Professional Documents
Culture Documents
Clinical Ophthalmic Oncology Collection Uveal Basic Retinoblastoma Eyelids Conjuctiva 2019 5 Books Retinal 3rd Ed Jacob Pe'Er
Clinical Ophthalmic Oncology Collection Uveal Basic Retinoblastoma Eyelids Conjuctiva 2019 5 Books Retinal 3rd Ed Jacob Pe'Er
https://ebookmeta.com/product/gynecological-oncology-basic-
principles-and-clinical-practice-kavita-singh-editor/
https://ebookmeta.com/product/advances-in-radiation-oncology-in-
lung-cancer-3rd-3rd-edition-branislav-jeremic-ed/
https://ebookmeta.com/product/berserk-
volume-1-5-collection-5-books-set-series-1-1st-edition-kentaro-
miura/
https://ebookmeta.com/product/the-rough-guide-to-cyprus-3rd-
ed-2019-3rd-edition-rough-guides/
Basic Radiation Oncology 2nd Edition Murat Beyzadeoglu
https://ebookmeta.com/product/basic-radiation-oncology-2nd-
edition-murat-beyzadeoglu/
https://ebookmeta.com/product/gynecologic-oncology-handbook-an-
evidence-based-clinical-guide-3rd-edition-michelle-benoit-md/
https://ebookmeta.com/product/abeloffs-clinical-oncology-john-e-
niederhuber-editor/
https://ebookmeta.com/product/presley-thurman-cozy-mystery-
collection-box-set-books-0-5-4-l-c-tuner/
https://ebookmeta.com/product/physics-for-clinical-oncology-
second-edition-amen-sibtain/
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.
i nflammation, impairment of host immunity, and, viral EBV has been identified in primary intra-
in some cases, a modulation of the balance ocular and ocular adnexal lymphomas, this find-
between proliferation and antiproliferation ing is of uncertain significance [13].
signals.
aposi Sarcoma Herpes Virus
K
Kaposi sarcoma herpes virus (KSHV), also
Viruses known as human herpesvirus 8, is the causative
agent of Kaposi sarcoma (KS). KS can affect the
Several DNA and RNA viruses are carcinogenic. skin of any site but can also involve the eyelid
Human papillomavirus (HPV), Epstein-Barr and the conjunctiva. This malignancy is rare in
virus (EBV), hepatitis viruses B and C, and immunocompetent individuals. Its incidence
Kaposi sarcoma herpesvirus (KSHV) exemplify increased significantly during the AIDS epidemic
the best-described oncogenic viruses. of the 1980s. Since the introduction of highly
active antiretroviral therapy (HAART), the inci-
uman Papillomavirus (HPV)
H dence KS has decreased significantly.
HPV was one of the first viruses to be identified as
causing human cancer. Almost all of the 200+ gen- Human T-Cell Lymphotropic Virus-1
otypes of HPV infect epithelial cells, causing Human T-cell lymphotropic virus type 1 (HTLV-
benign papillomas (warts) or squamous cell carci- 1) infection is endemic in Asia, Africa, and much
noma of the oral, laryngeal, cervical, and anogeni- of Latin America. The most well-described can-
tal regions. The strongest associations with cer associated with HTLV-1 is T-cell leukemia/
invasive cervical carcinoma are with HPV sub- lymphoma. Known ophthalmic manifestations of
types 16 and 18, which can be identified in approx- HTLV-1 include ocular infiltrates in patients with
imately 85% of cases. HPV causes uncontrolled adult T-cell leukemia/lymphoma as well as non-
cellular proliferation by inactivating the tumor malignant ocular complications, such as retinal
suppressor proteins p53 and pRB. DNA of HPV degeneration, neuro-ophthalmic disorders, uve-
with high oncogenic potential has been identified itis, necrotizing retinal vasculitis, and keratocon-
in dysplastic and malignant lesions of the conjunc- junctivitis sicca [14].
tiva and cornea suggesting that HPV may contrib-
ute to the development of OSSN [11, 12]. epatitis B and C Virus
H
Chronic infections by hepatitis B or hepatitis C
Epstein-Barr Virus viruses are well-defined risk factors for hepatocel-
Epstein-Barr virus (EBV) has been implicated in lular carcinoma. In addition, extranodal marginal
the pathogenesis of a number of cancers includ- zone lymphoma is increased in patients with hep-
ing the African form of Burkitt lymphoma, B-cell atitis C. Importantly, from the standpoint of oph-
lymphomas in immunosuppressed individuals thalmic cancers, there have been conflicting
(having a condition such as posttransplant lym- reports of an increased incidence of hepatitis C in
phoproliferative disorder [PTLD]), and nasopha- patients with ocular adnexal lymphoma [15, 16].
