A scoping review of proton radiation therapy and mutant-isocitrate dehydrogenase-inhibitors in IDH mutated lower-grade gliomas: pushing beyond surrogate end-points

Abstract

Background

Proton radiation therapy (PRT) and mutant isocitrate dehydrogenase inhibitors (mIDH-inhibitors) are emerging therapies for mIDH lower grade gliomas (LGGs). Despite their substantial theoretical benefits, comparisons with current standards – particularly pertaining to patient-centred outcomes – are limited.

Methods

Through PubMed and Scopus, a search strategy based on keywords focusing on PRT and mIDH-inhibitors was applied on December 3, 2024. Studies in English on at least 20 adult patients (≥ 18 years) with mIDH-LGG grade 2 or 3 and published between January 1, 2011 and August 31, 2024 were included. Review articles were excluded.

Results

Of 6383 identified articles, seven per treatment strategy were included. Overall survival was not reported for mIDH-inhibitors. The lack of high-quality studies comparing PRT to photon radiation therapy precludes conclusions regarding efficacy, effectiveness or even post-PRT radiological manifestations. For the mIDH-inhibitor Vorasidenib (AG-881), the radiological objective response rate was 10.0–42.9%, although lower for contrast-enhancing tumors. Vorasidenib significantly delayed tumor progression (27.7 versus 11.1 months, p < 0.001) and time to next intervention (not reached versus 17.8 months, p < 0.001) when compared to placebo. Adverse events were mostly mild, including elevated liver enzymes (15.6–44.2%) and headache (26.9–46.2%). Only 1/14 studies included assessments related to quality of life (QoL)-domains with inconclusive research on cognitive outcomes.

Conclusion

Studies reporting on patient-centred data including survival, cognition and QoL remain scarce. Larger, comparative prospective studies, preferably randomized controlled trials, with such outcomes are needed to inform clinicians whether the theoretical and radiological benefits can be translated to improved outcomes that matter to patients, i.e. living better and/or longer.

Supplementary information

The online version contains supplementary material available at 10.1007/s00701-025-06612-6.

Introduction

Isocitrate dehydrogenase (IDH)-mutant lower grade gliomas (LGGs) are slow growing, infiltrating brain tumors. With new treatments emerging, longer survival of affected patients is anticipated [4, 23, 25, 53]. Therefore, finding effective tumor therapies while simultaneously preserving a good quality of life (QoL) is a priority [51].

Surgical resection is often incomplete and both acute and late side effects are a concern after chemo- and/or radiotherapy [7]. Patients not only face shortened survival but also risk tumor- and treatment-related adverse effects [13, 26, 35, 61]. Nevertheless, current optimism in the field of LGG is present due to emerging therapies, namely proton radiation therapy (PRT) and mutant IDH-inhibitors (mIDH-inhibitors).

PRT holds promise to reduce radiation-related side effects given the beneficial dose-delivery to the healthy brain. Contrary to photons, protons have a determinate range and deposit most of their energy just as they come to a stop in the tissues at the so-called Bragg peak, thereafter, showing a steep dose fall-off. This abrupt stop helps to avoid the widespread low-dose radiation to healthy brain tissue [60].

mIDH-inhibitors such as Vorasidenib can halt the disease owing to their ability to decrease the elevated levels of the oncometabolite D-2-Hydroxyglutarate (2-HG) thereby facilitating cell differentiation, immune cell activation and decreasing tumor proliferation [9, 30, 31, 39].

These emerging therapies, both with a solid theoretical fundament, are increasingly used and reported in patients with mIDH-LGG. Before these approaches are universally or even widely used, it is important to illuminate their impact on aspects that matter to patients. Patients are typically concerned if their survival, cognition or QoL will be impaired. Similarly, patients without access to these treatments may wonder what they are missing out on. In this scoping review, we present the available clinical literature and emphasize the lack of studies demonstrating effects in these patient-centred aspects. Further, we discuss the importance of pushing for evidence beyond surrogate end-points also in slow-growing neoplasms like mIDH-LGGs.

The aim of this scoping review was to investigate the current clinical literature on the demonstrated effects of PRT and mIDH-inhibitors with regards to patient-centred outcomes, including improved survival, QoL or spared cognitive function, in adult patients diagnosed with mIDH-LGG grade 2 or 3.

Methods

Search strategy

Due to the expected limited, heterogenous literature along with the fair recency of these treatments in a group of patients requiring long follow-ups, a scoping review was deemed the most appropriate methodology in order to evaluate the current standing and potentially illuminate gaps in the literature where the need of future research can be directed. This scoping review was conducted, and results reported in alignment with guidelines from Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) [62]. Relevant articles were identified by searching the databases PubMed and Scopus. Assisted by a trained librarian, two search-blocks were formulated based on terms and keywords used in the Medical Subject Headings (MeSH) thesaurus and existing literature on PRT and mIDH-inhibitors (see Supplementary Information 1 for full search-blocks (Online Resource 1)). The search was performed on December 3, 2024.

