patients for clinical trials.

There are four different types of staging: (1) clinical staging: based on physical examination, imaging

tests, and appropriate biopsies, (2) pathologic stage: determined following surgical excision of cancer;

both clinical and pathologic characteristics contribute to the pathologic stage, (3) neoadjuvant or

posttherapy staging: following treatment with systemic chemotherapy or radiation therapy, relying on

clinical and/or pathologic staging guidelines, and (4) restaging: the extent of recurrent disease

following definitive therapy. The formal stage of a cancer does not change over time, even if the cancer

progresses. A cancer that returns or demonstrates progression is still referred to by the stage it was

given at first diagnosis. Restaged cancer is noted with a lower case “r” indicating a restaging

designation.

Table 14-13 TNM Staging Based on AJCC and UICC Classification

Tumor staging in most solid tumors is based on four main factors: (1) location of the primary or

original tumor, (2) tumor size and/or extent of tumors, (3) lymph node involvement or spread, and (4)

presence or absence of distant metastases. The TNM staging system developed by the American Joint

Committee on Cancer (AJCC) and Union for International Cancer Control (UICC) is the most widely

used system and is based on the extent of primary tumor (T), the extent of spread to lymph nodes (N),

and the presence of metastases (M). Table 14-13 demonstrates the general AJCC/UICC TNM staging

system; specific staging systems have been developed for each individual malignancy. Once the T, N,

and M individual stages are determined, they are combined, and an overall stage of 0, I, II, III, or IV is

assigned.

In addition to utilizing the traditional AJCC/UICC TNM staging systems, genomic data have been

recently incorporated to further fine tune and guide treatment decisions. The best-studied example is the

use of genomic tests for breast cancer staging. In four externally validated, commercially available tests,

genomic information is coupled with pathologic staging helping providers to determine the risk of

recurrence of early-stage, estrogen-receptor–positive breast cancers, and the benefit of adjuvant

chemotherapy. In addition, these genomic tests also have been validated in breast ductal carcinoma in

situ to determine the risk of recurrence following resection, the risk of a de novo cancer developing in

the same breast, and the benefits of adjuvant radiation therapy.178 Unfortunately, similar genomic tests

with proven clinical validation are lacking in other cancer subtypes and remain exploratory and useful

only in a research setting at this time.

Screening

The development of the periodic health examination in the late 19th century by medical providers was

the impetus linking cancer mortality to delay in diagnosis. In 1907, Dr. Charles Childe published the

first book, The Control of a Scourge, Or How Cancer is Curable, detailing cancer as a linear carcinogenesis

pathway that could be interrupted by identification of early warning signs and symptoms, thus

eradicating cancer.179 Although too simple, this publication and others led to the upsurge of a largescale health campaign in the 1940s within the United States aimed at advocating early cancer detection

programs, or cancer screening.

Technologic advances in screening and early detection are essential for progress in both the

prevention and treatment of cancer, but incorporation of such advances to useful clinical practice is

often challenging and requires careful consideration of all potential risks and benefits. For any screening

test to be useful, three tenets of screening should be met: (1) a test must exist that will detect the

470

http://surgerybook.net/

disease earlier than routine methods, (2) evidence must exist that earlier treatment leads to improved

outcomes, and (3) benefits of screening must outweigh the risks associated with any subsequent

diagnostic and therapeutic treatments.

The Canadian Task Force on the Periodic Health Examination and the United States Preventive

Services Task Force (USPSTF) in Figure 14-7 provide the basic analytical framework necessary to

delineate the steps for evaluating the worth of a screening test.180 The framework demands a clear

identification of the population at risk that is to be screened. The importance of this starting point is due

to the potential changing nature of the screened population and extrapolation of findings from a subset

of patients to a general population. More importantly, the framework details the adverse effects of both

screening and treatment of early-stage cancers. Because screening tests are performed on healthy

individuals there is significant potential harm associated with a false-positive test. A recent metaanalysis determined that after 3 years of screening tests, a man’s and woman’s risk of obtaining at least

one false-positive test was 60% and 50%, respectively.181,182 False-positive tests are a concern for three

reasons. First, they have the potential to generate negative psychological consequences. Second, they

can trigger a cascade of more invasive follow-up testing, which since the patient does not have cancer

and can obtain no benefit, represents pure harm to the individual. Third, false-positive tests and the

cascade of follow-up tests represent a significant burden to already strained healthcare resources.183

Figure 14-7. Analytic framework for evaluating a screening test. (Adapted from Harris RP, Helfand M, Woolf SH, et al. Current

methods of the US Preventative Task Force: a review of the process. Am J Prev Med 2001;20:21–35.)

Figure 14-8. A: With screening, the lead time in diagnosis prolongs survival even if death is not delayed. B: Screening is more

likely to detect indolent or slow-growing cancers therefore giving a length bias where survival appears improved with screening

but is secondary to the less aggressive tumor biology.

