| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Oak Ridge Conference |
1
Departments of Laboratory Medicine and Pathology and
2
Urology Research, Mayo Foundation for Medical Education and Research, 730 Hilton Bldg., Rochester, MN 55905.
a Author for correspondence. Fax 507-284-9758; e-mail okand{at}mayo.edu
Abstract
New assay development should be directed toward answering fundamental clinical questions. Caveats that must be considered before initiating assay development projects are: New assays should allow the clinician to interact with and treat a patient more effectively, thereby improving medical outcome; and new assays should facilitate recapture of system resources, enabling cost savings or reinvestment of resources. Defining the clinical questions and consideration of the caveats permit a means of prioritizing assay development activities. Laboratorians are faced with evaluating several types of development activities that lead to assay implementation in routine clinical testing. Assays can be prioritized for up-grading to newer cost-effective technologies, provided the changes maintain or improve analytical and clinical performance. Predicting which research assay will have future value is difficult when clinical performance is not fully validated. However, such assay development has the greatest potential for changing the delivery of healthcare by a clinician.
Assay conversion, assay development, and new assay implementation activities should be directed at answering fundamental clinical questions. These activities may have several identifiable objectives that focus on increasing clinical value of laboratory testing, which will enable clinicians to interact with and treat patients more effectively. Ultimately, these activities should improve the delivery of healthcare and improve patient outcome. Furthermore, assay conversion and implementation activities should facilitate resource allocation or reallocation and should be cost-effective. A large body of research has been developed by national and international organizations pertaining to the generalities and specifics of assay development (1)(2), minimum performance criteria (3)(4), defining reference ranges (5), and standardization of immunoassay (6)(7)(8), urinalysis (9), and molecular diagnostic assays (10)(11)(12). A review of this vast literature is beyond the scope of the present work. Instead, we take a top-down approach, in which general concepts regarding new assays are presented and illustrated with specific examples and actions taken at the Mayo Clinic that are performed within NCCLS and alternative methods guidelines (13). The perspectives presented are not to be construed as "the laboratorians' consensus statement" on new tests. Instead, this is an opportunity to discuss a unique, potentially divergent, practice pattern in which new approaches are attempted to overcome assay performance difficulties and to circumvent obstacles to change. The objectives are: to present specific results and experiences derived from assay conversion and new test implementation in an integrated practice setting that assays patients' samples from the clinic, hospital, and reference laboratories; to consider obstacles to change that impede new assay implementation; and to provide specific examples of new tests developed in basic science and clinical research laboratories that may have enhanced clinical utility. Although the approaches presented may be practiced in a unique setting, the general and specific examples may be modified or adapted for use in other practice settings.
Cost Effectiveness of Assay Conversion, Development, and
Implementation Activities
New assay development activities in the past two decades focused mainly on two issues: improving assay performance and quality through development of Reference Methods, defining reference materials, providing assay validation criteria, utilization of consistent nomenclature and terminology (14)(15), and improving analytical sensitivity through introduction of new detection methodologies and instrumentation (16)(17)(18)(19). Both of these efforts have lead to improved clinical utility. Recently, however, healthcare cost effectiveness has become a driver of assay utilization and change (20). Cost effectiveness is normally measured as the ratio between the changes in costs and changes in outcomes obtained after implementation of new procedures (21). Costs are separated into several categories, including:
• direct medical costs
• direct nonmedical costs, such as travel and accommodation costs for a patient's family members
• indirect costs, such as lost income from a patient's inability to work
