1. Post-operative therapy for G2 and G3 Gliomas
Jilly Maclean

2. Overview Background to G2/ G3 glioma non-surgical treatment
Recent updates RCT: RT+PCV G2 and G3 glioma
First line chemo alone?
TMZ instead of PCV?
Guidelines and caveats

3. Background
Maximal safe surgical resection as first treatment largely accepted (initial observation may be reasonable)
Most fit patients with G2 and G3 gliomas will require RT, PCV and TMZ at one point in the course of their disease
The question is timing

4. Previous post surgical treatment algorithm …........It used to be simple:
Low risk* G2 glioma
Observation
High risk G2 glioma
RT alone 50-54Gy
G3 glioma
RT alone 55-60Gy

5. “High risk” LGG previously defined as:
3 or more of the following:
>40yrs
Astrocytoma
>6cm max tumour diameter
Tumour crossing midline
Neurological deficit pre-op
According to multivariate analysis of EORTC studies

6. Recent Updates of RCT RT + PCV
G2 – RTOG 9802 (abstract only)
G3 – RTOG 9402 and EORTC 26951

7. RT + Adjuvant PCV for Grade 2 Gliomas RTOG 9802
251 pts randomised high-risk G2 (OD, OA and A)
(high risk is >40yrs or residual disease post-op)
RT (54Gy) +PCV x 6 (?how many cycles they got) v RT alone
2014 update ASCO (not published): OS 13.3yr v 7.8 yr; PFS 10.4yr v 4yr (SS)
51/15% G3/4 tox v 8/3%
Post-hoc subtype analysis: benefit in all subtypes but best for OD and OA (SS), A↑OS but NS (26% of pts)
No QoL data
No molecular data

8. RT + Adjuvant PCV – G3 ODs/ OAs (not AAs) RTOG 9402 – update 2013
291 AOD or AOA
PCV+RT v RT
PCV x 4 (8 week schedule)
11.3yrs median f/up
Median OS:
1p19q CD: med OS 14.7yr v 7.3yr (SS)
1p19q non CD: med OS 2.6 v 2.7yr (non SS)

9. RTOG 9402 – other genetic indicators
Subset post-hoc analysis:
IDH-mutant, nonCD tumours median OS 5.5 v 3.3 years (SS)
IDH wild-type tumors had poor prognosis and did not benefit from PCV: median OS 1.3 v 1.8 years (NSS)

10. EORTC 26951 – update 2013 G3 OD or mixed OA (at least 25% OD)
RT+PCV v RT
Median 3 cycles of PCV received (aim was 6)
Median 140/12 f/up
Median OS 42.3 v 30.6 months overall (NSS)
1p19q CD (n=80): OS not reached v 112/12; PFS 157/12 v 50/12; (SS)
Non CD1p19q: OS 25 v 21 months (p 0.185); PFS 15/12 v 9/12 (SS)
Post-hoc analysis shows benefit for IDH mutant also (but many 1p19qCD)
QoL: more short term nausea, reduced appetite and drowsiness with PCV but no long term effects

11. EORTC 26951 – other genetic indicators
Subset post hoc analysis
115 pts: MGMT methylation and CpG island hypermethylated phenotype (CIMP) also predictive of a benefit from PCV
MGMT meth patients med OS 8.65 v 1.98yrs (RT+PCV v RT alone)
CIMP tumors med OS 9.51 v 3.27 years (RT+PCV v RT alone)
178 pts: IDH1 mutation median OS 8.4 v 1.4 years
Firm conclusions can’t be drawn due to small subsets and cant ascertain independence from 1p19q status

12. Ongoing Studies CODEL: CD1p19q G3:RT+PCV v RT+TMZ
CATNON (BR 14): non CD 1p19q G3s: RT alone v RT+ concurrent TMZ v RT + adj TMZ v RT with conc and adj TMZ

13. Summary of RT + PCV studies
G2 - 1 RCT – strong survival benefit for OD, OA, prob not for A but unable to conclude this with certainty due to limited numbers. No molecular data.
This study does not advocate early post-op treatment of low risk LGG (not studied)
G3 – 2 RCTs – strong survival benefit for 1p19qCD tumours, maybe benefit for non-CD tumours with MGMT methylation +/or CpG island methylation +/or IDH-1 mutation, no SS survival benefit for nonCD tumours with none of those molecular features

14. 1st line chemo alone?
3 randomised studies - overall data immature, none compare to combined therapy

15. EORTC 22033 (BR13) –interim results
High risk LGG (>40 or PD) stratified for 1p status
Upfront RT alone v TMZ alone (477 pts)
2013 ASCO abstract:
Intact 1p: PFS RT = better than TMZ 41 v 30mo (p=0.06)
OS: no difference (few events)
Del 1p: PFS RT = TMZ 58 v 55 mo (p=0.95)
OS HR 0.47 (SS) suggests TMZ possibly better than RT (caution!)
Note that these outcomes are considerably less good than (RT + PCV) study for CD tumours
2015 ASCO abstract: IDH mutant/ non CD tumours shorter PFS after TMZ than RT; CD tumours no difference (but few events)

16. NOA-4 (2009) German RCT Anaplastic glioma (G3)
1st line RT v PCV (4 cycles) v TMZ (8 cycles)(2:1:1) – no combined tx arms
Max f/up 54/12 and < 50% progressed
No SS difference in median TTF (43 v 44 months) for AA, AOA and AO (not powered for to compare between types)
No SS difference in median TTF PCV v TMZ (but not powered for this comparison)
Methylation MGMT promoter, CD 1p/19q and mutation IDH1 = better prognosis regardless of tx

