients with lung cancer reached the excel= lent survival results seen in the I-ELCAP research. Long-term 80% survival represents a six-fold increase in the number of survivors compared with five-year total = lung cancer specific survival in the U.S. of 15% and a ten fold improvement in survival compared with overall survival of lung cancer patients at 10 years= in the U.S..

Results of I-ELCAP research:

There is now good information from a number of different studies from th= e U.S., Europe and Japan, that the low-dose, non-contrast spiral computerized tomogram (CT or CAT scan) is highly effect= ive in detecting lung cancer in an early stage in high risk persons.

http://www.ielcap.or= g/publications.html

CT scan detects 80% of lung cancers in Stage I; most in stage IA. This is about three times as sensitive as the chest x-ray. Most lung cancer is dete= cted in the United States= in advanced stages. Only about 7% of cases present in stage IA. The reason = that it is important to detect lung cancer early is that most cases of Stage I l= ung cancer can be cured by surgical resection alone without adjuvant radiation therapy or chemotherapy.

Recommendation:

Anyone wishing to participate in this lung cancer screening research tri= al can contact one of the 46 ELCAP sites near them through the links provided = at the individual center web sites, for further information about eligibility, costs and registration. A major advantage to screening at one of the I-ELCAP centers is that there is a management protocol, central data and image file management and quality control of radiologic interpretation, pathologic diagnosis and treatment.

People who would prefer to participate in a prospective randomized resea= rch project have limited options in the U.S. because the NCI/NLST stu= dy has fully accrued and is not accepting new patients.

A third alternative is to participate in one of the many non-research ba= sed community radiographic screening programs that have opened throughout the country in the past few years. To date, there have been few publications of results of screening in these centers. There are two main types of programs using either low-dose, non-contrast spiral CT or a different technology cal= led electron beam CT to test for both coronary artery disease as well as tumors= in the chest and other parts of the body. Until there is publication of data reporting the results of management of pulmonary nodules in these programs, this author recommends participation in a research-based screening program whenever possible.

If there is no availability of such research-based screening at a reason= able distance from your home, or you do not qualify for participation in a resea= rch trial (for example if you have had a prior lung or other tobacco caused can= cer) I believe that participation in a non-research community screening program = is preferable to no screening in individuals who are at high risk for lung can= cer i.e. current smokers and ex-smokers who are at least 40 years of age and who have smoked a minimum of ten-pack years of cigarettes i.e. one pack per day= for ten years or two packs per day for five

The following is a direct quote from the American Cancer Society.

“Testing for Early Lung Cancer Detection

At the current time, neither the ACS nor any other

medical/scientific organization recommends testing

for the detection of early lung canc= er in asymptomatic

individuals. However, the ACS historically h= as

recognized that patients at high risk of lu= ng cancer

due to significant exposure to tobac= co smoke or occupational

exposures may decide to undergo testing

for early lung cancer detection on an individual basis

after consultation with their physicia= n.38 Because of

the likelihood that a growing number= of individuals

would seek testing for early lung canc= er detection

with spiral CT,39-42 the ACS issued a narrative in

2001 emphasizing the importance of shared decision

making with regard to testing for early lung cancer

detection.6 The narrative not only emphasized the

importance of discussing potential benefits= and

harms, but also the importance of test= ing in settings

with multidisciplinary expertise in diagnostic workup

and treatment. “

The problem with this position is that very few physicians in the U.S.= have sufficient experience in and knowledge about lung cancer screening to = assist patients interested in screening with guidance in shared decision making, including accurate information on risks, potential benefits and harms. This web page will attempt to provi= de interested individuals to obtain accurate information regarding their risk = of lung cancer, their chance of surviving with standard treatment of lung canc= er following diagnosis after development of symptoms and following diagnosis in a screen= ing program. Subsequent web pages= will examine potential risks, alternatives and benefits of screening.=

The American Cancer Society statement goes on to state that =

“Currently, prospective trials to evaluate<= /span>

the efficacy of lung cancer screening are underway

in the United States and Europe,= with results

expected before the end of the decade.40 An update

to the current narrative concerning shared decision

making related to testing for early lung cancer detection

is not anticipated until results fr= om prospective

clinical trials currently underway are available.”

