The Washington Manual of Oncology 1st edition (June 15, ): by Ramaswamy, Md. Govindan (Editor), Matthew A., MD Arquette (Editor), Richard L . Lieber By. The Washington Manual of Surgery: Department of Surgery, Washington University School of Medicine, St. Louis, MO · Read more. download The Washington Manual of Oncology: Read 2 Books Reviews - nbafinals.info .
|Language:||English, Spanish, Hindi|
|Genre:||Fiction & Literature|
|Distribution:||Free* [*Sign up for free]|
Washington Manual of Oncology. Concise, clearly written sections on principles of oncology, site-specific diseases, and supportive therapy make The. govindan-the-washington-manual-of-oncology (2).pdf - Ebook download as PDF File .pdf), Text File .txt) or read book online. Concise, clearly written sections on principles of oncology, site- specific diseases , and supportive therapy make The Washington Manual® of.
Alternatively, the trial may attempt to determine whether the effects of two treatments differ by less than a clinically relevant, maximal allowable amount, or are equivalent. A cross-sectional trial gathers data from each participant during a brief interval, whereas a longitudinal trial measures the same quantities repeatedly from each participant over an extended time. Participants in a parallel trial study group may receive a single treatment before their results are compared, or participants in each group may cross over and receive the other's treatment before comparison within and between study groups.
If participants cross over to receive a second treatment, a washout or rest period is often included to ensure that the effect of the first treatment has ceased before the second begins. How many institutions will be involved? Small trials are often carried out at a single institution.
If large numbers of participants are needed or if the cancer of interest is rare, then several institutions may collaborate in a multicenter trial. The National Cancer Institute Cooperative Group program currently supports about a dozen cooperative oncology groups, each with an organizational framework for the conduct of multicenter clinical trials. At this time, cooperative group trials involve 8, researchers at 1, institutions and enroll more than 20, participants each year in clinical trials.
What are the end points being studied? Once the objectives are clearly stated, the next step is to choose measurements of the effects of interest, or end points. If these cannot be measured directly, then careful consideration is needed to find the best available surrogates.
Informative study end points are clearly related to the study hypotheses and objectives, unambiguously measurable with minimal error, and available within a reasonable period.
Information-rich measures are preferable, those that finely discriminate between degrees or states of the phenomena of interest. Common clinical end points in oncology trials may include survival or time to key events or response rates: 1. Complete response is the percentage of patients who have complete resolution of tumor lasting at least 4 weeks.
Stable disease is that which does not meet the definitions for response or progression.
Overall response is the sum of partial plus complete responses. Stable disease or tumor regression may be considered responses to treatment of very aggressive cancers. Event-free survival is the time from a clinically significant event such as diagnosis, treatment, or transplant until some defined event, such as overall survival, disease-free or progression-free survival, time to identification of distant metastases, or local recurrence-free survival.
These times may be expressed by a median for the group or a percentage at a particular time point e. Designing the trial A.
Study population. Equally important is defining the target population, which is that part of the human population to which the results will be applicable in a clinical setting. The sampling frame is that part of the target population from which the study sample will be drawn. The process by which study participants are identified and recruited is the sampling strategy.
A sound sampling strategy ensures that end points are represented fully in the study sample so that the results may be generalized to the population that may benefit from them. Reducing bias. Bias, which is the difference between a measured estimate and the true value of the quantity being measured, can arise at any stage and must be minimized as far as possible. Using a probabilistic sampling strategy, which gives each member of the sampling frame a predetermined chance of inclusion in the study, and randomization, which is the use of a formal probability model to assign participants to treatments, can reduce bias.
The randomization process must be documented as part of the trial and for audit if necessary. Assignments may be kept in sealed envelopes or be available from a randomization center. A randomized assignment usually is made only after the participant's eligibility is established and his or her informed consent has been given.
Knowledge of the treatment received by a participant can lead to significant alteration of his or her care and consequent bias of study results.
Restricting knowledge of individual treatment assignments can reduce bias arising from differential care of participants. In an unblinded, unmasked, or open-label study, assignments are known to study participants, treating physicians, and other study personnel.
In a single-blind study, either the study participant or the treating physician usually the former may be unaware of study group assignments. If both study participants and treating physicians do not know which treatment is being received, the study has a double blind. To maintain a double blind, it usually is necessary to extend the blind to other clinical personnel involved with participant treatment, data collection, and data management.
Data analysts also may be blinded. Any document linking individuals with treatments must be inaccessible to all blinded study members. Care must be taken to design treatments and controls to be as nearly indistinguishable as possible.
Unblinded study personnel must be discrete in discussing study-related information with those who are blinded. Other means of reducing bias. These include careful and consistent implementation of all study procedures by well-trained personnel, including maintenance of blinding see above and complete verified data collection.
Appropriate data analysis also is needed, controlling for confounding see later , avoiding nonhypothesis-driven searching for patterns, and including only planned interim analyses see later. At study end, bias can be reduced by thoughtful interpretation based on observed results and publication of all results, positive and negative, as completely as possible.