Eligibility criteria

Retrospective and prospective studies where patients with mIDH-LGG grade 2 or 3 received PRT and/or mIDH-inhibitors were included. Only papers written in English with a sample size of at least 20 adult patients (≥ 18 years old) were considered. Studies with at least 20 total patients of which ≥ 80% of patients had mIDH-tumors of grades 2 or 3 were accepted in studies with mixed populations. The inclusion period was between January 1, 2011, and August 31, 2024, in order to allow some years to pass since the identification of IDH in gliomas [42]. Review articles, studies focusing on laboratory research including animal or studies primarily in pediatric populations were excluded.

Study selection and presentation

Titles and abstracts were initially screened by two independent reviewers (AC and DH). Full-text assessments were made thereafter to ensure eligibility. A third reviewer (ASJ) was consulted when eligibility was doubted. As limited eligible literature was anticipated, no meta-analysis was performed. A PRISMA flowchart (see Fig. 1) was used to demonstrate identification and inclusion of the studies [41]. Through tabulation, the data extracted from included full-text articles was presented in Table 1 for PRT and Table 2 for mIDH-inhibitors.

Fig. 1.

PRISMA flow diagram showing inclusion of studies, IDH = Isocitrate dehydrogenase

Table 1.

Characteristics of studies concerning PRT, n:7

				Number of patients				Main Findings in relation to PRT
Study	Design	Follow-up: Median (range) months	Median Dose, Gy RBE (range)	Total, n	Tumor Grade (G)	IDH-status, n (%)	Aim	Survival	QoL	Toxicities	Radiological manifestations
Acharya et al. (2018) [1]	R	PRT: 19.6 (9.7–31.0) Photon: 34.5 (2.8–109.7)	PRT: 54.0 (50.4–59.4) Photon: 59.4 (50.4–63.0)	Total:160 PRT: 37 Photon:123	PRT: G2/3: n:19/18 %: 51/49 Photon: G2/3: n: 48/75 %: 39/61	PRT: -mIDH: 31(84) Photon: -mIDH: 73 (59) -Unknown: 22 (17)	Identify risk factors and incidence of developing clinically significant RN in PRT vs. photon treated gliomas	NR	NR	NR	2-year cumulative incidence for RN: -PRT vs. photon: 18.6% vs. 9.9% - Oligodendroglioma/Astrocytoma: 24.2%/6.2%
Dworkin et al. (2018) [14]	R	Overall: 58.0 (6.0–130.0) PRT + TMZ: 35.0 (NR) PRT only: 64.0 (NR)	PRT only: 54.0 (45.0–59.4) PRT + concurrent/adjuv-ant TMZ: 56.1 (54.0–60.0) PRT + adjuvant TMZ: 59.4 (54.0–59.4)	Total: 119 PRT + TMZ: 33 PRT only: 86	G2/3: n: 81/38 %: 62/29	mIDH: 81 (68)	Investigate whether administering PRT + TMZ vs. only PRT increases risk of PsP development	-Worse OS and FFP with larger target volumes -Better OS and FFP after 24 months with PsP presence	NR	NR	Increased risk for PsP with: -PRT + TMZ (hazard ratio:2.2) -G3 mIDH-tumors (63% vs. 35% in G2 tumors)
Eichkorn et al. (2023) [15]	R	PRT: 5.1 years (7 months-11 years)	Overall: 54.0 (50.4–60.0) G2: 54.0 (50.4–60.0) G3: 60.0 (54.0–60.0)	Total: 194	G2/3: n:128/66 %:66/34	mIDH: 194 (100)	Evaluate the efficacy, safety and long-term clinical and imaging outcomes of PRT in mIDH G2 and G3 gliomas	5-year OS: G2/3: %: 85/67 5-year PFS: G2/3: %: 60/30	NR	-Acute toxicities, CTCAE < 3a: Fatigue, Headache, Erythema, Focal alopecia -No CTCAE ≥ 3 nor late toxicities	Overall RICE: 25% of the cohort, of which 31% symptomatic and 35% required therapy RICE related to grade, G2/G3: %: 29/17 RICE: No effect on OS or PFS
Ek et al. (2023) [17]	R	Mean follow-up.: PRT: 26.0 (20.0–31.0) Photon: 69.0 (62.0–77.0)	PRT: 54.0 (54.0–60.0) Photon:59.0 (54.0–64.0)	PRT: 44 Photon: 98	PRT G2/3: n: 29/15 %: 66/34 Photon G2/3: n: 40/58 %: 41/59	PRT: -mIDH: 36 (82) Photon: -mIDH: 26 (27) -Unknown: 57 (58)	Compare treatment outcomes, patient characteristics and survival between PRT and photon therapy	Decreased PFS and OS with higher average doses Median PFS and OS not reached	NR	Acute alopecia CTCAE grade 2: PRT: 15.9% vs. Photon: 5.1% No difference in memory impairment in PRT vs. photon Less acute fatigue CTCAE grade 2 with PRT	PsP: -Total: n: 21/126 -PRT: n:4, 3% -Photon: n:17, 13%
Gómez Vecchio et al. (2024) [19]	P	PRT: 12.0 (NR) Photon: 12.0 (NR)	NR	Total: 51 PRT: 32 Photon: 6	G2/3:  n: 33/18 %: 65/35	mIDH: 51 (100)	Report on and find predictors of global health status and fatigue changes from preoperative setting to 12 month postoperatively	NR	PRT: Less unfavourable changes in fatigue No difference in global health status (PRT vs photon)	NR	NR
Qiu et al. (2022) [47]	R	PRT: 21.7 (NR)	G2: 54.0 (54.0) G3: 60.0 (60.0) G4: 60.0 (54.0–60.0)	PRT: 52	G2/3/4: n: 22/25/5 %: 42/48/10	mIDH: 52 (100)	Outcomes of PRT in mIDH gliomas	PFS: 12/24 months: 97.6%/78.4% OS: 12/24 months: 100%/91% Tumor grade significantly impacted both PFS and OS	NR	All acute toxicities: n:40 CTCAE grade 1, Alopecia, n:38 Late toxicities: n:9, CTCAE grade 2 and 3, n:5 including: Dizziness, Fatigue, Memory impairment, Epilepsy	RN:1.9%; observed in oligodendroglio-ma G3 only
Ritterbusch et al. (2021) [50]	R	PRT: 32.0 (17.0–57.0) Photon: NR	53.9 (42.0–60.0)	PRT: 57 Photon: 43	PRT G2/3: n:21/36 %: 37/63 Photon G2/3: n:12/31 %: 28/72	PRT: -mIDH: 49 (86) Photon: -mIDH: 24 (56) -Unknown: 8 (19)	To distinguish patterns of PsP post-PRT vs PsP after photon therapy	No difference in OS between: -Patients with/out PsP post-PRT -PRT and photon	NR	Asymptomatic PsP: n:5/14 Symptomatic PsP: n:9/14, including vision loss, memory loss	PsP: -PRT: n:14, 24.6% of which n:9, 64% were symptomatic -Photon: NR

