What are the explanations for rising incidence and falling mortality in prostate cancer An allIrelan

1
What are the explanations for rising incidence
and falling mortality in prostate cancer? An
all-Ireland study

• Frances Drummond
• National Cancer Registry, Ireland
• On behalf of the All-Ireland Prostate Cancer
  consortium

2
Background

• Prostate cancer is the most commonly diagnosed
  cancer in men in the US and Europe
• Incidence has increased significantly in the last
  20 years
• A major cause of cancer death - third in Ireland
• Mortality is decreasing in most high income
  countries

3
Prostate cancer incidence and mortality
Estimated incidence, Europe 2006
Ferlay J et al. Ann Oncol 2007 18(3) 581-92
4
Prostate cancer incidence and mortality
Estimated incidence, Europe 2006
Estimated mortality, Europe 2006
Ferlay J et al. Ann Oncol 2007 18(3) 581-92
5
Prostate Specific Antigen (PSA)

• Prostate Specific Antigen (PSA), a biomarker
• PSA testing has contributed to the increase in
  incidence(1)
• Whether PSA testing decreases mortality is
  heavily debated
• Randomized trials did not consistently show
  mortality decreases associated with prostate
  specific antigen (PSA) testing
• ERSPC reported a 20 decrease in prostate cancer
  mortality in the PSA screened group (2)
• PLCO observed no difference between screened and
  unscreened groups (3)

1. McDavid 2004, Public Health Rep
119(2)174-186 2. Schroder F 2009, N Engl J
Med3601320-8 3. Andriole G 2009, N Engl J
Med3601310-9
6
Comparing trends between countries have an
important role to play in explaining incidence
and mortality trends
7
Comparing trends between countries have an
important role to play in explaining incidence
and mortality trends
Contrast in health services Northern
Ireland (NI) publicly funded health care (NHS),
free at the point of delivery GPs
gate-keepers to tertiary care Republic
of Ireland (RoI) mixed public-private health
system 50 have private health
insurance 60 to visit GP/outpatient clinic in
public system
8
Comparing trends between countries have an
important role to play in explaining incidence
and mortality trends
Contrast in health services Northern
Ireland (NI) publicly funded health care (NHS),
free at the point of delivery GPs
gate-keepers to tertiary care Republic
of Ireland (RoI) mixed public-private health
system 50 have private health
insurance 60 to visit GP/outpatient clinic in
public system

• Contrast in PSA testing practices/policies
• Northern Ireland (NI) PSA screening not
  recommended in primary
• care (NHS Cancer Screening, PCRMP)
• but, PSA testing going on (1)
• Republic of Ireland (RoI) No guidelines NCF
  recommend against pop-based screening (2006)
• PSA testing widespread in primary care (2)
• Major variations in practice (3)

1. Gavin A, 2004 BJU Int, 3. Drummond FJ 2008
Ir J Med Sci. 2008 Dec177(4)317-23. 2.
Drummond FJ 2009 BMC Fam Pract 12103
9
Aim

• To investigate prostate cancer incidence and
  mortality trends and factors influencing these in
  the Republic of Ireland (RoI) and Northern
  Ireland (NI)

10
Subjects and Methods
11

• Prostate cancer incidence
• Data on invasive prostate cancers (ICD-O2C61)
  were obtained from the National Cancer Registry
  Ireland (NCRI) (1994-2005) and the Northern
  Ireland Cancer Registry (NICR) (1993-2005)

12

• Prostate cancer incidence
• Data on invasive prostate cancers (ICD-O2C61)
  were obtained from the National Cancer Registry
  Ireland (NCRI) (1994-2005) and the Northern
  Ireland Cancer Registry (NICR) (1993-2005)
• PSA
• NI
• Information on PSA tests performed in NI since
  1994 is routinely collected by the NICR
• RoI
• Data on all tests (1994-2005) were sought from
  the 36 laboratories which analyse PSA.
• Used information from 2006 lab survey to estimate
  missing data (1).
• We estimate that we collected information on 58
  of the total tests, 94-05.
• Data on PSA tests were linked to the NCRI
  database by name, DOB and address (where
  available). A similar linkage was performed in NI
  (2).
• PSA tests performed after the date of diagnosis
  with cancer were excluded.

1. Drummond FJ 2008 Ir J Med Sci. 2008
Dec177(4)317-23. 2. Connolly D, 2008 Cancer
Epidemiol Biomarkers Prev17(2)271-8.
13

• Prostate biopsy data
• RoI
• Numbers of prostatic biopsies (ICD9 60.11-60.15),
  by year and age-group, were obtained from the
  Hospital In-Patient Enquiry System (HIPE) -
  records all discharges from all public hospitals
  (1994-2005.
• Data on claims for all biopsies performed in
  private hospitals, by year and age group, VHI
  Healthcare and BUPA (1996 2005)
• NI
• Information on needle biopsies was obtained from
  the Directorate of Information Services which
  record procedure codes from all hospital
  discharges in NI (1999-2004).
• Total counts were provided (these data could not
  be broken down by age).

