In order to facilitate long-term treatment decisions, we aimed to define biomarkers defining the probability of receiving second-line (SL) targeted therapy (TT) in patients with metastatic renal cell carcinoma (mRCC) based on their characteristics present at first-line TT initiation. We analysed 152 consecutive mRCC patients treated and used multivariable binominal logistic regression to identify factors contributing to the probability of receiving SL TT. Final model was assessed with bias-corrected indices (Nagelkerke’s R and area under receiver operating characteristic curve [AUC]) and two bootstrap procedures were used for internal validation. Factors associated with the probability of SL TT eligibility were the presence of brain metastases (odds ratio [OR] 0.084, 95% confidence interval [CI] 0.010–0.707), number of metastatic sites (OR 0.740, 95% CI 0.575–0.953 per each site), platelet count (OR 0.971, 95% CI 0.947–0.997, per 10 /ml), lactate dehydrogenase level (OR 0.952, 95% CI 0.910–0.997 per 10 units/l), and albumin concentration (OR 1.924, 95% CI 1.057–3.503 per 1 g/dl). We developed on-line calculator that enables practicing clinicians to estimate SL treatment probability (http://www.r-calc.com). Keywords Metastatic renal cell carcinoma · Probability calculator · Second-line · Sequential treatment · Tyrosine kinase inhibitor Introduction or stable disease (SD) due to the development of treatment acquired resistance . Moreover, 20% of patients present Currently multiple antiangiogenic compounds includ- with initial endogenous resistance to TKIs [4, 5]. Out of all ing bevacizumab (anti- vascular endothelial growth fac- patients with mRCC that progress on first-line of treatment, tor [VEGF] antibody), sorafenib, sunitinib and pazopanib between 20 and 60% will receive second-line therapy [6–10]. (tyrosine kinase inhibitors [TKIs] targeting VEGF receptors) National cancer treatment programmes and/or drug pre- are first-line standard-of-care treatment options of metastatic scription registries that cover the whole country populations renal cell carcinoma (mRCC) providing progression-free provided data on new anti-mRCC therapies within real-life survival (PFS) benefit as proven in randomised phase III patients [8, 9]. Currently, patients who receive second-line trials . Nevertheless, these antiangiogenic therapies rarely therapy are expected to reach median overall survival (OS) provide complete or long-term responses . About 80% over 27 months, while those who are enrolled in three or of patients will experience disease progression after first more lines of treatment may obtain over 43 months of OS year of treatment, also despite initial partial response (PR) and are greatest beneficiaries of RCC targeted therapies (TT). Up to 85% patients are expected to receive sunitinib as first-line treatment . The percentage of patients who * Anna M. Czarnecka receive a second-line treatment is similar between sunitinib email@example.com (59%) sorafenib (52%) and bevacizumab (79%)  treated Department of Oncology, Military Institute of Medicine, patients. Majority of patients in SL are treated with everoli- Szaserow 128, 04-141 Warsaw, Poland mus (40–60%) or sorafenib (up to 30%) [11, 12]. Present Address: Maria Sklodowska-Curie Memorial Until now, the Memorial Sloan-Kettering Cancer Center Cancer Center and Institute of Oncology, Wawelska 15, (MSKCC) classification score (Motzer Score) and first- 00-001 Warsaw, Poland line treatment type were considered as only established Present Address: Medical University of Warsaw, Zwirki i predictive factors of receiving second-line therapy . Wigury 61, 02-091 Warsaw, Poland Vol.:(0123456789) 1 3 91 Page 2 of 9 Medical Oncology (2018) 35:91 Moreover, early progression is also significantly associated the time from the initiation of first-line TT to death from with a higher probability of not receiving second-line anti- any cause, (2) PFS which was defined as the time from the mRCC treatment . Nevertheless, still little is known on initiation of first-line TT to disease progression accord- predictive factors of second-line therapy enrolment in RCC ing to the Response Evaluation Criteria in Solid Tumours patients. Preclinical data suggest that the main downstream (RECIST), version 1.1, or death from any cause, and (3) effectors of mammalian target of rapamycin signalling cas- post-progression survival (PPS), which was defined as the cade—S6RP protein and its phosphorylated form—may time from disease progression on first-line TT to death become reliable predictive biomarkers of potential response from any cause. Medians and ranges were used to describe to everolimus , but for everyday practice