”Lærende succeshistorie”: Resultat af klassisk ... · Loco-regional control (%) 0 20 40 60 80...

Preview:

Citation preview

”Lærende succeshistorie”:Resultat af klassisk databasedrift og interaktion mellem forskning og kvalitetsmonitorering i DAHANCA

Jens OvergaardDanish Cancer Society, Department of Experimental Clinical

Oncology, Aarhus University Hospital, Denmark.Jens@oncology.dk

Regionernes nationale databasedag 2015København – 8 april

The Danish National Head and Neck Cancer Database

www.dahanca.dk

Jens OvergaardDanish Cancer Society, Department of Experimental Clinical

Oncology, Aarhus University Hospital, Denmark.Jens@oncology.dk

DAHANCA.dkDanish Head and Neck Cancer Group

Head and neck cancer

OropharynxLarynx

Epiglottic

EsophagusTrachea

Nasal cavity

Oral cavity

Supraglottic

Glottic

Subglottic

Nasopharynx

Laryngopharynx

Pharynx'

Head and Neck Cancer:

Probably the most diverse and heterogeneous cancer type in Europe

Major variations in:

IncidenceEpidemiologyEtiologyTreatmentPrognosis

1500 pts per year(4% of all cancer cases in Denmark)

Head and Neck cancer is a heterogeneous disease

… and so is the treatment strategy

established byTHE DANISH SOCIETY

FOR HEAD & NECK ONCOLOGY1976

DAHANCA.dkDanish Head and Neck Cancer Group

Organized by the Danish Society for Head and Neck Oncology

MULTIDICIPLINARY collaborative group involving

ALL relevant Danish departments and specialities

– and taking care of ALL head and neck cancer

patients in Denmark (and some in Norway)

National databaseRegistration and follow up of all H&N patientsNational treatment strategyNational (evidence-based) guidelinesClinical trials and studiesQuality assuranceBiobank Translational research infrastructureResearch projects (Ph.D students) Education and information (professionals – public)International collaboration (all issues above)

DAHANCA.dkDanish Head and Neck Cancer Group

DAHANCA.dk

DAHANCA –databaseIncludes: ICD-7 no. 141-148, 160-161,194.Since 1960: Larynx from 3 largest institutions (80%)

1971-91 Nat’l “official” larynx database (> 99%)

1992 - Nat’l “official” Head & Neck base (larynx, pharynx, oral cavity)

1996 - DATYRCA (thyroid cancer)

2000 - Unknown primary data base

2002 (1996)- DASPYTCA (salivary gland tumours)

2005- Biobank database (linked with clinical data)

2008- Nasal Sinus etc...

2010- Approved as “Nat’l Clinical Quality Database”

2010- Establisment of QA database for Radiotherapy DAHANCA.dk

per April 2015

>31.000 patient

s included

DAHANCA -databaseTYPE OF DATABASE.

Pre 1991 “Homemade” flat FORTRAN base (on the medical physicist’s computer).

DAHANCA.dk

DAHANCA -databaseTYPE OF DATABASE.

Pre 1991 “Homemade” flat FORTRAN base (on the medical physicist’s computer).

1991 Medlog (BMDP) Local recording with file to central collection

2003 - Internet based central database using Clarion on a Windows serverplatform.

Run from all computers (128 bit cryptation) Extractable to e.g. textfile, Excel, STATA (Bill Gates is the winner)

DAHANCA.dk

DAHANCA -databaseSTRUKTUR

DAHANCA.dk

Overordnet databaseansvarlig

(datatilsynet)

Onkol. center

RH

Onkol. center Herlev

Onkol. center SDU

Onkol. center Århus

Onkol. center Aalb

Onkol. center Oslo

DATHYRCA

Omfatter ICD-7 no. 141-148, 160-161,194

Lokaliseret: Afd. Eksp Klin Onkologi, AUH. (Aleks)

Tilknyttet: Kompetencecenter Nord (bruges ikke)

Dokumentationscenter for Kræftbehandling (KB)

Driftsprincip: entusiastisk, non-bureaukratisk

Økonomi: RKKP – resten forskningsfonde o.l.

DAHANCA -databaseWHY HAVE A DATABASE?

