Distinguishing Edema from Tumor Infiltration in High-Grade Glioma Patients: Initial Insights from quantitative MRI using Relaxometry

doi:10.21203/rs.3.rs-3434280/v1

Download PDF

Research Article

Distinguishing Edema from Tumor Infiltration in High-Grade Glioma Patients: Initial Insights from quantitative MRI using Relaxometry

https://doi.org/10.21203/rs.3.rs-3434280/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Introduction: The differentiation between edema and tumor infiltration in high-grade gliomas is fundamental for surgical planning. However, this distinction is challenging using conventional magnetic resonance imaging (MRI). Relaxometry is a new technique for quantitative diagnosis using MRI, currently under assessment for some neurological disorders.

Methods: In this study, we report our initial experience using T2 multiecho relaxometry to differentiate perilesional edema from tumor infiltration in patients with high-grade gliomas. In order to have values for a comparative analysis, we assumed T2-hyperintensity surrounding the enhancing tumor represented vasogenic edema on meningiomas or metastasis, while it could be vasogenic edema or tumor infiltration on high-grade gliomas.

Results: Twenty patients with high-grade gliomas and 10 patients with metastases or meningiomas were included. Images were analyzed using Relaxo software. Mean T2 value in regions of hypersignal for the metastasis group was 196.8, compared to 407.3 in high-grade gliomas.

Discussion: Difference was statistically significant (p<0.05). Our results suggest that the LNI relaxo software could be a helpful tool to differentiate edema and tumor infiltration in patients with high-grade gliomas, allowing for a more efficient preoperative planning and postoperative assessment of tumor infiltration resection rate.

Glioma infiltration

Edema

T2 relaxometry

Magnetic Resonance Imaging

Differentiating between edema and tumor infiltration in glioma patients through magnetic resonance imaging (MRI) can be challenging. Existing literature suggests that the T2 hyperintensity surrounding the tumor in patients with high-grade gliomas (HGG) includes both vasogenic edema and extensive tumor cell infiltration, whereas in patients with meningiomas or metastases, it primarily represents vasogenic edema [1]. Identifying these morphologically distinct structures is crucial for more precise preoperative planning and postoperative assessment of the extent of tumor resection. Moreover, accurately distinguishing the tumoral component within the non-enhancing lesion area is of significant clinical importance, especially when evaluating therapy response in HGG, given the frequent use of anti-angiogenic agents. The ability to differentiate vasogenic edema from infiltrative tumor could also have a positive impact on improving radiotherapy planning.

While several studies have explored differentiating between vasogenic-like edema and infiltrative tumor in HGG using MRI, there are limitations to current approaches. Some of these studies have utilized advanced imaging methods to classify the lesions that may not be universally available [1]. Additionally, a few studies have employed diffusion imaging to analyze the entire lesion area, but this approach is less suitable for HGG due to its high heterogeneity and is not readily applicable in routine clinical practice (2.3). Also, limited studies employed voxel-based classifications, although they usually can’t capture the heterogeneity of HGG or provide spatially relevant information (4).

MRI scans can be mapped by advanced techniques that quantify the relaxation time of different tissues, such as vasogenic edema and tumor infiltration, providing accurate information about tissue composition. This project utilized the T2 multiecho MRI relaxometry technique along with specialized software to compare two groups of patients. The first group consisted of high-grade glioma patients who typically exhibit areas of edema and tumor infiltration that are challenging to differentiate. The second group included patients with meningioma and metastases, where the T2 hypersignal is known to represent edema. To accomplish this, we employed the Relaxo LNI software, enabling the evaluation of multiecho T2 relaxometry in patients with brain tumors and/or edema [2].

2.1. Patient selection

This is a retrospective cohort study based on data obtained from electronic medical records of patients diagnosed with HGG, meningiomas, and metastases available in the database of the Neuroimaging Laboratory of the University of Campinas (UNICAMP) from January 2012 to 2022.

