Systematic review and meta-analysis of target terapies for the treatment of metastatic renal cancer

ARtIclE INFO _________________________________________________________ ___________________


INTRODUCTION
The best treatment for renal tumor is surgery, especially when the tumor is located in the kidney with no lymph node involvement or metastasis.In advanced renal cell carcinoma (ARCC), being the tumor incurable, most available treatments are palliative and the patients usually die.In those cases, the objective is to increase total survival (TS), free survival progression (FSP), response rate (RR) and quality of life (QF) of patients.
Bevacizumab is a recombinant humanized monoclonal antibody that combines to a VEGF and inhibits its biological activity.It is used intravenously.
Temsirolimus is an inhibitor of rapamicin--kinase (mTOR) with antiangiogenic effect, as well as everolimus.Everolimus is used orally.

MATERIAL AND METHODS
Systematic review (SR) was made by search of clinical randomized trials (CRT) that used TCT to treat ARCC compared to IFN-α as first and second lines of treatment.
The search was made at the databases EM- The following key words were used for the search: ("Kidney Neoplasms" [Mesh] OR kidney cancer OR renal carcinoma) AND ("Randomized Controlled Trial" [Publication Type] OR random* OR single blind OR double blind "First Line Therapy and Renal Cancer") AND ("First Line Therapy and Renal Cancer and Randomized Controlled Trials" [Mesh]) AND ("Second Line Therapy and Renal Cancer and Randomized") [MESH].
The selected studies were realized from January 2000 to December 2011, in English, Spanish or Portuguese.
The obtained summaries were evaluated by two independent reviewers and those who fulfilled the select criteria were pre-selected.When both reviewers disagreed, the article(s) was (re) reviewed by a third reviewer.
The inclusion criteria were: multicentric randomized double-blind studies that compared TCT with IFN-α.All studies that did not meet those criteria were excluded and those in duplicity were considered only one time.
The studies were identified by the reviewers by the first author's name and year of publishing.All data were directly obtained from the studies or calculated based on the available information: epidemiological data, methods and results, including FSP, TS and RR.Bias methodological aspects (5) were also analyzed, including randomization methods, double-blind aspects, intention of treatment, loss of patients, sample size, multicentricity and study sponsors.
Statistic analysis was based on the null hypothesis that TCT does not change the results of the treatment.It was used the software Review Manager 5.1 (Cochrane Collaboration Software) (6).In the hypothesis test the level of significance was α = 0.05 for the rejection of the null hypothesis, in a two-tailed test.

META-ANALYSIS
Meta-analysis of the quantitative variables related to TS, FSP and RR were done by Relative Risk (RRi) (6).

HETEROGENEITY ANALYSIS
In the present study we used the heterogeneity index I 2 based on the chi-square test calculus (7).When it was observed heterogeneity at meta-analysis with I 2 > 70% the reason was researched (5) using new analysis excluding discrepant studies in order to obtain new heterogeneity indexes (5).

Systematic Review
After the search using the selected key words and databases, 148.805 studies were initially selected, related to renal cancer.In order to select those randomized studies that used TCT, new key words were applied and it was obtained 375 references.Among these, after analysis of the summary, 33 were selected to detailed analysis of the complete article and 21 were excluded (Figure -1).

CASES
Most patients of those studies, although with different selection criteria, had the following characteristics: age greater than 18 years old, metastatic renal clear-cells tumor, no previous systemic therapy, good medullar response, good hepatic and renal function, good condition, ECOG (0-1) or Karnofsky ≥ 60 or 70%.Patients with metastasis, radicular compression, uncontrolled systemic hypertension, pregnancy and uncontrolled thyroid diseases were not included.

Quality of the Studies
The evaluated studies were heterogeneous and most were sponsored by the laboratories manufacturers.Most were randomized studies, double-blind and multicentric.The studies were classified as originals or complementary: the originals are initial studies of analysis of first publication (interim analysis) and the complementary when final analysis was described in a posterior publication (Table -1).

META-ANALYSIS
Not all selected studies were evaluated at meta-analysis due to the selected end points.Only seven articles of first line treatment and two of second line were evaluated.

Free Survival Progression (FSP)
Regarding FSP, meta-analysis used 1375 treated patients with target cell therapies (first line treatment) versus 1353 treated patients with interferon-alpha (control group).Among these patients, 628 progressed in group TCT and 746 in control group (Figure -2).