14

• Prostate biopsy data
• RoI
• Numbers of prostatic biopsies (ICD9 60.11-60.15),
  by year and age-group, were obtained from the
  Hospital In-Patient Enquiry System (HIPE) -
  records all discharges from all public hospitals
  (1994-2005.
• Data on claims for all biopsies performed in
  private hospitals, by year and age group, VHI
  Healthcare and BUPA (1996 2005)
• NI
• Information on needle biopsies was obtained from
  the Directorate of Information Services which
  record procedure codes from all hospital
  discharges in NI (1999-2004).
• Total counts were provided (these data could not
  be broken down by age).
• Prostate cancer mortality data
• Mortality data were extracted from World Health
  Organization mortality database for the period
  1979-2006

15
Statistical Analysis

• Age-standardised rates (ASR) in men aged 50
• incidence
• mortality
• PSA testing

16
Statistical Analysis

• Age-standardised rates (ASR) in men aged 50
• incidence
• mortality
• PSA testing
• Biopsy rates
• crude rates for NI
• rates for the RoI standardised to NI population

17
Statistical Analysis

• Age-standardised rates (ASR) in men aged 50
• incidence
• mortality
• PSA testing
• Biopsy rates
• crude rates for NI
• rates for RoI standardised to NI population
• Annual Percentage Change (APC)
• joinpoint regression log-linear model
• trends for all ages (50) and by age-group
  (50-74, 75)

18
RESULTS
19
Prostate cancer incidence rates, 1994-2005

• 19,844 prostate cancers in the RoI
• 7,388 in prostate cancers in NI.

20
Prostate cancer incidence rates, 1994-2005
All ages (50 years)

• 19,844 prostate cancers in the RoI
• 7,388 in prostate cancers in NI.

APC and joinpoint segment
21
Prostate cancer incidence rates, 1994-2005
All ages (50 years)

• 19,844 prostate cancers in the RoI
• 7,388 in prostate cancers in NI.
• Age-standardised incidence rate (1994-2005) was
  on average 41 higher in the RoI (346 per 100,000
  men aged gt50 years) than in NI (245 per 100,000
  men aged gt50).

APC and joinpoint segment
22
Age-standardised incidence rates by age, 1994-2005
Ages 50-74 years
APC and joinpoint segment
p-valuelt0.05
23
Age-standardised incidence rates by age, 1994-2005
Ages 50-74 years
Ages gt75 years
APC and joinpoint segment
p-valuelt0.05
24
Age at diagnosis

• The median age at cancer diagnosis was
  significantly lower in the RoI (71 years) compare
  to NI (73 years) (plt0.01).

25
Age at diagnosis

• The median age at cancer diagnosis was
  significantly lower in the RoI (71 years) compare
  to NI (73 years) (plt0.01).
• Median age decreased significantly over time
• RoI 1994, 74 years 2005, 68 years
  (p-trendlt0.01)
• NI 1994, 74 years 2005, 70 years
  (p-trendlt0.01)).

26
Age at diagnosis

• The median age at cancer diagnosis was
  significantly lower in the RoI (71 years) compare
  to NI (73 years) (plt0.01).
• Median age decreased significantly over time
• RoI 1994, 74 years 2005, 68 years
  (p-trendlt0.01)
• NI 1994, 74 years 2005, 70 years
  (p-trendlt0.01)).

Age at which Asymptomatic men are PSA tested by
RoI and NI GPs
27
Grade
RoI
NI
28
Age-standardised rates PSA testing
All ages (50 years)
APC and joinpoint segment
plt0.05
p-valuelt0.05 Excludes tests performed in those
with prostate cancer

• 412 tests per 1,000 men 50 years in RoI, 2004
• 206 per 1,000 50 years in 2004 in NI, 2004.

29
Age-standardised rates PSA testing by age
Ages 50-74, gt75 years
APC and joinpoint segment
p-valuelt0.05 Excludes tests performed in those
with prostate cancer
30
Median PSA level in tests within 6 months prior
to cancer diagnosis

• Data from 7,208 (36 of the prostate cancer cases
  1994-2005) in the RoI and 4,592 (66 of the
  prostate cancer cases 1994-2005)
• Plt0.001

31
Prostate biopsy rates
All men gt50 years
APC and joinpoint segment
p-valuelt0.05
Crude rates for NI rates for RoI standardised to
NI population
32
Prostate biopsy rates
All men gt50 years
Ages 50-74, gt75 years, RoI
APC and joinpoint segment
APC and joinpoint segment
p-valuelt0.05
Crude rates for NI rates for RoI standardised to
NI population
33
Age-standardised prostate cancer mortality rates
34
Prostate cancer treatment
prostate cancer patients receiving radical
prostatectomy and hormone therapy in 1996
Gavin A, 2005 Drummond F, 2007
35
Prostate cancer treatment
Treatment trends in RoI
prostate cancer patients receiving radical
prostatectomy and hormone therapy in 1996
Gavin A, 2005 Drummond F, 2007
36
Conclusions 1

