Abstract

Plain language summary

Summary of findings

Background

Craniopharyngiomas are usually slow‐growing, benign tumours that originate from epithelial nests or from areas of squamous metaplasia located in the hypothalamic and pituitary regions. Craniopharyngiomas are the most common intracranial tumours of non‐glial origin in the paediatric population, constituting 6% to 13% of all childhood brain tumours and peaking at five to 10 years of age (Sanford 1991; Bunin 1998). Although the tumour is of a benign histological nature, there is considerable morbidity and disability even when the tumour can be resected completely.

Objectives

To assess the benefits and harmful effects of intracystic bleomycin in children from birth to 18 years with cystic craniopharyngioma.

Specific objectives are to compare:

1. intracystic bleomycin with placebo (no treatment);

2. intracystic bleomycin with surgical treatment (with or without adjuvant radiotherapy);

3. intracystic bleomycin with other intracystic treatments.

Methods

Data collection and analysis

Selection of studies

Two review authors independently screened the titles and abstracts of studies identified through the searches and selected trials that met the inclusion criteria. We retrieved full articles for further assessment. We used discussion and consultation with a third review author to resolve any disagreement. The selection process is outlined in a flow chart (see Figure 1)

Open in a separate window

1

Study flow diagram.

Data extraction and management

For included studies, two review authors independently extracted the following information using a standard form:

1. General information: title, authors, published/unpublished, year of publication, language of publication, duplicate publications, study design, country, reference/source, contact address, urban/rural, sponsoring, setting.

2. Intervention: dose, route, timing, control intervention (placebo, no treatment, surgical treatment (with or without adjuvant radiotherapy), some other intracystic chemotherapies).

3. Participants: sampling, total number and number in comparison groups, sex, age, trial inclusion and exclusion criteria; withdrawals/losses to follow‐up (reasons/description), subgroups.

4. Outcomes: outcomes specified above, length of follow‐up, quality of reporting of outcomes.

If there were differences in data extraction, we resolved these by discussion, referring back to the original paper, or by consulting a third person.

Assessment of risk of bias in included studies

Two authors planned to assess independently the risk of bias in included RCTs and CCTs using Cochrane's 'Risk of bias' tool (Higgins 2011). The tool considers six domains of bias: sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting and other issues. For each domain, the study method is described using verbatim quotes and judged for adequacy (high, low, unclear risk of bias). We added items for the assessment of risk of bias as described in the module of Cochrane Childhood Cancer (Kremer 2010). We resolved any disagreements by discussion or using a third party arbitrator, and presented the results in the 'Risk of bias' table and also in both graph and written summary form.

Risk of bias assessment for non‐randomised clinical trials is a complex topic. However, since no non‐randomised trials (i.e. CCTs) were identified, using a modification of the Methodological Index for Non‐Randomised Studies (MINORS) was not applicable (Slim 2003).

Measures of treatment effect

For time‐to‐event data (e.g. overall survival and event‐free survival), we used the hazard ratio (HR). If HRs were not explicitly presented in the study, we planned to use Parmar's method (Parmar 1998).

For dichotomous outcomes (e.g. response of cyst to treatment, adverse events, visual outcome, neurological status and endocrine function), we calculated risk ratios (RR) with 95% confidence intervals (CI) for each trial.

For continuous outcomes (e.g. the rate of decrease of the size of the cystic component, quality of life), we intended to evaluate mean difference (MD) or standardised mean difference (SMD) with 95% CI. We only analysed one continuous outcome, reduction of cyst size, and for this we reported the MD.

Unit of analysis issues

We planned to consider each individual study included in the meta‐analysis as a unit for analysis.

Dealing with missing data

We analysed all data on an intention‐to‐treat basis. We did not contact authors for missing data.

Assessment of heterogeneity

Since only one eligible study was identified, assessing the presence of substantial heterogeneity (i.e. I² > 50% (Higgins 2011)) was not applicable.

Assessment of reporting biases

We planned to construct a funnel plot (Egger 1997), to ascertain graphically the existence of publication bias. However, as a rule of thumb, tests for funnel plot asymmetry should be used only when there are at least 10 studies included in the meta‐analysis, because when there are fewer studies the power of the tests is too low to distinguish chance from real asymmetry (Higgins 2011). Since only one RCT could be included in this review, we did not construct funnel plots.

Data synthesis

We performed data synthesis and analyses using the Cochrane Review Manager software, RevMan 5.3 (RevMan 2014). We performed the analyses according to the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We used the Mantel‐Haenszel method to perform the analysis. Since only one RCT could be included in this review, we used a fixed‐effect model. Also, because there was only one study available, we were unable to calculate a RR if one of the treatment groups experienced no events and we therefore used the Fisher's exact test instead (in SPSS 18.0). For each comparison we prepared a 'Summary of findings' table using the GRADE profiler software (www.gradepro.org), in which we presented the following outcomes: survival, size reduction of cyst, neurological status, total number of adverse effects, fever, third nerve paralysis and headache in combination with vomiting. For each outcome two review authors independently assessed the quality of the evidence by using the five GRADE considerations, i.e. study limitations, inconsistency, indirectness, imprecision and publication bias as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) and the GradePro Handbook (Schünemann 2013).

