Glioblastoma has a gigantic unmet medical need. Molecular knowledge has evolved substantially, including data on clonal selection with progression. Past trials for all-comers may have produced false negative results. Molecular precision at progression needs workup of new tissue, and revisiting drugs with a focus on brain tumor penetration may yield surprises. Clin Cancer Res; 24(2); 256–8. ©2017 AACR.

See related article by Byron et al., p. 295

In this issue of Clinical Cancer Research, Byron and colleagues from large U.S. brain tumor centers address feasibility of several relevant aspects of implementation of high-throughput biomarker assessment in the clinical decision process as well as revisiting older drugs with a potential in the brain tumor setting in one prospective cohort (1). Patients with newly diagnosed glioblastoma in trials receiving optimal therapy, including maximal safe surgical resection, modern radiotherapy, and chemotherapy with temozolomide, have a median survival of approximately 14 months, with about 27% of patients surviving up to 2 years and fewer than 10% of patients surviving up to 5 years (2). The addition of alternating electric fields, also referred to as tumor-treating fields, to maintenance temozolomide has shown promise, extending median overall survival in patients with newly diagnosed glioblastoma (3), but not if used at recurrence.

These concepts show that so far, the only options to prolong life for patients with glioblastoma are used in the newly diagnosed setting, whereas all attempts to prolong survival at progression largely failed, most recently with the combination of lomustine and bevacizumab (4).

The most intensively researched area in clinical neuro-oncology outside the different aspects of immunotherapy is precision therapy, largely focusing on targeted compounds.

Although it is acknowledged that glioblastoma is a molecularly diverse tumor type with multiple subgroups (5) and there are recent examples of successful implementation of biomarkers to dissect benefiting patients, at least in hindsight (6), outside of O6-methylguanine-DNA-methyltransferase promoter methylation, so far, biomarkers are not taken into consideration for clinical decision-making or trial entry.

The current work by Byron and colleagues focuses on patients at progression with representative postprogression tissue available (1). Treatment is supposed to be guided by molecularly informed decision based on exome and RNA sequencing data. The obstacle is the timely availability of the molecular data, development of standardized tests, and an algorithm for prioritization and combination or not of therapies in the molecular tumor board. In the study, for 20% of patients, there was no relevant tissue based on tumor cell content for the intended molecular diagnostics. Here, the need for fresh-frozen tissue might be overcome by focusing on a panel, which can be done from formalin-fixed paraffin-embedded (FFPE) tissue. The flexibility of the arrays and the confined set of actionable alterations may limit the need for the more complex approach to be performed online and instead perform an array, for example, as published by Sahm and colleagues. Given the data provided in the article, that is, the most common genetic alterations in EGFR (n = 10/16; 63%), PTEN (n = 9/16; 56%), CDKN2A (n = 7/16, 44%), NF1 (n = 7/16, 44%), RB1 (n = 5/16, 31%), and TP53 (n = 5/16, 31%), the array would have been able to catch the molecular changes used for decision-making (1, 7). The fast development of high-throughput exome- or genome-wide approaches and the quality of data even from FFPE material together with more comprehensive, subsequent biostatistics may allow the use of unsupervised methods and replace panels again within the next few years. As a drawback, growing evidence is also pointing toward a relevant heterogeneity within one and the same tumor specifically in spatially or temporally separated tumors, that is, multifocal tumors or tumor recurrences. Whereas data from a recent study do not support uniformity within spatially heterogenous tumors, it supports the current concept of using new tissue information for informed decisions (8).

As long as we lack approved in vitro tests, these experimental procedures need confirmation with a clinically accepted method, mainly IHC or Sanger sequencing, before we should use them for treatment assignments. The current cohort circumvented this topic by stating that treatment based on the tumor board recommendation should be to the discretion of the treating physician. Once, we are aiming at interventional trials, which we should in the near future, there will be a need for stringent assignments.

