Transcript Milena Sant

Workshop for a
EUROPEAN HIGH RESOLUTION STUDY
6th November 2012
Malpensa airport , Milano
Study proposal frame
Milena Sant, MD
EPAAC WP9 leader
Descriptive Studies and Health Planning Unit
Istituto Nazionale Tumori, Milano, Italy
EPAAC WP9 Objectives
1. To map the main sources of cancer data in Europe and to
identify the priority topics to be supported by the
Partnership
2. To unify under a common website cancer burden indicators
(incidence, mortality, survival, patterns of care and
prevalence) provided by existing European activities
3. To individuate indicators of cancer costs and socioeconomic
status to be used in population based studies
4. To develop a standardised approach for the collection of
data on survivorship using population based cancer
registries
5. To develop an inventory of statistical methods to analyse
population based cancer data
AIMS of the High-Resolution studies
To help understanding the reasons of differences in survival
highlighted by the EUROCARE main analyses
To describe and compare patterns of cancer care between
countries and regions
To study adherence to standard cancer care
To investigate the dissemination of innovative treatments in
current clinical practice
To use updated tumour classifications, also aking use of
biomolecular markers
To investigate the influence of comorbidity and metabolic
factors on the prognosis of cancer patients
By collecting more detailed clinical information than in the
usual registry activity
FURTHER AIMS of the High-Resolution studies
Update life status and clinical follow-up of the patients
included in past High resolution studies time of recurrences
and disease free interval
To investigate the feasibility of studying survivorship
Past EUROCARE high resolution studies
Year diagnosis 1987-89
1990-92
1996-98
Breast
Colorectal
Testis
Stomach
Prostate
Italian EUROCARE-5 HR study, cases 2003-2005, follow-up end 2007
(breast, colorectal, lung, melanoma, lymphoprolipherative)
EUROCARE HIGH RESOLUTION PUBLICATIONS
Ten-year survival and risk of relapse for testicular cancer: a
EUROCARE high resolution study Eur J Cancer 2007;43(3):585-92.
What reasons lie behind long term survival differences for gastric
cancer within Europe? Eur J Cancer. 2010 Apr;46(6):1086-92.
Operative mortality after gastric cancer resection and long term
survival differences across Europe. Br J Surg 2010 Feb;97(2):235-9.
Differences in stage and therapy for breast cancer across Europe
IntJCancer 2001; 93:894-90.
Stage at diagnosis is a key explanation of differences in breast
cancer survival across Europe IntJC 2003; 106: 416-422.
Breast Carcinoma Survival in Europe and the United States: A
Population-Based Study Cancer 2004; 100/4: 715-722.
Prognostic Value of Morphology and Hormone Receptor Status In
Breast Cancer – A Population-Based Study. BJC 2004 4;91(7):1263-8.
Variation in “standard care” for breast cancer across Europe: a High
Resolution study. Eur J Cancer. 2010 Jun;46(9):1528-36.
Salad vegetables dietary pattern protects against HER2 positive
breast cancer : a prospective Italian study. Int J Cancer 2007
15;121(4):911-4.
Do pre-diagnostic drinking habits influence breast cancer survival?
Tumori 2011;97(2):142-8
Understanding variation in survival for colorectal cancer in Europe: a
EUROCARE high resolution study. Gut 2000;47:533-8.
Comparison of regional patterns of care and survival for cancers of
breast and colorectum in Europe. IARC Technical Publication No. 37,
IARC Press Lyon 2003.
Survival differences between European and US patients with
colorectal cancer: role of stage at diagnosis and surgery. Gut 2005;
54: 268-273.
Patterns of care for European colorectal cancer patients diagnosed
1996-98: a EUROCARE high Resolution study. Acta Oncol. 2010
Aug;49(6):776-83.
Late outcomes of colorectal cancer treatment: a FECS –EUROCARE
study. J Cancer Surviv. 2007 Dec;1(4):247-54.
Prostate cancer treatment in Europe at the end of 1990s. Acta Oncol.
2009;48(6):867-73.
Regional inequalities in cancer care persist in Italy and can influence
survival. Cancer Epidemiol. 2012 Jul 5.
Breast cancer survival in the US and Europe: A CONCORD highresolution study. Int J Cancer. 2012 Jul 20.
CRITICAL points of the EUROCARE High Resolution studies
 Long
time interval between data collection, quality checks,
statistical analyses and publication of results
 Thus
published papers describe the past not the current
situation
 Very
expensive to carry out
 Representativeness
 Number of
 Long-term
with respect to incidence series
cases and statistical power, robustness of results
survival difficult to estimate (re-update life status
/recurrences, linkage with basic EUROCARE database not
possible/not allowed)
Strengths and achievements
Registries proved able to collect HR data allowing
generalized (population-based) conclusions:
Variation in stage at diagnosis explained most survival
variations for breast, colorectal and stomach cancer;
treatment was a major survival determinant for testicular
cancer
Strengths and achievements

