Transcript Dollars and Sense: Cost-Effectiveness in Clinical Medicine - HIV-DRI

HIV Resistance Testing:
Overview of Indications and Cost
Issues
Paul E. Sax, MD
Division of Infectious Diseases
Brigham and Women’s Hospital
Harvard Medical School
Outline
• Review of available resistance tests
• What tests to order when
• Review of cost analyses
• How cost issues relate to resistance testing
– USA and other developed countries
– Resource-limited settings
When to Use Resistance Testing
IAS-USA[1]
DHHS[2]
European[3]
Primary/acute
Recommend
Recommend
Recommend
Postexposure
prophylaxis
—
—
Recommend
Consider*
Recommend
Strongly consider*
Failure
Recommend
Recommend
Recommend
Pregnancy
Recommend
—
Recommend*
—
—
Recommend†
Chronic, Rx naive
Pediatric
*Especially if exposure to someone receiving antiretroviral drugs is likely or if prevalence of
drug resistance in untreated patients ≥ 5% (European: ≥ 10%).
1. Hirsch et al. Clin Infect Dis. 2003;37:113-28.
2. Available at: http://www.aidsinfo.nih.gov. Accessed May 4, 2006.
3. Vandamme et al. Antivir Ther. 2004;9:829-48.
Advantages and Disadvantages
of Genotype Testing
Advantages
Disadvantages
•
Rapid turnaround (1-2 wks)
• Indirect measure of resistance
•
Less expensive than phenotyping
•
Detection of mutations may
precede phenotypic resistance
• Relevance of some mutations
unclear
•
•
• Unable to detect minority
Widely available
variants (< 20% to 25% of viral
sample)
More sensitive than phenotype for
detecting mixtures of resistant
and wild-type virus
• Complex mutational patterns
may be difficult to interpret
Advantages and Disadvantages
of Phenotype Testing
Advantages
Disadvantages
• Provides direct and quantitative
measure of resistance
•
• Methodology can be applied to any
•
antiretroviral agent, including new
drugs, for which genotypic correlates of
•
resistance are unclear
• Uses 2 clinical cutoffs (CCO) derived
from clinical cohorts to define spectrum
of resistance
• Indicates which drugs have partial
activity
• Can assess interactions among
mutations
• Accurate with non-B HIV subtypes
Susceptibility cutoffs not standardized
between assays
Clinical cutoffs not defined for some
agents
May be unable to detect minority
variants for some mutations (< 20% to
25% of viral sample)
•
Complex technology with longer
turnaround
(~ 3 wks)
•
More expensive than genotyping
Advantages and Disadvantages
of “Virtual” Phenotype Testing
Advantages
Disadvantages
•
Similar advantages to genotype
•
(turnaround time, cost, sensitivity)
•
Defines resistance based on
database of in vivo responses in
treated patients
•
•
Uses 2 clinical cutoffs (CCO) to
define spectrum of resistance
Indicates which drugs have partial
activity
Is an estimated phenotype based
on the patient’s genotype, not an
actual measured phenotype
•
Reliability will depend on the
accuracy of the genotype
•
Available only from 1 vendor
•
More expensive than genotype
alone
•
Methodology of linking genotype
to phenotypic database not
intuitively obvious—uses a
proprietary “virtual phenotype
linear regression model engine”
Drug Resistance Testing in
Clinical Practice: Summary
• Indications for use of resistance testing have greatly
expanded
• Genotype preferred
– Treatment naive: acute or chronic infection
– Early virologic failure
– Patient no longer on therapy
• Phenotype, virtual phenotype, or combined
phenotype/genotype preferred
– High-level resistance to NRTIs or PIs on genotype
– Multiple regimen failure with limited treatment options
Cost Issues in Resistance Testing
Case Presentation
• 38 year-old MSM with community-acquired MRSA
• Treated with doxycycline, local care; recovered
uneventfully
• HIV test recommended; no prior test
• HIV+; initial CD4 180, HIV RNA 77,000
What Should be Standard of Care?
• Should a resistance test be done?
• If not, why not?
• If so, which one?
• Why do we care?
Who Decides?
• Doctor and/or patient?
• Medicaid or ADAP or VA?
• Kaiser or BC/BS or Harvard University Health Plan?
• USPHS or IAS or WHO guidelines?
• Resistance testing vendors?
• “Society”?
