[CANCER RESEARCH56. 4673-4678, October 15, 19961
diagnosed; over 50% will have advanced-stage IIIb or IV disease at
diagnosis, both of which are usually incurable by present therapy (2).
Reasons for this phenomenon reside in the fact that symptoms of early
lung cancer are often nonspecific, thus precluding many patients from
being motivated to seek medical attention during a time interval
wherein treatment could be advantageous (3, 4). Indeed, resectability
and survival rates are beuer among those with early stage lung cancer
(5—8).Such observations lend credence to the idea that a reduction in
lung cancer mortality is achievable if this disease can be diagnosed
and treated before metastatic spread.
The premise that early detection can produce a mortality benefit is
not new and, in fact, has served as the impetus for several studies
(9 —1 8) over the past four decades to evaluate the efficacy of lung
cancer screening by chest radiography and sputum cytology (the only
tests of established value for detecting presymptomatic bronchogemc
carcinoma). One of the most influential of these studies was the large,
randomized, controlled trial sponsored by the National Cancer Insti
tute in the early 1970s. Entitled the National Cancer Institute Coop
erative Early Lung Cancer Detection Program, this cooperative trial
was conducted at Johns Hopkins medical institutions, the Memorial
Sloan-Kettering Cancer Center, and the Mayo Clinic. The approxi
mately 30,000 participants were males 45 years of age who smoked
1 or more packs of cigarettes per day (or within 1 year of enrollment;
Ref. 13). At the Johns Hopkins medical institutions and Memorial
Sloan-Kettering Cancer Center, the study sample received annual
chest radiographs plus sputum cytological evaluations every four
months; the control sample received only yearly chest radiographs. In
the Mayo Clinic trial, the study sample received both chest radio
graphs and sputum cytological tests every four months; the control
group was simply advised that these two tests should be obtained
annually. Regardless of differences in the experimental designs em
ployed, these three trial centers uniformly failed to show a reduction
in mortality as a consequence of periodic chest radiographs and
sputum cytological examinations (14—17). This outcome prompted
the National Cancer Institute and other health policy and research
groups to conclude that large-scale radiological or cytological screen
ing for early lung cancer was not efficacious.
The National Cancer Institute cooperative trial had a debilitating
effect on early lung cancer detection research. In the 20 years since
that trial was initiated, however, significant advances have occurred in
the understanding of lung cancer biology, thereby renewing interest in
this topic. Intensive study of the molecular genetics of colon carci
noma have buttressed the multistep theory of carcinogenesis (19) and
suggest that invasive cancer of many types, including lung carcinoma
(20), is the result of an accumulation of mutations in precursor cells.
A corollary to this theory is the concept of “field cancerization.―In
lung cancer, field cancerization embodies the contention that the
entire respiratory epithelium is susceptible to carcinogens and that
repeated exposures may translate into the independent development of
precursor lesions with the potential for their progression at variable
rates to cancer (21). Precursor lesions may be identified by genetic or
4673
Cytopathological Analysis of Sputum in Patients with Airflow Obstruction
and Significant Smoking Histories'
Timothy C. Kennedy,2 Susan P. Proudfoot, Wilbur A. Franldin, Thomas A. Merrick, Geno Saccomanno,
Mary E. Corkill, Donna L. Mumma, Karim E. Sirgi, York E. Miller, Philip G. Archer, and Allan Prochazka
Lung Cancer Institute of Colorado, Denver, Colorado 80218 fT. C. K.. S. P. P.): Departments of Surgical Pathology 1W. A. F.), Pulmonary Sciences and Critical Care Medicine
and Division of Medical Oncology IY. E. Ml. Preventive Medicine and Biometrics (P. G. A.], and Section of Ambulatory Care (A. P.), University of Colorado Health Sciences
Center and Veterans Affairs Medical Center, Denver, Colorado 80209; Department of Pathology, Presbyterian/St. Luke ‘s Medical Center. Denver, Colorado 80218 (T. A. M.,
K. E. SI; Department ofPathology, St. Mary's Hospital and Medical Center, Grand Junction, Colorado 81501 1G. 5]; Department ofPathology, LutheranMedical Center, Wheat
Ridge, Colorado 80033 (M. E. C.]: and Department of Pathology, Memorial Hospital, Colorado Springs, Colorado 80909 (D. L M.)
ABSTRACT
Advances in the understanding of lung cancer biology have led to
observations that specific genetic changes occur in premalignant dyspla
ala. These observations have occurred predominantly in molecular studies
of resected lung tumors and consequently, they may not be fully repre
sentative of those biological abnormalities characterizing premalignant
lesions in individuals without overt lung cancer. Studies of premalignant
epithelial cell biology and chemoprevention are needed in this patient
subgroup. Such an initiative is now underway through the lung cancer
Specialized Program of Research Excellence (SPORE) grant awarded to
the University of Colorado Cancer Center (and affiliated institutions) by
the National Cancer Institute. To identify participants for the early de
tection and chemoprevention trials of the Colorado SPORE, we initiated
a sputum cytology screening program targeting persons with chronic
obstructive pulmonary diseaseand smokinghistories of 40 or more pack
years. During the first 26 months after activation of the screening pro
gram, sputum samples from 632 participants were evaluated. Of these,
533 (84%) of the subjects submitted specimens deemed adequate for
cytopathological interpretation; 99 (16%) provided sputum samples Un
suitable for cytodlagnesis. Of those participants who submitted adequate
samples, 48% had cytodiagnoses of mild dysplasia, 26% had moderate to
severe dysplasia, and 2% presented with carcinoma in situ or invasive
carcinoma. Logistic regression modeling was pursued to determine
whether selected demographic and/or clinical status variables could be
identified as statistically significant predictors of the specific cytological
outcome to be expected (mild dysplasia, moderate dysplasia, and so forth).
The only apparent associations found from both univariate and multiva
date analyses were that the total number ofpack-years ofsmoking history
decreased with severity of cytodiagnosis and that those individuals with
mild or moderate dysplasia were more likely to be ex-smokers than those
with grades of regular metaplasia or lower. Based on the initial results of
the Colorado SPORE sputum cytology screening program, we conclude
that persons with chronic obstructive pulmonary disease and 40 or more
pack-years of smoking history have a high prevalence of premalignant
dysplasia detectable through sputum cytology and should be targeted for
research programs focusing on lung cancer prevention, early detection,
and exploratory biomarker studies.
INTRODUCTION
Lung cancer is now the leading cause of cancer death among both
males and females in the United States (1). It is also one of the most
lethal types of cancers to acquire, as reflected through an overall
5-year survival rate of 14% (1). In part, the poor prognosis of lung
cancer is due to the late stage at which this disease is typically
Received 10/30/95; accepted 8/16/96.
Thecostsof publicationof thisarticleweredefrayedinpartbythepaymentof page
charges. This article must therefore be hereby marked advertisement in accordance with
18U.S.C.Section1734solelyto indicatethisfact.
1 Supported by National Grant No. P50 CA58187. All authors are either faculty or
affiliated researchers of the University of Colorado Cancer Center, University of Colorado
Health Sciences Center, Denver, CO 80209. Colorado SPORE pathologists: W. A. F.
(quality control reviewer), T. A. M., 0. S. (quality control adviser), M. E. C., D. L. M.,
and K. E. S.
2 To whom requests for reprints should be addressed. at Lung Cancer Institute of
Colorado, 1721 East 19th Avenue, Suite 366, Denver, CO 80218.
Research.
on October 1, 2021. © 1996 American Association for Cancer cancerres.aacrjournals.org Downloaded from