- Split View
-
Views
-
Cite
Cite
Bogdana Suchorska, Armin Giese, Annamaria Biczok, Marcus Unterrainer, Michael Weller, Mark Drexler, Peter Bartenstein, Ulrich Schüller, Jörg-Christian Tonn, Nathalie L Albert, Identification of time-to-peak on dynamic 18F-FET-PET as a prognostic marker specifically in IDH1/2 mutant diffuse astrocytoma, Neuro-Oncology, Volume 20, Issue 2, February 2018, Pages 279–288, https://doi.org/10.1093/neuonc/nox153
- Share Icon Share
Abstract
Stratification of glioma according to isocitrate dehydrogenase 1/2 (IDH1/2) mutation and 1p/19q codeletion status has gained major importance in the new World Health Organization (WHO) classification. Parameters derived from uptake dynamics of 18F-fluoro-ethyl-tyrosine PET (18F-FET-PET) such as minimal time-to-peak (TTPmin) allow discrimination between different prognostic glioma subgroups, too. The present study is aimed at exploring whether TTPmin analysis provides prognostic information beyond the WHO classification.
Three hundred patients with newly diagnosed WHO 2007 grades II–IV gliomas with 18F-FET-PET imaging at diagnosis were grouped into 4 subgroups (IDH1/2 mut–1p/19q codel; IDH1/2 mut–1p/19q non-codel; IDH1/2 wildtype WHO grade II and III tumors; and glioblastoma). Clinical and imaging factors such as age, Karnofsky performance score, treatment, TTPmin, and maximal tumor-to-brain ratio (TBRmax) were analyzed with regard to progression-free and overall survival (PFS and OS) via univariate and multivariate regression analysis.
PFS and OS were longest in the IDH1/2 mut–1p/19q codel subgroup, followed by IDH1/2 mut–1p/19q non-codel, IDH1/2 wildtype, and GBM (P < 0.001). Further, outcome stratified by TTPmin with a cutoff of 17.5 minutes revealed significantly longer PFS and OS in patients with TTPmin >17.5 minutes (P < 0.001 for PFS and OS). Lower TBRmax values or the absence of 18F-FET uptake was also associated with favorable outcome in the entire group. In the subgroup analyses, longer median TTPmin was associated with improved outcome specifically in the IDH1/2 mut–1p/19q non-codel group.
18F-FET-PET–derived dynamic analysis defines prognostically distinct subgroups of IDH1/2 mutant–1p/19q non-codel gliomas which cannot be distinguished as yet by molecular marker analysis.
Tailoring treatment options in glioma patients according to an individual risk profile is gaining increasing importance in the field of neuro-oncology.1 Recent analyses of large cohort studies have uncovered a dominant association of molecular markers with clinical outcome in glioma patients.2–5 As a consequence, the 2016 revision of the World Health Organization (WHO) classification of tumors of the central nervous system accounts for molecular markers for subclassification of gliomas.6 Isocitrate dehydrogenase 1 and 2 (IDH1/2) mutant tumors have a different biology and better outcome than IDH1/2 wildtype (wt) tumors, and IDH1/2 mutant tumors are further subdivided into 1p/19q codeleted (associated with oligodendroglial morphology and better outcome) and non-codeleted tumors (associated with astrocytic morphology and intermediate outcome).
Recent advances in molecular imaging via O-(2-18F-fluorethyl)-L-tyrosine PET (18F-FET-PET) have led to the establishment of several non-invasive, imaging-derived prognostic factors such as biological tumor volume and dynamic tracer uptake represented by time activity curves (TACs) and time-to-peak (TTP) evaluation.7–9 In particular the dynamic evaluation of uptake including TTP analysis is highly associated with prognosis across WHO grades II–IV gliomas.8,9 Aim of this study was to explore whether imaging-derived markers such as TTP still add to the profoundly improved prognostic classification realized within the framework of the updated WHO classification.6
Materials and Methods
Patient Evaluation
Patients with a supratentorial grade II–IV glioma (WHO 2007) diagnosed between 2004 and 2014 who had undergone 18F-FET-PET with static and dynamic analysis prior to histopathological diagnosis were retrospectively identified. The study was approved by the institutional review board (approval number: 604-16) and all subjects signed written informed consent as part of the clinical routine. PFS was measured from the date of surgical procedure to the first event of clinical deterioration—that is, new neurological symptoms, worsening as indicated by KPS, an increase in steroid medication, or tumor growth on conventional MRI according to the modified Response Assessment in Neuro-Oncology criteria.10 Overall survival (OS) was correspondingly calculated from date of surgery to date of death. Date of last follow-up was December 2016.
18F-FET-PET and MR Imaging
Dynamic 18F-FET-PET scans were acquired with an ECAT EXACT HR+ scanner (Siemens Healthcare) according to standard protocols after slow intravenous bolus injection of 180 MBq 18F-FET.8 Dynamic emission recording in 3D mode consisted of 16 frames (7 × 10 s, 3 × 30 s, 1 × 2 min, 3 × 5 min, and 2 × 10 min) and was conducted until 40 minutes post injection. For further evaluation, images were transferred to a HERMES workstation (Hermes Medical Solutions).
In the semiquantitative analysis, the maximal tumoral 18F-FET uptake corrected for mean background activity in the contralateral hemisphere (maximal tumor-to-brain ratio, TBRmax) was evaluated. Tumors were termed 18F-FET “positive” when TBRmax was above 1.6.11,12
For the dynamic analysis, which was performed using HERMES PET Display Dynamic, 90% isocontour regions of interest were semiautomatically drawn on each individual slice throughout the tumor in the 10–30 minutes summation images and afterward applied to the dynamic PET data in order to extract the TAC. For each extracted TAC within the tumor, the peak of the curve was identified and the time of peak uptake was noted and set as TTP. This procedure was repeated for all slices throughout the tumor. As explained in detail previously,8 the shortest TTP being present in at least 2 adjacent slices was defined as minimal TTP (TTPmin).
MR imaging was performed prior to tissue sampling according to standard protocols and included acquisition of axial T2-weighted sequences as well as 3D T1-weighted sequences before and after administration of intravenous contrast agent (0.1 mmol/kg gadobenate dimeglumine, MultiHance, Bracco Imaging). Progression was diagnosed whenever (i) any new contrast enhancement (CE) was noted in a previously non-enhancing tumor or (ii) T2 diameter of the tumor enlarged above 25%. Contrast enhancement (yes/no) as well as tumor size and volume (both CE and T2-based size/volume) were assessed on initial MRIs prior to surgical procedure using Brainlab software (iPlan, Brainlab).
Surgical Procedure
Biopsy or resection was performed according to interdisciplinary tumor board recommendations and patient’s preference. The stereotactic biopsy procedure at our institution has been described previously. Briefly, multiple tumor tissue specimens were obtained from the area of the highest 18F-FET uptake.13,14 The microsurgical resection procedure involved MRI- and 18F-FET PET–based neuronavigation (Brainlab).
Histopathology and Molecular Genetic Markers
Histological diagnosis was performed by experienced neuropathologists (U.S., A.G.) according to the WHO 2007 classification blinded for PET and MRI findings.15 Determination of O6-DNA methylguanine-methyltransferase (MGMT) promoter methylation was performed using methylation-specific PCR.16 Codeletion of 1p/19q was analyzed according to standard protocols17 with the following microsatellite markers: D1S548, D1S1184, D1S1608, D1S1592, D1S1161, D19S601, D19S559, D1S433, D19S718, and D19S431. Determination of IDH1 mutation was performed using pyrosequencing of an 88-bp-long fragment of the IDH1 gene including the mutation hotspot at codon 132, while for IDH2 mutations, pyrosequencing was performed on an 83-bp-long fragment of the IDH2 gene including the mutation hotspot at codon 172.
