1. Introduction

New technology will eventually find application for
screening and earlier cancer detection. Technology, for
instance, is developing the new field of proteomics, refining
the science of imaging, and revealing an increasing number
of genetic abnormalities and a wide range of cancer
biomarkers. As our progress continues, we are likely to
reach a point where technology, even at a molecular level,
can detect small precancerous lesions or even individuals at
high risk. For this reason, a Precancer Conference was
convened on November 8 - 9, 2004 to develop a very general
definition of precancerous lesions that would have wide
application. The definition of "precancer" was deliberately
made general so it could apply to multiple disciplines and
provide a common denominator for clinical and experimental
studies. The Conference was co-sponsored by the
National Cancer Institute and The George Washington
University Cancer Institute.

A definition takes on added meaning when "precancerous"
changes recognized by any method of detection
become the target of prevention interventions [1 - 3]. To
demonstrate the efficacy of these interventions, it becomes
imperative to have an appropriate definition of "precancer"
regardless of the method of detection. In theory, the
identification and elimination of cancer precursors would
lead to the near-eradication of cancer [1,3 - 8]. A second
objective of the Conference, which followed from the
definition, was to establish precancer as a formal area of
research with its own set of biologic principles, fundamental
research questions, and clinical goals. The emphasis of the
Conference was naturally on histopathology, because that is
the current method of diagnostic confirmation.

2. Methods
A National Cancer Institute-sponsored Conference on
precancer was convened on November 8 - 9, 2004, at The
George Washington University Medical Center, Washington,
DC. A definition of precancers was developed over 2
days of Conference discussions.

3. Results

3.1. The precancers

It was the opinion of the Conference organizers that there
was a general understanding among pathologists and other
physicians that precancers are the lesions that precede the
development of invasive cancers; however, there was
minimal agreement initially on a set of defining criteria
that could be consistently applied to distinguish precancers
from all other lesions, regardless of the method of detection.
Subsequently, after open discussion, the Conference
participants developed and agreed on a set of defining
criteria for the precancers. This report reviews the proposed
definition of precancers and suggests how pathologists,
oncologists, and cancer researchers can determine when
these criteria are satisfied.

3.2. Terminology

The Conference participants acknowledged that a number
of different terms are used to describe the lesions herein
referred to as "precancers." In many cases, these terms
include both diagnostic and prognostic implications. Terms
include "dysplasia," "atypical hyperplasia," "in situ
carcinoma," "intraepithelial neoplasia," "preinvasive cancer,"
"preinvasive lesion," "cancer precursor," "precancerous
lesion," "precancerous state," "premalignant lesion,"
premalignancy," and "incipient neoplasia" [1,5,9]. Currently,
the term "intraepithelial neoplasia" is widely used. Its
popularity may derive from the fact that the vast majority of
tumors arising in man originate from epithelial surfaces. It
should be noted that the term "intraepithelial neoplasia"
imposes limitations that have been discussed elsewhere [8]. It
should also be noted that the precursor of most malignant
germ cell tumors in men is designated intratubular germ cell
neoplasia. For purpose of discussion, the term "precancer"
was adopted by the Conference participants because it: (1)
considers all of the lesions commonly included in the term
(i.e. the intraepithelial neoplasias), (2) opens the door to
entities excluded by other terms (e.g. lymphoid, hematologic,
and soft tissue precancers), and (3) permits inclusion of
entities diagnosed by non-morphologic methods, for instance,
predysplastic molecular lesions. Again, for purposes of
discussion only, the term precancer was used to refer to any
change, at any level of analysis or detection, that serves as a
risk factor for the development of cancer. In the context of the
Conference, precancer does not imply that cancer is the
ultimate outcome.

