Summary of findings 1. Dexrazoxane versus no cardioprotective intervention or placebo for preventing or reducing cardiotoxicity in adults with cancer receiving anthracyclines.
| Dexrazoxane compared with no cardioprotective intervention or placebo for preventing or reducing cardiotoxicity in adults with cancer receiving anthracyclines | ||||||
|
Patient or population: adults with cancer receiving anthracyclines Settings: hospital Intervention: dexrazoxane Comparison: no cardioprotective intervention or placebo | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect (95% CI) | No of participants (studies) | Quality of the evidence (GRADE) | Comments | |
| Assumed risk | Corresponding risk | |||||
| No cardioprotective intervention or placebo | Dexrazoxane | |||||
|
Clinical heart failure Available case analysis Follow‐up ranged between 1 day and 5.1 years (nm for 5 studies) |
107 per 1000a | 24 per 1000 (12 to 46) | RR 0.22 (0.11 to 0.43) | 1221 (7 studies) | ⊕⊕⊕⊝ Moderateb,c |
In 1 study, none of the participants developed clinical heart failure; the relative effect for that study was not estimable. The available‐case, best‐case and worst‐case analyses showed identical results, including the GRADE assessment. In the worst‐case analysis, there was unexplained heterogeneity (I2 = 52%). |
|
Clinical heart failure and subclinical myocardial dysfunction combined Comparable definitions; see Characteristics of included studies for exact definitions. Available‐case analysis Follow‐up nm |
314 per 1000a | 116 per 1000 (75 to 176) | RR 0.37 (0.24 to 0.56) | 417 (3 studies) | ⊕⊕⊕⊝ Moderatec,d |
The available‐case, best‐case and worst‐case analyses showed identical results, including the GRADE assessment. |
|
Clinical heart failure and subclinical myocardial dysfunction combined Comparable definitions; see Characteristics of included studies for exact definitions. Available‐case analysis Follow‐up ranged between 1 day and 5.1 years |
312 per 1000a | 144 per 1000 (103 to 206) | RR 0.46 (0.33 to 0.66) | 534 (2 studies) | ⊕⊕⊕⊝ Moderatec,e |
The available‐case, best‐case and worst‐case analyses showed identical results, including the GRADE assessment. |
|
Overall survival (Illustrative comparative risks reported as number of alive participants) Follow‐up ranged between 1 day and 5.1 years (nm for 2 studies) |
233 per 1000f | 219 per 1000 (166 to 277) | HR 1.04 (0.88 to 1.23) | Unclear (4 studies) | ⊕⊕⊕⊝ Moderateg,h |
|
|
Progression‐free survival Defined as time from first date of complete response, partial response or stable disease until the date progressive disease was first noticed (Illustrative comparative risks reported as number of participants without progressive disease) Follow‐up nm |
0 per 1000i | 0 per 1000 (0 to 3) | HR 0.62 (0.43 to 0.90) | 164 (1 study) | ⊕⊕⊝⊝ Lowj,k |
All participants in the control group had progression at the end of follow‐up, but as the GRADEpro software was not able to calculate the corresponding risk with an assumed risk of 0%, we used 0.0001% as the assumed risk in the control group instead. |
|
Progression‐free survival Defined as time to progression; starting point nm (Illustrative comparative risks reported as number of participants without progression) Follow‐up nm |
150 per 1000l | 165 per 1000 (70‐297) | HR 0.95 (0.64 to 1.40) | Unclear (1 study) | ⊕⊕⊝⊝ Lowm,n |
|
|
Progression‐free survival Defined as time from randomisation to progression either on or off treatment (Illustrative comparative risks reported as number of participants without progression) Follow‐up ranged between 1 day and 5.1 years |
100 per 1000o | 66 per 1000 (37 to 107) | HR 1.18 (0.97 to 1.43) | Unclear (2 studies) | ⊕⊕⊕⊝ Moderatep,q |
|
|
Tumour response rate Defined as number of complete or partial remissions Available‐case analysis Follow‐up ranged between 1 day and 5.1 years (nm for 4 studies) |
533 per 1000a | 485 per 1000 (421 to 554) | RR 0.91 (0.79 to 1.04) | 956 (6 studies) | ⊕⊕⊕⊝ Moderater,s |
Due to the nature of this outcome (number of participants with a remission), a high event rate is favourable. The available‐case, best‐case and worst‐case analyses showed identical results, including the GRADE assessment. |
| Quality of life | No studies evaluated this outcome | |||||
