Review Reports - Can National Tests from the Last Year of Compulsory School Be Used to Obtain More Detailed Information about Academic Performance in Children Treated for Brain Tumours? A Nationwide, Population-Based Study from Sweden

Round 1

Reviewer 1 Report

The authors analyzed Swedish national school tests of oral, reading and writing skills in the first foreign language (English), the mother tongue (Swedish) and mathematics in order to evaluate school performance in children treated for brain tumours. They report results obtained from 475 children diagnosed with a brain tumour before their 15th birthday and 2,197 matched controls. Results showed that children treated for brain tumours had more difficulties with national tests than controls in almost all subtests.

Overall, this epidemiology study is well written and conceived. The two groups were matched for main demographics parameters and corrected for relatives cultural level. The results of the study provide a generic estimate of the impact of a diagnosis of brain tumor in children on the following school abilities and results.

Regrettably, from this study we do not know whether the population of cancer patients has also specific neuro cognitive impairment beyond specific school performance. I can understand that this information is difficult to obtain, but other relevant data are missing. In particular, what would be very important to know is:

Characteristics of the disease: tumor histology, treatment modalities, length of hospital stay and postoperative care
Causes of negative impact on neurocognitive performances: Neurological dysfunctions from tumor, from surgery, from radiation therapy or from long school absence

Some supplementary data would help to understand better the causes of the lower performances. The question is whether more detailed analysis of the disease of the cancer patient group could be provided.

Author Response

1. Characteristics of the disease: tumor histology, treatment modalities, length of hospital stay and postoperative care

2. Causes of negative impact on neurocognitive performances: Neurological dysfunctions from tumor, from surgery, from radiation therapy or from long school absence

Reply: Thank you for insightful comments. We agree with the reviewer that more information was needed, and now provide an updated version of Table 1, which now includes the distribution of tumour diagnoses according to the WHO diagnostic criteria at the time. Based on this, the tumours were classified as low- (WHO I or II) or high-grade (WHO III or IV). Specific details about individual treatment modalities, length of hospital stay, or postoperative care, were not available. It was assumed that the most appropriate treatment protocols and standard of care was applied for each patient and tumour diagnosis, and that treatment of high-grade tumours was more intense and included multiple modalities, for example radiotherapy. We have now added a sentence about treatment modalities in the text (lines 178-180).

Information about sequelae is unfortunately not systematically reported in the registry. Out of the 475 patients, there were free text notes about specific sequelae and other issues for 149, for example problems with hearing, vision, sensibility, motor coordination, endocrine function, memory, attention, cognition, posterior fossa syndrome, tumour relapse, epilepsy, hydrocephalus, etc., but since they are not systematically reported (all 149 are different) they are unfortunately not reliable for comparison with the others.

Reviewer 2 Report

The authors analyzed results the Swedish national school tests of children (<15yrs) who survived brain tumors. In total 475 tumor survivors were identified from the Swedish childhood cancer registry and compared to 2197 matched controls.

This is a well composed and interesting study. Some further information would however be of interest (if it can be extracted from the cancer registry): What kind of treatment did the patients receive (surgery, radiation, chemo)? As this could have practical implications for the choice and timing of treatment in some patients.

What percentage of the PBTS can be regarded as cured and how many require further or recurrent treatment?

Author Response

What percentage of the PBTS can be regarded as cured and how many require further or recurrent treatment?

Reply: Thank you for insightful comments. We agree with the reviewer that more information would be of interest, and now provide an updated version of Table 1, which includes the distribution of tumour diagnoses according to the WHO diagnostic criteria at the time. Based on this the tumours were classified as low- (WHO I or II) or high-grade (WHO III or IV). Specific details about individual treatment modalities were not available. It was assumed that the most appropriate treatment protocols and standard of care was applied for each patient and tumour diagnosis, and that treatment of high-grade tumours was more intense and included multiple modalities, for example radiotherapy. A sentence about treatment modalities is now added to the text (lines 178-180).

The overall long-term survival of all brain tumour patients in Sweden was approximately 80% [1]. Paediatric brain tumours are highly heterogeneous, and the majority are low-grade tumours, unlike brain tumours in adults. The risk of relapse is very different between tumour types and subtypes, and thereby also the treatments. As this is a population-based nationwide registry study over nine years, we consider the population representative for developed countries.

Gustafsson, G.; Kogner, P.; Heyman, M. Childhood Cancer Incidence and Survival in Sweden 1984-2010. Report 2013 From the Swedish Childhood Cancer Registry Available online: https://cceg.ki.se/documents/ChildhoodCancerIncidenceandSurvivalinSweden1984_2010.pdf (accessed on Nov 25, 2020).