Patients were divided in high-grade gliomas (glioma group) and metastases or meningiomas (non-glioma group) were included. All patients with HGG had a confirmed histopathological diagnosis based on WHO Classification, and all patients had a preoperative MRI scan. Patients with medical history of other brain diseases or incomplete data recorded were excluded. Data regarding brain tumor site, recurrence, anatomopathological results, date of the last T2 MRI were obtained.

2.2. Compliance with Ethical Standards

This study was approved by the Research Ethics Committee of our institution under the number CAAE 60250722.9.0000.5404. It was carried out in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. All persons gave their informed consent prior to their inclusion in the study. Patients' identity was preserved at all stages of the study.

2.3. Software Relaxo LNI

The Relaxo LNI software (https://www.lniunicamp.com/relaxometry) was used, an easy-to-use tool developed by the Neuroimaging Laboratory of UNICAMP to perform the analysis of relaxometry with multiple T2 echoes of the patients in the sample. The software processes T2 multi-echo images from the image in DICOM (Digital Imaging and Communications in Medicine) format, estimating a T2 map of the entire brain, in addition to enabling manual segmentation of sub-regions. Users can add a T1WI to register with T2 and draw Region of interest (ROIs) using one of the following options as background image: T2 (any echo), T1WI registered (if added), T2 with inverse contrast or quantized map T2. The software exports a quantitative table of the segmented area (*.xlsx format) and the quantified maps and drawn ROIs [10].

Multiecho T2 values relate to the analyzed brain tissue. Initially, it was used to contribute to the imaging diagnosis of patients with epilepsy to differentiate ROIs that eventually present gliosis, cortical dysplasia and developmental tumors [11]. The multiecho T2 value is specific to each tissue type. When defining the ROI, precautions were taken to exclude structures that exhibit contrast enhancement, such as blood vessels or the choroid plexus.

The software utilized in the study includes a tool that identifies regions with the highest hypersignal based on graphical representations and colors. The intensity of the color corresponds to the strength of the hypersignal, as depicted in Figures 1 and 2.

Figure 1. (a) T2-weighted MRI scans of patients with HGG, metastases, or meningiomas were selected. In the software, it is possible to highlight the image in red with white highlighting of the hypersignal region. (b) Next, we delimited the hypersignal region of the lesions. (c) Finally, a T2 relaxometry value and a histogram of the ROI are generated.

Figure 2. Example of ROI placed at the (a) center and (b) periphery of lesion patients in the high-grade glioma group and the corresponding intensity histogram.

After selecting the sample subjects, we manually delineated ROIs representing hypersignals on multiecho T2-weighted images of the included patients using Relaxo LNI. Subsequently, we conducted a comparative analysis between patient groups regarding edema and tumor infiltration.

2.4. T2 - Relaxometry

Relaxometry is the measurement of relaxation times on MRI (T1 or T2). Multiecho T2 relaxometry involves measuring the rate of relaxation within the ROI. The relaxation rate is a combination of exponential decay curves partly due to different water compartments in the tissue, but to simplify a first approximation using a single curve is used. While a minimum of three data points is required to define an exponential curve, more data points are preferred for accuracy. This process is applied to each selected ROI, resulting in a specific value [3].

To differentiate between glioma infiltration and vasogenic edema, we identified the voxel with the highest value on multiecho T2-weighted images within each chosen ROI for both groups.

2.5. Statistical methods

T2 values data were analyzed in IBM SPSS 20 with sex and age as covariates for possible statistical corrections of the samples. The Mann-Whitney U test was used as a two-tailed test to compare the median of each group, and is a non-parametric test applied to two independent samples. P values less than 0.05 were considered statistically significant.

We analyzed 20 patients with HGG, 6 patients with transitional meningioma and 4 patients harboring metastasis (total of 10 patients in non-glioma group). The mean age in the glioma group was 54 years, while in the non-glioma group, it was 57 years old. The most prevalent comorbidity in all patients was epilepsy, present in 20% of patients.