clinical factors continuous variables whereas frequencies and percentages seem to be more suited. The goal of our study was to ana- were used to describe categorical variables. The differ - lyse these questions in a series of subsequent RCC patients ences in baseline characteristics between the SL and non- treated in community-oriented treatment program at insti- SL groups were assessed using the U-Mann-Whitney test tution recognised for strong patient satisfaction scores and for continuous variables and the Pearson Chi-Square or standards compliance. We sought to identify pre-treatment the Fisher’s exact test (in the case of five or less expected clinical parameters that could help predicting the likelihood frequencies in each cell of a studied contingency table) of a patient receiving second-line therapy and to develop a for categorical variables. Distributions of OS, PFS and toll—calculator—enabling patients stratification. PPS were estimated using the Kaplan–Meier product-limit method; their medians with calculation of 95% confidence interval (CI) using log–log transformation were reported. Materials and methods The differences in survival probabilities between the SL and non-SL groups were assessed using the log-rank Patients test. The median follow-up time was calculated using the Schemper and Smith method . Patients’ data were last Consecutive mRCC patients who started treatment with first- updated on August 01, 2017. Patients, who were either line TT between November 2009 and March 2016, in the alive on that date or lost to follow-up, were censored in Department of Oncology, Military Institute of Medicine in survival analysis. Warsaw, Poland were included in the analysis. Patients with The identification of factors that independently predicted any histological RCC subtype with no other primary malig- receiving second-line therapy was conducted using a two- nancies and no adjuvant therapy were eligible. Additionally, step procedure based on binominal logistic regression. In the patients who were treated with interferon-based immuno- first step, all factors were included in univariable analysis therapy prior to the initiation of first-line TT were included; and these factors that reached P value less than 0.1 were however, IFN was not counted as a line of treatment. Patients included in the second step, i.e. multivariable analysis based were assigned to the second-line (SL) group if they had on step-wise forward selection with significance level of 0.05 received any of the second-line TT therapy, or to the non- for entering and removing variables. Factors that remained SL group, if they had not received any therapy beyond first- significant in the second step contributed to the final model. line. Patients with unknown status of second-line TT were The model performance was assessed with Nagelkerke’s R excluded from analysis. This group comprised of patients and bias-corrected Nagelkerke’s R as global goodness-of- who (1) continued treatment as had not progressed on first- fit measures, the Hosmer–Lemeshow test for calibration, an line TT at the time of the final data collection or (2) discon- area under receiver operating characteristic curve (AUC) and tinued treatment due to toxicity/consent withdrawn, not pro- bias-corrected AUC for discrimination. gression or (3) were lost to follow-up before second-line TT To assess the robustness of the model, internal validation initiation. Inclusion criteria for FL and SL covered adequate was performed using two bootstrap procedures that gener- organ function as described before [14–19]. ated new datasets by taking samples from original dataset The individual medical records were analysed. The insti- using random sampling with replacement. In the first proce- tutional ethics committee approved the study (agreement no. dure, 1000 new datasets were generated and binominal logis- 48/WIM/2014). Due to retrospective design of the analysis, tic regression was repeated for each sample, using variables individual informed consent was not required. selected in the final model. The odds ratios (ORs) with new 95% CIs and P values were produced and compared to those Outcomes and statistical methods of the model derived from original dataset. In the second procedure, another 1000 bootstrap datasets entered the same The status of second-line TT (received versus not received) modelling process used to derive the final model from the was a dependent binary variable for the main analysis. The original dataset. Factors that appeared in more than 50% of other assessed outcomes were (1) OS which was defined as computed models were considered to be significant [20, 21]. 