ryngeal carcinoma. The mechanism by which
EBV transforms B lymphocytes is not clearly
defined but most likely involves viral protein Bacteria
expression that leads to a viral latency program
associated with proliferation. This is particularly elicobacter pylori
H
relevant to carcinogenesis in the absence of a host Helicobacter pylori (H. pylori) infection is asso-
antiviral response, as with HIV-associated lym- ciated with peptic ulcer disease, gastric carci-
phoma and/or PTLD. EBV infection is not clearly noma, and mucosa-associated lymphoid tissue
linked to ophthalmic cancer. Although DNA from (MALT) lymphoma of the stomach [17, 18]. It is
2 Etiology of Cancer 15
hypothesized that chronic inflammation from the nodeficiency syndrome (AIDS), before the
infection leads to atrophic gastritis and conse- advent of highly active antiretroviral therapy
quent achlorhydria, gastrointestinal bacterial (HAART) [20]. AIDS causes several subtypes of
overgrowth, excess reduction of nitrates to lymphoma as well as Kaposi’s sarcoma.
nitrites, DNA damage, and, thus, carcinogenesis Furthermore, cancers arising in the context of
in gastrointestinal epithelium and mucosa- chronic inflammation as well as those with a high
associated lymphoid tissue. In the case of iso- mutational burden, such as UV-associated malig-
lated gastric MALT lymphoma, H. pylori nancies (e.g., melanoma) and smoking-related
eradication can result in complete and durable lung cancer, evade T-cell surveillance by MHC
remission, thereby obviating the need for chemo- downregulation or by upregulation of checkpoint
therapy or radiation. inhibitors [21].
hlamydia psittaci
C
A number of human diseases, including cancers, Dietary Factors
are strongly associated with infection by
Chlamydia. C. trachomatis and C. pneumoniae A significant proportion of cancer deaths can be
have been shown to be associated with cervical attributed at least partially to dietary factors.
carcinoma and lung cancer. In addition, C. psit- Diets with excessive calories from saturated fat
taci has been associated with ocular adnexal lym- are known to increase the risk of hormone-
phomas, although there appears to be geographic dependent cancers such as breast, ovarian, and
variation in the strength of this association [19]. endometrial cancers in females and prostate can-
cer in males. In contrast, a low-fiber diet is asso-
ciated with an increased risk of colorectal cancer,
Fungus whereas high dietary intake of red meat has been
strongly associated an increased risk [22].
Aspergillus flavus is a fungus found in grains and
peanuts. It produces aflatoxins that cause liver
cancer, with the highest incidence occurring in
the developing world, China and South Africa. Genetic Susceptibility
In patients who have received chemotherapy cult to quantitate the individual contribution of
alone, the rate of secondary malignancy may be several risk factors, but this should change, thanks
as high as 10%. Hodgkin lymphoma, testicular to the impressive genetic advances that are taking
cancer, small cell lung cancer, and other cancers place (Fig. 2.2). Immune evasion by cancer cells
with higher cure rates are associated with higher is an inherent requirement to their survival. It is
rates of secondary malignancy, because of longer noteworthy that, with few exceptions, most cancer
survival. The most carcinogenic agents are alkyl- treatment is independent of the etiology.
ating agents such as cyclophosphamide and
topoisomerase inhibitors such as etoposide.
References
Radiation Therapy 1. Nordling CO. A new theory on cancer-inducing
mechanism. Br J Cancer. 1953;7(1):68–72.
Therapeutic radiation causes DNA damage and 2. Mossman BT, Gee JB. Asbestos-related diseases. N
Engl J Med. 1989;320(26):1721–30.
can result in a variety of cancers, including sar- 3. International Agency for Research on Cancer. IARC
coma, breast cancer, and hematologic malignan- monographs on the evaluation of carcinogenic risks
cies such as myelodysplastic syndrome and acute to humans. Overall evaluation of carcinogenic risks
leukemia. In patients with retinoblastoma, the to humans. 1971-current. Monographs 1–76. http://
publications.iarc.fr/Book-And-Report-Series/Iarc-
increased risk of sarcomas after radiation is well Monographs-On-The-Evaluation-Of-Carcinogenic-
documented [23]. Risks-To-Humans?sort_by=year_desc&limit=
50&page=1.
4. Department of Health and Human Services, Public
Health Service, Office of Surgeon General. How
Summary tobacco smoke causes disease: the biology and behav-
ioral basis for smoking-attributable disease: a report
The development of cancer can only rarely be of the Surgeon General. Atlanta: Centers for Disease
attributed to a single cause. Genetic predisposi- Control and Prevention (US); 2010. ISBN-13: 978-0-
16-084078-4, PMID 21452462.
tion confers significant risks for developing can- 5. Kraemer KH, Lee MM, Andrews AD, et al. The role
cer in a minority of patients. Risk can be increased of sunlight and DNA repair in melanoma and non-
by carcinogen exposure, immunosuppression, and melanoma skin cancer. The xeroderma pigmentosum
environmental/iatrogenic causes. It can be diffi- paradigm. Arch Dermatol. 1994;130(8):1018–21.
2 Etiology of Cancer 17
6. Harris CC, Hollstein M. Clinical implications of 15. Arnaud P, Escande MC, Lecuit M, et al. Hepatitis C virus
the p53 tumor-suppressor gene. N Engl J Med. infection and MALT-type ocular adnexal lymphoma.
1993;329(18):1318–27. Ann Oncol. 2007;18(2):400–1; author reply 401–3.