In the current cohort (1), treatment recommendations have been made within an intended and clinically demanded period about 4 weeks from a pool of 180 FDA-approved agents without a marketing authorization for brain tumors or even cancer (repositioned drugs) but a known, acceptable safety profile, drug–drug interactions, and blood–tumor-barrier penetration. There are several exciting aspects regarding the current cohort: All patients showed at least one lesion that is generally deemed actionable. Combination treatments were regarded feasible given the known profiles of the administered drugs. This resulted in average of 3.4 treatments recommended. The actual therapy differed with drugs left out or a trial option considered more appropriate. Of note, drugs commonly recommended, but not necessarily first choices in neuro-oncology, are metformin (7 times of 15 recommendations), propranolol (8/15), minocycline (4/15), chlorpromazine (2/15) but also several general oncology agents (carboplatin, 3/15; lomustine; and temozolomide), or agents with an approval already focused to targets (palbociclib, pembrolizumab, vismodegib, erlotinib, and everolimus). Importantly, all decisions are based on proposed molecular mechanisms with associated data from other cancers or observations. However, some associations seem rather general, that is, tumors with the presumed oncometabolic mutation in the isocitrate dehydrogenase (IDH)1 gene (R132H) are considered to be very sensitive to metformin (9). Also, propranolol has been associated with decreased risk for cancer (10) and may inhibit vasculogenesis and invasiveness in various tumors. The ability to use multiple drugs in combination eases the challenge to provide a stringent algorithm to put recommendations into a hierarchy. However, when forced to be selective, because of presumed interaction or toxicity, a procedure to derive decisions is desirable, and Fig. 1 gives a schematic view on such an algorithm.

Figure 1.

Path toward precision. Paradigmatic workflow from a recurrent glioblastoma tissue sample is outlined. ALK, anaplastic lymphoma receptor tyrosine kinase; amp, amplification; BBB, blood–brain barrier; BRAF, B-raf proto-oncogene; CDK, cyclin dependent kinase; CDKN2A/B, cyclin dependent kinase inhibitor 2A/B; CNV, copy number variation; IDH, isocitrate dehydrogenase; MDM2, mouse double minute 2; MEK, mitogen-activated protein kinase kinase; MET, hepatocyte growth factor receptor; MGMT, O6-methylguanine-DNA-methyltransferase; mTOR, mechanistic target of rapamycin; N, no; NF1, neurofibromin 1; PALB2, partner and localizer of BRCA2; PARP, poly(ADP-ribose)-polymerase; RB, retinoblastoma; SHH, sonic hedgehog; SNV, single-nucleotide variants; TACC, transforming acidic coiled-coil; wt, wild type; Y, yes.

Figure 1.

Path toward precision. Paradigmatic workflow from a recurrent glioblastoma tissue sample is outlined. ALK, anaplastic lymphoma receptor tyrosine kinase; amp, amplification; BBB, blood–brain barrier; BRAF, B-raf proto-oncogene; CDK, cyclin dependent kinase; CDKN2A/B, cyclin dependent kinase inhibitor 2A/B; CNV, copy number variation; IDH, isocitrate dehydrogenase; MDM2, mouse double minute 2; MEK, mitogen-activated protein kinase kinase; MET, hepatocyte growth factor receptor; MGMT, O6-methylguanine-DNA-methyltransferase; mTOR, mechanistic target of rapamycin; N, no; NF1, neurofibromin 1; PALB2, partner and localizer of BRCA2; PARP, poly(ADP-ribose)-polymerase; RB, retinoblastoma; SHH, sonic hedgehog; SNV, single-nucleotide variants; TACC, transforming acidic coiled-coil; wt, wild type; Y, yes.

Close modal

However, other caveats are not prevented by the ability for relaxed recommendations. The design of the cohort and the wide availability of the drugs, as well as the very heterogenous level of biological evidence for activity, pose a risk of broad off-label use, false hopes, and unwanted interactions if used as an add-on to (other) regular chemotherapy. Also, the current data despite all effort fall short to initiate a development of a certain combination or even single drugs, as proper dose finding, specifically in the brain, and follow-up with focus on adverse effects or more subtle signs of efficacy are missing.