Presently many registries collect and analyze high resolution
data

There is growing interest in investigating the effectiveness of
new diagnostic and therapeutic procedures: HR studies can
help

Interest in Outcomes research -- collaboration with: OECI,
Alleanza Contro il Cancro, EuroCan Platform, EPAAC WP8on
research

interest to link population and clinical data
IS IT NOW THE TIME
TO LAUNCH AN UPDATED
EUROPEAN HIGH RESOLUTION STUDY?
General study design and organization proposal
Data management similar to EUROCARE- Survival
Uniform study protocol
 Centralised
data base, uniform quality checks
 Same data
access and publication rules, adapted to the
HR Working group
Cases included in the HR study: Sample of incident cancer
cases for which the relevant HR data could be collected
either
Retrospectively or prospectively
Prospective data collection:
Advantages:
Collection of clinical data could became part of the usual
registry procedure, with no need to recuperate clinical
documents that are archived elsewhere
Disavantages
Heavily dependent on the local registries procedures used
for completing their files
Appropriate methods should be studied in order to ensure
appropriate sampling and representativeness
Difficult to check data completeness
Need of long time interval to study survival
Retrospective data collection
Advantages
 It ensures
representativeness with respect to incidence
series (and population)
 Allows inspecting
and collecting the whole available clinical
information and checking its completeness
 Follow-up for life
status available from EUROCARE-Surv 
speed analyses
Disavantages
 More expensive than prospective data collection
 High proportion of missing data (?)
Retrospectve data collection:
Randomly sampling an appropriate number of cases from
the EUROCARE survival database (centralized)
From the latest available year of incidence, in most
registries 2007 or later
Send record tracks to the relevant cancer registry for
collection of HR clinical & Follow-up variables
Centralised data checking for format and variable
consistency
Invalid /defective records back to the registries for
appropiate corrections
Linking HR and survival individual records helps speed
analyses and reduces time lag between call for data and
availability of results
HR record structure & organisation
Eurocare-5 record:
Specific
Patient identification variables
High Resolution variables
Date life status follow-up
High resolution record
Transmission to the central repository
Quality Checks
adherence to protocol,
consistency,completeness
To CRs for
Revisions and
corrections
NO
OK
EU High
Resolution
Data Base
Specific HR variables common for all cancers
Clinical characteristics, diagnosis
 Way
of diagnosis: screening, symptomatic/asymptomatic
 Clinical
and pathological TNM stage at diagnosis (or other
cancer specific classifications)
 Diagnostic examinations
 Type
of nodal examination (sentinel, lymphadenectomy)
 Total/ metastatic N.
 Tumour
lymph-nodes
morphology, grading
 Molecular
biomarkers (cancer- specific)
Specific HR variables common for all cancers
Treatment & Follow-up

Surgery, chemo, radio, target , hormonal

Type of treatment (adjuvant, neoadjuvant)

Tumour stage after neo-adjuvant treatment

Type of relapse

Date relapse

Cause of death
Comorbidity

Presence of other diseases

Metabolic variables (BMI, glycaemia)
Cancers where experience on HR studies exists
Breast
•Most frequent cancers, represent public health
issue, increasing incidence and survival
Colorectal •Mass screening in course in many countries
•remarkable differences in care and survival across
and within countries
•New treatments available
•Existence of guidelines or protocols for diagnosis
and treament
Lung
•Frequent cancer, no overtime survival increase
•Uniformely poor prognosis, but strongly
dependent on stage and surgery
Cancers where experience on HR studies exists
Stomach •Incidence decreasing, but still highly frequent
cancer
•Poor prognosis
•Differences in survival largely explained by
subsite and stage
Prostate •Most frequest cancer in men, incidence and
survival increasing
•Large differences in survival across countries
•PSA diffusion and opportunistic screening impairs
interpretation
Cancers where experience on HR studies exists
Skin melanoma •Unfrequent cancer, but incidence increasing
in most EU countries
•Large differences in survival across countries
•Relatively favourable prognosis
•Differences in survival largely explained by
subsite and stage
•Screening campaigns in course in some
countries /opportunistic screeing
•New treatments available/ under evaluation
Cancers where experience on HR studies exists
Haematological •Changing diagnostic criteria and
classifications need accurate disease
malignancies
definition
•new effective treatments available
•Increase in survival, but mostly in wealthy
countries
•Long term prognosis still to be investigated
•Outcomes depend on availability and access
to good care
Cancers where experience on HR studies exists
Testis
•Unfrequent cancer, but incidence increasing
in most EU countries
•Prognosis good in most countries, low
survival largely depends on inadequate
treatment
•Outcomes reflect well the effectiveness of
health systems
•Death sentinel events (avoidable deaths)
Other Cancer sites to be investigated
Cervix uteri
Ovary
Orientative time plan
Early 2013. Preparation and circulation of study protocol
March – July 2013. Data collection by CRs
March – October 2013. Centralised data check & corrections
Within end 2013 – Preliminary data analyses
Incidence 2007, Follow-up 2011–2012
first results early 2014