Efficacy and safety of CyberKnife radiosurgery in elderly patients with brain metastases: a retrospective clinical evaluation

Predictors of local control after robotic stereotactic radiotherapy for brain metastases: 10-years-experience after Cyberknife installation

BACKGROUND

To evaluate the factors influencing brain metastases (BM) local control (LC) after stereotactic radiotherapy (SRT).

METHODS

Between 2010 and 2020, a cohort of 145 patients (246 BM) treated consecutively with robotic radiosurgery was analysed.

RESULTS

Median age was 61 years (range, 29-90 years). Median radiological follow-up of the lesions was 21.7 months (range, 3-115 months). The mean overall survival and LC were 33.0 and 82.7 months, respectively. On univariate analysis, sex, primary cancer site, histological type, use of systemic steroids, maximum diameter, volume, early MRI response, isodose line, number of fractions, BED10 value, and BED10 value proportional to volume and maximum diameter were significant factors for LC. On multivariate analysis, female sex (hazard ratio [HR]: 2.10 P: 0.035), adenocarcinoma histology (HR: 6.54 P: 0.001), no steroid use (HR: 3.60 P: 0.001), maximum diameter (≤1 cm) (HR: 2.64 P: 0.018), complete response of lesion at first follow-up MRI compared to stable or progressive disease (HR: 4.20, P = 0.024; HR: 19.15, P < 0.001), isodose line (≥90%) (HR: 2.00 P: 0.036), and tumour volume (PTV ≤2 cc) (HR: 5.19 P: 0.001) were independent factors improving LC.

CONCLUSIONS

SRT is an effective treatment for patients with a limited number of BM with a high LC rate. There are many factors related to the patient, tumour, and radiotherapy plan that have an impact on LC after SRT in brain metastases. These results warrant further investigation in a prospective setting.

Clinical Results and Hematologic Predictors of Linear Accelerator-Based Stereotactic Radiosurgery or Fractionated Stereotactic Radiotherapy for Brain Metastasis in Patients Aged 75 Years or Older: A Retrospective Study

OBJECTIVE

This study aimed to evaluate prognostic factors including pre-radiosurgical blood count in elderly patients (EPs) with brain metastasis (BM) who were treated using linear accelerator (LINAC)-based stereotactic radiosurgery (SRS) and fractionated stereotactic radiotherapy (fSRT) with a micro-multileaf collimator.

METHODS

Between January 2011 and November 2021, 101 consecutive EPs with BM were treated by LINAC-based SRS or fSRT using LINAC with a micro-multileaf collimator. EPs were defined as patients aged ≥75 years.

RESULTS

The tumors originated from the lungs (n = 90; 89.1%), colon (n = 2; 2.0%), and others (n = 9; 8.8%) in these EPs. The median pretreatment Karnofsky Performance Status was 80 (range, 40-100). The median follow-up time was 10 months (range, 0-76), as was the median survival. The 6-month, 1-year, and 2-year survival in the EP group was 58.3%, 43.2%, and 28.5%, respectively. Freedom from local failure at 6 months and 1 and 2 years was 97%, 95%, and 91.5%, respectively. Freedom from distant failure at 6 months and 1 and 2 years in EPs was 70.6%, 59.4%, and 54.2%, respectively. A high neutrophil/lymphocyte ratio >5.33 was an unfavorable predictor of prognosis for EPs with BMs treated with SRS and fSRT (P < 0.001). In the EPs, the prognostic factors associated with prolonged survival in the Cox proportional hazards model were being female and a good pretreatment Karnofsky Performance Status.

CONCLUSIONS

The findings of our study highlight the efficacy of LINAC-based SRS and fSRT with a micro-multileaf collimator in the treatment of EPs with BMs. Neutrophil/lymphocyte ratio can be an important factor in treatment decisions for EPs with BMs.

Single isocenter dynamic conformal arcs-based radiosurgery for brain metastases: Dosimetric comparison with Cyberknife and clinical investigation

Purpose

To compare the dosimetric quality of automatic multiple brain metastases planning (MBM) with that of Cyberknife (CK) based on the clinical tumor condition, such as the tumor number, size, and location.

Methods

76 treatment plans for 46 patients treated with CK were recalculated with the MBM treatment planning system. Conformity index (CI), homogeneity index (HI), gradient index (GI), lesion underdosage volume factor (LUF), healthy tissue overdose volume factor (HTOF), geometric conformity index (g) and mean dose to normal organs were compared between CK and MBM for tumor number, size, shape and distance from the brainstem or chiasm.