ResearchProtocols, pro- and retrospective studies attracts youngsters

Quality control (“law”)

Health care management

“Political reasons”. (knowledge is power)

DAHANCA.dk

DAHANCA -databaseWHAT DO WE RECORD

“Relevant” baseline data (> 300 variables)

“On Study, Treatment, Follow-up, Recurrence, Death”.

Flexible (ad hoc data if needed)

Longitudinal (time trends) – contineous variables

Exchange/International(AJC, UICC, ICD, EORTC, INKA, etc)

Data must be verified to be applicable and useful)

DAHANCA.dk

DAHANCA -databaseDATA MUST BE USED (updated and verified)

National Cancer Registry

(Eurocare, Nordcan)

Hospital records

CPR (unique personnumber)

EPJ, e-journal

Course of death registry

Pathology registry

Other …….

DAHANCA.dk

DAHANCA -databaseDATA MUST BE USED (and verified) and GOOD

> 99% coverage of national patient cohort (compared with Danish Cancer registry)

DAHANCA.dk

Problem:Den ”ægte” varevs. ”industrielt genbrug”

Problem:Vi kan ikke ”tilbagerette”forkerte data(så de går igen)

DAHANCA -databaseWHAT DO WE USE IT FOR?

A FEW EXAMPLES:

DAHANCA.dk

EXAMPLE 1

DAHANCA.dk

Time after treatment (months)0 6 12 18 24 30 36 42 48 54 60

Loco

-regi

onal

con

trol (

%)

0

20

40

60

80

100

27%

DAHANCA Database Stage 3-4 Larynx and Pharynx

5 fx/wk

The DAHANCA strategy: progression through consecutive clinical trials

Standard 1985

EXAMPLE 1

Primary RT of HN

Cancer

larger RT dose

Hyperfx

Neo-adjuvant Chemo-

RT

Smaller RTvolume

IMRT

Hypoxic modification

of RT

Basic “virtues”

Waiting time etc.

Con-commitant Chemo-RT

Reduced RT time

Accl fx

Changed RT-Surg balance

Better diagnosis

and imaging

Dahanca 2

Dahanca 5

Dahanca 10

Dahanca 1

Dahanca 9

Dahanca 11

Dahanca 13

Dahanca 15

Dahanca 6

Dahanca 7

Dahanca study

DAHANCA.dk

Hypoxic modification

of RT

Reduced RT time

Accl fx

Dahanca 14

Dahanca 18

Dahanca 19Dahanca 16

Con-commitant Chemo-RT

Dahanca 24

Dahanca 27

Dahanca 29

Dahanca 30

Dahanca 28

Comorbidity

Biological modifiers

Anti-EGFr

DAHANCA.dk

Time after treatment (months)0 6 12 18 24 30 36 42 48 54 60

Loco

-regi

onal

con

trol (

%)

0

20

40

60

80

100

27%

DAHANCA Database Stage 3-4 Larynx and Pharynx

5 fx/wk

5 fx/wk + NIM44%

Benefit of Hypoxic Modification

The DAHANCA strategy: progression through consecutive clinical trials

Standard 1985

DAHANCA.dk

Time after treatment (months)0 6 12 18 24 30 36 42 48 54 60

Loco

-regi

onal

con

trol (

%)

0

20

40

60

80

100

27%

DAHANCA Database Stage 3-4 Larynx and Pharynx

5 fx/wk

5 fx/wk + NIM44%

62% 6 fx/wk + NIM

Benefit of Hypoxic Modification

Benefit of Accelerated Fract.

The DAHANCA strategy: progression through consecutive clinical trials

Standard 1985

DAHANCA.dk

Time after treatment (months)0 6 12 18 24 30 36 42 48 54 60

Loco

-regi

onal

con

trol (

%)

0

20

40

60

80

100

27%

DAHANCA Database Stage 3-4 Larynx and Pharynx

5 fx/wk

5 fx/wk + NIM44%

62%

83%

6 fx/wk + NIM

6 fx/wk + NIM + cis-P

Benefit of Hypoxic Modification

Benefit of Accelerated Fract.

