- Education and Science»
- Medicine & Health Science
How are drugs developed and approved? The drug development process - ePharmacology
Welcome to ePharmacology! To we will discuss about the different stages of drug/medicine development!
Scientists all over the world are in a continuous effort to develop new drugs although drug development is an extremely arduous, highly technical and enormously expensive operation. Among the scientists, pharmacologists and chemists are more concerned with drug development who work jointly in evaluating new chemical entities (NCEs) and gaining better understandings of active agents. Until about one hundred years ago, most drugs used for the treatment of disease were derived from naturally occurring substances of plant or animal origin, for example, opium from poppy capsule, quinine from cinchona tree and digitalis from foxglove. In 1874, sodium salicylate was synthesized by Kolbe, introducing the era of synthetic therapeutics. Today the large majority of new therapeutic substances in pharmaceutical laboratories and only few are obtained from natural sources.
What you will learn today:
- Preclinical development
- Clinical development
- Clinical trials
- Clinical trial designs
- Multicenter trials
- Randomization of clinical trials
- Protocols in clinical trials
- Control of bias in clinical trials
- Intention to treat analysis
- Prospective design or not
- Record forms
- Data handling in clinical trials
- Statistical aspects of clinical trials
The continuing accumulation of fundamental knowledge about disease and technological progress in molecular biology and biotechnology such as die genetic manipulation of bacteria has led to the development of much better animal and laboratory models for many diseases This makes the drug screening process more efficient and may also lead to decrease in the use of animals for research and preclinical development.
Completion of the human genome project in 2001 has yielded a minimum of 30.000 potential drugs targets. The pharmacogenetics opened up a new horizon to study genetic sequences of target tuch as those coding for ion channels involved in drug response, either desired or undesired.
Synthesis and evaluation of thousands of NCEs are usually necessary for new drugs to be introduced in the market. Successful development must carry the cost of the failure. So, it is highly expensive which in USA, is therefore, ranged from 35 to 50 million dollars. Research and development of new drugs have been done under strict government regulations which have greatly increased over the past couple of decades.
Synthesis of new chemical entities is done as per research policy decision which is based on:
a) random synthesis of new compounds
b) structural variation of compounds with known pharmacological activity
c) biochemical and pharmacological insight
d) chance finding (serendipity).
The aim of the preclinical development phase for a potential new medicine is to explore the drug's efficacy and safety before it is administered to patient. In this preclinical form, varying drug doses are tested in animals and/or in vitro systems such yeasts and bacteria. If active compounds are found, then studies in animal are done. Studies in animal include pharmacodynamics, pharmacokinetics and toxicology. In toxicological study, single dose is for acute toxicity and repeated doses are for subacute, intermediate and chronic toxicity. Full scale carcinogenicity study will only be required if the drug is to be given to man for more than one year. Most of the preclinical tests have to be conducted in accordance with the guidelines prescribed.
About I in 1,000 NCEs reach this stage. At this stage, studies are along several lines. These lines are:
a) pharmaceutical study
b) pharmacological study
c) studies in human or clinical trial.
Pharmaceutical study covers stability, formulation and compatibility of the NCEs with other tablets or infusion ingredients.
Pharmacological study includes further chronic toxicological study in animal, initial animal metabolic and pharmacokinetic study. When studies in animals predict that a NCE may be a useful medicine i.e., effective and safe in relation to its benefits, then the time has come to put it to the test in man, i.e., clinical trial.
Fun Video of Clinical Trial
Studies on human or clinical trial
Clinical trial is a means by which the efficacy of a drug is tested on human beings. It may also give some idea about the risk involved, but usually this is not the primary consideration. Clinical trial is conventionally divided into four phases with each phase progressively testing the safety and efficacy of the compound.
Phase 1 (clinical pharmacology):
This is the first exposure of a new drug on man which is usually conducted in healthy volunteers or patients and is designed to test the tolerability, dose, duration of action and other preliminary information of pharmacokinetic and pharmacodynamic aspects rather than efficacy. Trial of this phase is usually carried out in only one center on 20 to 50 subjects.
Phase 2 (therapeutic exploration):
This phase comprises of small scale trials on patients used to determine dose level and establish that the treatment offers some benefit. It usually involves 100 to 500 patients and is usually conducted in several centers.
Phase 3 (formal therapeutic trial):
Full scale evaluation of treatment comparing it with the standard treatment is done in this phase. It involves randomized control trial on 250 to 2000 patients and is done in.multiple centers.