7. Cleaver JE, Crowley E. UV damage, DNA repair and 16. Ferreri AJ, Viale E, Guidoboni M, et al. Clinical
skin carcinogenesis. Front Biosci. 2002;7:d1024–43. implications of hepatitis C virus infection in MALT-
8. Singh AD, Rennie IG, Seregard S, et al. Sunlight type lymphoma of the ocular adnexa. Ann Oncol.
exposure and pathogenesis of uveal melanoma. Surv 2006;17(5):769–72.
Ophthalmol. 2004;49(4):419–28. 17. Parsonnet J, Friedman GD, Vandersteen DP, et al.
9. Pierce DA, Shimizu Y, Preston DL, et al. Studies of the Helicobacter pylori and the risk of gastric carcinoma.
mortality of atomic bomb survivors. Report 12, Part N Engl J Med. 1991;325:1127–31.
I. Cancer: 1950–1990. Radiat Res. 1996;146(1):1–27. 18. Parsonnet J, Hansen S, Rodriguez L, et al. Helicobacter
10. Williams D. Cancer after nuclear fallout: lessons from pylori infection and gastric lymphoma. N Engl J Med.
Chernobyl accident. Nat Rev Cancer. 2002;2:543–9. 1994;330(18):1267–71.
11. McDonnell JM, Mayr AJ, Martin WJ. DNA of human 19. Ferreri AJM, Guidoboni M, Ponzoni M, et al.
papillomavirus type 16 in dysplastic and malignant Evidence for an association between Chlamydia psit-
lesions of the conjunctiva and cornea. N Engl J Med. taci and ocular adnexal lymphomas. J Natl Cancer
1989;320(22):1442–6. Inst. 2004;96:586–94.
12. Nakamura Y, Mashima Y, Kameyama K, et al. Detection 20. Boshoff C, Weiss R. AIDS-related malignancies. Nat
of human papillomavirus infection in squamous tumours Rev Cancer. 2002;2(5):373–82.
of the conjunctiva and lacrimal sac by immunohisto- 21. de Visser KE, Eichten A, Coussens LM. Paradoxical
chemistry, in situ hybridisation, and polymerase chain roles of the immune system during cancer develop-
reaction. Br J Ophthalmol. 1997;81(4):308–13. ment. Nat Rev Cancer. 2006;6(1):24–37.
13. Chan CC. Molecular pathology of primary intra- 22. Chan DS, Lau R, Aune D, et al. Red and processed
ocular lymphoma. Trans Am Ophthalmol Soc. meat and colorectal cancer incidence: meta-analysis
2003;101:275–92. of prospective studies. PLoS One. 2011;6(6):e20456.
14. Buggage RR. Ocular manifestations of human T-cell 23. Wong FL, Boice JD Jr, Abramson DH, et al. Cancer
lymphotropic virus type 1 infection. Curr Opin incidence after retinoblastoma. Radiation dose and
Ophthalmol. 2003;14(6):420–5. sarcoma risk. JAMA. 1997;278(15):1262–7.
Cancer Pathology
3
Gustav Stålhammar, Katarina Bartuma,
Charlotta All-Eriksson, and Stefan Seregard
The earliest known descriptions of tumors (Latin As noted above, tumors may be classified as
for swelling. In Greek: Onkos) or tumorous benign or malignant depending on their potential
growths as a disease appear on papyri from for metastatic spread. Both types of neoplasm
Ancient Egypt. Hippocrates described several might however lead to morbidity and death,
distinct types of such tumors, referring to them depending on factors such as cell of origin,
by the Greek word for cray or crayfish (karkinos immunological responses, nutritional conditions,
or carcinos) by appearance of their cut surface. and extent and location of the growth.
Later, this was translated to its Latin equivalent Consequently, one should not confuse benign
(cancer). The term Cancer now usually signifies tumors for being uniformly harmless.
a malignant neoplasm. On the histological-morphological level, fur-
Currently, the field of oncology spans the ther characteristics separate malignant from
research, prevention, diagnosis, and treatment of benign lesions. Besides metastasis, no single
several hundreds of different types of tumors, characteristic can however reliably isolate on
including the ones originating from the eye and from the other without exceptions. For example,
its immediate surroundings. basal cell carcinoma – the most frequently occur-
This text is intended as an overview of cancer ring human cancer – can display significant cel-
pathology with particular reference to classifica- lular dedifferentiation and pleomorphism and
tion, tumorigenesis, microscopic features, sam- most certainly invades the stroma across the
pling, and diagnostic techniques for ocular basement membrane but exceedingly rarely
cancer. metastasizes. Typical differences between benign
and malignant tumors are summarized
(Table 3.1).
Benign Tumors
G. Stålhammar (*) · K. Bartuma · C. All-Eriksson ·
S. Seregard
Department of Ophthalmic Pathology and Oncology Benign tumors are usually labeled by the suffix
Service and Department of Clinical Neuroscience, -oma. Some exceptions tend to cause confusion;
St. Erik Eye Hospital and Karolinska Institutet, lymphoma and melanoma are by definition
Stockholm, Sweden malignant irrespective of the -oma suffix. To
e-mail: gustav.stalhammar@ki.se
Table 3.1 Typical histological differences between in other organs such as the lacrimal gland or in
benign and malignant tumors peripheral blood.
