Comprehensive geriatric assessment in older people undergoing cancer treatment
Dr Danielle HarariConsultant Physician, Senior Lecturer
Guys & St Thomas’ Hospital Foundation NHS Trust, Kings College London
Improving Cancer Treatment Assessment and Support for Older People Project: partly funded by the Department of Health and Macmillan Cancer Support (registered charity no 261017), supported by Age UK (registered charity no
1128267)
What is the problem? Cancer Reform Strategy, NCEPOD, National Chemotherapy Advisory Group, NICE
'Britain's cancer shame as 15,000 elderly patients could be saved every year' Daily Mail June 2009
Overall cancer survival in the UK is improving but not for older people (National Cancer Intelligence Network 2010)
Older people (with same cancer & comorbidity profile as younger) receive less curative or adjuvant treatments
Lack of evidence to guide treatment in older peopleClinical trials include small nos. fit older people - benefit
from therapy as much as younger patients (survival, QOL)
BUT exclude frailer OP (often those seen in clinical practice especially in myeloma)
What is needed?
Risk assessment methods to provide guidance on appropriate levels of treatment in older people
Comprehensive support to optimise outcomes in frailer patients
Trials of modified treatment in older and frailer patients (does dose reduction limit toxicity, but at a cost to tumour response?)
DH/Macmillan/AgeUK funded 5 national ‘Older Persons Pilots’ (including SELCN)
What is Comprehensive Geriatric Assessment (CGA)?
STRUCTURED ASSESSMENT of older patients to identify comorbidities, physical, psychological and social functional problems plus
INTERVENTION - addressing these issues through ongoing patient-centred management plans (often multidisciplinary)
Domains covered by variety of tools (not prescriptive, can be adapted to diff settings)
Improves outcomes in geriatric literature
Role of CGA in oncology:current situation
Oncologists usually use Life Expectancy & Performance Status
PS gives little info beyond mobility and does not assess reasons underlying functional difficulties
Comorbidities rarely formally assessedLife expectancy – meaningless without comorbidity
assessmentNo assessment or support specific to the needs of
older people in NHS cancer services
Role of CGA in oncology:current situation
Growing interest (SIOG, DH, Macmillan, NCEPOD) in integrating CGA into pre-treatment assessment to
- avoid age-based treatment decision making
- inform treatment choices to optimise outcomes Existing oncology studies show CGA
can predict morbidity and mortality is feasible cancer outcomes and toxicity can be predicted by CGA domains
such as functional dependency, depression and comorbidity
Increasing use of brief ‘frailty’ scores (e.g. Balducci) and prescriptive ‘CGA’ tools to decide if patients are ‘fit’ for chemotherapy
BUT dangers of using CGA assessment without intervention…
Extra issues identified by CGA scores may lead oncologists to overestimate treatment risk
Women 70+ breast cancer CGA-screened: Treatment plan changed by oncologists in 39% to less active treatment (most influenced by depression and low weight)
Use of briefer tools may also overestimate riskCGA assessment should aim to accurately:
- identify ‘fit’ patients for full cancer Rx - identify at risk patients for optimisation by geriatricians or other providers to improve fitness for cancer treatment
‘POPS-GOLD’ – Improving cancer treatment in older people South-East London Cancer NetworkProject Lead: Dr Danielle HarariProject Team: Dr Tania Kalsi (Spr fellow), Gordana Babic-Illman (CNS)Collaborators (haemoncology): Dr Paul Fields
Project funding from Department of Health (Health Care
Inequalities, Cancer Strategy), Macmillan, GST CharityObservational: what factors (age, comorbidity) influence whether
or not older people are offered evidence-based care? Can geriatric-oncology liaison improve (a) appropriate
treatment decisions (b) treatment tolerance (c) patient-reported outcomes (QOL) (d) healthcare processes (e.g. transport to hospital, unplanned admissions, LOS)?