CTCAE; Common terminology criteria for adverse events, FFP; Freedom from progression, G; Tumor Grade, Gy; Gray, IDH; Isocitrate dehydrogenase, mIDH; mutant- isocitrate dehydrogenase, NR; Not reported, OS; Overall survival, P; Prospective, PFS; Progression Free Survival, PRT; Proton Radiation Therapy, PsP; Pseudoprogression, QoL; Quality of Life, R; Retrospective, RBE; Relative biological effectiveness, RICE; Radiation-induced Contrast Enhancement, RN; Radiation necrosis, TMZ; Temozolomide, vs.; versus

^aRates unspecified for each adverse event

Table 2.

Characteristics of studies on mIDH-inhibitors, n:7

				Number of patients				Main Findings in relation to mIDH-inhibitor administration
Study	Design	mIDH-inhibitor	Median treatment duration (range)	Total, n	Tumor Grade (G),	Previous therapy, n (%)	Aim	Survival	QoL	Toxicities	Treatment Response/ORR (%)
Cho et al. (2022) [9]	R	-Ivosidenib (AG-120) -Vorasidenib (AG-881)	Follow-up: Up to 4 months post-treatment	Ivosidenib: 18 Vorasidenib: 11	G2/3/4: n:19/7/3 %: 66/24/10	NR	To investigate MRI changes in gliomas during mIDH-inhibitor treatment at 3–6 weeks and/or 2–4 months after treatment start	NA	NR	NR	Radiological observations: At 3–6 weeks: Increase in nrCBV and median nrCBV/ADC but not in ADC or FLAIR At 2–4 months: No changes from baseline
Mellinghoff et al. (2021) [32]	P, Open-label	Vorasidenib (AG-881)	Enhancing: 3.3 months (0.2–53.6) Non-enhancing: 26.8 months (1.0–50.9)	Total Glioma: 52 Enhancing: 30 Non-enhancing: 22	G2/3/4: n:25/22/4 %: 48/42/8 Unknown: n:1 %: 2	Enhancing: -RT: 22 (73) -Systemic: 25 (83) -Only Surgery: 4 (13) Non-enhancing: -RT: 8 (36) -Systemic: 14 (64) -Only surgery: 7 (32)	Investigate safety and outcomes of Vorasidenib	Median PFS: -Overall Cohort: 7.5 months -Enhancing: 3.6 months -Non-enhancing: 36.8 months	NR	Any CTCAE, n: 52, most common: Headache 46.2% ALT elevation 44.2% CTCAE ≥ 3, n:10 ALT/AST elevation, Fatigue, Nausea, Seizure, Vomiting, Decreased neutrophil count	Enhancing: ORR (0.0) Non-enhancing: ORR (18.2)
Mellinghoff et al. (2023) [30]	RCT, Open-label	Ivosidenib (AG-120) Vorasidenib (AG-881)	Vorasidenib: 14.3 months (0.9 −22.6) Ivosidenib: 15.1 months (1.8–22.1)	Vorasidenib: 24 Ivosidenib: 25	Vorasidenib: G2/3: n: 22/2 %: 92/8 Ivosidenib: G2/3: n:21/4 %: 84/16	Vorasidenib:-Surgery:24 (100)-RT:7 (29)-Systemic:10 (42)Ivosidenib:-Surgery:25 (100)-RT: 7 (28)-Systemic: 14 (56)	To investigate the mechanism of action of Vorasidenib and Ivosidenib	Median PFS not reached in either mIDH-inhibitor	NR	Vorasidenib: -Any CTCAE, n: 24, most common: Nausea, Headache 41.7%, Diarrhea, Fatigue 29.2% -CTCAE ≥ 3, n:7 Anemia, ALT-elevation, hyperglycemia, hypophosphatemia Ivosidenib: -Any CTCAE, n:25, most common: Headache, Anemia 36.0%, Diarrhea, Seizure 28.0% -CTCAE ≥ 3, n:6, hyponatremia	Vorasidenib −50mg q.d.: ORR (42.9) −10mg q.d.: ORR (10.0) Ivosidenib −500 mg q.d.: ORR (35.7) −250 mg b.i.d.: ORR (12.5) 2-HG concentration reduction: -Vorasidenib 50 mg q.d.: 92.6% reduction -Ivosidenib 500 mg q.d.: 91.1% reduction
