An unexpected blank peak needs an explanation
A peak in an HPLC blank does not by itself prove carryover or invalidate every result. First apply the approved method: a blank that breaches a defined criterion may be a system suitability or other method-acceptance failure. Passing the remaining system suitability test (SST) checks does not resolve an unexpected blank signal. An unexplained discrepancy that could affect release or stability results needs documented investigation and an assessment of affected data before those data support batch disposition. 21 CFR 211.192.
Consider a quality control (QC) impurity run in which precision, tailing, and resolution pass, but the blank after a concentrated standard has a small peak near the analyte retention time. The immediate question is where that signal came from. Residue from an earlier injection, contamination in the blank preparation, and interference from the chromatographic system call for different corrective actions. The signal also has to be evaluated against the analytical decision it could affect, rather than judged by its appearance beside a much larger standard peak.
This discussion concerns US finished-pharmaceutical release and stability testing under current good manufacturing practice (CGMP). FDA’s July 2015 analytical-procedures guidance recommends describing the injection sequence, including blanks, and the applicable suitability criteria. Its examples of tailing, precision, and resolution do not establish a universal numerical blank limit or prevent a particular method from including a blank criterion. FDA guidance provides recommendations; the cited CFR provisions impose regulatory requirements. FDA analytical-procedures guidance, sections IV.G–H.
The broader question of what a passing SST establishes is covered in our guide to HPLC system suitability and sample validity. Here the investigation stays with the blank peak: its origin, the observations that discriminate among possible sources, and the evidence needed to evaluate affected results. HPLC system suitability and sample validity.
Carryover, contamination, and system interference
Carryover means that material from a preceding injection contributes to a later measurement. FDA’s May 2018 Bioanalytical Method Validation guidance describes analyte appearing from a preceding sample; ICH M10 similarly describes residual analyte remaining in the instrument. Those definitions help identify the mechanism, but the documents’ bioanalytical acceptance criteria do not become QC release criteria. FDA bioanalytical guidance. ICH M10, section 3.2.6.
For troubleshooting, separate material retained from an earlier injection from material introduced with the blank or mobile phase. These categories can overlap: the term ghost peak is not a separate, proven root cause. Lab Manager’s technical discussion includes carryover among ghost peaks and notes that strongly retained material can persist through several blanks. Its instrument-vendor observations are technical context, rather than regulatory requirements. Lab Manager: Minimizing HPLC Carryover.
McNally and Usher’s 2013 LCGC article reports method-development experiments with agricultural products. It shows that wash-solvent choice can affect carryover and precision, and identifies autosampler hardware, fittings, tubing, and the wash station as areas to examine. These findings justify checking the installed configuration; they do not supply a universal wash recipe for pharmaceutical methods. LCGC method-development experiments.
| Working hypothesis | Useful comparison | Limit on the conclusion |
|---|---|---|
| Residual material from an earlier injection | Compare blanks before and after a high-level standard or sample, including any successive blanks. | Appearance after exposure and declining response support carryover; they do not identify a particular component or prove chemical identity. |
| Contamination introduced with the blank | Compare the original preparation with independently prepared diluent and suitable fresh vials. | A changed result narrows the source only if the other conditions are controlled. Changing solvent and vial together cannot distinguish their individual contributions. |
| Mobile-phase or system interference | Review appropriate mobile-phase controls, prior instrument use, and the conditions under which the peak appears. | A peak in a mobile-phase control warrants further investigation; it does not automatically exclude retained material elsewhere in the system. |
A useful investigation statement distinguishes an observation from an inference. “The peak followed the high standard” records sequence evidence. “The needle caused it” names a component and needs additional evidence. Retention time alone is also insufficient to label an unexpected peak as the product, an impurity, or a degradation product. Keep the original observation available while testing the proposed explanation.
