Neuropsychiatric Adverse Drug Reactions with Tyrosine Kinase Inhibitors in Gastrointestinal Stromal Tumors: An Analysis from the European Spontaneous Adverse Event Reporting System

Round 1

Reviewer 1 Report

This is a comprehensive analysis of the major ADRs associated with a panel of RTKIs used in the treatment of GIST. This manuscript is overall well written and conclusions are supported by both the data provided and considerations of treatment pathways (eg. dosing, date of approval, characteristics of cohort). My minor point in relation to the flow of the manuscript, is that it contains a lot of abbreviations. I appreciate the reasons for this, such they may have been defined in pre-existing classfication systems (eg. MedDRA classification levels) and they may be recognisable terms to readers typically familar with such data. However as someone who isn't as familar with these terms (which may be many other readers as well), I had to return to earlier parts of the manuscript multiple times as I could not remember what a particular abbreviation stood for. For example page 3 has 11 abbreviations.  I appreciate that there is an abbreviation list and the authors included reference to abbrevations in figure legends, but constantly referring back to it broke up my reading flow (a point particulary impactful when reading online manuscripts) and made it more difficult for me to follow results. I also appreciate that this isn't necessarily something that can be resolved, but perhaps the meaning behind some of the most critical abbreviations could be explained further for those unfamilar eg. what is a high level term versus a preferred term?

Specific points:

Line 57 - more information on the D842V mutation of PDGFR would be welcomed eg. where is this mutation in the PDGFRA structure (presumably the kinase region, which of the aforementioned exons)? what is the molecular importance of it?  Is poor prognostic outcome seen for these GIST patients?

Line 67 - AVA is described as a 'highly type I RTKI'. A highly what?

Results:

1. What was the justification for removing some of the TKI-related ICRSs with other indications? As these were by far the highest proportion of ICSRs, and are still related to TKI treatment, therefore what specific criteria were used to define/exclude by 'other indications'?

I appreciate that this may not be possible due to data availability, but is it possible to sort ICSRs even further using a second characterisitic? eg. sex and age? (pre vs postmenopausal females?) Similarly, the increased neuropsychiatric ICSRs seen in the eldery could be refelcted of increased dementia risk (as acknowledged by the authors in the discussion). To investigate any age related correlation, could the adult population be subdivided further into more discrete age catagories?

A higher frequency of admisitration site ADRs were seen with SU and RIP - are these drugs delivered differently? 

Figure 3 - details of data presetnation are needed in the legend eg. what does the horizontal line represent through the bars (median presumably?) and the 'error bars' are CI? Is this data for the whole cohort of patients on each TKI? What happens if you compare across age?

Following this point, could further statistical comparisons be made between cohorts/ADR prevalence beyond ROR, CIs?

SADRs is missing from the abbreviations list and was not defined (line 173).

Minor point - line 59 'leads' should be 'led'

 

Author Response

Dear Reviewer,

We provided a point-by-point response to your comments, accordingly. All changes are reported in our manuscript on-track changes with all markup. 

• The manuscript in  This is a comprehensive analysis of the major ADRs associated with a panel of RTKIs used in the treatment of GIST. This manuscript is overall well written and conclusions are supported by both the data provided and considerations of treatment pathways (eg. dosing, date of approval, characteristics of cohort). My minor point in relation to the flow of the manuscript, is that it contains a lot of abbreviations. I appreciate the reasons for this, such they may have been defined in pre-existing classfication systems (eg. MedDRA classification levels) and they may be recognisable terms to readers typically familar with such data. However as someone who isn't as familar with these terms (which may be many other readers as well), I had to return to earlier parts of the manuscript multiple times as I could not remember what a particular abbreviation stood for. For example page 3 has 11 abbreviations.  I appreciate that there is an abbreviation list and the authors included reference to abbrevations in figure legends, but constantly referring back to it broke up my reading flow (a point particulary impactful when reading online manuscripts) and made it more difficult for me to follow results. I also appreciate that this isn't necessarily something that can be resolved, but perhaps the meaning behind some of the most critical abbreviations could be explained further for those unfamilar eg. what is a high level term versus a preferred term?