Subgroup analysis and investigation of heterogeneity

As only one eligible study was identified and the sample size was small, subgroup analysis was not applicable. Otherwise we would have looked at:

1. age at onset of intracystic bleomycin treatment;

2. gender;

3. dosage (or the concentration of the dose) of treatment;

4. multicystic craniopharyngioma.

Sensitivity analysis

Since only one eligible study was identified, performing sensitivity analyses for the quality criteria used was not applicable.

Results

Risk of bias in included studies

The risk of bias is summarised in the 'Risk of bias' summary (Figure 2) and 'Risk of bias' graph (Figure 3). For more detailed information see the 'Risk of bias' table in the Characteristics of included studies.

Open in a separate window

2

'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study.

Open in a separate window

3

'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

The generation of allocation sequence and allocation concealment was unclear.

Blinding

Blinding of participants was performed. However, blinding of personnel and outcome assessors for the clinical outcomes was not performed.

Incomplete outcome data

The numbers and reasons for dropouts and withdrawals in all intervention groups were not described. We considered the trial to have an unclear risk of bias.

Selective reporting

The protocol for this study was unavailable, therefore the trial was at unclear risk of bias.

Other potential sources of bias

Baseline imbalance

The patients in the bleomycin group were all multicystic, while the patients in the ³²P group were not. Hence, the trial did have a risk of baseline imbalance.

Early stopping

The sample size calculations for the trial were not reported. Hence, it was unclear whether the trial was stopped early.

Summary of quality assessment of included trial

We considered the trial to have a high risk of performance bias, detection bias and baseline imbalance bias, and an unclear risk of selection bias, attrition bias, reporting bias and early stopping bias.

Effects of interventions

See: Table 1

Summary of findings for the main comparison

Intracystic bleomycin compared to other types of intracystic treatment for cystic craniopharyngiomas in children

Intracystic bleomycin compared to intracystic radiotherapy with 32P for cystic craniopharyngiomas in children
Patient or population: cystic craniopharyngiomas in children from birth to 18 years old Setting: inpatients from 3 hospitals in China Intervention: intracystic bleomycin Comparison: intracystic radiotherapy with 32P
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Quality of the evidence (GRADE)	Comments
	Risk with intracystic 32P	Risk with intracystic bleomycin
Survival	NA	NA	‐	‐	‐	There was no information on overall survival and event‐free survival reported in this trial
Change in cyst size	Mean reduction in cyst size was 65.5%	Mean reduction in cyst size was 50.7%	MD ‐15% (‐69% to 39%)	7 (1 RCT)	VERY LOW 1	MD was 15% reduction in cyst size; the 95% CI mean a 69% reduction to a 39% increase in cyst size
Neurological status	See comment	See comment	Not estimable	Not estimable	VERY LOW 1	1 patient in the bleomycin group had left paralysis, while none of the patients in the 32P group had severe sequelae. There was no significant difference between the 2 groups (Fisher's exact P value = 0.429).
Total number of adverse effects	500 per 1000	875 per 1000 (340 to 1000)	RR 1.75 (0.68 to 4.53)	7 (1 RCT)	VERY LOW 1	Adverse effects including fever, 3rd nerve paralysis, headache and vomiting were included in this outcome measure
Fever	250 per 1000	730 per 1000 (183 to 1000)	RR 2.92 (0.73 to 11.70)	7 (1 RCT)	VERY LOW 1	‐
3rd nerve paralysis	See comment	See comment	Not estimable	7 (1 RCT)	VERY LOW 1	1 patient in the 32P group had 3rd nerve paralysis, while none of the patients in the bleomycin group suffered from this. There was no significant difference between the 2 groups (Fisher's exact P value = 1.00).
Headache and vomiting	See comment	See comment	Not estimable	7 (1 RCT)	VERY LOW 1	All 3 patients in the bleomycin group had headache and vomiting, while none of the patients in the 32P group did. There was a significant difference in favour of the 32P group (Fisher's exact P value = 0.029).
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; NA: not available; RR: risk ratio; OR: odds ratio
GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

Open in a separate window

¹We downgraded a total of 3 levels due to study limitations (the trial had a high risk of performance bias, detection bias and baseline imbalance bias, and an unclear risk of selection bias, attrition bias, reporting bias and early stopping bias), small sample size (the current review includes only one trial with small sample size (n = 7)), and wide confidence intervals including 'no effect', 'appreciable harm' and 'appreciable benefit'.