Eligibility and exclusion criteria. The research hypotheses and sampling frame are used to identify characteristics of participants who may benefit from the treatment, the eligibility criteria for the trial. Characteristics of participants who are unlikely to benefit or who may be at unusually high risk if enrolled define the exclusion criteria. Eligibility and exclusion criteria are usually specific to the condition and treatment under study, although the presence of unknown or poorly estimable risks e.
Participants should never be included or excluded automatically e. A consecutive series of participants from a single clinic or practice, even if they represent all-comers, is biased by the nature of the clinic, its location, the mechanisms of referral, and many other factors. Such a single-institution sample may not be easily generalized to the target population, and any results may require subsequent confirmation before they are accepted.
Study design. Once the sample is defined, the study design is written. The design is a plan for assignment of participants to treatment; measurement of study end points; and collection, organization, and analysis of the resulting data. The design ensures that the results produced by the study represent its objectives in an accurate and unbiased manner in all parts of the study sample.
An observational or natural history trial measures study end points without attempting to relate them to a baseline or alternative treatment. Such uncontrolled studies can be useful when little is known about the condition of interest in the target population. They may be used to collect data on treatment safety and study feasibility, as well as to estimate study parameters for the planning of future, controlled trials.
A controlled study compares two or more treatments given concurrently under similar conditions. Such a study has at least two arms, or participant groups receiving different treatments.
The control arm may receive a placebo, an inactive or dummy treatment that resembles as closely as possible the experimental one. Alternatively, the control arm may use an active control, a different but active alternative to the experimental treatment, usually the current standard therapy. Use of a placebo is ethically justified if there is no established standard of care or known effective treatment under the circumstances that surround the trial.
Otherwise, it is unethical to offer any participants less than the standard of care.
Patients who have been followed up for the same condition and whose outcomes are known at the outset of the study may be used as a historical or external controls. Conclusions from comparison with historical control are questionable because of the many unknown differences, temporal changes, and uncontrolled sources of bias that may occur between the two sets of measurements, however.
A concurrent control, whether active or placebo, is preferred. Covariates and confounding variables. Study hypotheses and objectives usually will make clear which characteristics of the participants, their environment, or the condition of interest may have effects on the end points. These covariates often include demographic and clinical characteristics of the participants, aspects of the disease process under study, and any striking features of past or current treatments received.
The study hypotheses define how covariates will be included in the analysis and interpretation of study results. If covariates of interest are related to one another, as well as to the study end points, then their interrelations may distort, mask, or confound their effects on the outcome measures. If these relations are understood, they can be included in and adjusted for during subsequent analyses.
The term confounding also refers to the inability to separate effects of two or more covariates on an end point. If the study arms contain very unequal numbers of participants with differing disease-related characteristics, then it may be impossible to separate the effect of treatment from the effect of the characteristics. For example, if a pulmonary-function study contains one treatment group composed largely of urban residents and another of rural residents, the effect of treatment will be difficult to separate from the effect of residence.
To avoid such confounding, assignment of treatments may be stratified so each treatment group contains approximately equal numbers of participants with each characteristic e. Treatment effects are compared within strata. Participants also may be matched on the basis of covariate values e. Study end points are the differences observed within the matched sets. Close matching may reduce the number of participants, as appropriate matches become difficult to find. Whether covariates are effects of interest or factors to be adjusted, their definition and role need careful planning.
New drug development and trial design. An efficient study design obtains the fullest information possible about study end points from the fewest possible study participants over the shortest possible time. A variety of formal designs are available to maximize information, observe nested effects, or manage the impact of missing data. These are routinely used in a variety of scientific disciplines and are discussed in the references. An Investigational New Drug Application IND includes the data collected in preclinical trials in vitro and animal studies and early human studies.
Human trials begin with a phase I or dose-finding trial. In such a trial, a small cohort of patients is treated with a small dose e. This cohort is observed for toxicity. In a classic phase I design, if no unacceptable or dose-limiting toxicities are observed in the first cohort, then another cohort of patients may be treated with a higher dose.
This process is continued until toxicity is demonstrated. If only one patient demonstrates toxicity in a cohort, then that group may be expanded and additional patients treated before escalating the dose further. Once dose-limiting toxicity is demonstrated in more than one patient in a cohort, then the trial is completed, and the next lowest dose is considered the maximal tolerable dose MTD The MTD is recommended for further testing.
Although efficacy is not a traditional end point of a phase I trial, patient responses, if observed, may indicate directions for further testing. One problem with the traditional phase I design described here is that too many patients may be treated at low doses, wasting resources by needlessly enlarging the trial and treating patients with subtherapeutic doses.
This has led to increasing use of novel phase I designs with more rapid dose escalation. Studies of toxicity, absorption, activity, and clearance of the drug safety, pharmacokinetics, and bioavailability also are carried out.
