22. Pseudo-Progression (PsP) and Radiation Necrosis
(RN) vs Tumor/Progressive Disease (PD)
• Histopathology gold standard:
- often not obtained due to morbidity of
surgery.
Pathology not always straight forward:
-Tumor can have necrosis
-Mixed pathology seen [sampling error]
• Often studies presume PsP/RN, PD based on
f/u imaging and clinical
Many studies have been loose in distinguishing
between these entities.
23. Pathology: PsP, RN
• Despite proposed differences in
pathophysiology between PsP and RN:
PsP & RN have many overlapping pathologic
features, and seem to have as such many
similar imaging features, potentially
exploitable to distinguish from PD
24. Limitations of the literature
• Most studies of treatment effects are high
grade glial (HGG) tumors in adults.
• [Small numbers, not prospective]
• Children have different spectrum of
malignant tumors with differing biologic
behavior than adult high grade glioma.
• Reaction to radiation (animal studies suggest
endothelial/oligodendrocyte/microglia
differences between children and adults)
25. ATRT: Gross total resection;
Proton Beam: Radiation Necrosis
Initial tumor 1 year post PBT
26. Incidence of Radiation Necrosis
• Photon radiation necrosis rates in adults:
5% 72Gy, 10% 90Gy [Lawrence 2010]
• Few studies in children:
• Photons;
-Plimpton 2015: n=101, 54-59.4Gy;
5% ‘rad necrosis’ [mean 1.2m post XRT]
-Spreafico 2008 n=49, chemo & PBSC rescue;
HHG 24%, PNET 75% ‘abnormal imaging’
(8m [2-39m) post XRT]
• Proton Beam (PBT):
-Kralik 2015: n=60; 31% rad necrosis
-Sabin 2013: n=17 (young); 47% rad necrosis
44. Diffusion weighted imaging (ADCratio)
• ADC ratio = ADC lesion /ADC normal contralateral white matter
• Hein 2004: ADC ratio better then ADC lesion
-ADC ratio PD 1.43±.11; Tx effect 1.82±.07 p= <.001
Treatment effect > 1.62 > Progressive Disease
45. RN/PsP vs PD: ADClesion , ADCratio
• Matsusue 2010: (n=10/15 tumor recurrence; mixture
LGG & HGG;); 3/15 MRI ≤6 months post XRT;
-ROI lowest ADClesion/ ADC normal contralateral white matter
-ADCratio > 1.3; sens 9/10, specific 4/5 PsP,RN vs PD.
• Prager 2015: HGG, n=68; 10/68 TX effects;
[8/10=PsP, (<6months post XRT)]
-ADC lesion threshold <0.00149; sens 73.7%, spec 70%= PD
-ADC ratio difference not significant
46. DWI/ADC: Pseudoprogression
early diffusion restriction?
• Young 2011; n=93, adult HGG with worsening
lesion on MRI 2-4 wks post XRT: PsP= 30/93
Qualitative assessment: reported diffusion
restriction: 16/30 of those with PsP
PsP pathophysiology associated with diffusion
restriction?
47. Post. Fossa ATRT 15 months post PBT:
Radiation necrosis (acute?); low ADC
Initially ischemia?
48. DWI/ADC: summary
• High ADClesion (>.0015mm/s2): probably RN/PsP
• Low ADClesion (<.0011mm/s2): probably PD
• ADCratio ? PD < 1.3-1.6 < RN/PsP
• Limitations:
• Low ADC can be seen in RN; (acute/ ischemic
phase?)
• Low ADC may be more common early with PsP
(awaiting further data )
49. Distinguishing treatment effects/
recurrent tumor
• Conventional MRI:
-Timing/relationship to radiation-chemotherapy
-Tumor type: pattern
HGG-infiltrative spread likely vs CSF spread with
many ped tumors
• Advanced MR:
-Diffusion Weighted Imaging
-Perfusion Weighted Imaging
-MRS
50. Perfusion (PWI)
Theory
• Detect microvasculature, angiogenesis
• Tumor: Increased CBV
• RN/PsP: Decreased CBV
• Most studies use DSC; Parametric maps: CBV
• ‘Normalized’ rCBVratio= CBVlesion/CBV nl contralateral WM
• Technique/analysis more difficult, potential issues
with heterogeneity, capillary leak.
51. 7 y.o. HGG 2 years post PBT
Recurrence: rCBV
rCBV= 4
52. PF ATRT: 4m post PBT
Radiation Necrosis (Brainstem)
53. Treatment effect versus progressive
disease: DSC
Matsusue 2010: n=15 [adult HGG+LGG]
- rCBVratio threshold 2.1: PPV 9/10, NPV 4/5
Prager 2015: n=68 [adult HGG]
- rCBVratio ≥ 1.27: sens 86.5%, specific 83.3% PD
- Sub analysis predicting PsP (<6 m post XRT)
- rCBVratio < 1.07: sens 100%, specific 75% PsP
54. Progressive disease versus PsP:
PWI/DSC
• Wan et al 2017 meta analysis;
• 11 studies with 20+ subjects between 2011-
2016 employing DSC
• AUC of the SROC = 0.8899
• ‘good but not excellent diagnostic accuracy’
55. Distinguishing treatment effects/
recurrent tumor
• Conventional MRI:
-Timing/relationship to radiation-chemotherapy
-Tumor type: pattern
HGG-infiltrative spread likely vs CSF spread with
many ped tumors
• Advanced MR:
-Diffusion Weighted Imaging
-Perfusion Weighted Imaging
-MRS
56. MRS: Theory
• Tumor/PD:
Cho [ NAA]
• Treatment effects/Radiation necrosis (RN):
Cho, NAA [lipid, lac].
• Typically use metabolic ratios; Cr denominator.
• Limitations:
volume averaging [SVS vs MVS]
variable TE employed; peak height vs peak area
60. MRS: summary
• Cho/Cr: <1.3 = RN
• Cho/Cr > 1.6 -1.8 = PD.
• Limitations:
Volume averaging [multi-voxel preferred]
Overlap:
Intra-tumoral necrosis
Occasionally Cho elevated in RN
Cho elevation possible in PsP
61. Tumor MGMT promoter
methylation status:
• MGMT (O6 methylguanine DNA methyl
transferase) is responsible for repairs of DNA
errors during replication.
• Methylated MGMT promoter =
decreased MGMT production [‘turned off’]
decreased DNA damage repair (--radiation--)
more tumor cell death
66. Conclusions: ‘Team Sport’
• Prospective diagnosis of radiation necrosis /
pseudoprogression in children, as in adults
‘remains challenging’
• Imaging interpretation most useful including the
clinical context; collaboration with a
multidisciplinary team:
- XRT (type/dose/margins/timing)
- Chemo-radiotherapy regimen
- MGMT promoter methylation status (HGG)