What blank placement can and cannot establish
FDA’s analytical-procedures guidance places blanks within a described sequence. ICH M10 uses a blank after the highest calibrator to challenge carryover in chromatographic bioanalysis. The table below adapts that placement logic into investigation questions for QC; it is an editorial framework, not a regulator-prescribed release sequence. Any additional injections should follow an authorized investigation plan. FDA analytical-procedures guidance, section IV.G. ICH M10, section 3.2.6.
| Position or comparison | Question it helps answer | What it cannot establish alone |
|---|---|---|
| Leading diluent blank, before the current run’s standards and samples | Was the signal present before exposure to the current sequence? Record previous instrument and column use as well. | It cannot be caused by an injection that has not yet occurred. It can still reflect residue from earlier work, blank contamination, or system interference. |
| Blank immediately after a high-level standard or sample | Does the signal appear after exposure to a potential source? Preserve the preceding injection identity and method conditions. | A matching retention time is not proof of analyte identity or of a needle-specific mechanism. One blank cannot separate all competing sources. |
| Successive blanks under recorded conditions | Does the response decrease, persist, or vary, and were conditions held constant? | There is no general percentage-decay rule in the cited sources. Persistent response does not rule out carryover, and disappearance does not establish why it occurred. |
| Blanks interspersed through a sequence | Where was the signal observed relative to the samples whose results may be affected? | A clean blank at one position does not demonstrate absence of interference throughout the run or in every sample matrix. |
| Independent diluent and appropriate mobile-phase controls | Does the observation follow a preparation, solvent source, or system condition? Specify what actually changed. | A mobile-phase injection and a zero-volume instrument cycle are different experiments. Neither automatically excludes the autosampler, column, or other flow-path surfaces. |
Sequence position is strongest when linked to a controlled comparison. Record the vial identity, preparation, instrument method, preceding injection, and time for each observation. A spreadsheet of blank areas without that context can conceal a changed wash setting or replaced solvent bottle. Conversely, a sequence map with no response information cannot show whether the suspected interference was relevant to a low-level result.
Plan comparisons to answer competing explanations. If a fresh preparation is clean, the evidence may favor a preparation-related source, but it does not identify the original contaminant. If a signal persists after several blanks, retain carryover among the possibilities. The practical value is a narrower set of hypotheses and a justified next comparison, rather than a binary pass/fail diagnosis from one pattern.
Collect evidence before changing the system
FDA’s May 2022 OOS guidance recommends prompt assessment of unexplained results, retention of preparations if stable, supervisor notification, and raw-data review. A blank peak is not automatically an out-of-specification (OOS) product result: classification depends on the applicable specifications and method criteria. The initial-assessment principles are useful for a blank discrepancy, while the firm’s procedure determines its investigation route. An investigation can therefore follow the firm’s discrepancy procedure without automatically becoming an OOS Phase I event. FDA OOS guidance, section III.
Preserve the original state and define the question
Before routine washing, replacing reagents, or starting a new sequence, preserve the acquired data and record the conditions associated with the observation. Retain relevant preparations when their stability and handling requirements permit. Record any necessary intervention and why it could not wait. The objective is to avoid losing the ability to distinguish the original condition from the condition after maintenance or preparation changes.
State the competing explanations and the proposed comparison before collecting more results. For example, a post-standard blank signal can lead to a controlled comparison using independently prepared diluent, alongside review of the leading blank and prior sequence history. Record which variable changes, what observation would support each explanation, and the limits of the comparison. A clean follow-up chromatogram is evidence to interpret, not an automatic reason to discard the original finding.
Compare chromatograms without overstating identity
Compare the blank, relevant standards, and potentially affected samples using the acquisition and processing conditions actually applied. Record retention behavior, peak response, integration, and any available spectral information. Use method-supported identification criteria. A retention-time match or a low-level UV spectrum is not, by itself, conclusive identification or exclusion. ICH Q2(R2) discusses selectivity through interference assessment and, where needed, complementary procedures. ICH Q2(R2), section 3.1.
Test preparation and instrument hypotheses separately
For a preparation hypothesis, document what is different about the independent diluent or vial comparison. If several items change simultaneously, record that limitation rather than assigning the result to one item. For an instrument hypothesis, compare the actual wash configuration and maintenance history with the approved procedure and applicable instrument instructions. The cited technical literature supports these lines of inquiry; it does not guarantee that a stronger solvent or a replaced seal will resolve every case.
