Page 83 - 2019_12-Haematologica-web

Hydroxyurea in polycythemia vera
Table 2. Thrombosis incidence by age and history of thrombosis. Average
Age 70 years
95% CI
2.4 4.8
1.5 5.0
1.0 8.9
80 years Risk 95% CI
6.8% 2.6
6.1% 2.0
9.3% 0.0
Average
No previous thrombosis
Previous thrombosis
1.9% 0.7 3.2
1.8% 0.3 3.2
2.7% 0.6 4.7
11.1
10.2
19.7
Risk 95% CI
3.3% 2.2 4.4
3.0% 1.3 4.6
4.5% 1.1 7.9
60 years Risk 95% CI
Risk
3.6%
3.3%
5.0%
Second cancer and side effects
fore, advisable to promote new pharmacological strategies and to consider our reported thrombosis rate as a bench- mark for future comparative studies.
With regard to hematologic transformations, we observed that annual incidence of AML is fairly constant and the cumulative 10-year incidence is approximately 4% (0.4% patients/year).
In contrast, annual incidence of evolution into MF, as predicted by meta-regression, increases steeply after five years of follow up. Therefore, in the 0-5/5-10 years of observation periods, the average annual rate of MF evolu- tion was 1.0% and 5.7%, respectively. Mortality followed a similar pattern as MF, although the divergence between the two meta-regression models was much less remark- able, with an overlap in 95%CI. We retrieved an incidence of second cancer of 1.7% patients per year. However, this may not be a reliable estimate given the limited number of events and the very large between-study heterogeneity for this outcome.
The first major strength of our work is the remarkable sample size we were able to put together, which allowed us to obtain robust estimates for the most relevant out- comes in PV. However, a possible limitation of our analy- sis is that most reports did not specifically address our study questions, and consequently the relative estimates are based on raw frequency data extracted from descrip- tive tables or text. Furthermore, we cannot exclude bias in reporting events in individual studies, since most of these were not specifically designed to answer our primary questions. On the other hand, the fact that the studies did not address our question makes publication bias in favor of certain results very unlikely.
A second strength of our approach is that we managed to greatly reduce the issue of study heterogeneity by using adequate statistical methods, namely a logistic GLMM. In this way we mitigated any possible distortion. Furthermore, by adjusting for study-specific co-variates, we were able to account for the effect of the most relevant confounders, which for some outcomes (namely MF and AML) allowed us to reduce heterogeneity to negligible values. Interestingly, for most studies, we were able to extract data on study-specific confounders stratified by treatment; this was to be expected to greatly reduce the effect of “ecological bias”, which is a common issue in meta-analysis of aggregated data. Another limitation is that while our methods supposedly reduce “ecological bias”, it is probably impossible to entirely remove its effect in a meta-regression on aggregate data. Some known predictors of clinical outcomes, such as history of thrombosis (which is a well-known risk factor for recur- rences) turned out to be not significant in meta-regression. This may suggest that, under HU treatment, history of thrombosis is no longer a risk factor for recurrences; but it
The number of second cancers was too small and between-study heterogeneity too high to allow for reli- able inference on this outcome. Based on 59 events on 755 patients, pooled incidence of second cancer was 1.7% per- sons/year (95%CI: 1.3-2.2%), mainly comprising non- melanoma skin cancer.
Only two studies in our selection reported HU-associat- ed adverse events, which does not allow reliable estimates to be made.
Discussion
We systematically collected literature on the benefit-risk profile of HU treatment in patients diagnosed with PV published in the 2008-2018 period. Out of 429 records, we selected 16 reports which allowed retrieval of incidence of specific clinical outcomes in these patients: namely major thrombosis, bleeding, evolution into MF and/or AML, mortality.
Concerning thrombosis, in previous studies, the inci- dence of thrombosis in high-risk PV patients candidates to cytoreductive treatment was estimated from large patient cohorts including both patients under HU and patients not receiving cytoreduction or taking drugs other than HU,34,35 so that the effect of HU was not clearly evidenced. Overall incidence of thrombosis in our population was approxi- mately 3% per year, obtained by pooling together event rates from each study. This estimate does not account for heterogeneity across studies, yet a meta-regression analy- sis accounting for study-specific confounders, such as median age, antithrombotic therapy, CV risk factors and history of thrombosis, provides a slightly lower estimate (2.8%). This rate does not seem to change over follow-up time, as shown by a comparison between a logistic and a negative binomial model, and depends on age. Based on 2,552 patients and 469 events, estimates of thrombosis incidence rate in patients with a median age of 60, 70 and 80 years under HU treatment are 1.6%, 3.6% and 6.8%, respectively.
Contrary to the commonly held view, we did not find a statistically significant effect of history of thrombosis on incidence of new vascular events. However, this is not sur- prising in meta-regression analysis, since it is prone to the “ecological bias”, i.e. the loss of information that follows from dealing with aggregate data.36 This mirrors the effect of increasing age on the thrombotic risk of the general population observed either for arterial or thrombotic events.37,38 However, we highlight the fact that the residual incidence of thrombosis in HU-treated PV patients is still elevated, corresponding to approximately 3-fold higher than that estimated in the general population.37 It is, there-
haematologica | 2019; 104(12)
2397