17. 2015 ASCO abstract update – 11.8 yr f/up – “no difference” between upfront chemo or RT in any of the molecular subgroups of glioma
TTF 4.6 v 4.4 yr (RT v chemo); OS 8 v 6.5yr – No sig values given

18. CODEL study – interim analysis
G3 anaplastic glioma 1p19q codeleted
Initially RT v (RT + TMZ) v TMZ
Revised to (RT+PCV) v (RT + TMZ) v TMZ
Interim analysis SNO 2015 abstract
36 pts, median f/up 3.4yrs: TMZ alone worse
PD in 50% of TMZ alone arm (6/12) v 8% RT arms (2/24) p=0.002
Death from PD 25% TMZ alone (3/12) v 4% RT arms (1/24)
TMZ alone arm closed: now (RT + PCV) v (RT + TMZ)
Now includes G2 CD gliomas

19. Summary of Chemo Alone
Initial chemo alone may be as good as RT alone but unlikely to equal to RT+chemo
Follow-up, detail and molecular data lacking

20. Adjuvant TMZ (Stupp) instead of PCV?

21. Summary of PCV v TMZ studies
No reported direct comparison studies for G2 or G3 gliomas (NOA-4 not powered)
Extrapolation from ”high grade” studies suggest TMZ may = PCV
CODEL should answer for CD G3 and CATNON will be compared to RT+PCV studies – but will take years
Hard to get away from the impressive results of RT+PCV studies for G2 and G3 AOD/ AOA
Possibly reasonable for G3 A but no evidence

22. Overall Summary Currently:
G2: RCT data for RT+PCV versus RT alone but ?subtypes (abstract only)
G2: new data has no implications for timing of tx in low risk pts
G3: Strong RCT data shows RT+PCV better than RT alone for 1p19q CD, but other genetic factors may influence
G3: AA were not included
No randomised data to show 1st line chemo can delay RT regardless of genetics (BR13 mature results awaited)

23. Guidelines for treatment

24. NCCN guidelines for G2 glioma

25. NCCN guidelines for G3 glioma

26. Proposed treatment algorithm post maximal safe resection
*Low risk: <40 years with no residual disease on post-op MRI
Low risk* G2 glioma
Observation
High risk G2 glioma
RT + PCV
(caveat 1 and 2)

27. G3 non-CD tumours with favourable genetics* RT + PCV (caveat 2) or
G3 1p19qCD glioma
RT + PCV (caveat 2)
G3 non-CD tumours with favourable genetics*
RT + PCV (caveat 2) or
RT alone
(depending on genetic and patient factors)
G3 1p19q non-CD gliomas with no favourable genetics and AAs
RT alone (caveat 3)
*Favourable genetics:
1p OR 19q deletion
MGMT promotor methylation
CpG island methylation
IDH-1 mutation
RT dose: (50-54Gy for G2; 60Gy for G3)

28. Caveats
Where a G2 glioma has no favourable genetics, RT alone may be appropriate for select patients
RT + TMZ (as Stupp) may be considered in patients unable to have PCV
It is reasonable to consider treating select AAs and G3 OA/ OD with no favourable genetics as GBM
There is no RCT data to support chemo alone as initial treatment (versus RT+PCV) but it may be appropriate in select patients appropriately counselled

31. Issues Interobserver variability in grading and biopsy sampling
How reliable is the genetic testing (different methods used)?
Pretty clear that chemo is beneficial for 1p19q CD tumours but need more data on other genetic profiles
Defining high risk/ low risk G2s
AA – should these just be treated as GBMs? – paradox of AA RT only then GBM RT+chemo
Does the old mantra of delaying RT for “low-risk” G2 gliomas still hold?
Is there a “low-risk” G3 group where tx can be delayed?

32. Upfront or delayed treatment G2
EORTC (accrual 1986 – 1997), med f/up 93/12
311 pts randomised upfront RT v RT at progression
MPFS 3.7yr v 5.4yr
MOS 7.9yr v 7.1yr

33. G3 Historical studies showing benefit for RT for “high grade gliomas”
No study of delayed treatment

34. Summary of chemo alone studies
2 completed RCT chemo only v RT only (1 in G2, 1 in G3)
Follow-up largely immature or in abstract only, molecular evaluation lacking
BR13 (for G2 CD tumours) and NOA-4 (G3 all) suggest that upfront chemo alone may = upfront RT alone BUT results not as good as RT+chemo studies for CD tumours
CODEL closed chemo only arm after interim analysis indicated pts do less well – we will not get a comparison of chemo alone v RT+chemo in CD G3 tumours
Difficult to advocate for chemo alone as standard initial treatment

35. The Case for TMZ TMZ less toxic
Vincristine doesn’t cross BBB (but probably disrupted)
Theoretically concurrent + adjuvant chemo may be more effective than just adjuvant (not proven in gliomas)
NOA-04: no difference in TTF with PCV v TMZ in G3s but not powered for this comparison.
MRC study 2001: HGG RT v (RT + PCV) – no difference in OS – compare with TMZ Stupp study
BR12 TMZ v PCV for recurrent HGG (103pts G3) - no survival advantage for PCV.
Retrospective analysis of AA treated with TMZ or PCV post RT – no difference in outcomes, TMZ better tolerated
Standard in North America to treat G3 gliomas the same as GBM (Stupp protocol with TMZ) – pragmatic but no randomised data to support it

36. The Case for PCV
3 RCTs show large survival advantage for (RT + PCV) v RT
Non-comparator chemo-alone studies for PD tend to show higher response rates to PCV than TMZ and longer duration of response (although lots of caution needed in cross-comparison)
Multi-chemo often more effective than single agent for other tumour types
CODEL study should answer for 1p19q deleted patients (RT+PCV v RT+TMZ)