• intangible costs, such as patient anxiety and stress
Outcomes considered are typically:
• decreased morbidity attributed to assay changes, e.g., implementing assay changes that improve nutrition testing for geriatric presurgery patients
• years of life saved (postponed mortality)
• improved quality of life
This cost-effectiveness approach is useful for retrospective health sciences evaluations that ascertain clinical utility regarding screening populations applied to large populations. Furthermore, this approach drives replacement of tests for older analytes by those for newer analytes, which have improved clinical utility. However, the appropriateness of this cost-efficacy approach is not immediately apparent when the clinical laboratory is confronted with making decisions regarding assay conversion and test implementation. One problem is that the laboratorian may not be involved in assessing these outcomes. Outcomes are generally assessed by health sciences or statistic units. Costs can be quantified or estimated from this model, but some of the outcome endpoints are not readily obtained or may not be amenable to quantification. Although decreased morbidity resulting from an assay change would be easy to quantify, postponed mortality data are more difficult to obtain on a useful time scale. For example, survival data in oncology research on slow-growing cancers may be difficult to acquire. Quality of life outcomes in oncology, for example, are second in importance to years of life saved (22)(23). However, quality of life is difficult to quantify and may be assessed differently by different patients. Consequently, the impact of laboratory testing on the outcomes described above may be difficult to document. It may be better to consider cost effectiveness of assay changes by using different endpoints. Changes in costs could include:
• assay-associated costs
• systems costs
• direct nonmedical costs
• intangible costs such as anxiety
Outcomes in cost-effectiveness studies are improvements realized from the assay conversion activities that can be monitored in terms of:
• changed analytical and clinical performance of the assays
• changes in turnaround time (TAT)1 for performing the new assays
• ability to utilize resources more effectively
With this model, changes in cost are readily quantified and outcomes are readily identified and are objectifiable. Changes in analytical performance include improved detectability limits, improved linearity upon dilution, improved precision, and reduced CVs. Reducing CVs is important in guarding against generating requests for unnecessary patient testing. With some assays in the Mayo practice setting, an increased assay CV may result in results being reported for some patients that erroneously exceed a decision threshold. This may automatically trigger these patients to be scheduled for unnecessary additional testing by a subspecialty area. Consequently, emphasis has been placed on improving analytical performances of assays.
Emphasis has also been placed on decreasing TAT: At Mayo, at least 10% of the multiday tests are required to be replaced by same-day testing each year. The changes in TAT are important for several reasons. The requested TAT for assay results from clinicians varies with the setting. Immunosuppressive drug monitoring results are not needed until 1600, so a TAT of 6 h is acceptable for cyclosporin and FK506 assay reports. In contrast, glycohemoglobin results are requested to be reported to the Diabetes Clinic within 15 min of blood draw. In this case, patient monitoring and follow-up are closely linked temporally. This linkage avoids writing follow-up letters and scheduling revisits, and provides an opportunity for assuring tighter compliance with treatment and diet. Improved TAT may also permit clinicians to deliver healthcare to patients more rapidly. It permits assessment of whether a patient should remain in the hospital for additional care or may be legitimately dismissed. It may permit a clinician in a clinic to perform initial patient test ordering and reschedule a follow-up patient's visit on the same day. It may reduce unnecessary waiting caused by delayed test results and also reduce the number of patients' complaints. In this sense, improved TAT is a real-time quality of life indicator for the patient and a real-time customer service marker for the medical institution that can be readily identified and quantified.
Types of Assay Conversion, Development, and Implementation
Activities
Several different types of activities contribute to assay conversion, development, and implementation, including:
1. Replacing an assay for a particular analyte with an assay for a more useful analyte that has improved clinical value.
2. Converting existing assays for an analyte to new methodologies.
3. Reducing assay testing by introducing a screening procedure that eliminates patients' samples from the workstream.
4. Developing new assays through translational research activities.
The first three assay conversion activities can be regarded as continuous improvement projects. These activities improve the quality and cost effectiveness of assays but do not offer the greatest chance at changing how a clinician will interact with a patient. The fourth, translational research activities, may be defined as the process of implementing promising research discoveries for clinical intervention through prevention, improved diagnosis or prognosis, or implementing smart, rational therapies, and, within the context of this discussion, new assays (24)(25)(26). Translational research offers an excellent chance of changing how the clinician will either treat a patient or be able to interact with a patient through the introduction of tests for new clinical markers or through introduction of new technologies (27). The potential gains offered through translational research activities are balanced by the risks that the research may subsequently be demonstrated to be nonproductive. Introducing new tests or new methodologies has to be regarded as a managed risk situation in which many assays that are developed may ultimately prove not to be useful. An example of this is translational research on cytokine growth factors used to reduce neutropenia after chemotherapy (28). Although this process is now widely utilized, a recent report indicates that granulocyte colony-stimulating factor should not be administered to afebrile patients because of the lack of clear-cut clinical benefit (29).
replacing an existing assay by an assay for a more clinically
useful analyte
Established analytes may be slowly displaced from use by assays
for new analytes that have greater clinical utility. This assay
replacement activity takes place over a long period and is driven by
health sciences research evaluation of clinical utility and cost
effectiveness. Clinical utility of assays for established analytes
should be reassessed periodically to determine whether a more useful
test exists that should be considered for introduction. This
reassessment function is performed in our practice by a Clinical
Pathology Work Group (CPWG in Fig. 1
) composed of representatives from all aspects of laboratory
medicine, administration, core and reference laboratories, and
technology. In addition, experts in clinical areas are invited to
contribute to the development of clinical utility guidelines for
assays. This offers a mechanism to introduce assays for a new, more
clinically useful analyte, to replace existing analyte assays. This
assay implementation activity frequently sparks vigorous, healthy
debate on the clinical utility of the old and the new assays. Obstacles
encountered in replacing assays for older analytes with new analyte
assays include entrenched personal agendas and biases that must be
capable of being overturned if these positions are not supported by
impartial data. Classical examples of this type of assay conversion
activity are: use of total creatine kinase (CK) and CK-MB as early
markers of myocardial infarction instead of lactate dehydrogenase
isoenzyme analysis (30), and replacement of the LE (lupus
erythematosus) cell assay by testing for antibodies against
double-stranded DNA for diagnosis of systemic lupus erythematosus
(31). An evolving example of this activity may be the use
of assays for serum tryptase in mast cell disease (32) in
place of assays for histamine or 11ß-prostaglandin F2
.
|
converting existing assays for analytes to new methodologies
The forces driving conversion of existing assays to newer methods
may differ with the analyte or medical environment, but generally this
activity is directed at reducing assay-associated costs and improving
analytical performance and improving TAT. The obstacles to implementing
this type of change may include embedded personal agendas, as well as
lack of support from clinicians if the new assay reports in units
different from those in the old assay, or if there is a change in
reference intervals. Another obstacle to change that may be unique to
large organizations is that multiple laboratories may be interested in
utilizing assays for a single analyte. This may result in competing
assay conversion efforts, which will inappropriately consume
development resources. To reduce the obstacles to assay conversion, we
have developed a defined process to coordinate and facilitate these
activities. An Assay Conversion Team has been implemented to identify
assays that have high priority for conversion from manual RIAs to
nonisotopic and automated immunoassays (Fig. 1
). After discussions
between the clinical chemist or laboratory leader and the potentially
affected clinicians, the process is started with a decision about
whether the current assay for the analyte has sufficient clinical value
to warrant continued performance of the test. Advice is solicited from
the CPWG if concerns are raised about the clinical utility of the
existing test.
Assays judged to have sufficient clinical utility are then assessed for
their potential for automation, either with an existing commercial kit
or a reagent system or with in-house assay development activities when
necessary. Assays compatible with automation are then prioritized for
conversion/development on the basis of several factors, including the
annual volume of the test, potential TAT improvement, and potential
reductions in assay and system costs. Analytical and platform
performance characteristics of the old assay are compared with those of
the new format, and clinician input is solicited at this stage. This is
similar to the method selection, evaluation, and development loop
presented in Fig. 1
of Koch and Peters (33). The
clinicians who utilize the assay must agree that the new assay and its
performance characteristics are satisfactory and meet the clinical
need. Subsequently, a preliminary financial impact analysis is
performed for the several automated immunoassay platforms available.
