paediatrics Brussels 17

Articles

50 days (range 52–81). One patient had chemotherapy after delayed second-look surgery. 51 (57%) completed all seven cycles of chemotherapy. A further ten patients stopped protocol treatment as they had been on treatment for over 1 year without completing all cycles, ninewere in remission, and one had stable residual disease. Chemotherapy was stopped early in 27 patients: 11 had disease progression, ten had unacceptable toxicity, one had residual disease and was irradiated, one patient had no tumour on imaging, whereas for four patients no specific reason was given. There was one postoperative death. Haematological toxicity was the most common treatment-related problem, with either grade 3 or 4 occurring in all patients (table 3). Only two, one metastatic and one non-metastatic, patients had grade 4 audiological toxicity. 50 of the 80 patients with non-metastatic disease progressed, of whom 34 were irradiated for progression (figure 1); eight were aged less than 3 years), and 16 patients with progressive disease were not irradiated for reasons determined by the physician or parent. In addition, two were irradiated (one younger than 2·5 years) despite no progression being reported. Thus, 44 (55%) patients with non-metastatic disease were not irradiated. All nine patients who had metastatic disease at diagnosis progressed. Six of these nine patients were irradiated. As would be anticipated, the radiotherapy rate was lower in those with a complete tumour resection: 18 of 44 (41%) compared to 24 of 44 (55%) of those not fully resected. Given that not all patients who progressed were irradiated, to reflect more accurately the need for radiotherapy, we used cumulative incidence methodology (CIM) 19 to calculate radiotherapy-free survival. For all 89 patients, the 3-year and 5-year cumulative radiotherapy rates were 44·6% (95% CI 34·5–56·2) and 49·3% (38·8–61·0), respectively. The median time from surgery to radiotherapy was 20·3 months (range 7·8–123·6) and median age at irradiation was 3·6 years (range 1·5–11·9). In non-metastatic patients, combining all who were irradiated with those that were eligible for this treatment modality (but did not receive it) perhaps reflects more accurately the success of this chemotherapy protocol in avoiding radiotherapy. The 5-year cumulative incidence rate of freedom from radiotherapy for non-metastatic patients was 42% (95% CI 32–53; 1-sum of all the curves in figure 2). In total, 59 patients including all nine with metastatic disease progressed, 37 of whom subsequently died. 47 relapsed locally, four at a metastatic site, six at both local and metastatic sites, one was unknown, and one child died during surgery. In all 40 patients who received radiotherapy for their progression, 23 also had surgery. Of the remainder who did not receive radiotherapy, seven patients underwent surgery alone, two received alternative chemotherapy, one was treated under a palliative care regimen, eight had no further treatment, and one died perioperatively (table 4).

After a central review, 59 (66%) of cases were designated classic (WHO grade II) ependymoma. The local histopathological grade was changed in 26 (29%) tumours—14 cases changed from grade II to III, and from grade III to II in 12. Cerebrospinal fluid (CSF) cytology was examined in 37 patients (42%), with no malignant cells detected. 44 (49%) had a complete resection of the primary tumour, 41 (46%) had subtotal resection, and biopsy only was done in three (3%), and one (1%) died perioperatively. The concordance between the surgical report and central radiology reviewwith respect to completeness of resection was 68%. The median start of chemotherapy was 23 days (range 0–81) after surgery. Five patients had a delay of more than

1·00

Metastatic Non-metastatic

0·75

0·50

0·25 Event-free survival (%)

0·00

12

0

2

4

6

8

10

Event-free survival (years)

Numbers at risk

·· ··

3

Metastases

9

0

0

0

0 9

51

No metastases 80

34

21

15

Figure 3: Event-free survival by presence or absence of metastases at diagnosis

1·00

Metastatic Non-metastatic

0·75

0·50

Overall survival (%)

0·25

0·00

0

2

4

6

8

10

12

Time from surgery (years)

Numbers at risk

·· ··

Metastases

9

7

3

0

0

0 9

72

47

28

17

No metastases 80

Figure 4: Overall survival by presence or absence of metastases at diagnosis

700

http://oncology.thelancet.com Vol 8 August 2007