3.3. Definition of precancer

It was accepted by the Conference participants that
during cancer progression, distinctive lesions occur that
persist for a time and that they have a set of characteristic
properties that permit them to be detected, diagnosed,
prevented, and treated. On this basis, the participants
developed a working definition for the precancers that
clinicians and researchers could use to distinguish precancers
from non-neoplastic changes and from other types
of changes that might be encountered during "cancer
progression." This definition modifies and generalizes a
definition initially proposed for endometrial intraepithelial
neoplasia [10 - 12].

The following five defining criteria must all apply:

(1) Evidence exists that the precancer is associated with an
increased risk of cancer.

(2) When a precancer progresses to cancer, the resulting
cancer arises from cells within the precancer.

(3) A precancer is different from the normal tissue from
which it arises.

(4) A precancer is different from the cancer into which it
develops, although it has some, but not all, of the
molecular and phenotypic properties that characterize
the cancer.

(5) There is a method by which the precancer can be
diagnosed.

Each of these defining criteria is discussed.

3.3.1. Evidence exists that the precancer is associated
with an increased risk of cancer.

This criterion is designed to exclude lesions that have
some features often found in precancers (e.g. cellular atypia,
monoclonality, cytogenetic markers, and specific mutations);
but, which by themselves, do not have any propensity
for progressing to cancer. According to the Conference
participants, the term "precancer" must apply only to
lesions or changes that potentially will be clinically
significant. Lesions that can be distinguished from normal
tissues and from cancers using analytic techniques (e.g.
mathematical clustering algorithms on gene expression
datasets) but which lack any proven propensity to progress
toward cancer, do not qualify as precancers.
The group discussed at length the lines of evidence that
would be needed to prove that an identified lesion has a
known risk of progression towards cancer. After a long
debate, no consensus was reached on specific methods to
establish risk. It is the responsibility of the investigator to
provide convincing evidence of risk, using whatever means
are appropriate. Epidemiologic, morphologic, molecular,
and biologic methods might all be acceptable approaches
when justified by the investigator. It should be possible to
prove the existence of a precancer without formally
following the "lesion" through all stages of progression,
only that it has a risk of progression at any level of detection
or investigation.

3.3.2. When a precancer progresses to cancer, the
resulting cancer arises from cells within the precancer.

Precancers need not progress to cancer and may have a
high rate of regression, especially if the offending agent is
removed. That progression to invasive cancer can take many
years, providing an opportunity for treatment or prevention
of progression [3,13 - 18]. However, when a precancer does
progress, cancer is the obligate outcome (i.e. precancers
never progress into types of lesions other than cancer). This
biological property allows us to infer that agents that induce
precancers are carcinogens. Furthermore, precancers
become the morphological or molecular surrogates for
invasive cancer [19,20].

The definition of precancer was further limited by
exclusions. Some conditions are associated with an increased
risk of cancer but which do not contain discrete identifiable
lesions composed of cells from which the cancer arises.
Inherited conditions that raise the overall likelihood of
affected individuals developing cancer (e.g. Li-Fraumeni
syndrome, Lynch syndrome, and multiple endocrine neoplasia
syndromes) were not considered precancers. Medical
conditions that are associated with the development of cancer
are not precancers. Immunodeficiency, for instance, is not a
precursor for Kaposi’s sarcoma, malignant lymphoma, or
leiomyosarcoma [21]. Carcinogenic exposures are not
precancers. Smoking, for example, is not a precancer for
lung cancer and radiation is not a precursor for papillary
thyroid carcinoma [22]. Likewise, specific pathologic
conditions that may result in the occurrence of cancers are
not precancers. For example, cholelithiasis is not a precancer
of gallbladder carcinoma. Chronic atrophic gastritis is not the
cancer precursor of carcinoid tumors of the fundus of the
stomach. Cryptorchidism is not a precancer for germ cell
tumors. Similarly, discrete and non-discrete lesions that
portend an existing cancer that arose from another tissue (e.g.
acanthosis nigricans, sign of Leser-Trelat, and other paraneoplastic
phenomena) are not precancers.