Approval to proceed with assay conversion is given with assignment of
development personnel after having achieved consensus that the assay
and clinical performance characteristics are acceptable for a given
platform and that there is a positive financial impact.
Not infrequently, problems are encountered in converting an old assay to a new methodology. This topic has been dealt with regarding specific analyte assay conversions (cortisol and human chorionic gonadotropin), and summaries of the many difficulties and approaches taken to overcome them are provided (34)(35). Often, there is a change in reference interval values because epitope recognition by antibodies in the old and new immunoassays may differ. A second problem is that the new assay methodology may not adequately assay all of the patients' samples from a given patient population. Some patients' samples may fall below the range that the new assay was designed to measure. This problem is aggravated by the use of 2 SD above the mean of the zero standard as the detection limit in kit information inserts, which is inappropriate for analytical validation. The functional sensitivity of the assay, i.e., the concentration of analyte at an interassay CV of 20%, provides a better estimate of the limit of quantification in routine practice. Furthermore, some patients may have interfering substances that the manufacturer has not considered.
A specific example of this is a recent conversion of testosterone from
an organic extraction-based assay to a direct immunoassay on a
random-access automated platform. The existing assay, involving
extraction with methylene chloride, eliminates interferences from many
of the water-soluble testosterone conjugates. The direct immunoassay
for testosterone that we recently implemented correlates well with the
extraction assay except at testosterone concentrations <650 ng/L (Fig. 2
). Accuracy in quantification of these low concentrations of
testosterone is required in several clinical contexts, including
deciding the sex of a child born with ambiguous genitalia and in
suppression of precocious puberty. Part of the inaccuracy of the direct
assay, and poor correlation with the extraction assay, is attributable
to the presence of testosterone conjugates in patients' samples.
Testosterone sulfate competes stoichiometrically with testosterone in
this direct immunoassay (unpublished). A second problem is that
triglycerides (36) show a small but significant
interference in the direct assay after extraction. These problems can
be minimized if the samples are pretreated with lipase to reduce
triglyceride interference and an organic extraction step is implemented
on samples with low amounts of testosterone (unpublished). This
eliminates the reporting of high outliers for samples containing
water-soluble conjugates of testosterone and reduces the positive bias
noted with lipemic samples (36).
|
Evaluation of the testosterone assay conversion from an extraction-based RIA to chemiluminescence on a random-access platform showed the conversion was cost-effective. The cost reduction of ~45% came predominately from a decrease in personnel utilization on this assay. Furthermore, TAT was improved from second-day reporting to a same-day reporting with a TAT of ~<4 h, which meets institutional goals and clinicians' expectations. Analytical performance was improved, the imprecision (CV) improving slightly from 20% at 100 ng/L with the original extraction assay to 13% at 50 ng/L with the new direct immunoassay. There was better utilization of the random-access platform, and clinicians were able to interact more readily with patients on a same-day basis. Overall, this was a cost-effective assay conversion, although imperfect because sample preprocessing (lipase treatment and extraction) could not be eliminated and so were incorporated into the assay protocol. The preprocessing steps are now the subject of an automation feasibility study.
introduction of new screening procedures
Introduction of inexpensive prescreening techniques to eliminate
normal samples from the work flow is the last type of continuous
improvement activity now being considered. A specific example of this
is in urine screening. A routine urinalysis may be performed on every
urine sample submitted even with no evidence of clinical justification.