Patients aged 70+ being considered for cancer treatment
Complete CGA/comorbidty questionnaire
Observational ‘pre’ groupUsual care
POPS-ONCOLOGYLow-risk patients identified as ‘fit’ At risk patients assessed for comorbidity optimisation pre-treatmentCGA ‘holistic’ supportFollow-through during treatment including liaison on oncology wards
OUTCOMES% undergoing treatment with curative intentTreatment tolerance (toxicity, completion of planned protocol, decompensation of chronic conditions)Hospitalisations (emergency, length of stay)Patient reported quality of life, function, mood
Findings from observational work (‘pre’ group) – all patients
completed GOLD-CGA questionnaire:
Why may older people be ‘under-treated’
GOLD-CGA questionnaire
All questions source-referencedComorbidities questions nuanced e.g. is
BP usually high when checked, breathless on walking on flat surfaces
Evidence-based functional scoresEORTC-QLQ-C30 (cancer-specific QOL
tool validated in older people)
CGA screening in patients with lymphoma BSH 2012
o 74 older patients (aged ≥65) attending lymphoma clinic (mean age 74) o Mean questionnaire completion time was 11.5 + 7.4 minutes.o Comorbidities included: BP usually high when checked 23%, diabetes 21% (6%
poorly controlled), angina/previous MI 11%, breathless on flat surfaces 27%o Cognition: confusion episodes 12%, significant memory problems 11%o Polypharmacy ( 4 medications) 30%o Function: Difficulties with 1 basic activity of daily living (ADL) 48%, with 1
instrumental ADL 53%, fatigue 71%, pain 38%, incontinence 26%o 34% lived alone, 14% had noone to look after them for a few days if needed
o Questionnaire responses were used to categorise as low or high risk:o Low risk = no functional difficulties, no active comorbidity, mild QOL difficultieso High risk = functional difficulties &/or active comorbidity &/or severe QOL difficulties.
o 64% of patients aged 70+ and 48% of those aged 65-70 were high risk, often with a combination of comorbidities, functional difficulties & QOL issues
Frailty- a comparison of diagnostic criteria SIOG 2013
108 patients judged fit for chemotherapy by usual clinical oncological practice, had frailty categorisation assigned retrospectively. This enabled a comparison between clinical judgement of fitness and the 2 frailty criteria for fitness.
Participants were defined as "fit" or "frail" using the Balducci criteria and a frailty index:
The Balducci criteria defined frail:age 85+ &/or functional deficit (≥1 ADL dependency) &/or serious comorbidity (serious cardiovascular, respiratory or
cerebrovascular disease or 3+ comorbidities) &/or presence of any geriatric syndrome
• The frailty index was derived from 43 items from the CGA-GOLD screening questionnaire using methodology as described by Rockwood.
Frailty- a comparison of diagnostic criteria SIOG 2013
The frailty index classified 33.0% (35/106) as frail compared with 72.6% (77/106) by the Balducci criteria
There was poor agreement in who was fit or frail between the 2 diagnostic criteria (kappa=0.25)
The use of Balducci criteria to define frailty to aid treatment decision-making may risk under-treatment of older people with cancer. Frailty indices (based on CGA screening data) may provide a more comprehensive approach.
Chemotherapy treatment decision-making should not be based on the result of frailty scores whilst existing tools do not reliably agree on who is “frail” in this setting. The optimal measure of frailty to apply to clinical practice with proven abilities to accurately detect frailty has yet to be identified.
Low grade toxicity in older people undergoing chemotherapy ECCO 2013
N=108 patients aged 65+ recruited at start of chemotherapy
Research questionTo identify which level of toxicity (and how
many toxicities) trigger a) treatment modification
• defined as dose reductions, delays or drug omissions
b) early discontinuation of chemotherapy
Results: treatment modifications due to toxicity N=60 (55%)
35% (21/60) had no greater than grade 2 toxicity
Of these 21: Mean 2.19+/-1.33 grade 2 toxicities 7 patients had only one grade 2 toxicity Range of G2 toxicity types
Most common: Fatigue (8), haem (8), GI (6) & infections (5)
Results: Toxicity grade trigger to treatment modification (N=60) by comorbidity
Few Comorbidities (<4) N=41
Multiple comorbidities (4+)N=19
Low grade toxicity57.9%(N=11)
High grade toxicity42.1%(N=8)
High grade
toxicity75.6%(N=31)
Low grade toxicity24.4%(N=10)
Statistically significant: p=0.011, 2=6.41
Results: Early discontinuation due to toxicity N=23 (21%)
39.1% (9/23) had no greater than grade 2 toxicity.