A mobile-phase control may help direct attention to solvent contributions. If an instrument-specific zero-volume or no-injection cycle is proposed, describe its flow path and valve behavior separately. Do not call it an injection of pure mobile phase or assume it bypasses every potential carryover site. The investigation plan should explain what the control exposes and what remains untested.
Keep processing history linked to the finding
Review relevant sequence and processing records for changes that affect whether the peak is visible or reported. Preserve the original record and document the scientific reason for any change. Do not suppress an unexpected peak through result-driven integration. Section 211.194 requires complete test records and records of method modifications; the Tender letter also addresses data controls and unofficial injections. 21 CFR 211.194(a)–(b). Tender Corporation warning letter.
The detailed record requirements after reprocessing are discussed in our guide to chromatogram reintegration. Here, the practical check is whether a reviewer can reconstruct the blank observation, each follow-up comparison, and the basis for the conclusion. Chromatogram reintegration records.
Acceptance criteria must match the method and purpose
For chromatographic bioanalytical assays, ICH M10 section 3.2.6 recommends assessing carryover with blanks after the upper-limit-of-quantitation calibrator. The blank response should be no greater than 20% of the analyte response at the lower limit of quantitation and 5% of the internal-standard response. M10 also discusses validated mitigation when carryover is unavoidable. These are bioanalytical recommendations, not finished-drug QC release limits. ICH M10, scope and section 3.2.6.
FDA’s May 2018 bioanalytical guidance separately lists a carryover criterion of 20% of the lower limit of quantitation in Table 1 and calls for assessing effects on study-sample quantitation. That table does not provide M10’s separate 5% internal-standard carryover criterion. Its recommendation addresses bioanalysis and supplies no numerical limit for the QC impurity-run scenario here. FDA Bioanalytical Method Validation, Table 1.
ICH Q2(R2) applies to release and stability analytical procedures and addresses fitness for purpose, including specificity/selectivity. It does not state a universal chromatographic carryover percentage. Its detection- and quantitation-limit discussion is not a blank carryover acceptance rule. FDA’s 2015 guidance refers readers to USP <621> for chromatographic suitability; that reference alone cannot establish the current chapter’s or an individual monograph’s requirements. ICH Q2(R2). FDA analytical-procedures guidance, section IV.H.
The regulatory basis for an appropriate control is more general. Section 211.160 requires scientifically sound laboratory controls, quality-unit approval, and recorded justification for procedural deviations. Section 211.165 requires appropriate acceptance criteria and documented test-method performance. The firm should translate those obligations into a justified blank evaluation within the applicable method and quality system, while observing any binding application or compendial requirements. 21 CFR 211.160(a)–(b). 21 CFR 211.165(d)–(e).
Identify the decision at risk. An impurity reporting decision and a high-concentration assay result need different evaluations of potential interference.
Specify the blank’s role, preparation, placement, response measure, and applicable criterion. Record where the supporting method-performance evidence is held.
Explain how the criterion protects the intended use across relevant conditions. A percentage copied from a different analytical scope is not that explanation.
Define the response to a breach or unexpected signal. Acceptance criteria and investigation instructions should not be improvised after the desired sample result is known.
The cited sources provide no default percentage of an impurity reporting threshold, fraction of the assay-standard area, or required rate of decay across blanks. Where the method lacks an adequate evaluation, document the gap and establish a scientifically justified approach through the quality system. A new number selected only because the observed peak falls below it cannot resolve the underlying question.
What the Tender Corporation letter actually shows
FDA’s July 23, 2020 warning letter to Tender Corporation described an inadequately supported sodium bicarbonate assay for topical products. Under 21 CFR 211.160(b), FDA identified abnormal peaks, missing integration, and reliance on blanks between samples to prevent carryover. The letter also described six months of suitability failures attributed to several causes without comprehensive investigation and timely corrective and preventive action (CAPA). Separately, FDA criticized unofficial injections and incomplete data controls. Tender Corporation, MARCS-CMS 599789.
This is a site-specific enforcement example. It supports investigating recurring discrepancies and retaining complete data; it does not prohibit blank injections. The useful distinction is whether a blank is part of a justified control or documented experiment, or an unexplained workaround that leaves method reliability unresolved. Labeling an injection a wash does not remove its relevance to the investigation.