In addition to potential harms, at least two important biases, lead-time and length, need to be

accounted for prior to advocating for a screening test. The intent of screening is to advance the date of

diagnosis to an earlier point in time than it would otherwise been made. Therefore, lead-time refers to

the amount of time between screen-detected and symptom-detected diagnosis (Fig. 14-8A). This leadtime in diagnosis appears to prolong survival in screened individuals, although mortality in this group

may not actually be delayed, creating a lead-time bias. The use of 5-year survival rates to judge the

efficacy of a cancer screening test must be used with caution. For example, the Mayo Lung Project used

chest x-ray and sputum cytology as screening modalities in a randomized controlled trial and

demonstrated that 5-year survival rates increased from 19% to 36% in the screened population.

However, lung cancer mortality rates were not significantly different between the two groups indicating

that although lung cancer was diagnosed sooner in the screened group no overall benefit was noted.184

Length-time bias refers to the tendency of screening to detect cancers that are indolent and slower

471

http://surgerybook.net/

growing because of a longer detectable preclinical phase compared with faster-growing, more

aggressive forms of cancer (Fig. 14-8B). Consequently, in a group of screened individuals this

phenomenon creates the appearance that screening is extending survival, when in fact extended survival

is due to the more indolent nature of the cancers found in this group and not necessarily to the

screening itself. Over diagnosis is one extreme form of length-time bias and can occur in two situations:

(1) cancers that are so benign that they have virtually no growth potential and (2) cancers that grow so

slowly that the person would die of another competing cause of death before the tumor generated

symptoms. The classic indication of an over diagnosis bias is a rise in early-stage cancers coupled with a

minor or even nonexistent decline in incidence of late-stage disease. An example of over diagnosis is

seen in prostate cancer where although there is a decrease in the rate of late-stage disease, the absolute

rate of decline makes up only a tiny fraction of the associated increase in early-stage disease. Thereby,

there are more early cases being identified than late-stage cases being prevented.

4 Currently, population-based screening tests are available for the following cancers: cervical

(Papanicolaou test), colon (colonoscopy, fecal occult blood test, flexible sigmoidoscopy, and doublecontrast barium enema), breast (mammogram), and prostate (PSA test). The current recommendations

for screening by the USPSTF and/or American Cancer Society are listed in Table 14-14.185

Surveillance

Improving cancer survival rates have generated an increased focus on survivorship programs, or the

care of the cancer patient following treatment. One of the goals of these follow-up care programs is the

early detection of tumor recurrence and new primary cancers at a point where curative treatment is still

possible. Unfortunately, due to the heterogeneity of cancer treatment and outcome measures there is a

paucity of evidenced-based data concerning follow-up care, surveillance protocols, and secondary

prevention measures for survivors of cancer. Currently, there are recommended guidelines for only two

cancers, colorectal and breast, detailing survivorship programs.186,187

Surveillance following curative resection of colorectal cancer is guided by the presumed risk of

recurrence and functional status of the patient and is generally heightened in the first 4 years following

surgery.187 Current ASCO guidelines recommend the following: (1) a medical history, physical

examination, and carcinoembryonic antigen (CEA) testing every 3 to 6 months for 5 years, (2) annual

abdominal and chest imaging using computed tomography for 3 years, PET imaging is not recommend,

and (3) surveillance colonoscopy approximately 1 year following surgery with repeat procedures every

5 years if no abnormal findings.

Table 14-14 Current Population-Based Screening Recommendations

The length of surveillance following breast cancer treatment corresponds to the risk of recurrence. As

breast cancer recurrences can occur decades following curative treatment, current continual surveillance

is recommended for at least 15 years.186 Current ASCO guidelines also recommend the following: (1)

history and physical examination every 3 to 6 months for the first 3 years, then every 6 to 12 months

for the next 2 years, and then annually, (2) referral to genetic counseling for women at high risk for

familial breast cancer syndromes, (3) mammography beginning 6 months after definitive radiation

therapy, then annual mammograms following stability of mammographic findings, and (4) regular

gynecologic follow-up.

472

http://surgerybook.net/

Cancer Biostatistics

The driving force behind the maturation of an epidemiologic approach to oncology has been the

incorporation of statistical analysis in modern medical research. This is no more evident than when

discussing survival rates following diagnosis with cancer. The most common survival statistic is OS,

which is typically defined as the time from diagnosis or treatment to death from any cause and is

measured either as a median (months) or as a percentage over 5 years (5-year OS). Disease-specific

survival (DSS) is similar to OS and is measured in months or in a 5-year percentage but where the end

measure is death due to cancer, which is oftentimes difficult to ascertain retrospectively. PFS and time

to progression (TTP) are commonly used to assess efficacy in cancer drug development. PFS is defined

as time of treatment start (or randomization in randomized clinical trials) to the time of disease

progression or death from any cause. Similarly, TTP measures the interval from time to treatment start

but concludes at time of disease progression. Unlike PFS, deaths are censored from analysis in TTP

calculations. Although, used interchangeably, PFS is likely more suitable to situations where a therapy

or intervention can produce an adverse event directly leading to death. Whereas, TTP is more

commonly used in clinical trials with high underlying patient comorbidity, where early patient death

from nontreatment-related events could negatively skew the studied treatment effects. Other commonly

used survival rates include DFS and recurrence-free survival (RFS). DFS is defined as the time from

treatment to recurrence of disease either from the treated cancer or a new primary. RFS is defined as

the time from treatment to recurrence of disease only related to the treated cancer and not a new

primary. Again, both RFS and DFS can be measured in months or time interval percentage, commonly 3-

or 5-year increments.