Mellinghoff et al. (2023) [31]	RCT, Double-blinded	Vorasidenib (AG-881)	Median follow-up (Interquartile range): Vorasidenib: 14.0 months (10.1 to 17.9) Placebo: 14.3 months (10.0 to 18.1)	Vorasidenib: 168 Placebo: 163	G2: n:331 %: 100	Surgery: 331 (100)	Investigate PFS and time to next intervention in grade 2 tumors receiving Vorasidenib	Vorasidenib vs placebo: -PFS: 27.7 vs. 11.1 months -Time to next intervention: not reached vs. 17.8 months	NR	-Any CTCAE, n:141, most common: ALT elevation 38.9%, Fatigue 32.3% -CTCAE ≥ 3, n:27, ALT/AST/gamma-glutamyl transferase elevation, Fatigue, Diarrhea, Seizure	Vorasidenib 40 mg q.d.: ORR (10.7) Placebo: ORR (2.5)
Natsume et al. (2022) [39]	P, Open-label	Safusidenib (DS-1001)	Enhancing: 7.3 weeks (0.0–190.0) Non-enhancing 91.2 weeks (15.0–207.0)	Total: 47 Enhancing: 35 Non-enhancing: 12	G2/3/4 n:17/23/7 %36/49/15	RT:47 (100)Chemotherapy:38 (81)-TMZ:35 (75)-Nimusutine14 (30)-Bevacizumab 7 (15)	To determine the safety, efficacy and pharmacology of Safusidenib/ DS-1001	Median PFS Non-enhancing vs. Enhancing: not reached vs.10.4 weeks	NR	-All CTCAE, n:45, most common: Skin hyperpigmentation 53.2% Diarrhea 46.8% -CTCAE 3, n: 20, Diarrhea, Arthralgia, headache, decreased neutrophil count, ALT/AST elevation	-Significantly lower 2-HG in Safusidenib (DS-1001) tumor samples -Enhancing: ORR (17.1) (CR: n: 2, 5.7%, PR: n: 4, 11.4%) -Non-enhancing: ORR (33.3) (PR: n: 1, 8.3%, mR: n: 3, 25.0%)
Peters et al. (2023) [43]	R	Ivosidenib (AG-120)	43.7 weeks (5.5- 74.1)	Total: 30 Enhancing: 8 Non-enhancing: 22	G2/3/4: n:21/8/1 %:70/27/3	RT: 8 (27)TMZ: 18(60)mIDH-peptide vaccine: 2 (7)Bevacizumab1 (3)	To investigate the outcomes of Ivosidenib treatment	Deaths: n:2 (of total 30)	NR	CTCAE 1, n:23, most common: Diarrhea 26.7%, Elevated creatine kinase 33.3%	Enhancing: ORR (0.0) Non-enhancing: ORR (36.3)
Wick et al. (2021) [65]	P	BAY1436032	NR	Total glioma:39 Enhancing: 33 Non-enhancing: 2 Unknown: 4 Dose expansion glioma: 25 Dose escalation glioma: 14	NR	Dose expansion: -RT: 23 (92)Dose escalation:NR separately	To investigate the safety profile and pharmacology of BAY1436032 in mIDH solid tumors	PFS rate at 3 months: 0.31	NR	NA due to analysis with other tumor types	ORR: (11) (CR: n:1, 3.0%, PR: n:3, 9.0%)

2-HG; 2-hydroxyglutarate, ADC; Apparent diffusion coefficient, ALT; alanine aminotransferase, AST; aspartate aminotransferase, b.i.d.; bis in die, twice a day, CR; Complete Response, CTCAE; Common terminology criteria for adverse events, FLAIR; Fluid-attenuated inversion recovery, G; Tumor grade, mIDH; mutant-isocitrate dehydrogenase, mR; minor Response, MRI; Magnetic Resonance Imaging, NA; Not Applicable, NR; Not reported, nrCBV; normalized relative cerebral blood volume, ORR; Objective response rate, P; Prospective, PFS; Progression Free Survival, PR; Partial Response, q.d.; quaque die, once a day, QoL; Quality of Life, R; Retrospective, RCT; Randomized controlled trial, RT; Radiotherapy, TMZ; Temozolomide, vs.; versus

Results

Our search identified a total of 6383 articles from PubMed and Scopus with gradual assessments leading to the final inclusion of seven articles per treatment (see Fig. 1).