Feature Benign Malignant • Melanomas are malignant tumors that origi-
Cellular pleomorphism None or Mild to nate from melanocytes, i.e., cells containing
mild severe
intracytoplasmic pigment lodged in melano-
Cellular dedifferentiation None or Mild to
mild severe somes. These cells appear in the skin, uvea,
Necrosis Rare Occasionally conjunctiva, and a variety of other tissues.
Basement membrane Never Frequent
invasion
Proliferation and growth Low High
rate
Tumorigenesis
Metastatic spread Never Occasionally
Advances in techniques for studying cancer
genetics, gene expression, proteomics, metabo-
emphasize this, terms like malignant lymphoma lism, and tumor microenvironment have revolu-
and malignant melanoma are sometimes used. tionized our understanding of the tumorigenesis
process, in which normal cells and tissues pro-
• Adenomas are composed of cells originating gressively evolve to a neoplastic state [1]. The
from glandular epithelium. rapidly accumulating information on the genetic
• Hamartomas are composed of physiologic and epigenetic constitution of malignancies has
cells normally occurring at the affected site. made it possible to tailor novel therapeutic agents,
• Choristomas are composed of physiologic several of which are in clinical use. Hanahan and
cells not normally occurring at the affected Weinberg structured “hallmarks” of cancer, which
site. mark the steps of tumorigenesis [1]. These hall-
• Teratomas are composed of pluripotent cells marks include genomic instability and mutations,
forming different types of tissue originating evading immune destruction, proliferative signal-
from one or more of the three germ cell layers. ing and reprogramming energy metabolism,
A teratoma may be either benign or resisting growth suppressors, escaping cell death,
malignant. replicative immortality, angiogenesis, invasion
and metastasis, and the tumor microenvironment.
The hallmarks are briefly outlined below; the
Malignant Tumors genes involved in tumorigenesis are discussed in
greater detail elsewhere (Chap. 7).
• Carcinomas are malignant neoplasms of epi-
thelial origin. For example, an adenocarci-
noma of the lacrimal gland is a cancer derived Genomic Instability and Mutations
from the glandular epithelium of the lacrimal
gland. Believed to underlie many of the hallmark capa-
• Sarcomas are derived from mesenchymal bilities of neoplastic cells, genomic instability
tissue. generates random mutations including chromo-
• Blastomas are malignant tumors of embryonic somal rearrangements (Fig. 3.1). Conceptually
origin. For example, retinoblastomas are and in a very broad sense, tumors progress by
derived from retinoblasts in the developing stepwise accumulation of gain-of-function muta-
retina. tions of oncogenes that promotes cell survival
• Leukemias are malignancies of blood cells and proliferation or loss-of-function mutations of
that arise from bone marrow precursor cells tumor suppressor genes. Genetic aberrations
and are present in the peripheral blood. such as replication errors, deletions, DNA dam-
• Lymphomas are malignancies derived from age and methylation, and histone modifications
lymph nodes but may occasionally be present and amplifications can occur randomly, be inher-
3 Cancer Pathology 21
Chromosomal abnormalities
b Insertion
c Inversion
Fig. 3.1 Chromosomal abnormalities in human cancer. A versa. (b) Insertion: a chromosomal segment is inserted
multitude of different chromosomal aberrations underlie into another chromosome. (c) Inversion: a 180° rotation of
tumorigenesis. These may lead to the formation of fusion a segment. (d) Copy number gain can occur through, e.g.,
gene or over- or underexpression of a structurally normal complete or partial trisomy, intrachromosomal, and extra-
gene. The most common abnormalities are included in chromosomal amplifications. (e) Copy number loss
this figure. (a) Translocation: one chromosomal segment includes copy number neutral loss of heterozygosity
is translocated from one chromosome to another and vice (LOH), submicroscopic, or large-scale deletions
ited, or be induced by external factors such as ferentiation that follows genomic instability. The
radiation, oxidative stress, viral infection, or a latter is one of the mechanisms proposed as an
breakdown of the genomic maintenance machin- explanation to refractory treatment responses,
ery. Mutations that give a subclone of cells a after an initial period of tumor regression [4].
selective advantage will accumulate throughout Studies have shown that mutations of telomerase
cell divisions and drive tumor evolution. and telomeres are required to achieve endless
Similarly, subclones of cells with increased replication and survival [5]. However, more
genomic instability will more frequently generate recent evidence suggests that cells that survive
events that either promote or impede expansion, the telomere erosion enter breakage-fusion-
further increasing both the genomic instability bridge cycles with resulting gross genetic aberra-
and number of candidates for selective advan- tions. Tumor cells with these complex genetic
tage. Non-tumorigenic regions may harbor “pas- alterations exemplify accelerated acquisition of
senger” mutations that are not pathogenic but mutations and rapidly progress to more malig-
may serve as biomarkers [2]. Recurrent genetic nant states [6].
