Professional Documents
Culture Documents
[FREE PDF sample] Clinical Ophthalmic Oncology Basic Principles Arun D. Singh ebooks
[FREE PDF sample] Clinical Ophthalmic Oncology Basic Principles Arun D. Singh ebooks
OR CLICK LINK
https://textbookfull.com/product/clinical-
ophthalmic-oncology-basic-principles-arun-d-singh/
Read with Our Free App Audiobook Free Format PFD EBook, Ebooks dowload PDF
with Andible trial, Real book, online, KINDLE , Download[PDF] and Read and Read
Read book Format PDF Ebook, Dowload online, Read book Format PDF Ebook,
[PDF] and Real ONLINE Dowload [PDF] and Real ONLINE
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
https://textbookfull.com/product/imaging-in-clinical-oncology-
athanasios-d-gouliamos/
https://textbookfull.com/product/allergy-and-sleep-basic-
principles-and-clinical-practice-anna-fishbein/
https://textbookfull.com/product/surgical-ophthalmic-oncology-a-
collaborative-open-access-reference-1st-edition-sonal-chaugule-
editor/
https://textbookfull.com/product/clinical-gynecologic-oncology-
philip-j-disaia/
Abeloff's Clinical Oncology 6th Edition Niederhuber
https://textbookfull.com/product/abeloffs-clinical-oncology-6th-
edition-niederhuber/
https://textbookfull.com/product/clinical-radiation-
oncology-4e-fourth-edition-bogart/
https://textbookfull.com/product/gunderson-teppers-clinical-
radiation-oncology-joel-tepper/
https://textbookfull.com/product/international-manual-of-
oncology-practice-imop-principles-of-medical-oncology-1st-
edition-ramon-andrade-de-mello/
https://textbookfull.com/product/practical-radiation-oncology-
physics-a-companion-to-gunderson-teppers-clinical-radiation-
oncology-1e-sonja-dieterich-phd/
Clinical Ophthalmic
Oncology
Basic Principles
Arun D. Singh
Bertil E. Damato
Editors
Third Edition
123
Clinical Ophthalmic Oncology
Arun D. Singh • Bertil E. Damato
Editors
Clinical Ophthalmic
Oncology
Basic Principles
Third Edition
Editors
Arun D. Singh Bertil E. Damato
Department of Ophthalmic Oncology Nuffield Department of Clinical
Cole Eye Institute, Cleveland Clinic Neurosciences, University of Oxford
Cleveland, OH John Radcliffe Hospital
USA Oxford, UK
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
Ophthalmic tumors are rare and diverse so that their diagnosis can be quite
complex. Treatment usually requires special expertise and equipment and in
many instances is controversial. The field is advancing rapidly, because of
accelerating progress in tumor biology, pharmacology, and instrumentation.
Increasingly, the care of patients with an ocular or adnexal tumor is provided
by a multidisciplinary team, consisting of ocular oncologists, general oncolo-
gists, radiotherapists, pathologists, psychologists, and other specialists.
For all these reasons, we felt that there was a need for the new edition of
the textbook providing a balanced view of current clinical practice. Although
each section of Clinical Ophthalmic Oncology, 3rd Edition, now represents a
stand-alone volume, each chapter has a similar layout with boxes that high-
light the key features, tables that provide comparison, and flow diagrams that
outline therapeutic approaches.
The enormous task of editing a multiauthor, multivolume textbook could
not have been possible without the support and guidance by the staff at
Springer: Caitlin Prim, Melanie Zerah, ArulRonika Pathinathan, and Karthik
Rajasekar. Michael D. Sova kept the pressure on to meet the production
deadlines.
It is our sincere hope that our efforts will meet high expectation of the
readers.
v
Acknowledgments
Arun D Singh
Bertil E Damato
vii
Contents
ix
x Contents
Index���������������������������������������������������������������������������������������������������������� 317
Contributors
xi
xii Contributors
Research question(s)
Other considerations
• Ethical aspects
• Practical limitations
• Budget constraints
quality of life measures have become increasingly hand. Broadly speaking, mortality rate refers to the
popular. In ophthalmic oncology, visual acuity is incidence of death, and survival rate is its comple-
also an important outcome measure. ment, i.e., survival rate = 100 – mortality rate.
