1. Treatment escalation in patients with early stage Hodgkin lymphoma and a positive PET scan after initial chemotherapy is not always required A subsidiary analysis of the UK NCRI RAPID study
S. F. Barrington, E. H. Phillips, N. Counsell, B. Hancock, R. Pettengell, P. Johnson, J. Wimperis, D. Culligan, B. Popova, L. Clifton-Hadley, A. McMillan, A. Brownell, A. Kruger, A. Lister, P. Hoskin, M. O'Doherty, T. Illidge, J. Radford

2. Risk Stratification in Early Stage HL
Goal is to maximise cure and minimise toxicity
Needs accurate risk stratification to individualise treatment
Usually reliant on pre-treatment clinical risk factors
Interim 18FDG-PET assessment identified as one of the strongest prognostic indicators
EORTC
GHSG
Age ≥50 years
Extra-nodal disease
ESR >50 (A)
ESR >30 (B)
Large mediastinal mass
4 or more involved sites
3 or more involved sites .

3. Interim PET in Early Stage Hodgkin Lymphoma
Positron Emission Tomography
Radioactive glucose tracer (18fludeoxyglucose or FDG) to show areas of high metabolic activity .
CT scan
PET scan
PET-CT scan
Zinzani et al EJNMMI 2012

4. PET-Adapted Treatment in Hodgkin Lymphoma
Johnson et al N Engl J Med 2016
Radford et al N Engl J Med 2015

5. H10 study (LYSA/EORTC/FIL)
Risk adapted treatment by interim PET results
Patients with a +ve PET after 2 ABVD
benefit from intensification to eBEACOPP
But
Greater toxicity in eBEACOPP arm
Link slide by outlining advantages to defining which group benefit from treatment escalation
Andre et al J Clin Oncol 2017

6. H10 study (LYSA/EORTC/FIL)
Risk adapted treatment by interim PET results
Patients with a +ve PET after 2 ABVD
benefit from intensification to eBEACOPP
But
Greater toxicity in eBEACOPP arm
PET ‘positivity’ defined by IHP criteria
Now replaced by Deauville score
Score 3 regarded as PET negative
with standard treatment
Link slide by outlining advantages to defining which group benefit from treatment escalation
Barrington et al J Clin Oncol 2017

7. Aim of this subsidiary analysis of RAPID
Do all PET positive patients have equivalent outcomes?
Can we identify a higher risk group within PET+?
Assess prognostic value of:
PET score after 3 ABVD
Pre treatment clinical risk factors
EORTC
GHSG
Radford et al N Engl J Med 2015

8. RAPID - Trial Design
Patients: Stages IA/IIA HL with no mediastinal bulk
Treatment: ABVD chemotherapy x 3
PET -ve: N=426 (75%)
PET +ve: N=145 (25%)
Randomisation
4th cycle ABVD then IFRT
NC – 6 not randomised: 3 patient choice, 2 physician choice, 1 error
30 Gy IFRT
No further therapy

9. Methods
PET scans centrally reviewed at St Thomas’ Hospital- NCRI PET network Prospectively assigned PET score 1-5 PET positive equivalent to Deauville score 3, 4, 5 Cox regression used for association between PET score, baseline risk stratification, treatment arm and HL-specific EFS EFS defined as progression or HL-related death

10. Patient Characteristics
PET -ve: IFRT arm
(N=209)
PET -ve: no further therapy (N=211)
PET +ve
(N=145)
Age
34 (16 – 74)
34 (16 – 75)
36 (18 – 75)
Stage IA
69 (33.0%)
70 (33.2%)
48 (33.1%)
Stage IIA
140 (67.0%)
141 (66.8%)
97 (66.9%)
EORTC favourable
118/184 (64.1%)
122/185 (65.9%)
70/126 (55.6%)
GHSG favourable
114/175 (65.1%)
136/184 (73.9%)
77/123 (62.6%)
67.8% GHSG favourable
62.3% EORTC favourable
For paper will also include proportion >50y, ESR >50, number of nodal sites and extranodal disease
Will also look at ECOG PS- I suspect RAPID included a higher number of older/less fit patients than other early stage HL studies
Sally: note more GHSG favourable patients in the NFT arm, proponents of CMT may point this out ….

11. Outcomes According to Treatment Arm
Primary endpoint: PFS in PET negative randomised patients (ITT)
PET negative: PET score 1 or 2
5y PFS 91.2% (95% CI: )
IFRT: 93.2% (95% CI: )
NFT: 89.2% (95% CI: )
Will mention that there was no significant difference between any of the treatment arms and that some commentators have interpreted this as PET not being prognostic for early stage HL
Radford et al N Engl J Med 2015

12. Outcomes According to Treatment Arm
Primary endpoint: PFS in PET negative randomised patients (ITT)
PET negative: PET score 1 or 2
5y PFS 91.2% (95% CI: )
IFRT: 93.2% (95% CI: )
NFT: 89.2% (95% CI: )
PET positive: PET score 3, 4 or 5
5y PFS 87.2% (95% CI: )
Add DS
Argue that PET does not have significant discrimination in early stage HL
Radford et al N Engl J Med 2015

