59 patients with local GBM recurrence were treated with fSRT following the initial treatment. Patient characteristics are presented in Table 1. For the primary treatment, 20 (34%), 28 (47%) and 11 (19%) patients underwent stereotactic biopsy, subtotal resection and gross total resection respectively. All patients completed the initial chemoradiation procedure. The median KPS score at recurrence was 90% (60–100%). 31 (53%) patients received systemic treatment following reirradiation (Table 1). Forty-five (76%) patients were reirradiated on the first recurrence of GBM, 10 (17%) on the second recurrence (6 patients had systemic therapy and 4 patients had surgery on the first recurrence), 3 (5%) on the third recurrence (systemic therapy) and 1 (2%) on the fourth recurrence (systemic therapy). The reirradiation characteristics are presented in Table 2. The median time from the end of primary radiotherapy to the first recurrence was 9.5 months (range: 1.3–65 months). The median time from the recurrence diagnosis and fSRT initiation was 1 month (range: 0.1–3.5 months). All patients completed the reirradiation course. The median follow-up appointment after fSRT initiation was at 8.8 months (range: 3–54 months). Two patients were still alive at the time of analysis.
Table 1
Patient characteristics at diagnosis and at the time of recurrence for the whole cohort
At diagnosis
|
Age
|
61 ± 10
|
Male
|
32(54)
|
Surgical procedure
Stereotactic Biopsy
Subtotal resection
Complete resection
|
20 (34)
28 (47)
11 (19)
|
Adjuvant systemic therapy
Temozolomide
Bevacizumab
Nivolumab
Temozolomide with bevacizumab
Temozolomide with nivolumab
|
52 (88)
1 (2)
1 (2)
4 (6)
1 (2)
|
At recurrence
|
Number of previous recurrences
0
1
2
3
|
45(76)
10(17)
3(5)
1(2)
|
KPS
60
70
80
90
100
|
1 (2)
10 (17)
9 (15)
35 (59)
4 (7)
|
ADL
6
5
4
3
2
|
53 (90)
2 (3)
1 (2)
2 (3)
1 (2)
|
Multifocality
Yes
No
|
5 (8)
54 (92)
|
Steroids
Yes
No
|
14 (24)
45 (76)
|
Antiepileptic medication
Yes
No
|
35 (59)
24 (41)
|
Headaches
Yes
No
|
14 (25)
45 (75)
|
Motor deficiency
Yes
No
|
9 (85)
50 (15)
|
Intracranial hypertension
Yes
No
|
0 (0)
59 (100)
|
Adjuvant systemic treatment after reirradiation
Temozolomide
Bevacizumab
Fotemustine
Temozolomide with bevacizumab
Fotemustine with bevacizumab
Lomustine with bevacizumab
Nivolumab
|
31(52)
12 (20)
5 (8)
4 (7)
3 (5)
4 (7)
2 (3)
1 (2)
|
Footnote: Result: mean ± standard deviation or frequency (percentage); RTCT: Concomitant adjuvant chemotherapy plus radiotherapy; KPS: Karnofsky Performance Status; ADL: Activity of Daily Living |
Table 2
Treatment characteristics of reirradiation.
|
Median (minimum – maximum)
|
Total dose (Gy)
|
25 (25–25)
|
Maximum dose at plan (Gy)
|
31.25 (31.13–35.71)
|
Number of beams
|
101 (56–164)
|
Conformity Index
|
1.28 (1.14–2.1)
|
GTV-PTV margin (mm)
|
2 (1–3)
|
PTV (cc)
|
11.4 (1.1–60.7)
|
Footnote: PTV: Planning target volume; GTV: Gross tumor volume; Cc: cubic centimeter |
The incidence of KPS impairment in all patients was 51.9% [CI-95%, 37.3–68.3] 5 months after fSRT (Fig. 1). Using univariate analysis, KPS impairment was significantly associated with headaches (HR = 2.44 [CI-95%, 1.07–5.59], p = 0.035) and a larger PTV (continuous data for an increase of 1 cc) (HR = 1.05 [CI-95%, 1.02–1.08], p = 0.016) (Table 3). A PTV value ≥ 10cc was observed for 33 patients (55.9%) and significantly increased the risk of KPS impairment (HR 2.82 [CI-95%, 1.12–7.14], p = 0.028). Using multivariate analysis, only a larger PTV was associated with KPS impairment. Six months after reirradiation, 37.8% [CI-95%, 23.5–56.7] of patients had ADL impairment (Fig. 1), the median institutionalization-free survival time was 7.7 months [CI-95%, 6.1–8.8] (Fig. 2). For the 45 patients who received fSRT at the first recurrence, the incidence of KPS impairment was 52.6% [CI-95%, 35.9–71.4] 5 months after fSRT and an increase of PTV (continuous data for an increase of 1 cc) was also significantly associated with KPS impairment (HR = 1.05 [CI-95%, 1.01–1.09], p = 0.024). Six months after reirradiation, 36.7% [CI-95%, 21.7–57.5] of patients had ADL impairment. Twenty-four patients experienced toxicity within three months following reirradiation: 17 patients presented with grade 1 (12 headache, 3 seizure, 1 motor defect and 1 intracranial hypertension), 5 patients experienced grade 2 (2 headache, 2 intracranial hypertension, 1 motor deficiency) and 2 patients suffered from grade 3 toxicity (2 intracranial hypertension).