FDA’s OOS guidance separately criticizes repeated testing until a passing result is obtained while earlier OOS findings are disregarded without scientific justification. Additional blanks do not automatically meet that description, but a clean blank after repeated washing cannot alone invalidate earlier sample results. The investigation must connect the proposed cause with the data being evaluated. FDA OOS guidance, section IV.B.1.
Document closure and assess affected results
The closure record should let a reviewer move from the original observation to the conclusion without filling gaps from memory. Section 211.194 addresses complete laboratory data and calculations. Section 211.192 requires a written investigation with conclusions and follow-up, extending to other batches or products potentially associated with the discrepancy. That scope depends on the evidence; shared equipment or reagents are investigation leads, not proof that every associated batch is affected. 21 CFR 211.194(a). 21 CFR 211.192.
The following checklist organizes the evidence for this particular blank-peak decision. It is not a prescribed federal form, and it does not require every listed diagnostic experiment in every case. Record which comparisons were appropriate, what they showed, and what uncertainty remains.
| Evidence domain | What the record should explain | Basis or boundary |
|---|---|---|
| Original finding | Blank identity, sequence position, response, method criterion, and the original chromatogram. | Complete test data: 21 CFR 211.194(a). |
| Competing explanations | Why carryover, preparation contamination, or system interference was considered; which controlled comparisons supported or weakened each. | Scientifically sound controls: 21 CFR 211.160(b). The diagnostic plan is method-specific. |
| Follow-up work | Purpose, conditions, preparations, results, and limitations of additional injections or maintenance. Include changed conditions. | Record deviations and justify them: 21 CFR 211.160(a). |
| Affected results | Which samples and related work could share the exposure, how impact was assessed, and why the proposed scope is adequate. | Discrepancy investigation and extension: 21 CFR 211.192. |
| Closure and prevention | Supported conclusion, unresolved uncertainty, corrective action where needed, review, and follow-up to evaluate control. | 21 CFR 211.192; Tender illustrates the problem of recurring uninvestigated failures. |
Connect the blank observation to the reporting decision
Map potentially affected samples against the actual sequence and the source hypothesis. A post-standard blank does not necessarily measure the interference in every later sample. If a numerical impact estimate is used, identify its calibration basis, dilution factors, response assumptions, and uncertainty. Explain why that estimate applies to the samples in question. Do not treat an observed blank area as a universal correction factor.
In particular, a worst-case assessment is not permission to subtract a blank peak from an OOS impurity result to create a passing result. Evaluate data validity under the approved procedure and retain the original result. When an OOS result exists, its investigation and any proposed invalidation need the scientific support described in FDA’s OOS guidance. The distinction between an initial result and a confirmed failure is examined separately in our OOS guide. FDA OOS guidance. OOS results and confirmed failures.
For a receiving laboratory, relevant method-transfer records may help explain whether the observed behavior differs from the established method performance. For stability work, the assessment should identify any reported time points that could be affected. Those are routes for examining the consequence of the blank finding, rather than automatic conclusions about a transfer or a shelf-life trend. Analytical method transfer criteria. Stability out-of-trend signals.
Document the supported explanation, the evidence for controlling it, and the implications for results that might have been exposed. If the cause remains unresolved, record that uncertainty and the resulting limits on data use. Neither a passing SST summary nor a final clean blank substitutes for that evaluation.
Sources
ICH. Q2(R2) Validation of Analytical Procedures (adopted November 1, 2023)
FDA. Bioanalytical Method Validation (May 2018), including Table 1; bioanalytical scope
FDA. Investigating Out-of-Specification Test Results for Pharmaceutical Production (May 2022)
FDA. Tender Corporation warning letter, MARCS-CMS 599789 (July 23, 2020)
21 CFR 211.160 — General requirements, reproduced by Cornell LII
21 CFR 211.194 — Laboratory records, reproduced by Cornell LII
21 CFR 211.165 — Testing and release for distribution, reproduced by Cornell LII
21 CFR 211.192 — Production record review, reproduced by Cornell LII
DePalma. Lab Manager: Minimizing HPLC Carryover; technical trade reporting, not regulatory authority