Population mortality is the chance that a per-
son in the general population will die from a spe-
Prevalence cific disease over a specified time frame. It is a
useful concept for measuring the burden of dis-
Prevalence refers to the proportion of the study ease in a population. For example, the population
population with the condition of interest. Usually mortality for heart disease was 197.2 per 100,000
prevalence is given for a specific moment in time population per year in 2015. Therefore, it is a
(point prevalence), but sometimes it is estimated measure more important for public health policy-
for a period of time (e.g., 1 year or lifetime preva- makers as opposed to clinicians. This measure is
lences). For example, the lifetime prevalence of calculated from death certificates, where cause of
uveal melanoma in a Caucasian population with death is known [6].
oculo(dermal) melanocytosis is estimated to be Overall mortality is the chance that a person
0.26% [4]. with a disease will die within a time period after
diagnosis. It is important to specify a time period
for the mortality statistics – in the long run, mor-
Incidence tality is 100% for any condition. Of note, this
definition is indifferent to cause of death. Overall
Whereas prevalence relates to existing cases, mortality is the most common measure of mortal-
incidence relates to the proportion of new cases ity in the literature and is often used to guide
in the study population. It is important that the prognosis. This measure is helpful in identifying
population under investigation is at risk of devel- risk factors for poor prognosis, as well as measur-
oping the condition. For example, persons with ing disparities between populations. The inter-
bilateral enucleation are no longer at risk of pretation of this measure is complicated by
developing uveal melanoma. There are two dif- biases, including lead time, length, and overdiag-
ferent measures of incidence: cumulative inci- nosis biases.
dence (CI) and incidence density (ID). CI is the Cause-specific mortality is the chance that a
proportion of new cases in a population at risk person with a disease will die within a time
over a specified period of time. For example, the period after diagnosis due to the disease. This is
CI of second malignant neoplasms in hereditary in contrast to overall survival, which does not dis-
retinoblastoma patients is 17% at the age of tinguish between causes of death. Cause-specific
35 years [5]. ID refers to the rate of developing mortality most closely measures the “deadliness”
the condition during follow-up, usually expressed of a disease, but similar to overall mortality, it
as a proportion per person-year at risk. can be affected by lead time, length, and overdi-
agnosis bias.
Relative mortality is a proportion that com-
Mortality pares the overall mortality of people with a dis-
ease to that of an unaffected, but otherwise
Cancer is among the leading causes of mortality. identical population. Relative mortality measures
In order to understand the processes that either the excess mortality associated with a diagnosis
hasten or delay this outcome, it is necessary to rig- compared to the general population. It is a conve-
orously define the burden of disease. Clinical epi- nient measure to calculate because it does not
demiologists have created several concepts of require cause of death to be recorded. It is also a
mortality, all with their own definition, interpreta- helpful measure for understanding the deadliness
tion, and uses. Unfortunately, many of these con- of a disease that affects the elderly, where patients
cepts use similar nomenclature, so it is important are at risk from dying from other causes. To cal-
to always clarify the definition of mortality at culate relative survival, overall survival of a
4 A. C. Moll et al.
cohort of patients is compared to life tables of the endpoints argue that they are preferable for rea-
general population, matched by age, sex, race, sons of feasibility, cost, and length of study, with
and other important demographic features. possible expediency of the drug approval pro-
Conditional mortality is another clinically cess. Critics of surrogate endpoints point out that
informative measure of mortality, which mea- the evidence base for surrogate endpoints corre-
sures the chance that a person with a disease will lating closely with clinical outcomes is often
die within a specified time period after having tenuous.
already lived with the disease for a certain amount
of time. It is a helpful measure to assess how
prognosis changes over time. Conditional sur- Quality of Life
vival requires a cohort of patients where long-
term follow-up is close to complete. With increasing survival rates and with severe
side effects of some treatment modalities, quality
of life measures have become increasingly
Time to Event important in ophthalmic oncology. These mea-
sures encompass symptoms and physical, social,
Another class of outcomes that is of interest to and psychological functioning from a patient’s
clinical epidemiologists is time to event. Time to perspective. Usually, quality of life is assessed
event measures the length of time that elapses with a structured questionnaire, and scores are
from some start point to a defined endpoint, the summarized assuming an interval scale. Several
most common time-to-event measure being sur- questionnaires have recently been developed for
vival, which is defined as time from diagnosis to patients with ocular diseases, such as the measure
death (not to be confused with survival rate). of outcome in ocular disease [MOOD]) [9].