3.3.3. A precancer is different from the normal tissue
from which it arises.

This criterion requires that a precancer has some property
that distinguishes it from normal tissue. Examples include
acquisition of a genetic, karyotypic, structural, morphologic,
or functional change in a focus of cells that offsets them from
the surrounding normal field. When cells are morphologically
atypical or when they have some cytogenetic or
mutational marker, they are different from normal. However,
it should be recognized that monoclonality or genetic
mutations do not always translate into precancerous or
cancerous lesions. Monoclonality has been demonstrated in
normal tissue, such as the breast and in subendothelial
fibroblasts and myofibroblasts. K-ras mutations are known
to occur in normal ductal epithelium and in simple ductal
hyperplasia of the pancreas [23,24].

This criterion is difficult to evaluate when a focus of cells
stands out from surrounding cells due to features considered
variant of normal. For instance, regenerative lesions in
response to inflammation or a focus of hyperplastic cells
(proliferating at a high rate) may be associated with an
increased risk of cancer. Should a hyperplastic lesion with no
other distinguishing properties (i.e. no molecular markers and
polyclonal) be considered a precancer? It was the opinion of
the Conference participants that focal, transient hyperplastic
lesions are variants of normal and should not be considered
precancers. Only if a hyperplasic focus fulfills all five
precancer criteria listed above, including the acquisition of
characteristic features that distinguish it from normal cells,
should it be reclassified as a precancer. Examples of specific
"hyperplastic" entities that can be reclassified as bona fide
precancers according to the criteria considered in this
publication include: AIDS-associated lymphoid hyperplasia,
Helicobacter-associated gastric MALT hyperplasia, atypical
ductal hyperplasia of the breast, and atypical endometrial
hyperplasia [1].

3.3.4. A precancer is different from the cancer into
which it develops, although it has some, but not all, of
the molecular and phenotypic properties that
characterize the cancer.

If it is not possible to distinguish a putative precancer
from a cancer, then the lesion should not be considered a
precancer. The inability to distinguish these two lesions is a
non-trivial requirement and addresses a specific criticism
raised by individuals who assume that cancer is a continuous
transformation from normal to malignant cells. Assumptions
include:

(1) There is no abrupt transition between biological states.

(2) Any neoplastic lesion is a heterogeneous collection of
cells representing many phenotypic states.

(3) The differences between a population of precancer cells
and a population of cancer cells tend to be superficial
and unrelated to the cancer phenotype.

(4) The features of the cancer phenotype (e.g. aneuploidy,
population atypia, and defects in differentiation) are
found in precancers.

These assumptions may be true, but in order to satisfy the
criterion that precancer cells are different from cancer cells,
it would be necessary to show a specific property that is
present in all of the precancer cells and is absent from the
cancer cells, or vice versa. The properties most often used to
distinguish precancers from cancers on a practical level are
stromal invasion and ability to spread. In general, cancer
cells are invasive and precancer cells are not invasive.
Another characteristic of many epithelial precancers is their
propensity to extend into adjacent normal glands or tubular
structures, mimicking invasion. When a precancer has
progressed to the stage where any focus of stromal invasion
is present, it is generally accepted that it has progressed to
become a fully malignant cancer, even though most of the
cells in the lesion remain non-invasive. It should be noted,
however, that in some anatomic sites the clinical behavior of
microinvasive cancers is similar to that of precancers
because they do not metastasize and can be successfully
treated with conservative surgery [4].
The criterion chosen to distinguish a precancer from a
cancer need not be morphologic. Any property that is
representative of the entire population of precancer cells that
is generally not representative of the cancer (that develops
from the precancer) would suffice.