| Secondary malignant neoplasms | No studies evaluated this outcome | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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). GRADE Working Group grades of evidence High quality: further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: we are very uncertain about the estimate. | ||||||
CI: confidence interval CTCAEv2: Common Terminology Criteria for Adverse Events, version 2 e.g.: for example HR: hazard ratio LVEF: left ventricular ejection fraction LVFS: left ventricular fractional shortening MUGA: multigated acquisition scan NCI: National Cancer institute nm: not mentioned P: P‐value RR: risk ratio
aThe assumed risk is based on the overall prevalence in the control groups of the included studies. bUnclear risk of selection bias in 5 (71%) studies, high risk of performance bias in 4 (57%) and unclear risk in 1 (14%) of the studies, unclear risk of detection bias in 2 (29%) studies, high risk of selective reporting in 1 (14%) study, unclear risk of other bias in all studies (downgraded 1 level). cWe did not downgrade for imprecision; the total number of events was fewer than 300 (the threshold rule‐of‐thumb value stated in the GRADEpro handbook (GRADEpro handbook), but the effect was large and the 95% CI is small and below no effect. dUnclear risk of selection and other bias in all studies, high risk of performance bias in all studies, unclear risk of detection bias in 1 (33%) study, high risk of attrition bias in 1 (33%) study (downgraded 1 level). eUnclear risk of other bias in all studies (downgraded 1 level). fThe assumed risk is based on the approximate mean percentage of participants alive in the control groups at the final point of the survival curves presented in the included studies. gUnclear risk of selection bias in 2 (50%) studies, high risk of performance bias in 2 (50%) studies, high risk of attrition bias in 1 (25%) study and unclear in 3 studies (75%), unclear risk of other bias in all studies (downgraded 1 level). hWe did not downgrade for imprecision; the number of events and total available participants in the 4 studies was unclear, but based on the maximum number of participants and the assumed baseline risk, we assumed that it was above 300 (the threshold rule‐of‐thumb value stated in the GRADEpro handbook (GRADEpro handbook); the 95% CI includes no effect, but was small. iThe assumed risk is based on the percentage of participants without progression in the control group at the final point of the survival curve presented in the included study (see comments for more information). jUnclear risk of selection bias, detection bias and other bias and a high risk of performance bias in the included study (downgraded 1 level). kAs this was a small study with a total number of events fewer than 300 (the threshold rule‐of‐thumb value stated in the GRADEpro handbook (GRADEpro handbook) without a large effect, we downgraded 1 level, even though the 95% CI was below no effect. lThe assumed risk is based on the approximate percentage of participants without progression in the control group at the final point of the survival curve presented in the included study. mUnclear risk of selection bias, detection bias, attrition bias and other bias and a high risk of performance bias in the included study (downgraded 1 level). nAs this was a small study with a total number of events fewer than 300 (the threshold rule‐of‐thumb value stated in the GRADEpro handbook (GRADEpro handbook), we downgraded 1 level. oThe assumed risk is based on the approximate mean percentage of participants alive in the control groups at the final point of the survival curves presented in the included studies. pUnclear risk of attrition and other bias in both studies (downgraded 1 level). qWe did not downgrade for imprecision; the number of events and available participants in the 2 studies was unclear, but based on the maximum number of participants and the assumed baseline risk we assumed that it was above 300 (the threshold rule‐of‐thumb value stated in the GRADEpro handbook (GRADEpro handbook); the 95% CI includes no effect, but was small. rUnclear risk of selection and detection bias in 4 (67%) studies, high risk of performance bias in 4 (67%) studies, high risk of attrition bias in 3 (50%) studies, unclear risk of other bias in all studies (downgraded 1 level). sWe did not downgrade for imprecision; the total number of events was more than 300 (the threshold rule‐of‐thumb value stated in the GRADEpro handbook (GRADEpro handbook); the 95%CI includes no effect, but was small.