The present investigation is a systematic review and meta-analysis conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

The review question was formulated according to the PICO (P: patients; I: intervention; C: comparison; O: outcomes) scheme, as follows: in case of newly diagnosed or recurrent glioblastoma multiforme (P), is 5-ALA (I) a useful tool for increasing the extent of resection (O), considering the tumor fluorescence positivity and the false-positive rates (C)?

Three different medical databases (PubMed, Scopus, and Cochrane Library) were screened using the following search terms: “5-ALA”, “aminolevulinic”, “recurrent glioma”, “recurrent glioblastoma”, “radionecrosis”, “pseudoprogression”, “glioma recurrence”, “glioblastoma recurrence”, “radiation necrosis” [MeSH], combined using Boolean operators “AND” and “OR”.

Titles and abstracts were screened in the first search round. In the second round, full text of eligible papers and their reference lists (forward search) were evaluated. Then, papers were considered for data reporting and availability in the third round of search. Studies matching our inclusion criteria were finally included in the present systematic review and meta-analysis. Two authors (GDP and AS) independently conducted the first two search rounds, and any discordance was solved by consensus with a third senior author (LR).

Comparative studies in English language on newly diagnosed and recurrent GBM, reporting data on the intraoperative 5-ALA fluorescence status (positive/negative), and histological findings of 5-ALA-positive regions were considered for eligibility. True positive was considered as 5-ALA-positive fluorescence and histological confirmation of GBM; false positive, instead, as 5-ALA-positive fluorescence and no histological diagnosis of GBM. Reviews, case reports, letters, technical notes, video articles, and studies on pediatric population (<18 years old) were not considered. The inclusion and exclusion critera are summarized in Table 1 .

Included studies were screened for the number of newly diagnosed and recurrent GBM patients, their demographics, type of operative microscope, and percentage of true and false positives in each group. Furthermore, histological findings in case of false-positive samples were also collected.

We performed a meta-analysis of proportions to estimate the pooled rates of each outcome. Proportion meta-analyses were not used when the frequency of an outcome was reported in <1% of the sample [raw proportions and 95% confidence interval (95% CI) were reported in such cases], and a random-effects model was adopted to account for the inter-study heterogeneity.

A total of 448 titles and abstracts were firstly screened. Eighteen papers were considered as eligible, and their full text was evaluated. After the full-text analysis and the forward search, 3 papers were finally included for meta-analysis. The search strategy is summarized in Figure 1 . Papers excluded with reason after their full-text examination are summarized in Supplementary Table S1 .

We included 331 patients from 3 studies, 1 retrospective (10) and 2 prospective (11, 12).

A newly diagnosed GBM was reported in 212 patients while a recurrent GBM in 119. Patient’s gender and age were reported in 1 out of the 3 (33%) included studies.

All patients received histological diagnosis according to the WHO 2016 guidelines. All patients who underwent second surgeries for GBM recurrence received the Stupp protocol between the two procedures.

The use of 5-ALA fluorescence was used to perform multiple biopsies in each included study, and the definition of “false positive” was homogeneously considered as the absence of GBM cells at least in one of the analyzed samples.

The type of operative microscope used for conducting the surgical procedures was reported in 1 out of the 3 included studies, and it was Pentero (Carl Zeiss, Oberkochen, Germany).

A positive 5-ALA fluorescence was reported in 169 (79.7%) out of the 212 newly diagnosed GBM and in 89 (74.8%) out of the 119 recurrent GBM.

The rate of positive 5-ALA fluorescence was not significantly different between newly diagnosed and recurrent GBM [p = 0.45; odds ratio (OR): 1.23; 95% CI: 0.72–2.09; I² = 0%]. The heterogeneity was very low, confirming the reliability of data ( Figure 2 ).

The histopathology did not confirm the presence of GBM cells in 7 (4.1%) out of the 169 5-ALA fluorescence-positive newly diagnosed cases and in 9 (10.1%) out of the 89 5-ALA fluorescence-positive recurrent GBM.

The incidence of false positives was significantly higher in recurrent GBM cases (p = 0.04; OR: 0.24; 95% CI: 0.06–0.91; I² = 19%) with a relatively low heterogeneity rate.

Histological report was available in 13 (81.3%) out of the 16 false-positive cases. It reported “abnormal brain tissue characterized by reactive astrocytes and scattered inflammatory cells” in 11 (68.8%), “normal brain tissue” in 1 (6.3%), “infiltrating neutrophils” in 1 (6.3%), and not reported in 3 (18.8%). No cases of RN were reported in false-positive cases among the recurrent GBMs ( Figure 3 ).