Other events can be used as the endpoint, such as
relapse or radiographic progression. Time-to-
event data is typically graphically presented with M easures of Association
a Kaplan-Meier plot, which displays a nonpara-
metric estimation of the rate of survival in a pop- In epidemiological research, we are usually inter-
ulation as a function of time [7]. ested in how certain interventions or exposures
are associated with outcomes; for example, is
there an association between paternal age and
Surrogate Endpoints retinoblastoma in the offspring? [10]. There are
several statistical approaches that can be used to
While overall mortality and cause-specific mor- quantify associations, either as a ratio or as a dif-
tality are the two gold standard measures of mor- ference, depending upon the study design and
tality when evaluating the efficacy of statistical method used (Table 1.2).
interventions, other measures of efficacy are
being proposed, such as surrogate endpoints,
which are biomarkers that are intended to substi- Relative Risk
tute for a clinical endpoint. Of note, these are not
clinical quantities of interest but are correlated The ratio of cumulative incidences of exposed
with them. Examples of surrogate endpoints and unexposed individuals (or between treated
include evidence of radiographic progression, and untreated patients) is the relative risk (RR).
biochemical markers, and physical signs [8]. For example, in the Netherlands, the RR of reti-
These are controversial measures and are not uni- noblastoma in children conceived by in vitro fer-
formly accepted within the scientific and regula- tilization is between 4.9 and 7.2. This implies that
tory community. Proponents of surrogate the risk of getting retinoblastoma is between 4.9
1 Principles of Cancer Epidemiology 5
Table 1.2 The relation between outcome, measures of association, study designs, and statistical methods
Measure of
Outcome association Computation Study designs Statistical methods
Prevalence Prevalence rate P1/P2 Cross-sectional Chi-square test
Logistic regression
analysis
Prevalence P1 − P2 Cross-sectional Chi-square test
difference
Odds of exposure Odds ratio Odds of exposure group 1/ Case-control study Chi-square test
odds of exposure group 2 (cohort study, RCT) Logistic regression
Cumulative Relative risk CI1/CI2 Cohort study/RCT Chi-square test
incidence (CI) Risk difference CI1 − CI2 RCT
Incidence density Hazard ratio ID1/ID2 Cohort study/RCT Kaplan-Meier
(ID) Cox regression
Risk difference ID1 − ID2 RCT Kaplan-Meier
O/E ratio Observed ID/expected ID in Cohort study/registry
general population study
Quality of life Difference in X1 − X2 Cohort study/RCT Independent t-test
mean score Linear regression
analyses
P1 prevalence group 1, P2 prevalence group 2, CI cumulative incidence, CI1 CI group 1, CI2 CI group 2, ID incidence
density, ID1 ID group 1, ID2 ID group 2, O/E ratio observed to expected ratio, RCT randomized controlled trial,
X1 = mean score group 1; X2 = mean score group 2
and 7.2 times higher for children conceived after expected. This data can be expressed as stan-
IVF than naturally conceived children. dardized mortality ratio of 5.41 [11].
The ratio of incidence densities of unexposed The odds ratio (OR) is the most commonly
and exposed patients (or between treated and reported measure of association in the literature,
untreated patients) is the hazard ratio (HR), due to the fact that this is the statistic that can be
which has a similar interpretation as the RR. This derived from the popular logistic regression anal-
measure is often used in relation to mortality, ysis. The OR is the ratio of the odds of outcome
because we are generally interested not only in of interest between the exposed and the
the proportion of patients that die but also in the unexposed. Generally speaking, the OR is a good
time from baseline (diagnosis or start of treat- approximation of the RR or HR.
ment) until death. A special application of the
HR is the ratio of the observed to the expected
number of cases (O/E ratio). In this case the Differences in Risk
observed incidence density is calculated for the
study population, and this is compared to the Differences in risks (RD) are preferably reported as
expected incidence density derived from a popu- outcome in randomized controlled trials. The RD is
lation registry (e.g., cancer registration). For easy to interpret and can be used to calculate the
example, in a study of lifetime risks of common number of patients needed to treat (NNT) to pre-
cancers among 144 hereditary retinoblastoma vent one extra event (e.g., death) compared to the
survivors, 41 cancer deaths were observed, standard treatment or placebo. The NNT can be
whereas only 7.58 deaths due to cancer were calculated as inverse of RD (1/RD). A related
6 A. C. Moll et al.
concept is that of the number needed to screen Therefore, interpretation of findings should never
(NNS). This refers to the number of patients needed solely rely on statistical significance.
to screen to prevent one extra event compared to
the situation without a screening program. The
NNS thus depends on the predictive probability of Bias
the screening test as well as on the efficacy of treat-
ment for people that are diagnosed with that screen- An estimate can be very precise but nevertheless
ing test. The value of routine neuroimaging inaccurate, because of bias. Three main sources
screening of pineoblastoma in retinoblastoma of bias exist: confounding, selection, and infor-
patients is uncertain and a point of discussion [12]. mation bias.