3.3.5. There is a method by which the precancer can be
diagnosed.

At first glance, this criterion may seem superfluous. If we
can distinguish a precancer from normal tissue (criterion 3)
and from cancerous tissue (criterion 4), can we not infer that
we can diagnose the lesion? Actually, no. The problem for
the diagnostician is to distinguish precancers from normal
tissue, cancerous tissue, and from every other lesion that is
neither normal nor cancer, including atypical reactive
lesions. If there are no available modalities that permit us
to distinguish a precancer from a proliferative lesion that has
no risk of progressing to cancer, then we have failed to
diagnose the precancer. In effect, the fifth criterion imposes a
clinical (diagnostic) attribute on an otherwise hypothetical
construct. This requirement is not constrained by the
diagnostic method, which today is primarily routine
histopathologic examination of excisional biopsies. Any
single or combinatorial diagnostic modality with sufficient
sensitivity and specificity that is clinically useful may fulfill
this requirement. Functional imaging and molecular analysis
of biologic samples are examples of emergent technology
that might be useful in this regard.

4. Discussion

These five criteria represent the minimal set of conditions
for a lesion detected by any method to be considered a
precancer. These conditions are necessary and sufficient. All
of the criteria must apply, and they all must apply
concurrently (i.e. not additive over time). The different
kinds of precancers may vary in every biologic feature
except those specified in the definition (identifiable lesions
that precede the development of cancer).

The definition reached has several interesting features
that are intended to dispel some of the confusion in the field.
First, the definition has no required morphologic criteria.
Most definitions of precancers presume that these lesions
have specific morphologic features that permit them to be
recognized. The Conference participants preferred not to
comment on any specific diagnostic criteria, allowing for
lesions characterized by cytogenetic, molecular, and even
behavioral (phenotypic) properties. This opens the possibility
of developing taxonomic annotations and even new
classifications based on an assortment of criteria.
The Conference participants briefly discussed (but did
not answer) a number of fundamental questions.

4.1. Does every cancer of man have an identifiable
precancer?

Some common examples are listed in Table 1.

The list of common tumors with known precursors
provided above contains only a few examples of tumors
derived from hematopoietic or lymphoid tissues. It is our
opinion that more information regarding putative nonepithelial
precancers will emerge as newgenomic, proteomic,
and functional data are generated in these non-epithelial
models. Although the best examples of precancers today are
epithelial, our definition is sufficiently general and open
ended to be applicable to non-epithelial precancers.

4.2. Are precancers the obligate lesions preceding
cancers?

As an example, we know that adenomas are precancerous
lesions that may lead to the development of colorectal
carcinoma. Is it true that every colorectal carcinoma is
preceded by an adenoma? Are there cases of colorectal
carcinoma arising ab initio from a single transformed cell
that appeared within a population of normal cells, and which
was not associated with an identifiable precancerous lesion?
If they occur, can they be recognized in light of the tiny size
of the initial lesion?

Pathologists occasionally encounter very small cancers
that are obviously invasive and that are not accompanied by
visible remnants of precancerous lesions. Small cell
carcinoma of the lung is a rapidly growing cancer that
characteristically is not associated with a precancerous
lesion identified morphologically. Most likely, this is due to
overgrowth of the precancerous lesion by the small cell
carcinoma. In addition, there are rodent models of rapidly
induced tumors. Cancers may arise very soon after
immunosuppression of humans, sometimes in only a few
months [29]. When the induction time for viral-induced
tumors is short, it may be impossible to identify a
precancerous lesion. The practical benefits of precancer
detection and therapy are diminished when the interval
between the appearance of a precancer and its progression to
an invasive cancer is brief.

Although we know very little about precancers for nonepithelial
tumors, there is reason to hope that a subset caused
by or associated with infectious agents can be prevented
with vaccines or effective anti-viral drug therapy, such as
those that exist for epithelial tumors. Some of these viral
associated tumors can be averted if the viral infection is
prevented. Hepatitis B vaccination has greatly reduced the
incidence of liver cancer. Successful human papilloma virus
vaccines are expected to reduce the incidence of cervical
cancer.We should anticipate similar results for other human
viruses that have been associated with cancer (Epstein-Barr
virus, Herpes virus 6, HIV, and human T-cell leukemia
virus). Although we know very little about tumors that do
not arise from precancers, there is reason to suspect that
some tumors in this group also have a viral etiology and can
be controlled with vaccines or anti-viral drug therapy.