aberrations and mutations in across several dif-
ferent tumor types indicate that certain subsets of
genetic regions are important drivers of tumori- Evading Immune Destruction
genesis [3]. The advent of deep sequencing of
entire cancer cell genomes has increased our The immune system plays an important role in
knowledge of the tumorigenic potential and hier- eliminating both micrometastases and late-stage
archical arrangement of tumor subclones and the tumors [1]. Studies on patients with acquired
substantial plasticity in differentiation and dedif- immunodeficiency syndrome as well as mouse
22 G. Stålhammar et al.
models have revealed that deficiencies of NK and of these is downregulated in a wide range of
T cells result in a higher risk of developing can- tumors [13]. Approximately 70% of women who
cer, which suggests that both the adaptive and inherit BRCA1 mutation will develop breast can-
innate immune systems serve as barriers to tumor cer by the age of 80, and individuals with germline
progression [7, 8]. Tumor cells must therefore be mutations of RB1 suffer a >90% risk of developing
able to evade immunological killing in order to retinoblastoma, as well as an increased risk for
thrive. This can be achieved, e.g., by altering several other tumor types. In normal cells, cell-to-
immunological reactions by secreting immuno- cell contact mediates suppression of growth [1]. If
suppressive factors such as TGF-β or inhibitory normal cells lose the cell-to-cell contact, they
checkpoint molecules such as BTLA, CTLA-4, undergo programmed cell death. This inhibitory
PD-1, IDO and TIGIT [9–11]. signal is often lost in tumor cells. A set of cell-
surface adhesion molecules, most notably the cad-
herins (E-cadherin and N-cadherin), are commonly
Proliferative Signaling altered in tumor cells enabling the tumor cell to let
and Reprogramming Energy go and then reattach, thus serving as an important
Metabolism step in metastatic spread [14, 15].
Cancer-associated fibroblast
(CAF) Cancer Cell (CC)
Immune
Pericyte (PC) inflammtory cells
(ICs)
Fig. 3.2 The cells of the tumor microenvironment. invade normal tissue and thereafter seed and colonize dis-
Upper: an assemblage of distinct cell types constitutes tant tissues. The abundance, histologic organization, and
most solid tumors. Both the parenchyma and stroma of phenotypic characteristics of the stromal cell types, as
tumors contain distinct cell types and subtypes that col- well as of the extracellular matrix (hatched background),
lectively enable tumor growth and progression. Notably, evolve during progression, thereby enabling primary,
the immune inflammatory cells present in tumors can invasive, and then metastatic growth. The surrounding
include both tumor-promoting as well as tumor-killing normal cells of the primary and metastatic sites, shown
subclasses. Lower: the distinctive microenvironments of only schematically, likely also affect the character of the
tumors. The multiple stromal cell types create a succes- various neoplastic microenvironments. (Reprinted from
sion of tumor microenvironments that change as tumors Hanahan and Weinberg [1] with permission from Elsevier)
(Fig. 3.3a). Cell proliferation markers like pro- is usually balanced by the presence of many
liferating cell nucleolar antigen (PCNA), phos- apoptotic cells. Not all cancers show high cell
phohistone H3 (PHH3), Ki-67, and the cyclins proliferation rates; typically uveal melanoma
often reveal a larger proportion of proliferating features comparatively low counts but nonethe-
cancer cells than detected by mitotic counts less may metastasize. However, in many can-
alone (Fig. 3.3b). A high mitotic index (mitotic cers, including uveal melanoma, a high cell
count per unit of microscopic area) is often proliferation rate is associated with poor prog-
found in rapidly growing tumors although this nosis [34, 35].
26 G. Stålhammar et al.
a Cellular Differentiation
b
Tissue Sampling and Processing
Cytological Sampling
Exfoliative Cytology
Exfoliative cytology involves sampling of cells
that are dispersed in fluids (e.g., cerebrospinal
fluid sampled by lumbar puncture) or forcibly
removed by a spatula, brush, or some other instru-
ment or filter paper. Cells may also be dislodged
Fig. 3.4 Highly differentiated retinoblastoma with tumor
from a surface by a touch preparation, for exam-
cells arranged in a radiating, flowerlike pattern referred to
as Flexner-Wintersteiner rosettes (a). Uveal melanoma ple, imprint cytology for conjunctival tumors. The
specimens featuring spindle-type tumor cells (b) and cells exfoliative sample is then spread on a glass slide
with epithelioid appearance (c) and stained. Vitrectomy samples may be filtered
through a membrane (e.g., Millipore filter), cen-
irrelevant for malignant classification. Carcinomas trifugated in a pellet (cytospin preparation), or
with minimally invasive features are sometimes paraffin-embedded as a cell block.
referred to as microinvasive.