Mean T2 values in regions of hypersignal in Relaxo software for the non- glioma group were 196.8, with a median of 199,18 and a standard deviation (SD) of 19,18. In the glioma group, the value was 407.3, with a median of 371,37 and a SD of 154,33. In the boxplot (Figure 3), it is possible to visualize the distribution of T2 values in the group of patients with HGG and in the non-glioma group. T2 values data were analyzed in IBM SPSS 20 with sex and age as covariates for possible statistical corrections of the samples. In the pairwise test (pairwise comparison), a significant difference was found in the multiecho T2 values between the glioma and non- glioma groups, with significance < 0.05.

Figure 3. Boxplot showing the distribution of T2 values by group.

Mean T2 relaxation times were longer for the high-grade glioma group when compared to the control group (p<0.05). The mean relaxation times of the central areas of HGGs were also longer (mean value 498.68), where tumor infiltration is the most common component, in relation to the peripheral areas (mean value 136.13), p< 0.05, where edema tends to be predominant.

Figure 4 shows that the T2 values of the peripheral areas of glioma are closer to the values of metastasis and meningiomas.

Figure 4. Boxplot showing comparison between the glioma group, the periphery of the glioma and the non-glioma group.

The T2 values obtained in the investigation of the neuroimaging exams of the patients in the present study are shown in table 1.

	Birth date (AGE)	Sex	Histological type	T2 mean values
Patient 1	25/04/1948	F	glioblastoma (grade 4)	295
Patient 2	06/06/1943	F	glioblastoma (grade 4)	274,33
Patient 3	31/08/1963	M	glioblastoma (grade 4)	366,14
Patient 4	27/02/1975	M	glioblastoma (grade 4)	278,27
Patient 5	31/03/1982	M	glioblastoma (grade 4)	660
Patient 6	1/12/1960	F	glioblastoma (grade 4)	376,6
Patient 7	22/06/1958	M	glioblastoma (grade 4)	383,84
Patient 8	24/12/1982	F	glioblastoma (grade 4)	601,48
Patient 9	20/04/1973	F	glioblastoma (grade 4)	223
Patient 10	22/12/1959	F	glioblastoma (grade 4)	300,67
Patient 11	9/11/1946	M	glioblastoma (grade 4)	709,43
Patient 12	12/11/1985	M	glioblastoma (grade 4)	501,92
Patient 13	19/04/1973	F	anaplastic astrocytoma (grade 3)	618,09
Patient 14	05/10/1978	F	anaplastic astrocytoma (grade 3)	222
Patient 15	16/09/1957	M	anaplastic astrocytoma (grade 3)	519,13
Patient 16	09/03/1979	M	anaplastic astrocytoma (grade 3)	258,18
Patient 17	05/03/1968	M	anaplastic astrocytoma (grade 3)	481,47
Patient 18	08/04/1978	M	anaplastic astrocytoma (grade 3)	268,02
Patient 19	18/01/1981	M	anaplastic astrocytoma (grade 3)	496,33
Patient 20	03/09/1984	M	anaplastic astrocytoma (grade 3)	312,68
Patient 21	12/02/1984	F	transitional meningioma (grade 1)	197,36
Patient 22	16/02/1960	M	transitional meningioma (grade 1)	174,42
Patient 23	08/04/1971	F	transitional meningioma (grade 1)	160
Patient 24	21/02/1965	M	transitional meningioma (grade 1)	191,2
Patient 25	18/03/1951	M	transitional meningioma (grade 1)	204,07
Patient 26	07/06/1971	F	transitional meningioma (grade 1)	229,42
Patient 27	7/10/1964	M	metastasis (adenocarcinoma)	212
Patient 28	22/10/1967	F	metastasis (adenocarcinoma)	194,65
Patient 29	18/10/1957	F	metastasis (adenocarcinoma)	204,27
Patient 30	12/09/1964	M	metastasis (melanoma)	201

Table 1. The anatomopathological classification was made according to the WHO brain tumor classification.