Table 3
Prognostic factors of KPS impairment after reirradiation by bivariate analysis
Variable
|
Hazard ratio [95% CI]
|
p-value
|
Male sex
|
0.88 [0.40 ; 1.98]
|
0.763
|
Initial brain surgery
|
1.20 [0.48 ; 3.03]
|
0.697
|
Motor deficiency
|
1.32 [0.45 ; 3.88]
|
0.615
|
Antiepileptic medication
|
0.71 [0.32 ; 1.59]
|
0.405
|
Headaches
|
2.44 [1.07 ; 5.59]
|
0.035
|
Age (years) > 60
|
1.0 [0.45 ; 2.24]
|
1
|
Steroids (mg)
|
1.47 [0.64 ; 3.38]
|
0.363
|
PTV (cc) *
|
1.57 [1.19 ; 2.08]
|
0.028
|
KPS (%) ≥ 90
|
1.51 [0.60 ; 3.81]
|
0.388
|
Time from RTCT to fSBReRT (months) < 12
|
1.36 [0.59 ; 3.10]
|
0.473
|
Number of recurrences before reirradiation ≥ 1
|
1.27 [0.50 ; 3.20]
|
0.617
|
Multifocality
|
1.66 [0.49 ; 5.62]
|
0.412
|
Brain topography
Cortical
Subcortical
|
1
2.54 [1.00 ;6.44]
|
0.050
|
Footnote: CI: confidence interval; * HR computed for an increase of 10 cc; RTCT: Concomitant adjuvant chemotherapy plus radiotherapy; fSBReRT: Fractionated stereotactic body reirradiation |
The median OS from diagnosis and from fSRT was 25.8 months [CI-95%, 22-29.7] and 8.8 months [CI-95%, 7.4–10.9], respectively (Fig. 2). The median PFS time was 3.9 months [CI-95%, 3.3–5.2] (Fig. 2). Using univariate analyses, surgery (HR = 0.48 [CI-95%, 0.27–0.86], p = 0.013) and steroids administration (HR = 1.81 [CI-95%, 1.04–3.16], p = 0.037) were significant associated with PFS (Table 4). Only initial stereotactic biopsy was significantly associated with PFS deterioration using multivariate analysis. For patients with treatment at the first recurrence, the median OS from diagnosis and from fSRT was 23 months [CI-95%, 20.9–29.7] and 10.8 months [CI-95%, 8.2–12], respectively. Median PFS time was 4 months [CI-95%, 3.4–5.7], and the initial stereotactic biopsy was significantly associated with decreased PFS (HR = 0.41 [CI-95%, 0.21–0.81], p = 0.011).
Table 4
Prognosis factors for decreased PFS by bivariate analysis
Variable
|
Hazard Ratio [95% CI]
|
p-value
|
Male sex
|
0.85 [0.50 ; 1.47]
|
0.566
|
Initial brain surgery
|
0.48 [0.27 ; 0.86]
|
0.013
|
Motor deficiency
|
1.42 [0.68 ; 2.94]
|
0.350
|
Antiepileptic medication
|
1.17 [0.69 ; 1.98]
|
0.567
|
Headaches
|
0.86 [0.46 ; 1.61]
|
0.645
|
Age (years) > 60
|
0.67 [0.40 ; 1.13]
|
0.135
|
Steroids (mg)
|
1.81 [1.04 ; 3.16]
|
0.037
|
PTV (cc) *
|
1.14 [0.90 ; 1.44]
|
0.269
|
KPS (%) ≥ 90
|
1.01 [0.58 ; 1.73]
|
0.985
|
Time from RTCT to fSBReRT (months) < 12
|
0.60 [0.35 ; 1.01]
|
0.055
|
Number of recurrences before reirradiation ≥ 1
|
1.84 [0.98 ; 3.44]
|
0.056
|
Multifocality
|
1.18 [0.47 ; 2.99]
|
0.715
|
Brain topography
Cortical
Subcortical
|
1
1.58[0.77 ;3.26]
|
0.214
|
Footnote: CI: confidence interval; * HR computed for an increase of 10 cc; RTCT: Concomitant adjuvant chemotherapy plus radiotherapy; fSBReRT: Fractionated stereotactic body reirradiation |