4.3. What criteria will we use to diagnose non-epithelial
precancers?

At the histological level, epithelial precancers are
relatively easy to define and to diagnose. Most are
characterized as foci of atypical cells confined within the
normal anatomic boundary of the epithelial compartment
(i.e. the basement membrane). Atypical cells that have
penetrated the basement membrane are considered malignant
because they are invasive. The term "intraepithelial
neoplasia" describes these lesions and includes specific
criteria for their diagnosis. How can we expect to distinguish
between the precancerous and cancerous stages of tumors
that derive from tissues without a confining basement
membrane (i.e. all connective tissue tumors)?
Though the Conference did not answer the questions
posed above, it would seem that precancers have distinctive
biological properties that serve to separate them from the
cancers, even if there is no intraepithelial compartment that
can be examined for invasion. Some of the general properties
of precancers that would apply to non-epithelial and
epithelial precancers are described.

4.3.1. Regression

Not all precancers will progress to cancer. Regression, in
and of itself, is a relatively uncommon phenomenon. When a
clonal, proliferative lesion has a high rate of regression,
particularly when it has a generally indolent (slow-growing)
behavior, one should suspect that the lesion is a precancer. In
a well conducted study, the regression rate of all
precancerous lesions of bronchial epithelium was 54%
[30]. The regression rate of these precancerous bronchial
lesions was found unrelated to various risk factors, such as
smoking, past history of cancer, or chronic obstructive
pulmonary disease. In cervical intra-intraepithelial neoplasia,
on the other hand, lack of progression or regression is
directly related to the degree of cytologic atypia, mitotic
activity, and type of human papillomavirus infection [31].
For unknown reasons, most in situ neuroblastomas do not
evolve into clinically apparent tumors [1]. In some nonepithelial
malignant tumors, regression may be a common
phenomenon [32]. For instance, clinical regression has been
reported in 30% of cases of untreated follicular lymphomas
[33]. Regression has rarely been reported in testicular germ
cell neoplasms, neuroblastoma, melanoma, and other
invasive cancers [34].
Regardless of the method of detection the lesions that we
designate as precancers are often members of a biologically
heterogeneous group comprised of some lesions that
progress to cancer and other lesions, usually the majority,
that persist without developing into invasive cancer or that
regress. It can be argued that the non-progressing lesions are
not true precancers in the sense that they did not progress to
cancer. Does it follow that many of the lesions in the nonprogressing
group that were designated precancers have
been misdiagnosed? Is there any value in experimental
observations (i.e. genetic markers, metabolic pathways, and
novel treatment protocols) on the regressing subset of
precancers? What is the value of finding a cure for lesions
that would have regressed without treatment?
These questions are thought-provoking, but they are also
hypothetical. At present, we have no way of distinguishing
precancers that progress from precancers (of similar
morphology) that do not progress or that regress. As far
as we know at present, the progressing and the regressing
precancers are equivalent, and factors that determine
progression or regression are most likely external to the
biology of the precancer. Even if we had markers to
distinguish the progressing from the regressing lesions,
factors that alter the balance between progression and
regression would be an important research subject.
Precancer regression is a complex and poorly understood
biological phenomenon that requires further study.
4.3.2. Precancer progression
Even though it is difficult to distinguish precancers that
progress from those that do not progress, on a practical level
again, epithelial precancers that do progress usually show
greater cytologic atypia, more mitotic activity, and more
genetic abnormalities than those that persist or regress. If a
proliferative lesion typically transforms, over time, into a
more aggressive lesion with identifiable features of the
malignant phenotype not observed in the original lesion, this
would be another reason to suspect that the original lesion is
a precancer. Examples of non-epithelial proliferative lesions
that occasionally transform into a more aggressive and
morphologically malignant tumor are shown in Table 2.

