Trial

Risk of bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

NCT03141177 Total trial population (combined all risk groups)

Low risk of bias

Interactive Response Technology was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 3 participants randomised to the experimental arm and 8 participants randomised to the control arm did not receive any treatment. The method of analysis was appropriate.

Low risk of bias

1.7% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

Adverse events (AEs) were reported by the participants and the investigator was responsible for detecting, documenting and reporting events. Both were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Low risk of bias

A study protocol with SAP available. Safety analysis was pre‐specified in the protocol and SAP.

High risk of bias

Overall judged high risk of bias due to the outcome assessors’ awareness of assigned intervention.

NCT02811861 Comparison 1 (LEN+PEM vs. SUN) Total trial population (all combined risk groups)

Low risk of bias

Interactive voice and web response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 3 participants randomised to the experimental arm and 17 participants randomised to the control arm did not receive any treatment. The method of analysis was not appropriate (as‐treated).

Low risk of bias

2.8% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

AEs were most likely reported by the participants and assessed by the investigator. Both were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Low risk of bias

A study protocol with SAP available. Safety analysis was pre‐specified in the protocol and SAP.

High risk of bias

Overall judged high risk of bias due to inappropriate method of analysis and the outcome assessors’ awareness of assigned intervention.

NCT02811861 Comparison 2 (LEN+EVE vs. SUN) Total trial population (combined risk groups)

Low risk of bias

Interactive voice and web response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 2 participants randomised to the experimental arm and 17 participants randomised to the control arm did not receive any treatment. The method of analysis was not appropriate (as‐treated).

Low risk of bias

2.7% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

AEs were most likely reported by the participants and assessed by the investigator. Both were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Low risk of bias

A study protocol with SAP available. Safety analysis was pre‐specified in the protocol and SAP.

High risk of bias

Overall judged high risk of bias due to inappropriate method of analysis and the outcome assessors’ awareness of assigned intervention.

NCT00065468 Comparison 1 (TEM vs. IFN) Total trial population (only intermediate and poor risk groups included in the trial)

Some concerns

Participants were randomised, but no information provided about the allocation concealment. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 1 participant randomised to the single‐drug arm and 7 participants randomised to the control arm did not receive any treatment. No precise information provided about the method of analysis (as‐treated is indicated).

Low risk of bias

1.9% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to the National Cancer Institute Common Toxicity Criteria (NCICTC), version 3. Measurement of AEs could have differed between intervention groups due to differences in number of visits to the healthcare provider. AEs were assessed by the participants who were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about allocation concealment, method of analysis and method of outcome measurement; probable differences in outcome measurement between intervention arms; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT00065468 comparison 2 Comparison 2 (IFN+TEM vs. IFN) Total trial population (only intermediate and poor risk groups included in the trial)

Some concerns

Participants were randomised, but no information provided about the allocation concealment. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 2 participants randomised to the combination arm and 7 participants randomised to the control arm did not receive any treatment. No precise information provided about the method of analysis (as‐treated is indicated).

Low risk of bias

2.2% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 3. AEs were assessed by the participants who were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about allocation concealment, method of analysis and method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT00081614 Total trial population (only favourable and intermediate risk groups included in the trial)

Low risk of bias

Interactive voice response service was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was double‐blind: both participants and those delivering the intervention were not aware of assigned interventions. No information provided about the method of analysis.

Low risk of bias

Only 3 randomised did not have outcome data and 1 from the control group was lost to follow‐up.

Some concerns

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 3. Outcome assessors were not aware of the assigned intervention.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk due to lack of information about method of analysis and method of outcome measurement; missing study protocol and SAP.

NCT01024920 Total trial population (combined risk groups)

Low risk of bias

Interactive voice randomisation system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 3 participants randomised to the experimental arm did not receive any treatment. No precise information provided about the method of analysis.

Low risk of bias

3% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 3. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about method of analysis and method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT01835158 Total trial population (only intermediate and poor risk groups included in the trial)

Low risk of bias

Randomisation was performed centrally. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 1 participant randomised to the experimental arm and 6 participants randomised to the control arm did not receive any treatment. No precise information provided about the method of analysis.

