Intracystic interferon-alpha in pediatric craniopharyngioma patients-reply

On behalf of all the contributors to the published article “Intracystic interferon-alpha in pediatric craniopharyngioma patients: an international multicenter assessment on behalf of SIOPE and ISPN,” Neuro-Oncology 2017;19(10):1398–1407, we would like to thank Dr Cavalheiro for his response to “Intracystic interferon-alpha in pediatric craniopharyngioma patients,” highlighting some important points which we hope to address below.

We share Dr Cavalheiro’s view of risk when evaluating treatment among retrospective cohorts containing treatment-naïve and pretreated patients. However, this was precisely the rationale behind analyzing the cohort as published. The cohort was typically assessed in its entirety wherever the influence of previous therapy was deemed inconsequential—areas such as presenting features, symptom interval, and concordance with the recommended interferon treatment protocol, while clinical outcomes across the 56 children were based exclusively on physical evaluations prior to and following the completion of interferon therapy. Moreover, when assessing the comparative efficacy of intracystic interferon at delaying disease progression, treatment-naïve patients were removed from analysis. Progression rates following interferon were assessed across the entire cohort, rather than dividing patients according to specific treatments already received, as limited patient numbers did not allow statistically viable results. Nevertheless, the progression-free survival rate post interferon between treatment-naïve and pretreated patients was not significant (0.6 y [95% CI: 0.2–1 y] vs 1.2 y [95% CI: 0.7–1.1 y], P = 0.82; data not in manuscript).

The guidance for intracystic interferon administration that was adhered to by all contributing centers in the analysis was defined as the Toronto protocol as it had followed the Toronto Hospital for Sick Children’s standard operating procedure. This published institutional protocol¹ was selected, as it was deemed to demonstrate an increased level of detail for clinical users compared with previous descriptions of the technique, to thereby improve standardization of approach and reduce operator bias. This was specifically addressing technical issues of pretherapeutic imaging studies, volumes of cystic fluid aspirated prior to each therapy, and the use of a posttherapeutic saline flush solution. However, all contributing authors are the first to recognize that this more detailed guidance is derived from the published experiences of Dr Cavalheiro and the São Paulo group, whose findings were fundamental to its development—any suggestion otherwise was wholly unintended nor implied in the article, which indeed cites all 3 papers referenced by Dr Cavalheiro in his response.^2–4

We thank Dr Cavalheiro for his suggestions highlighting the importance of analyzing patient subgroups according to preceding treatment modalities and lesional/fluid biology. We agree that performing survival analyses according to these subdivisions would have been interesting.

However, performing such an analysis on a relatively small total cohort size was not feasible. We faced the challenge of historical retrospective data collection, spanning several years, from an array of global centers, each with a spectrum of scientific infrastructures allowing biological tumor analysis. Consequently, our remit was precise: through sharing the clinical experiences of these specialist treatment centers, to evaluate the efficacy of intracystic interferon in delaying or preventing both disease progression and the need for subsequent definitive therapy, along with evaluating toxicity and clinical outcomes following treatment. As can be seen in Supplementary Table S2, there was a large and varied range of preceding treatment modalities. To further subdivide these therapies according to lesional composition and subsequent biological strata and then perform a survival analysis across all data streams would result in an inability to draw sound statistical conclusions, and hence such analysis was dismissed.

Likewise, the criteria for repeating a cycle of interferon therapy or indeed changing to another therapeutic modality was at the discretion of individual treating centers, and as Dr Cavalheiro states, such decisions were non-uniform. Typically, therapy was extended past 12 doses if a complete radiological response had not been achieved, while Supplementary Table S2 demonstrated the types of progression reported (most commonly new cysts and solid lesional growth) and the subsequent management strategies employed. Given we were only reporting the experiences of these contributing centers, it was beyond the scope of the study to clarify each center’s reasoning for the decisions made. Nevertheless, the suggestion that moving away from interferon hastily may cause false-negative findings of drug efficacy is well taken. Of the 10 cases progressing only as cystic reaccumulation, only 6 underwent a rechallenge with interferon.

We thank Dr Cavalheiro for his considered suggestions and concur with several, as it again highlights the need for prospective international clinical trial work in this field, using randomized, sizable patient cohorts treated according to standardized protocols to evaluate many of the issues raised both in the published manuscript and in the response article.