This is an interesting topic for which several confusing points have arisen and which merits an in depth look.
These concepts are frequently used in the discussion of causality and regulatory agencies pay particular attention to them in assessing individual case safety reports particularly in the post-marketing setting. First some definitions (from FDA: https://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM299138.pdf)
Challenge – In our pharmacovigilance world, this refers to the giving of the drug to the patient during the AE or treatment in question. That is, a patient is started today on, say, ampicillin orally. This is the “challenge”.
Dechallenge – This refers to the stopping of the drug, usually after an adverse event (AE) or at the end of a planned treatment (e.g. a two week course of ampicillin). Dechallenges may be complete or partial. That is, the drug is fully stopped or decreased in dose and the AE may fully disappear or only partially decrease. The results of the dechallenge can be confusing:
- A positive dechallenge – This refers to the AE disappearing after the stopping of the drug. Thus, the AE (which may really be an adverse reaction – AR) of diarrhea disappeared a day after the patient stopped the ampicillin.
- A negative dechallenge – This refers to the AE NOT disappearing after the stopping of the drug. In our example, the diarrhea continued even after the ampicillin was stopped.Note that these can be a little confusing as a “positive” dechallenge refers to the disappearance of the problem.
Rechallenge – This refers to the restarting of the same drug after having stopped it, usually for an AE. Rechallenges may also be complete or partial. Thus the patient may have restarted ampicillin a week later after having stopped it.
- A positive rechallenge – This refers to the AE recurring after restarting the drug. To have this occur, the AE had to have previously disappeared after the dechallenge in order for it to restart.
- A negative rechallenge – This is the case where the AE does not recur after the drug is restarted.Note the confusion here: With a positive dechallenge the AE disappears but with a positive rechallenge the AE comes back. And vice versa.
Prechallenge – This is a newish term that refers to the use of the same drug at some point in the past. For example, the patient may have taken ampicillin 2 years ago for an infection. If the patient took the drug in the past and nothing adverse occurred this argues against the drug producing the AE in the current instance. However, this suffers from lots of weaknesses including forgetting what actually occurred if it happened years ago. Alternatively, if an allergic situation, the first use “primes the pump” so to speak and the allergic reaction only shows up on the subsequent uses. This is a weak concept.
So, the general principles around these concepts run as follows:
If a patient takes a drug and has an AE the question then becomes whether the AE is due to the drug (thus it would be an AR or Adverse Drug Reaction ADR) or not. Obviously this is a critical point in determining the safety of the drug and whether it can or should be continued in the particular patient. Usually, as FDA has stated, it is difficult to assess causality in a single individual having an AE. But, we must do it in clinical trials and in signaling of post-marketing cases.
After the AE has occurred, it is common for the physician to stop the drug to see if the AE disappears. If it does disappear this is a positive dechallenge and suggests, at least in regard to the temporal aspects of the case, that the drug might be the cause of the AE. If it does not disappear, then it may be less likely that the drug caused the event unless the drug produced an irreversible change that cannot disappear (e.g. blindness).
If the AE does disappear and if the patient really does need the drug and there no adequate substitutes, the physician may consider restarting the drug if the AE was not serious or severe (one would avoid restarting if the AE was, say, laryngospasm or ventricular tachycardia). This is the rechallenge.
If the use of the drug the second time produces the same AE this is a positive rechallenge and is taken by many as a fairly strong indicator of causality: the drug is a likely suspect cause of the AE. If the AE does not recur then many would say that the drug did not cause the AE and causality is less likely.
Finally, a second dechallenge may be done. Again, the same scenario. If the AE disappears (another positive dechallenge) that is again evidence that the drug was a possible cause of the AE.
The use of the challenge-dechallenge-rechallenge concept is called by some as Koch’s Postulates. This refers to a concept developed by the great German microbiologists Robert Koch and Friedrich Loeffler in the 19th century which states, in their original form, the requirements to establish the etiology of certain infections:
- The microorganism must be found in abundance in all organisms suffering from the disease, but should not be found in healthy organisms.
- The microorganism must be isolated from a diseased organism and grown in pure culture.
- The cultured microorganism should cause disease when introduced into a healthy organism.
- The microorganism must be reisolated from the inoculated, diseased experimental host and identified as being identical to the original specific causative agent. (from Wikipedia: https://en.wikipedia.org/wiki/Koch%27s_postulates)
Although these concepts are similar to our challenge definitions, it is probably not appropriate to refer to drug pharmacovigilance cases as fulfilling Koch’s postulates.
There are many areas in medicine where challenges, dechallenges and rechallenges are useful. For example, in testing for food allergies or sensitivities, having the patient stop all possible offending foods and adding them back one at a time over weeks may show which food or foods cause the problems.
There is another lovely example in gastroenterology of the use of challenge/ dechallenge/rechallenge called the Bernstein Test or the Acid Perfusion Test. This is occasionally done when there is a question about whether chest pain is due to reflux esophagitis or cardiac pain (angina pectoris) and runs as follows. The patient has an NG tube placed midway down the esophagus so that the tip is several inches from the esophageal-gastric junction. The top of the tube is connected to a “Y tube” and two bottles of water are hung – one on each of the “Y” tubes. One bottle contains dilute hydrochloric acid with a pH similar to natural gastric fluid (which contains HCl secreted by the stomach) and the other a saline (salt water) solution. Each bottle is clamped such that one clamp can be opened at a time to allow that fluid to run into the esophagus.
The bottles are covered and hung behind the patient so he/she can’t see them. Then the doctor tells the patient that he/she should raise their hand when they feel their typical heartburn pain. The doctor then alternates the drips for about 1-2 minutes each without telling the patient when the drip changes from saline to acid and back. The issue about dechallenge here is that as soon as the acid touches the tender, inflamed part of the esophagus pain should start if there is typical esophageal inflammation. And then as soon as the saline hits the esophagus it washes away the acid and the pain should disappear within 30 to 60 seconds. The doc does this a few times to see if the patient can tell when the acid is dripped and when not. If there is typical heartburn pain that the patient usually experiences when the acid is dripped and which disappears with saline you have made the diagnosis of esophagitis without having to do an endoscopy etc.
The point here is that this is a lovely use of the concept of dechallenge/rechallenge because the “adverse event” (heartburn) is treated (with the pure water) and should disappear in a minute or so but the dechallenge is clear.
Unfortunately, life does not always run as smoothly as a laboratory test! And in the Bernstein test sometimes the patient has pain but the pain is different from the pain that the patient usually has and thus the test is equivocal because the usual pain may be cardiac pain and dripping acid into the esophagus just confused the matter!!
Conclusion
So is the dechallenge-rechallenge concept useful in assessing causality in drug safety? Yes if good clinical judgment is also applied. This should take into account the timing of the disappearance/reappearance, the clinical setting, other comedications, the disease in question etc. For example, if an allergic type reaction (e.g. urticarial) disappears over a few hours after stopping the drug this is consistent with a drug effect. If the urticarial reaction reappears in minutes after a rechallenge this too is consistent with a drug effect. If the urticaria occurs hours or days later it is less consistent. Similarly, many confounding comedications may make dechallenge/rechallenge hard to interpret.
Health agencies expect that questions about dechallenge and rechallenge will be asked in the work up an ICSR that is submitted. By no means should the company necessarily suggest that the patient be dechallenged and/or rechallenged just to see if the drug caused the AE. This is the judgment of the physician, investigator, patient etc. not the company. But if the dechallenge/rechallenge is done this data should be sought and included in the report.