Results

The results showed that the mean brain dose was significantly smaller in MBM than CK. CI did not differ between MBM and CK; however, HI was significantly more ideal in CK (p = 0.000), and GI was significantly smaller in MBM (P = 0.000). LUF was larger in CK (p = 0.000) and HTOF and g was larger in MBM (p = 0.003, and 0.012). For single metastases, CK had significantly better HTOF (p = 0.000) and g (p = 0.002), but there were no differences for multiple tumors. Brain dose in MBM was significantly lower and CI was higher for tumors < 30 mm (p = 0.000 and 0.000), whereas HTOF and g for tumors < 10 mm were significantly smaller in CK (p = 0.041 and p = 0.016). Among oval tumors, brain dose, GI and LUF were smaller in MBM, but HTOF and g were smaller in CK. There were no particular trends for tumors close to the brainstem, but HTOF tended to be smaller in CK (0.03 vs. 0.29, p = 0.068) for tumors inside the brainstem.

Conclusions

MBM can reduce the brain dose while achieving a dose distribution quality equivalent to that with CK.

Survey of Transfer Learning Approaches in the Machine Learning of Digital Health Sensing Data

Machine learning and digital health sensing data have led to numerous research achievements aimed at improving digital health technology. However, using machine learning in digital health poses challenges related to data availability, such as incomplete, unstructured, and fragmented data, as well as issues related to data privacy, security, and data format standardization. Furthermore, there is a risk of bias and discrimination in machine learning models. Thus, developing an accurate prediction model from scratch can be an expensive and complicated task that often requires extensive experiments and complex computations. Transfer learning methods have emerged as a feasible solution to address these issues by transferring knowledge from a previously trained task to develop high-performance prediction models for a new task. This survey paper provides a comprehensive study of the effectiveness of transfer learning for digital health applications to enhance the accuracy and efficiency of diagnoses and prognoses, as well as to improve healthcare services. The first part of this survey paper presents and discusses the most common digital health sensing technologies as valuable data resources for machine learning applications, including transfer learning. The second part discusses the meaning of transfer learning, clarifying the categories and types of knowledge transfer. It also explains transfer learning methods and strategies, and their role in addressing the challenges in developing accurate machine learning models, specifically on digital health sensing data. These methods include feature extraction, fine-tuning, domain adaptation, multitask learning, federated learning, and few-/single-/zero-shot learning. This survey paper highlights the key features of each transfer learning method and strategy, and discusses the limitations and challenges of using transfer learning for digital health applications. Overall, this paper is a comprehensive survey of transfer learning methods on digital health sensing data which aims to inspire researchers to gain knowledge of transfer learning approaches and their applications in digital health, enhance the current transfer learning approaches in digital health, develop new transfer learning strategies to overcome the current limitations, and apply them to a variety of digital health technologies.

Comparison of the accuracy of Monte Carlo and Ray Tracing dose calculation algorithms for multiple target brain treatments on CyberKnife

Single plan multiple brain targets (MBT) stereotactic radiosurgery dose difference between Monte Carlo (MC) and Ray Tracing (RT) algorithms has not been studied. A retrospective study and dose measurements were performed to access factors influencing dose differences. Fifty-three RT treatment plans with a total of 209 brain metastases were extracted from Precision Treatment Planning System (TPS). These plans were generated using fixed cones and were delivered using the CyberKnife M6 system. The same treatment plans were recalculated using MC algorithm and keeping the beam parameters unchanged. MC calculated plan parameters were extracted and dose differences were normalised to MC calculated dose. Correlations were investigated. RT and MC calculated off-centre-ratio (OCR) and tissue-phantom-ratio (TPRs) were exported from the TPS and compared with measured. Plans with 5 gross tumour volumes (GTVs) were created on a phantom and dose measured using a CC04 ionisation chamber and microdiamond detector for comparison with calculated doses. Calculated and measured TPR agreed within ± 1% beyond depth of maximum dose. The OCR showed differences up to 4.3% in the penumbra and out-of-field (OOF) regions. Largest RT and MC calculated GTV mean dose difference was - 5.7%. An increase in the number of GTVs and reduction in the geometric separation of metastases were associated with increased differences between RT and MC calculated doses. In conclusion, calculated dose disagreement in MBT depends on the number of GTVs per plan, number of GTVs within a certain separation distance and plan complexity. MC dose calculation is recommended for complex CyberKnife SRS of MBT.