At Mayo, ~80% of urine specimens have normal values for erythrocytes
and leukocytes, whereas ~10% contain abnormal numbers of either cell
type. However, all urine specimens are evaluated either by automated
imaging or by a manual microscopic evaluation. Routine urinalysis
represents an opportunity to reduce the number of specimens that
require additional testing, i.e., to reduce the number of normal
samples that enter the work stream and require additional work-up by
automated imaging or manual microscopic assessment. The second
objective is to obtain a cost-effective improvement in clinical
sensitivity of leukocyte detections in urine. Leukocytes may be present
in urine for a variety of reasons, including cystitis, various forms of
nephritis, autoimmune diseases, reaction to drugs, and
Chlamydia infections (37). The goal is to
detect leukocytes that may indicate conditions other than urinary tract
infections.
Although leukocyte esterase used in conjunction with nitrite dipsticks may sensitively detect urinary tract infections (38), use of the dipsticks alone may not be sufficiently sensitive for detecting urinary leukocytes in the absence of infection (39). The chemistry used to detect leukocytes involves dehydrogenation of an acridan substrate by peroxidase and hydrogen peroxide, giving rise to an acridinium molecule that then undergoes thermal decomposition and gives rise to chemiluminescence (40). A microhematuria/hemoglobinuria assay performed at the same time utilizes the pseudoperoxidase activity of hemoglobin. In this assay, the erythrocytes are first lysed and then reacted with hydrogen peroxide and a luminol derivative that also gives rise to chemiluminescence emission (41). These screening assays are set in parallel at present.
In terms of sensitivity and specificity, the leukocyte screen has 88% clinical sensitivity for detecting a minimun of 5 leukocytes per high-powered field and a specificity of 92%. Of those samples tested for leukocyte chemiluminescence, only 20% are sent on for additional analysis. Of these, 62% have >5 leukocytes per high-powered field, but the other 38% are false positives. The microhematuria assay has a clinical sensitivity of 99% and a specificity of 96%. Of those samples tested, only 15% are sent on for additional analysis, including microscopic examination or automated imaging. Of these 15%, 74% are abnormal, having >3 erythrocytes per high-powered field, and 26% are false positives. The minimum number of samples eliminated by utilizing the leukocyte and erythrocyte prescreens is ~68%—a marked reduction in the amount of work that has to be redirected to definitive testing by manual or automated imaging.
The cost effectiveness of the urine prescreens has allowed deferring the purchasing of additional automated imaging equipment, reduced reagent costs compared with dipsticks costs, and reduced personnel costs. In addition, resource utilization of the automated imaging system is improved. TAT is improved slightly and there is no degradation of analytical performance. Overall, this is a cost-effective development project.
Implementation of New Tests Through Translational Research
The greatest potential for changing how a clinician fundamentally interacts with a patient can be provided through introduction of new assays. This can be achieved by translating results from basic research or clinical research into the clinical protocols (24)(25)(26). Translational research activities related to new test introduction can be divided into two areas: new markers for clinical conditions, and utilization of new technologies.
new markers
Clinical research produces a large number of markers that have
potential clinical utility. A recent Medline search for prognostic
markers for breast cancer management produced a list of >30
potential candidates that might have clinical impact. If the past is an
indicator of future laboratory utilization, most of these potential
markers will not be implemented in clinical practice (42).
However, a few may find acceptance based on clinical utility within a
limited medical context. An example of this is the present situation
with c-erbB2/HER2/neu. Although this marker
has been demonstrated to be a prognostic marker in breast cancer, it
has since been determined not to be an independent prognostic marker,
in effect providing redundant information (43).
c-erbB2/HER2/neu has not been implemented
widely for prognosis of molecular staging in breast cancer. More
recently, clinical utility for
c-erbB2/HER2/neu assays may have been
demonstrated in a more tightly constrained medical context, in that a
cytotoxic antibody to c-erbB2/HER2/neu has
demonstrated therapeutic utility in a subpopulation of women with
recurrent breast cancer (44). Implementation of the assay
for c-erbB1/HER2/neu may be considered
clinically useful, pending evaluation of the ongoing clinical trial.