This statement is inaccurate, in that it has now been conced= ed by the managing staff of the National Cancer Institute’s National Lung Screen Study that final information on the data from this study will not be published by the end of the decade. Deferring decisions on screening for a number of years will inevitably translate into hundreds of thousands of unnecessary lung cancer deaths.

1. What= is my risk of lung cancer?

This is the first and most important questi= on that an individual must ask before making an informed decision about whethe= r to be screened for lung cancer. = To find an answer to this important question, go to the next web page.

What i= s my risk of lung cancer?

If you would like further information on the I-ELCAP research continue reading below.

What has been the response to t= he New England Journal publication?

Many reviewers state that the I-ELCAP results are too good to be true and result from a failure to control for multiple biases rather than a true improvement in survival. Edit= orial and media commentary has been generally negative and has tended toward confrontation and hostility. = The consensus opinion reached by guideline groups (NCCN, ACCP) has been that th= is study does not have sufficient power to convince epidemiologists and health= care providers (Medicare, Medicaid, HMOs, Blue Cross, Blue Shield or other medic= al insurance companies) that they should pay for CT screening for lung cancer. Critics of the I-ELCAP research insist that a final decision on population screening of high-risk = patients cannot be recommended until completion of a prospective randomized study.

In my opinion the delay incumbent on these recommendations represents a catastrophic error that will inevitably result in the deaths of tens of thousands of lung cancer patients who might otherwise have been salvaged by= treatment of screen-detected, early-stage lung cancer.

How was (and is) the I-ELCAP St= udy conducted?

Because much of the criticism of the I-ELCAP study comes from individuals who appear to have little understanding of how the research was conducted, = it is important to address this topic in some detail.

The I-ELCAP study is grounded upon consideration of terrible survival results following treatment of symptomatic lung cancer, prior chest x-ray screening research and upon prior research conducted by Japanese investigat= ors that demonstrated that CT scanners are capable of detecting tiny nodules in= the lungs even when the voltage of the machines is turned down to low levels, to reduce radiation exposure.

The primary investigator of the I-ELCAP study is Dr. Claudia Henschke MD= of Cornell University in New York City. Her group completed a preliminary = study (ELCAP) on 1000 high risk patients in 1999. This study, which was initially re= jected by American medical journals and published in the British journal Lancet, demonstrated that low-dose, non-contrast (no IV or drugs required) CT scans were able to find pulmonary nodules as small as 2 mm. in diameter. http://www.ielcap.org/rp/lanc= et.html

The CT scan was shown to b= e much more sensitive than the chest x-ray in detecting lung nodules. 27 (2.7%) high risk individuals (&= gt;60 years; >10 pack/years) were diagnosed with lung cancer. Of critical importance, 80% of can= cers were diagnosed in Stage I compared to only 15-20% of total lung cancers in = the U.S. di= agnosed in stage I. The duration of t= he study was not sufficiently long to provide information on long-term surviva= l, but because survival in stage I has generally been found to be in the range= of 40-60%, the ELCAP study strongly suggested that CT screening, in the contex= t of the ELCAP protocol, could save many lives.

Because only a small percentage of the pulmonary nodules detected by screening CT scans were cancers, the Cornell investigative team developed s= ets of rules (algorithms or protocols) based upon the principle that lung cancer screening is not a single test, but rather a process of medical management designed to detect and diagnose = lung cancers at a small size and early curable stage, while simultaneously avoid= ing unnecessary further testing, radiation exposure, needle biopsy, surgical operations, anxiety and expense. Diagnosis in early stage should result in timely application of cura= tive treatment, improved survival and reduced lung cancer mortality. These I-ELCAP protocols are review= ed every six months based upon continuing evaluation of data on now more than 40,000 study patients. Revisions in the guideline management recommendations are made by a panel of study investigators based upon this review. These protocols are available onli= ne at www.ielcap.org

Critics challenged the study results on the grounds that results were insufficient to support recommendations for population lung cancer screening. In response, the C= ornell investigators invited researchers, first within the state of New York (NY-ELCAP) and later in the rest of the U.S. and internationally (I-E= LCAP), to participate in a larger study prospective, single arm study of CT screen= ing using the same set of principles incorporated in the original study.