Proton radiation therapy

Across all included studies, a total of 535 patients received PRT [1, 14, 15, 17, 19, 47, 50]. All but one article had a retrospective study design, and the single prospective study focused on longitudinal changes in patient-reported outcomes where radiation therapy was a covariate [19]. No eligible randomized controlled trial (RCT) was identified. Median follow-up ranged between 12–64 months and median PRT-dose prescribed was 54–60 Gy RBE across studies. For a detailed overview, see Table 1.

Survival

While median overall survival (OS) was not reached in two studies, [15, 17], one study reported a progression free survival (PFS) and OS at 24-months after PRT to 78.4% and 91.0%, respectively [47]. In the same study, tumor grade impacted OS and PFS the most as shown by univariate (OS: p = 0.002, PFS: p = 0.004) and multivariate analysis (OS: p = 0.024, PFS: p = 0.022) [47]. Another study also indicated the importance of tumor grade, with the 5-year OS in grade 2/3 tumors being 85.0%/67.0%, respectively (p = 0.0088) [15]. Similarly, the 5-year PFS in grade 2 and 3 tumors were 60.0% and 30.0%, respectively (p < 0.0001) [15].

Two other studies compared survival in patients receiving PRT versus photon radiation therapy [17, 50]. None of the studies showed differences in OS between radiation modalities. Moreover, the proportion of mIDH-tumors in the photon therapy groups in both studies did not satisfy this review’s eligibility criteria.

Toxicities

Four studies included reports on adverse events [15, 17, 47, 50]. Toxicities mostly occurred early in the post-PRT period and were mild, grading 1–2 on the common terminology criteria for adverse events (CTCAE) (see Table 1). The reporting on specific mild toxicities varied between 1.9% and 73.1%. The most common toxicity was alopecia (15.9–73.1%), but adverse events also included headache, erythema and fatigue. Late, severe toxicities were reported by two studies [47, 50].

The prospective study by Gómez Vecchio et al. indicated a lower proportion of PRT patients experiencing unfavorable changes in fatigue in comparison to photon therapy (66.7% vs. 95.7%, p = 0.03) [19]. No differences related to changes in global health status as a QoL-construct were noted (p = 0.61) [19]. Another study reported a higher rate of fatigue in photon therapy than PRT up to 3 months posttreatment (p = 0.016) thereafter stabilizing in both modalities. The same study reported no difference in subjective memory impairment between radiation modalities [17]. However, the photon therapy group here was not molecularly defined to the same extent, precluding valid conclusions. Other studies reported individual cases developing impaired memory [47, 50].

Radiological manifestations

Pseudoprogression (PsP) was discussed in three studies [14, 17, 50]. In one study, no difference was found in PsP-rate between patients treated with photons versus protons (p = 0.38) [17]. Another study showed PsP in 43.6% (52/119) of PRT-patients with an increased risk of PsP-development in grade 3 versus grade 2 tumors (p = 0.003) [14]. However, more grade 2 tumors here had unknown IDH-status. Therefore, caution should be exerted when interpreting these findings [14]. Ritterbusch et al. observed unique imaging patterns of PsP following PRT. There, PsP developed in 24.6% of patients and PsP-lesions were characterized by being small (< 1 cm), transient, multifocal and patchy with a location at the Bragg peak. The mean time to PsP-occurrence following PRT was 15.4 months while photon-related PsP occurred within 3 months [50]. None of the studies showed association between PsP and tumor subtypes [14, 17, 50]. Finally, Eichkorn et al. assessed radiation induced contrast enhancement (RICE), developing in 25.3% of their overall cohort, and in 29.7% of grade 2 versus 16.7% of grade 3 patients (p = 0.11) [15]. The presence of RICE did not influence OS (p = 0.15) nor PFS (p = 0.67) [15]. However, astrocytomas were associated with earlier RICE development in comparison to oligodendrogliomas (p = 0.047) and RICE risk increased independently with older age (p = 0.017) [15].

Despite the large proportion of IDH-wildtype tumors in the photon therapy group, Acharya et al. reported the 2-year cumulative incidence for radiation necrosis (RN) following PRT to be 18.6% compared with 9.9% after photon therapy (p = 0.16) [1]. Further, the incidence of RN was higher in oligodendrogliomas compared to astrocytomas (p = 0.01) with no difference in radiation modality observed (p = 0.68) [1]. Another study reported a single case of severe RN (CTCAE grade 3) in their cohort, observed in a patient with oligodendroglioma grade 3 treated with PRT [47].

Mutant IDH-inhibitors

Out of seven studies on mIDH-inhibitors, two had a retrospective design, [9, 43], while the rest were prospective studies. A total of 73 patients received Ivosidenib (AG-120), 255 received Vorasidenib (AG-881), 47 received Safusidenib (DS-1001), and 39 received BAY1436032. For a detailed overview, see Table 2.