How we treat octogenarians with brain metastases

Biologically younger, fully independent octogenarians are able to tolerate most oncological treatments. Increasing frailty results in decreasing eligibility for certain treatments, e.g., chemotherapy and surgery. Most brain metastases are not an isolated problem, but part of widespread cancer dissemination, often in combination with compromised performance status. Multidisciplinary assessment is key in this vulnerable patient population where age, frailty, comorbidity and even moderate additional deficits from brain metastases or their treatment may result in immobilization, hospitalization, need for nursing home care, termination of systemic anticancer treatment etc. Here, we provide examples of successful treatment (surgery, radiosurgery, systemic therapy) and best supportive care, and comment on the limitations of prognostic scores, which often were developed in all-comers rather than octogenarians. Despite selection bias in retrospective studies, survival after radiosurgery was more encouraging than after whole-brain radiotherapy. Prospective research with focus on octogenarians is warranted to optimize outcomes.

What if: A retrospective reconstruction of resection cavity stereotactic radiosurgery to mimic neoadjuvant stereotactic radiosurgery

Introduction

Neoadjuvant stereotactic radiosurgery (NaSRS) of brain metastases has gained importance, but it is not routinely performed. While awaiting the results of prospective studies, we aimed to analyze the changes in the volume of brain metastases irradiated pre- and postoperatively and the resulting dosimetric effects on normal brain tissue (NBT).

Methods

We identified patients treated with SRS at our institution to compare hypothetical preoperative gross tumor and planning target volumes (pre-GTV and pre-PTV) with original postoperative resection cavity volumes (post-GTV and post-PTV) as well as with a standardized-hypothetical PTV with 2.0 mm margin. We used Pearson correlation to assess the association between the GTV and PTV changes with the pre-GTV. A multiple linear regression analysis was established to predict the GTV change. Hypothetical planning for the selected cases was created to assess the volume effect on the NBT exposure. We performed a literature review on NaSRS and searched for ongoing prospective trials.

Results

We included 30 patients in the analysis. The pre-/post-GTV and pre-/post-PTV did not differ significantly. We observed a negative correlation between pre-GTV and GTV-change, which was also a predictor of volume change in the regression analysis, in terms of a larger volume change for a smaller pre-GTV. In total, 62.5% of cases with an enlargement greater than 5.0 cm3 were smaller tumors (pre-GTV < 15.0 cm3), whereas larger tumors greater than 25.0 cm3 showed only a decrease in post-GTV. Hypothetical planning for the selected cases to evaluate the volume effect resulted in a median NBT exposure of only 67.6% (range: 33.2-84.5%) relative to the dose received by the NBT in the postoperative SRS setting. Nine published studies and twenty ongoing studies are listed as an overview.

Conclusion

Patients with smaller brain metastases may have a higher risk of volume increase when irradiated postoperatively. Target volume delineation is of great importance because the PTV directly affects the exposure of NBT, but it is a challenge when contouring resection cavities. Further studies should identify patients at risk of relevant volume increase to be preferably treated with NaSRS in routine practice. Ongoing clinical trials will evaluate additional benefits of NaSRS.

Stereotactic radiotherapy of brain metastases: clinical impact of three-dimensional SPACE imaging for 3T-MRI-based treatment planning

Purpose

For planning CyberKnife stereotactic radiosurgery (CK SRS) of brain metastases (BM), it is essential to precisely determine the exact number and location of BM in MRI. Recent MR studies suggest the superiority of contrast-enhanced 3D fast spin echo SPACE (sampling perfection with application-optimized contrast by using different flip angle evolutions) images over 3D gradient echo (GE) T1-weighted MPRAGE (magnetization-prepared rapid gradient echo) images for detecting small BM. The aim of this study is to test the usability of the SPACE sequence for MRI-based radiation treatment planning and its impact on changing treatment.

Methods

For MRI-based radiation treatment planning using 3T MRI in 199 patients with cerebral oligometastases, we compared the detectability of BM in post-gadolinium SPACE images, post-gadolinium MPRAGE images, and post-gadolinium late-phase MPRAGE images.

Results

When SPACE images were used for MRI-based radiation treatment planning, 29.8% and 16.9% more BM, respectively, were detected and included in treatment planning than in the post-gadolinium MPRAGE images and the post-gadolinium late-phase MPRAGE images (post-gadolinium MPRAGE imaging: n_total = 681, mean ± SD 3.4 ± 4.2; post-gadolinium SPACE imaging: n_total = 884, mean ± SD 4.4 ± 6.0; post-gadolinium late-phase MPRAGE imaging: n_total = 796, mean ± SD 4.0 ± 5.3; P_{post-gadolinium SPACE imaging versus post-gadolinium MPRAGE imaging} < 0.0001, P_{post-gadolinium SPACE imaging versus post-gadolinium late-phase MPRAGE imaging}< 0.0001).

Conclusion

For 3T MRI-based treatment planning of stereotactic radiosurgery of BM, we recommend the use of post-gadolinium SPACE imaging rather than post-gadolinium MPRAGE imaging.