• Prostate cancer Incidence was consistently higher
  in the RoI than NI

37
Conclusions 1

• Prostate cancer Incidence was consistently higher
  in the RoI than NI
• The difference in incidence mainly due to the
  relative intensity of cancer investigation via
  prostatic biopsy, rather than PSA testing

38
Conclusions 1

• Prostate cancer Incidence was consistently higher
  in the RoI than NI
• The difference in incidence mainly due to the
  relative intensity of cancer investigation via
  prostatic biopsy, rather than PSA testing
• 1994-2000, PSA rates similar, but incidence
  higher in the RoI
• PSA testing was increasingly used in NI before
  1999, but no rise in incidence until 1999
• very low biopsy rate in NI in 1999 incidence
  rose as biopsy rate rose
• higher biopsy rate in the RoI and higher
  incidence
• in RoI, age-specific trends in incidence mirror
  those for biopsies
• evidence that threshold for biopsy lower in RoI
• lower median PSA level in those with cancer
• studies among primary care physicians (Connolly,
  2007 MD thesis Drummond et al. BMC Fam Pract
  2009) and urologists are consistent with this
• consistent with differences in healthcare system

39
Conclusions 1

• Prostate cancer Incidence was consistently higher
  in the RoI than NI
• The difference in incidence mainly due to the
  relative intensity of cancer investigation via
  prostatic biopsy, rather than PSA testing
• 1994-2000, PSA rates similar, but incidence
  higher in the RoI
• PSA testing was increasingly used in NI before
  1999, but no rise in incidence until 1999
• very low biopsy rate in NI in 1999 incidence
  rose as biopsy rate rose
• higher biopsy rate in the RoI and higher
  incidence
• in RoI, age-specific trends in incidence mirror
  those for biopsies
• evidence that threshold for biopsy lower in RoI
• lower median PSA level in those with cancer
• studies among primary care physicians (Connolly,
  2007 MD thesis Drummond et al. BMC Fam Pract
  2009) and urologists are consistent with this
• consistent with differences in healthcare system
• Information on PSA testing alone not sufficient
  to assess impact of screening activity on
  incidence need biopsy information

40
Conclusions 2

• PSA testing is not the reason for decreasing
  mortality rates in Ireland
• mortality rates were falling from 1995 - before
  PSA testing became widespread
• change in mortality essentially equivalent in the
  two countries although PSA testing and biopsy
  rates much higher in RoI than NI

41
Conclusions 2

• PSA testing is not the reason for decreasing
  mortality rates in Ireland
• mortality rates were falling from 1995 - before
  PSA testing became widespread
• change in mortality essentially equivalent in the
  two countries although PSA testing and biopsy
  rates much higher in RoI than NI
• Other possible explanations
• changes in treatment (e.g. wide-spread use of
  hormonal therapy)
• attribution bias

42
All-Ireland Prostate Cancer Research Group
National Cancer Registry Ireland Anne-Elie
Carsin statistician Harry Comber director France
s Drummond study co-ordinator Linda
Sharp epidemiologist
Northern Ireland Cancer Registry/ Queens
University Belfast Amanda Black research
fellow David Connolly urologist Anna
Gavin registry director Liam Murray
epidemiologist
Collaborators Erasmus University Medical Centre
Pim van Leeuwen International Agency for
Research on Cancer Philippe Autier Mathieu
Boniol Lars Egevad
43
Acknowledgments

• laboratories who provided data on PSA tests
• HIPE, VHI and BUPA who provided biopsy data
• GPs, Urologists and radiologists for completing
  questionnaires
• those at the NICR and NCRI for collecting and
  processing the data and reviewing death
  certificates
• Funded by
• NI Research Development,
• Health Research Board,
• National Cancer Screening Service,
• Irish College of General Practitioners

44
Comparative study between the Republic of Ireland
(RoI) and Northern Ireland (NI)
A-E Carsin1, FJ Drummond1, A Black2, PJ van
Leeuwen3, L Sharp1, LJ Murray4, D Connolly5, L
Egevad6, M Boniol6, P Autier6, H Comber1, A
Gavin7

• National Cancer Registry Ireland
• Cancer Prevention, National Cancer Institute
• Erasmus University Medical Centre, Rotterdam,
  Netherlands
• Cancer Epidemiology and Prevention Research
  Group, Queen's University Belfast
• Department of Urology, Belfast City Hospital,
  Belfast, Northern Ireland
• International Agency for Cancer Research, Lyon
• Northern-Ireland Cancer Registry, Belfast

45

• Thank you !!!!
• f.drummond_at_ncri.ie