A STERIS-authored article in Pharmaceutical Technology reports that in-line UV spectrometry at 220 nm can follow a biopharmaceutical cleaning cycle continuously, from heavily soiled rinse to baseline. The data are encouraging, the framing is vendor-supplied, and the distance between process monitoring and a release decision is where our own thinking has to do the work.
What happened
On 29 July 2025, Pharmaceutical Technology (pharmtech.com) published "In-line UV Spectrometry Monitoring in Cleaning Validation" by John Schallom, Brian Bosso, Paul Lopolito, Katie Frey and Amy Thanavaro — all affiliated with STERIS, a supplier of cleaning chemistries and equipment (1). The article reports feasibility studies on in-line UV absorbance at 220 nm through a sanitary flow cell, used to follow a cleaning cycle continuously rather than confirm cleanliness after the fact.
The motivation is familiar: manufacturers lean heavily on non-specific at-line testing of final rinse water, which the authors say can generate false positives. They cite EU GMP Annex 15 Sections 10.6.1 and 10.6.2, which permit non-specific methods such as TOC and conductivity where specific residue testing is not feasible (2). A justified real-time in-line method belongs in that category — and covers the full cycle, which a grab sample cannot.
Performance was established against formulated alkaline and acid cleaners. A 10 cm flow path instead of a 1 cm cuvette raises absorbance ten-fold (Beer–Lambert), lowering detection capability. Reported limits: alkaline cleaner LOD 5 ppm / LOQ 25 ppm; acid cleaner 5 / 10 ppm; upper range 1000 ppm. Interference testing with monoclonal antibody drug substance (native and denatured), insulin drug product and bovine serum albumin (BSA) showed cleaner and soil responses to be additive — soil-to-cleaner absorbance ratios of 0.60 to 0.96, no enhancement or depression.
A dynamic test mounted a PX2 photometer on the drain port of a 60 L vessel soiled with BSA and insulin. A one-minute 1% alkaline wash, a one-minute 1% acid wash and a DI water rinse gave absorbances above 4.0 during cleaning and a return to baseline (<0.02) within a minute on rinse. The authors conclude in-line UV supports quality by design, PAT and Pharma 4.0 goals.
Why it matters
Earlier work used on-line A280 to track protein dissolution kinetics, and UV rinse analysis has been combined with direct-surface FTIR to reduce routine swab sampling (3,4). The value here is that the practical objections — does cleaner absorb? does product plus cleaner distort the signal? can an in-line probe match a bench instrument? — are treated as experiments rather than assertions.
FDA's 2004 PAT guidance defines at-line, on-line and in-line measurement and describes endpoints based on attribute attainment rather than fixed time (5). Cleaning sits awkwardly in that frame: it releases no product, so what PAT offers is control and continuous process verification, not a release decision.
What is particularly interesting
The additive behaviour of cleaner and product across a 0.60–0.96 ratio range is what makes a non-specific method defensible — without it, a rinse signal could be attributed to the wrong contributor and the approach collapses. The authors also concede the trade-off: because UV is non-specific, all signal must be attributed to the single worst-case component, which "may lead to a bias to failing."
They add that with the process not fully optimised, foam in the line could reduce effective pathlength — a physical failure mode of an optical measurement that a qualification protocol tends to miss.
What could we learn
The most useful lesson is a discipline about vocabulary. In-line UV gives us a process signal: it shows the cycle behaved as it did during qualification, that soil and cleaner cleared, and that a deviating cycle can be stopped and investigated in real time. That is valuable continuous process verification.
It does not, on its own, give us a release decision for the equipment. A 2024 peer-reviewed study using an in-line UV-Vis PAT tool on a film-flow cleaning rig found that only dissolution is quantifiable and that undissolved API passed the sensor undetected (6). A drain-line probe also samples the fluid stream, not the surface — a location that never fully wets cannot be shown clean by a rinse signal. And UV at one wavelength says nothing about the microbiological state of the equipment.