Patients’ diseases were categorized as either glioblastoma (GBM) or glioma WHO grades II and III, the latter 2 groups differentiated by mutational status of IDH1/2 and codeletion of 1p/19q.
Statistical Analysis
SPSS for Windows v21.0 was used for statistical calculations. Progression-free survival (PFS) and OS were analyzed with the Kaplan–Meier method; when median PFS or OS times were not reached, mean values were given and used for analysis. The distribution of patient- and tumor-related variables was analyzed by chi-squared statistics (for categorical variables) and the Mann–Whitney U-test (for continuously scaled variables). The median was used as threshold for dichotomization of parameters. For univariate prognostic analyses, all parameters were evaluated using Cox regression. Covariates significant in one-variable models were then evaluated in multivariate analyses using a stepwise backward exclusion model. In case of an intercorrelation of most relevant covariates, alternative models were tested and compared by computing the maximized likelihoods. A 2-tailed P-value <0.05 was considered significant.
Results
Patient Characteristics
Three hundred patients (median age, 47.6 y [range, 8.1–84.0]; 166 males) with primary diagnosis of a glioma and 18F-FET PET scan at diagnosis were evaluated. Clinical data of all patients are listed in Table 1. Median follow-up time was 67.5 months for the survivors. During this time period, 178 patients experienced tumor progression and 144 patients died. Median PFS was 23.6 months (95% CI: 19.9–27.0) and median OS was 59.3 months (95% CI: 29.9–88.6). An IDH1/2 mutation was found in 142 patients, a 1p/19q codeletion in 60 patients. In 22 patients, information on the molecular tumor profile was not available. Outcome by histological grade versus diagnosis of molecular markers is summarized in Supplementary Table S1. Kaplan–Meier curves for OS by age, gender, WHO grade, IDH1/2 mutation, and 1p/19q status are summarized in Supplementary Figure S1. Detailed data on each patient are given in Supplementary Table S2.
Factor . | Number . |
---|---|
Patients | 300 |
Gender (m, f) | 166/134 |
Age, y (median, range) | 47.6 (8.1–84.0) |
KPS (median, range) | 90 (60–100) |
Surgical procedure | |
Biopsy | 238 |
Surgery | 62 |
WHO grade | |
II | 121 |
III | 106 |
IV | 73 |
Tumor location | |
Frontal | 87 |
Temporal | 129 |
Parietal | 42 |
Occipital | 9 |
Midline/basal ganglia/corpus callosum | 33 |
Eloquent brain area involved | |
Yes | 170 |
No | 130 |
Contrast enhancement | |
Yes | 172 |
No | 128 |
T2 volume (mean, median), mL | 71.0; 49.0 |
T2 diameter (mean, median), cm | 5.9; 5.5 |
CE diameter (mean, median), cm | 1.5; 0.8 |
Molecular markers | |
IDH1/2 mut–1p/19q codel | 58 |
IDH1/2 mut–1p/19q non-codel | 79 |
IDH1/2 wt (WHO II+III) | 76 |
GBM | 73 |
First-line therapy in addition to surgery/biopsy | |
Wait-and-see | 67 |
Chemotherapy | 87 |
Radiotherapy | 46 |
Radiochemotherapy | 100 |
Factor . | Number . |
---|---|
Patients | 300 |
Gender (m, f) | 166/134 |
Age, y (median, range) | 47.6 (8.1–84.0) |
KPS (median, range) | 90 (60–100) |
Surgical procedure | |
Biopsy | 238 |
Surgery | 62 |
WHO grade | |
II | 121 |
III | 106 |
IV | 73 |
Tumor location | |
Frontal | 87 |
Temporal | 129 |
Parietal | 42 |
Occipital | 9 |
Midline/basal ganglia/corpus callosum | 33 |
Eloquent brain area involved | |
Yes | 170 |
No | 130 |
Contrast enhancement | |
Yes | 172 |
No | 128 |
T2 volume (mean, median), mL | 71.0; 49.0 |
T2 diameter (mean, median), cm | 5.9; 5.5 |
CE diameter (mean, median), cm | 1.5; 0.8 |
Molecular markers | |
IDH1/2 mut–1p/19q codel | 58 |
IDH1/2 mut–1p/19q non-codel | 79 |
IDH1/2 wt (WHO II+III) | 76 |
GBM | 73 |
First-line therapy in addition to surgery/biopsy | |
Wait-and-see | 67 |
Chemotherapy | 87 |
Radiotherapy | 46 |
Radiochemotherapy | 100 |
Factor . | Number . |
---|---|
Patients | 300 |
Gender (m, f) | 166/134 |
Age, y (median, range) | 47.6 (8.1–84.0) |
KPS (median, range) | 90 (60–100) |
Surgical procedure | |
Biopsy | 238 |
Surgery | 62 |
WHO grade | |
II | 121 |
III | 106 |
IV | 73 |
Tumor location | |
Frontal | 87 |
Temporal | 129 |
Parietal | 42 |
Occipital | 9 |
Midline/basal ganglia/corpus callosum | 33 |
Eloquent brain area involved | |
Yes | 170 |
No | 130 |
Contrast enhancement | |
Yes | 172 |
No | 128 |
T2 volume (mean, median), mL | 71.0; 49.0 |
T2 diameter (mean, median), cm | 5.9; 5.5 |
CE diameter (mean, median), cm | 1.5; 0.8 |
Molecular markers | |
IDH1/2 mut–1p/19q codel | 58 |
IDH1/2 mut–1p/19q non-codel | 79 |
IDH1/2 wt (WHO II+III) | 76 |
GBM | 73 |
First-line therapy in addition to surgery/biopsy | |
Wait-and-see | 67 |
Chemotherapy | 87 |
Radiotherapy | 46 |
Radiochemotherapy | 100 |
Factor . | Number . |
---|---|
Patients | 300 |
Gender (m, f) | 166/134 |
Age, y (median, range) | 47.6 (8.1–84.0) |
KPS (median, range) | 90 (60–100) |
Surgical procedure | |
Biopsy | 238 |
Surgery | 62 |
WHO grade | |
II | 121 |
III | 106 |
IV | 73 |
Tumor location | |
Frontal | 87 |
Temporal | 129 |
Parietal | 42 |
Occipital | 9 |
Midline/basal ganglia/corpus callosum | 33 |
Eloquent brain area involved | |
Yes | 170 |
No | 130 |
Contrast enhancement | |
Yes | 172 |
No | 128 |
T2 volume (mean, median), mL | 71.0; 49.0 |
T2 diameter (mean, median), cm | 5.9; 5.5 |
CE diameter (mean, median), cm | 1.5; 0.8 |
Molecular markers | |
IDH1/2 mut–1p/19q codel | 58 |
IDH1/2 mut–1p/19q non-codel | 79 |
IDH1/2 wt (WHO II+III) | 76 |
GBM | 73 |
First-line therapy in addition to surgery/biopsy | |
Wait-and-see | 67 |
Chemotherapy | 87 |
Radiotherapy | 46 |
Radiochemotherapy | 100 |
18F-FET-PET and MRI Correlates of Outcome
Initial median tumor size as assessed by T2-MRI was 49 mL; there was no significant difference in tumor size within the 4 tumor groups (P = 0.33). Contrast enhancement was observed in 172/300 (57%) of tumors; all GBM patients presented with CE, while presence of CE was equally distributed among the 3 remaining molecular groups (P = 0.25). In 18F-FET-PET, 255 of 300 (85%) tumors had an elevated 18F-FET uptake and thus were classified as 18F-FET positive. The rates of 18F-FET− lesions per molecular subgroup were: 3/58 (5%) for IDH1/2 mut–1p/19q codel tumors, 28/79 (35%) for IDH1/2 mut–1p/19q non-codel tumors, 9/76 (11.8%) for IDH1/2 wt tumors, and 0/73 for GBM (Table 2). In the entire group, a negative 18F-FET-PET was associated with prognosis: PFS and OS times in patients with 18F-FET− tumors were much longer compared with patients with 18F-FET+ tumors (P = 0.01 and 0.001).