Of these 9: Mean 1.78+/-1.2 grade 2 toxicities One grade 2 toxicity n=3 Most common grade 2 toxicities: fatigue
(5) and haemotological toxicity (4)
Key questions & future research in low grade toxicity
Truly have a greater clinical impact on older people?
Is this related to differences in the clinical interaction between dr & older patient?Lower threshold for modifying/discontinuing
treatment in older people? If so, why?Reporting behaviour?
Additional support (e.g. geriatrician liaison) improve treatment tolerance?
Fatigue in older people undergoing chemotherapy SIOG 2013
Baseline fatigue is rarely documentedFatigue toxicity was cited by treating oncologists
in 69.1% (n=75) of all patients during chemotherapy, with grade 2+ occurring in 36.1% (39) and grade 3+ occurring in 11.1% (11)
Fatigue severity from EORTC-Q30 as part of CGA-GOLD questionnaire
Improved fatigue % (N)
No change % (N)
Fatigue worse % (N)
At 2 months follow up (n=89) 14.6 (13) 71.9 (64) 13.5% (12)
At 6 months follow up (n=68) 14.7 (10) 76.5 (52) 8.8 (6)
Findings from interventional work (‘post’ group) :
Impact of geriatric-oncology liaison in outpatients and
inpatients (oncology wards)
GOLD PATHWAYS DEVELOPEDOLDER PATIENT WITH CANCER
SELF REPORTING CGA SCREENING QUESTIONNAIRE
IN DEPTH REVIEW BY GERIATRICIAN TO
OPTIMISE/REVERSE CGA
INFORM ONCOLOGY
HIGH RISKLOW RISK
TREATMENT DECISION
CONTINUED GERIATRICS SUPPORT & RE-REVIEW AS NEEDED
ONCOLOGY REFERRAL
NO CGA REQUIRED
SERVICE DEVELOPMENT – CLINIC PATHWAYS
Tailor CGA intervention to cancer treatment Optimise in relation to tx and plan proactively
for anticipated cancer treatment toxicityDeveloped to fit in within existing oncology
pathwaysTailor to individual needs of the tumour groups
bladder cancer - joint clinic with a walk-in CGA colorectal and prostate cancer - fast track review
typically within 1 week of referral
Examples of targeted interventions
Cardiac and cardiac risk optimisation in patients receiving anthracyclines
Improving renal function in those to receive platin based chemo – polypharmacy etc
Treating pre-existing anaemia – iv iron, B12 and folateDiabetes management with steroidsNutritional supportPain and mobility optimisation (osteoarthritis)Fatigue investigation and management plan –
protocolised fatigue pathway developedManaging continence (QOL)Transport assistance esp for people having outpatient
chemo/RT
Screening QuestionnaireRECRUITED n=177
BEXLEY GP GROUP n = 31
GSTT GROUP n=146
SCREENING QUESTIONNAIRE NOTE REVIEW AND TELEPHONE CLINIC FOR CGA NEED
IN DEPTH CGA CLINICN=73 (50%)
NO CGA CLINIC AS PER NEED OR WISHES N=73 (50%)
Questionnaire Validity & Reliability (EUGMS 2013, BGS 2103)
Inter-rater reliability Subgroup of 71 patients, 2 clinicians (SPR & CNS) review same
screening questionnaires Same decision in 87.3% (n=62/71) of questionnaires
Reliability: against clinical notes review Clinician 1 (SPR): notes changed decision of CGA need in
10.9% (n=9/82) patients Clinician 2 (CNS) notes changed decision in 9.6% (n=8/83)
patients
Acceptability: patient responses o 80.2% (n=142) did not need help to completeo Mean time to complete: 14.5 mins +/- SD 9.3
Outpatients - Comorbidities
IN DEPTH REVIEW BY GERIATRICIAN TO
OPTIMISE/REVERSE CGA
HIGH RISKLOW RISK
NO CGA REQUIRED
COMORBIDITIES MEDIAN 3.0
MEAN 2.51 +/- SD 1.9.