From the above, one may infer that clinical research and translational research carry with them certain fiscal risks. However, managed risks must be taken to improve disease detection, management, and patient care. The key to successfully managing these risks is to first define the question to be addressed by an assay for a new marker in a tightly constrained clinical context. Assuming collection of favorable clinical data, the tight constraints imposed permit an estimation of potential changes in clinical outcomes resulting from utilization of the new marker assay.
As an example of ongoing work in this area, let us consider telomerase, a new oncologic marker that has received considerable interest in the last 3 years (45)(46)(47)(48). A ribonucleoprotein, telomerase extends the ends of chromosomes (telomeres) both in vivo and in vitro through reverse-transcription. Telomeres are degraded during each cell division because of replication problems inherent to linear chromosomes; the continued loss of telomeric DNA is believed to ultimately trigger cell death. The absence of telomerase from most normal cells results in these cell lineages having a finite life span. However, telomerase activity is present in most immortal cells, including most malignancies (typically >90%) (45)(46). For reasons similar to those discussed above, telomerase would be difficult to introduce as a new oncologic marker in a broad clinical context, despite its prevalence in malignancies. A major reason for this is that many oncologic diagnoses are rendered after visual examination for differences in histology and nuclear morphology of either paraffin-embedded sections or frozen sections intraoperatively. Frozen sectioning takes only a few minutes to perform, and diagnostic information is conveyed to the surgical team by telecom or face to face (49). The rapid TAT facilitates scheduling appointments for surgery patients with medical and radiation oncologists. It would be difficult to replace evaluation of permanent or frozen sections with a telomerase assay because:
1. Telomerase assay results may provide redundant information when compared with conventional histologic and microscopic evaluations of samples obtained by invasive sampling; meanwhile, microscopic detection of neoplastic cells is efficacious.
2. The error rate for pathological diagnosis by intraoperative frozen section evaluation is ~0.1%; telomerase activity, however, is present in at most 95% of tumors, resulting in a false-negative rate of at least 5%.
3. TAT for the telomerase assay would be ~6 h, including transport, sample preparation, and report generation and verification.
However, telomerase may be a useful marker if the clinical context is constrained. Telomerase detection may yet find use with fine-needle biopsies. Telomerase may find utility for detection of micrometastases to the axillary lymph nodes in "node-negative" breast cancer (48). Micrometastases are frequently missed with frozen sections and necessitate an additional day for inspection of permanent fixed sections for confirmation. Potentially, the detection of telomerase in this context may alter how the medical oncologist will treat the patient with hormonal, chemo-, or adjuvant therapy.
A more immediate example of how telomerase may prove clinically useful is in the limited context of noninvasive sampling (50)(51): Can telomerase be used for minimally invasive or noninvasive early detection of malignant neoplasia? Primary transitional cell carcinoma (TCC) is usually detected by using urine cytology and cystoscopy after a patient presents with hematuria or discomfort. Although FDA-approved tests are available for detecting recurrent TCC (52)(53)(54)(55), none for detecting primary TCC is available. Telomerase activity in bladder washes and spot urines containing shed uroepithelium is under evaluation for detection of primary TCC (50)(56)(57). Telomerase activity is measured in vitro by the enzymatic extension of a synthetic DNA primer, followed by PCR amplification (58). The telomerase amplification products are separated by electrophoresis on miniDNA sequencing gels and visualized by Sybr Green staining. The analytical time is 4.5–5 h. Typically, a 6-bp ladder is visualized in the presence of active telomerase, whereas no visible gel products are observed in the absence of telomerase activity. Urines containing cells from TCC produce easily detectable telomerase-extension ladders, whereas these are absent when unaffected urines are examined. Clinical sensitivity for detecting TCC with telomerase as a marker in urines or bladder washes collected at the time of cystoscopy range from 62% to 73%. In a more limited ongoing study of primary TCC, we obtained 94% sensitivity (15 of 16) by utilizing telomerase detection with bladder washes (57). This suggests that telomerase should be further investigated in a larger study to ascertain if it is be a useful marker for minimally invasive identification of primary TCC in patients with a consistent presentation. If successful when used with voided urines, positive telomerase assay results may initiate a reflex of those patients for cystoscopy. Other potential applications based on the clinically constrained context of noninvasive or minimally invasive sampling may be: early detection or stratification of pancreatic cancer patients by using ductal cells in secretin-induced pancreatic secretions; early detection or stratification of cholangiocarcinoma by using ductal cells in cholecystokinin-induced bile duct secretions; and colorectal cancer screening by using shed colonic epithelium (51).