Differentiating tumor infiltration from brain edema is essential to adequately treat patients with HGG, but it presents challenges due to the limitations of standard MRI techniques. However, precise target delineation is crucial for effective treatment planning, allowing maximizing the extent of surgical resection as well as defining the high-dose field in radiation therapy [5]. From a clinical perspective, the interface between the contrast-enhanced region of the glioma and the surrounding peritumoral region holds significant importance [12]. Surgical interventions strive to achieve maximal tumor resection while minimizing potential postoperative complications for the patient. Nonetheless, despite aggressive resection attempts, tumor recurrence frequently occurs at the resection margin [12]. Currently, conventional MRI techniques widely employed in routine practice do not offer the capability to visualize the non-enhancing components of HGGs.

Although there are no large randomized controlled trials to date, there is strong evidence supporting a positive correlation between the extent of tumor resection and prognosis in patients with HGG [4,5,6]. Consequently, surgical interventions strive to achieve maximal tumor resection while minimizing potential postoperative complications for the patient. [7]. More than removing all contrast-enhancing areas, infiltrating tumor cells ideally should also be removed [8, 12]. Nonetheless, despite aggressive resection attempts, tumor recurrence frequently occurs at the resection margin [13]. Currently, conventional MRI techniques widely employed in routine practice do not offer the capability to visualize the non-enhancing components of HGGs. Some advanced imaging modalities are currently being assessed for creating infiltration probability maps, but still with limitations preventing it from being widely used [1]. In this study, we describe our initial experience with a promising MRI-based tool, T2 relaxometry, to distinct brain edema versus neoplastic infiltration in high-grade gliomas.

Previous studies also using relaxometry showed that T2 values of HGG are higher than T2 values of low-grade gliomas, possibly due to hypermetabolism [9]. Another study used a T2 mapping technique to quantify edema reduction in recurrent glioblastoma treated with bevacizumab, and the result suggested a correlation between post-treatment T2 values and survival from disease progression [10]. Blystad et al, in a quantitative T1 and T2 mapping survey, showed that tissue changes in the peritumoral region that are not visible on conventional MRI can be detected, describing relaxation values in peritumoral edema with a heterogeneous pattern in the first 10 mm of the portion of the contrast-enhanced tumor, with a gradient in relaxation values from the contrast-enhanced part of the tumor to the peritumoral edema [14]. Note that T2 mapping by software is a non-invasive test that can provide additional information about the composition of the analyzed lesion.

Recently, Blystad et al. also showed that the relaxometry detects contrast enhancement in the peritumoral area of malignant gliomas, which can represent infiltrative tumor growth [14]. These observations align with prior research. Quantitative MRI, such as T1- and T2-relaxometry, have demonstrated their ability to identify subtle tissue alterations within the peritumoral edema of HGGs before surgery, as evidenced by previous study by Ellingson et al. [15]. Furthermore, these techniques have proven valuable in detecting early tissue changes associated with glioblastoma recurrence during follow-up, even before such changes become apparent in conventional imaging [16]. Moreover, the identification of non-visible peritumoral contrast enhancement is consistent with the discoveries made previously by Müller et al. [17] who employed quantitative T1-mapping to detect peritumoral contrast enhancement that remained imperceptible in standard MRI scans. Their work also demonstrated that a reduction in this imperceptible contrast enhancement during treatment was indicative of a favorable treatment response.

The MRI technique employed in this study is quantitative, contrasting with conventional MRI, which depends on visual evaluation. The advantage of a quantitative technique lies in its measurability and the foundation it offers for longitudinal comparisons. In this study, we used Relaxo LNI software to delimit the central ROIs of tumors and peritumoral areas. Relaxation times of T2 mapping in HGG differed from those in patients with pure vasogenic edema.

The tumors have a heterogeneous appearance on conventional MRI, and also exhibit different relaxation patterns depending on the tumor region. Mean relaxation times of HGG showed a variable T2 curve progression from the center of the tumor to the periphery. T2 values in the center of the tumor in HGG were higher than in the periphery, which is in accordance with the tumor's behavior. It is expected to find more infiltration and tumor in the core and more edema in the boundaries, supporting the idea that T2 relaxometry can differentiate between tumor infiltration and edema. Another factor in favor of our software is that the values in the periphery of the glioma group were closer to those of the non-glioma group with pure vasogenic edema.

