paediatrics Brussels 17

Articles

N Event-free survival (%)

Overall survival (%)

5 years (95% CI)

7 years (95% CI)

HR (95% CI)

p

5 years (95% CI)

7 years (95% CI)

HR (95% CI)

p

Tumour grade Differentiated

68 86·4 (76·8–96·0)

79·2 (66·1–92·3) 61·3 (38·8–83·8)

1·0

0·005 91·9 (84·3–99·5)

89·4 (79·6–99·2) 71·8 (52·6–91·0)

1·0

0·006

Anaplastic

85

61·3 (46·4–76·2)

2·58 (1·30–5·12)

··

78·3 (66·3–90·3)

3·56 (1·37–9·22)

··

Tumour location Infratentorial

122

71·1 (60·5–81·7) 82·9 (66·6–99·2)

65·8 (52·7–78·9) 82·9 (57·6–100·0)

1·0

0·16 84·0 (75·6–92·4)

80·5 (69·5–91·5)

1·0

0·6

Supratentorial Ethnic origin White

31

0·52 (0·20–1·32)

··

89·5 (76·8–100·0)

83·1 (59·4–100·0) 0·75 (0·25–2·22)

··

126 75·5 (66·3–84·7)

70·4 (57·7–83·1) 64·5 (30·8–98·2)

1·0

0·26 87·7 (80·6–94·8)

84·5 (74·7–94·3) 60·7 (27·4–94·0)

1·0

0·017

Other

27

64·5 (30·8–98·2)

1·55 (0·71–3·38)

··

72·9 (44·9–100·0)

2·84 (1·16–6·92)

··

Sex Female

58 84·7 (73·9–95·5)

81·0 (66·3–95·7) 61·0 (43·4–78·6)

1·0

0·018 91·8 (83·8–99·8)

88·6 (76·8–100·0) 1·0

0·091

Male

95

66·7 (53·4–80·0)

2·40 (1·13–5·06)

··

81·1 (70·1–92·1)

76·0 (61·1–90·9)

2·20 (0·86–5·61)

··

Age at CRT (years) ≥3

75

79·0 (66·8–91·2)

69·4 (52·2–86·6) 68·6 (52·1–85·1)

1·0

0·37

90·1 (81·1–99·1) 80·4 (69·8–91·0)

81·7 (68·0–95·4) 80·4 (66·1–94·7)

1·0

0·46

<3

78 68·6 (55·7–81·5)

1·34 (0·71–2·52)

··

1·37 (0·60–3·12)

··

Total dose (Gy) 54

22

80·7 (61·5–99·9) 72·4 (62·4–82·4)

70·6 (44·1–97·1) 68·8 (55·5–82·1)

1·0

0·67

85·4 (68·9–100·0) 85·0 (77·0–93·0)

77·7 (53·8–100·0) 1·0

0·82

59·4

131

1·04 (0·87–1·24)

··

81·6 (70·8–92·4)

0·98 (0·80–1·19)

··

Number of surgical procedures 1 87

79·7 (69·3–90·1)

74·4 (60·3–88·5) 62·0 (41·2–82·8)

0·55 (0·29–1·02)

0·056 90·1 (82·7–97·5)

83·9 (72·3–95·5) 78·4 (60·6–96·2)

0·56 (0·24–1·26)

0·15

2–4

66 65·6 (49·5–81·7)

1·0

··

78·4 (64·5–92·3)

1·0

··

Surgical extent GTR

125

81·5 (72·7–90·3)

77·3 (65·0–89·6) 34·2 (12·1–56·3)

0·21 (0·11–0·40)

<0·0001 93·0 (87·3–98·7)

88·0 (78·8–97·2) 52·4 (25·5–79·3)

0·16 (0·07–0·36)

<0·0001

NTR or STR

28 41·0 (17·7–64·3)

1·0

··

52·4 (25·5–79·3)

1·0

··

Pre-CRT chemotherapy Yes 35

59·4 (39·6–79·2)

48·7 (26·0–71·4) 75·9 (62·8–89·0)

1·0

0·008 73·6 (55·6–91·6)

66·9 (43·0–90·8) 85·3 (75·1–95·5)

1·0

0·038

No

118 78·1 (68·3–87·9)

0·43 (0·22–0·81)

··

88·6 (81·3–95·9)

0·42 (0·18–0·98)

··

HR=hazard ratio. CRT=conformal radiotherapy. GTR=gross-total resection. NTR=near-total resection. STR=subtotal resection.

Table 2: Univariate analysis of event-free survival and overall survival according to different variables

81·0% (71·0–91·0), respectively (figure 2). Median time to progression was 22·5 months (range 5·0–90·9) from diagnosis and 20·3 months (3·1–75·4) from the start of conformal radiotherapy. Univariate analyses of overall survival by various clinical variables are presented in table 2. Multiple regression analysis showed overall survival was affected by tumour grade, extent of resection, and ethnic origin: gross-total resection was associated with a lower risk of death from any cause than was near-total or subtotal resection (HR 0·16 [95% CI 0·07–0·37]; p<0·0001), while the risk of death was greater in patients with anaplastic tumours than in those with differentiated tumours (HR 3·98 [1·51–10·48]; p=0·0052) and in non-white patients versus white patients (HR 3·0 [1·21–7·44]; p=0·018). However, death from necrosis accounted for the lower overall survival in non-white patients, compared with white patients: when we excluded the two patients who died of necrosis, the comparison of ethnic origin was not significant for overall survival (HR 2·1 [0·8–5·7]; p=0·16 by univariate analysis). The use of chemotherapy before conformal radiotherapy was associated with a lower overall

of brainstem ischaemia on postoperative T2-weighted MRI, required tracheostomy, and had postoperative hypertension needing medication. Two of the three also had a history of a postoperative seizures. There were no other cases of necrosis and no other patients had a similar constellation of clinical signs and symptoms before or during radiotherapy. Including all three cases of necrosis, the cumulative incidence of brainstem necrosis at 7 years was 2·5% (95% CI 0·0–5·2); excluding the patient who died after a seizure, it was 1·6% (0·0–4·0). Seizure disorders required chronic medication in five patients with supratentorial tumour location. Two needed surgery for epilepsy and were able to reduce or stop medication. There was one case of radiation-related cerebral vasculopathy in a patient with infratentorial tumour location that required revascularisation surgery. The patient was aged 12 months at the time of irradiation and the high-dose volume encompassed the Circle of Willis. 7-year estimates of local control, EFS, and overall survival were 87·3% (95% CI 77·5–97·1), 69·1% (56·9–81·3), and

262

www.thelancet.com/oncology Vol 10 March 2009