Purpose and Introduction
Recovery studies are a requirement for demonstrating that if a residue is present on equipment surfaces after a cleaning cycle, it can be effectively removed from the surface via the specified sampling method and subsequently accurately detected via the specified analysis method. Manufacturing equipment typically includes numerous materials of construction (e.g., stainless steel, glass, polypropylene, acrylic, silicon, PTFE, etc.). Typically, it is not practical for a firm to perform recovery studies on all materials of construction present in the equipment, and a strategy should be in-place to select the materials of construction for which recovery studies will be performed.
This document provides Hyde recommendations for which materials of construction (MOCs) should be chosen for evaluation when performing recovery studies.
Scope / Background
Several 483 observations have been made by the FDA regarding recovery studies excluding various MOCs that are present in the manufacturing equipment. Below are some examples:
FEI Number
3003981475 (2022) Observation 6 point A was made for “The QC rinse recovery & swab recovery study… failed to include glass (for viewing port) for viewing ports that is swabbed during cleaning validation.”
FEI Number
3000209996 (2023) Observation 3 point C.4 was made for “TOC swabbing recovery studies… failed to access all representative product contact surfaces. For example, gaskets [redacted] were excluded from the studies.”
PDA Technical Report No. 49 (TR 49):
Points to Consider for Biotechnology Cleaning Validation, Section 5.4.1,
General Considerations (for sampling recovery studies), recommends selecting the MOCs for recovery studies by evaluating the overall contribution to total surface area for each MOC. The guidance states that if the overall surface contribution is under a small percentage then the material may be excluded from the studies.
In addition, PDA TR 49 states that recovery studies should be performed for all MOCs sampled. This is corroborated by the observation captured in FEI 3003981475, where the firm was cited for not performing recovery studies for a sampled MOC.
Results / Discussion
Selecting Materials of Construction Based on Contribution to Overall Surface Area
When new equipment is implemented, an assessment should be performed on what MOCs the new equipment is comprised of, how much of each MOC is present, and what components are made with each MOC. In addition, a basic understanding of what samples will be collected for cleaning validation is needed. As stated in the above sections, all MOCs that will be sampled during cleaning validation should be included in the recovery studies. For MOCs that will not be sampled during cleaning validation, the firm may opt not to perform recovery studies, especially for MOCs that constitute a low percentage of the overall equipment train as discussed in PDA TR 49. For MOCs that will not be sampled or included in recovery studies, it is important and expected to assess the overall risk to the process, product, and patient. As an example consideration to be made during the risk assessment, if an equipment component is not included in the wetted surfaces of the equipment train, then a recovery study may not be needed for that component and associated MOC since the risk of carryover from a non-contact surface is low.
As for the risk associated with not sampling low-percentage MOCs (and therefore not performing recovery studies on these MOCs), it should be considered that a particular MOC may have a lower recovery than the other MOCs that comprise the equipment train. If all equipment MOCs are not tested during recovery studies, a low recovery factor from a particular MOC may not be known. Maximum allowable carryover calculations are based on the assumption that any residual product or other residue of concern (e.g., cleaning agent) is evenly spread over the surface of the equipment. Under this assumption, if the unknown low recovery factor (e.g., 50% recovery) is not factored into the carryover calculations, the amount of residue possibly present on the low-percentage MOCs may present a higher risk of carryover than theoretically expected. As an example, for highly potent compounds with low carryover limits, the amount of possible carryover may not be acceptable even on MOCs that constitute a fairly low percentage of the equipment train.
The above scenario presents one example of a consideration to be made when opting not to perform recovery studies on all MOCs. All factors that a firm considers when identifying MOCs that will not be sampled and not included in recovery studies should be captured and justified in a risk assessment.
Selecting Materials of Construction for Swab Recovery Studies
As discussed in PDA TR 49, the FDA expects that if an MOC in the equipment train is sampled via swabbing, the particular MOC should have a swab recovery study performed on it, regardless of the contribution of the MOC to the overall surface area. Hyde recommends following this practice, as the recovery of the same soil from differing MOCs can be significantly different due to surface characteristics such as roughness or porosity. These types of characteristics can impact the residue retention and recovery. When using swab sampling to assess the cleanliness of a material type within processing equipment, it should be studied and understood how well the soil can be expected to be recovered. This provides confidence that the assessment of cleanliness from swabbing is accurate.
Selecting swabbing sites for the full-scale equipment should be based on a variety of considerations; examples are provided below:
- Worst-case locations, i.e., hard to clean areas, should be swabbed. This includes locations that are likely to have residue if cleaning is inadequate. For example, any locations that may have shown riboflavin during spray coverage testing, dead legs, bubble traps, etc.
- Swabbing sites representative of the state of clean of the equipment (e.g., side wall of a vessel below the air-liquid interface) should be selected.
- Swabbing sites should include representative functional locations. For example, transfer tubing, vessel dome or lid, agitator blades, valves, ports, etc.
Note that in the provided examples, each of the selected sampling sites may have a different material of construction. For example, 316L stainless steel for the main body of the equipment (vessels and pipping), EPDM for diaphragms on valves, borosilicate glass for a sight glass, acrylic for columns, PTFE for outlet valves, etc. Any site selected for swabbing should have the corresponding MOC tested during recovery studies.
Selecting Materials of Construction for Rinse Recovery Studies for CIP Processes
When assessing equipment surface cleanliness via rinse sampling, the rinsate to be collected for analysis flows across all surfaces of the equipment during the final clean-in-place (CIP) rinse. There is typically no way to sample a particular single MOC in a system that is cleaned in place and all MOCs get sampled together. Generally, equipment trains have a predominant MOC so the risk of carryover from a non-dominant MOC is much lower. As there is no way to collect a discrete sample from a singular MOC, Hyde typically recommends performing rinse recovery studies from the single predominant MOC in the equipment train. However, there are scenarios in which multiple MOCs should be tested during rinse recovery studies. If the equipment train is comprised of multiple MOCs that are present in large percentages, i.e., there is not a single predominant MOC, multiple MOCs should be assessed. An example of this is a chromatography skid that is often comprised of comparable percentages of acrylic and stainless steel. In this example scenario, it is necessary to test rinse recovery on both noted MOCs.
Selecting Materials of Construction for Rinse Recovery Studies for Parts Washers and COP Processes
Unlike the CIP process described above, when equipment is cleaned-out-of-place (COP) or in a parts washer, it is possible to segregate by material during rinse sampling for the individual equipment. As with swab recovery MOC selection, if a particular MOC is individually sampled via rinse under these scenarios, Hyde recommends that MOC should be evaluated during rinse recovery studies. For example, a glass graduated cylinder and a stainless steel filter housing may be within the same load of a parts washer. If both small parts are assessed for cleanliness via rinse sample collection, both the glass and the stainless steel should have rinse recovery studies performed on them.
Conclusions
It is not always acceptable to perform swab and/or rinse recovery studies on only the primary material of construction of the equipment train. Hyde recommends a comprehensive approach to selecting MOCs for swab and/or rinse recovery studies based on the specific scenario. The approach for typical scenarios is summarized below:
- Swab recovery should be performed on all MOCs that are sampled at full scale.
- Rinse recovery for CIP processes should be performed on the predominant MOC(s), which may include more than one MOC.
- Rinse recovery for COP processes and parts washers should be performed on all MOCs that are sampled at full scale.
If any MOCs that align with the approach defined above are to be excluded from recovery studies, a risk assessment should be performed. This includes MOCs that are excluded based on their contribution to the overall equipment train surface area.
References
- PDA Technical Report 49, Points to Consider for Biotechnology Cleaning Validation 2010
