Radiopharmaceutical Release Before All Tests Finish

A release packet must answer two separate questions: what does the evidence establish, and who has authorized the batch for use? An approaching dispatch time answers neither. Under § 212.1, final release is the authoritative decision by a responsible person in a PET production facility to permit a batch's use in humans. A completed checklist, an instrument result, or a delivery booking is not that decision.

This guide covers PET manufacture and associated release evidence under Part 212 only. Section 212.5 limits that part to PET drugs; other human drugs fall under Parts 210 and 211. Medical use, radiation licensing, transport qualification, USP chapters, EU GMP Annexes, and Part 11 electronic-record requirements are not evaluated here. Nothing below establishes compliance with those separate frameworks.

For the broader legal foundation, use the Part 212 PET drug CGMP guide. For finite production and laboratory scheduling, use Radiopharmaceutical manufacturing planning. Here, the focus is narrower: the evidence available at the decision, the pending obligations, and the records that connect release to subsequent events.

Build the packet around a single batch identifier and an index, rather than around whichever report arrives first. Section 212.50(c) requires a unique batch production and control record with specified information, including production steps, critical times, equipment, testing results, labeling, significant-step initials or signatures, and investigation results. A release-facing index can point to those records without replacing them.

As a site practice, identify the specification and procedure revisions used, the source location of each result, and the person responsible for resolving each open item. A summary saying 'QC passed' is too compressed to explain which tests were performed or whether their results were reviewed. Conversely, attaching every document without an index makes omissions harder to detect. The useful middle ground is a short decision view with traceable supporting records.

For a shared laboratory or contractual handoff, make the evidence transfer equally explicit. Section 212.10 requires sufficient personnel with the necessary capabilities and adequate resources; it does not supply a special 'gateway approval' mechanism. A practical handoff identifies who performs the work, who receives the complete records, and who escalates missing information. Do not treat an external report or a contractual milestone as the batch-release decision itself.

A result can exist without being acceptable, and it can be acceptable without having been reviewed. Keep those distinctions visible. Section 212.70(c) requires an appropriate laboratory determination of conformance before final release, except for sterility. Section 212.70(d) separately requires review of associated laboratory data and documentation and authorization by a designated qualified individual's dated signature, subject to the conditional-release exception.

Use separate fields for availability, comparison with acceptance criteria, and review. For summary counting, define mutually exclusive buckets. In this guide, 'available passing' means a completed result meeting its criterion; 'failed' means an available nonconforming result; 'pending' means no completed result yet; and 'invalid/unresolved' means no usable determination is established. Review is an overlay, not another bucket to add to the total.

Laboratory reliability belongs in that assessment. Section 212.60 requires written testing procedures, suitable analytical methods and equipment, and documented routine calibration, inspection, checks, and maintenance. Its test-record requirements include sample identification, methods, calculations, sample measurement, all test data, test date and time, comparisons with acceptance criteria, and the performer's initials or signature and date.

As a practical review method, check the equipment actually used for the result in question rather than attaching a general calibration inventory. If an equipment problem calls a result into question, preserve the original evidence and explain the unresolved status. Section 212.61 concerns stability; it is not the source of the laboratory-equipment calibration requirement.

Section 212.70(e) says sterility testing need not be completed before final release. It does not create general permission to defer other specifications. For ordinary final release, § 212.70(c) and (d) still require the nonsterility laboratory determination, review demonstrating conformance except for sterility, and the designated qualified individual's dated authorization.

The sterility test must start within 30 hours after completion of production. The rule allows that requirement to be exceeded due to a weekend or holiday. If the sample is held longer than 30 hours, the producer must demonstrate both that the longer period does not adversely affect the sample and that results will be equivalent to results obtained had testing started within 30 hours. Weekend or holiday status alone does not supply that demonstration.

Anchor the practical timing check to completion of production, not to laboratory receipt, dispatch, or release. Record sample collection and receipt times so the chain is understandable, but do not invent a separate sampling deadline from this provision. Section 212.70(e) specifies the testing-start clock and longer-hold conditions. It also requires tested samples to come from individual batches and not be pooled.

A useful pending-test entry names the batch-specific sample, the test-start record or arrangement, the current status, and the person monitoring the result. If a longer hold is relevant, link the weekend or holiday circumstances and the equivalence evidence. These are recommended ways to expose the legal conditions to review; the regulation does not prescribe this particular worksheet.

The follow-up obligations are equally important. If the product fails a sterility criterion, immediately notify all facilities that received it of the test results and provide any appropriate recommendations. Document the notification. When the investigation is complete, notify all those facilities of its findings. The second communication is not optional merely because an initial alert was sent.

The decision should be understandable without reconstructing an informal conversation. For ordinary final release, including sterility-pending release, § 212.70(d) requires the designated qualified individual's dated signature. Section 212.1 defines final release through a responsible person's authoritative decision. Job titles alone do not explain how a site's designated authority is assigned.