Progression free survival

None of the seven articles reported on OS while six studies reported PFS [9, 30–32, 39, 65]. In an RCT examining Ivosidenib in non-enhancing LGG, median PFS was not reached after a median postoperative treatment duration of 15.1 months [30]. Another study reported two deaths (of 30 patients in total) after a median time on Ivosidenib of 43.7 weeks [43].

With Vorasidenib, a median PFS of 7.5 months was seen in a broad mIDH glioma cohort, with longer PFS in non-enhancing versus enhancing tumors (36.8 vs. 3.6 months, respectively) [32]. This difference in PFS between enhancing and non-enhancing lesions was seen also with Safusidenib (DS-1001). There, median PFS was 10.4 weeks in enhancing tumors while median PFS was not reached in non-enhancing gliomas after a median treatment duration of 7.3 and 91.2 weeks, respectively [39]. At 3 months, BAY1436032 showed a PFS of 31.0% [65]. The study by Cho et al. included joint data on Ivosidenib and Vorasidenib [9]. A shorter PFS was associated with: 1) continuously decreasing diffusion coefficient on imaging, 2) Fluid-attenuated inversion recovery (FLAIR) volume increase > 4cm³ and 3) elevated perfusion values at 2–4 months posttreatment. However, the authors stress not to interpret this as prognostic effects of mIDH-inhibitors, rather as preliminary radiographic observations [9].

Lastly, in the INDIGO double-blinded RCT, Vorasidenib showed significantly prolonged median PFS versus placebo (27.7 vs. 11.1 months; p < 0.001) in grade 2 non-enhancing tumors after a median follow-up of 14 months [31]. Similarly, Vorasidenib improved median time to next intervention (not reached vs. 17.8 months, p < 0.001) [31].

Toxicities

Mild toxicities (CTCAE grade < 3) dominated in patients treated with mIDH-inhibitors (see Table 2). With Vorasidenib, headache, nausea and liver enzyme elevation were common. Other infrequent toxicities included diarrhea, constipation and electrolyte and/or glucose disturbances [30–32]. Fatigue was reported at rates around 30% and subjective memory impairment at 16.7% [9, 30–32]. Ivosidenib reported analogous toxicity rates to Vorasidenib, but more electrolyte and/or glucose disturbances and anemia were reported with Ivosidenib. Seizure frequency increased in 13.3% of patients and fatigue constituted 12.0% of all toxicities of Ivosidenib [30, 43].

Safusidenib (DS-1001) exclusively caused frequent dermatological complications in up to 53.2% of patients, alopecia/arthralgia in 27.7% and backaches in 21.3%. Diarrhea was more common while rates of headache and nausea were comparable to other mIDH-inhibitors. Limited electrolyte disturbances or elevation of liver enzymes were noted [39]. Toxicities of BAY1436032 were analysed alongside other tumor types and included diarrhea, nausea, fatigue and liver enzyme elevation [65].

Therapeutic response

Neither Vorasidenib nor Ivosidenib showed complete responses (CR). In non-enhancing tumors, the highest objective response rate (ORR) was noted following 50 mg Vorasidenib (ORR 42.9%) [30]. The ORR decreased to 10.0% when administering 10 mg Vorasidenib, indicating a dose–response effect. Another Vorasidenib trial showed an ORR of 18.2% in non-enhancing tumors, with a partial response of 4.5% occurring at 50 mg once daily and minor response of 13.6% at 200 mg once daily [32]. No radiographic response was noted in enhancing tumors [32]. The INDIGO trial reported an ORR of 10.7% (vs. 2.5% in placebo) with the administration of 40 mg Vorasidenib once daily [31]. Thus, the range of ORR with Vorasidenib across studies was 10.0–42.9%, at least partially explained by administered doses. Lastly, a 92.6% reduction in 2-HG concentrations in resected tumor tissues was noted in addition to decreased tumor proliferation, more activated immune cells and suppressed genes related to cell-cycle progression. Of note, these analyses were performed in patients who received Vorasidenib 50 mg once daily for 28 (+ 7) days, up until surgery [30].

Ivosidenib also showed better response in non-enhancing versus enhancing tumors (ORR 36.3% vs. 0.0%) [43]. With the administration of 500 mg Ivosidenib once daily, an ORR of 35.7% was reported. The ORR decreased almost threefold with the administration of 250 mg Ivosidenib twice a day instead (ORR 12.5%) [30]. The concentrations of 2-HG in resected tumor tissues were consequently decreased by 91.1% following the administration of 500 mg Ivosidenib once daily for 28 (+ 7) days, up until surgery [30].

Among all mIDH-inhibitors, Safusidenib (DS-1001) reported the highest CR of 5.7% and exclusively in enhancing tumors. Following Safusidenib (DS-1001) administration twice per day in 21-day cycles, the ORR observed was 33.3% in non-enhancing tumors versus 17.1% in enhancing tumors. Further, tumor samples of patients still receiving Safusidenib (DS-1001) but underwent surgery upon progression showed lower 2-HG concentrations after treatment [39]. BAY1436032 specifically targeted enhancing tumors and demonstrated CR in 3.0% of cases [65].

