Originally Published IVD Technology June 2002
In Person
Adapting to change
Overcoming
new industry challenges
 |
|
Joanne
Spadoro, PhD, is senior
vice president of operations for Roche Molecular Systems Inc. (Pleasanton, CA). |
IVD companies are
continually improving their manufacturing processes and organizational structures
to meet the changing regulatory and technological needs of the marketplace.
As one of the major players in diagnostics manufacturing, Roche Molecular Systems
Inc. (Pleasanton, CA) conducts continuous monitoring of its manufacturing processes
in order to provide reproducible test methods for its customers.
In this interview with IVD Technology editor Steve Halasey, Joanne Spadoro,
PhD, senior vice president of operations for Roche Molecular Systems, explains
how the company reviews its manufacturing approach in accordance with the dynamics
of the industry. She also discusses the new challenges presented by emerging
molecular diagnostics technologies and more-complex instrumentation.
IVD Technology: In terms of sheer technical complexity, what are the most troublesome areas involved in manufacturing IVDs?
Joanne Spadoro: I would not call anything troublesome. We do, however, have processes of varying complexity within our manufacturing operation. Manufacturing an enzyme or an oligonucleotide, for example, is more complex and challenging than manufacturing a bulk reagent.
What in particular makes enzyme and oligonucleotide manufacturing more complex than the processing of other products?
The development of reproducible methods that yield highly purified primers, probes, and enzymes is more involved and technically challenging than the methods involved in optimizing the manufacture of a bulk reagent. We perform extensive validation of all our processes and must be absolutely certain that they're reproducible. We have developed very high quality standards of manufacture in order to provide our customers with a reliable product.
What challenges
does ensuring lot-to-lot reagent consistency present?
We have extensive quality control built into our products. The raw materials are tested, then we perform in-process testing on our bulk material, do component testing as needed and, finally, we test the kits. We trend all of our data over time and continually compare and review the in-house data and the product release data.
What do you make of the IVD Directive's requirements for traceable calibrators and reference materials? How is Roche making use of the directive?
We participated
in some of the studies to characterize the World Health Organization standards
for heptatitis C virus (HCV) and HIV.
Once such reference standards are available and accepted in the worldwide community, we use them in the development of new products and for the improvement of existing kits.
Manufacturers
of products for which reference materials don't exist can have trouble
figuring out what to do. How do you handle that?
We had that type of situation early on with HCV. We couldn't compare results from our test with those from our competitors' tests because a copy is not a copy from one test to the next.
We had our methods of making the controls and assessing the copy number, which we were quite comfortable with. It was difficult, however, for customers to compare test results from one kit to another and to draw conclusions. Having a reference standard makes it easier for the customers, and easier for us, because then we have an accepted reference standard against which to test the performance of our kits.
Northern European laboratories have provided the driving force behind the IVD Directive's requirements for reference standards. Are U.S. customers as concerned as European customers are with being able to compare their results from lab to lab?
In the United States, reference labs use either home-brew tests or FDA-approved kits.
Organizations such as the College of American Pathologists (Northfield, IL) actually run proficiency programs where they send the same samples out to multiple labs and look at the concordance of results. Many of the labs that use our products participate in these types of studies and look at how well the kits are performing from lab to lab and how reproducible the results are with a given set of standards.
As a scientist by training, you probably weren't accustomed to handling regulatory issues. What was your reaction when you were first told that you had to meet FDA or IVD Directive requirements?
My first experience
with FDA was when I was still part of the development organization and we were
developing the first PCR-based tests going to FDA. We had several meetings with
FDA where we discussed the products under development, the various technical
aspects of PCR, as well as some of the manufacturing hurdles.
Now, as head of the operations organization, I have had to get accustomed to my organization being the focus of the inspection. But we have a very good experience with FDA. We have had several 510(k) and PMA products approved, and now we have biologic license application (BLA) submissions at the agency as well.
Do you also
conduct in-house audits to prepare for all of the various inspections?
Absolutely. Our quality organization conducts internal audits of many departments within Roche, not just the manufacturing organization.
