Renal cell carcinoma (RCC), is the most common subtype of kidney cancer (Choueiri, 2011). Approximately 90% of these tumors are classified as RCC and up to 80% as “clear cell” RCC (Janzen et al., 2003; Choueiri, 2011). Most patients (>75%) are diagnosed when the disease is already advanced or metastatic, mRCC, hence no more curable (Facchini et al., 2009). Recurrent and/or mRCC is associated with a poor 5-years survival, roughly 10%; however, in the last decade, a series of novel agents have been introduced in clinical practice and the outcomes are slowly improving. Despite an increasing knowledge about the genetic and signaling abnormalities involved in the RCC carcinogenesis, such as VHL, PBRM1, SETD2, and BAP1 genes, no biomarkers are currently available, thus there are no molecular factors, which may guide the clinicians in choosing the therapeutic strategy (Brugarolas, 2014).

The advent of the target therapies (TTs) has revolutionized the mRCC treatment with an impressive effect on the overall survival (OS), which increased from an average of 9 months in 1995, when the only option in first-line was interferon alfa (IFN-α), to a median of 28–29 months in 2013, the TTs era (Motzer and Russo, 2000; Chowdhury et al., 2008; Albiges et al., 2015; de Velasco et al., 2015). The TTs include the tyrosine kinase inhibitors (TKIs), targeting the vascular endothelial growth factor receptors (VEGFRs), the mammalian target of rapamycin inhibitors (mTORis), and bevacizumab, a monoclonal antibody targeting the VEGF ligand.

Sunitinib, a TKI targeting the angiogenesis, is currently considered the standard therapy in newly diagnosed, mRCC patients, who have a good-intermediate prognosis (Albiges et al., 2015). Nevertheless, pazopanib (PAZ), a multi-targeted TKI, has been recently approved in first-line mRCC treatment and in second-line, after treatment with cytokines. The approval comes from the results of the phase III, pivotal trial that compared PAZ to placebo in a heterogeneous population, including both therapy-naïve and cytokine-pretreated patients. PFS was significantly prolonged in the PAZ arm, in the first-line subset (median 11.1 vs. 2.8 months; p < 0.0001) as well as in the second-line setting (median, 7.4 vs. 4.2 months; p < 0.001) and in the overall cohort (median, 9.2 vs. 4.2 months; p < 0.0001); the median OS was 22.9 months with PAZ and 20.5 months with placebo, with an ORR of 30 vs. 3% with placebo (Sternberg et al., 2010).

A large non-inferiority study, the COMPARZ, compared PAZ to sunitinib in first-line, showing no statistically significant difference in terms of PFS, OS, and response rate (RR), with a slightly different toxicity profile in favor of PAZ (Motzer et al., 2013a). Subsequently, the PISCES (Patient Preference Study of PAZ vs. Sunitinib in Advanced or Metastatic Kidney Cancer) trial confirmed that PAZ had a better safety profile and was associated to a better quality of life, compared to sunitinib (Escudier et al., 2014).

Although some authors have questioned the results of the above-mentioned studies, the evidences are robust enough to support the use of PAZ in first-line as valuable alternative to sunitinib. To date there are not validated prognostic and predictive biomarkers of response, and none of the available studies has provided any straight, unbiased factor, which might guide the clinician in choosing PAZ rather than sunitinib, or vice versa. In the front-line setting, hence the decision is made exclusively on the basis of the safety profile and patients medical history; in particular, PAZ is generally preferred in patients with severe cardiovascular co-morbidities or poor performance status (D'Alterio et al., 2010, 2012; Sonpavde et al., 2012; Choueiri et al., 2013; Ravaud et al., 2013; D'Aniello et al., 2014; Cavaliere et al., 2015).

Several “real world” studies have showed the efficacy and safety of PAZ in unselected populations (Galvis et al., 2013; Matrana et al., 2013a; Jonasch et al., 2014; Rizzo et al., 2015; Vogelzang et al., 2015), we thought to further reinforce such evidences publishing our own experience with the drug.

Randomized clinical trials are the gold standard to evaluate the efficacy of a new treatment within a specified patient population and the available international guidelines for physicians for treatment of mRCC essentially reflect the evidences for the different agents emerged from the respective clinical trial (AIOM, 2015; ESMO, 2015; NCCN, 2015). Nevertheless, the populations included in pivotal clinical trials are highly selected (no relevant co-morbidities, no poor performance status, no atypical metastases [e.g., brain metastases] or no histo-types different from clear cell) and generally have better outcomes compared to patients who are ineligible for such studies (Molina and Motzer, 2011; Heng et al., 2014). Therefore, it is not always simple to transfer the clinical trials results into clinical practice due to the large proportion of patients who are routinely treated in our clinics but would be judged ineligible in the pivotal trials. The optimization of the disease management should be based on clinical trials but also on real experiences (Schmidinger et al., 2013). Such registries, also known as “real world” or “real life” studies, providing insights into treatment patterns and outcomes, are emerging as a crucial tool for the clinicians because of a major reproducibility compared with randomized trials (Shek et al., 2012; Ta et al., 2013; Vaishampayan et al., 2014).