In our view, the authors are to be commended for the attempt, the concept, and the setup of the whole process, including the high-throughput diagnostics, the molecular tumor board, and the stringent treatment suggestions. On the other hand, given the growing interest in neuro-oncology toward repositioned agents, we need to follow up on this first trial with a clear set of necessary laboratory or circumstantial evidence for action, validated and potentially more simple molecular tests, a hierarchy for decision-making, and proper multiarmed (umbrella) and controlled trials. Finally, there will be a debate to what extent biomarkers will be considered ready for preselection of patients or whether a selection of drugs with different presumed biomarkers to be selected for will be randomly assigned and putative biomarkers only addressed in retrospect based on clinical success.

W. Wick reports receiving commercial research grants from Apogenix, Boehringer Ingelheim, MSD, Pfizer, and Roche. No potential conflicts of interest were disclosed by the other author.

Conception and design: W. Wick, T. Kessler

Development of methodology: W. Wick

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): W. Wick, T. Kessler

Writing, review, and/or revision of the manuscript: W. Wick, T. Kessler

This work was supported by German Cancer Aid, 70111980.

1.
Byron
SA
,
Tran
NL
,
Halperin
RF
,
Phillips
JJ
,
Kuhn
JG
,
de Groot
JF
, et al
Prospective feasibility trial for genomics-informed treatment in recurrent and progressive glioblastoma
.
Clin Cancer Res
2018
;
24
:
295
305
.
2.
Weller
M
,
van den Bent
M
,
Tonn
JC
,
Stupp
R
,
Preusser
M
,
Cohen-Jonathan-Moyal
E
, et al
European Association for Neuro-Oncology (EANO) Task Force on Gliomas
. 
European Association for Neuro-Oncology (EANO) guideline on the diagnosis and treatment of adult astrocytic and oligodendroglial gliomas
.
Lancet Oncol
2017
;
18
:
e315
e29
.
3.
Stupp
R
,
Taillibert
S
,
Kanner
AA
,
Kesari
S
,
Steinberg
DM
,
Toms
SA
, et al
Maintenance therapy with tumor-treating fields plus temozolomide vs. temozolomide alone for glioblastoma: a randomized clinical trial
.
JAMA
2015
;
314
:
2535
43
.
4.
Wick
W
,
Gorlia
T
,
Bendszus
M
,
Taphoorn
M
,
Sahm
F
,
Harting
I
, et al
Lomustine and bevacizumab in progressive glioblastoma
.
N Engl J Med
2017
;
377
:
1954
63
.
5.
Kessler
T
,
Sahm
F
,
Sadik
A
,
Stichel
D
,
Hertenstein
A
,
Reifenberger
G
, et al
Molecular differences in IDH wildtype glioblastoma according to MGMT promoter methylation
.
Neuro Oncol
2017 Aug 24
.
[Epub ahead of print]
.
6.
Wick
W
,
Fricke
H
,
Junge
K
,
Kobyakov
G
,
Martens
T
,
Heese
O
, et al
A phase II, randomized, study of weekly APG101+reirradiation versus reirradiation in progressive glioblastoma
.
Clin Cancer Res
2014
;
20
:
6304
13
.
7.
Sahm
F
,
Schrimpf
D
,
Jones
DT
,
Meyer
J
,
Kratz
A
,
Reuss
D
, et al
Next-generation sequencing in routine brain tumor diagnostics enables an integrated diagnosis and identifies actionable targets
.
Acta Neuropathol
2016
;
131
:
903
10
.
8.
Lee
JK
,
Wang
J
,
Sa
JK
,
Ladewig
E
,
Lee
HO
,
Lee
IH
, et al
Spatiotemporal genomic architecture informs precision oncology in glioblastoma
.
Nat Genet
2017
;
49
:
594
9
.
9.
Cuyas
E
,
Fernandez-Arroyo
S
,
Corominas-Faja
B
,
Rodriguez-Gallego
E
,
Bosch-Barrera
J
,
Martin-Castillo
B
, et al
Oncometabolic mutation IDH1 R132H confers a metformin-hypersensitive phenotype
.
Oncotarget
2015
;
6
:
12279
96
.
10.
Chang
PY
,
Huang
WY
,
Lin
CL
,
Huang
TC
,
Wu
YY
,
Chen
JH
, et al
Propranolol reduces cancer risk: a population-based cohort study
.
Medicine
2015
;
94
:
e1097
.