Many regulatory
requirements and international standards are implemented to prevent manufacturers
from producing bad products. In what ways do those requirements simplify complicated
processing matters for manufacturers?
When there is reference
material available from which we can make a secondary standard or additional
reference material, it allows us to test our kits with a well-characterized
and accepted standard. When we say something is "X" number of international
units, customers have a lot more confidence in the result since it is measured
against a known standard.
With regard to
regulatory requirements and international standards, the discussion is a little
more complex. These requirements are put into place, as you indicate, so that
all manufacturers are using consistently acceptable practices. This consistency
moves industry toward a standardized approach for manufacturing and design of
products.
Do you rely on in-house expertise, or product-testing guidelines, to determine
statistical reliability?
We do a combination
of things. We reference statistical methods and we have spoken with statisticians
to make sure we have sound statistical analyses.
Are regulations and standards equally useful when dealing with cutting-edge
products like molecular diagnostics?
Yes, they are helpful. With cutting-edge technologies, the regulatory requirements are not always clear until guidelines and standards are established.
Are there areas
where regulatory agencies are expecting too much data?
Our experience
with the various regulatory bodies has been good. Yes, there's a lot of
information that is required. We were the first ones to really go in front of
FDA with an amplification-based technology. That posed challenges for both FDA
and for us. So we did have to provide a lot of information on all the various
aspects of the kit performance and our manufacturing processes, which I don't
think is any different from what is provided for an immunochemistry test.
I think the challenge also comes in determining the clinical utility of these kits. In the United States, a test doesn't get approved simply because it detects the number of copies of RNA that it claims to detect. A manufacturer has to prove that there's a clinical use for the test. That is probably the biggest challenge because those types of studies generally take a longer period of time to conduct.
What areas of
the IVD Directive, which will become effective in December 2003, have you found
most difficult to deal with?
Much of the information
required is similar to what we put together for an FDA submission. Since I don't
have any familiarity with what the inspection process is going to be for the
IVD Directive, I am not sure what to expect or how challenging it will be.
Has the language requirement, which introduces some complexity into the packaging
lines of manufacturing facilities, been an issue for Roche?
Actually, we are in the process of discussing many of these issues internally. Because Roche already makes blood-screening and other IVD products that are used worldwide, we already have some of the requirements incorporated into our labeling. There are some additional requirements, such as those governing the use of symbols. However, as with any government requirements, regulated industry continually monitors the state of the requirements and moves toward implementation, as applicable, for their product line.
Have you streamlined
your activities to eliminate redundancies between what's required for FDA
and for the IVD Directive?
We have to do separate
submissions. The way we deal with our nonclinical studies, however, is by designing
them such that they're applicable to numerous regulatory submissions.
We try to design
our studies knowing and accommodating the various requirements. There may be
situations in which there are some unique requirements and we've simply completed
these requirements as needed.
Platform Flexibility
Are challenges different for companies that are involved more in instrument
manufacturing than in biochemical processing?
Yes. The Roche
Instrument Center (Rotkreuz, Switzerland) manufactures our instruments. We face
many challenges because we are developing instruments and reagents simultaneously.
We want both the instruments and the reagents to be of high quality.
One area that is a big challenge in terms of instrument development is validation. If you speak with the instrument people, I'm sure they will say that the reagent challenge pales by comparison with the challenges of instrument development.
When you organize
a test to be run on a particular instrument, how early in that process do the
engineers and the chemists start working side by side with one another?
The engineers and
chemists work together very early in the process. For instance, when we developed
our Cobas Amplicor instrument, which was our first automated system for amplification
and detection, the lead engineer from Rotkreuz, Switzerland, actually moved
to the United States and worked with us in the reagent development facility
for several years.
Since the reagent
and instrument development cycles are integrally linked, it is critical that
the engineers and the development scientists work together from the beginning
of the development cycle. During each stage of the process, we verify that the
reagents and instruments are performing as anticipated and within specification.
Once the instrument
is available, we develop our assays somewhat independently of the instrument
engineers unless there are software or hardware upgrades that are required for
new products.
Once a platform has been developed, do new tests have to fit the platform?