For scores on interval scales, such as quality of Confounding occurs when the association
life, differences in mean score between exposed between exposure and outcome is influenced by a
and unexposed participants are the most impor- third variable that is related both to the exposure
tant measure of interest. These can be derived and the outcome (Fig. 1.2). A recent study found
from independent samples t-test of general linear an association between cooking (as occupation)
models (e.g., linear regression analysis). and the incidence of ocular melanoma [13]. It
could be argued that as many cooks work at night,
it is possible that they could have relatively high
Precision of the Estimate exposures to sunlight due to daytime leisure
activities compared to people working during the
When interpreting an outcome, we do not only daytime. It is implied that the association between
want to know the numerical value of the point cooking and ocular melanoma could potentially
estimate but also the precision with which it has (in part) be explained by a higher exposure to
been assessed. In other words, can we be confi- sunlight by cooks.
dent that the outcome is not just a chance find-
ing? The usual standard for accepting an outcome
as being beyond chance is p (probability) <0.05. Selection Bias
A more informative description is provided by
the 95% confidence interval (CI). The rough Selection bias may occur when the chance of
interpretation of the 95% CI is that there is a 95% being included in the study population is not ran-
probability that the real value lies within the con- dom for all members of the source population. For
fidence interval. example, patients with advanced tumor stage are
Statistical significance and the width of confi- more likely to be referred to a special cancer cen-
dence intervals are strongly dependent on the
sample size of a study. This means that in very
Exposure
large samples, weak (and potentially unimport- Disease
(intervention)
ant) associations can be statistically significant.
In contrast, in small samples, strong (and poten- r≠0
tially important) associations are sometimes not
r≠0
statistically significant. Such findings should of
course not be dismissed as being irrelevant. Confounder
Instead they should be replicated in larger study
r = correlation
populations. Associations, although statistically
significant, need not be clinically important. Fig. 1.2 Schematic representation of confounding
1 Principles of Cancer Epidemiology 7
ter than patients with a less advanced tumor stage. adopted in order to address a research question.
This form of selection bias is called referral bias. Each of the study designs has its advantages and
Selection bias could also be introduced in a study disadvantages (Table 1.3).
by choosing the wrong control group, especially
if controls are selected from hospital patients.
Case Series
associated histologically with increased density pants are randomly assigned to the intervention
of choroidal melanocytes (p = 0.005). The authors group (treatment under investigation) or a control
concluded that increased choroidal pigmentation, group (no treatment, placebo, or standard treat-
as a result of an increase in the density of pig- ment). After the start of a treatment, patients
mented choroidal melanocytes, is not protective often get better. This may be due to the treatment
but may actually be a risk factor for the develop- or to other circumstances such as spontaneous
ment of posterior uveal melanoma in white resolution, effective co-interventions, and pla-
patients [14]. cebo effects. Only a sufficiently large random-
The cross-sectional study design has the ized, blinded trial is useful to estimate the efficacy
advantage that it is relatively easy to plan, that of drugs and other treatments. The best compari-
only one measurement is needed, and that it is son is often between the new treatment and the
inexpensive and quick to perform. From a meth- best available one, not the sham treatment [15].
odological point of view, however, the design has The randomization, if successful, ensures that
some disadvantages. As both exposure and out- confounding factors are evenly distributed
come are measured at the same time, we cannot between the intervention and control groups. As
be sure that the exposure preceded the outcome with the cohort studies, incident cases in both
(the most important criterion for causality). groups are determined during or at the end of
Moreover, the outcome is always measured in follow-up, allowing for the risk estimates to be
terms of prevalent cases, and prevalent cases may calculated.
have a relatively better prognosis (they are still For clinicians interested in evidence pertain-
alive) than the incident cases. Therefore, the ing most directly to a particular class of patients,
associations found in a cross-sectional study can subgroup analyses can be very informative. The
only rarely be interpreted as being causal. strength of evidence for subgroup effects depends
on the question whether hypotheses have been
defined prior to analysis, whether potential prob-
Cohort Study lems regarding multiple comparisons have been
considered, and whether there is biological plau-
Some of the problems listed above can be over- sibility of the effects found. Using these guide-
come by conducting a (prospective) cohort study. lines, the reader of a trial report should be able to
At baseline, one starts with a cohort of people decide if presented subgroup effects are of clini-
free from the outcome of interest. During or at cal importance or if the overall result is a better
the end of follow-up, incident cases in both the estimate of treatment effect [16].