The participants are usually cancer patients with terminal disease for whom no standard therapy or salvage treatment exists, rather than healthy volunteers, as is often the case in other branches of medicine.
Phase II trials enroll approximately 20 to 40 participants to investigate the drug's efficacy and the better to assess toxicity of the drug in a larger group of patients.
These may be composed of a single, nonrandomized group or include one or more control groups. Study results may be analyzed before the trial is complete interim analysis to determine whether there is evidence of benefit, lack of benefit futility , or unacceptable toxicity.
The trial may be stopped for any of these reasons. Interim analyses and early-stopping rules must be planned before starting the trial because testing procedures and significance levels are adjusted to preserve the overall validity of the trial. Results of interim analyses are communicated in an abbreviated form to avoid alteration of treatment and bias of the final study results.
Phase III trials. If the results are encouraging, larger, multiarm, controlled, randomized phase III trials will compare treatment effects with standard therapy. Phase IV trials.
After a New Drug Application NDA has been approved and a drug released by the FDA, additional postmarketing phase IV trials may be carried out to observe treatment effects with long-term follow-up in a broader clinical setting or to examine issues of cost of therapy or quality of life.
Power and statistical significance. Once the design is clear, the study power and sample size can be determined. Power is the probability of detecting the effects of interest if they exist. The complement of power 1 power is the probability of failing to find an effect when one does exist, a type II error.
Common values for type II error are 0. The study significance level is the probability of detecting an apparent effect when none exists and differences observed are the result of chance, a type I error. Common values for type I error are 0. Too small a sample leads to increased probability of both errors and low power. Too large a sample wastes resources, extends the study duration, and exposes participants unnecessarily to the risks associated with the trial.
It also may detect differences that are statistically significant but too small to be clinically important. To calculate sample size, the investigator must know the desired power, significance level, and expected magnitude of the effect to be observed based on the best information available at the outset of the trial. If the required sample size is large, the study may recruit participants from more than one institution, lengthen the recruitment time, or consider other, more information-rich end points to provide finer discrimination between effects of interest.
Data collection, management, and analysis A. Data-collection forms. Once the end points and study design are defined, data-collection forms can be created. Well-designed forms smooth the daily operation of the study, save time and effort, and produce complete, clean, unambiguous data for subsequent analysis.
A comparatively small investment in planning at this stage saves a great deal of time and effort later. Any discrete encounter with a participant that produces study data should have a form. Information collected at different times is kept on separate forms, so that forms are not routinely left partly blank. All forms containing information to be gathered at a particular stage of the study are gathered together into a discrete packet.
Forms should be printed ahead of time, kept up to date with any changes in data collection, and made readily accessible to those who use them. Their content is determined largely by the study design. The format and organization are best worked out collaboratively with staff members experienced in data collection, entry, and analysis. If the study requires that a specific script be followed when asking questions or presenting information to participants, then the text is included at the appropriate point on the form.
Individual data items are presented as clearly as possible, offering selection lists and check boxes rather than free-text fields when possible. The format and units of dates and numbers are printed on the form to reduce later confusion. Required data fields are printed in bold or otherwise clearly identified.
Data items are presented as nearly as possible in the order in which they will be recorded, so that data fields are not routinely skipped. Data fields are never routinely left blank. Codes or check boxes are used to indicate why values are missing to make clear that data collection has been completed.
The person filling out a form initials and dates each one in case questions arise later about the information on the form. Completed forms are sorted by participant and stage of the project and kept in locked filing cabinets in a secure area to preserve confidentiality of participant information.
Data analysis. Once the study hypotheses and end points are defined, a data-analysis plan specifies the exploratory analyses, comparisons, modeling procedures, and hypothesis tests to be carried out. Clinical and demographic characteristics of participants and their environment are identified, and the roles of these covariates in data analyses are specified. This edition has been updated to include new standards in the treatment of malignancies and hematologic disorders, mechanisms of action of new therapeutic agents, and current use of molecular prognostic factors.
The information in each chapter is now presented in a consistent format, with important references cited. Our goal is to provide a concise, practical reference for fellows, residents, and medical students rotating on hematology and oncology subspecialty services.
Most of the authors are hematology—oncology fellows or internal medicine residents, the physicians who have recent experience with the issues and questions that arise in the course of training in these subspecialties. Primary care practitioners and other health care professionals also will find this manual useful as a quick reference source in hematology and oncology.
As the practice of hematology and oncology continues to evolve, changes in dosing and indications for chemotherapy and targeted therapies will occur, and staging systems will be modified. We recommend a handbook of chemotherapy regimens and an oncology staging manual to complement the information in this book.
And of course, clinical judgment is imperative when applying the principles presented here to the care of individual patients. If you found this book helpful then please like, subscribe and share.
Save my name, email, and website in this browser for the next time I comment. Notify me of follow-up comments by email.
Notify me of new posts by email. This site uses Akismet to reduce spam. Learn how your comment data is processed.