Some malignant tumors show a particular Aspiration Cytology
affinity for spread along peripheral nerves Aspiration cytology may be applied to palpable
extending a considerable distance away from the lesions, guided by ultrasound or computerized
primary site (e.g., perineural growth in squamous tomography imaging. Intraocular fine-needle
cell carcinoma or adenocystic carcinoma of the aspiration biopsy (FNAB) is performed with nee-
lacrimal gland). In some of these tumors, pain, dles between 21 and 25 gauges [38]. A pars plana
tingling, or hypoesthesia due to invasion or com- approach guided by indirect ophthalmoscopy
pression of sensory nerve fibers may be the first may be used, but clear cornea and transscleral
clinical presentation. routes have also been advocated [39–41]. The
28 G. Stålhammar et al.
Fig. 3.6 Illustration of digital image analysis. Two mark nuclei positive for CkMNF116 but negative for
adjacent tumor sections stained with a marker for the Ki67. The proportion of blue polygons to the sum of
type of cell to be analyzed (e.g., an epithelial pancyto- blue and green polygons constitutes the Ki67 prolifera-
keratin marker such as CkMNF116 in breast cancer, a tion index. Bottom: illustration of heat map function
biomarker of choice (such as estrogen receptors (ER)) where the digital image analysis system has analyzed
are aligned on top of each other. Thereby, the software the digitally scanned glass slide (left) for tumor area
can distinguish the fraction of Ki67/ER-positive cells with highest concentration of cells stained by both the
within the cell population of interest. Middle: green dot- pancytokeratin marker and Ki67, marked in red (right).
ted line marks part of a region of interest, automatically Scale bar, middle = 50 μm. Scale bar, lower = 500 μm.
scored for Ki67 index. Blue polygons mark nuclei posi- (Reprinted from Stålhammar et al. [45] with permission
tive for both Ki67 and CkMNF116. Green polygons from Springer Nature)
3 Cancer Pathology 31
vast amount of data on protein expression. Next- uveal melanoma. Cancer Cell. 2017;32(2):204–220.
e215.
generation sequencing of the whole genome has 12. Jones RG, Thompson CB. Tumor suppressors and cell
become a groundbreaking tool in the study of metabolism: a recipe for cancer growth. Genes Dev.
cancer cells. It is used for the characterization 2009;23:537–48.
and identification of, e.g., DNA, RNA, exome, 13. Burkhart DL, Sage J. Cellular mechanisms of tumour
suppression by the retinoblastoma gene. Nat Rev
and transcriptome sequences of tumor cells. It Cancer. 2008;8:671–82.
enables identification of, e.g., copy number vari- 14. Onken MD, Ehlers JP, Worley LA, et al. Functional
ants, sequence variants, mutations, and structural gene expression analysis uncovers phenotypic
changes such as chromosomal translocations, switch in aggressive uveal melanomas. Cancer Res.
2006;66:4602–9.
and fusion genes. Various open-source projects 15. Berx G, van Roy F. Involvement of members of the
aim at collecting sequence variants and muta- cadherin superfamily in cancer. Cold Spring Harb
tions critical in the development of human can- Perspect Biol. 2009;1:a003129.
cers. Public sharing of data has enabled inclusion 16. Adams JM, Cory S. The Bcl-2 apoptotic switch
in cancer development and therapy. Oncogene.
in larger data sets for reanalysis and construction 2007;26:1324–37.
of databases of cancer profiles. 17. Shay JW, Wright WE. Hayflick, his limit, and cellular
Described here are only a handful of the tech- ageing. Nat Rev Mol Cell Biol. 2000;1:72–6.
niques developed for research and clinical study 18. Mehlen P, Puisieux A. Metastasis: a question of life or
death. Nat Rev Cancer. 2006;6:449–58.
of tumors. In the near future, many of the tech- 19. Valastyan S, Weinberg RA. Tumor metastasis:
niques used for experimental purposes will delin- molecular insights and evolving paradigms. Cell.
eate central pathways in cancer cells with 2011;147:275–92.
prognostic, diagnostic, and treatment implica- 20. Nagrath S, Sequist LV, Maheswaran S, et al. Isolation
of rare circulating tumour cells in cancer patients by
tions in human cancer. microchip technology. Nature. 2007;450:1235–9.
21. Husemann Y, Geigl JB, Schubert F, et al. Systemic
spread is an early step in breast cancer. Cancer Cell.
References 2008;13:58–68.
22. Kessenbrock K, Plaks V, Werb Z. Matrix metallopro-
teinases: regulators of the tumor microenvironment.
1. Hanahan D, Weinberg RA. Hallmarks of cancer: the
Cell. 2010;141:52–67.
next generation. Cell. 2011;144:646–74.
23. Foss AJ, Alexander RA, Jefferies LW,
2. Stratton MR. Exploring the genomes of cancer cells:
et al. Microvessel count predicts survival in uveal
progress and promise. Science. 2011;331:1553–8.
melanoma. Cancer Res. 1996;56:2900–3.
3. Korkola J, Gray JW. Breast cancer genomes – form
24. Al-Mehdi AB, Tozawa K, Fisher AB, et al.
and function. Curr Opin Genet Dev. 2010;20:4–14.
Intravascular origin of metastasis from the prolif-
4. Meacham CE, Morrison SJ. Tumour heterogeneity
eration of endothelium-attached tumor cells: a new
and cancer cell plasticity. Nature. 2013;501:328–37.
model for metastasis. Nat Med. 2000;6:100–2.
