Research
What a company keeps telling the agency while a study runs
Permission to test a drug in people is not a one-time event. It opens a file that has to be fed, on defined channels and defined clocks, for as long as the work continues.
The file does not close when the study starts
It is easy to picture drug regulation as a gate: a company files, the agency reads, the study begins. The gate is real, and it is the smaller part of the arrangement.
Once an application is in effect, the sponsor has to keep amending it so the investigations are conducted according to the protocols in it.
That duty runs on three ordinary channels plus a safety channel, each with its own trigger, its own label, and its own clock. Knowing them makes an obvious question askable of any seller: what is the equivalent duty here.
Which channel a piece of news travels on
The rules sort what happens during a study into four routes, and one of them is defined by exclusion.
Anything about the design or conduct of a study goes as a protocol amendment. Anything urgent about harm goes as a safety report on its own clock.
The yearly account of the whole program goes as an annual report.
Everything else goes as an information amendment, which is defined as essential information that is not within the scope of the other three. It is the channel for what nobody anticipated.
A new study: two conditions, either order
Where a sponsor intends to run a study not covered by a protocol already in the file, it submits a protocol amendment containing that protocol.
The study may begin once two conditions are met. The sponsor has submitted the protocol to the agency for review, and the protocol has been approved by the review board responsible for the study.
Then a small sentence with real consequences: the sponsor may comply with these two conditions in either order.
There is no waiting period like the one that applies to the original application. Submission plus board approval is the whole gate for an added study.
A change to a study, and what counts as big enough
Changing an existing protocol has a threshold, and it moves with the stage of the work.
For a first-phase protocol, an amendment is required for any change that significantly affects the safety of subjects.
For a second or third phase protocol, the trigger is wider: safety of subjects, the scope of the investigation, or the scientific quality of the study.
The rule then gives examples rather than leaving significance to argument. Any increase in dosage or in the duration of an individual subject's exposure beyond the current protocol. Any significant increase in the number of subjects.
Any significant change in design, such as adding or dropping a control group. And the addition of a test or procedure meant to improve monitoring for a side effect, or the dropping of a test intended to monitor safety.
That last example cuts both ways deliberately. Adding safety monitoring is a filed change, and so is removing it.
The exception for something already going wrong
The ordinary sequence is submit, get board approval, then implement. One situation reverses it.
A protocol change intended to eliminate an apparent immediate hazard to subjects may be implemented immediately.
The paperwork still happens, just afterward. The agency is notified by protocol amendment, and the reviewing board is notified under its own rules.
The structure is worth noticing. The only permission to act first is a permission to act faster in the direction of protecting people.
A new person running the study
Adding an investigator is itself a filed change, which follows from the fact that the original application named each one.
A sponsor submits a protocol amendment when a new investigator is added to carry out a previously submitted protocol.
The work does not have to stop for it. Once the investigator is added, the drug may be shipped to them and they may begin participating.
A clock then runs backward from the act. The sponsor must notify the agency of the new investigator within 30 days of the addition.
These particular amendments may also be grouped and submitted at 30-day intervals, which is why the file shows batches of investigators rather than a stream.
The labels on the envelope
The rules govern the outside of a submission as well as its contents, and the reason is navigability rather than ceremony.
A protocol amendment must be prominently identified as one of three things: a new protocol, a change in protocol, or a new investigator.
Each carries a specified payload. A new protocol goes in with a brief description of the most clinically significant differences from previous protocols.
A change in protocol goes in with a brief description of the change and a reference, by date and number, to the submission that contained the original.
A new investigator goes in with their name, their qualifications, a reference to the protocol, and the same additional information the original filing required about investigators.
Information amendments are labeled the same way, by subject: chemistry and manufacturing, pharmacology and toxicology, or clinical.
The channel for everything unclassifiable
The information amendment carries essential information that does not fit the other channels, and the rule gives two examples.
New toxicology, chemistry or other technical information is one.
The other is a report regarding the discontinuance of a clinical investigation. A study that simply stops is reportable, on the same channel as new laboratory data.
The contents are prescribed loosely but not vaguely: a statement of the nature and purpose of the amendment, and an organized submission of the data in a format appropriate for scientific review.
Timing is throttled rather than fixed. These should be submitted as necessary but, to the extent feasible, not more than every 30 days.
The yearly account, and what is in it
Once a year the sponsor files a progress report, within 60 days of the anniversary of the date the application went into effect.
For each study in progress or completed during the year it gives the title and identifiers, the purpose, a brief statement of the patient population, and whether the study is complete.
Then it counts. The total number of subjects initially planned. The number entered to date, tabulated by age group, gender and race. The number who completed as planned. And the number who dropped out for any reason.
Where results are known, even interim, it gives a brief description of them.