In terms of mPFS, we observed a median of 12.7 months, which is longer than the one reported in the COMPARZ (Motzer et al., 2013a) trial (8.4 months) and similar to the one detected in the pivotal phase III trial for the group of treatment-naïve patients (Sternberg et al., 2010). Looking at the different “real world” experiences, the mPFS was 8.5 months in the Vogelzang et al. (2015) analysis, 13.7 months in the MD Anderson Cancer Center study 17 (Matrana et al., 2013a), and 13.0 months in the Christie study (Galvis et al., 2013), in which the 74% of the patients receiving PAZ was treatment-naïve. The variations in the mPFS observed in these studies could have resulted from varying operational definitions of PFS such as different timing and frequency of disease assessment.

Comparing the data about survival, we noted that our median OS, 26.2 months, was longer than the ones observed in other studies, the pivotal trial (22.9 months), in the Christie study (19 months; Galvis et al., 2013), Vogelzang NJ et al analysis (22 months) (Vogelzang et al., 2015), and similar to the ones detected in the COMPARZ trial (28.4 months; Motzer et al., 2013a) and in the MD Anderson “real life” study (29.1 months; Matrana et al., 2013a). ORR and DCR appeared similar when compared to the outcomes of the randomized, phase III trial. We have been comforted by PAZ activity, therefore is not obvious to consider that patients with greater DCR and objective response obtained an improvement in OS and PFS. It is interesting to note that Karnofsky PS not influenced the OS, in contrast previous nephrectomy and an age ≤ 70 year old impact positively.

Overall our findings are very encouraging especially considering that the population was totally unselected, with a large proportion of frail patients, presenting more unfavorable prognostic factors (multiple co-morbidities, high tumor burden, worse PS and lower number of nephrectomies) compared to the available literature; our efficacy data are substantially consistent with those reported in both the prospective phase III trials and the retrospective analyses (Galvis et al., 2013; Matrana et al., 2013a,b; Motzer et al., 2013a,b; Sternberg et al., 2013; Jonasch et al., 2014; Vogelzang et al., 2015).

In regards to tolerability, we reported a safety profile slightly better compared with the other experiences. In our cohort, no grade 4 toxicity occurred and no permanent drug discontinuations due to toxicity were required. Furthermore, the rate of dose reductions due to AEs was 28%, but up to 45% received a subsequent re-escalation of dose.

The most frequent AE was hypertension observed in the 40.6% of the cases, such rate is similar to the ones observed in the COMPARZ (46%; Motzer et al., 2013a) and in the phase III pivotal trial (40%) (Sternberg et al., 2010), but superior to the one reported in the MD Anderson study (Matrana et al., 2013a) (21%) and in the analysis by Vogelzang et al. (2015) (27%). It is important to underline that about 25% of patients had poorly controlled hypertension at the times beginning PAZ, compared to the phase III trials. This medical condition was not related to a higher risk of cardiac toxicity during PAZ treatment, supporting its use even in this subset. The differences in hypertension incidence across the studies might have arisen from closer monitoring of vital signs during our study protocol according to the clinical trials. In contrast, the rates of fatigue (10.5%) and diarrhea (15.8%) were lower than previously reported: the MD Anderson study (Matrana et al., 2013a), 58 and 39%, respectively; Vogelzang analysis (Vogelzang et al., 2015), 56 and 52%, respectively; COMPARZ (Motzer et al., 2013a): 55 and 63%, respectively; Phase III trial: 19 and 52%, respectively (Sternberg et al., 2010). These differences in the incidence of fatigue are likely to be due both to the variability of the evaluation criteria adopted and to the retrospective nature of our data which probably led to an underestimation of the toxicities, especially the ones that are subjectively evaluated such as fatigue/asthenia.

The incidence of hepatic impairment was as expected (26.3%), the elevation of the liver function tests of grade 3 involved only 6 patients, and this event has been reversible and was not related to liver metastasis or other liver illness.

Six patients were treated with a lower dose of PAZ for concomitant diseases for the entire duration of treatment; we found that dose reduction was not associated with reduced efficacy, suggesting that the personalization of the drug schedule and the early management of the emerging toxicities represent crucial steps to maximize the therapeutic benefit.

We failed to identify predictive factors of response, although the occurrence of a thyroid dysfunction and grade 3 hypertension seemed to be related to a prolonged PFS and OS, respectively.

The use of PAZ out of clinical trial allowed us to evaluate its activity also in a subset of patients with histology other than clear cell, usually excluded from academic studies for their poor prognosis and limited sensibility to anticancer drugs. Our patients were not candidate for sunitinib and or temsirolimus treatment due to several specific co-morbidities. PAZ demonstrated to be still active also in this subset of patents with a mPFS of 10.6 months (95% CI, 4.4–16.8 months), data that are longer compared to those from pivotal phase III trial of sunitinib, and that will probably contribute to enrich the very few anecdotal literature data.

Our results are consistent with those reported in prospective phase III trials and the published retrospective “real world” experiences. This analysis confirm the safety and efficacy of PAZ in first-line setting, both in frail patients with multiple co-morbidities and Karnofsky PS < 80% and in younger, healthier patients with number of metastatic sites < 6. Thyroid dysfunction and hypertension may represent prognostic factors for longer PFS and OS, respectively. More extensive data and larger sample are needed in order to appropriately guide real-world practice.

SC, AC, GF, CD conceived the study, performed the statistical analysis, and wrote the paper. SR, CDP conceived the study, collected the data and wrote the paper. CC, MD, DS, GC, PMa, PMu, SiP, MB, SaP conceived the study and wrote the paper. GI, RP provided assistance with patients, wrote the paper and participate in the design of the study.