4.3.3. Multiplicity of lesions

In animal models, carcinogenic agents often produce
multiple precancers. Over time, some of these precancers
develop into cancers [35 - 37]. This is presumably an
expression of the relative ease of producing precancerous
lesions and may be accountable by a lower threshold of
mutational or oncogene-activating events needed for
precancers, as compared with cancers. The occurrence of
multiple precancers seems to hold, in many instances, in
man. An individual who has hundreds of actinic keratoses is
more likely to have one or more squamous cell carcinomas
than an individual with only a few keratoses. An individual
with hundreds of nevi will likely have a smaller number of
atypical nevi and a very small number of malignant
melanomas. Colon adenomas that develop in familial
adenomatous polyposis are often synchronous and multiple,
but it is unusual to find patients with multiple colon
carcinomas. When a proliferative lesion is multiple, it may
well be a precancer. A number of human cancers are
components of inherited neoplastic syndromes, such as
MEN type IIa. Patients with this syndrome develop a
precancerous lesion, known as C-cell hyperplasia or
medullary thyroid carcinoma in situ [38]. This precancerous
lesion is often multicentric, nearly always bilateral, and can
be detected by identification of the specific RET germline
mutation. Other genetically determined syndromes are
characterized by precancerous lesions that are multicentric
and diagnosed preoperatively by genetic testing [1,27].

4.3.4. Chronologic precedence

Progression of precancers to cancer, if it occurs, takes place
over time. Thismeans that for any given precancer, the average
age of individuals in whom the precancer occurs should be
younger than the average age of individuals in which the
developed cancer occurs. The property of chronological
precedence seems to be an inescapable truth. If this is so, then
why is chronological precedence not included among the
required defining properties of precancers?
If populations were screened at regular intervals, and if
there were methods to reliably detect precancers and
cancers, it might be feasible to use epidemiologic data to
determine the chronologic precedence of precancers.With a
few exceptions, this type of study has not been carried out.
An actinic keratosis, or an atypical nevus, may not
precipitate a trip to the dermatologist as easily as a
squamous cell carcinoma or melanoma. Precancers may be
an incidental finding in patients who are newly diagnosed
with cancer (i.e. synchronous diagnosis), which would bias
the data against finding precedence. Nonetheless, it would
be interesting to have data on the average age of first
appearance of a putative precancer in a patient population, to
compare with the average age at first appearance of the
corresponding invasive cancer in the same population.
Strong evidence of precancer precedence might serve as a
useful indicator of lesion progression for both non-epithelial
and epithelial putative precancers.

5. Conclusion

The Conference participants developed a biological
definition for precancers, providing a consistent and clinically
useful way of distinguishing precancers fromall other lesions
regardless of the level of analysis ormethod of detection. The
consistent application of the proposed criteria is intended to
support the creation of a clinically relevant taxonomy of
precancer lesions. Because every precancer included in the
taxonomy will be annotated with defining biological criteria,
it will permit the construction of a classification of precancers
based on shared features. It is expected that class properties
for the precancers will lead to the rational selection of
methods to effectively prevent or treat these important lesions.
The Conference participants urge investigators who describe
new precancer entities to include evidence indicating that
each defining criterion is satisfied.

References

[1] Henson DE, Albores-Saavedra J, eds. Pathology of incipient neoplasia.
3rd ed. New York: Oxford University Press, 2001.

[2] Berman JJ, Henson DE. Classifying the precancers: a meta data
approach. BMC Med Inform Decis Mak 2003;3:8.