Antiretroviral & Prophylaxis Costs:
United States
Zidovudine
$3,300 TMP-SMX
$
105
Tenofovir
$5,500 Dapsone
$
60
Lamivudine
$4,000 Atovaquone
$ 9,560
Indinavir
$7,000 Azithromycin
$ 1,450
Nelfinavir
$9,125 Fluconazole
$
Efavirenz
$5,900 Ganciclovir
$15,600
Lopinavir/r
$8,500 Enfuvirtide
$20,000
*Wholesale cost per person for one year
510
Resources are Limited
– Even Here (US)
• ADAPs with waiting lists
– Alabama: 6 on waiting list
– Alaska: 10 on waiting list
– Indiana: 33 on waiting list
– Montana: 20 on waiting list
– South Carolina: 209 on
waiting list
– West Virginia: 24 on
waiting list
Source: www.tiicann.org, Oct 6 2006.
• ADAPs with other costcontainment strategies
– Mississippi: Medical
eligibility restrictions
– Oklahoma: Annual per
capita expenditure limit
– South Carolina: Reduced
formulary
• Six states expecting to start
cost-containment in 2007
Quiz Question
• Has HIV become more or less costly since the
introduction of potent therapy in 1996?
Cost Analyses: HIV Care
is Becoming More Expensive
• What does it cost/year to care for an HIV patient in the
USA?
–
–
–
–
HCSUS,1992: $14,700
HCSUS, 1998: $20,000
Johns Hopkins, 1999: $15,660
CEPAC Collaboration, 2004:
$26,800
• What is the lifetime cost?
– 1992: $100,000 (survival 6.8 years)
– 2004: $649,000 (survival 24.2 years)
Bozzette et al, NEJM 1998;339:1897-904.
Gebo et al, AIDS 1999;13:963-9.
Schackman et al. Med Care. 2006;44:990-7.
Cost-Benefit Analysis
• One outcome measure
• Costs and benefits both valued in dollars
• Difficult to assign a value to clinical outcomes (PCP,
stroke, etc.)
• Often viewed as morally distasteful by MDs …
“I’ve received your credit report,
and you seem to be a person worth saving.”
Cost-effectiveness Analysis
• Two different outcome measures:
– Cost in dollars
– Effectiveness: Years of life saved (YLS) or qualityadjusted life years (QALY)
• Cost-effectiveness ratio:
– Resource use ($)/Health benefit (QALY)
Incremental Cost-Effectiveness Ratio
Net increase in health care cost
Net gain in health effect
• measure of “value” for resources
• a basis for comparing options
• cost-saving cost-effective
The “$50,000” Threshold:
Often Cited, Often Ignored
$/YLS
Propanolol, mild HTN
14,000
TPA vs streptokinase
33,000
Rx hypercholesterolemia
47,000
Dialysis, ESRD
51,000
Screening mammography:
Annual 50-69
Annual 40-49
YLS = years of life saved
57,500
168,400
Antiretroviral Therapy is
Very Cost Effective
C-E Ratio
Strategy
Costs ($)
QALM
($/QALY)
No ART
59,790
47.52
---
AZT/3TC/EFV
94,290
79.56
13,000
No ART
54,150
35.04
---
AZT/3TC/IDV
80,460
53.16
17,000
Dupont 006 (CD4 350)
Johns Hopkins (CD4 217)
Freedberg et al. NEJM 2001;344:824-31.
What Does HIV Lab Testing Cost?
Test
HIV RNA
CD4
Costs in $
119
83
Genotype
“Virtual” phenotype
Phenotype
Phenotype + genotype
(Tropism assay
355-676
550
700-1148
800-1690
710)
Sources: BWH hospital lab, private vendors
Resistance Testing is
Cost-effective after Treatment Failure
Separate study: 22,510 euros/life-year gained.
Weinstein et al. Ann Int Med. 2001;134:440-50.
Corzillius et al. Antivir Ther. 2004;9:27-36.
Why is this Expensive Test
So Cost-Effective?
“The cost of the resistance test itself is, in effect,
‘amortized’ over several months of added survival.
Compared with the cost of HAART and other HIV-related
care in the added months of survival, the cost of the
resistance test is modest. Even if each resistance test
cost $1000, resistance testing would be relatively good
value for money.”
Weinstein et al. Ann Int Med. 2001;134:440-50
Should Resistance Testing be
Done in Antiretroviral-naïve Patients?
• Hypothetical cohort of HIV+ patients, newly diagnosed
• Outcomes compared with and without initial genotype
resistance testing
• Sensitivity analyses:
–
–
–
–
Rate of resistance
Cost of test
Efficacy of test
Proportion starting NNRTI vs PI-based therapy
Sax et al. Clin Infect Dis. 2005; 41:1316-23.
Resistance Testing at Diagnosis
Improves Outcome at Reasonable Cost
Sax et al. Clin Infect Dis. 2005; 41:1316-23.