Dynamic analysis of tracer uptake could be performed in only 18F-FET positive tumors. TTPmin analysis was therefore not available in the 45 18F-FET− cases and in a further 5 patients for technical reasons. Thus, the following calculations refer to a study population of 255 patients. Correlation of continuously scaled TBRmax and TTPmin with PFS or OS revealed shorter (P < 0.001) PFS and OS in patients with higher TBRmax and shorter TTPmin values. Longer TTPmin was highly correlated with the presence of IDH1/2 mutation (P < 0.001). In order to identify different prognostic subgroups, median values for TBRmax and TTPmin were calculated within the entire 18FET positive patient population as well as in subgroups defined by molecular markers. PFS and OS were compared between patient groups split by this median.
Median TBRmax was 2.6 for all patients. Median PFS and OS times were shorter in patients with TBRmax values >2.6 (15.4 mo [95% CI: 12.4–18.4] / 34.2 mo [CI: 95% 23.0–45.5]) than in patients with a TBRmax ≤2.6 (PFS, 50.1 mo; 95% CI: 30.0–70.2; P < 0.001). Mean OS was 89.4 months (95% CI: 79.9–98.9) in patients with a TBRmax ≤2.6 (Table 2).
. | Factor . | N . | PFS (mo) . | 95% CI . | Log-rank P . | OS (mo) . | 95% CI . | Log-rank P . |
---|---|---|---|---|---|---|---|---|
All (including patients without complete molecular profile) | 18F-FET negative | 45 | 50.1 (median) | 24.0–76.7 | 0.01 | 97.9 (mean)** | 83.1–112.7 | 0.001 |
18F-FET positive | 255 | 20.0 (median) | 15.9–24.3 | 46.1 (mean)** | 26.6–65.6 | |||
TBRmax ≤2.6 | 147 | 50.1 (median) | 30.0–70.2 | <0.001 | 89.4 (mean)** | 79.9–98.9 | <0.001 | |
TBRmax >2.6 | 153 | 15.4 (median) | 12.4–18.4 | 34.2 (median) | 23.0–45.5 | |||
TTPmin ≤17.5 | 162* | 14.2 (median) | 11.1–17.3 | <0.001 | 26.2 (median) | 21.6–30.8 | <0.001 | |
TTPmin >17.5 | 88* | 74.7 (mean)** | 62.7–86.7 | 116.3 (mean)** | 106.4–126.3 | |||
IDH1/2 mut–1p/19q codel (n = 58) | 18F-FET negative | 3 | 26.3 (mean)** | 18.3–34.2 | 0.60 | All cases censored | na | 0.73 |
18F-FET positive | 55 | 92.2 (mean)** | 78.3–106.2 | All cases censored | na | |||
TBRmax ≤2.7 | 29 | 87.9 (median) | 10.5–165.4 | 0.58 | All cases censored | na | 0.07 | |
TBRmax >2.7 | 29 | 95.7 (mean)** | 77.5–113.9 | All cases censored | na | |||
TTPmin ≤25 | 17* | 84.2 (mean)** | 61.0–107.5 | 0.54 | 127.5 (mean)** | 114.5–140.6 | 0.92 | |
TTPmin >25 | 38* | 98.4 (mean)** | 84.0–112.7 | 131.9 (mean)** | 122.4–141.3 | |||
IDH1/2 mut–1p/19q non-codel (n = 79) | 18F-FET negative | 28 | 53.8 (median) | 41.3–66.3 | 0.42 | 99.2 (mean)** | 83.7–115.2 | 0.65 |
18F-FET positive | 51 | 37.1 (median) | 27.5–46.6 | 98.1 (mean)** | 111.8–110.6 | |||
TBRmax ≤1.7 | 40 | 56.8 (median) | 40.8–72.8 | 0.01 | 104.2 (mean)** | 91.5–116.9 | 0.18 | |
TBRmax >1.7 | 39 | 34.9 (median) | 18.0–51.9 | 92.4 (mean)** | 76.8–107.9 | |||
TTPmin ≤25 | 31* | 32.3 (median) | 16.4–48.2 | 0.02 | 75.1 (median) | 58.1–92.0 | 0.002 | |
TTPmin >25 | 19* | 77.0 (mean)** | 56.1–97.9 | 125.6 (mean)** | 115.0–136.2 | |||
IDH1/2 wt (n = 76) | 18F-FET negative | 9 | 35.2 (median) | 15.8–33.3 | 0.09 | 40.9 (median) | 14.7–67.0 | 0.15 |
18F-FET positive | 67 | 12.0 (median) | 6.5–17.3 | 24.5 (median) | 20.1–28.8 | |||
TBRmax ≤2.5 | 41 | 14.4 (median) | 10.8–17.9 | 0.52 | 26.2 (median) | 20.7–31.6 | 0.79 | |
TBRmax >2.5 | 35 | 10.4 (median) | 1.6–19.1 | 24.0 (median) | 17.4–30.6 | |||
TTPmin ≤12.5 | 44* | 10.4 (median) | 4.9–19.9 | 0.51 | 24.0 (median) | 20.2–27.9 | 0.14 | |
TTPmin >12.5 | 23* | 11.9 (median) | 6.3–17.6 | 26.0 (median) | 15.8–36.1 | |||
GBM (n = 73) | 18F-FET negative | 0 | Not applicable | NA | NA | Not applicable | NA | NA |
18F-FET positive | 73 | 10.3 (median) | 9.1–11.5 | 14.0 (median) | 10.4–17.7 | |||
TBRmax ≤3.6 | 36 | 11.9 (median) | 7.6–16.1 | 0.18 | 15.8 (median) | 7.1–24.5 | 0.67 | |
TBRmax >3.6 | 37 | 8.7 (median) | 6.0–11.4 | 13.4 (median) | 10.4–16.3 | |||
TTPmin ≤12.5 | 49* | 9.9 (median) | 7.8–12.0 | 0.36 | 12.8 (median) | 11.2–14.3 | 0.29 | |
TTPmin >12.5 | 20* | 11.9 (median) | 9.1–14.6 | 23.1 (median) | 11.7–34.6 |
. | Factor . | N . | PFS (mo) . | 95% CI . | Log-rank P . | OS (mo) . | 95% CI . | Log-rank P . |
---|---|---|---|---|---|---|---|---|
All (including patients without complete molecular profile) | 18F-FET negative | 45 | 50.1 (median) | 24.0–76.7 | 0.01 | 97.9 (mean)** | 83.1–112.7 | 0.001 |
18F-FET positive | 255 | 20.0 (median) | 15.9–24.3 | 46.1 (mean)** | 26.6–65.6 | |||