Utilization of telomerase cannot be viewed as cost-effective at this time; clinical utility has yet to be fully demonstrated and only potential changes in cost are available. There are potentials for major cost-reductions in healthcare from early detection capabilities and high clinical sensitivity. The major anticipated outcome, however, would be increased quality-adjusted years. Unfortunately, no evidence at this time indicates that early detection will decrease mortality from these cancers.
new technologies
How a clinician interacts with a patient or the patient's family
may also be affected by the implementation of new assay technologies.
The detection of heterozygous carriers of genetic diseases is an
example. Provided patient's record confidentiality can be maintained,
there is promise of reducing medical costs by identifying those
patients who have a genetic predisposition to a disease, or who are
carriers of such traits, and providing the appropriate medical
surveillance. The difficulty is eliminating a large number of normal
patients from the screening protocol to make the testing procedures
cost-effective.
Hemochromatosis presents an example of the need to reduce the number of patients evaluated for a particular mutation. Carriers of the trait for hemochromatosis are not rare (~1 in 14 Caucasians), but penetrance is low. It would be difficult to justify screening every Caucasian for the polymorphism(s) giving rise to hemochromatosis. A more effective approach might be to prescreen the population to enrich for patients who may benefit from monitoring and treatment by prophylactic phlebotomy (59). The approach taken once again is to define a narrow clinical context and to implement an inexpensive screening assay that reduces the number of potential positive samples sent on for additional definitive mutational analyses.
Cystic fibrosis (CF) is the model genetic disease we are using to assess the utility of the mutation detection systems. CF is a rather common autosomal recessive genetic disease with a carrier frequency of ~1 in 30 in European populations. Although many mutations result in the CF phenotype, 70% of patients have mutations in exon 10 of the cftr gene. Most CF patients are diagnosed readily by clinical presentation, confirmed by sodium sweat-conductance tests, whereas heterozygous carriers are asymptomatic. The clinical question becomes: Is a family member of a patient a carrier? The approach taken is hierarchical: Develop an inexpensive, rapid, highly sensitive/moderately specific screening procedure that will identify likely carriers, and evaluate potential carrier samples by definitive mutational analysis with a more costly, exacting technique such as direct sequencing.
Several new methodologies are available for mutational screening. The
characteristics of the most promising technologies are rapid TAT,
compatibility with future automation, and sensitive detection of a wide
variety of mutations, including single-base mismatches, insertions, and
deletions. A homogeneous mismatch detection technique has been
developed by combining the mutational mismatch-stabilization of
cruciform DNA structures (Holliday Junctions) with photoexcited
chemiluminescence. Branch migration, the process by which homologous
DNA duplexes exchange strands, is inhibited by mismatches in the
presence of high concentrations of Mg2+
(60). Consequently, alleles can be detected by the
inhibition of branch migration. Following exon-specific amplification,
cohybridization of the allelic amplicons will result in formation of
cruciform DNA structures (Fig. 3
). Separation of the cruciform structures into two DNA duplexes
is inhibited if a mismatch is present in both strands of the cruciform
structure. Detection of the mutation in the heterozygous carrier is
accomplished through using modified primers with biotin and digoxigenin
incorporated at the 5'-ends. This permits attaching photosensitizer
beads through streptavidin bridges and chemiluminogenic beads through
anti-digoxigenin to the cruciform structures, allowing homogeneous
detection through photoexcitation (16). Laser excitation
generates singlet oxygen. If a chemiluminogenic bead is in close
proximity because of the presence of a mutation, then a
chemiluminescence signal is generated (Fig. 4
). If branch migration has occurred, only a background signal is
obtained because of the short mean diffusional path of singlet oxygen.