Reviewer 3 Report

Lönnerblad and colleagues describe school results from a large sample of brain tumor survivors and matched controls in Sweden. This type of study is important for the field as further knowledge is needed regarding school functioning in brain tumor survivors. The primary strengths of the manuscript are related to the large, representative sample of survivors of brain tumors and the various school measures that were analyzed in the study.

Major concerns:

The primary research questions of this study (page 2) are not necessarily different from what is known in the literature (e.g., difference between brain tumor survivors and controls in school performance; effect of sex, tumor grade, and age on school outcomes). There should be stronger clarification in the introduction of why a national test (vs. school test) could provide further details on a child’s functioning.
The clinical utility of “missing information” from the national academic tests is unclear. The authors note that this could be due to school absence or fatigue, however, this information was not collected in the sample. Missing information was a strong predictor of academic performance, and thus the authors conclude that regular educational follow-ups would be needed. Given the unknown details, this conclusion does not seem warranted.

Minor points:

How was matching between survivors and controls performed (page 3)? (e.g., based on age, sex, parent education)
Children were grouped into ages 0-5, 6-9, and 10-14 years. Do authors have access to specific ages? It is suggested to explore continuous variables rather than categorical (to obtain greater power).
The authors examined sex, age at diagnosis, and tumour grade as potential covariates. What about treatment or other complications? Patients who received cranial radiation or experienced posterior fossa syndrome (cerebellar mutism syndrome) are at highest risk for impairments. Do these survivors have worse school performance? These factors may explain some of the non-significant results for group comparisons (Table 8).
There were some grammatical errors and long sentences throughout the manuscript. The manuscript would benefit from English formatting and proofreading.

Author Response

Major concerns:

The primary research questions of this study (page 2) are not necessarily different from what is known in the literature (e.g., difference between brain tumor survivors and controls in school performance; effect of sex, tumor grade, and age on school outcomes). There should be stronger clarification in the introduction of why a national test (vs. school test) could provide further details on a child’s functioning.

Reply: Thank you for insightful comments. This is a very good point and we have now tried to further explain this (lines 65-73). There are several reasons why we have analysed the Swedish national test results. One reason is that in the Swedish neuropsychological assessments, no diagnostic school-related tests are included. Another reason is that although follow-ups in Sweden after brain tumour treatment sometimes include diagnostic school tests, they are not the same across the country. Thus, the national tests are the only tests that all children in Sweden perform. The advantage of using these tests is that results can be compared to the control group, which had not been possible if we had used diagnostic school tests performed only by the children treated for brain tumours. A third reason is the possibility to assess different skill sets (oral, reading/listening and writing) in a school setting, which we consider adds to the ecological validity.

The clinical utility of “missing information” from the national academic tests is unclear. The authors note that this could be due to school absence or fatigue, however, this information was not collected in the sample. Missing information was a strong predictor of academic performance, and thus the authors conclude that regular educational follow-ups would be needed. Given the unknown details, this conclusion does not seem warranted.

Reply: We agree that this conclusion did not seem warranted. We have removed this sentence (line 319 ff). We also explain that missing information is defined as PBTS or controls where we did have information about their final grades in a particular subject, but for unknown reasons lacked information about their national test grades from one or more subtests. The national tests are compulsory for students enrolled in the regular curriculum and it is mandatory for the schools to report the results. This means that there must be very good reasons for a child not to take one or more of the national tests, or for the school not to report the results (lines 317-322). We have previously explored the final grades in these three subjects (Lönnerblad et al., Pediatr Blood Cancer 2020, 67, doi:10.1002/pbc.28014), and expected the grades from the national tests to be similar, but to our surprise the differences between PBTS and controls were smaller for the national test results than for the final grades. Compared with our study of final grades in these subjects, odds ratios for PBTS to fail national tests compared with controls were smaller than for having fail as their final grade. For example, as mentioned in the discussion (lines 688-695), in mathematics the odds ratio for fail as the final grade was 2.33 times higher for PBTS compared with controls, whereas the odds ratio for fail as the composite grade in the national tests, as investigated in this study, was only 1.50 times higher for PBTS compared with controls. In the present study, PBTS had three times more missing information compared with controls from the subtests in English. PBTS had twice as much missing information compared with controls from the subtests in Swedish and mathematics. If we had excluded all children with missing information, the exclusion rate of PBTS would have been 25.9 % (123/475) and of controls 12.4 % (273/2,197). By not excluding children with missing information we could make the comparison with their final grades in the subjects and found that missing information from the three national tests in a subject was a strong predictor for obtaining fail as the final grade in the subject. We therefore concluded that the mere existence of missing information is an important warning signal that warrants attention.