Resistance Testing in Naives:
Conclusions
• CE ratio remained < $50,000/QALY until:
– Prevalence of resistance fell to 1% or
– Cost of test increased to $3000 or
– Efficacy of test was 14% of baseline
• Conclusions
– Single test that improves outcome over lifetime of patient is
highly cost-effective ($23,900/QALY)
– Substantial benefit for those with resistance (esp NNRTI)
overrides no benefit for those without
– Current rates of NNRTI resistance are higher than when
analysis done – baseline genotype testing likely to be more
cost-effective now
Sax et al. Clin Infect Dis. 2005; 41:1316-23.
Genotype versus
Phenotype + Genotype
• Industry-sponsored CE analysis
• Hypothetical cohort of treatment-experienced patients,
CD4 50-200
• Susceptibility score estimated from “GUESS III” study
– higher for pheno + geno (PTGT) than geno (GT)
• Outcomes projected using state transition Markov
model
Coakley et al. ICAAC 2005, Abstract #H1054
Genotype versus
Phenotype + Genotype: Results
•
Results
– Costs of GT strategy slightly lower than PTGT
– Survival longer with PTGT
– Incremental CE ratio = $28,812/QALY
•
Limitations:
– benefits of PTGT over GT likely to be much smaller in those with
limited resistance
– Industry-sponsored
Coakley et al. ICAAC 2005, Abstract #H1054
Resistance Issues in
Resource-limited Settings
HIV Drug Resistance is
Becoming More Important in
Resource-Limited Settings
• Treatment started with more
advanced disease
• Fewer agents available
• Some older treatments have
long-term toxicity that reduces
adherence
• Supply chain for medications
inconsistent
• Viral load usually not used for
monitoring  prolonged
treatment with virologic failure
• Resistance testing not
available
Hospital laboratory, Rwanda
HIV RNA Monitoring: Desired
but Unavailable
“Future directions to improve access to treatment in resource -limited settings
Considerable progress has been made in the past four years towards making ART scale-up in the
developing world a reality. However, much remains to be done....
There are immediate needs in the area of diagnostics. Making affordable and
accurate CD4 cell counting widely available is a high priority. Simultaneously, the
field needs to move towards the development and implementation of affordable
viral load testing. CD4 and plasma HIV-1 RNA testing are not luxuries. They are
important tools supporting the delivery of optimal care and, in the setting of the
public health approach, are invaluable measures of programme monitoring and
performance.”
Slide courtesy William Rodriguez, M.D.
How to Select MDR HIV:
Lessons from the Past
Highly adherent, aggressively treated
patients with non-suppressive
regimens led to selection of
multidrug-resistant HIV
Sequential NRTI
monotherapy and
dual-NRTI therapy
No
ART
ZDV
monotherapy
Early
80s
Late
80s
“Hit hard,
hit early”
Earlier initiation
of therapy with
better rx
“Sequential
monotherapy”
with PIs/NNRTIs
Early
90s
Mid
90s
Deferral
of therapy
Late
90s
Early
00s
Late
00s
Where is Resistance
Testing Being Performed in
Resource-Limited Settings?
• Brazil
– Available at all sites after panel reviews indication
• Botswana
– Limited access; recommended for “second-line” treatment
failure
• All other sites surveyed
– Highly-limited access (e.g., private payors only) or no
access at all
Schechter M, Shapiro R, Rodriguez W, Marconi V,
Haubrich R, Cahn P, Antunes F, Libman H, Eisenberg
M, Cosimi L, Mayer K. Personal communications.
Botswana:
Guidelines for Resistance Testing
4.2 Role of Resistance Testing
Since resistance testing is costly and background level
resistance is probably still negligible in our setting,
resistance testing at first failure is not justifiable.
However, genotypic resistance testing should be
definitely done to guide choice of drugs for third line
regimen as it may then be possible to recycle some of
the drugs used in previous regimens. Pregnant women
who have taken monotherapy for PMTCT purposes
may be considered for resistance testing.
Botswana Antiretroviral Treatment Guidelines, 2005 Revision.
•
WHO plan for setting up labs
and sentinel surveillance sites
•
Goal is to describe both
transmitted and on-treatment
prevalence of resistance
•
No recommendations for the
use of resistance testing in
clinical practice
http://www.who.int/hiv/drugresistance/en/
WHO Guidelines: Only Mention
of Clinical Use of Resistance Testing
“For highly treatment experienced patients, individual
management is necessarily tailored to the availability of
alternative ARVs, for which there is very limited provision
in the public sector in resource-limited settings, and to
additional laboratory investigations, such as individual
drug resistance testing.”
Antiretroviral Therapy For HIV Infection In Adults
And Adolescents, WHO, 2006 Revision
Resistance Testing and Cost:
Conclusions and Future Directions
• Resistance testing: costly but cost-effective
• Major challenge: how can we apply our understanding
of resistance prevention and management to resourcelimited settings?
• Additional questions
– What is the clinical utility and cost-effectiveness of
looking for minority variants?
– How will existing testing be adapted to new drug classes?