Methods and materials
As the VRL program expanded at the author's single site, a larger pool of observers came in. The expanded pool of observers
included 20–30 scientists. In addition, equipment washers and supervisory personnel were trained in VRLs and visual inspection
of clean equipment. Data from the additional VRLs were compared with the original work. The data set was also analyzed for
API, excipient, and formulation correlations.
The multisite study consisted of the company's three international sites: West Point in the United States, Montréal in Canada,
and Hoddesdon in England. Each site determined VRL for the same five APIs. Visual determinations were based on two grades
of stainless-steel coupons from a common source: milled or 316-finish stainless steel (the grade of material on most manufacturing
equipment) and mirror-finish stainless steel. The experimental conditions, as defined previously, were the same as those used
for the pilot plant: a distance of 18 in., a viewing angle of 30°, and a light intensity level of >200 lux.
Personnel at each site weighed approximately 20 mg of the respective test material and placed it into a 25-mL volumetric flask.
They dispersed the material using a newly opened bottle of high-performance liquid chromatography (HPLC) grade methanol, added
the methanol to completely dissolve or disperse the material, and brought the flask to volume. The resulting sample concentration
was approximately 800 μg/mL.
Table I: Residue target concentrations.
|
After verifying that the coupons were visually clean before use, personnel cleaned the coupons with methylene chloride, chloroform,
or methanol to remove any residue from the coupon that might interfere with the residue-spotting process or visual observations.
Spot residues were prepared according to conditions listed in Table I. The range of spots was targeted to span between ARL
at 4 μg/cm2 to below VRL and intended to cover approximately 25 cm2. The target spot size was a 5-cm diameter circle, or 20 cm2. The consistency of spotting areas was important to the eventual spot concentration and, therefore, personell made every
attempt to keep residue spots consistently close to the intended size. Residues were arranged in order of descending concentration
for consistency among sites and spotted by pipetting 100 μL of the appropriate sample onto the stainless-steel coupon. The
residue was then dried and the area of each residue spot was determined.
If prepared correctly, each of the residue spots generally approximated a circle. The area was calculated by measuring the
circle's diameter. If the spot resembled more of an oval, two diameter measurements provided the area. The amount of residue
was divided by the spot area to find the concentration of residue.
The coupons were placed on a flat background and the ambient light level was measured to verify accuracy. The identity and
concentration of each spot was unknown to the observers. The coupons were oriented so that each spot could be observed from
the same distance and angle, and the ability of each observer to see each of the residue spots was recorded.
The highest standard spot for APIs was 100 μL of an 800-μg/mL solution or suspension in methanol. This resulted in a circular
residue with an approximate diameter of 5 cm, which was approximately the size of the swab area (25 cm2). Various decreasing concentrations with the same volume resulted in VRL when the residue was no longer visible to all observers.
If the lowest spotted residue was visible to the observers, then the VRL was reported as "less than" the lowest spotted residue
concentration. For VRLs with the "less than" designation, the lowest tested residue was sufficiently lower than the ARL as
to not pose a significant risk of cleaning failure.
To address the phenomenon of residue appearance, various volumes of the same concentration solution near to VRL were spotted
to complement existing data and to determine if there was an effect on the appearance of residue and subsequent VRL determination.
|