Low risk of bias

4.5% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 4. AEs were assessed by the participants and the investigator. Both were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

Study protocol available with some statistical considerations briefly described, but no separate SAP available to fully check the pre‐planned analyses.

High risk of bias

Overall judged high risk of bias due to lack of information about method of analysis and method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing SAP.

NCT01984242 Comparison 1 (ATE vs. SUN) Total trial population (combined risk groups)

Low risk of bias

Interactive voice/web response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 1 participant randomised to the control arm did not receive any treatment. No precise information provided about the method of analysis, but data for crossed‐over participants were reported separately. We assume participants were analyzed as randomised in period 1.

Low risk of bias

0.3% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are very low and probably did not have an effect on the outcome.

High risk of bias

No information provided about method of measuring AEs. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT01984242 Comparison 2 (ATE+BEV vs. SUN) Total trial population (combined risk groups)

Low risk of bias

Interactive voice/web response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 1 participant randomised to the control arm did not receive any treatment. No precise information provided about the method of analysis, but data for crossed‐over participants were reported separately. We assume participants were analyzed as randomised in period 1.

Low risk of bias

0.3% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are very low and probably did not have an effect on the outcome.

High risk of bias

No information provided about method of measuring AEs. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT02684006 Total trial population (combined risk groups)

Low risk of bias

Interactive voice response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 8 participants randomised to the experimental arm and 5 participants randomised to the control arm did not receive any treatment. The method of analysis was appropriate.

Low risk of bias

1.5% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 4.03. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

Study protocol and SAP available. However, the exact time point of outcome measurement is unclear.

High risk of bias

Overall judged high risk of bias due to lack of information about method and time point of outcome measurement; the outcome assessors’ awareness of assigned intervention.

NCT00719264 Total trial population (combined risk groups)

Some concerns

Participants were randomised in a 1:1 ratio, but no information provided about who conducted the randomisation and whether the allocation was concealed. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 1 participant randomised to the experimental arm and 2 participants randomised to the control arm did not receive any treatment; 1 participant randomised to the experimental arm had no post baseline safety assessment. No precise information provided about the method of analysis.

Low risk of bias

1.1% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 3. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about randomiz sation process, allocation concealment, method of analysis and method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT00720941 Total trial population (combined risk groups)

Low risk of bias

Interactive voice response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

Some concerns

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 3 participants randomised to the experimental arm and 5 participants randomised to the control arm did not receive any treatment. The method of analysis was not appropriate (as‐treated), but this probably did not have an effect on the outcome as there is evidence that participants actually received the assigned intervention.

Low risk of bias

Less than 1% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are small and probably did not have an effect on the outcome.

High risk of bias

AEs were most likely reported by the participants. The investigator and site staff were responsible for detecting, documenting and reporting events. All were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Low risk of bias

CSR and study protocol with SAP available. Safety analysis was pre‐specified in the protocol and SAP.

High risk of bias

Overall judged high risk of bias due to inappropriate method of analysis and the outcome assessors’ awareness of assigned intervention.

NCT00732914 Total trial population (combined risk groups)

Low risk of bias

Randomisation was performed centrally. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 5 participants randomised to the experimental arm and 7 participants randomised to the control arm did not receive any treatment. No precise information provided about the method of analysis, but data for first period reported separately. We assume participants in first period received their allocated intervention.

Low risk of bias

3.3% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 3. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about the method of outcome measurement; outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT01613846 Total trial population (combined risk groups)

Some concerns

Participants were randomised in a 1:1 ratio, but no information provided about who conducted the randomisation and whether the allocation was concealed. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 6 participants randomised to the experimental arm and 5 participants randomised to the control arm did not receive any treatment. The method of analysis was appropriate.

