Studies
25
Imaging studies
2016 – 2026
Lesions
23
Lesions catalogued
7 brain · 5 spine · 6 ventricular · 5 other
PF-1
+29%
PF-1 baseline change
10.3 to 13.3 mm over 37 days on CAR-T
ST-1
+46%
ST-1 single-interval change
7.07 to 10.3 mm in 22 days
CSI
5+
Lesions with reduced enhancement on CSI
First non-progression scan since 12 Apr
01
Imaging timeline
Select a marker for scan detail
MRI Brain
MRI Spine
PET / Functional
Surgery / SRS
Systemic Tx
Stable / baseline
Progression
Response / stable
Therapy / surgery
×
02
Quantitative trajectories
Lesions with caliper measurements at 2 or more timepoints
Lesion size over time
Four of the 23 catalogued lesions have caliper measurements documented at 2 or more timepoints: PF-1 (R posterior fossa), ST-1 (bilateral thalamic / periventricular), R frontal horn ependymal nodule, and R parietal parafalcine mass. Other lesions (most ventricular, all spinal LMD, R parietal precuneus) are tracked qualitatively only.
Interval change
Percentage change in maximum dimension between consecutive timepoints. Positive values indicate progression (lesion enlargement); negative values indicate response (lesion shrinkage).
Lesion count by region
23 catalogued lesions by anatomical compartment
PET tracer uptake by lesion
Each bar is a different lesion measured on a single PET study. Tracers and institutions vary, so values are not directly comparable across bars — use as orientation, not for response assessment.
Cumulative new lesions
Count of distinct new lesions documented over time
03
Anatomical map
Select a region or lesion marker
Anatomical distribution
Lesions span every major anatomical compartment
Primary site: right frontal lobe (resected at MGH Feb 2016 and at UCLA Oct 2024; irradiated 60 Gy 2016 and 30 Gy boost 2024; craniospinal proton irradiation Apr–May 2026). Disease has subsequently been documented in: supratentorial parenchyma (R cingulate / corpus callosum, R parietal parafalcine, R parietal precuneus, L cingulate, R external capsule), lateral and fourth ventricles, cerebral aqueduct, posterior fossa, R brachium pontis, and the entire craniospinal axis (C, T, L). On craniospinal irradiation the dominant C5–C6 spinal nodule shows reduced enhancement intensity and there are no new spinal lesions on the 30 Apr 2026 scan.
Brain parenchymal: 7 lesions
Ventricular / aqueductal: 6 lesions
Spine: 5 lesions
Latest assessment: Mixed response
04
Lesion catalogue
23 distinct lesions · select a row for detail
| ID | Region | Specific location | Type | First seen | Latest | Status |
|---|
05
Functional imaging
PET · perfusion · spectroscopy
FET-PET tumour-to-background ratio across studies
TBR above approximately 2 suggests viable tumour; above approximately 4 suggests high-grade tumour. Each point on this chart is a separate FET-PET study and refers to a different lesion (the May 2025 reading is an aggregate of 3 lesions of concern; November 2025 is the R external capsule lesion specifically; March 2026 is the posterior cingulate at Huashan). The R external capsule TBR of around 5 in November 2025, despite stable lesion size, was the signal that informed Stanford CyberKnife SRS.
PET tracers used
Three tracers acquired across UCSF, Suzhou, and Huashan. Each measures a different aspect of tumour biology.
06
Disease phases
12 named periods of the imaging course
07
Technical caveats
Cross-vendor comparison limitations
Slice thickness changed mid-series. The Suzhou MRI scans changed from 6 mm to 5 mm slice thickness between 11 March and 2 April. Linear measurements of small lesions are less affected than volume calculations, but the 9 percent change in PF-1 reported on 2 April may be partly explained by this technical difference rather than true growth.
Imaging protocol differed between the two UMiami brain scans. On 1 May the post-contrast spin-echo sequence was performed before the MP RAGE — the reverse of the 12 April sequence order. The apparent reduction in enhancement on 1 May may be partly explained by this timing difference.
The 12 April MP RAGE was degraded by motion. Comparison between the two UMiami brain scans was therefore based on the spin-echo sequence rather than the higher-resolution MP RAGE, which limits the resolution of subtle changes.
Both UMiami scans were acquired during active radiation. The 30 April and 1 May scans were performed in the middle of fractionated craniospinal proton irradiation. A reduction in enhancement during active radiation can reflect a transient radiation effect rather than a durable treatment response.
UMiami findings are based on the radiologist's report only. The 30 April and 1 May findings come from the written radiology reports, which describe the lesions in qualitative terms (stable, decreased, similar, mildly increased) without numerical measurements. Direct comparison to the China measurement series would benefit from access to the UMiami scan files themselves so that calipers can be applied to the same lesions.
Magnet strength varied between sites. The 9 February UMiami study was performed on a 1.5 Tesla scanner; all other studies in this tracker were at 3 Tesla. Smaller leptomeningeal nodules can be harder to detect at lower field strength, so a small new finding on a 3 T study may not be a true new lesion.
Contrast agent varied between sites. UMiami used gadopiclenol (6 mL); UCSF used gadoterate (Dotarem, 12 mL); the Suzhou agent was not specified. Comparing how brightly a lesion enhances across studies from different sites should be done with caution.
08
Data sources
Hide sources
▾
All references in this dashboard are derived from the source reports listed above. Filenames have been anonymised; original reports remain on the patient's institutional records. Click any institution or document for the full attribution.