Prodded by the formation of I-ELCAP, NCI investigators quickly cobbled together a prospective randomized trial comparing CT to CXR out of two exis= ting studies. NCI funded this rese= arch with $200,000,000.

I-ELCAP Study Design:

I-ELCAP investigators chose the single arm model rather than a prospecti= ve randomized model for a number of reasons.

First, the prospective randomized trial requires a control arm of study subjects who get standard care compared with the trial intervention, in this case, CT screening. Since the National Cancer Institute has concluded and repeatedly advised the public t= hat chest x-ray (CXR) screening is ineffective, standard care in the U.S. is= no screening with beginning efforts to diagnose and treat lung cancer only upon the onset of symptoms. In add= ition it had already been shown that CT was far more sensitive in early diagnosis= of lung cancer than CXR (ELCAP) and that retrospective data from Japanese screening studies suggested substantially higher survival in CT detected lu= ng cancer (in the range of 80%) than in CXR detected lung cancers (approximate= ly 50%). Furthermore, data from literally hundreds of series of lung cancer treatment had consistently shown markedly improved survival in patients with Stage I lung cancer compared wi= th Stages II, III and IV.

Ethics of Human Research Considerations:

In order for a medical researcher to design a randomized prospective clinical trial, he/she should first decide that they have “equipoise”. Equi= poise is an ethical concept that means that the researcher truly does not know whether the new intervention (treatment arm, or in this case screening arm) will be better than the standard approach (control arm). Given the known markedly different= survival rates between the two approaches (above) the ability of a rational individu= al to claim equipoise for such a study is questionably ethical. I for example am an ex-smoker with= a substantive risk of developing lung cancer. Knowing what I do, I would not agr= ee to participate in a randomized clinical trial compared to doing no screening or chest x-ray screening only. (I should note that I do believe that the data from the Mayo Lung Trial indica= tes that chest x-ray screening is probably better than no screening.) Accordingly, I could not ethically advise a patient of mine to be screened in a study that offered no screenin= g or in a study that offered CXR screening (like the NCI’s National Lung Screen Study (NLST)).

Another more practical reason against such a study is that potential stu= dy subjects are typically reluctant to participate in a study that offers them= no foreseeable benefit and it would be difficult or impossible to find 50,000 people who would agree to participate in such a study. One of the problems with prospecti= ve randomized trials is that, depite their theoret= ical advantages, they often incorporate bad compromises because of their size and expense that detract from the integrity of the research.

Although the NCI recommends against CXR screening, their investigators c= hose to make CXR the control arm of their NLST study. This improves the ethical basis of= the study somewhat and assisted them to quickly accrue 50,000 study subjects, b= ut represents bad science and violates the principles underlying prospective controlled research.

Another problem with this research design is “crossover”. Consider this. If you are an impoverished or frug= al individual at risk of lung cancer who has looked at the NLST design and dec= ided that you would prefer a CT screen, but can’t afford or don’t= 217; want to pay for the test ($150-1000), agreeing to participate in the study = you have a 50% chance of getting the CT scan.&= nbsp; If you lose the coin flip and are put in the CXR arm, you just drop = out of the study or get together the money for the CT scan and “cross ove= r: to the CT arm by paying for a CT scan.&nbs= p; Such crossover compromises the integrity of the study.

Although a single-arm study has theoretical disadvantages in that there = is no control group, we have very precise information on what happens when a person is diagnosed with lung cancer based on the size and stage at which i= t is diagnosed. There is an enormo= us data base of tens of millions of cases of lung cancer existing in national cancer registries around the world that can be used as “historical= 221; controls against which to compare results in a screening study. Using such a control is a bit problematic because a somewhat different population is addressed, but has an enormous advantage in the sense that no patients are exposed to the increas= ed risk of lung cancer death by inclusion in a no-screen or CXR-screen study arm. To put this a different = way, there is no reason to cause the death of a large number of nice people just= to satisfy the intellectual curiosity of an epidemiologist. Perhaps this is why epidemiology h= as been styled “medicine without the tears”.