TBRmax ≤2.6 | 147 | 50.1 (median) | 30.0–70.2 | <0.001 | 89.4 (mean)** | 79.9–98.9 | <0.001 | |
TBRmax >2.6 | 153 | 15.4 (median) | 12.4–18.4 | 34.2 (median) | 23.0–45.5 | |||
TTPmin ≤17.5 | 162* | 14.2 (median) | 11.1–17.3 | <0.001 | 26.2 (median) | 21.6–30.8 | <0.001 | |
TTPmin >17.5 | 88* | 74.7 (mean)** | 62.7–86.7 | 116.3 (mean)** | 106.4–126.3 | |||
IDH1/2 mut–1p/19q codel (n = 58) | 18F-FET negative | 3 | 26.3 (mean)** | 18.3–34.2 | 0.60 | All cases censored | na | 0.73 |
18F-FET positive | 55 | 92.2 (mean)** | 78.3–106.2 | All cases censored | na | |||
TBRmax ≤2.7 | 29 | 87.9 (median) | 10.5–165.4 | 0.58 | All cases censored | na | 0.07 | |
TBRmax >2.7 | 29 | 95.7 (mean)** | 77.5–113.9 | All cases censored | na | |||
TTPmin ≤25 | 17* | 84.2 (mean)** | 61.0–107.5 | 0.54 | 127.5 (mean)** | 114.5–140.6 | 0.92 | |
TTPmin >25 | 38* | 98.4 (mean)** | 84.0–112.7 | 131.9 (mean)** | 122.4–141.3 | |||
IDH1/2 mut–1p/19q non-codel (n = 79) | 18F-FET negative | 28 | 53.8 (median) | 41.3–66.3 | 0.42 | 99.2 (mean)** | 83.7–115.2 | 0.65 |
18F-FET positive | 51 | 37.1 (median) | 27.5–46.6 | 98.1 (mean)** | 111.8–110.6 | |||
TBRmax ≤1.7 | 40 | 56.8 (median) | 40.8–72.8 | 0.01 | 104.2 (mean)** | 91.5–116.9 | 0.18 | |
TBRmax >1.7 | 39 | 34.9 (median) | 18.0–51.9 | 92.4 (mean)** | 76.8–107.9 | |||
TTPmin ≤25 | 31* | 32.3 (median) | 16.4–48.2 | 0.02 | 75.1 (median) | 58.1–92.0 | 0.002 | |
TTPmin >25 | 19* | 77.0 (mean)** | 56.1–97.9 | 125.6 (mean)** | 115.0–136.2 | |||
IDH1/2 wt (n = 76) | 18F-FET negative | 9 | 35.2 (median) | 15.8–33.3 | 0.09 | 40.9 (median) | 14.7–67.0 | 0.15 |
18F-FET positive | 67 | 12.0 (median) | 6.5–17.3 | 24.5 (median) | 20.1–28.8 | |||
TBRmax ≤2.5 | 41 | 14.4 (median) | 10.8–17.9 | 0.52 | 26.2 (median) | 20.7–31.6 | 0.79 | |
TBRmax >2.5 | 35 | 10.4 (median) | 1.6–19.1 | 24.0 (median) | 17.4–30.6 | |||
TTPmin ≤12.5 | 44* | 10.4 (median) | 4.9–19.9 | 0.51 | 24.0 (median) | 20.2–27.9 | 0.14 | |
TTPmin >12.5 | 23* | 11.9 (median) | 6.3–17.6 | 26.0 (median) | 15.8–36.1 | |||
GBM (n = 73) | 18F-FET negative | 0 | Not applicable | NA | NA | Not applicable | NA | NA |
18F-FET positive | 73 | 10.3 (median) | 9.1–11.5 | 14.0 (median) | 10.4–17.7 | |||
TBRmax ≤3.6 | 36 | 11.9 (median) | 7.6–16.1 | 0.18 | 15.8 (median) | 7.1–24.5 | 0.67 | |
TBRmax >3.6 | 37 | 8.7 (median) | 6.0–11.4 | 13.4 (median) | 10.4–16.3 | |||
TTPmin ≤12.5 | 49* | 9.9 (median) | 7.8–12.0 | 0.36 | 12.8 (median) | 11.2–14.3 | 0.29 | |
TTPmin >12.5 | 20* | 11.9 (median) | 9.1–14.6 | 23.1 (median) | 11.7–34.6 |
*TTP was calculated in 18F-FET positive patients only (independent of TBRmax); **median times were not reached; na: not applicable
. | Factor . | N . | PFS (mo) . | 95% CI . | Log-rank P . | OS (mo) . | 95% CI . | Log-rank P . |
---|---|---|---|---|---|---|---|---|
All (including patients without complete molecular profile) | 18F-FET negative | 45 | 50.1 (median) | 24.0–76.7 | 0.01 | 97.9 (mean)** | 83.1–112.7 | 0.001 |
18F-FET positive | 255 | 20.0 (median) | 15.9–24.3 | 46.1 (mean)** | 26.6–65.6 | |||
TBRmax ≤2.6 | 147 | 50.1 (median) | 30.0–70.2 | <0.001 | 89.4 (mean)** | 79.9–98.9 | <0.001 | |
TBRmax >2.6 | 153 | 15.4 (median) | 12.4–18.4 | 34.2 (median) | 23.0–45.5 | |||
TTPmin ≤17.5 | 162* | 14.2 (median) | 11.1–17.3 | <0.001 | 26.2 (median) | 21.6–30.8 | <0.001 | |
TTPmin >17.5 | 88* | 74.7 (mean)** | 62.7–86.7 | 116.3 (mean)** | 106.4–126.3 | |||
IDH1/2 mut–1p/19q codel (n = 58) | 18F-FET negative | 3 | 26.3 (mean)** | 18.3–34.2 | 0.60 | All cases censored | na | 0.73 |
18F-FET positive | 55 | 92.2 (mean)** | 78.3–106.2 | All cases censored | na | |||
TBRmax ≤2.7 | 29 | 87.9 (median) | 10.5–165.4 | 0.58 | All cases censored | na | 0.07 | |
TBRmax >2.7 | 29 | 95.7 (mean)** | 77.5–113.9 | All cases censored | na | |||
TTPmin ≤25 | 17* | 84.2 (mean)** | 61.0–107.5 | 0.54 | 127.5 (mean)** | 114.5–140.6 | 0.92 | |
TTPmin >25 | 38* | 98.4 (mean)** | 84.0–112.7 | 131.9 (mean)** | 122.4–141.3 | |||
IDH1/2 mut–1p/19q non-codel (n = 79) | 18F-FET negative | 28 | 53.8 (median) | 41.3–66.3 | 0.42 | 99.2 (mean)** | 83.7–115.2 | 0.65 |
18F-FET positive | 51 | 37.1 (median) | 27.5–46.6 | 98.1 (mean)** | 111.8–110.6 | |||
TBRmax ≤1.7 | 40 | 56.8 (median) | 40.8–72.8 | 0.01 | 104.2 (mean)** | 91.5–116.9 | 0.18 | |
TBRmax >1.7 | 39 | 34.9 (median) | 18.0–51.9 | 92.4 (mean)** | 76.8–107.9 | |||
TTPmin ≤25 | 31* | 32.3 (median) | 16.4–48.2 | 0.02 | 75.1 (median) | 58.1–92.0 | 0.002 | |