Benefit of Chemo-Radiotherapy

The DAHANCA strategy: progression through consecutive clinical trials

Standard 1985

Standard 2007

Larynx/oropharynx/ hypopharynx

HPV/p16 negativeStage III - IV (T1-4, N0-3)

RANDOMIZATION

Accl RT (70 Gy, 6 fx/wk) + cddp (40 mg/m2

weekly x 5 or 100 mg/m2 x 2)+ placebo (daily)

Stratification:•Institution•Localization•T-stage•N-stage•WHO-PS•Hypoxic gene-profile

+ Nimorazole (1.2 g/m2 daily)

Accl RT (70 Gy, 6 fx/wk) + cddp (40 mg/m2

weekly x 5 or 100 mg/m2 x 2)

1219 ROG-HNCG/DAHANCA 29: Study Design

DAHANCA.dk

Charlotte Rotbøl Bøje, MD PhD

Comorbidity among 12.623

Head and Neck Cancer patients from the 

DAHANCA database: 

EXAMPLE 2

ESTRO Barcelona 2012

The CPR Registry

Danish CancerRegistry

The National Patient Registry

Diabetes Registry

Registry ofCauses of Death

DAHANCA

Materials and methods

Prescription registry

National Health Insurance Registry

99.4%

ESTRO Barcelona 2012

• Resulted in more than 1.000.000 diagnoses on the 12.623 patients.

• Diagnoses from 10 years prior to the HNSCC diagnosis were included

Materials and methods

ESTRO Barcelona 2012

Comorbidity and overall survival

CCI=0 50%CCI=1 45%CCI=2 39%CCI=3+ 33%

P<0.001

Crude HR:CCI=1: HR=1.16 (1.08 - 1.25)CCI=2: HR=1.34 (1.22 - 1.46)CCI=3: HR=1.63 (1.51 - 1.80)

N= 12,596

025

5075

100

Ove

rall

surv

ival

(%)

1237 799 549 410 291 221Severe comorbidity1196 854 604 473 366 286moderate comorbidity2110 1582 1199 902 723 587mild comorbidity8013 6266 4843 3943 3330 2831no comorbidity

Number at risk

0 1 2 3 4 5Time after treatment (years)

Comorbidity and SurvivalN=12.623

Presence of comorbidity is associated with increased risk of death,

highly statistically significant

ESTRO Barcelona 2012

Comorbidity and disease specific survival

Crude HR:CCI=1: HR=1.00 (0.9 - 1.1)CCI=2: HR=1.04 (0.93 - 1.17)CCI=3: HR=1.00 (0.89 - 1.13)

N= 12,623

025

5075

100

Dis

ease

spe

cific

sur

viva

l(%)

1232 719 498 370 269 206Severe comorbidity1183 754 515 395 311 245moderate comorbidity2105 1387 1023 783 643 520mild comorbidity7980 5442 4171 3422 2891 2458no comorbidity

Number at risk

0 1 2 3 4 5Time after treatment (years)

Comorbidity and Survival

No association between comorbidity and disease specific survival

HPV in Head and Neck Squamous Cell Carcinoma (HNSCC)

• Oropharynx cancer

• Increasing incidence

• Non‐keratinizing SCC

• p16 expression, p53wt

• Younger age/more fit

• Less smoking

• Less alcohol

• N+ (advanced disease)

• Improved prognosis Updated from Lassen Radiother Oncol 2010

Year

1977 1982 1987 1992 1997 2002 2007 2012

Num

ber o

f pat

ient

s

0

50

100

150

200

250

300

350

400

Oropharynx

Larynx

55%

37%42%

33%

70%

74%

p16‐pos*

*

Denmark 1977‐2013DAHANCA database

EXAMPLE 3

Impact of HPV/p16 on RT‐outcome by tumor siteOropharynx N=1002

Time after treatment (months)0 6 12 18 24 30 36 42 48 54 60

Loco

-regi

onal

con

trol (

%)

0

20

40

60

80

100

81%

55%

DAHANCA DatabaseOropharyngeal tumors

HPV status

HPVneg

HPVpos

p<0.0001

432 pts

569 pts570pts

HR: 0.32 [0.25‐0.42]

Loco‐regional controlHPV/p16 neg

p<0.0001

HPV/p16 pos

0

25

50

75

100

Dis

ease

-spe

cific

sur

viva

l (%

)

0 12 24 36 48 60Time after treatment (months)

