Rouge is a common phenomenon in water-for-injection (WFI) and clean-steam systems and can often be found in equipment and
piping connected to these systems. The root cause of rouging in high-grade stainless steel equipment is a contentious issue
among pharmaceutical engineers and definitions of rouge range widely. What is generally agreed is that rouge is a form of
surface corrosion that can form on stainless steel when exposed to high-purity water at elevated temperatures. This corrosion
can manifest itself as a glazed, immobile discoloration or as a powdery reddish deposit (1).
Opinions on how to investigate and assess the impact of rouge and choose appropriate corrective and preventive actions (CAPA)
are diverse and sometimes conflicting. Essentially, however, the quality assurance (QA) specialist or product manager must
assure that the equipment is not reactive, additive, or absorptive, as per 21 CFR 211.65(a) (2). Questions to investigate may include:
- Is this rouged metal a good spot for biofilm development?
- Is iron oxide particulate present in the final solution?
- Could iron oxide affect the stability or effectiveness of the product?
Investigating impact
Rouging on a stainless-steel tank. (IMAGES COURTESY OF THE AUTHOR)
|
The team investigating rouge can collect and analyze existing data from their process. For example, surface roughness (Ra)
would have been measured during installation qualification of the WFI still and piping. Further data would be available from
the operational and performance qualifications on clean-in-place system tanks. The team can also analyze water testing and
monitoring data over the system lifetime, and look at the impurity profile for products made in these tanks or with this WFI.
Using quality systems and release criteria from years of operation, investigators have many of the tools needed to assess
any potential effect to the product or intermediates from the observed rouge.
Rouging on a water-for-injection pump impeller.
|
Investigators should first determine whether the observed problem is indeed rouge and not product residue or another type
of contamination. If a swab from the surface comes away with reddish-brown residue, the residue can be sent to a laboratory
that performs Fourier transform infrared spectroscopy (FTIR) analysis. This analysis will show the presence of iron oxide
from a sample of rouge. Close visual inspection should also be used to assure that there is no pitting or flaking in the affected
area, which would indicate an advanced level of corrosion beyond the scope of rouging.
If rouge is present, the most critical determination is whether the rouge is additive to the process or not. Is the rouge
firmly adhered to the surface of the equipment or could it instead come off into the process stream either slowly or as a
result of water hammer or another shock? If the rouge is glazed and smooth in appearance and does not wipe off or is not found
on downstream filters or gaskets, then it would be safe to conclude that it is not an additive problem. A second critical
determination is whether the rouged location is cleanable. If it is not cleanable, it could be a spot where biofilm could
develop, which could impact the process. One should assess the cleanability of the rouged area by comparing the surface roughness
to areas with no rouge in the same tank or piece of equipment. A review of any cleaning or monitoring data from the system
can be used to corroborate whether the rouge is additive and cleanable.
If, on the other hand, the rouge wipes off easily on a swab or a glove, or is found downstream on filters or gaskets, then
it can also be found in the WFI or process stream. The equipment can indeed be said to be "additive, reactive, or absorptive,"
which is a violation of 21 CFR 211.65(a). In this case, one must investigate the impact of iron oxide particulate in the WFI or other process fluids.
Factors such as dilution or downstream filtration may contribute to an impact assessment. The engineering team can calculate
the surface area of equipment effected by rouge and the amount of rouge particulate possible in the system. This calculation
may reveal that the dilution factor is so small that no further investigation is needed. WFI added to any critical downstream
process will most likely have a sterile filter in line as well. Although rouge particulate may be as small as 10 nm, particulate
normally clumps into larger sizes, and is trapped in sterile filters (3). Conductivity, which is a monitored criteria in WFI
loops, is not elevated by rouge because the volume and usage of the WFI storage and distribution system is too great in comparison
to the volume of iron oxide particulate that could be formed and released into the circulating WFI in a measurable amount.