TTPmin >25 | 19* | 77.0 (mean)** | 56.1–97.9 | 125.6 (mean)** | 115.0–136.2 | |||
IDH1/2 wt (n = 76) | 18F-FET negative | 9 | 35.2 (median) | 15.8–33.3 | 0.09 | 40.9 (median) | 14.7–67.0 | 0.15 |
18F-FET positive | 67 | 12.0 (median) | 6.5–17.3 | 24.5 (median) | 20.1–28.8 | |||
TBRmax ≤2.5 | 41 | 14.4 (median) | 10.8–17.9 | 0.52 | 26.2 (median) | 20.7–31.6 | 0.79 | |
TBRmax >2.5 | 35 | 10.4 (median) | 1.6–19.1 | 24.0 (median) | 17.4–30.6 | |||
TTPmin ≤12.5 | 44* | 10.4 (median) | 4.9–19.9 | 0.51 | 24.0 (median) | 20.2–27.9 | 0.14 | |
TTPmin >12.5 | 23* | 11.9 (median) | 6.3–17.6 | 26.0 (median) | 15.8–36.1 | |||
GBM (n = 73) | 18F-FET negative | 0 | Not applicable | NA | NA | Not applicable | NA | NA |
18F-FET positive | 73 | 10.3 (median) | 9.1–11.5 | 14.0 (median) | 10.4–17.7 | |||
TBRmax ≤3.6 | 36 | 11.9 (median) | 7.6–16.1 | 0.18 | 15.8 (median) | 7.1–24.5 | 0.67 | |
TBRmax >3.6 | 37 | 8.7 (median) | 6.0–11.4 | 13.4 (median) | 10.4–16.3 | |||
TTPmin ≤12.5 | 49* | 9.9 (median) | 7.8–12.0 | 0.36 | 12.8 (median) | 11.2–14.3 | 0.29 | |
TTPmin >12.5 | 20* | 11.9 (median) | 9.1–14.6 | 23.1 (median) | 11.7–34.6 |
. | Factor . | N . | PFS (mo) . | 95% CI . | Log-rank P . | OS (mo) . | 95% CI . | Log-rank P . |
---|---|---|---|---|---|---|---|---|
All (including patients without complete molecular profile) | 18F-FET negative | 45 | 50.1 (median) | 24.0–76.7 | 0.01 | 97.9 (mean)** | 83.1–112.7 | 0.001 |
18F-FET positive | 255 | 20.0 (median) | 15.9–24.3 | 46.1 (mean)** | 26.6–65.6 | |||
TBRmax ≤2.6 | 147 | 50.1 (median) | 30.0–70.2 | <0.001 | 89.4 (mean)** | 79.9–98.9 | <0.001 | |
TBRmax >2.6 | 153 | 15.4 (median) | 12.4–18.4 | 34.2 (median) | 23.0–45.5 | |||
TTPmin ≤17.5 | 162* | 14.2 (median) | 11.1–17.3 | <0.001 | 26.2 (median) | 21.6–30.8 | <0.001 | |
TTPmin >17.5 | 88* | 74.7 (mean)** | 62.7–86.7 | 116.3 (mean)** | 106.4–126.3 | |||
IDH1/2 mut–1p/19q codel (n = 58) | 18F-FET negative | 3 | 26.3 (mean)** | 18.3–34.2 | 0.60 | All cases censored | na | 0.73 |
18F-FET positive | 55 | 92.2 (mean)** | 78.3–106.2 | All cases censored | na | |||
TBRmax ≤2.7 | 29 | 87.9 (median) | 10.5–165.4 | 0.58 | All cases censored | na | 0.07 | |
TBRmax >2.7 | 29 | 95.7 (mean)** | 77.5–113.9 | All cases censored | na | |||
TTPmin ≤25 | 17* | 84.2 (mean)** | 61.0–107.5 | 0.54 | 127.5 (mean)** | 114.5–140.6 | 0.92 | |
TTPmin >25 | 38* | 98.4 (mean)** | 84.0–112.7 | 131.9 (mean)** | 122.4–141.3 | |||
IDH1/2 mut–1p/19q non-codel (n = 79) | 18F-FET negative | 28 | 53.8 (median) | 41.3–66.3 | 0.42 | 99.2 (mean)** | 83.7–115.2 | 0.65 |
18F-FET positive | 51 | 37.1 (median) | 27.5–46.6 | 98.1 (mean)** | 111.8–110.6 | |||
TBRmax ≤1.7 | 40 | 56.8 (median) | 40.8–72.8 | 0.01 | 104.2 (mean)** | 91.5–116.9 | 0.18 | |
TBRmax >1.7 | 39 | 34.9 (median) | 18.0–51.9 | 92.4 (mean)** | 76.8–107.9 | |||
TTPmin ≤25 | 31* | 32.3 (median) | 16.4–48.2 | 0.02 | 75.1 (median) | 58.1–92.0 | 0.002 | |
TTPmin >25 | 19* | 77.0 (mean)** | 56.1–97.9 | 125.6 (mean)** | 115.0–136.2 | |||
IDH1/2 wt (n = 76) | 18F-FET negative | 9 | 35.2 (median) | 15.8–33.3 | 0.09 | 40.9 (median) | 14.7–67.0 | 0.15 |
18F-FET positive | 67 | 12.0 (median) | 6.5–17.3 | 24.5 (median) | 20.1–28.8 | |||
TBRmax ≤2.5 | 41 | 14.4 (median) | 10.8–17.9 | 0.52 | 26.2 (median) | 20.7–31.6 | 0.79 | |
TBRmax >2.5 | 35 | 10.4 (median) | 1.6–19.1 | 24.0 (median) | 17.4–30.6 | |||
TTPmin ≤12.5 | 44* | 10.4 (median) | 4.9–19.9 | 0.51 | 24.0 (median) | 20.2–27.9 | 0.14 | |
TTPmin >12.5 | 23* | 11.9 (median) | 6.3–17.6 | 26.0 (median) | 15.8–36.1 | |||
GBM (n = 73) | 18F-FET negative | 0 | Not applicable | NA | NA | Not applicable | NA | NA |
18F-FET positive | 73 | 10.3 (median) | 9.1–11.5 | 14.0 (median) | 10.4–17.7 | |||
TBRmax ≤3.6 | 36 | 11.9 (median) | 7.6–16.1 | 0.18 | 15.8 (median) | 7.1–24.5 | 0.67 | |
TBRmax >3.6 | 37 | 8.7 (median) | 6.0–11.4 | 13.4 (median) | 10.4–16.3 | |||
TTPmin ≤12.5 | 49* | 9.9 (median) | 7.8–12.0 | 0.36 | 12.8 (median) | 11.2–14.3 | 0.29 | |
TTPmin >12.5 | 20* | 11.9 (median) | 9.1–14.6 | 23.1 (median) | 11.7–34.6 |
*TTP was calculated in 18F-FET positive patients only (independent of TBRmax); **median times were not reached; na: not applicable
Median TTPmin was 17.5 minutes for all patients; patients with a TTPmin >17.5 minutes had longer PFS and OS than patients with a TTPmin ≤17.5 minutes: median PFS not reached versus 14.2 months (95% CI: 11.1–17.3)/(log-rank P < 0.001) and median OS not reached versus 26.2 months (95% CI: 21.6–30.8)/(log-rank P < 0.001).