The summary half of the report covers the year across the whole program, and it has seven parts.
The seven summaries, and the four extras
The first is a narrative or tabular summary showing the most frequent and most serious adverse experiences by body system.
The second is a summary of all safety reports submitted during the past year. The third and fourth are the two lists a companion article covers: subjects who died, with the cause of death for each, and subjects who dropped out in association with any adverse experience.
The fifth is a brief description of anything learned that is pertinent to understanding the drug's actions, expressly including dose response, information from controlled trials, and bioavailability.
The sixth is a list of the preclinical studies, including animal studies, completed or in progress during the year, with a summary of the major findings.
The seventh is a summary of any significant manufacturing or microbiological changes made during the year.
Four further elements close the report. A general investigational plan for the coming year, replacing last year's. A description of any revision to the investigator brochure, with a copy of the new one. Any significant first-phase protocol modifications not previously reported. And a brief summary of significant foreign marketing developments, such as approval in any country, or withdrawal or suspension from marketing in any country.
What the upkeep produces, and what it does not
Read together, these three sections describe a file that is continuously reconciled against reality.
A study that changes is documented before it changes, except where the change protects people, in which case it is documented immediately after.
A study that stops is reported. A person who dies is named in a list with a cause. A manufacturing change is summarized. A brochure revision travels with a copy.
None of it is public while it happens, which is the honest limit. What a reader outside the system gets is the shape of the obligation, not the contents of any particular file.
That shape is still usable. It is a list of the questions a supervised program answers on a schedule, and a fair set of questions to ask about any program that answers none of them.
Key takeaways
- An application in effect is a live file: the sponsor must keep amending it so studies are run according to the protocols in it.
- Four channels exist, and the information amendment is defined as everything essential that the other three do not cover.
- A new study may start once the protocol has been submitted and the review board has approved it, in either order.
- A protocol change is filed before implementation, with a narrow exception for changes eliminating an apparent immediate hazard.
- Adding an investigator is a filed change, notified within 30 days, though the investigator may begin at once.
- The annual report is due within 60 days of the anniversary and tabulates enrollment by age group, gender and race, alongside dropouts and deaths.
- It also requires foreign marketing developments, including any withdrawal or suspension from marketing in any country.
Frequently asked questions
Can a company add a new study to an existing application without waiting?
Yes, once two conditions are met. The sponsor has submitted the protocol to the agency for its review, and the protocol has been approved by the review board responsible for it. The rule expressly allows the sponsor to satisfy those two conditions in either order. There is no separate waiting period of the kind that applies to the original application.
What kind of change to a study has to be filed?
For a first-phase protocol, any change that significantly affects the safety of subjects. For a second or third phase protocol, any change significantly affecting safety, the scope of the investigation, or the scientific quality of the study. The rule gives examples. An increase in dosage, or in an individual's duration of exposure. A significant increase in subject numbers. A significant design change, such as adding or dropping a control group. And adding or dropping a safety monitoring test.
Can a company act before filing if something is going wrong?
In one situation. A protocol change intended to eliminate an apparent immediate hazard to subjects may be implemented immediately, provided the agency is subsequently notified by protocol amendment and the reviewing board is notified under its own rules. The permission to move first exists only in the direction of protecting participants.
What is an information amendment?
The channel for essential information that does not belong in a protocol amendment, a safety report or the annual report. The rule gives two examples: new toxicology, chemistry or other technical information, and a report regarding the discontinuance of a clinical investigation. They are submitted as necessary but, to the extent feasible, not more than every 30 days.
What is in the annual report?
A per-study section and a summary section. Per study, it gives the title and purpose, the patient population, and whether the study is complete. It then gives the number initially planned, the number entered tabulated by age group, gender and race, the number completing as planned, the number dropping out, and any known results. The summary covers adverse experiences by body system, all safety reports, deaths with causes, and dropouts associated with an adverse experience. It also covers findings about the drug's actions, preclinical studies and their major findings, and significant manufacturing or microbiological changes.
Does the annual report mention what happened in other countries?
Yes. It requires a brief summary of significant foreign marketing developments with the drug during the past year, such as approval of marketing in any country, or withdrawal or suspension from marketing in any country. That duty repeats yearly, separately from the disclosures required in the original application.
Sources
Each document below is named as it names itself, with the date printed on that document rather than the day it was read.
- 21 CFR 312.30 — Protocol amendments — Office of the Federal Register, Electronic Code of Federal Regulations, August 2026
- 21 CFR 312.31 — Information amendments — Office of the Federal Register, Electronic Code of Federal Regulations, August 2026
- 21 CFR 312.33 — Annual reports — Office of the Federal Register, Electronic Code of Federal Regulations, August 2026