Discussion

In this scoping review, we demonstrate promising results on surrogate outcomes following both PRT and mIDH-inhibitors. Although the studies involving mIDH-inhibitors were more often prospective and randomized than in PRT, some of the same shortcomings exist. Thus far, no solid evidence exists with regards to patient-centred outcomes such as improved survival, QoL or spared cognitive function following these therapies especially in relation to other treatment options.

Patient-centred outcomes

To date, only PFS and time to next intervention has been reported for mIDH-inhibitors. Substantial data on OS, QoL and cognition is lacking. Survival comparisons between PRT and photon therapy were limited by the low proportion of mIDH-tumors in the photon groups, often included at an earlier time-point [17, 50]. It is yet to be explored what influences neurocognition in mIDH-LGG; whether it is the disease itself, the treatments applied or a combination of both [11, 58, 59, 67]. Dosimetric studies assessing PRT showed a comparable target volume to photon therapy and delivery of lower radiation doses to organs at risk [2, 22, 68]. This spared critical neurological structures, including areas of neurogenesis (subventricular zone and hippocampus), thereby showing potential in decreasing long term sequelae and improving QoL [2, 22]. Only one included study associated PRT with less unfavourable changes in fatigue affecting QoL [19]. Other included studies reported fatigue, memory impairment and other adverse events without elaborating on specific QoL-domains. Studies not included in this review, assessed QoL with PRT and showed development of neuroendocrine deficiencies with patients retaining cognitive function overtime [55, 57]. In pediatric populations, PRT was associated with less cognitive impairment [20, 33, 38, 45].

Surrogate end-points

Surrogate end-points applied in this context included PFS and dosimetric advantages. In recent years, cancer drugs showing only a PFS benefit were approved by the Food and Drug Administration [44]. Due to limited therapies for mIDH-LGG, the application of surrogate end-points might accelerate the availability of novel treatments in the market. However, the long-term benefits of these treatments remain unknown and surrogate end-points do not necessarily reflect on outcomes that matter to patients [52, 64].

It remains unclear whether an improved PFS translates into improved survival, QoL or cognition. Deterioration in clinical symptoms does not always correlate with pre-defined radiological progression criteria [34, 52]. A comparable case involved Bevacizumab in glioblastoma-treatment which, despite a beneficial PFS, did not ultimately prolong OS nor improve QoL but was more frequently associated with serious toxicities [8, 12, 18]. Similarly, another trial comparing early versus delayed radiotherapy in LGG showed a PFS benefit with early post-surgical radiotherapy that could not, however, be translated to improved OS [63]. Even in grade 2 mIDH-glioma, predicting OS or post-progression survival based upon PFS has proven difficult [52]. The dosimetric advantages of PRT may be offset by post-PRT manifestations such as PsP, RICE or RN, especially as these per se may be difficult to differentiate from true tumor progression, thereby leading to unnecessary, and sometimes deleterious, interventions due to misdiagnosis [1, 16, 24, 40, 59]. Perhaps assessing clinical outcomes could better guide treatment decision-making [52].

Post-PRT radiological manifestations

Contrast-enhancing brain lesions, RN, PsP and RICE were apparent in several included studies [1, 14, 15, 17, 47, 50]. However, the heterogeneity of the results, populations and terminology used to describe these manifestations complicates comparisons and prevents conclusions from being drawn [17, 27, 29]. These inconsistencies in reporting were even apparent in other studies not included in this review.

In similarity to the included study by Acharya et al. [1], another study, not included in this review according to the eligibility criteria, reported a higher proportion of tissue necrosis following PRT compared to photon therapy (18% vs. 6% respectively) [66]. However, consistent with findings from another included study [17], Bronk et al. observed no difference in PsP rate between radiation modalities [6].

Among the included studies, PsP was more frequently observed in grade 3 tumors while RICE rate was higher in grade 2 tumors [14, 15]. PsP was not associated with tumor subtypes [14, 17, 50]. However, astrocytomas showed earlier RICE development, while oligodendroglioma patients -regardless of treatment modality- expressed a higher incidence of RN with one case of severe RN following PRT in grade 3 oligodendroglioma [1, 15, 47]. Despite not meeting this review’s eligibility criteria, Bronk et al. showed earlier occurrence of PsP in oligodendroglioma following PRT and no correlation between PsP rate and tumor grade [6]. In another photon-dominated cohort, all RN cases following PRT occurred in patients with oligodendroglioma [3]. However, too few patients were present in both studies to be able to conclude on this much debated topic [3, 6]. Older age was shown to be an independent risk factor for RICE in one included study [15]. While oligodendrogliomas usually present at higher median ages [25], another study showed slightly higher RICE rate in astrocytomas (50% vs. 46% in oligodendrogliomas) [16]. Hence, inconclusive research is currently present regarding associations between tumor subtype and post-PRT manifestations.

Limitations

Our search was dominated by single-institutional, retrospective, uncontrolled studies with small sample sizes, mixed IDH-mutational status and short follow-ups. Nonetheless, this reflects the present state of evidence. The latest 2021 World Health Organization classification of tumors was not always applied [28]. For PRT, selection was biased to younger patients, less pre-radiation cognitive impairment and lower tumor grades with lower administered doses in accordance with the current guidelines in order to minimize neurocognitive effects [1, 17, 19]. Further, inconsistent definitions and different nomenclature used to describe post-PRT radiological manifestations prevents current comparisons between proton studies further complicating the clinical application of these manifestations and the gathering of results on common ground [17, 27, 29].

