Patients’ Selection
In this study,63 patients with multiple brain metastases had undergone HA-WBRT + SIB between January 2016 and December 2020 at the Department of Radiation Oncology of the Third Hospital of Zhangzhou City and XiaMen ChangGung Hospital in the observation group: Patients received HA-WBRT (30Gy in 12 fractions, the maximum dose of the hippocampus ≤ 14Gy) and a SIB (48Gy in 12 fractions).The control group(a HA-WBRT retrospective cohort):189 patients being retrospectively selected had been treated with WBRT between January 2016 and December 2020.among them, 98 patients had received HA-WBRT and 91 patients were treated with WBRT without Hippocampus-Avoidance, who were not fit for further analysis. There were 56 patients in each group( HA-WBRT + SIB,n = 56;HA-WBRT, n = 56) through 1:1 propensity score matching(Fig. 1).There was significant difference in the small cell lung cancer(SCLC)(HA-WBRT + SIB,n = 7>HA-WBRT, n = 42,χ2 = 210.00,P<0.0001).therefore,patients with SCLC were excluded. Primary focuses of all patients were confirmed by pathology and enhanced-brain-MRI
-proven multiple brain metastases(range 4–15), without hippocampal metastases or 7mm away from hippocampus. Those patients with one to three brain metastases and maximum diameter of BM less than 5mm were excluded. Those patients who had no clear records at diagnosis or incomplete treatment were also removed. Baseline characteristics of patients was listed Table 1. The average characteristics of the patients include age, gender, karnofsky performance status(KPS score), number and maximum size of brain metastasis, brain edema, symptoms of BM, primary tumor, type of radiotherapy, extracranial metastases and systemic treatment.
Table 1
Dose constraints of Organs-at-Risk
Organs at Risk
|
Dose constraint
|
Hippocampus
|
Dmax ≤ 14Gy
|
Dmean ≤ 9Gy
|
Lens 1-sided
|
Dmax ≤ 9Gy
|
Brain stem
|
Dmax ≤ 40Gy
|
Inner ear 1-sided
|
Dmax ≤ 40Gy
|
Retina, 1-sided
|
Dmax ≤ 35Gy
|
Optic nerves, chiasm
|
Dmax ≤ 35Gy
|
Optic nerve, 1-sided
|
Dmax ≤ 35Gy
|
Abbreviations: Dmax, Maximum dose of Organs-at-Risk; Dmean, Mean dose of the hippocampus. |
Treatment
All patients were diagnosed as multiple brain metastases by contrast-enhanced magnetic resonance imaging(MRI). Among them, 56 patients were treated with HA-WBRT + SIB and another 56 patients received HA-WBRT alone. 112 patients received radiotherapy-planning computed tomography (CT) with 1.5mm slice thickness in thermoplastic mask, c-pillow and head frame immobilization as well as contrast-enhanced transversal T1-weighted magnetic resonance imaging(MRI). The scanned images were transmitted to Varian's planning system ( Varian Eclipse 8.0 and 15.0 version), and then the CT images were fused with the transversal T1-weighted MRI images and served for the target volume and organ-at-risk delineation.
The whole brain excluding the hippocampus-avoidance region (HAR, a 7mm 3-dimensional margin around the hippocampus) was defined as the clinical target volume(CTV), and an extension of 3mm on CTV was defined as planning tumor volume(PCTV) in the observation group. PCTV was given 30Gy in 12 fractions,per fraction one day,5 fractions per week, maximum dose of the hippocampus (Dmax) ≤ 14
Gy, mean dose(Dmean) ≤ 9Gy. The gross tumor volume (GTVmetastases,i.e. GTVm) was contoured on the fused images(CT images and MRI images), GTVm with 2mm extension formed PGTVm,with total dose of 48Gy in 12 fractions,one fraction every day, 5 fractions one week in the observation group.The whole brain excluding HAR with 3mm margin was defined as planning tumor volume of the brain (PTVwhole brain), with the same dose of PCTV of HA-WBRT + SIB in the control group. The dose was prescribed to cover the 95% isodose.SRT, SRS and SBRT have not been implemented in many places due to unbalanced development in various regions of China. Therefore, the dose escalation to PGTV was used to treat patients with multiple brain metastases (4–15) by linear accelerator(VARIAN CLINAC IX) in this study.Dose constraints of organs at risk are showed in Table 1.
On the basis of the situation of patients and clinical experience of physicians, 36 patients(HA-WBRT + SIB) and 34 patients(HA-WBRT) underwent more than 2 cycle of systemic therapy, 20 patients(HA-WBRT + SIB) and 22 patients(HA-WBRT) did not undergo systemic therapy.
Follow‑up And Study Endpoint
Follow-up was scheduled for examinations in the first month after radiotherapy,inclu-
ding contrast enhanced MRI,clinical examination,and adverse reaction evaluation on the basis of version 4.0 of the Common Terminology Criteria for Adverse Events (CTCAE),and repeated at 3 months interval for the first 2 years, then repeated at 6 months interval in years 3 to 5, and once a year thereafter. The last follow-up was November 2021. Overall survival (OS) ranged from the first day of brain radiotherapy to death or the day on which the patient was last known to be alive in the case of loss to follow-up. Median survival means that only 50% of the individuals can live through this time.Intracerebral control (IC) included complete response (CR), partial response (PR), or stable disease (SD) of brain metastases after radiotherapy. Furthermore, pseudoprogression and radionecrosis related to radiotherapy were excluded from local failures of IC. If the total maximum diameter of treated lesions increased by 20%, or increase the absolute value ≥ 5 mm, or new lesions appeared in the brain, which were considered progressive disease (PD). Intracranial progression-free survival (iPFS) was counted from the time to start radiotherapy until brain metastases progression or death for any reason or the day on which the patient was last known to be alive in the case of loss to follow-up.
Evaluation of treatment-related toxicity included alopecia, radiation dermatitis, encephaledema, headache, emesis, sicchasia, vertigo, fatigue, focal neurologic deficits, epilepsia, neurocognitive dysfunction, and radionecrosis.
Statistical analysis
Statistical analysis was adopted SPSS 20.0 (IBM Corporation, Chicago, IL, USA) statistical software.To manage the unbalance of potential interference factors, propensity score matching (PSM) was adopted to set up two treatment groups with an even distribution of original characteristics. The propensity score matching analysis was used between HA-WBRT + SIB(observation) and HA-WBRT(control) groups to control confounding factors of patients, and performed with a logistic regression that considered the following factors:age, gender, intracranial symptoms, Karnofsky performance status, primary tumor, extracranial metastases, maximum size and number of brain metastases.OS,median survival and iPFS were calculated by Kaplan–Meier method(Fig. 2).1-year-IC rates and adverse events were calculated by χ2 test (fisher’s exact test).The baseline characteristics of the patients were counted by χ2 test after matching.P<0.05 (two-tailed) of all analysis results were considered statistically significant.