In addition, survival can be compared between the treatment arm, in this case the I-ELCAP study and a control arm in a different prospective trial.<= span style=3D'mso-spacerun:yes'> For example, an international meet= ing has recently convened in Washington D.C. when cancer advocacy groups requested permissi= on to use the data from the NCI’s Prostate, Lung, Colon, Ovary screening study (PLCO). This study includes an enormous nu= mber of individuals who are similar to study subjects in the IELCAP study but who have had either CXR or no-screening. Although this PLCO study has now been in process for 15 years and has not yet produced any survival data, the NCI refused to allow use of their d= ata.

The PLCO study represents another classic problem with prospective randomized research trials. I= t is enormously expensive and, if and when it ever produces survival data, the d= ata will be relatively useless because it will only provide information about a= technology, the CXR that is hopelessly obsolete, at least in the U.S.. By refusing use of PLCO data in th= is manner, NCI wastes an opportunity to regain some value from the $200,000,00= 0 or so spent to date on this study.

Other Publications:<= /b>

In the time period between the Lancet 1999 and New England Journal of Medicine articles, members of the I-ELCAP consortium published a large numb= er of reports describing experiences with various aspects of lung cancer screening. http://ielcap.org= /ielcap/pubs.html

Information from these studies, that explain many sidelights and minor issues of the research have been generally ignored by critics of the resear= ch.

Study Participants:<= /b>

A list of 46 medical centers in 16 states and 8 nations outside of the <= st1:place w:st=3D"on">U.S. that participate in the I-ELCAP study is available online at http://ielcap.org/members/me= mbers.php

I have been a principle investigator in the I-ELCAP study centered at Ci= ty of Hope National Medical Center in Duarte CA, 2001-2008.

Warning:

I emphasize that I am not recommending non-research study based lung cancer screening outside of the specific context of the I-ELCAP protocols.= Screening done by inexperienced individuals without proper attention= to safety features can result in poorer results and increased harm to study subjects. In fact screening research publications from groups that do not incorporate carefully structu= red protocols have demonstrated lower survival and increased rates of biopsies = and surgical operations for non-cancerous lung nodules.

This disclaimer does not imply that lung cancer screening can only be do= ne in the setting of a cancer center or university hospital. The I-ELCAP consortium includes pr= ograms from excellent community medical centers that have demonstrated excellence = in lung cancer screening.

Study structure and safeguards:=

It must be emphasized that all participating I_ELCAP centers are careful= ly monitored for quality control of their adherence to technical, diagnostic a= nd treatment standards during the course of the study.

One important difference between the I-ELCAP protocol and other screening approaches is an emphasis on the importance of demonstrating growth of scre= en detected nodules smaller than 15 mm in diameter. This differs, for example, from br= east cancer screening where there is a higher reliance upon needle biopsy.

Each participating institution has to meet standards with regard to the quality of their CT scanners. In the last decade, major improvements in the quality of CT scanners have allo= wed reduced radiation exposure, fewer repeat radiographic studies and biopsies, improved assessment of growth of nodules.&= nbsp; PET scanning has assisted in documenting metabolic activity in screen detected nodules.

Only individuals with well documented risk factors known to result in a = high risk of lung cancer participate in the study. These patients include older adult= s with a long history of cigarette smoking. They include both current smokers and people who had quit smoking in= the past.

Each center is allowed a degree of latitude with regard to which patients are considered high-risk. In = our City of Hope study, we included patients over age 40 who had at least 10 pack-years of exposure (i.e. patients who had smoked 1 pack/day for ten years or ½ pack/day for 20 years). Other centers studied patients with exposure to asbestos and/or radon gas.

A small percentage of non-smokers were studied in Japanese I-ELCAP centers. There is currently no evidence documenting whether individuals who are never smokers would benefit from lung cancer screening. T= his is an important subject for future research, now under way in I-ELCAP centers, since about 15% of lung cancer patients have not smoked.

Only asymptomatic patients with only “smokers cough” and no other symptoms suggestive of lung cancer were included in the study. One study center is known to have accrued patients with symptoms of lung cancer into their arm of the study, = with important adverse impact that will be discussed in detail in a later “= ;WWW page entitled Lung Cancer Screening: Arguments Con and Pro.”

