Belinostat (PXD101) br Materials and methods br Results
Materials and methods
Discussion Patients with pT2N0 MIBC who are RC candidates should receive neoadjuvant CHT, if eligible . Based on this international guideline recommendation, neoadjuvant CHT was adopted into clinical practice, albeit at a slowly increasing rate. For example, SEER-Medicare and NCDB data showed neoadjuvant CHT rates between 17 and 19% [, , , ]. The suboptimal neoadjuvant CHT use rate may be related to equivocal survival benefits, when its efficacy was examined in pT2N0 patients. For example, Grossman et al.  reported a median survival of 105 vs 75 months (p = 0.05) to pT2N0 patients respectively treated with neoadjuvant CHT or no CHT. Nonetheless, these authors demonstrated a significantly higher efficacy, when neoadjuvant CHT was administered to pT3N0/N + patients (median survival 65 vs. 24 months). A similar lack of specific protective effect was reported by Sherif et al. . Unfortunately, several other prospective trials of neoadjuvant CHT failed to report their results after Belinostat (PXD101) according to T and N stages [6,7]. In consequence, only two neoadjuvant CHT trials allow to formally distinguish the effect of neoadjuvant CHT vs. no neoadjuvant CHT between pT2N0 versus more advanced stage patients. However, both analyses where such distinction is possible demonstrate absence of statistically significant survival benefit of neoadjuvant CHT in pT2N0 patients [4,5]. Lack of benefits may represent an historical phenomenon and may be at least in part attributable to small sample sizes and selection biases. In consequence, we performed an observational study focusing on neoadjuvant CHT in a large cohort of pT2N0 MIBC patients treated with RC. We hypothesized that contemporary rates of CHT increased over time and that neoadjuvant CHT is associated with a clinically meaningful and statistically significantly lower OM and CSM. Our analyses revealed several noteworthy findings. First, the rates of neoadjuvant CHT use were higher than in four previous reports [, , , ]. Specifically, an increase of CHT administration rates from 32 to 56% between 2004 and 2015 was recorded. These observations are very encouraging with respect to adherence to perioperative CHT use guideline recommendations and indicate greater confidence in neoadjuvant CHT on behalf of involved clinicians. Fourth, it may be considered noteworthy that a difference in OM exists with respect to results addressing this endpoint within the subgroup of contemporary patients. Specifically, after multivariable and IPTW adjustment but prior to testing for immortal-time bias, the association between CHT and OM demonstrated a protective effect and was statistically significant. However, when 3-month landmark analysis was applied to this comparison (CHT vs no-CHT), the resulting OM effect was no longer statistically significant. It is of note that this methodology was uniformly applied in all survival analyses that addressed CSM and OM. This approach was based on the notion that immortal-time bias may result from consideration of immediate survival that is recorded after a specific procedure, in this case neoadjuvant CHT in RC patients, by virtue of favorable selection. Landmark analyses control for the systematic inflation of the survival benefit and should ideally be applied when a bias in survival can be operational. In the current manuscript, the application of landmark analyses indeed revealed an association between CHT and tested survival outcomes that decreased in magnitude and was no longer statistically significant, relative to results presented prior to landmark analyses. In consequence, the results of the landmark analyses indicate two important points: 1) an immortal-time bias was indeed operational in the population of contemporary patients; 2) after controlling for that bias, results no longer reached statistical significance. This methodology was applied in several previous survival analyses [14,16], even in smaller sample size analyses with more limited effect size. In these comparisons, the adjustment for immortal-time bias did not result in decreased magnitude of the protective effect or loss of statistical significance. In consequence, such methodology does not invariably eliminate the putative benefit of a treatment modality. However, its application in the current analysis did obliterate the beneficial effect of CHT on OM, as well as its significance and this result should be interpreted as final.