The study has some limitations that should be acknowledged. Firstly, the sample size in this study was relatively small, which may have limited the generalizability of the findings. Additionally, this study had a retrospective design, which can introduce inherent biases and limitations associated with the collection and availability of historical data. Despite these limitations, the promising results obtained in this study provide valuable insights and motivation for future research with a larger and more diverse cohort of participants. A larger sample size would allow for more robust statistical analyses and enhance the generalizability of the findings. Next steps include determining cut-off values in the relaxation time where it is safe to state the absence of tissue infiltrated by tumor cells, contributing to preoperative planning and postoperative evaluation of the extent of resection of these lesions.

Our findings indicate that the LNI Relaxo software holds promise as a valuable tool for distinguishing between edema and tumor infiltration in individuals with high-grade gliomas. This discovery can be interpreted as a quantitative reflection of tumor infiltration. As a result, this method shows potential for improving surgical planning, equipping surgeons to achieve maximal tumor resection while preserving regions primarily affected by peritumoral edema, ultimately leading to improved patient outcomes. Future investigations may yield additional insights into the diagnostic and prognostic utility of T2 relaxometry mapping within the field of neuroimaging.

Funding and acknowledgments

This work was supported by São Paulo Research Foundation (FAPESP, grant number 2021/12749-2) and the Universtiy Clinics Hospital of Campinas State University.

Acknowledgements

We would like to thank the support from the staff of Clinics Hospital of the State University of Campinas.