COMORBIDITIESMEDIAN 6
MEAN OF 5.75 +/- SD 2.4
Did POPS-GOLD influence oncology treatment decision-making BGS 2012
60% (n=24) of oncologists responded to semistructure questionnaire (21% consultants, 63% registrars, 17% clinical nurse specialists)
All respondents had read the CGA assessment letter at the patient’s next cancer appointment.
63% (n=15) reported the assessment had influenced their decision-making.
Of these, 67% (n=10) reported CGA assisted the evaluation of fitness for treatment, more often in favour of active treatment (8 versus 2 patients).
Common themes reported as beneficial were: medical review (n=5) increased information (n=3) facilitated communication (n=2) increasing confidence (n=3).
Did POPS-GOLD influence oncology treatment decision-making BGS 2012
“it was so helpful.....we thought he might have had a cardiac problem related to the chemo but you have identified the culprit drug. Based on your consultation, we decided to continue chemotherapy without any dose reductions”
“Overall, POPS review was a very helpful and precise holistic assessment of the patient”
“Partly......altering medications had improved her symptoms. But balance is to control disease vs toxicity and she was relatively symptom free”
“Confirmed impression that not fit for further systemic therapy and that efforts should be palliative. It was really useful to confirm co-morbidities and their impact on symptoms. Also useful to clarify modifiable factors...”
“No. We knew what treatment the patient needs to be on. However, the pt did mention he found the POPS review helpful particularly with respect to medications”
“increased confidence in proceeding with chemo with knowledge of optimal medical management”
Of the 9 who reported no influence on decision-making, 5 found it useful for other reasons:
“the reduction in antihypertensives is likely to mean he will tolerate radiotherapy”
Did POPS-GOLD influence oncology treatment decision-making BGS 2012
To impact on decision-making, CGA needs to be delivered within a tight timeframe to fit in with existing cancer targets. This could be a challenge for an already busy geriatric medicine department. However, the CGA screening questionnaire allowed us to assess for CGA need. This meant clinic time could be utilised effectively to enable rapid CGA delivery for those that needed it most.
Within limitations, this evaluation highlights the potential benefits of geriatrician-led CGA, more often in favour of more actively treating older people
o Early CGA can influence oncology decision-making.
o Feedback suggests this relates not only to improved medical support and the information provided, but by increasing confidence to actively treat older people with cancer.
Patient & Carer Feedback
“Nice to know GOLD are there to give advice and help with possible problems.”
“There is time to talk and the Doctor looks at you as a person and how you can cope with the medical problems”.
“The clinic is very relaxed and you feel there is time to talk, whereas other clinics are so busy and the Doctor is catching up with information on the computer.”
‘They saw my mother a few weeks ago and did a fantastic job in sorting her out for chemo. Consultant haematologist
In-patient Liaison
Service & Pathway Development for geriatric liaison on oncology wards
Identified patients morning board rounds (CNS)MDT (CNS/SPR)Case note review (CNS/SPR)Patients were stratified according to risk- pathways
Clinical ReviewFor patients in need Optimised in a similar way to in the CGA clinic. Discharge planning
GOLD Intensity of Input
GOLD Intensity of Input N = 113% (n)
Not involvedLight touchMedium touchHeavy Very heavy
37% (42)25% (28)11% (13)20% (22)
7% (8)
Impact on quality of information across to primary care and community and coding
Oncology Discharge letter GOLD ENHANCEDPRINCIPAL DIAGNOSIS1. AML
COMORBIDITIES2. Myelodysplasia