As site practice, maintain a clear assignment of release authority and an absence-cover arrangement. Distinguish the analyst who generated a result, the person who reviewed its associated evidence, the person authorizing release, and the person executing dispatch. These functions may be assigned according to the site's arrangements; this guide does not impose a universal staffing pattern.

A recommended decision entry includes the batch identifier, selected path, evidence index, unresolved obligations, decision date and time, and the authorizing person's identity. For a malfunction conditional release, record the justification and all follow-up required by § 212.70(f); do not claim that the ordinary procedure alone establishes eligibility.

Preserve the historical decision when later information arrives. A subsequent failure should change the current batch assessment and trigger action, not erase the fact that sterility was pending when release was authorized. If source timestamps disagree, document the discrepancy and its resolution. Consistent time-zone notation is a useful record practice, not a substitute for the underlying evidence.

Section 212.70(f) applies when a required finished-product test cannot be completed because of a malfunction involving analytical equipment. A late sample, unavailable analyst, delivery pressure, or an unexplained missing report does not by itself establish that condition. Identify the actual malfunction, the affected test, and the evidence showing why the test could not be completed.

There are two additional hard boundaries. Under § 212.70(f)(2), conditional final release is prohibited if the malfunction prevents a radiochemical identity/purity test or determination of the product's specific activity, even if the other conditions are met. Under § 212.70(f)(3), another batch of that PET drug product cannot be released until the malfunction problem is corrected and the omitted finished-product test is completed.

The supplied rule does not specify a universal number of preceding consecutive batches. Do not invent one or imply that any fixed count guarantees eligibility. The recorded data must support the determination required by the rule. Likewise, 'promptly' correcting equipment is not a permission to continue releasing later batches while waiting for the omitted result.

A practical control is to give the further-release restriction a named owner and an explicit removal record. The responsible person reviewing the next batch should be able to find evidence of both correction and completed testing, rather than relying on a verbal assurance that the instrument is working again.

Release and distribution are connected records, not interchangeable events. Section 212.80 requires every statement on each label to be contained in the batch production record and requires control of labeling and packaging operations to prevent mix-ups. A practical handover checks the released batch identity against the label and the shipment instruction before the product leaves.

Under § 212.1, a common or contract carrier transporting the product is not the receiving facility. A courier contact therefore cannot replace the recipient's name, address, and telephone number. Distribution records may include or refer to the required information, but the reference should be usable when an urgent notification is necessary.

Section 212.90(a) requires written distribution procedures ensuring that the shipping method will not adversely affect identity, purity, or quality. This guide does not establish transport qualification. Similarly, § 212.61 requires documented stability results to determine appropriate storage conditions and expiration dates and times; a release worksheet should reference those established conditions, not invent a new expiry to accommodate dispatch.

As a handover practice, pass forward the release path and any pending-test communication responsibilities. Record actual dispatch separately from authorization. If quantities are split across recipients, retain the recipient-level trace rather than only a batch-total shipment figure.

When a later result arrives, first connect it to the batch, sample, original decision, and distribution record. For sterility failure, the immediate notification and subsequent investigation-findings notification are the specific obligations in § 212.70(e). For an out-of-specification reserve-sample result after malfunction conditional release, § 212.70(f)(1)(v) requires immediate notification of the receiving facility.

Do not wait for an investigation's conclusion before making the immediate notification. A recommended communication record captures the result communicated, the recipients, the time, any appropriate recommendations, the communicator, and attempts or acknowledgments. Recipient reconciliation is especially useful: list every receiving facility from the distribution record and show whether the initial notice and later sterility-investigation findings have each been communicated.

Section 212.71 requires rejection of a nonconforming batch, procedures to identify and segregate product to avoid mix-ups, and investigation of the causes. The investigation must examine processes, operations, records, complaints, and other relevant information. Its documentation must include the results and what happened to the rejected product, and identified problems must be corrected to prevent recurrence.

For already distributed product, a practical record distinguishes what remains under the producer's control from what has left. Record known disposition facts and coordinate appropriate action through the site's procedures. Clinical assessment and any obligations under other notification regimes are outside this guide; a complete Part 212 record should not be represented as proof that those separate obligations have been discharged.

Reprocessing is limited. Section 212.71(d) permits it if appropriate; when material failing acceptance criteria is reprocessed, the procedures must be those stated in the product's approved application, and the finished product must meet specifications except for sterility before final release. For rejected product that has not been returned, destruction may instead be selected as a site disposition. That is a practical disposition choice, not an additional destruction command in § 212.71.

Complaints also have their own record pathway. Section 212.100 requires written complaint-handling procedures and designated-person review and investigation of complaints involving possible specification failure. As site practice, link a related complaint to the batch investigation rather than allowing separate files to produce inconsistent accounts.

