STATINS for CANCER PREVENTION

1
STATINS forCANCER PREVENTION???

• Lindsay Sher
• GIM Journal Club
• January 22, 2008

2
The Association Between Statins and Cancer
Incidence in a Veterans Population

• Journal of the National Cancer Institute
• January 16, 2008

3
Background

• Cancer is the second leading cause of death in
  the US (first is heart disease)
• Estimated number of deaths attributable to
  cancer 559,650
• Estimated number of new cancer diagnoses in 2007
  1,444,920
• Lifetime risk of receiving a diagnosis of cancer
  almost 1 in 2 for Men, greater than 1 in 3 for
  women
• Statistics from the American Cancer Societys
  Cancer Facts and Figures 2007

4
Background

• Statins, or HMG-CoA reductase inhibitors, are
  drugs used to lower cholesterol through their
  inhibition of the rate-limiting step in the
  mevalonate pathway of cholesterol synthesis
• The first statin, Mevastatin, was isolated from a
  mold, Penicillium citrinium, in 1976
• Lovastatin (Mevacor) was the first statin to
  appear on the market
• Developed by Merck
• Approved in 1987 by the FDA
• Isolated from a species of Aspergillus
• Forbes Magazine listed Lipitor as the most
  profitable prescription drug in 2005, making 12.9
  billion dollars in just one year (Zocor comes in
  5th, making 5.3 billion dollars for Merck )

5
Why study Statins and Cancer?

• Several laboratory studies have shown that
  statins have the ability to inhibit molecular
  pathways utilized in cancer cells
• Trials of statin therapy to prevent cancer have
  produced mixed results
• Confounding variables such as age and
  healthy-user effect are reasons sited against the
  validity of prior studies

6
Method

• Retrospective Cohort Study
• Study population Veterans 18 yrs or older
  enrolled in the VA health-care system between
  January 1, 1997, and December 31, 2005

7
Who made the cut?

• Cohorts were assembled from patients who filled
  at least 2 Rxs for any antihypertensive or statin
  within one year, continued to fill Rx for med of
  interest at least once per year, and who were
  seen in a VA OP clinic at least once per year.
• Cohort 1 Patients who never filled a Rx for a
  cholesterol-lowering medication, but did fill a
  Rx for an antihypertensive med (meant to correct
  for compliance).
• Cohort 2 Patients who filled Rxs for a statin
  (defined list). These Pts may or may not have
  filled Rxs for antihypertensives.

8
Definitions

• Entry Date first recorded date that a Rx was
  filled for a med of interest
• Observation period began 2 yrs after entry date
  and until
• First occurrence of a cancer diagnosis in Pts
  record
• 1 yr after the last date that a Rx was filled for
  a med of interest
• Pts death
• End of analysis period

9
Whos Out?(ie Exclusions)

• Patients diagnosed with cancer on or within 2 yrs
  of their entry date onto the study
• Patients who discontinued their statin within 2
  yrs of entry date (based on assumption that
  long-term exposure to statins necessary)

10
Outcomes

• Primary outcome defined as cancer incidence,
  excluding non-melanomatous skin cancer
• Cancer incidence was defined by a set of ICD-9
  codes for cancer type within the VA EMR during
  both inpatient and outpatient encounters
• The diagnosis of cancer was validated by a
  physician blinded to medication use for 300
  randomly selected charts (confirmation rates were
  equal amongst statin users and non-users)

11
Statistical Analysis

• Characteristics of each group were compared using
  t tests (continuous variables) and chi-squared
  tests (categorical variables)
• Variables age, weght, female, smoking history,
  aspirin use, cholesterol (Total, LDL, HDL),
  history of colorectal endoscopy, disease states
  (thyroid, DM, HTN, renal failure, mental
  illness, substance abuse, CV, Lung, GI,
  prostate)
• Constructed age- and multivariable-adjusted Cox
  proportion hazards models for total cancer
  incidence and incidence of 5 most frequently
  diagnosed cancers
• Investigators also calcuated a propensity score
  for being prescribed a statin and constructed
  models that used this score.

12
Statistical Analysis

• To study the relationship between statin dose and
  cancer incidence, tertiles of equivalent
  simvastatin doses were defined
• Equivalent doses of other statins were calculated
• Tests of trend were calculated across the
  tertiles
• Compounding variables were controlled for in the
  same manner

13
Results

• Total Referent group 25,594
• Total Statin user group 37,248

14
How about dose?

• Total Referent group 25,594
• Total Statin user group 37,248

15
Discussion

• Use of statins correlated with an decrease overal
  risk of cancer, especially lung and colorectal
  cancer
• A dose-response relationship may exist between
  statins and cancer incidence

16
Strengths of the Study

• Large study size (ability to control for
  confounders)
• Data available on the VA EMR regarding patient
  characteristics, prescription information, and
  lab work
• Healthy-user effect controlled for compliance
  and cancer screening

17
Weaknesses of the Study

• Very small proportion of women and minorities
• No information on diet, exercise, quantity of
  cigarette or EtOH use
• Reliance on ICD-9 coding
• Not all patients first-time statin users
• Inability to verify all new cancer diagnoses
• Healthy-user effect
• Statin users may be more likely to receive
  healthy lifestyle counseling
• Other medications confounding results
  (antihypertensives, etc)

18
Does stand for Statin?
19
The FUTURE!

• The results of this study are exciting, but not
  readily generalizable
• Further studies, including radomized trials, must
  be conducted in a more controlled manner to
  further delineate the possible preventative
  effect of statins on cancer incidence

20
For Our Practice

• As the numbers point out, we use statins A LOT!
• There is no indication at this time to prescribe
  statins explicitly for cancer prevention
• We should continue to prescribe statins for
  cholesterol lowering and keep our fingers crossed
  that future studies proved Statins to truly be
  wonder drugs!

21
(No Transcript)