We apologize for all abbreviations included in our manuscript, but we have complied with Cancers Template and Instructions for Authors. For sake of clarity, we added the meaning of System Organ Class, High Level Term and Preferred Term in Materials and Methods section, Data Analyses subsection, page 4 lines 157-163 “The ADRs were analyzed according to MedDRA® classification level by System Organ Class (SOC) which are groupings by aetiology and manifestation site, High Level Term (HLT) based upon anatomy, pathology, physiology, aetiology, or function, and Preferred Term (PT) considered a distinct descriptor for a symptom or sign. Moreover, the individual ADRs were sorted into the equivalent SOC and the PTs corresponding to the same clinical condition were grouped under a unique term after a careful clinical evaluation.”

Specific points:

• Line 57 - more information on the D842V mutation of PDGFR would be welcomed eg. where is this mutation in the PDGFRA structure (presumably the kinase region, which of the aforementioned exons)? what is the molecular importance of it?  Is poor prognostic outcome seen for these GIST patients?

We added more information on the D842V mutation of PDGFR as the reviewer suggested (see Introduction section, page 2, lines 61-63).

• Line 67 - AVA is described as a 'highly type I RTKI'. A highly what?

We agree with the referee’s comments and carefully modified the text in “The approval of avapritinib (AVA), a highly selective type I TKI, in 2020 proved to be a milestone in PDGFRA D842V mutant GISTs” (see Introduction section, page 2, line 82).

Results:

• What was the justification for removing some of the TKI-related ICRSs with other indications? As these were by far the highest proportion of ICSRs, and are still related to TKI treatment, therefore what specific criteria were used to define/exclude by 'other indications'?

We apologize for the lack of clarity. The aim of our study is to describe better ADRs and to evaluate the reporting frequency of some toxicities through the analysis of individual case safety reports (ICSRs) among TKIs approved for GISTs collected into the European SRS database with a focus on neuropsychiatric ADRs. To avoid some therapeutic bias with other approved indications we decided to include and focus only ICSRs having GIST indication. For sake of clarity, we modified exclusion criteria in Materials and Methods section, Data sources and selection process subsection, page 3, lines 133-135.

• I appreciate that this may not be possible due to data availability, but is it possible to sort ICSRs even further using a second characterisitic? eg. sex and age? (pre vs postmenopausal females?) Similarly, the increased neuropsychiatric ICSRs seen in the eldery could be refelcted of increased dementia risk (as acknowledged by the authors in the discussion). To investigate any age related correlation, could the adult population be subdivided further into more discrete age catagories?

We thank the reviewer for this suggestion. However, it is well known that results from SRS database cannot be considered as measures of risk. The number of cases in SRS database does not correspond to the number of exposed people. This cannot be considered a pharmacoepidemiological study. Indeed, we presented our results describing better ADRs and evaluating the reporting frequency of some toxicities through the analysis of individual case safety reports (ICSRs) among TKIs approved for GISTs collected into the European SRS database with a focus on neuropsychiatric ADRs. In particular, this study supports the relevance of the SRS database to better characterize the safety profile of newly approved drugs in a real world setting. This is crucial to prevent/minimize the occurrence of clinically relevant ADRs related to the use of TKIs in GIST patients, thus improving their compliance. We are aware that spontaneous data are fraught with limitations to draw any meaningful conclusion around head-to-head comparison as we have already reported in strengths and limitation subsection. The lack of information about the total number of drug-exposed patients, as well as the poor quality of information listed in each ICSR including the missing data on age and sex, is a limit that makes impossible to calculate event rates in the absence of denominators. Moreover, the database contains only reports by patient age group: <18 years, 18-65 years, and >65 years that makes impossible to investigate any age related subdivision.

• A higher frequency of admisitration site ADRs were seen with SU and RIP - are these drugs delivered differently?

In our Results and Discussion sections, we reported a higher frequency of ADRs belonging to the System Organ Class (SOC) “General disorders and administration site conditions”. As reported in Table 2, High Level Terms involved in the SOC “General disorders and administration site conditions” were “General signs and symptoms NEC”, “Asthenic conditions”, and “Death and sudden death”. No administration site ADRs were reported. Indeed, all TKIs are administered orally, including SU and RIP.