Batch SIM-F18-261005-A is an invented F-18 PET product with five shippable containers, excluding laboratory samples. Production completes at 2026-10-05 08:00 UTC. A batch-specific sterility sample is collected at 08:03 UTC and testing starts at 08:20 UTC that same date. No samples are pooled. The start interval is 20 minutes, which is less than 30 hours, or 1,800 minutes; no weekend or holiday extension is used.

In this simulated packet, R01–R08 each point to the method, sample identification, underlying data, calculations where applicable, acceptance comparison, and performer record. LAB-EQ-01 links the relevant instrument checks and maintenance evidence. LBL-01 is reconciled with the batch record. The last nonsterility result is available at 08:30 UTC; associated-data review is completed at 08:35 UTC.

At 08:40 UTC, the fictional designated qualified individual signs a dated final-release decision under the ordinary procedures, recording sterility as pending under § 212.70(e). This is not conditional final release under § 212.70(f): no analytical-equipment malfunction or omitted nonsterility test is involved. The pending entry assigns ST09 monitoring to the laboratory follow-up owner and links the receiving-facility notification procedure.

The status accounting is 9 entries = 8 available passing + 1 pending + 0 failed + 0 invalid/unresolved. All eight completed results are reviewed. Checking the sterility sample and start record does not turn the unfinished sterility determination into a ninth completed result. The release record preserves that distinction.

All production, review, release, and dispatch events occur on the same stated UTC date; later follow-up dates are explicit. The later result changes the accounting to 9 entries = 8 passing + 0 pending + 1 failed + 0 invalid/unresolved. It does not retroactively change the original decision into malfunction conditional release. The original release evidence, subsequent nonconformance, initial notices, and investigation notices remain separately retrievable.

For the drill, the investigation records an identified production-control lapse and its correction; it does not claim to establish a real microbiological cause. No product is returned in this scenario. The recipient directories and full supporting records would need to exist in a complete training dataset before attempting an end-to-end retrieval exercise.

An internal audit can test whether the evidence tells a consistent story. The exercise below is recommended site practice, not a Part 212 requirement for a particular audit format, frequency, or completion time. Select an actual historical batch appropriate to the audit and ask someone other than its original assembler to retrieve the evidence.

1. Start from a distribution entry. Find the batch record, release decision, and applicable result inventory.
2. Reconstruct the state at authorization: which results were available, which met criteria, which were reviewed, and what remained pending?
3. Follow one completed result back to sample identity, method, original data, calculations, acceptance comparison, and performer record.
4. For sterility-pending release, retrieve production completion, test start, batch-specific sample evidence, and any longer-hold justification that was needed.
5. For malfunction conditional release, retrieve every condition, the excluded-test assessment, and evidence that the next release waited for repair and completed omitted testing.
6. Follow any later failure to the receiving-facility list, notifications, investigation, corrective action, and disposition evidence.

Record what made retrieval difficult: a broken reference, a missing recipient telephone number, a summary that obscured an unreviewed result, or an unresolved timestamp discrepancy. Measure retrieval time if it helps improve the process, but label the target as a local objective. A quick retrieval is not evidence of conformance if the underlying record is incomplete.

Section 212.110 requires records at the production facility or another location reasonably accessible to responsible officials and designated FDA inspectors. Records must be legible, protected from deterioration or loss, and readily available for FDA review and copying. It requires all referenced records and documentation to be retained for at least one year from final release, including conditional final release.

As a practical retention check, include linked laboratory records and later follow-up, not only the signed release page. Verify that an exported or archived packet still exposes the referenced evidence. This article does not assess other retention obligations or determine whether a particular electronic-record arrangement meets Part 11.

A useful readiness review ends with specific repairs, not a generic statement that the site is ready. If the specification list cannot be reconciled, fix the inventory. If authority is unclear, resolve the assignment. If recipient contacts cannot be retrieved, repair the distribution reference. If a pending test has no owner, establish the handover before treating the release packet as operationally complete.

• Use a distinct recorded path for ordinary, sterility-pending, and malfunction conditional final release.
• Keep result availability, acceptance comparison, and review status separate.
• Make the human authorization and its evidence basis visible.
• Carry pending tests and further-release restrictions into the next handover.
• Preserve actual recipient and dispatch information for later notifications.
• Close investigations and follow-up without overwriting the historical decision.

For training, ask the reviewer to explain the release basis aloud using only the packet. Then remove one assumed convenience: the original analyst is unavailable, the shipping coordinator is absent, or a report must be retrieved from an archive. The purpose is to expose dependencies on memory, not to add an artificial emergency deadline.

Return broader CGMP questions to the Part 212 foundation guide and capacity conflicts to the manufacturing planning guide. The release-evidence task remains the same: establish the permitted path, show the applicable evidence, obtain the responsible human decision, and keep later obligations connected. Neither urgency nor a reassuring status display can substitute for those steps.