Low risk of bias

2.9% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 4.03. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about randomistion process, allocation concealment and method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT02420821 Total trial population (combined risk groups)

Low risk of bias

Interactive voice and web response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 3 participants randomised to the experimental arm and 15 participants randomised to the control arm did not receive any treatment. Conflicting information about method of analysis in the protocol.

Low risk of bias

2% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

Adverse events were reported by the participants and/or study personnel was responsible for detecting, documenting and reporting events. Both were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Low risk of bias

A study protocol with SAP available. Safety analysis was pre‐specified in the protocol and SAP.

High risk of bias

Overall judged high risk of bias due to conflicting information about method of analysis; the outcome assessors’ awareness of assigned intervention.

NCT00920816 Total trial population (combined risk groups)

Low risk of bias

A centralized registration system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 3 participants randomised to the experimental arm did not receive any treatment. The method of analysis was appropriate.

Low risk of bias

1% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT01030783 Total trial population (combined risk groups)

Low risk of bias

Interactive voice response system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 1 participant randomised to the experimental arm did not receive any treatment. The method of analysis was not appropriate (as‐treated).

Low risk of bias

0.2% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are very low and probably did not have an effect on the outcome.

High risk of bias

Measurement of AEs could have differed between intervention groups due to differences in number of visits to the healthcare provider. AEs were assessed by the participants and the investigator. Both were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No SAP available. Study protocol available, but unclear whether it was finalized before unblinded outcome data were available.

High risk of bias

Overall judged high risk of bias due to inappropriate method of analysis; probable differences in outcome measurement between intervention arms; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT00738530 Total trial population (combined risk groups)

Low risk of bias

Interactive voice recognition system was used for randomisation. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was double‐blind: both participants and those delivering the intervention were not aware of assigned interventions. Only 2 participants from the control arm and 6 participants from the intervention arm did not receive any treatment. The method of analysis was not appropriate (as‐treated).

Low risk of bias

1.2% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

Low risk of bias

Outcome assessors were not aware of the assigned intervention.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to inappropriate method of analysis; missing study protocol and SAP.

NCT01274273 Total trial population (combined risk groups)

Low risk of bias

Participants were randomised in a 1:1 ratio. Allocation was probably controlled by an external unit. There were no baseline imbalances that would suggest a problem with randomszation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. No information provided about the method of analysis.

Low risk of bias

We assume all participants received the intended interventions.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 3. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about method of analysis and method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT01108445 Total trial population (combined risk groups)

Low risk of bias

Participants were randomised in a 1:1 ratio. Randomisation was done under allocation concealment. There were no baseline imbalances that would suggest a problem with randomisation.

High risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned intervention. Only 1 participant withdrew after consent, but before randomisation and before study drug was assigned. No precise information provided about the method of analysis.

Low risk of bias

All 108 participants were evaluable.

High risk of bias

No precise information provided about method of measuring AEs. However, AEs were defined according to NCICTC, version 4. Outcome assessors were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

No study protocol or SAP available.

High risk of bias

Overall judged high risk of bias due to lack of information about method of analysis and method of outcome measurement; the outcome assessors’ awareness of assigned intervention; missing study protocol and SAP.

NCT00903175 Total trial population (combined risk groups)

Low risk of bias

Interactive voice response system was used for randomsation. There were no baseline imbalances that would suggest a problem with randomisation.

Low risk of bias

The study was open‐label: both participants and those delivering the intervention were aware of assigned interventions. Only 2 participants randomised to the control arm did not receive any treatment. No precise information provided about the method of analysis, but data for crossed‐over participants were reported separately. We assume participants were analyzed as randomised in period 1.

Low risk of bias

Less than 1% did not receive the intended interventions and therefore did not have outcome data. However, these numbers are low and probably did not have an effect on the outcome.

High risk of bias

AEs were assessed by the participants who were aware of the assigned intervention. Knowledge of intervention received could have affected outcome measurement.

Some concerns

Study protocol available with some statistical methods described, but no separate SAP available to fully check the pre‐planned analyses.

High risk of bias

Overall judged high risk of bias due to the outcome assessors’ awareness of assigned intervention; missing SAP.