Impact of fractionated stereotactic radiotherapy on activity of daily living and performance status in progressive/recurrent glioblastoma: a retrospective study

Demogeot, Nicolas; Salleron, J.; Rech, Fabien; Taillandier, Luc; Royer, P.; Vogin, Guillaume

doi:10.1186/s13014-022-02169-1

Cited by 3 publications

(2 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[4] Moreover, in clinical practice, when managing recurrent glioblastoma, their performance status can significantly influence therapeutic decision-making, which may involve options like surgical re-intervention, re-irradiation, or best supportive care. [22] Due to the limited availability of publicly accessible, precise clinical databases, research predicting the course of performance status in patients with malignant tumors is much scarcer compared to the development of prediction models for overall survival or progression-free survival. However, in recent years, several studies have reported that machine learning approaches have successfully predicted changes in KPS 6 months postoperatively in patients with glioblastoma, as well as poor performance status in patients with cancer 6 months after diagnosis.…”

Section: Discussionmentioning

confidence: 99%

Postoperative Karnofsky performance status prediction in patients withIDHwild-type glioblastoma: a multimodal approach integrating clinical and deep imaging features

Sasagasako,

Ueda,

Mineharu

et al. 2024

Preprint

View full text Add to dashboard Cite

Background and Purpose: Glioblastoma is a highly aggressive brain tumor with limited survival that poses challenges in predicting patient outcomes. The Karnofsky Performance Status (KPS) score is a valuable tool for assessing patient functionality and contributes to the stratification of patients with poor prognoses. This study aimed to develop a 6-month postoperative Karnofsky Performance Status (KPS) prediction model by combining clinical data with deep learning-based image features from pre- and postoperative MRI scans, offering enhanced personalized care for glioblastoma patients. Materials and Methods: Using 1,476 MRI datasets from the Brain Tumor Segmentation Challenge 2020 public database, we pretrained two variational autoencoders (VAEs). Imaging features from the latent spaces of the VAEs were used for KPS prediction. Neural network-based KPS prediction models were developed to predict scores below 70 at 6 months postoperatively. In this retrospective single-center analysis, we incorporated clinical parameters and pre- and postoperative MRI images from 150 newly diagnosed IDH wild-type glioblastoma, divided into training (100 patients) and test (50 patients) sets. In training set, the performance of these models was evaluated using the area under the curve (AUC), calculated through fivefold cross-validation repeated 10 times. The performance of these models in the training set was evaluated using the area under the curve (AUC) from a fivefold cross-validation repeated 10 times. The final evaluation of the developed models assessed in the test set. Results: Among the 150 patients, 61 had 6-month postoperative KPS scores below 70 and 89 scored 70 or higher. We developed three models: a clinical-based model, an MRI-based model, and a multimodal model that incorporated both clinical parameters and MRI features. In the training set, the mean AUC was 0.785±0.051 for the multimodal model, which was significantly higher than the clinical-based model (0.716±0.059, P=0.038) using only clinical parameters and MRI-based model (0.651±0.028, P<0.001) using only MRI features. In the test set, the multimodal model achieved an AUC of 0.810, outperforming the clinical-based (0.670) and MRI-based (0.650) models. Conclusion: The integration of MRI features extracted from VAEs with clinical parameters in the multimodal model substantially enhanced KPS prediction performance. This approach has the potential to improve prognostic prediction, paving the way for more personalized and effective treatments for patients with glioblastoma. Abbreviations: KPS, Karnofsky performance status. IDH, isocitrate dehydrogenase. VAE, variational autoencoder. BraTS, Brain Tumor Segmentation challenge

show abstract

Section: Discussionmentioning

confidence: 99%

Postoperative Karnofsky performance status prediction in patients withIDHwild-type glioblastoma: a multimodal approach integrating clinical and deep imaging features

Sasagasako,

Ueda,

Mineharu

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…The treatment was described in a previous study conducted in our hospital [20], with fSRT delivered by a dedicated stereotactic radiation therapy machine (Cyberknife® (Accuray Inc.)), with 5 sessions of 5 Gy, distributed over 5-7 days, prescribed to the 80% normalized isodose line. The gross tumor volume (GTV) corresponded to the T1-enhanced lesion(s) on a recent brain MRI (maximum 4 weeks before treatment), matched with simulation computed tomography (CT).…”

Section: Methodsmentioning

confidence: 99%

Pattern of recurrence after fractionated stereotactic reirradiation in adult glioblastoma