Competing interests

The authors have no relevant financial or non-financial interests to disclose

Ethics approval and consent to participate

Artzi M, Bokstein F, Blumenthal DT, Aizenstein O, Liberman G, Corn BW, Ben Bashat D (2014) Differentiation between vasogenic-edema versus tumor-infiltrative area in patients with glioblastoma during bevacizumab therapy: a longitudinal MRI study. Eur J Radiol. ;83(7):1250–1256. doi: 10.1016/j.ejrad.2014.03.026. Epub 2014 Apr 12. PMID: 24809637
Pavlisa G, Rados M, Pavlisa G, Pavic L, Potocki K, Mayer D (2009) Mar-Apr;33(2):96–101 The differences of water diffusion between brain tissue infiltrated by tumor and peritumoral vasogenic edema. Clin Imaging. 10.1016/j.clinimag.2008.06.035. PMID: 19237051
Kinoshita M, Goto T, Okita Y, Kagawa N, Kishima H, Hashimoto N, Yoshimine T (2010) Diffusion tensor-based tumor infiltration index cannot discriminate vasogenic edema from tumor-infiltrated edema. J Neurooncol 96(3):409–415. 10.1007/s11060-009-9979-0Epub 2009 Aug 21. PMID: 19696968
Matthieu Lê Hervé, Delingette J, Kalpathy-Cramer E, Gerstner HA, Shih et al (Sep 2014) Multimodal Analysis of Vasogenic Edema in Glioblastoma Patients for Radiotherapy Planning. Workshop Image-Guided Adaptive Radiation Therapy. Boston, United States. ⟨hal-01093397⟩
Jena R, Price SJ, Baker C, Jefferies SJ, Pickard JD, Gillard JH, Burnet NG (2005) Diffusion tensor imaging: possible implications for radiotherapy treatment planning of patients with high-grade glioma. Clin Oncol (R Coll Radiol). ;17(8):581 – 90. 10.1016/j.clon.2005.04.012. PMID: 16372482
Shah AS, Yahanda AT, Sylvester PT, Evans J, Dunn GP, Jensen RL, Honeycutt J, Cahill DP, Sutherland GR, Oswood M, Shah M, Abram SR, Rich KM, Dowling JL, Leuthardt EC, Dacey RG, Kim AH, Zipfel GJ, Limbrick DD Jr., Smyth MD, Leonard J, Chicoine MR (2021) 'Using Histopathology to Assess the Reliability of Intraoperative Magnetic Resonance Imaging in Guiding Additional Brain Tumor Resection: A Multicenter Study', Neurosurgery, 88(1), E49+, available: https://link.gale.com/apps/doc/A672656501/AONE?u=anon~30c40209&sid=googleScholar&xid=25c54508 [accessed 04 Oct 2023]
Li YM, Suki D, Hess K, Sawaya R (2016) The influence of maximum safe resection of glioblastoma on survival in 1229 patients: Can we do better than gross-total resection? J Neurosurg 124(4):977–988 Epub 2015 Oct 23. PMID: 26495941
Saitta L, Heese O, Forster AF, Matschke J, Siemonsen S, Castellan L et al (2021) A intensidade do sinal na ressonância magnética T2' está relacionada ao grau de glioma cerebral. Eur Radiol 1068–1076. 10.1007/s00330-010-2004-3
Ellingson BM, Cloughesy TF, Lai A, Nghiemphu PL, Lalezari S, Zaw T et al (2012) Quantificação da redução do edema usando mapeamento de relaxometria quantitativo diferencial T2 (DQT2) em glioblastoma recorrente tratado com bevacizumabe. J Neurooncol. Janeiro; 106 (1):111–9. 10.1007/s11060-011-0638-x
Giacomini L, de Souza JPSA, Formentin C, de Campos BM, Todeschini AB, de Oliveira E, Tedeschi H, Joaquim AF, Cendes F, Ghizoni E (2020) Temporal lobe structural evaluation after transsylvian selective amygdalohippocampectomy. Neurosurgical Focus FOC 48(4):E14. https://doi.org/10.3171/2020.1.FOCUS19937
Molinaro AM, Hervey-Jumper S, Morshed RA et al (2020) Association of Maximal Extent of Resection of Contrast-Enhanced and Non–Contrast-Enhanced Tumor With Survival Within Molecular Subgroups of Patients With Newly Diagnosed Glioblastoma. JAMA Oncol 6(4):495–503. 10.1001/jamaoncol.2019.6143
Petrecca K, Guiot MC, Panet-Raymond V, Souhami L (2013) Failure pattern following complete resection plus radiotherapy and temozolomide is at the resection margin in patients with glioblastoma. J Neurooncol. ;111(1):19–23. 10.1007/s11060-012-0983-4. Epub 2012 Oct 10. PMID: 23054563
Blystad I, Warntjes JBM, Smedby Ö et al (2020) Quantitative MRI using relaxometry in malignant gliomas detects contrast enhancement in peritumoral oedema. Sci Rep 10:17986. https://doi.org/10.1038/s41598-020-75105-6
Ellingson BM, Lai A, Nguyen HN, Nghiemphu PL, Pope WB, Cloughesy TF (2015) Quantification of Nonenhancing Tumor Burden in Gliomas Using Effective T2 Maps Derived from Dual-Echo Turbo Spin-Echo MRI. Clin Cancer Res 21(19):4373–4383. 10.1158/1078-0432.CCR-14-2862Epub 2015 Apr 21. PMID: 25901082; PMCID: PMC4592379
Lescher S, Jurcoane A, Veit A, Bähr O, Deichmann R, Hattingen E (2015) Quantitative T1 and T2 mapping in recurrent glioblastomas under bevacizumab: earlier detection of tumor progression compared to conventional MRI. Neuroradiology 57(1):11–20. 10.1007/s00234-014-1445-9Epub 2014 Oct 7. PMID: 25287076
Müller A, Jurcoane A, Kebir S, Ditter P, Schrader F, Herrlinger U, Tzaridis T, Mädler B, Schild HH, Glas M, Hattingen E (2017) Quantitative T1-mapping detects cloudy-enhancing tumor compartments predicting outcome of patients with glioblastoma. Cancer Med 6(1):89–99 Epub 2016 Nov 28. PMID: 27891815; PMCID: PMC5269700

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Distinguishing Edema from Tumor Infiltration in High-Grade Glioma Patients: Initial Insights from quantitative MRI using Relaxometry

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and methods

3. Results

4. Discussion

5. Conclusion

Declarations

References

Additional Declarations

Status:

Version 1