Total Survival
Four studies evaluated TS.Rini 2010 (16) did not inform how many patients survived or died in each arm, and the study was excluded at final

Respost Rate
RR was analyzed in 1375 patients treated with TCT and 1353 patients with interferon-alpha.Among these patients, 384 and 141 showed complete response respectively (Figure -4).

Heterogeneity (I 2 )
Due to the high level of I 2 = 94% at meta--analysis of FPS and of I 2 = 68% at objective RR, several analysis were made excluding each work of the respective meta-analysis.TS meta-analysis showed I² = 0%, so the analysis of heterogeneity was unnecessary.
During FSP analysis the maximum value obtained for heterogeneity was observed when the study of Rini et al (2010 (16) was excluded: I 2 = 89%, showing great heterogeneity among the studies.After excluding the study of Motzer et al., 2007 (13) during RR analysis heterogeneity value was I 2 = 0%.

Second line studies
It was not possible to perform meta-analysis, since populations of both studies were quite different.One of the studies included patients  previously treated with systemic therapy and the other patients treated previously with sunitinib, sorafenib or both (Escudier et al., 2007 (18) and Motzer et al., 2008 (19), respectively).

SURVIVAL CURVES
FSP curves of first line treatments showed little improvement, in order of importance, with the use of sunitinib, bevacizumab and interferon--alpha.TS curves were very similar and with no statistical differences (Figure -5).
In relation to first line treatments of high risk patients, FSP and TS curves showed a little better improvement with temsirolimus, compared to interferon-alpha (Figure -6).
TS curves of second line treatment were very similar and with no significant difference, while FSP curves showed some advantages, in order of importance, with the use of everolimus,  ).As previously related, second line treatments were very heterogeneous and it was not possible to perform meta-analysis.

DISCUSSION
The lack of suitable studies that evaluate efficiency of target cell therapies (TCT) for advanced renal cell carcinoma (ARCC) was noted during systematic review and only a few were submitted to meta-analysis: seven first line treatments and two second line treatments.
It was also observed difficulty to collect data from the studies, since some presented not clear and non-objective results.Among the revised studies, there were several difficulties regarding interpretation of results, such as randomization, quantification of response rate, number of deaths during the study, that impaired TS calculus.
Many selected works were excluded since they were not prospective and with a control group, randomized and with the necessary information for analysis.Subgroup analysis studies were also excluded as well as those with duplicated populations.
In relation to RR meta-analysis, treatment with sunitinib was superior (RR 3.83 and CI [2.86-5.12]),compared to other drugs.This difference was so bigger that, when the heterogeneous analysis was performed, the initial I² was 68% and became 0% after the exclusion of that paper (Motzer et al., 2007 (13)).
All simulations of heterogeneity analysis of FSP showed high I² (> 50%) demonstrating that the five first line treatment studies are not homogeneous.The differences may be related to varied results of efficiency of the four drugs, population characteristics and methodologies.It was not possible to prove that the five studies presented homogeneously results of FSP.In the future, further studies are necessary with a bigger and more homogeneous population.

CONCLUSIONS
For first line treatment, sunitinib is the most efficient therapy regarding FSP; the studies did not show any improvement of TS and RR was low, but significant, using sunitinib and bevacizumab.Available studies could not conclude the use as second line therapies.Those results must be carefully analyzed due to the small number of available studies.

Figure 2 -
Figure 2 -Meta-analysis of free survival progression.

Figure 5 -
Figure 5 -Free survival progression and total survival curves for first line treatment.

Figure 6 -
Figure 6 -Free survival progression and total survival curves for first line treatment of high risk patients.

Figure 7 -
Figure 7 -Free survival progression and total survival curves for second line treatment.
BASE, LILACS, MEDLINE, The Cochrane Controlled Trials Register, Cochrane Database of Systematic Reviews and Cochrane Clinical Trials.Other sources included American Society of Clinical Oncology, European Society of Medical Oncology, Brazilian Society of Clinical Oncology, Cancer Brazilian Society, American Society of Urology and Brazilian Urological Society.

Table 1 -Metodological characteristics of the included studies.
A -Acceptable; Y -Yes N -No; O -Original study; C -Complementar study; NC -not clear