The honest structure is therefore: in-line UV as continuous process verification, retained swab or surface sampling for what the probe cannot reach, and visual inspection where it remains the accepted criterion. Moving a release decision onto in-line data requires the full apparatus — method validation to ICH Q2, documented worst-case attribution, and a justified argument that the measurement is at least equivalent to the sampling it replaces.
Questions worth asking in our own programme
- What would an in-line UV baseline look like on our worst-case product and cleaner combination, and have we measured the soil-to-cleaner absorbance ratio rather than assumed additivity?
- Which acceptance criteria are surface-based (µg/cm², visually clean) and therefore not addressable by a fluid-stream measurement at all?
- If a cycle's UV profile deviates, do our procedures define what happens next — hold, investigate, re-clean — or is there no written response to a monitoring signal?
- Have we characterised foam, air entrainment and particulate scattering in the flow path as potential pathlength and signal artefacts?
- Is our target process monitoring, or a release decision? If the latter, what is the validated equivalence argument and who has reviewed it?
- Does our qualification of an in-line probe address blanking, drift, cleaning of the sensor itself, and calibration over the commercial lifecycle?
Wider context
This is industry-press material written by a supplier of cleaning chemistries, in a trade magazine whose editorial mix includes peer-reviewed papers alongside technical articles and application notes (7). Read it as vendor-authored content: the experiments are detailed enough to be useful, but they were designed by parties with a commercial interest in the conclusion and have not been independently replicated. Treat the numbers — LODs, LOQs, the ratio range, the under-a-minute clearance — as reported findings, not independent evidence.
The directional trend is nonetheless real: analytical monitoring is migrating from the QC bench toward the process stream, and cleaning is following. The barrier is not sensitivity but demonstrating equivalence to the sampling it would displace.
Sources
- Schallom J, Bosso B, Lopolito P, Frey K, Thanavaro A. In-line UV Spectrometry Monitoring in Cleaning Validation. Pharmaceutical Technology 2025;49(6). Online 29 July 2025. https://www.pharmtech.com/view/in-line-uv-spectrometry-monitoring-in-cleaning-validation
- European Commission. EU GMP Annex 15: Qualification and Validation, Section 10.6.
- Cleaning Characterization of Protein Drug Products Using UV-vis Spectroscopy. Biotechnology Progress (AIChE). https://aiche.onlinelibrary.wiley.com/doi/10.1021/bp070427a
- Rethinking Cleaning Validation for API Manufacturing. Pharmaceutical Technology (2018).
- FDA. Guidance for Industry: PAT — A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance. October 2004. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/pat-framework-innovative-pharmaceutical-development-manufacturing-and-quality-assurance
- Experimental design of a film flow cleaning rig equipped with in-line PAT tool for real-time monitoring. Heliyon 2024. https://doi.org/10.1016/j.heliyon.2024.e34679
- About PharmTech.com and Pharmaceutical Technology. https://www.pharmtech.com/view/about-pharmtech-com-and-pharmaceutical-technology
Validation takeaway
In-line UV spectrometry is credible as continuous process verification of a cleaning cycle, but a fluid-stream optical measurement is not a substitute for surface-based acceptance criteria: it cannot distinguish dissolved from entrained residue, cannot reach unwetted locations, and says nothing about the microbiological state. Keep the release decision on validated swab or surface sampling until a documented equivalence argument supports moving it.
Questions worth asking in our own programme
- What would an in-line UV baseline look like on our worst-case product and cleaner combination, and have we measured the soil-to-cleaner absorbance ratio rather than assumed additivity?
- Which of our acceptance criteria are surface-based (µg/cm², visually clean) and therefore not addressable by a fluid-stream measurement at all?
- If a cycle's UV profile deviates, do our procedures define what happens next — hold, investigate, re-clean — or is there no written response to a monitoring signal?
- Have we characterised foam, air entrainment and particulate scattering in the flow path as potential pathlength and signal artefacts?
- Is our target process monitoring, or a release decision? If the latter, what is the validated equivalence argument and who has reviewed it?
- Does our qualification of an in-line probe address blanking, drift, cleaning of the sensor itself, and calibration over the commercial lifecycle?