In a typical clinical trial or study, survival outcome data are represented as a Kaplan–Meier plot (Fig.

14-9). In constructing a Kaplan–Meier survival curve, the probability of surviving in a given length of

time is calculated. For each time interval, survival probability is calculated as the number of subjects

surviving divided by the number of patients at risk. Subjects who have died or dropped out are no

longer considered part of the at-risk population and are removed from the denominator. Within the

graphical representation of the plot, if a patient is removed or no longer participating in the study

before the final outcome is observed, death in OS analysis, the patient is censored and a small vertical

tick mark is seen. The most common statistical method comparing Kaplan–Meier estimates is the log

rank test, which calculates the chi-square for each event time for each group and sums the results giving

a p-value. In addition to the log rank test, median values of an outcome measure can be calculated from

a Kaplan–Meier plot.188

Figure 14-9. Kaplan–Meier survival curve of a hypothetical cancer. Median overall survival is plotted by the time in months of the

median or 50% patient (red line). Deaths represent the step down lines in the curve. Censored events including lost to follow-up

are represented by the vertical tick marks.

Clinical Trials

Clinical trials, in their purest form, are designed to observe outcomes of human subjects under

“experimental” conditions controlled by the researcher. Prior to initiation of human clinical trials,

preclinical investigations include the following animal studies: studying the treatment or drugs safety in

animals at doses equivalent to human exposure, pharmacodynamics, and pharmacokinetics need to be

completed. Pharmacodynamics is the study of what a drug or treatment does to the body, whereas

pharmacokinetics is the study of the body’s effect on the drug.

473

http://surgerybook.net/

The first studies done in human are phase I trials. These are usually open-labeled studies performed in

a small number of “healthy” or “diseased” individuals with the intent on obtaining the MTD useful for

further trials. The MTD is determined by dose escalation of the study drug or treatment and closely

following patients for predetermined, dose-limiting toxicities or adverse events. Typically, there are no

therapeutic end points including survival or progression included in phase I trials.

Phase II trials, also referred to as “therapeutic exploratory” trials, are usually larger than phase I

trials, and are conducted in patients with the disease in question. They are designed to test safety,

pharmacokinetics, and pharmacodynamics as well as providing preliminary data on optimal doses and

frequencies for future phase III trials. Trial design can include a single arm with comparison to historical

controls or a randomized design comparing the study drug or treatment directly with either placebo or a

known drug or treatment. Usually the narrow scope of phase II trials prevents full approval for a study

drug or treatment and requires validation in a phase III trial.

Phase III trials, are typically large-scale studies performed in a more diverse target population in

order to confirm efficacy of earlier trials and to identify and estimate the incidence of common adverse

events. The most common type of phase III trial is comparing the intervention of interest with either a

standard therapy or placebo with a balance in treatment allocation typically through randomization.

Another feature of phase III trial design is stratification, which balances study arms by ensuring that

specific prognostic factors of presumed clinical importance are properly balanced in the arms of a

clinical trial. Following drug or treatment approval, a phase IV trial may be completed. These trials also

referred to as “postmarketing” studies are observational in nature and are aimed at identifying less

common adverse reactions and evaluating cost and/or drug effectiveness in diseases, populations, doses

similar to or markedly different from the original population. The results of phase IV studies can lead to

new black box warnings and even withdrawal of the drug or treatment for safety reasons.189

References

1. Siegel R, Ma J, Zou Z, et al. Cancer statistics, 2014. CA Cancer J Clin 2014; 64:9–29.

2. Siegel R, Desantis C, Jemal A. Colorectal cancer statistics, 2014. CA Cancer J Clin 2014;64:104–117.

3. Edwards BK, Ward E, Kohler BA, et al. Annual report to the nation on the status of cancer, 1975–

2006, featuring colorectal cancer trends and impact of interventions (risk factors, screening, and

treatment) to reduce future rates. Cancer 2010;116:544–573.

4. Cress RD, Morris C, Ellison GL, et al. Secular changes in colorectal cancer incidence by subsite,

stage at diagnosis, and race/ethnicity, 1992–2001. Cancer 2006;107:1142–1152.

5. Weiss W. Cigarette smoking and lung cancer trends. A light at the end of the tunnel? Chest

1997;111:1414–1416.

6. Jemal A, Thun MJ, Ries LA, et al. Annual report to the nation on the status of cancer, 1975–2005,

featuring trends in lung cancer, tobacco use, and tobacco control. J Natl Cancer Inst 2008;100:1672–

1694.