Next, we assessed the correlation with outcome of 18F-FET-PET parameters within the molecular subgroups defined by IDH1/2 mutation, 1p/19q codeletion status, and GBM histology according to the WHO 2016 classification.6
IDH1/2 Mutant–1p/19q Codeleted Group
In this subgroup, neither presence nor absence of 18F-FET uptake nor TBRmax nor TTPmin discriminated between outcome of patients. Median TBRmax was 2.7 in this subgroup and median TTPmin was 25 minutes. There was no significant OS difference between tumors with a TTPmin >25 and those with a TTPmin of ≤25 minutes (log-rank P = 0.54 for PFS and P = 0.92 for OS; Table 2, Fig. 1A/B).
IDH1/2 Mutant–1p/19q Non-Codeleted Group
In this subgroup, 18F-FET-uptake per se (positive versus negative) did not correlate with outcome, while the magnitude of 18F-FET uptake did. Patients with TBRmax values ≤1.7 (in-group median) had a significantly longer PFS time and slightly longer OS time compared with those with TBRmax values >1.7 (Table 2). Furthermore, outcome analysis according to the in-group median TTPmin of 25 minutes revealed patients with a TTPmin >25 minutes to have a significantly better outcome for both PFS and OS (Table 2, Fig. 1C/D). A comparison of clinical parameters between these 2 patient groups with different outcomes revealed no distinguishing feature apart from a different distribution of TTPmin between the WHO grades (Supplementary Table S3). Interestingly, patients with a TTPmin >25 minutes had a better outcome than patients with a TTPmin ≤25 minutes irrespective of WHO grade II or III (Fig. 3B). Median TBRmax did not provide a comparable separation (Fig. 3C). An example of a patient with grade III tumor/TTPmin >25 and favorable outcome as opposed to poor outcome in a patient with a grade II tumor and TTPmin ≤25 minutes is illustrated in Supplementary Fig. S2.
IDH1/2 Wildtype Group (WHO grades II and III)
Neither 18F-FET uptake (positive versus negative) nor median TBRmax of 2.5 was associated with outcome (Table 2). Furthermore, outcome analysis according to median TTPmin of 12.5 minutes in IDH1/2 wt tumors showed no difference in PFS and OS times between patients with a TTPmin ≤12.5 compared with those with a TTPmin >12.5 minutes (P = 0.51 and P = 0.14, respectively; Table 2, Fig. 2A/B).
GBM Group
All tumors within this subgroup were 18FET positive. Neither the median TBRmax of 3.6 nor the median TTPmin of 12.5 minutes was associated with outcome. However, while not reaching statistical significance, with 23.1 months compared with 12.8 months, patients with a TTPmin >12.5 minutes had a considerably longer OS time (P = 0.29; Fig. 2D).
Univariate and Multivariate Survival Analysis (all patients)
Univariate analysis revealed lower age, higher KPS, delay of cytotoxic therapy, lower WHO grade, and presence of IDH1/2 mutation to be highly associated with both PFS and OS in the entire group (Table 3). Absence of CE on initial MRI, absence of 18F-FET uptake, lower TBRmax, and a TTPmin >17.5 minutes were also associated with longer PFS and OS.
Univariate Analysis . | ||||||
---|---|---|---|---|---|---|
Factor . | PFS . | OS . | ||||
. | P-value . | Hazard Ratio (HR) . | 95%CI: HR . | P-value . | HR . | 95% CI: HR . |
Age | <0.001 | 0.41 | 0.31–0.56 | <0.001 | 0.27 | 0.19–0.38 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.03 | 0.71 | 0.52–0.97 | <0.001 | 0.46 | 0.33–0.64 |
≥90 | ||||||
<90 | ||||||
Surgical procedurea | 0.36 | 0.85 | 0.60–1.21 | 0.54 | 0.88 | 0.60–1.31 |
Adjuvant therapyb | <0.001 | 0.62 | 0.54–0.71 | <0.001 | 0.49 | 0.42–0.58 |
WHO gradec | <0.001 | 0.44 | 0.36–0.53 | <0.001 | 0.31 | 0.25–0.39 |
IDH 1/2 mutationd | <0.001 | 0.17 | 0.12–0.23 | <0.001 | 0.09 | 0.06–0.14 |
18FET-negative vs positive | 0.01 | 0.58 | 0.38–0.89 | 0.001 | 0.39 | 0.22–0.69 |
TBRmax | <0.001 | 0.47 | 0.35–0.64 | <0.001 | 0.47 | 0.34–0.66 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | <0.001 | 0.29 | 0.20–0.42 | <0.001 | 0.14 | 0.08–0.24 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | <0.001 | 0.45 | 0.33–0.63 | <0.001 | 0.34 | 0.23–0.48 |
No | ||||||
Yes | ||||||
T2 volume | 0.10 | 0.78 | 0.58–1.04 | 0.39 | 0.87 | 0.62–0.83 |
≤49 mL | ||||||
>49 mL | ||||||
Multivariate analysis | ||||||
Factor | PFS | OS | ||||
P-value | HR | 95% CI: HR | P-value | HR | 95% CI: HR | |
Age | 0.80 | 0.95 | 0.65–1.40 | 0.11 | 0.71 | 0.47–1.07 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.49 | 0.94 | 0.78–1.12 | 0.75 | 0.95 | 0.69–1.31 |
≥90 | ||||||
<90 | ||||||
Adjuvant therapyb | 0.23 | 0.85 | 0.66–1.11 | 0.24 | 0.87 | 0.61–1.13 |
WHO gradec | 0.04 | 0.65 | 0.44–0.97 | 0.007 | 0.52 | 0.33–0.84 |
IDH 1/2 mutationd | <0.001 | 0.24 | 1.15–0.40 | <0.001 | 0.19 | 0.11–0.34 |
TBRmax | 0.08 | 0.70 | 0.47–1.05 | 0.72 | 0.93 | 0.59–1.43 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | 0.19 | 0.72 | 0.44–1.18 | 0.01 | 0.43 | 0.22–0.82 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | 0.83 | 0.95 | 0.61–1.49 | 0.22 | 0.71 | 0.42–1.22 |
No | ||||||
Yes |
Univariate Analysis . | ||||||
---|---|---|---|---|---|---|
Factor . | PFS . | OS . | ||||
. | P-value . | Hazard Ratio (HR) . | 95%CI: HR . | P-value . | HR . | 95% CI: HR . |
Age | <0.001 | 0.41 | 0.31–0.56 | <0.001 | 0.27 | 0.19–0.38 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.03 | 0.71 | 0.52–0.97 | <0.001 | 0.46 | 0.33–0.64 |
≥90 | ||||||
<90 | ||||||
Surgical procedurea | 0.36 | 0.85 | 0.60–1.21 | 0.54 | 0.88 | 0.60–1.31 |
Adjuvant therapyb | <0.001 | 0.62 | 0.54–0.71 | <0.001 | 0.49 | 0.42–0.58 |
WHO gradec | <0.001 | 0.44 | 0.36–0.53 | <0.001 | 0.31 | 0.25–0.39 |
IDH 1/2 mutationd | <0.001 | 0.17 | 0.12–0.23 | <0.001 | 0.09 | 0.06–0.14 |
18FET-negative vs positive | 0.01 | 0.58 | 0.38–0.89 | 0.001 | 0.39 | 0.22–0.69 |
TBRmax | <0.001 | 0.47 | 0.35–0.64 | <0.001 | 0.47 | 0.34–0.66 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | <0.001 | 0.29 | 0.20–0.42 | <0.001 | 0.14 | 0.08–0.24 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | <0.001 | 0.45 | 0.33–0.63 | <0.001 | 0.34 | 0.23–0.48 |
No | ||||||
Yes | ||||||
T2 volume | 0.10 | 0.78 | 0.58–1.04 | 0.39 | 0.87 | 0.62–0.83 |
≤49 mL | ||||||
>49 mL | ||||||
Multivariate analysis | ||||||
Factor | PFS | OS | ||||
P-value | HR | 95% CI: HR | P-value | HR | 95% CI: HR | |
Age | 0.80 | 0.95 | 0.65–1.40 | 0.11 | 0.71 | 0.47–1.07 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.49 | 0.94 | 0.78–1.12 | 0.75 | 0.95 | 0.69–1.31 |
≥90 | ||||||
<90 | ||||||
Adjuvant therapyb | 0.23 | 0.85 | 0.66–1.11 | 0.24 | 0.87 | 0.61–1.13 |
WHO gradec | 0.04 | 0.65 | 0.44–0.97 | 0.007 | 0.52 | 0.33–0.84 |
IDH 1/2 mutationd | <0.001 | 0.24 | 1.15–0.40 | <0.001 | 0.19 | 0.11–0.34 |
TBRmax | 0.08 | 0.70 | 0.47–1.05 | 0.72 | 0.93 | 0.59–1.43 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | 0.19 | 0.72 | 0.44–1.18 | 0.01 | 0.43 | 0.22–0.82 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | 0.83 | 0.95 | 0.61–1.49 | 0.22 | 0.71 | 0.42–1.22 |
No | ||||||
Yes |
a: surgery vs biopsy; b: adjuvant therapy: wait-and-see vs chemotherapy vs radiation vs radiochemotherapy; c: WHO grade II vs III vs IV; d: IDH1/2 mut vs IDH wt.