For mIDH-inhibitors, prior therapies and short follow-ups complicate effects on survival, while placebo cross-over in the INDIGO-trial will complicate long term survival analysis. In the INDIGO trial, patients receiving placebo could cross-over to Vorasidenib upon progression. Given the aim to delay radio-chemotherapy, the threshold to receive Vorasidenib as second-line therapy for placebo patients was probably lower than to receive radio-chemotherapy, which was the option in the Vorasidenib arm of this trial [31]. Further, the INDIGO trial compared treatment-naïve patients receiving Vorasidenib monotherapy to placebo [31]. This may limit comparisons to the current standards of LGG treatment. On the other hand, it provides the opportunity for studying a watch-and-wait strategy that is applied in selected centres following surgical resection and compare it with active treatment with presumably less neurotoxic effects [5, 13, 31, 32].

Despite the current caveats, the cumulative evidence suggests that lower-grade, non-enhancing tumors are responsive to mIDH-inhibitors with regards to progression and metabolic alterations such as reduced 2-HG concentrations. For mIDH-LGGs having undergone malignant transformation, studies incorporating combination therapies and trials in enhancing tumors are coveted. Previous studies in acute myeloid leukemia showed significant advantage when mIDH-inhibitors were combined with the hypomethylating agent, Azacitidine [10, 37]. Interestingly, Safusidenib (DS-1001) and BAY1436032 produced CR in enhancing tumors, even providing preliminary prospects for systematic targeted therapy in mIDH gliomas of higher grade as well [39, 65].

Future directions

Assessing OS in mIDH-LGGs might take decades, but relying on surrogate end-points is only a short-term solution. Given the long-expected survival of patients, often undergoing multiple interventions over time, therapies directed at the early stages of disease should prove long-term benefits [52]. Prospective studies with longer follow-ups examining patient-centred domains are, therefore, highly encouraged. Awaiting us in the field of PRT in mIDH-glioma are three prospective, randomized trials; PRO-GLIO, NRG-BN005, and APPROACH; leveraging on survival and cognitive outcomes [21, 23, 56]. Additionally, the production of consensus-based definitions may ease the reporting, analysis and application of post-PRT radiological changes in future studies, a work that has already been started by the European Particle Therapy Network [27]. The long-term OS, cognitive and QoL analyses from the INDIGO trial and further trials on Safusidenib (DS-1001) and Vorasidenib through the VIGOR trial are even forecasted [31, 46, 54]. Finally, to facilitate the holistic reporting and analysis of trial outcomes aligning with patient concerns and priorities, the development of core outcome sets in future trials concerning these treatments is worth considering. So far for adult gliomas, the COBra-study developed a core outcome set where some of the main outcome domains included survival, health-related QoL and cognitive function [36, 48, 49].

Conclusion

PRT and mIDH-inhibitors are two therapies promising precision in targeting mIDH-LGG. Current reports lack profound assessments of patient-centred outcomes despite their prognostic importance in the young, affected patient group. Prospective, patient-centred studies, preferably RCTs, with longer follow-ups and larger populations, are strongly encouraged to aid treatment decisions.

Supplementary information

Below is the link to the electronic supplementary material.

Abbreviations

2-HG

D-2-Hydroxyglutarate

CR

Complete Response

CTCAE

Common terminology criteria for adverse events

IDH

Isocitrate-dehydrogenase

LGGs

Lower grade gliomas

mIDH-inhibitors

Mutant isocitrate-dehydrogenase-inhibitors

ORR

Objective response rate

OS

Overall survival

PFS

Progression free survival

PRT

Proton radiation therapy

PsP

Pseudoprogression

QoL

Quality of life

RCT

Randomized controlled trial

RICE

Radiation induced contrast enhancement

RN

Radiation necrosis

Author contributions

A.S.J. conceptualized and designed the framework of this scoping review. A.C. and D.H conducted the literature search and screening of titles and abstracts. A.C., A.S.J. and D.H. were involved in the eligibility assessment of full-texts. D.H. drafted the manuscript. A.C., A. M., A.S.J., M.B. and P.B. critically revised the manuscript. All authors reviewed and approved the final version of the manuscript for publication.

Funding

Open access funding provided by University of Gothenburg. A.S.J. received grants from the Swedish state under the agreement between the Swedish government and the county councils, the ALF-agreement (ALFGBG-1006089). M.B. received grants from Jubileumsklinikens Cancerfond, the Swedish Society of Medicine and the Gothenburg Society of Medicine. The funders had no role in the preparation of this scoping review.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethical approval

No ethical approval from an ethics committee or institutional review board was required as no access to sensitive patient information nor any direct involvement of human/animal participants was included in this review.

Consent to participate

Not applicable.

Consent to publish

Not applicable.

Competing interests

Author P.B. declares honoraria for lectures, consultation or advisory board participation from Servier, Ekspertpanelet and Moloklinikken.

The remaining authors have no conflicts of interest to declare that are relevant to the content of this article.