Overall survival or event‐free survival

There was no information on overall survival and event‐free survival reported in this trial.

The response of the cyst to treatment

There was no information about the response of the cyst to treatment.

The size of the cystic component

There was no significant difference in cyst size reduction between the bleomycin and phosphorus³² (³²P) groups (mean difference (MD) ‐0.15, 95% confidence interval (CI) ‐0.69 to 0.39, P value = 0.59, very low quality of evidence) (see Analysis 1.1 and Table 1).

Open in a separate window

1.1

Analysis

Comparison 1 Intracystic bleomycin versus intracystic ³²P, Outcome 1 Reduction of cyst size.

Endocrine function

The data for endocrine function could not be extracted.

Neurological status

One child in the bleomycin group had left paralysis after two episodes of epilepsy. None of the children in the ³²P group had severe sequelae. There was no significant difference between the groups (Fisher's exact P value = 0.429, very low quality of evidence).

Visual outcome

There was no visual outcome reported in this trial.

Quality of life

There was no quality of life outcome reported in this trial.

Adverse effects

All three children in the bleomycin group had complications of fever, headache and vomiting, however, in the ³²P group there was only one patient who had transient fever; none of the patients in that group suffered from headache or vomiting. There was no significant difference in fever between the two groups (risk ratio (RR) 2.92, 95% CI 0.73 to 11.70, P value = 0.13, very low quality of evidence) (see Analysis 1.2 and Table 1), but there was a significant difference in headache (Fisher's exact P value = 0.029, very low quality of evidence) and vomiting (Fisher's exact P value = 0.029, very low quality of evidence) in favour of the ³²P group. A patient in the ³²P group had third nerve paralysis, while none of the patients in the bleomycin group suffered from this. There was no significant difference between the two groups (Fisher's exact P value = 1.00, very low quality of evidence). In summary, there was no significant difference in total adverse effects between the groups (RR 1.75, 95% CI 0.68 to 4.53, P value = 0.25, very low quality of evidence) (see Analysis 1.3 and Table 1).

Open in a separate window

1.2

Analysis

Comparison 1 Intracystic bleomycin versus intracystic ³²P, Outcome 2 Fever.

Open in a separate window

1.3

Analysis

Comparison 1 Intracystic bleomycin versus intracystic ³²P, Outcome 3 Total adverse effects.

Subgroup analysis

It was not possible to perform subgroup analysis due to the small sample size.

Discussion

The occurrence of cysts is one of the typical characteristics of craniopharyngioma, occurring in more than 90% of tumours (Blacklund 1994). The treatment of cystic craniopharyngioma is controversial. Puget 2007 used a preoperative classification system to grade hypothalamic involvement and stratify treatment. For tumours that do not involve the hypothalamus and that are amenable to complete resection without hypothalamic injury, radical resection should be attempted. For tumours that appear to invade the hypothalamus on magnetic resonance imaging (MRI), especially if children with these tumours already have clinical evidence of hypothalamic dysfunction, the risks of radical resection are higher and partial resection followed by radiotherapy might be feasible. However, for any type of radiotherapy, the potential complications are higher the younger the child is, therefore intracystic treatment that delays the use of radiotherapy or is followed by resection may be beneficial. Another systematic review has concluded that intracystic interferon currently seems to have the best benefit‐risk ratio, but the evidence on which this was based was from non‐randomised controlled trials (Bartels 2012). This is an update of the previous systematic reviews evaluating the use of intracystic bleomycin for cystic craniopharyngiomas in children (Fang 2012; Zheng 2014).

There were no eligible randomised controlled trials (RCTs), quasi‐randomised trials or controlled clinical trials (CCTs) in which only the use of intracystic bleomycin in children differed between groups. Even though RCTs are the highest level of evidence, it should be recognised that data from non‐randomised studies on the use of intracystic bleomycin in cystic craniopharyngioma are available. The results are promising (Frank 1995; Hader 2000; Mettolese 2001; Park 2002; Lena 2005; Kim 2007; Hukin 2007). Most of these studies were retrospective cohort studies and they mentioned that intracystic bleomycin has potential advantages, such as inducing the shrinkage of the cyst size and allowing the delay of radiotherapy or radical resection.
There was only one eligible randomised trial that compared two different types of intracystic treatments: intracystic bleomycin and intracystic phosphorus³² (³²P). No survival or duration of follow‐up data were provided to compare the two groups. There was no significant difference in the size reduction of the cyst, neurological status, fever or third nerve paralysis. There was a significant difference in the occurrence of headache and vomiting in favour of the ³²P group. There was no significant difference in total adverse effects. However, the quality of the evidence is very low; the groups were small and relevant patient characteristics (i.e. characteristics of the cysts) were different between the two groups. Furthermore, there is a high risk of bias. We considered the trial to have a high risk of performance bias, detection bias and baseline imbalance bias and an unclear risk of selection bias, attrition bias, reporting bias and early stopping bias. In addition, the trial did not calculate the sample size, involved only seven patients and was therefore grossly underpowered to detect clinically important differences in outcome.