Univariate Analysis . | ||||||
---|---|---|---|---|---|---|
Factor . | PFS . | OS . | ||||
. | P-value . | Hazard Ratio (HR) . | 95%CI: HR . | P-value . | HR . | 95% CI: HR . |
Age | <0.001 | 0.41 | 0.31–0.56 | <0.001 | 0.27 | 0.19–0.38 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.03 | 0.71 | 0.52–0.97 | <0.001 | 0.46 | 0.33–0.64 |
≥90 | ||||||
<90 | ||||||
Surgical procedurea | 0.36 | 0.85 | 0.60–1.21 | 0.54 | 0.88 | 0.60–1.31 |
Adjuvant therapyb | <0.001 | 0.62 | 0.54–0.71 | <0.001 | 0.49 | 0.42–0.58 |
WHO gradec | <0.001 | 0.44 | 0.36–0.53 | <0.001 | 0.31 | 0.25–0.39 |
IDH 1/2 mutationd | <0.001 | 0.17 | 0.12–0.23 | <0.001 | 0.09 | 0.06–0.14 |
18FET-negative vs positive | 0.01 | 0.58 | 0.38–0.89 | 0.001 | 0.39 | 0.22–0.69 |
TBRmax | <0.001 | 0.47 | 0.35–0.64 | <0.001 | 0.47 | 0.34–0.66 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | <0.001 | 0.29 | 0.20–0.42 | <0.001 | 0.14 | 0.08–0.24 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | <0.001 | 0.45 | 0.33–0.63 | <0.001 | 0.34 | 0.23–0.48 |
No | ||||||
Yes | ||||||
T2 volume | 0.10 | 0.78 | 0.58–1.04 | 0.39 | 0.87 | 0.62–0.83 |
≤49 mL | ||||||
>49 mL | ||||||
Multivariate analysis | ||||||
Factor | PFS | OS | ||||
P-value | HR | 95% CI: HR | P-value | HR | 95% CI: HR | |
Age | 0.80 | 0.95 | 0.65–1.40 | 0.11 | 0.71 | 0.47–1.07 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.49 | 0.94 | 0.78–1.12 | 0.75 | 0.95 | 0.69–1.31 |
≥90 | ||||||
<90 | ||||||
Adjuvant therapyb | 0.23 | 0.85 | 0.66–1.11 | 0.24 | 0.87 | 0.61–1.13 |
WHO gradec | 0.04 | 0.65 | 0.44–0.97 | 0.007 | 0.52 | 0.33–0.84 |
IDH 1/2 mutationd | <0.001 | 0.24 | 1.15–0.40 | <0.001 | 0.19 | 0.11–0.34 |
TBRmax | 0.08 | 0.70 | 0.47–1.05 | 0.72 | 0.93 | 0.59–1.43 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | 0.19 | 0.72 | 0.44–1.18 | 0.01 | 0.43 | 0.22–0.82 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | 0.83 | 0.95 | 0.61–1.49 | 0.22 | 0.71 | 0.42–1.22 |
No | ||||||
Yes |
Univariate Analysis . | ||||||
---|---|---|---|---|---|---|
Factor . | PFS . | OS . | ||||
. | P-value . | Hazard Ratio (HR) . | 95%CI: HR . | P-value . | HR . | 95% CI: HR . |
Age | <0.001 | 0.41 | 0.31–0.56 | <0.001 | 0.27 | 0.19–0.38 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.03 | 0.71 | 0.52–0.97 | <0.001 | 0.46 | 0.33–0.64 |
≥90 | ||||||
<90 | ||||||
Surgical procedurea | 0.36 | 0.85 | 0.60–1.21 | 0.54 | 0.88 | 0.60–1.31 |
Adjuvant therapyb | <0.001 | 0.62 | 0.54–0.71 | <0.001 | 0.49 | 0.42–0.58 |
WHO gradec | <0.001 | 0.44 | 0.36–0.53 | <0.001 | 0.31 | 0.25–0.39 |
IDH 1/2 mutationd | <0.001 | 0.17 | 0.12–0.23 | <0.001 | 0.09 | 0.06–0.14 |
18FET-negative vs positive | 0.01 | 0.58 | 0.38–0.89 | 0.001 | 0.39 | 0.22–0.69 |
TBRmax | <0.001 | 0.47 | 0.35–0.64 | <0.001 | 0.47 | 0.34–0.66 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | <0.001 | 0.29 | 0.20–0.42 | <0.001 | 0.14 | 0.08–0.24 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | <0.001 | 0.45 | 0.33–0.63 | <0.001 | 0.34 | 0.23–0.48 |
No | ||||||
Yes | ||||||
T2 volume | 0.10 | 0.78 | 0.58–1.04 | 0.39 | 0.87 | 0.62–0.83 |
≤49 mL | ||||||
>49 mL | ||||||
Multivariate analysis | ||||||
Factor | PFS | OS | ||||
P-value | HR | 95% CI: HR | P-value | HR | 95% CI: HR | |
Age | 0.80 | 0.95 | 0.65–1.40 | 0.11 | 0.71 | 0.47–1.07 |
≤48 y | ||||||
>48 y | ||||||
KPS | 0.49 | 0.94 | 0.78–1.12 | 0.75 | 0.95 | 0.69–1.31 |
≥90 | ||||||
<90 | ||||||
Adjuvant therapyb | 0.23 | 0.85 | 0.66–1.11 | 0.24 | 0.87 | 0.61–1.13 |
WHO gradec | 0.04 | 0.65 | 0.44–0.97 | 0.007 | 0.52 | 0.33–0.84 |
IDH 1/2 mutationd | <0.001 | 0.24 | 1.15–0.40 | <0.001 | 0.19 | 0.11–0.34 |
TBRmax | 0.08 | 0.70 | 0.47–1.05 | 0.72 | 0.93 | 0.59–1.43 |
≤2.6 | ||||||
>2.6 | ||||||
TTPmin | 0.19 | 0.72 | 0.44–1.18 | 0.01 | 0.43 | 0.22–0.82 |
>17.5 min | ||||||
≤17.5 min | ||||||
CE | 0.83 | 0.95 | 0.61–1.49 | 0.22 | 0.71 | 0.42–1.22 |
No | ||||||
Yes |
a: surgery vs biopsy; b: adjuvant therapy: wait-and-see vs chemotherapy vs radiation vs radiochemotherapy; c: WHO grade II vs III vs IV; d: IDH1/2 mut vs IDH wt.