• Figure 3 - details of data presetnation are needed in the legend eg. what does the horizontal line represent through the bars (median presumably?) and the 'error bars' are CI? Is this data for the whole cohort of patients on each TKI? What happens if you compare across age?

Figure 3 is a box plot, a method for graphically demonstrating the time to onset of neuropsychiatric ADRs grouping by each TKI through their median and quartiles. For sake of clarity, we added the description of the figure as “Data are reported as box plot with the box drawing from Q1 to Q3 and the horizontal line drawing in the middle to denote the median.” (see Figure 3, page 10, lines 273-275). In accordance with the aim of our study, the box plot defines all neuropsychiatric ADRs reported in ICSRs from the European SRS database, EudraVigilance. No data of cohort of patients are reported, making comparison by age impossible.

• Following this point, could further statistical comparisons be made between cohorts/ADR prevalence beyond ROR, CIs?

We thank the reviewer for this suggestion. However, it is well known that results from SRS database cannot be considered as measures of risk. The number of cases in SRS database does not correspond to the number of exposed people. This cannot be considered a pharmacoepidemiological study. The lack of information about the total number of drug-exposed patients is a limit that makes impossible to calculate prevalence rates in the absence of denominators.

• SADRs is missing from the abbreviations list and was not defined (line 173).

We added the definition of SADRs, accordingly (see Results section, Characteristics of ICSRs subsection, page 6, line 206).

• Minor point - line 59 'leads' should be 'led'

We modified leads in led, accordingly (see Introduction section, page 2, line 64).

Reviewer 2 Report

Dear authors,

In the present manuscript is presented a descriptive analysis of neuropsychiatric adverse drug reactions (ADRs) for TKIs used in GISTs. Data were extracted from the European database, EudraVigilance. Additionally, authors performed a disproportionality analysis of frequency of the neuropsychiatric ADRs reported for each TKI compared to all other.

Please find my suggestions and comments below:

·       Authors should detail why they choose the neuropsychiatric ADRs for the present descriptive analysis

·       Authors should provide a list of PTs related to neuropsychiatric ADRs

·       Information regarding the period of reporting for each TKI should be added in the Materials and methods section

·       It would be better if authors provide additional information regarding the average duration of treatment for each TKI (e.g. clinical studies)

·       Can the authors present some statistical data regarding the number of TKIs prescription in EU?

Author Response

Dear Reviewer,

We provided a point-by-point response to your comments, accordingly. All changes are reported in our manuscript on-track changes with all markup. 

Dear authors,

In the present manuscript is presented a descriptive analysis of neuropsychiatric adverse drug reactions (ADRs) for TKIs used in GISTs. Data were extracted from the European database, EudraVigilance. Additionally, authors performed a disproportionality analysis of frequency of the neuropsychiatric ADRs reported for each TKI compared to all other.

Please find my suggestions and comments below:

• Authors should detail why they choose the neuropsychiatric ADRs for the present descriptive analysis

We implemented our choice of neuropsychiatric ADRs, as the reviewer suggested (see Introduction section, page 3, lines 104-108). As we have already reported in the Introduction section “Currently, no original research on adverse drug reactions (ADRs) related to TKIs for the treatment of GISTs from the SRS database is available, and some questions remain to be answered, such as neurocognitive and psychiatric AEs. Premarketing studies have suggested a possible correlation between AVA and cognitive impairment [16,19], and the risk management plan of the European Medicines Agency (EMA) states that intracranial haemorrhage and cognitive effects are important identified risks for AVA [20]. However, no postmarketing data have evaluated neuropsychiatric AEs for TKIs.” Neuropsychiatric ADRs have a strong impact on patients' quality of life, especially those with GISTs. Considering the large use of TKIs in GIST patients and the important identified risk of neuropsychiatric ADRs with AVA, the aim of our study is to describe better all ADRs with TKIs. Moreover, we evaluated the frequency of reporting of neuropsychiatric ADRs not only for AVA but for all TKIs. Indeed, our results showed some neuropsychiatric ADRs not reported in literature and worthy of discussion: a higher frequency of nerve inflammation, olfactory dysfunction, and hallucinations for AVA, a small increased reporting of cerebrovascular accident for REG, and disturbances in consciousness for RIP. Moreover, we describe better the selection process in Materials and Methods section, Data sources and selection process subsection (see page 3, lines 136-140).