5. Blasco MA. Telomeres and human disease: ageing,
25. Chambers AF, Groom AC, MacDonald IC.
cancer and beyond. Nat Rev Genet. 2005;6:611–22.
Dissemination and growth of cancer cells in metastatic
6. Artandi SE, DePinho RA. Telomeres and telomerase
sites. Nat Rev Cancer. 2002;2:563–72.
in cancer. Carcinogenesis. 2010;31:9–18.
26. Bidard FC, Madic J, Mariani P, et al. Detection rate
7. Teng MW, Swann JB, Koebel CM, et al. Immune-
and prognostic value of circulating tumor cells and
mediated dormancy: an equilibrium with cancer. J
circulating tumor DNA in metastatic uveal melanoma.
Leukoc Biol. 2008;84:988–93.
Int J Cancer. 2014;134:1207–13.
8. Kim MY, Oskarsson T, Acharyya S, et al. Tumor
27. Straussman R, Morikawa T, Shee K, et al. Tumour
self-seeding by circulating cancer cells. Cell.
micro-environment elicits innate resistance to
2009;139:1315–26.
RAF inhibitors through HGF secretion. Nature.
9. Shields JD, Kourtis IC, Tomei AA, et al. Induction
2012;487:500–4.
of lymphoidlike stroma and immune escape by
28. Grivennikov SI, Greten FR, Karin M. Immunity,
tumors that express the chemokine CCL21. Science.
inflammation, and cancer. Cell. 2010;140:883–99.
2010;328:749–52.
29. Dirat B, Bochet L, Dabek M, et al. Cancer-associated
10. Topalian SL, Drake CG, Pardoll DM. Immune
adipocytes exhibit an activated phenotype and
checkpoint blockade: a common denomina-
contribute to breast cancer invasion. Cancer Res.
tor approach to cancer therapy. Cancer Cell.
2011;71:2455–65.
2015;27(4):450–61.
30. DeNardo DG, Barreto JB, Andreu P, et al. CD4(+)
11. Robertson AG, Shih J, Yau C, et al. Integrative analy-
T cells regulate pulmonary metastasis of mammary
sis identifies four molecular and clinical subsets in
32 G. Stålhammar et al.
carcinomas by enhancing protumor properties of 43. Malhotra R, Chen C, Huilgol SC, et al. Mapped
macrophages. Cancer Cell. 2009;16:91–102. serial excision for periocular lentigo maligna
31. Navin N, Kendall J, Troge J, et al. Tumour evolu- and lentigo maligna melanoma. Ophthalmology.
tion inferred by single-cell sequencing. Nature. 2003;110:2011–8.
2011;472:90–4. 44. Folberg R, Salomao D, Grossniklaus HE,
32. Cho RW, Clarke MF. Recent advances in cancer stem et al. Recommendations for the reporting of tissues
cells. Curr Opin Genet Dev. 2008;18:48–53. removed as part of the surgical treatment of common
33. Singh A, Settleman J. EMT, cancer stem cells and malignancies of the eye and its adnexa. Am J Surg
drug resistance: an emerging axis of evil in the war on Pathol. 2003;27:999–1004.
cancer. Oncogene. 2010;29:4741–51. 45. Stålhammar G, Fuentes Martines N, Lippert M,
34. Seregard S, Spangberg B, Juul C, et al. Prognostic et al. Digital image analysis outperforms manual
accuracy of the mean of the largest nucleoli, vascu- biomarker assessment in breast cancer. Mod Pathol.
lar patterns, and PC-10 in posterior uveal melanoma. 2016;29:318–29.
Ophthalmology. 1998;105:485–91. 46. Paik S, Shak S, Tang G, et al. A multigene assay to
35. Seregard S, Oskarsson M, Spangberg B. PC-10 as a predict recurrence of tamoxifen-treated, node-nega-
predictor of prognosis after antigen retrieval in pos- tive breast cancer. N Engl J Med. 2004;351:2817–26.
terior uveal melanoma. Invest Ophthalmol Vis Sci. 47. Onken MD, Worley LA, Ehlers JP, et al. Gene expres-
1996;37:1451–8. sion profiling in uveal melanoma reveals two molecu-
36. Seregard S, Kock E. Prognostic indicators following lar classes and predicts metastatic death. Cancer Res.
enucleation for posterior uveal melanoma. A multivar- 2004;64:7205–9.
iate analysis of long-term survival with minimized loss 48. Onken MD, Worley LA, Char DH, et al. Collaborative
to follow-up. Acta Ophthalmol Scand. 1995;73:340–4. ocular oncology group report number 1: prospective
37. Makitie T, Summanen P, Tarkkanen A, et al. Tumor- validation of a multi-gene prognostic assay in uveal
infiltrating macrophages (CD68(+) cells) and progno- melanoma. Ophthalmology. 2012;119:1596–603.
sis in malignant uveal melanoma. Invest Ophthalmol 49. Field MG, Harbour JW. Recent developments in prog-
Vis Sci. 2001;42:1414–21. nostic and predictive testing in uveal melanoma. Curr
38. Char DH, Kemlitz AE, Miller T. Intraocular biopsy. Opin Ophthalmol. 2014;25:234–9.
Ophthalmol Clin North Am. 2005;18:177–85, x. 50. Shields CL, Ganguly A, Bianciotto CG, et al.