Phase 4 (post-marketing surveillance):
Information from all studies are received by the committee of safety of Medicine (CSM). If satisfied, product license is issued, then the drug is marketed. Reports about the efficacy and toxicity are received from medical practitioners and reviewed by the committee of Review of Medicine. Renewal or cancellation of the product license depends on the comment of the review committee.
These phases are arbitrary divisions of a continuous expanding process beginning with a single subject closely observed in the laboratory and proceeding to tens of volunteers through hundreds of patients to thousands before the drug is agreed to a medicine by a National Regulatory Authority and is licensed for general prescription. Studies in each phase should be adequately designed and controlled but the phases have no intrinsic merit.
Clinical trial design
Parallel group study:
This is the simplest form of design. Patients are allocated to one of the different treatment groups. The trial size will have been decided in advance. When this trial has reached approximately that desired size, recruitment will be stopped and the data are analyzed.
Historical (retrospective) controlled trial:
This is simply to give new treatment to all patients and to compare the results with previous one. This method is very unreliable because:
a) the patients in the control group are very unlikely to be comparable with the new group
b) patients records are notoriously badly kept so the data extracted from these records will be of inferior quality
c) patient's care may be different
d) severity of disease may fluctuate
e) there may be other environmental differences
Sometimes in chronic stable conditions (such as essential hypertension, Parkinsonism, diabetes mellitus) that can not be cured but can be alleviated, it is possible to give each of the treatments including placebo under test, to each patient. Thus, conveniently using them as their own controls. This is the cross-over study which can not be used in the treatment aimed at curing disease. In the simplest cross-over design, patients are randomized to one of two groups. In one group, they receive treatment A first, then treatment B. In other group, the order of administration is just reversed. In this type of design, the advantage is that fewer subjects are needed to demonstrate an effect.
Sequential trial design:
In sequential trial design, there is scope for analyzing the data when the trial is in progress. When it becomes evident that there is a difference between the treatments, the trial is then stopped. The number of subjects to be studied can be kept to a minimum in this design. However, it is easy to miss small differences. The procedure requires expert statistical supervision.
Multicenter trials can be use at any phase in clinical development, but are particularly valuable when used to continue therapeutic value in phase 3. It provides the possibility of recruiting subject from a wide population and of clinical settings. The main problem of a multicenter trial is that the heterogeneity of treatment effect but more enters may be cleared difficulty in arriving of a single interpretation.
Randomization in clinical trials
If the treatment groups differ significantly in age, sex, race, duration of disease, severity of disease or any other possible relevant factors, it will not he possible to attribute differences in outcome of the treatments under investigation unless there is some way of eliminating the bias.
The best way of getting equivalent groups is by random allocation. The purpose of randomization is to act as a safeguard against selection bias. Randomization should be carried out immediately before treatment. There may be a delay between the time when the patient is approached and the time of randomization. It allows the patient and the investigator to have a second thought about participating or admitting the patient to the study respectively. Once randomization has occurred the investigator is committed to include that patient's data in the final analysis.
Simple methods of randomizing can be designed by tossing a coin or by using published tables of random numbers. Simple randomization is not adequate for large trials (more than 100 patients), multicenter trials and trials within planned interim analysis- other methods of randomization are recommended in these eases.
The sampling interval (n) is first determined by dividing the estimated size of study population by required sample size. Then every size unit is selected till total sample size is drawn.
In this technique, the total population is first subdivided regularly into more homogenous subgroups on the basis of the variable which are known to influence more on dependant variable. Next, sampling units are selected from each group using simple random sampling or systematic sampling technique.
MULTI STAGE SAMPLING:
When the total population is large and diverse, it is convenient to consider it in naturally occurring geographic or arbitrary hierarchical levels or stages. Units at each level are selected using simple random sampling or systematic sampling till the required number of sample size is obtained.
In this sampling technique, population is first divided into groups or within ‘clusters’ and these clusters are then selected using simple random sampling or systematic sampling. Here, natural geographical clusters are commonly used.
Learn more about randomization in clinical trials:
Protocol in clinical trials
Before starting a trial one should design a protocol. Writing a protocol is a very valuable discipline. It is not just a recipe, but serves as a record of the trial and provide continuity for investigators and statisticians.
Aim of the trial:
The aim of the trial should be carefully formulated before, embarking on any other aspects. It is generally best to ask one or two specific questions and to design the trial to answer those questions only. Usually the main aim is to answer the question how effective is this therapy in this condition compared with another therapy? If one tries to answer too many in one study then the trial design becomes more and more complex, and organization becomes difficult.