Yes, but we have some flexibility. The Cobas Amplicor was developed initially to perform qualitative polymerase chain reaction (PCR) assays. However, over the course of time we realized that there was a great need for quantitative tests. So we worked with the engineers at Rotkreuz and modified the software on the instrument so that it had the flexibility to do both qualitative and quantitative tests.
If the initial
product has a certain limit to it, you are kind of stuck with that, aren't
you?
Yes. I think any
platform has its limitations. As the technology evolves we have to decide whether
the tests can be adapted to the platform, whether the platform can be modified,
or whether we're simply constrained and have to think about an alternative
platform.
We have to be sensitive, however, to the number of platforms that we introduce into the field.
Molecular Diagnostics
What special difficulties are presented by manufacturing molecular diagnostics
as opposed to traditional immunoassays?
Molecular diagnostics,
of course, is a relatively new field. Compared with more-classical immunochemistry
kits, our kits contain numerous reagents.
There are additional
technical challenges. We deal with the detection of nucleic acids, and RNA,
in particular, degrades quite easily. As a result, we manufacture many of our
reagents under RNAse-free conditions.
In addition, when
we make our control material, we make it at very low copy levels—between
10 and 20 copies per PCR. Initially, it was a bit of a challenge to figure out
how to make these controls at such low copy levels and maintain their stability.
Additionally, we're
not making sterile products, but we want to minimize DNA contamination due to
the sensitivity of the technology.
Many companies license PCR from Roche and a lot of other companies have tried
to find other methods of amplification to get around using PCR. What is your
expectation for the future of such automated molecular systems?
The technology will evolve the same way immunochemistry evolved to become routine and walkaway automated systems, where there's very little manual intervention on the part of the laboratory analyst. Automated sample extraction will replace manual methods, and rapid PCR amplification cycles will reduce the total time to a result.
Organizational Structure
You have responsibility for a number of Roche's facilities in the United States.
What role does a company's organizational structure play in simplifying or complicating
product manufacturing?
I think the biggest
challenge for us in terms of our organization is that we have product development
on the West Coast and the manufacturing and quality activities on the East Coast.
Previously, it was all in a single facility.
Although a challenge, we have developed processes that minimize the impact of the geographic divide.
Despite Roche's
vertical integration, does Roche outsource any of its development or manufacturing,
and what would be the business rationale for that?
Actually, we do outsource some of our manufacturing activities. For example, we outsource the manufacture of our positive controls. We also outsource the manufacture of our Amplicor HBV test. When we have scientific or other compelling reasons, we outsource. In general, however, we develop and manufacture the vast majority of our products.
In the current
economy, is there more or less pressure on companies to look for a better deal
outside?
The decision to outsource is really collectively made based on the business needs. I'm not pressured to outsource. We do ask under certain circumstances, however, if it is a better solution. Many times we find that the amount of work that goes into outsourcing coupled with the cost do not result in the most effective solution. Even if it involves additional investments, we generally find that doing it internally makes the most sense.
Would a smaller
IVD company have a different solution?
We are fortunate
to be part of a larger pharmaceutical company and do not have some of the financial
challenges that small independent start-up companies may have. We are in a highly
regulated business, as a lot of your questions referred to, and it's not
as simple as manufacturing and sticking something in a package. It's about
adhering to a quality system that results in reproducible high-quality products.
Considering this, outsourcing may be a good alternative for some smaller companies.
Looking Ahead
What actions is Roche taking to prepare for future markets?
It is, of course,
necessary for us to adhere to worldwide regulatory requirements. As a result,
we have expanded our thinking and planning during the product development phase
to include the requirements defined under the IVD Directive in addition to FDA
requirements. We want to be in a position to rapidly obtain CE registration
for our products.
To meet our changing customer demands, we have recently begun introducing another line of platforms, which includes automated specimen preparation and our TaqMan amplification technology. These instruments will hopefully help drive nucleic acid amplification–based testing in the direction of fully automated walkaway systems allowing the laboratory analysts to focus on other activities.
Copyright ©2002 IVD Technology