Multivariate analysis was conducted using all parameters with P < 0.05 in the univariate analysis. Lower WHO grade and presence of IDH1/2 mutation were associated with prolonged PFS. In addition to WHO grade and IDH1/2 mutation, TTPmin >17.5 minutes was an independent prognostic factor for improved survival (see Table 3). Subgroup analysis could not be performed due to low number of events in the 2 IDH1/2 mutated groups.
Discussion
Since the molecular markers IDH1/2 mutation and 1p/19q codeletion have been identified to be strongly associated with prognosis,1–5,18–20 the 2016 revision of the WHO Classification of Tumors of the Central Nervous System has implemented a classification scheme for gliomas based on these molecular markers.6 This might affect therapeutic approaches in the future by allowing stronger emphasis on individual, tumor-tailored therapies based on molecular profiling.
Accordingly, amino acid PET has been shown to provide valuable information regarding differential diagnosis of cerebral lesions as well as prognosis among gliomas.8,9,12,21,22 Dynamic analysis of 18F-FET uptake using TTPmin analysis discriminates patients with poor or favorable prognosis at the time of diagnosis in gliomas across WHO 2007 grades II to IV.8,9 In light of the revision of the WHO classification, we sought to reassess the information derived from dynamic analysis of 18F-FET-PET within the framework of a glioma classification in adults largely based on IDH1/2 mutation and 1p/19q codeletion.
As a principal observation, longer TTPmin correlates with longer OS independently of grading and the presence of IDH1/2 mutation in our entire study population. Notably, in IDH1/2 mut–1p/19q non-codel gliomas, TTPmin provides an additional prognostic marker, emphasizing the value of PET in these tumors.
The biological mechanism of tracer kinetics leading to short or long TTPmin is not fully understood yet.2318F-FET uptake depends on a bidirectional L-type amino acid transporter (LAT1/2) expressed in the cell membrane and vasculature of gliomas. Its expression level was found to correlate with the degree of malignancy according to the WHO 2007 classification: Pooled grades III and IV gliomas had much higher LAT1 expression than grade II gliomas.24 Moreover, overexpression of LAT1 in glioma cells with low endogenous LAT1 expression enhanced tumor growth in nude mice.24 After intracellular uptake, 18F-FET is not incorporated into proteins or trapped within the tumor cell but washed out after a certain period of time.23 The faster the 18F-FET uptake, the faster it is washed out of the tumor cell. A higher tracer turnover might be influenced by either a higher LAT1/2 expression or higher tracer availability due to increased tumor vascularity/perfusion. High vascularity and an elevated ratio of LATs have been reported for grades III and IV gliomas, and both might contribute to this observation. Both patients with IDH1/2 wt grades II and III tumors as well as patients with GBM have the shortest median TTPmin of 12.5 minutes. In the GBM subgroup, patients with a TTPmin >12.5 minutes have a considerably longer survival time of 23.1 months compared with 12.8 months in patients with a TTPmin ≤12.5. In the other group also known to display higher perfusion, namely the IDH1/2 mut–1p/19 q codel tumors, we did not detect TTPmin as a prognostic factor, which might be due to a limited number of events. One could speculate whether high perfusion might interfere with effects of TTPmin in these 2 highly vascularized tumor groups. Interestingly, different perfusion properties were shown to be associated with both outcome and presence of IDH1/2 mutation in astrocytic tumors by Kickingereder et al.25 These authors found an overactivation of pro-angiogenic pathways in IDH1/2 wt tumors, well explaining the observed difference in relative cerebral blood volume in the different molecular subgroups and demonstrating the potential additional value of imaging biomarkers. As TTPmin was associated with survival in tumors being IDH1/2 mutated without 1p/19q codeletion, both TTPmin and perfusion might be surrogates of a distinct biological tumor property. Further studies combining dynamic PET and perfusion-based MRI may help elucidate the interaction between perfusion, vascularity, and TTPmin.
In contrast to TTPmin, TBRmax was associated with prognosis in the 2 IDH1/2 mutant groups. OS was longer, albeit not significant, in IDH1/2 mut–1p/19q codel patients with TBRmax ≤2.7 (P = 0.07) and PFS was significantly longer in the non-codel tumors with TBRmax ≤1.7 (P = 0.01). Hence, the magnitude of 18F-FET uptake might be associated with histological features such as cell density, mitotic index, and vascularization and thus most likely reflects WHO grade.26
The factor “18F-FET-negative” was associated with favorable outcome in the entire group; however, this could be attributed to the high intercorrelation with the molecular subtype: 18F-FET− tumors were most often found among the IDH1/2 mut–1p/19q non-codel (astrocytic) tumor type, while 95% of IDH1/2 wt tumors were 18F-FET positive.
Although counterintuitive, the factor “18F-FET-negative” lost its significance within the group of IDH1/2 mut–1p/19q non-codel tumors, and could not be evaluated in the remaining 2 groups due to small number of 18F-FET− cases. So far, the underlying mechanisms leading to complete lack of 18F-FET uptake are not understood; one explanation might be the lack or an inactivity of LATs in a proportion of IDH1/2 mut–1p/19q non-codel tumors (approximately one-third of our astrocytic tumor population) and remains to be addressed in further studies.
Limitations of the study arise from the retrospective study design and heterogeneous surgical and postsurgical management strategies.
Furthermore, TBRmax values as well as determination of TTPmin are dependent on scanner resolution and data processing, with the consequence that absolute values may not be comparable between different centers.22,27 A standardization of data processing and evaluation will help to improve comparability.
Altogether, dynamic analysis of 18F-FET tracer uptake using TTPmin discriminates patients with favorable and poor prognosis within the molecular defined subgroup of IDH1/2 mut–1p/19q non-codel tumors, most of which are now classified as astrocytomas. Thus, this might be an imaging biomarker providing additional prognostic information to stratify astrocytoma patients into low-risk and high-risk groups.
Supplementary Material
Supplementary material is available at Neuro-Oncology online.
Funding
We thank the Remark-Moolenaars family for a financial grant to support this study.
Conflict of interest statement. B.S, A.G., A.B., M.U., M.D., U.S. and N.L.A. report no confict of interest. M.W. has received research grants from Acceleron, Actelion, Bayer, Isarna, MSD, Merck & Co, Novocure, OGD2, Piqur, and Roche and advisory board/consulting honoraria from BMS, Celldex, Immunocellular Therapeutics, Isarna, Magforce, MSD, Merck & Co, Northwest Biotherapeutics, Novocure, Pfizer, Roche, Teva and Tocagen. P.B. has received honoraria/consulting fees from GE and Siemens, J.C.T. has received research grants from BrainLab and honoraria for advisory board participation from Celldex, Roche, BrainLab, and Siemens.
References
Author notes
Present address: Institute of Neuropathology, University Medical Center, Hamburg-Eppendorf, Research Institute Children’s Cancer Center, Department of Pediatric Hematology and Oncology, University Medical Center, Hamburg-Eppendorf, Hamburg, Germany
Equal contribution