|
|
An initial trial of this technology was performed by assaying a panel of 50 encoded DNA samples that included samples heterozygous for exon 10; homozygous for normal exon 10 (no mutations); homozygous for exon 10 mutation; and homozygous for exon 11 mutation. All heterozygous carriers were correctly identified with this technology. Homozygous mutations should not be detected directly by using the carrier identification format because of the lack of mismatch potential, but one homozygous exon 10 mutation was detected as a false positive.
The excellent sensitivity and specificity obtained in the pilot study indicates that this technology merits investment of additional effort and resource. The mismatch-stabilized inhibition of the branch migration assay has potential for automation by laboratory robots for performance in 96-well microplates. This would decrease the costs associated with genetic screening. Potential improvements in outcomes would be improved TAT as well as better resource utilization of the more costly DNA sequencing equipment. These potentials will be more readily apparent once DNA extractions are automated.
Obstacles to Assay Conversion, Development, and Implementation
Several obstacles to assay conversion/implementation activities may be raised, in part from a general resistance to change. Objections to changing analytes or assay methodologies are frequently raised in part because such changes may negatively affect routine laboratory testing in the short-term. The conversion of existing assays to modern technologies is often viewed as consuming valuable development resources in an effort to reinvent the wheel: Why redevelop an assay that already works well? Furthermore, personal agendas and biases may predispose an assay conversion activity to failure. The most effective argument to overcoming these objections is compelling data that validate the cost-effectiveness of the activity proposed. This should be performed at least for the first several assay change projects to help develop easier acceptance for additional activities.
Other arguments may be raised in objection to translation research activities aimed at implementing new tests. The argument that new markers and new technologies are unproven may be countered with the generation of validation data. However, the major obstacle limiting new test implementation is the concern by manufacturers of legal liability. Few clinical DNA amplification-based tests have been marketed today because of this concern. One approach to avoid this obstacle is to utilize new technologies for in-house development activities, in which the requirement for assay performance validation, and liabilities, become the responsibility of the medical institution performing the assay.
In summary, several assay implementation activities may impact healthcare in a cost-effective manner. The easiest manner to perform these activities successfully is to:
1. Define the clinical question that is driving the specific assay conversion/development/implementation process.
2. Tightly constrain the clinical context in which the impact of assay changes will be evaluated.
3. Evaluate potential cost-effectiveness resulting from the assay changes.
4. Obtain supporting analytical performance data.
5. Perform outcomes and cost accounting of the new vs old assay after implementation of the new assay.
Following these steps will maintain the focus of laboratory testing on improving the delivery of healthcare for the patient.
Acknowledgments
We gratefully acknowledge the contributions to this work from several collaborators: Pai C. Kao and George G. Klee, Mayo Clinic, for assistance with steroid assay conversion; David M. Wilson, Timothy S. Larson, Robert Liedke, and Bernice Stevens, Renal Function Laboratory, Mayo Clinic, for their help with developing urine prescreening assays; Steven N. Thibodeau, Molecular Genetics Laboratory, for provision of samples for mutation screening; Henry H. Homburger for helpful discussion; and Edwin Ullman, Alla Lishansky, and Nurith Kurn, Behring Diagnostics, San Jose, CA, for their generous provision of reagents and instrumentation for mutation screening. Telomerase assay development activities were supported in part by an R21 grant from the National Cancer Institute.
Footnotes
1 Nonstandard abbreviations: TAT, turnaround time; CPWG, Clinical Pathology Work Group; TCC, transitional cell carcinoma; and CF, cystic fibrosis. 
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