Each I-ELCAP center has to agree to provide smoking cessation advice in conjunction with screening.

Each I-ELCAP center agrees to perform an annual CT scan for a minimum of three years and to provide long-term follow up for all patients diagnosed w= ith lung cancer. All current information is based on screening at one-year intervals. Future research may indicate that screening at longer intervals produces acceptable results, but no informati= on currently exists on this question.

Each center must enter data for each study patient via the internet into= a central computer repository at Cornell U. The quality of the data is continu= ously reviewed.

Each center transmits DICOM electronic files of each CT image to a centr= al computer repository at Co= rnell U. This allows central review of the quality of radiographic technique and interpretation at new participating centers during their early experience.&nbs= p; Feedback on interpretation is provided to study radiologists. Special three-dimension computer programs aid in determination of small amounts of growth in lung nodules.
Each center sends pathology specimens from each cancer case for review b= y a Pathology Panel constituted by an international group of seasoned experts in the pathological evaluation of lung cancer.

In the past, and in many locales, the management plan for pulmonary nodu= les found on chest radiographic studies has been surgical i.e. “When in doubt; cut it out.” This approach has resulted in many unnecessary surgical operations for benign pulmonary nodules. I-ELCAP protocols contain a strong emphasis on the importance of obtaining tissue d= iagnosis by needle biopsy rather than primary surgical resection.

Investigators meet in symposia every six months to hear presentations of study data and new developments in technology and medical science.

Each cancer case diagnosed in the study is presented to a panel of prima= ry investigators at I-ELCAP meeting each six months.

The I-ELCAP protocol is modified and updated each year based upon experience.

A teaching program that allows medical students, residents and staff physicians to learn and practice diagnostic case management within the cont= ext of the I-ELCAP protocol is available on the I-ELCAP website courtesy of Isr= aeli radiologist Dorith Shaham<= /span> M.D.. http://ielcap.org/pr= ofessionals/teaching.html

Results of I-ELCAP research:

There is now good information from a number of different studies from th= e U.S., Europe and Japan, that the low-dose, non-contrast spiral computerized tomogram (CT or CAT scan) is highly effect= ive in detecting lung cancer in an early stage in high risk persons.

http://www.ielcap.or= g/publications.html

CT scan detects 80% of lung cancers in Stage I; most in stage IA. This is about three times as sensitive as the chest x-ray. Most lung cancer is dete= cted in the United States= in advanced stages. Only about 7% of cases present in stage IA. The reason = that it is important to detect lung cancer early is that most cases of Stage I l= ung cancer can be cured by surgical resection alone without adjuvant radiation therapy or chemotherapy.

Recommendation:<= /p>
Anyone wishing to participate in this lung cancer screening research tri= al can contact one of the 46 ELCAP sites near them through the links provided = at the individual center web sites, for further information about eligibility, costs and registration. A major advantage to screening at one of the I-ELCAP centers is that there is a management protocol, central data and image file management and quality control of radiologic interpretation, pathologic diagnosis and treatment.

People who would prefer to participate in a prospective randomized resea= rch project have limited options in the U.S. because the NCI/NLST stu= dy has fully accrued and is not accepting new patients.

A third alternative is to participate in one of the many non-research ba= sed community radiographic screening programs that have opened throughout the country in the past few years. To date, there have been few publications of results of screening in these centers. There are two main types of programs using either low-dose, non-contrast spiral CT or a different technology cal= led electron beam CT to test for both coronary artery disease as well as tumors= in the chest and other parts of the body. Until there is publication of data reporting the results of management of pulmonary nodules in these programs, this author recommends participation in a research-based screening program whenever possible.

If there is no availability of such research-based screening at a reason= able distance from your home, or you do not qualify for participation in a resea= rch trial (for example if you have had a prior lung or other tobacco caused can= cer) I believe that participation in a non-research community screening program = is preferable to no screening in individuals who are at high risk for lung can= cer i.e. current smokers and ex-smokers who are at least 40 years of age and who have smoked a minimum of ten-pack years of cigarettes i.e. one pack per day= for ten years or two packs per day for five

The most important question that an individual must ask before making an informed decision about whether to be screened for lung cancer is

“What is my risk of developing lung cancer.”