unexposed and the exposed groups are identified,
and RRs or HRs can be calculated. Despite theo-
retical advantages of a cohort design, there are Case-Control Study
some practical disadvantages. The cohort studies
often need large sample sizes and/or long follow- In contrast to cohort studies, the starting point in
up to accumulate enough incident cases for
case-control studies is not to assess the exposure
meaningful analyses. These studies are often status, but the disease status. People with the dis-
expensive. From a methodological point of view, ease of interest are selected, and a control group
the potential bias of (residual) confounding can of people without the disease is subsequently
never be totally excluded. recruited. The control group should include peo-
ple from the same source population as the cases,
implying that if any of the controls had developed
Randomized Controlled Trial the disease, they would have been eligible for
inclusion in the study as a case.
Randomized controlled trials are a specific type The selection of a valid control group is
of cohort study. At the start of the study, partici- important in case-control studies and has there-
1 Principles of Cancer Epidemiology 9
the Dutch retinoblastoma register 1862–1994. Hum analysis of studies including 1,066 patients. Acta
Genet. 1996;98:109–12. Ophthalmol Scand. 1999;77:414–7.
11. Fletcher O, Easton D, Anderson K, et al. Lifetime
risks of common cancers among retinoblastoma sur-
vivors. J Natl Cancer Inst. 2004;96:357–63.
12. Moll AC, Imhof SM, Schouten-van Meeteren AYN, Further Readings
et al. Screening for pineoblastoma in patients with
retinoblastoma. Arch Ophthalmol. 2001;120:1774. Bouter LM, Zielhuis GA, Zeegers MP. Textbook of epi-
13. Harbour JW. What is the best treatment for retinoblas- demiology. Houten: Bohn Stafleu van Loghum; 2018.
toma? Am J Ophthalmol. 2004;138:471–3. Fletcher RH, Fletcher SW, Wagner EH. Clinical epidemi-
14. Harbour JW, Brantley MA Jr, Hollingsworth H, et al. ology: the essentials. Baltimore: Williams & Wilkins;
Association between choroidal pigmentation and 1996.
posterior uveal melanoma in a white population. Br J Guyatt G, Rennie D, editors. Users’ guide to the medical
Ophthalmol. 2004;88:39–43. literature: a manual for evidence-based clinical prac-
15. Abel U, Koch A. The role of randomization in clini- tice. Chicago: AMA Press; 2002.
cal studies: myths and beliefs. J Clin Epidemiol. Rothman KJ. Epidemiology: an introduction. New York:
1999;52:487–97. Oxford University Press; 2002.
16. Oxman AD, Guyatt GH. A consumer’s guide to sub- Sackett DL, Strauss SE, Richardson WS, et al. Evidence-
group analyses. Ann Intern Med. 1992;116:78–84. based medicine. Edinburgh: Churchill Livingstone;
17. Lasky T, Stolley PD. Selection of cases and controls. 2000.
Epidemiol Rev. 1994;16:6–17. User’s guide to evidence-based practice. Educative series
18. Friedman LM, Furberg CD, De Mets of articles from the JAMA. www.cche.net/users-
DL. Fundamentals of clinical trials. St. Louis: Mosby; guides/main.asp.
1996. Website of the centre for evidence-based medicine,
19. Seregard S. Long-term survival after ruthenium Oxford, United Kingdom. www.cebm.net.
plaque radiotherapy for uveal melanoma. A meta-
Etiology of Cancer
2
Brian T. Hill
Cancer is a pathologic accumulation of clonally Although a clearly defined cause is not apparent
expanded cells derived from a common precur- in the majority of de novo malignancies, it is well
sor. The fundamental cause of all cancers is documented that carcinogenic agents contribute
genetic damage, which is usually acquired but is to many human cancers. These fall into several
sometimes congenital. In general, the genetic broad groups, which can be categorized as infec-
dysregulation that gives rise to uncontrolled cell tive (e.g., viruses), chemical (including occupa-
proliferation results from activation of growth- tional, environmental, and therapeutic),
promoting oncogenes and/or deletion/inactiva- electromagnetic radiation (i.e., ultraviolet [UV],
tion of growth-inhibiting tumor suppressor genes. X-ray, gamma ray), and immunosuppressive
Additional contributions to carcinogenesis come (HIV, immunosuppressive medications). These
from genes that regulate programmed cell death are shown in Table 2.1.