PP is defined as peritumoral inflammatory tissue that may be erroneously reported as tumor progression or recurrence on MRI images (18). The PP phenomenon is well documented, and its MRI aspect is justified by its histological findings, consisting of inflammation, neutrophil and macrophage infiltration, reactive astrocytes, high mitotic rate, and neoangiogenesis (19). This may justify the 5-ALA fluorescence in these peritumoral areas, as confirmed in our results. In fact, “abnormal brain tissue characterized by reactive astrocytes and scattered inflammatory cells” was found in 68.8% of false-positive specimens and “infiltrating neutrophils” in 6.3%. The histological report was “normal brain” in 6.3%, while it was not available in 18.8% of the included cases. Accordingly, inflammatory findings were found in 82.1% of false-positive samples, confirming that PP could represent the underlying mechanism of 5-ALA accumulation in non-malignant GBM peritumoral areas.

Post-surgical or post-radiation infiltration of reactive astrocytes (responsive astrocytosis), immune cell presence, and 5-ALA extracellular accumulation can be responsible for false-positive cases. Indeed, histiocytes/macrophages (which have function of phagocytosis) and lymphocytes can internalize 5-ALA: this may lead to a significant buildup of porphyrin precursors, making the tissue fluorescent (8). Regarding 5-ALA leakage and extracellular fluorescence accumulation, the latter is true mostly for lesions with pronounced perifocal edema.

RN is a relatively frequent finding in postoperative follow-up imaging from GBM patients. It consists of local necrosis and fibrous tissue spreading, usually contiguously to the surgical cavity. Although RN consists of a relatively low-activity metabolic tissue, its enlargement may determine some grade of compression on the surrounding brain tissue, resulting in inflammatory phenomena (20). Conversely, RN is not supposed to be 5-ALA positive due to its metabolic status and histological characteristics (8).

This meta-analysis confirms this evidence, as no cases of RN were reported among false-positive samples.

These data are of importance, as there is not a clear picture about its behavior in literature with 5-ALA. Indeed, an aspect to be critically appraised concerning this issue, especially in oldest reports, is the fact that neuropathological examination in many reports could underestimate the presence of tumor cells. As a matter of fact, in many cases, hematoxylin–eosin staining depicting reactive changes only can be actually converted to the diagnosis of infiltrating tissue after additional immunohistochemistry investigations. Furthermore, while RN can occur in a relatively small area, the surrounding tissue has been radiated itself, then some grade of radiation-induced inflammation has to be considered. Therefore, while proper RN tissue should not display fluorescence, the surrounding brain may be characterized by inflammatory phenomena and histologically related findings and hence might display some degree of 5-ALA fluorescence.

The present analysis demonstrated that the incidence of false positives in recurrent glioma surgery is not negligible. This is especially true when PP is considered for differential diagnosis. PP has been reported to occur predominantly (almost 60% of cases) within the first 3 months after completing adjuvant treatments, although it may occur later, as reported after medical administration of lomustine and temozolomide. In addition, methylguanine-DNA methyltransferase (MGMT) methylation tumor’s status has been associated with PP occurrence (21). Hence, if timing and molecular status are consistent, it is important for the neurosurgeon performing 5-ALA-guided resections to be aware of concerns regarding the accurate diagnosis of this phenomenon and that fluorescence status might hinder non-neoplastic tissue. On the other hand, in this setting, non-tumor-related 5-ALA positivity can be used as a guide to target surgical excision to areas of inflammatory infiltrations or reactive gliosis when surgery is indicated for edema relief.

Nevertheless, 5-ALA in the recurrent setting is valuable to the surgeon when RN is suspected: as demonstrated in the present analysis, 5-ALA-related fluorescence is rather more specific. Hence, this can provide surgeon guidance for proper histopathological sampling to increase diagnostic yield and tailor resection.

Indeed, 5-ALA can provide a feasible guidance especially in the recurrent setting, where MRI-based information often fails to preoperatively identify proper oncological tissue.

The “real-life” intraoperative picture corresponds to a substantial lack of textural feedback (tumor is often friable) and presence of scar, inflammation, neo-angiogenesis, and gliosis that further compromise a surgeon’s ability to distinguish tumor from non-tumoral tissue. This corresponds to a mixture of reactive/regressive tissue that shows areas with different degrees or absence of fluorescence along with proper neoplastic tissue that display fluorescence (17).

Therefore, a robust awareness of 5-ALA potentialities and pitfalls in recurrent glioma surgery is therefore paramount for a cognizant surgical strategy, especially when in proximity to eloquent brain areas where oncological benefit and functional cost associated with an aggressive resection should be considered.