There are usually several possible measures of treatment efficacy. It should be recorded which one is the primary interest and perhaps which one the secondary.
Patient's eligibility criteria:
One will have to list a set of inclusion criteria and a set of exclusion criteria. A patient will not be entered into the trial unless he or she fulfils the conditions in the former and none in the latter.
The decision to start a formal trial depends on there being a genuine doubt as to whether there is true difference between the two treatments; or where there is an existing treatment, the proposed treatment has any efficacy. The theoretical basis for the study is to test the null hypothesis, i.e., the regimens to be compared are equally effective. Where it is genuinely reasonable to propose the null hypothesis, then a study is ethical and indeed necessary. Patient must give written consent and the protocol should be reviewed by an ethical committee.
A sufficient number of patients must be studied to give a realistic chance of detecting clinically important differences. Many studies fail to produce clear results because too few patients are entered. If sample size is inadequate, there is every possibility of error.
In type I error, trial treatment effects will achieve statistical significance whereas clinical difference may not achieve statistical significance. So, one must be very careful about the detection of sample size which is dependent on the type of study.
Consideration of statistical aspect is also very important to calculate the sample size. The factors which influence the sample size are:
a) rate of prevalence of the disease
b) size of the error acceptable to the researcher
c) the degree of confidence which the researcher wishes to have.
Control of bias in clinical trials
In clinical trials, the control is usually done by blind trial and using placebo.
It means that the individuals involved in the trial do not know what treatment is being administered. The purpose of blinding is to eliminate the bias.
Blind trial is of two types: single blind and double blind.
In single blind trial the patient is unaware of the identity of the treatment being given. The purpose of blinding the patient is to eliminate the differences in responses that can occur because of differences in the patient’s expectation of what a particular treatment do. For example, a patient receiving pain killer and expecting relief of pain may report an effect of pain relief excess of the true effects.
In double blind trial neither the patient nor the investigator knows the identity of the treatment. Sometimes it may be difficult to keep the investigator blind. For example, it is impossible for an investigator comparing the effects of a beta-adrenoceptor antagonist with those of a placebo to be blind for the obvious clinical effects of the former e.g. bradycardia. This can sometimes be overcome by having the outcome evaluated by different investigators who is not in the team in charge of treating the patients.
The word placebo is derived unchanged from the Latin word Placebo meaning "I shall please”.
It refers originally to those substances given merely to please the patient when no specific treatment was available.
Quite often there is no standard treatment for the disease under study, and therefore the control group will have to receive no active treatment. An oral drug is given, and then the patient might feel better merely because of the fact that something is being done.
So, the use of a placebo, a tablet looking exactly like the active treatment but containing no active component, is very useful to avoid bias. It removes from the comparison mentioned above the possibility that any difference is due to the mere act of swallowing any sort of tablet.
A further advantage of placebo is that it allows a trial where there is no standard therapy to be carried out double blind. It is important that the placebo is made to match the active treatment as closely as possible regarding the size, shape, color, texture, weight, taste and smell.
The power of placebo!
Intention to treat analysis video 1
Intention to treat analysis video 2
Analysis by intention to treat in clinical trials
Some patients will have to be withdrawn from the treatment due to serious side effects and non-compliance with their treatment. Such events can be considered under the heading of deviation from protocol.
The main problem with such patients is what to do with the results of them particularly at the analysis stage. Suppose that patient receiving a new treatment had to be withdrawn from that and placed on the standard treatment due to intolerable side effects. How should such a patient be treated in analysis? The dictum of intention to treat states that patient should be retained in the new treatment group for the analysis regardless of the fact that the patient receive the standard therapy for most of the trial.
The basic reason for this seemingly bizarre approach is that the treatment groups are strictly comparable only after randomization. Patients who can not tolerate the treatment are not randomly chosen individuals, so swooping their treatment groups destroys the comparability of the groups.
The safe alternative, dropping the patient from the analysis may be appropriate, but is not without problem. It is likely that a patient who develops side effects would not be doing too well on the primary response variable. Therefore, dropping such a patient will enhance the treatment effect.