HR: 0.19 [0.14‐0.26]

88%

54%

Disease‐specific survival

HPV/p16 neg

p<0.0001

HPV/p16 pos

0

25

50

75

100

Ove

rall

surv

ival

(%)

0 12 24 36 48 60Time after treatment (months)

HR: 0.24 [0.19‐0.32]

80%

50%

Overall survival

Time after treatment (months)0 6 12 18 24 30 36 42 48 54 60

Loco

-regi

onal

con

trol (

%)

0

20

40

60

80

100

57%51%

DAHANCA DatabaseNon-oropharyngeal tumors

HPV status

HPVneg

HPVpos

p=0.34

528 pts

77 pts

HR: 0.89 [0.60‐1.30]

p=0.53

Impact of HPV/p16 on RT‐outcome by tumor site

Time after treatment (months)0 6 12 18 24 30 36 42 48 54 60

Loco

-regi

onal

con

trol (

%)

0

20

40

60

80

100

81%

55%

DAHANCA DatabaseOropharyngeal tumors

HPV status

HPVneg

HPVpos

p<0.0001

432 pts

569 pts

Loco‐regional controlLoco‐regional control

Oropharynx N=1002 Non‐Oropharynx N=604

HR: 0.32 [0.25‐0.42]

N=1606 stage III‐IV

6 marts 2007EXAMPLE 4

Year

1968 1978 1988 1998 2008

'Wai

ting

time'

in d

ays

7

14

21

28

35

”Ventetid” for start på strålebehandling ved hoved-hals kræft i Danmark 1968-2008

Konklusion 6 marts 2007:

Hvad har vi lært?

Der er ingen data der støtter at ventetid er harmløs og uden forringelse af prognosen – derimod er der stort set entydige data som klart viser at selv kort ventetid indebærer en risiko for en markant reduktion af mulighederne for helbredelse

– det er derfor videnskabeligt bevist (hvor det er undersøgt) at ventetid medfører en forringelse for helbredelsesmulighederne og dermed en dårligere prognose

– formodentlig skal en væsentlig del af det efterslæb vi har i Danmark vedrørende resultater af cancerbehandling ses i lyset heraf – både i form af ventetider for nydiagnostiserede patienter og pga. manglende udnyttelse af de screenings-muligheder der foreligger

Ventetid (d

er er ingen

nedre græ

nse) medfør

er

dårligere t

umor kontrol

og deraf d

årligere

overlevelse

- Cancer er e

n akut sygd

om – og skal beh

andles

som sådan

Politiken juli 2007

10 august 2007

Pakkeforløb

Year

1968 1973 1978 1983 1988 1993 1998 2003 2008 2013

Med

ian

"wai

ting

time"

in d

ays

7

14

21

28

35

Median time from 1st contact with oncological centerto 1st day of radiotherapy/surgery

DAHANCA -databaseJakob Axel Nielsen

Bent Hansen

DAHANCA -databaseWHAT’S IMPORTANT

COLLECTION with a “Carrot” – not a stick

Obviously benefit for the persons who input data

Only collect what will be used (we miss it anyway)

Access for the users (they “own” the data)

Alert and control system

Feedback (control, verification, revision)

(remember: a data base is a “living” thing)

DAHANCA.dk

DAHANCA -databaseWHAT’S IMPORTANT (cont..)

Easy input and (research) access (to own data)

Webbased – accessible from all computers

Flexible and not to big

Quality

Connected with biobank (DCB?)

Good clinical data from protocolized studies with associated biobank material is worth more than gold)

DAHANCA.dk

DAHANCA -databaseWHAT’S IMPORTANT (cont..)

USE IT(e.g. annual question)

Ph.d. studies, publications…

DAHANCA.dk

>150 publicationer based on the database4 DMSc and 18 (6 ongoing) Ph.d studies.

Web site for extract of database

Meget begrænset mulighed for elektronisk opfølgning af data!

Paradox: Vi har lov til at kende data, - vi har de praktiske og tekniske muligheder for at fådata - men det må vi ikke elektronisk (og de er ikke længere på papir).

Dette er de kliniske (kræft) databasers største trussel –og den er tæt på at være fatal.

Danish Head and Neck Cancer Group

DAHANCA

Thank you

Recommended