13. EFS by PET Score
p < 0.01
a high score of 5 was associated with risk of progression or HL-related death

14. Alive with progression
Outcomes by PET Score
PET Score N
Alive with progression
n (%)
HL deaths
5-year EFS
(95% CI) 1
299
19 (6.4%)
0 (0.0%)
93.9%
(91.2 – 96.6) 2
121
9 (7.4%)
1 (0.8%)
91.8%
(86.1 – 97.5) 3 90
3 (3.3%)
1 (1.1%)
95.3%
(90.8 – 99.8) 4 32
2 (6.3%)
93.5%
(84.9 – 100) 5 23
5 (21.7%)
3 (13.0%)
61.9%
(41.1 – 82.7)
For reference PET Score 4-5 together as per Lugano classification: 5y PFS = 74.3% (95% CI: 62.7 – 85.9)

15. Outcomes by PET Score PET Score N Alive with progression n (%)
HL deaths
5-year EFS
(95% CI)
5-year PFS 1
299
19 (6.4%)
0 (0.0%)
93.9%
(91.2 – 96.6)
91.2%
(87.9 – 94.5) 2
121
9 (7.4%)
1 (0.8%)
91.8%
(86.1 – 97.5)
91.1%
(85.2 – 97.0) 3 90
3 (3.3%)
1 (1.1%)
95.3%
(90.8 – 99.8) 4 32
2 (6.3%)
93.5%
(84.9 – 100)
87.5%
(76.1 – 98.9) 5 23
5 (21.7%)
3 (13.0%)
61.9%
(41.1 – 82.7)
56.5%
(36.3 – 76.7)
PET Score N
Alive with progression
n (%)
HL deaths
5-year EFS
(95% CI) 1
299
19 (6.4%)
0 (0.0%)
93.9%
(91.2 – 96.6) 2
121
9 (7.4%)
1 (0.8%)
91.8%
(86.1 – 97.5) 3 90
3 (3.3%)
1 (1.1%)
95.3%
(90.8 – 99.8) 4 32
2 (6.3%)
93.5%
(84.9 – 100) 5 23
5 (21.7%)
3 (13.0%)
61.9%
(41.1 – 82.7)
For reference PET Score 4-5 together as per Lugano classification: 5y PFS = 74.3% (95% CI: 62.7 – 85.9)

16. EFS According to Risk Stratification
Outcome by EORTC risk group
Favourable (n=310, events=26)
5y EFS = 91.7% (95% CI: )
Unfavourable (n=185, events=13)
5y EFS = 92.5% (95% CI: )
Mention similar results looking at PFS
p = 0.68

17. EFS According to Risk Stratification
p = 0.58
No evidence of a difference when adjusted for PET score: EORTC p=0.46, GHSG p=0.51
p = 0.68

18. Multivariable analysis
Comparison
Unadjusted
Adjusted for EORTC group
PET score 1 versus 5
HR=0.14 (95% CI: , p<0.001)
HR=0.15 (95% CI: , p<0.001)
PET score 2 versus 5
HR=0.20 (95% CI: , p<0.01)
HR=0.22 (95% CI: , p<0.01)
PET score 3 versus 5
HR=0.11 (95% CI: , p<0.001)
HR=0.13 (95% CI: , p<0.01)
PET score 4 versus 5
HR=0.15 (95% CI: , p=0.02)
HR=0.19 (95% CI: , p=0.04)
Adjusted for PET score
EORTC
favourable vs unfavourable
HR=1.15 (95% CI: , p=0.68)
HR=1.29 (95% CI: , p=0.46)
However data is analysed, only PET score 5 predicted. GHSG and EORTC risk stratification did not predict outcomes
Will also mention there was no significant difference between PET scores 1-4
NC – similar results for GHSG (in terms of both PET after adjusting for GHSG, and GHSG after adjusting for PET)
NC – similar results when also including study grouping (IFRT, noFT, or PET+), but would be more difficult to present given the relationship between PET score and study grouping (i.e. noFT = PET1 + PET2 – IFRT cohorts)
Similar results with GHSG risk stratification
and when corrected for study group (NFT, IFRT or PET+)

19. Conclusions 1
Early PET score has greater prognostic impact than clinical risk stratification in early stage HL
In patients with stage IA/IIA HL (and no mediastinal bulk) receiving PET guided treatment only PET score 5 predicted poor outcome
‘Positive’ PET score does not carry uniform prognostic weight as reported in other lymphoma subtypes 1-3
1. Ceriani et al Int J Radiat Oncol Biol Phys 2017
2. Hertzberg et al Haematologica 2017
3. Mikhaeel et al ICML 2015

20. Conclusions 2
Patients with a PET score of 3 have excellent outcomes with ABVD and radiotherapy alone
Some patients with early stage HL and PET score 4 may be adequately treated with ABVD and radiotherapy and not require treatment escalation
Future trials should consider treatment escalation based on PET score 5 for early stage HL
Will also state that Deauville 3s do not need treatment escalation
Will state verbally that although there were only small numbers of patients with score 4, our data suggest that some patients with early stage HL and PET score 4 may be adequately treated with ABVD and IFRT and not require treatment escalation such as an H10 approach (have clarified in the slide before these are patients without B symptoms and no med bulk)

21. Contact: ctc.RADAR@ucl.ac.uk
*in some circumstances pts in CMR may receive ISRT if intention declared prior to randomisation

22. Acknowledgements
Patients and their families All RAPID investigators Core Lab at St Thomas’ and participating PET Centres CRUK and UCL Cancer Trials Centre Funders