39. Zorn KK, Bonome T, Gangi L, et al. Gene expression Prognosis of uveal melanoma in 500 cases using
profiles of serous, endometrioid, and clear cell sub- genetic testing of fine-needle aspiration biopsy speci-
types of ovarian and endometrial cancer. Clin Cancer mens. Ophthalmology. 2011;118:396–401.
Res. 2005;11:6422–30. 51. Danielli R, Ridolfi R, Chiarion-Sileni V, et al.
40. Shields JA, Shields CL, Ehya H, et al. Fine- Ipilimumab in pretreated patients with metastatic
needle aspiration biopsy of suspected intraocular uveal melanoma: safety and clinical efficacy. Cancer
tumors. The 1992 Urwick Lecture. Ophthalmology. Immunol Immunother. 2012;61:41–8.
1993;100:1677–84. 52. Tarhini AA, Frankel P, Margolin KA, et al. Aflibercept
41. Bechrakis NE, Foerster MH, Bornfeld N. Biopsy in (VEGF Trap) in inoperable stage III or stage IV mela-
indeterminate intraocular tumors. Ophthalmology. noma of cutaneous or uveal origin. Clin Cancer Res.
2002;109:235–42. 2011;17:6574–81.
42. Spencer JM, Nossa R, Tse DT, et al. Sebaceous car-
cinoma of the eyelid treated with Mohs micrographic
surgery. J Am Acad Dermatol. 2001;44:1004–9.
Pathology Specimen: Handling
Techniques
4
Hardeep Singh Mudhar
Fig. 4.1 (a) A skin ellipse excision of a basal cell carcinoma form the lower lid, containing orientation sutures placed
by the surgeon, to mark the superior and lateral margins. (b) A conjunctival mapping biopsy (tissue with brown hue)
placed on a sponge to keep it as flat as possible by the surgeon in theatre. (c) Pathology request form that shows a clear
diagram of the origin of the conjunctival mapping biopsies. (d) Fresh tissue sampled from the eyelid destined for frozen
sectioning. (e) The surgeon placing marker sutures into the tissue in (d). (f) The surgeon marking the new margin with
a marker pen. (g) The surgeon carefully sandwiching the fresh tissue between two pieces of sponge to keep it flat during
transit to the pathology laboratory. (h) The tissue being placed in a vessel for transport to the pathology laboratory. (i)
An enucleation specimen in a standard plastic pot containing 10% buffered formalin for fixation. (j) A short 25-gauge
needle attached to short tubing is inserted through the scleral bed into the tumour and aspiration is performed (left). The
contents of the needle, tubing and syringe are then rinsed into a tube of CytoLyt or other suitable transport medium
(right). (k) Vitreous washings sent up to the pathology laboratory in the vitreous bag. The surgeon has fixed the speci-
men in theatre with an equal volume of CytoLyt alcohol-based fixative. (l) A plastic impression cytology ring in a
standard plastic pot containing 10% buffered formalin for fixation. (m) The surgeon has drawn a very helpful diagram
on the pathology request form to aid orientation of the specimen. (n) A setup of the cut-up bench in a pathology labora-
tory for grossing ophthalmic pathology specimens. (o) A lower lid resection of a basal cell carcinoma, painted with
yellow and blue tissue dyes in the pathology laboratory to facilitate margin assessment during microscopy (p) The
specimen in figure o sliced into three pieces. (q) Crystal violet dye used to stain a conjunctival mapping biopsy in the
pathology laboratory. (r) The dyed conjunctival mapping biopsy from (q) being placed in the centre of some tracing
paper, to be wrapped up and then placed in a tissue net for processing. (s) Fresh tissue destined for frozen section being
unwrapped in the pathology laboratory. (t) The fresh tissue from s being measured. (u) The fresh tissue being placed in
cryo-matrix prior to freezing. (v) The frozen tissue being sectioned on the cryostat. (w) The frozen sections stained with
H&E ready for microscopic examination. (x) An iridectomy specimen containing a central brown melanoma. (y) An
irido-cyclectomy sample for a brown, dome-shaped ciliary body melanoma. (z) The gross sampling of the specimen in y.
(j: Reprinted from Singh et al. [13]. Copyright © 2011, Karger Publishers. y, z: Reprinted from Ford et al. [3]. With
permission from Elsevier.)
Another random document with
no related content on Scribd:
The Project Gutenberg eBook of Cairo to Kisumu
This ebook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this ebook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.
Language: English
CAIRO TO KISUMU
EGYPT—THE SUDAN—KENYA COLONY
ON THE GREAT ASWAN DAM
“The dam serves also as a bridge over the Nile. I crossed on a car, my motive
power being two Arab boys who trotted behind.”
BY
FRANK G. CARPENTER
LITT.D., F.R.G.S.