However, the role of intracystic bleomycin as a primary curative technique compared with radical therapy, partial resection plus radiotherapy, or other intracystic treatment in children with cystic craniopharyngioma can only be adequately determined through evaluation within prospective RCTs. Intracystic bleomycin is associated with a significant risk of morbidity. The acute adverse effects of intracystic bleomycin include headache, nausea, vomiting and transient mild fever, which occur in as many as 70% of patients, typically 24 hours after each instillation, and are self‐limiting (Hukin 2007). Even more serious are the reported delayed complications. There have been cases where MRI revealed signs of bleomycin leakage consistent with the extensive vasogenic oedema surrounding the cyst, which resulted in hypothalamic injury (Lafay‐Cousin 2007). The following complications have also been reported: diabetes insipidus (Park 2002), progressive panhypopituitarism (Hukin 2007), precocious puberty (Hukin 2007), and hypothalamic dysfunction resulting in transient hypersomnolence and poor memory (Park 2002), visual deficits (Mettolese 2001; Park 2002), neurocognitive deficits (Hukin 2007), sensorineural hearing loss (Broggi 1989; Frank 1995), peritumoral oedema (Hukin 2007), cerebral ischaemia (Broggi 1989), cerebral artery stenosis/occlusion (Cho 2012), hemiparesis (Jiang 2002; Park 2002), moyamoya disease after the combined treatment of intracystic bleomycin and radiotherapy (Hukin 2007), and death possibly related to a high individual and cumulative dose (Savas 2000).

Hence, surgeons have to realise that intracystic bleomycin for cystic craniopharyngioma in children currently has to be regarded as an experimental treatment that should only be performed in the context of a trial and requires close clinical monitoring. Imaging evaluation should be performed using MRI during treatment to ensure the safety of the therapy.

Comparing two intracystic treatments (with or without post‐treatment therapy) is an appropriate consideration for a future randomised trial. For example, the use of intracystic interferon alpha (IFNα) (Cavalheiro 2005; Ierardi 2007) and beta‐emitting radionuclides (Cáceres 2005), such as phosphorus³² (³²P), yttrium⁹⁰, rhenium¹⁸⁶ and aurum¹⁹⁸, are also available to provide control of the cystic tumour. Compared to the morbidity associated with intracystic bleomycin, IFNα seems to have similar advantages, but does not appear to have any significant major toxicity, even if it spills into the subarachnoid space. Therefore, further trials on intracystic treatment of cystic craniopharyngiomas are very much needed.

Authors' conclusions

What's new

History

Protocol first published: Issue 12, 2010
Review first published: Issue 4, 2012

Acknowledgements

Appendices

Notes

New search for studies and content updated (no change to conclusions)

Data and analyses

Characteristics of studies

Differences between protocol and review

We added "The dose, frequency and duration could vary" under the heading Types of interventions.

We added "or via neuroendoscopic technique" under the heading Types of interventions; this technique was not used for this procedure when we wrote the protocol and the original version of the review, but it is now.

We did not contact authors for missing data under the heading Dealing with missing data.

We deleted "We scanned the proceedings abstracts of the International Symposium of Pediatric Neuro‐Oncology (ISPNO) conferences and paediatric neurosurgical conferences (from 2005 to 2010)" under the heading Search methods for identification of studies and Results of the search.

We added "visual outcome" under the heading Secondary outcomes.

We added "cerebrovascular event" under the heading Secondary outcomes; since writing the protocol this adverse effect has been shown in non‐randomised studies.

As opposed to the earlier version of the review we now have prepared a 'Summary of findings' table and we have performed a GRADE assessment of the quality of the evidence.

Contributions of authors

S Zhang: searching for trials (this updated review), GRADE assessment and 'Summary of findings' table (this updated review).

Y Fang: searching for trials (previous publication and this updated version), GRADE assessment and 'Summary of findings' table (this updated review), 'Risk of bias' assessment (previous publication), data analyses (previous publication).

BW Cai: searching for trials (previous publication), 'Risk of bias' assessment (previous publication).

JG Xu: searching for trials (previous publication).

C You: protocol and review development, searching for trials (previous publication), 'Risk of bias' assessment (previous publication), data analyses (previous publication).

All authors approved the final version of this updated review.

Sources of support

Declarations of interest

S Zhang: nothing to declare.

Y Fang: nothing to declare.

BW Cai: nothing to declare.

JG Xu: nothing to declare.

C You: nothing to declare.

References