[3] O’Shaughnessy JA, Kelloff GJ, Gordon GB, Dannenberg AJ, Hong
WK, Fabian CJ, et al. Treatment and prevention of intraepithelial
neoplasia: an important target for accelerated new agent development
recommendations of the American Association for Cancer Research
Task Force on the treatment and prevention of intraepithelial neoplasia.
Clin Cancer Res 2002;8:314 - 46.

[4] Grundmann E, Beck L, eds. Minimal neoplasia, in recent results in
cancer research. Berlin: Springer-Verlag, 1988.

[5] Rohan T, Franco EL. Introduction. In: Franco EL, Rohan TE, eds.
Cancer precursors: epidemiology, detection, and prevention. New
York: Springer-Verlag, 2002. 1 - 3.

[6] Berman JJ, Henson DE. The precancers: waiting for a classification.
Hum Pathol 2003;34:833 - 4.

[7] Berman JJ, Henson DE. Reply Hum Pathol 2004;35:137 - 8.

[8] Seidman JD, Berman JJ. Premalignant non-epithelial lesions: a biological
classification. Mod Pathol 1993;6:544 - 54.

[9] Ponten J. Cell biology of precancer. Eur J Cancer 2001;37:S97 - 113.

[10] Mutter GL. Diagnosis of premalignant endometrial disease. J Clin
Pathol 2002;55:326 - 31.

[11] Mutter GL. Endometrial intraepithelial neoplasia: a new standard for
diagnosing precancers. Contemp Ob/Gyn 2001;46:92 - 8.

[12] Mutter GL, Duska L, Crum CP. Endometrial intraepithelial neoplasia.
In: Crum CP, Lee K, eds. Diagnostic gynecologic and obstetric
pathology. Philadelphia: Saunders, 2006.

[13] Correa P, Fontham ET, Bravo JC, Bravo LE, Ruiz B, Zarama G, et al.
Chemoprevention of gastric dysplasia: randomized trial of antioxidant
supplements. J Natl Cancer Inst 2000;92:1881 - 8.

[14] Wargovich MJ, Jimenez A, McKee K, Steele VE, Velasco M,Woods J,
et al. Efficacy of potential chemopreventive agents on rat colon
aberrant crypt formation and progression. Carcinogenesis 2000;21:
1149 - 55.

[15] Brewer MA, Johnson K, Follen M, Gershenson D, Bast Jr R. Prevention
of ovarian cancer: intraepithelial neoplasia. Clin Cancer Res
2003;9:20 - 30.

[17] Giardiello FM, Hamilton SR, Krush AJ, Piantadosi S, Hylind LM,
Celano P, et al. Treatment of colonic and rectal adenomas with
sulindac in familial adenomatous polyposis. N Engl J Med
1993;328:1313 - 6.

[18] Boone CW, Bacus JW, Bacus JV, Steele VE, Kelloff GJ. Properties of
intraepithelial neoplasia relevant to the development of cancer chemopreventive
agents. J Cell Biochem 1997;(Suppl. 28 - 29):1 - 20.

[19] Kelloff GJ, O’Shaughnessy JA, Gordon GB, Hawk ET, Sigman CC.
Counterpoint: because some surrogate end point biomarkers measure
the neoplastic process they will have high utility in the development of
cancer chemopreventive agents against sporadic cancers. Cancer
Epidemiol Biomarkers Prev 2003;12:593 - 6.

[20] Kelloff GJ, Sigman CC, Johnson KM, Greenwald P, Crowell JA, Hawk
ET, et al. Perspectives on surrogate end points in the development of
drugs that reduce the risk of cancer. Cancer Epidemiol Biomarkers
Prev 2000;9:127 - 37.

[21] Monforte-Munoz H, Kapoor N, Albores-Saavedra J. Epstein-Barr
virus associated leiomyomatosis and post-transplant lymphoproliferative
disorder in an immunocompromised patient: case report and
review of the literature. Pediatr Dev Pathol 2003;6:449 - 57.