Minor points:

How was matching between survivors and controls performed (page 3)? (e.g., based on age, sex, parent education).

Reply: The Swedish Childhood Cancer Registry sent information about the PBTS to the national public agency Statistics Sweden, where all PBTS were matched to five controls each by birth year, sex and place of residence at diagnosis. PBTS were not eligible as controls and each control only appeared for one PBTS. When we reviewed background factors, we found no significant differences between the PBTS group and the control group for sex (p=0.931), parents’ education (mothers p= 0.245, fathers p=0.284) or Swedish as the first or second language (p= 0.396). This information is now added (lines 103-107).

Children were grouped into ages 0-5, 6-9, and 10-14 years. Do authors have access to specific ages? It is suggested to explore continuous variables rather than categorical (to obtain greater power).

Reply: Thank you for the suggestion. This is a very good point. However, the age groups we are referring to follow the stages in the Swedish school system and therefore represent ages when certain skills and knowledge are acquired. The knowledge and skills are then prerequisite for the next stage. Ages 0-5 are the years before school, ages 6-9 are the first years of compulsory school and ages 10-14 are the middle and later years of compulsory school. This comparison would not have been possible if we had explored age as a continuous variable. When PBTS diagnosed at ages 0-5 start school, their teachers may not have any knowledge about the treatment they have undergone. Children diagnosed at ages 6-9 were treated during the years when basic, and for most children new, skills are taught, for example English and mathematics, but many years before final grades and national tests. Children diagnosed at ages 10-14 had their diagnoses and treatments closer in time to the national tests and final grades, and hence more vivid awareness in the minds of both families and teachers. Hence, the awareness of the treatments and possible difficulties and deficits resulting from the treatments, likely vary across the stages.

The authors examined sex, age at diagnosis, and tumour grade as potential covariates. What about treatment or other complications? Patients who received cranial radiation or experienced posterior fossa syndrome (cerebellar mutism syndrome) are at highest risk for impairments. Do these survivors have worse school performance? These factors may explain some of the non-significant results for group comparisons (Table 8).

Reply: The reviewer is completely right that patients who received cranial radiation or experienced posterior fossa syndrome (cerebellar mutism syndrome) are at the highest risk for impairments. Our hypothesis was, therefore, that PBTS treated for high-grade tumours would perform less well in school. To our surprise, we did not observe a difference between PBTS treated for high- and low-grade tumours. There may be many reasons for this, as discussed in Lönnerblad et al., 2020 (Pediatr Blood Cancer 2020, 67, doi:10.1002/pbc.28014). For example, the deficits of PBTS treated for low-grade tumours may be underestimated, and some PBTS treated for high-grade tumours are not able to attend regular schools.

There were some grammatical errors and long sentences throughout the manuscript. The manuscript would benefit from English formatting and proofreading.

Reply: We have now made some changes in the manuscript. If you and the editor think the text would benefit from English formatting and proofreading, we are open to this.

Round 2

Reviewer 1 Report

This study has several limitations, but still provides interesting information that can be useful to predict scholar performances of children undergoing brain surgery.

Author Response

Thank you for your new comments. In this copy, we have made the following changes:

The word ‘have’ is changed to ‘suffer from’ (line 12).

The words ‘Swedish and English’ were changed to ‘English and Swedish’ (line 263).

The word ‘with’ is changed to ‘where’ (line 669).

The word ‘were’ was added (line 670).

The words ‘having certain’ have been deleted (line 675).

The word ‘given’ has been replaced by the word ‘provided’ (line 676).

The word ‘rather’ has been inserted (line 686).

We have also added a sentence about missing treatment details (lines 687-690)

Kind regards,

Malin Lönnerblad

Reviewer 3 Report

The authors have adequately addressed most of my comments. For the strengths/weaknesses section in the discussion, it is suggested to add a sentence or two about the missing treatment details (e.g., cranial radiation, complications) and how this could add to future research in the area of school/brain tumor research. The authors suggest that the high vs. low grade tumors address this point, but not all details can be elucidated from tumor grade alone.

Author Response

Thank you for your new comments. We agree with the reviewer that the details about missing treatment details should be commented and we have now inserted a sentence about this in the strength and weaknesses section (lines 687-690).

We have also made the following changes:

The word ‘have’ is changed to ‘suffer from’ (line 12).

The words ‘Swedish and English’ were changed to ‘English and Swedish’ (line 263).

The word ‘with’ is changed to ‘where’ (line 669).

The word ‘were’ was added (line 670).

The words ‘having certain’ have been deleted (line 675).

The word ‘given’ has been replaced by the word ‘provided’ (line 676).

The word ‘rather’ has been inserted (line 686).

Kind regards,

Malin Lönnerblad