• Authors should provide a list of PTs related to neuropsychiatric ADRs

We provided a list of all HLTs and PTs included in neuropsychiatric ADR analysis as the reviewer suggested (see Supplementary Table S1).

• Information regarding the period of reporting for each TKI should be added in the Materials and methods section

Thanks for this suggestion. We added each period of reporting for all TKIs (see Materials and Methods section, Data sources and selection process subsection, page 3, lines 126-135).

• It would be better if authors provide additional information regarding the average duration of treatment for each TKI (e.g. clinical studies)

We provided additional information on the median duration of treatment for each TKI, accordingly (see Introduction section, page 2, lines 59-60, 67-68, 70, 76-77, 83).

• Can the authors present some statistical data regarding the number of TKIs prescription in EU?

Thanks for this suggestion. However, no free data are available for antineoplastic drug prescriptions in Europe. For this reason, it cannot be possible to present some statistical data regarding the number of TKI prescriptions. It is well known that results from SRS database cannot be considered as measures of risk. The number of cases in SRS database does not correspond to the number of exposed people. This cannot be considered a pharmacoepidemiological study. The lack of information about the total number of drug-exposed patients is a limit that makes impossible to calculate prevalence rates in the absence of denominators.

Reviewer 3 Report

1. "The aim of this study was to evaluate" should be changed to "The aim of this study is to evaluate". or change it to "The evaluation of the reporting frequency of neuropsychiatric adverse drug reactions (ADRs) for TKIs through the analysis of European individual case safety reports (ICSRs) is the aim of this study."

In the keywords part, please change "tyrosine kinase inhibitors" to "tyrosine kinase inhibitors (TKI)".

I suggest that the authors provide the chemical structures of TKIs discussed in this article by citing completely relevant review articles such as https://link.springer.com/article/10.1007/s11030-022-10406-8. It would be interesting for medicinal chemists. In this way, they will cite this article more.

"Conclusions" should be changed to "Conclusion".  

Are the TKIs reported in this study FAD-approved? or other reputed agencies please mention them and highlight them within the text. 

 

Author Response

Dear Reviewer,

We provided a point-by-point response to your comments, accordingly. All changes are reported in our manuscript on-track changes with all markup.

• "The aim of this study was to evaluate" should be changed to "The aim of this study is to evaluate". or change it to "The evaluation of the reporting frequency of neuropsychiatric adverse drug reactions (ADRs) for TKIs through the analysis of European individual case safety reports (ICSRs) is the aim of this study."

We thank the reviewer for this suggestion. We modified “was” with “is” in introduction and abstract section, accordingly (see Abstract section, line 22; Introduction section, page 2, line 109).

• In the keywords part, please change "tyrosine kinase inhibitors" to "tyrosine kinase inhibitors (TKI)".

We modified the keyword, as the reviewer suggested.

• I suggest that the authors provide the chemical structures of TKIs discussed in this article by citing completely relevant review articles such as https://link.springer.com/article/10.1007/s11030-022-10406-8. It would be interesting for medicinal chemists. In this way, they will cite this article more.

We added all chemical structures and synthesis of TKIs by citing completely relevant review articles as the reviewer suggested (see Introduction section, page 2, lines 58-59, 65-67, 68-69, 73, 83-86).

• "Conclusions" should be changed to "Conclusion". 

We modified “Conclusions” in “Conclusion”, accordingly. 

• Are the TKIs reported in this study FAD-approved? or other reputed agencies please mention them and highlight them within the text. 

All TKIs included in the analysis are approved by Food and Drug Administration (FDA) for GIST treatment. However, considering the aim of our study, we decided to cite only the Summary of Product Characteristics for each TKI related to European Medicines Agency approval.

Round 2

Reviewer 3 Report

I have no more comments.