[22] LiVolsi VA, Albores-Saavedra J, Asa SL, Baloch ZW, Sobrinho-
Simoes M, Wenig B, et al. Papillary carcinoma. In: DeLellis RE,
ed. Pathology and genetics. Tumours of endocrine organs. Lyon: IARC
Press, 2004[World Health Organisation Classification of Tumours; 8].

[23] Tada M, Ohashi M, ShiratoriY,OkudairaT,KomatsuY, Kawabe T, et al.
Analysis of K-ras mutation in hyperplastic duct cells of the pancreas
without pancreatic disease. Gastroenterology 1996;110:227 - 31.

[24] Luttges J, Diederichs A, Menke MA, Vogel I, Kremer B, Kloppel G.
Ductal lesions in patients with chronic pancreatitis show K-ras mutations
in a frequency similar to that in the normal pancreas and lack
immunoreactivity for p53. Cancer 2000;88:2495 - 504.

[25] Greenberg AK, Yee H, Rom WN. Preneoplastic lesions of the lung.
Respir Res 2002;3:20 - 30.

[26] Bostwick DG, Qian J. High-grade prostatic intraepithelial neoplasia.
Mod Pathol 2004;17:360 - 79.

[27] Henson DE, Albores-Saavedra J. Pathology of incipient neoplasia. In:
Kelloff G, Hawk E, Sigman CC, eds. Cancer chemoprevention
strategies for cancer chemoprevention, vol. 2. Totowa, NJ: Humana
Press, 2005: 69 - 96.

[28] Hruban RH, Takaori K, Klimstra DS, Adsay NV, Albores-Saavedra J,
Biankin AV, et al. An illustrated consensus on the classification of
pancreatic intraepithelial neoplasis (and intraductal papillary mucinous
neoplasms. Am J Surg Pathol 2004;28:977 - 87.

[29] Chang YS, Kim Y, Kim DY, Kim HJ, Ahn CM, Lee DY, et al. Two
cases of post-transplant lymphoproliferative disorder in lung transplant
recipients. Korean J Intern Med 2004;19:276 - 81.

[30] Breuer RH, Pasic A, Smit EF, van Vliet E, Vonk Noordegraaf A, Risse
EJ, et al. The natural course of preneoplastic lesions in bronchial
epithelium. Clin Cancer Res 2005;11:537 - 43.

[31] Nasiell K, Nasiell M, Vaclavinkova V. Behavior of moderate cervical
dysplasia during long term follow-up. Obstet Gynecol 1983;61:
609 - 14.

[32] Krikorian JG, Portlock CS, Cooney P, Cooney P, Rosenberg SA.
Spontaneous regression of non-Hodgkin’s lymphoma: a report of nine
cases. Cancer 1980;46:2093 - 9.

[33] Horning SJ, Rosenberg SA. The natural history of initially untreated
low grade non-Hodgkin’s lymphoma. N Engl J Med 1984;311:1471 - 5.

[34] Simpson K, Albores-Saavedra J. Unusual findings in papillary thyroid
microcarcinoma suggesting partial regression. A study of two cases.
Ann Diag Pathol, in press.

[35] McDonnell TJ, Korsmeyer SJ. Progression from lymphoid hyperplasia
to high-grade malignant lymphoma in mice transgenic for the t(14;18).
Nature 1991;349:254 - 6.

[36] Solt DB, Medline A, Farber E. Rapid emergence of carcinogeninduced
hyperplastic lesions in a new model for the sequential analysis
of liver carcinogenesis. Am J Pathol 1977;88:595 - 618.

[37] Kirkpatrick CJ, Alves A, Kohler H, Kriegsmann J, Bittinger F, Otto M,
et al. Biomaterial-induced sarcoma: a novel model to study preneoplastic
change. Am J Pathol 2000;156:1455 - 67.

[38] Albores-Saavedra J, Krueger J. C-cell hyperplasia and medullary
thyroid microcarcinoma. Endocr Pathol 2001;12:365 - 77.