Prospective design or not in clinical trials
Clinical trial designed to evaluate the efficacy of new drugs are always prospective (forward looking). In other words, the characteristics of the population to be studied are identified before the study begins and the results of treatment are subsequently observed. For example, if one randomized all patients with heart failure to treatment with either digoxin or a new positive drug and then study the outcome over the following six months that would be a prospective trial
It is possible, however, to study aspects of existing treatment retrospectively (backward looking) by a case control study. In such a study the outcome is first identified and then comparisons are made retrospectively between the characteristics of patients who did or did not have that outcome. Case control studies are cheap and easy to carry out. However, it is not possible to ensure random allocation to treatment groups or blindness and the results may be less reliable than those of a randomized prospective trial. Nevertheless the results of control study may prompt a proper prospective trials in older to confirm the original findings and to extend the Investigation of the problem.
Record forms in clinical trials
Studies require many types of record forms, e.g. write up of the raw data, in structure to the patients. Desirable features of record forms are:
a) simple language
b) neat and orderly structure
c) clear print
d) consistent layout throughout
e) no complex question and answer
I) separate sheet for each assessment
g) to avoid bias
Data handling in clinical trials
Every item of raw data must be reviewed. Before data analysis, they should be checked either manually or by computer for completeness and legibility, and to ensure that recorded values are within reasonable limits. Errors are corrected by reference to the trial record sheets and, if necessary, to the doctor who complete the form.
Read the book "The Prevention and Treatment of Missing Data in Clinical Trials" published by the National Academy of Sciences.
You can either buy the hardcopy or download the softcopy as pdf for free!
Statistical aspect of clinical trial
Almost 500 years ago Leonardo said “No investigation can be called true science without passing mathematical test". Now-a-days it is almost impossible to publish results or understand reports in reputable journals without at least a simple statistical analysis. So, statistical analysis of a trial is of utmost importance lo make the results scientifically true.
Clinical trial is essentially of two type- positive and negative.
A positive trial demonstrates a significant treatment benefit. A negative trial demonstrates a lack of benefit and is of little interest. Well designed good trias are always informative whether positive or negative, whereas poorly designed bad trials are uninformative and at worst seriously misleading. The false impression of the difference between the positive and the negative trials can be minimized by statistical analysis.
INTERPRETIVE FACTORS FOR STATISTICAL ANALYSIS
Null hypothesis:It denotes that there is no true difference between the observed value and the expected or specified value, and the difference is just due to chance.
The alternative to the Null hypothesis is that the difference is more than can be expected by chance.
After completion of a trial, when it Is suspected that treatment A may be superior to treatment B, then, to find out the truth, it is convenient to set about it by testing the hypothesis that the treatments are equally effective or ineffective, as the case may be the hypothesis of no difference. Thus, when two groups of patients have been treated or each patient has had a course of each drug and it has been found that improvement has occurred more often with one treatment than with the other, it is necessary to decide how likely it is that this difference is due to a real superiority of one treatment over the other.
Statistical significance: Statistical significance of a trial can be done by several tests. Selection of test depends upon the type of data. Statistical significance test will tell how often a difference of the observed size would occur due to chance if there is, in reality, no difference between the treatments.
Most clinical trials consist of a comparison of two groups in terms of some measure of response to treatment. The aim of a statistical significance test is to assess whether there is evidence of a true difference, or whether this is a plausible explanation for the difference.
Using an appropriate statistical test, the probability (chance) of observing a difference as extreme or more extreme than one actually observed can be derived assuming that there is really no difference between the groups. This probability is known as the P value and is a measure of whether the observed data are consistent with the null hypothesis of no difference.
A very small P value suggests that the null hypothesis should be rejected and the results are interpreted as providing evidence of real difference. There is a widely accepted but quite arbitrary convention, that a P value below 0.05 is sufficiently small to be taken as evidence of a true difference and a test resulting in such a P value is said to be significant at the 5% level. If the P value does not fall below the 0.05 limit, then the observed difference can easily be explained as a chance finding and so no true difference is established.
Share your feedback!
Help me develop this article!
Confidence interval: Confidence intervals are expressed as a range of values within which we may be 95% certain that the true values lies. This interval may be broad indicating uncertainty or narrow indicating mere certainty. Confidence intervals reveal the precision of an estimate. A wide confidence interval points to a lack of information, whether the difference is statistically significant or not, and is a warning. Confidence intervals are extremely helpful in interpretation, particularly of small studies, as they show the degree of uncertainty related to a result. Their use in case of non-significant finding of a trial may be especially enlightening.
That's it for now. Take the self-assessment interactive quiz below and find out how much you have understood!
view quiz statistics
© 2016 epharmacology