(apoptosis) and genes involved in DNA repair.
The most widely accepted theory of cancer devel-
opment is the Knudson “2-hit” hypothesis, which Chemical Carcinogens
posits that a mutation in one predisposing gene is
necessary but not sufficient for malignancy and Several hundred chemicals have been shown to
that only after development of a second mutation be carcinogenic in humans. Harm from these
will invasive cancer develop (Fig. 2.1) [1]. In chemical carcinogens can be a result of occupa-
addition, it has become increasingly apparent that tional (asbestos, aniline dyes), environmental
cancer cells must evade host immune responses (alcohol, tobacco), or iatrogenic (chemotherapy)
and this can sometimes be exploited to treat can- exposure. While many carcinogens are directly
cer. In this chapter, we discuss the major catego- mutagenic to DNA, most carcinogens undergo
ries of carcinogens and their role in the etiology activation after exposure to reactive metabolites
of cancers, with an emphasis on common oph- that are responsible for genetic damage.
thalmic cancers.
Environmental Exposure
B. T. Hill (*)
Department of Taussig Cancer Institute, Cleveland Of the known environmental carcinogens, per-
Clinic, Cleveland, OH, USA haps the most well-documented agent known to
e-mail: hillb2@ccf.org
cause cancer is asbestos. This is a naturally bladder cancer [3]. Aflatoxin and vinyl chloride
occurring mineral that has a broad range of are associated with liver cancers. An exhaustive
industrial and commercial applications, but aero- list of known environmental carcinogens is
solized fibers of asbestos are known to lodge in beyond the scope of this text but should be con-
small airways, causing tissue damage and signifi- sidered when obtaining the occupational and
cantly predisposing exposed individuals to both social history of patients with newly diagnosed
lung cancer as well as mesothelioma [2]. cancer.
A number of other agents have been associ-
ated with elevated risks of various cancers
through epidemiologic studies. For example, sev- Behavioral Exposure
eral agents have been implicated in lung cancer,
including heavy metals [arsenic, cadmium, chro- Tobacco
mium, and nickel, as well as BCME (bischloro- Tobacco is responsible for over 30% of the cancer
methl ether)] [3]. Aromatic amines such as deaths and represents the single most common
4-AMP, 4-aminobiphenyl, naphthylamine, and cause of preventable cancer. Cigarettes as well as
benzidine are associated with an increased risk of smokeless tobacco are associated with cancers of
2 Etiology of Cancer 13
multiple organs, including the oral cavity, phar- cer, including squamous cell carcinoma, basal
ynx, larynx, lung, esophagus, stomach, pancreas, cell carcinoma, and melanoma [7]. In the eye,
colon, rectum, kidney, bladder, ureter, and cervix. UV exposure increases the risk of ocular surface
Over 60 carcinogens have been identified in squamous neoplasia (OSSN), which includes
tobacco, with the highest carcinogenic potency dysplasia, carcinoma in situ, and squamous cell
attributed to polycyclic aromatic hydrocarbons, carcinoma of the conjunctiva or cornea.
nitrosamines, and aromatic amines [4]. The risk of In the case of uveal melanoma, sunlight expo-
tobacco-associated cancer is proportional to total sure has not been implicated as a cause of this con-
lifetime cigarette smoke exposure. The risk of lung dition, known risk factors including Caucasian
cancer among long-term heavy smokers is approx- race, light skin color, blond hair, and blue eyes [8].
imately 10–20 times greater than non-smokers.
• You pay a royalty fee of 20% of the gross profits you derive
from the use of Project Gutenberg™ works calculated using
the method you already use to calculate your applicable
taxes. The fee is owed to the owner of the Project
Gutenberg™ trademark, but he has agreed to donate
royalties under this paragraph to the Project Gutenberg
Literary Archive Foundation. Royalty payments must be paid
within 60 days following each date on which you prepare (or
are legally required to prepare) your periodic tax returns.
Royalty payments should be clearly marked as such and
sent to the Project Gutenberg Literary Archive Foundation at
the address specified in Section 4, “Information about
donations to the Project Gutenberg Literary Archive
Foundation.”
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.F.
1.F.4. Except for the limited right of replacement or refund set forth in
paragraph 1.F.3, this work is provided to you ‘AS-IS’, WITH NO
OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.
Please check the Project Gutenberg web pages for current donation
methods and addresses. Donations are accepted in a number of
other ways including checks, online payments and credit card
donations. To donate, please visit: www.gutenberg.org/donate.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.