PRINCIPAL DIAGNOSIS1. Neutropenic Sepsis 2. Anaemia secondary to UGI (gastric ulcers) and
AML - needing blood transfusion3. Pancytopenia4. AML - end of life - fast-tracked to hospice 5. Pulmonary oedemaCOMORBIDITIES1. MDS2. AML3. Gastric ulcers4. Barrett Oesophagus5. Hypertension 6. B12 deficiency7. Folate deficiency8. Angiodysplasia,9. Lives alone
Impact on length of stay
5 6 7 8 9 10 11 12 13
LOS IN DAYS
OCT 12 NO POPS
SEPT 12 POPS CNS & SPR
AUG 12 POPS CNS & SPR
JULY 12 POPS CNS & SPR
JUN 12 POPS CNS & SPR
MAY 12 POPS -CNSMAINLY
APRIL 12 NO POPS(HOLIDAY/CONFERENCES)
Mar 12 POPS CNS MAINLY
FEB12 POPS - CNS MAINLY
JAN 12 NO POPS
DEC 11 NO POPS
NOV 11 NO POPS
MO
NT
H W
ITH
/WIT
HO
UT
PO
PS
LOS WITH AND WITHOUT POPS
Series1 9.8 7.2 7.2 9.4 8.7 10.6 11.5 9.1 9.5 11.7 11.5 12.5
OCT 12 NO POPS
SEPT 12
POPS
AUG 12
POPS
JULY 12
POPS
JUN 12 POPS CNS &
MAY 12
POPS -
APRIL 12 NO POPS
Mar 12 POPS CNS
FEB12 POPS -
CNS
JAN 12 NO
POPS
DEC 11 NO
POPS
NOV 11 NO POPS
Impact on LOSLOS in patients aged 65+ reduced with GOLD
Pre-GOLD LOS: 11.7-14.0 days (Oct 11-Jan 12)
Partial GOLD LOS: 9.1 - 9.5 days (Feb 12 – March 12)
GOLD LOS: 7.2 - 9.4 days (Jun – Aug)
In addition, a number of younger patients with complex needs and lengthy hospitalisations would benefit from this approach.
Our scoping would suggest that at least half of all inpatients fall into the category of requiring GOLD input
Dissemination to oncology training bodies
Survey of medical oncology trainees
Kalsi T, Payne S, Brodie H, Wang Y, Mansi JL, Harari D. Are UK oncology trainees adequately informed about the needs of older people with cancer? British Journal of Cancer 1–6 | doi: 10.1038/bjc.2013.204
Survey currently being considered in the revision of the national medical oncology curriculum
Geriatric Oncology Training During Specialist Training66.1% never received any training on the needs of
older people with cancer19.4% had only ever received this training once
Training in geriatrics specific issues common in oncology patients (eg delirium, falls)Of those who had received training, the majority
received it 3 years ago Want training
cognitive impairment/delirium (n=18)polypharmacy (n=17) discharge planning (n=7).
Practice in cognitive impairment
Cognitive assessments45.9% rarely/never assessed
Consent and Mental Capacity Assessment27.3% never consent patients with cognitive
impairment50.9% would rarely consent38.9% MCA never/rarely used to decide about
the patient’s understanding
Confidence in risk assessment
81.4% confident for younger pts 27.1% for older patients 10.2% for older patients with dementia
25.4% confident/extremely confident managing
multiple comorbidities
Macmillan/DOH/Age UK report: Cancer Services Coming of Age, Dec 2012
http://www.macmillan.org.uk/Aboutus/Healthprofessionals/Improvingservicesforolderpeople/Pilots/PilotSites.aspx
Department of health recommendations
improving survival rates in the population aged 75 years and over
to deliver high quality services to increasing numbers of older patients with cancer, including age appropriate assessment, for example the Comprehensive Geriatric Assessment (CGA)
involvement of elderly care specialists
http://cno.dh.gov.uk/2012/12/20/cancer-services-coming-of-age-report-published/
How can oncologists, surgeons and geriatricians work together?
CGA / comorbidity screening with identification of low and at risk patients can be done in oncology clinic
In-depth CGA for at risk patients (outpatient) – ideally joint oncology/geriatric clinics
Assessment is part protocolised so could also be done by oncology with geriatrician support
Inpatient liaison – medical optimisation, rehabilitation goal setting, early discharge planning – dedicated geriatric liaison team is preferred model (if funded…)
Could be done by oncologists with consultative support and geriatrician sitting in on ward MDM