Dynamic 18F-FET PET/CT to differentiate recurrent primary brain tumor and brain metastases from radiation necrosis after single-session robotic radiosurgery

OBJECTIVE

Cyberknife robotic radiosurgery (RRS) provides single-session high-dose radiotherapy of brain tumors with a steep dose gradient and precise real-time image-guided motion correction. Although RRS appears to cause more radiation necrosis (RN), the radiometabolic changes after RRS have not been fully clarified. ¹⁸F-FET-PET/CT is used to differentiate recurrent tumor (RT) from RN after radiosurgery when MRI findings are indecisive. We explored the usefulness of dynamic parameters derived from ¹⁸F-FET PET in differentiating RT from RN after Cyberknife treatment in a single-center study population.

METHODS

We retrospectively identified brain tumor patients with static and dynamic ¹⁸F-FET-PET/CT for suspected RN after Cyberknife. Static (tumor-to-background ratio) and dynamic PET parameters (time-activity curve, time-to-peak) were quantified. Analyses were performed for all lesions taken together (TOTAL) and for brain metastases only (METS). Diagnostic accuracy of PET parameters (using mean tumor-to-background ratio >1.95 and time-to-peak of 20 min for RT as cut-offs) and their respective improvement of diagnostic probability were analyzed.

RESULTS

Fourteen patients with 28 brain tumors were included in quantitative analysis. Time-activity curves alone provided the highest sensitivities (TOTAL: 95%, METS: 100%) at the cost of specificity (TOTAL: 50%, METS: 57%). Combined mean tumor-to-background ratio and time-activity curve had the highest specificities (TOTAL: 63%, METS: 71%) and led to the highest increase in diagnosis probability of up to 16% p. - versus 5% p. when only static parameters were used.

CONCLUSIONS

This preliminary study shows that combined dynamic and static ¹⁸F-FET PET/CT parameters can be used in differentiating RT from RN after RRS.

Applying pytorch toolkit to plan optimization for circular cone based robotic radiotherapy

Background

Robotic linac is ideally suited to deliver hypo-fractionated radiotherapy due to its compact head and flexible positioning. The non-coplanar treatment space improves the delivery versatility but the complexity also leads to prolonged optimization and treatment time.

Methods

In this study, we attempted to use the deep learning (pytorch) framework for the plan optimization of circular cone based robotic radiotherapy. The optimization problem was topologized into a simple feedforward neural network, thus the treatment plan optimization was transformed into network training. With this transformation, the pytorch toolkit with high-efficiency automatic differentiation (AD) for gradient calculation was used as the optimization solver. To improve the treatment efficiency, plans with fewer nodes and beams were sought. The least absolute shrinkage and selection operator (lasso) and the group lasso were employed to address the “sparsity” issue.

Results

The AD-S (AD sparse) approach was validated on 6 brain and 6 liver cancer cases and the results were compared with the commercial MultiPlan (MLP) system. It was found that the AD-S plans achieved rapid dose fall-off and satisfactory sparing of organs at risk (OARs). Treatment efficiency was improved by the reduction in the number of nodes (28%) and beams (18%), and monitor unit (MU, 24%), respectively. The computational time was shortened to 47.3 s on average.

Conclusions

In summary, this first attempt of applying deep learning framework to the robotic radiotherapy plan optimization is promising and has the potential to be used clinically.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13014-022-02045-y.

The Time for Chronotherapy in Radiation Oncology

Five decades ago, Franz Halberg conceived the idea of ​​a circadian-based therapy for cancer, given the differential tolerance to treatment derived from the intrinsic host rhythms. Nowadays, different experimental models have demonstrated that both the toxicity and efficacy of several anticancer drugs vary by more than 50% as a function of dosing time. Accordingly, it has been shown that chemotherapeutic regimens optimally timed with the circadian cycle have jointly improved patient outcomes both at the preclinical and clinical levels. Along with chemotherapy, radiation therapy is widely used for cancer treatment, but its effectiveness relies mainly on its ability to damage DNA. Notably, the DNA damage response including DNA repair, DNA damage checkpoints, and apoptosis is gated by the circadian clock. Thus, the therapeutic potential of circadian-based radiotherapy against cancer is mainly dependent upon the control that the molecular clock exerts on DNA repair enzymes across the cell cycle. Unfortunately, the time of treatment administration is not usually considered in clinical practice as it varies along the daytime working hours. Currently, only a few studies have evaluated whether the timing of radiotherapy affects the treatment outcome. Several of these studies show that it is possible to reduce the toxicity of the treatment if it is applied at a specific time range, although with some inconsistencies. In this Perspective, we review the main advances in the field of chronoradiotherapy, the possible causes of the inconsistencies observed in the studies so far and provide some recommendations for future trials.