Transcript Continuous Flow Mfg Skip-Lot Sampling Closing Presentation

Continuous Flow Mfg
Skip-Lot Sampling
Closing Presentation
Topic A
Lynn Torbeck
Frank Gomez
Domenick Amato
Overview







Continuous flow manufacturing
Batches and lots
Suppliers
“Skip Test” vs. “Skip Lot”
Statistical Sampling Plans
USP
Supplier Relationships
Continuous Flow Mfg

Defining Continuous Process:
• Long term vs. short term
• The size of the material exceeds the
capacity of the equipment
• The reactor full vs. ongoing
production
• Discrete activity
Continuous Flow


Common in chemicals to campaign
large batches
Sampling and testing procedures
may help define a continuous vs.
batch process.
Batches and Lots


210.3(b)(2)
See definition of a batch
• Discrete volume vs. ongoing
production

See definition of a lot
• Lot is equal to a batch
• Or, lot is a sub portion of a batch
Batches and Lots


Within the context of the GMP’s a
batch or lot can be defined as part
of the quality agreement between
the supplier and customer
The batch must have uniform
characteristics regardless of
contract.
Batches and Lots



A change in the manufacturing, a
process upset, may cause a
change in the definition of the
batch size.
A batch can be defined by a
predetermined interval, i.e. 6 hrs.
Drug manufacturer is responsible
for assuring compliance.
Suppliers



Excipient suppliers are not generally
qualified, but the Excipient
manufacturers are qualified/
certified.
Need to know the supply chain
including shipping and distribution.
Audit the excipient manufacturer for
compliance.
Suppliers


Depending on the number of lots
purchased choose an interval for
certifying the reliability of excipient
manufacturer’s C of A
Interval of
• 1, 3 times/yr
• Test ten then every tenth lot or one per yr.

Drug manufacturer must test critical
performance criteria. It may or may not
be on the C of A.
Critical Excipients


Drug manufacturer determines
what are critical excipients.
The extent to which GMP’s controls
are applied depend on criticality
and usage or intended use as
determined by the drug
manufacturer.
Defining a Batch


Different drug manufacturer have
different acceptance criteria for
raw material delivery and batch
definition.
For some every shipment is a
different batch, test all
characteristics.
“Skip Testing”


Other companies do an ID test and
accept other tests from the CoA of
the previous shipments of the
same supplier batch.
Quality history should not be used
to sunset testing of critical
attributes
Sampling Plans

Sampling
• Variables with a spec range
• Little or no attribute testing
• Z1.4 and Z1.9, S1 are not used.

Square root of (N) +1 is used
• Composite testing for variables test
• Individual samples for ID but not for
variables test

All specification testing must be done.
Square root of (N) +1






Can be a valid sampling plan
Lot size
Sample size
Accept on zero
Reject on one defect or failure
Matches Z1.4 GL I sample sizes
Sampling Plans


In practice, testing is accept on
zero and reject on one.
Statistical “Skip Lot” testing is not
being done because an ID must be
done on all batches. Other tests
may not be done as the supplier is
doing the tests, “skip testing.”
Sampling Plans


If the supplier doesn't do the
specification tests then the drug
manufacturer must do the test.
Sampling
• Pre delivery sample Very little
• Tail gate sampleVery little
• Pull on site
Common
Skip Testing



API, 10 lots initial, then one a year
Excipients, 5 lots initial then one a
year
A few companies use statistics to
determine the skip testing.
Trending


Some companies are using control
charts and process capability, Cpk,
to monitor and trend data
collected.
Some suppliers allow the
customers to see the control
charts.
USP



The USP can change the
monograph given adequate
scientific justification. The drug
manufacturer may have to change
the filing.
Time span for a change is 9-24
months.
Committees are volunteers.
Changes to Methods


Change to a non monograph
method may require an update of
the drug master file.
If a customer makes a change to a
test for a critical parameter they
may have to update the drug
filing.
Supplier Relationships




Need a strong trusting relationship
between supplier and customer.
Supplier relationship is critical for
supplier certification.
Excipient manufacturer notifies
customer of changes.
ISO document has criteria for
supplier grades, A, B, and C.
Critical Quality Attributes


If a drug mfg identifies a critical quality
attribute that is not a USP test then the
mfg must develop a test.
Look at ICH and FDA guidances to
determine what are the critical
attributes, and FDA SUPAC guidances to
determine the levels of changes stated
therein.
Conclusion


Continuous flow processes can
meet the GMP definitions of
batches and lots.
Within the GMPs for continuous
flow processes, a batch or lot can
be defined by an agreement
between the supplier or
manufacturer and customer.
Conclusions



Square root of (N)+1 is being used
correctly for composite samples.
Square root of (N)+1 can be a
valid sampling plan.
Statistical Skip Lot testing can not
be used since each lot must at
least be tested for ID.
Conclusions


Many companies are doing “skip
test” procedures.
Did not find any practice that
needs to be changed or modified.
Closing Questions / Comments



Confusion exists because most excipient
manufacturers do not conduct all tests
because of their controls. If they do not
run the test it must be clearly indicated
since some one will need to run that test.
(21 CFR 211.84)
In today’s environment of PAT do all tests
really need to be run if excipient
manufacturers have the systems under
control. (21 CFR 211.84)
Does 21 CFR 211.84 need to be revised to
allow for PAT?