Margulies,

Sahki,

Rech

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Background Glioblastomas all eventually relapse after initial treatment, and an option to treat these recurrences is fractionated stereotactic reirradiation (fSRT). The location of recurrences following reirradiation has not been studied for fSRT delivered by a dedicated stereotactic device. We aimed to analyze these locations to better elucidate safety margins, dose and fractionation regimens. Methods We retrospectively analyzed the data of patients with glioblastoma recurrence that had been reirradiated by fSRT in October 2010-December 2020, in 25 Gy in 5 fractions delivered by a CyberKnife® at Institut de Cancérologie de Lorraine. We matched the images of the relapse post-fSRT relapse with the stereotactic radiation treatment planning scan to determine the relapse location. Results The location of recurrences after fSRT was “out-field” in 43.5%, “marginal” in 40.3%, and “in-field” in 16.1% of patients (N = 62). A GTV-PTV margin of 1 mm (versus 2 or 3 mm, HR = 0.38 [0.15–0.95], p = 0.037) and a PTV volume of ≥ 36cc (HR = 5.18 [1.06–25.3], p = 0.042) were significantly associated with the “marginal” recurrences. Being ≥ 60 years old at initial treatment (HR = 3.06 [1.17–8.01], p = 0.023) and having one or more previous recurrences (HR = 5.29 [1.70–16.5], p = 0.004) were significantly associated with “out-field” recurrences. The median PFS from fSRT was 3.4 months, and OS from diagnosis and from fSRT were 25.7 and 10.8 months respectively. Conclusion Reirradiation of glioblastoma recurrence by fSRT with 25 Gy in 5 fractions provides good local control.

show abstract

Postoperative Karnofsky performance status prediction in patients with IDH wild-type glioblastoma: A multimodal approach integrating clinical and deep imaging features

Sasagasako,

Ueda,

Mineharu

et al. 2024

PLoS ONE

View full text Add to dashboard Cite

Background and purpose Glioblastoma is a highly aggressive brain tumor with limited survival that poses challenges in predicting patient outcomes. The Karnofsky Performance Status (KPS) score is a valuable tool for assessing patient functionality and contributes to the stratification of patients with poor prognoses. This study aimed to develop a 6-month postoperative KPS prediction model by combining clinical data with deep learning-based image features from pre- and postoperative MRI scans, offering enhanced personalized care for glioblastoma patients. Materials and methods Using 1,476 MRI datasets from the Brain Tumor Segmentation Challenge 2020 public database, we pretrained two variational autoencoders (VAEs). Imaging features from the latent spaces of the VAEs were used for KPS prediction. Neural network-based KPS prediction models were developed to predict scores below 70 at 6 months postoperatively. In this retrospective single-center analysis, we incorporated clinical parameters and pre- and postoperative MRI images from 150 newly diagnosed IDH wild-type glioblastoma, divided into training (100 patients) and test (50 patients) sets. In training set, the performance of these models was evaluated using the area under the curve (AUC), calculated through fivefold cross-validation repeated 10 times. The final evaluation of the developed models assessed in the test set. Results Among the 150 patients, 61 had 6-month postoperative KPS scores below 70 and 89 scored 70 or higher. We developed three models: a clinical-based model, an MRI-based model, and a multimodal model that incorporated both clinical parameters and MRI features. In the training set, the mean AUC was 0.785±0.051 for the multimodal model, which was significantly higher than the AUCs of the clinical-based model (0.716±0.059, P = 0.038) using only clinical parameters and the MRI-based model (0.651±0.028, P<0.001) using only MRI features. In the test set, the multimodal model achieved an AUC of 0.810, outperforming the clinical-based (0.670) and MRI-based (0.650) models. Conclusion The integration of MRI features extracted from VAEs with clinical parameters in the multimodal model substantially enhanced KPS prediction performance. This approach has the potential to improve prognostic prediction, paving the way for more personalized and effective treatments for patients with glioblastoma.

show abstract

Impact of fractionated stereotactic radiotherapy on activity of daily living and performance status in progressive/recurrent glioblastoma: a retrospective study

Cited by 3 publications

References 51 publications

Postoperative Karnofsky performance status prediction in patients withIDHwild-type glioblastoma: a multimodal approach integrating clinical and deep imaging features

Postoperative Karnofsky performance status prediction in patients withIDHwild-type glioblastoma: a multimodal approach integrating clinical and deep imaging features

Pattern of recurrence after fractionated stereotactic reirradiation in adult glioblastoma

Postoperative Karnofsky performance status prediction in patients with IDH wild-type glioblastoma: A multimodal approach integrating clinical and deep imaging features

Contact Info

Product

Resources

About