1 3 Medical Oncology (2018) 35:91 Page 3 of 9 91 Cases with variables that contained missing data were 6.1–10.2) for all analysed patients. The median PFS was 11.7 excluded from analyses that involved those variables. P months (95% CI 9.0–14.1) and 4.9 months (95% CI 3.5–5.6) values less than 0.05 (two-sided) indicated statistical sig- for the SL-group and the non-SL group, respectively. The nic fi ance for all tests, except univariable logistic regressions median OS was 30.4 months (95% CI 26.2–37.8) and 7.4 where the cut-off level of 0.1 was used. All statistical proce- months (95% CI 5.5–10.3) for the SL-group and the non-SL dures were performed using Stata, version 14.2 (StataCorp, group, respectively. The median PPS was 14.9 months (95% College Station, Texas, USA) and R, version 3.2.5 (The R CI 13.5–16.7) and 1.9 months (95% CI 1.2–3.0) for the SL- Foundation for Statistical Computing, Vienna, Austria) with group and the non-SL group, respectively. The SL-group the rms package, version 5.1-0. had significantly longer PFS (P < 0.001), OS (P < 0.001) and PPS (P < 0.001) than the non-SL group (Fig. 1A-C). Results Model building and validation Characteristics of the two groups After performing a series of univariable binominal logistic regressions, 17 factors were found to have an influence on Overall, 326 patients treated with first-line TT were the probability of having second-line TT (Table 2). On the screened. Two hundred and sixty-seven (267 [100%]) multivariable analysis, five factors remained significant and patients had known second-line TT status and, therefore, contributed to the final model. Four of them were associ- were included in the analysis. One hundred and fifty-two ated with decreased probability of having second-line ther- (152 [57%]) patients had received second-line TT (everoli- apy: the presence of brain metastases (OR 0.084, 95% CI mus − 117/152 [77%], axitinib − 32/152 [21%], and cabo- 0.010–0.707), number of metastatic sites (OR 0.740, 95% zantinib − 3/152 [2%]) and contributed to the SL group. The CI 0.575–0.953 per each site), platelet count (OR 0.971, remaining 115 [43%] patients were not eligible to receive 95% CI 0.947–0.997, per 10 /ml) and lactate dehydrogenase any subsequent systemic treatment and were assigned to (LDH) level (OR 0.952, 95% CI 0.910–0.997 per 10 units/l), the non-SL group. The detailed characteristics collected at while albumin concentration was associated with increased the time of first-line TT initiation are presented in Table 1. probability (OR 1.924, 95% CI 1.057–3.503 per 1 g/dl). The Patients in the SL-group had less frequent diagnosis-to- model showed satisfactory calibration (the Hosmer–Leme- treatment interval < 1 year and Fuhrman grade 3–4 than show test P value = 0.133), discrimination (AUC = 0.750, patients in the non-SL group. At the same time, patients bias-corrected AUC = 0.736) and global fit (Nagelkerke’s 2 2 receiving second-line TT presented with better performance R = 0.277, bias-corrected Nagelkerke’s R = 0.231). In the status and were more frequently assigned to the International first validation procedure, all five model covariates remained Metastatic Database Consortium (IMDC) favourable- and statistically significant after repeating the regression on 1000 intermediate-risk groups at treatment initiation than patients bootstrap samples. In the second validation procedure, four with no systemic treatment beyond first-line. The SL-group factors: number of metastatic sites, LDH, platelet count and was characterised with lower total number of metastatic albumin concentration appeared in more than a half of 1000 sites, and lower proportion of patients had bone, liver and newly constructed models (51, 52, 56 and 52%, respec- brain metastases. Patients in the SL-group at treatment ini- tively), whereas the brain metastases status did not (45%) tiation had not only higher levels of haemoglobin and albu- (Table 3). min concentration, but also lower levels of corrected calcium The regression equation was used to construct a calcula- concentration and platelet count. There were no significant tor, named MRCCSECLINE, which gives the probability of differences between the two groups in terms of first-line tar - having second-line TT in MRCC patients. A free version of geted drug or other characteristics. the calculator is available at http://www.r-calc.com. Survival results Discussion The median follow-up time for the whole cohort of patients was 69.3 months (95% CI 64.1–73.1). The median follow-up Between 2006 and 2011, the use of TTs in patients with time was 69.2 months (95% CI 58.5–74.7) in the SL-group mRCC increased from below 23% to over 70% . The and 71.2 months (95% CI 65.0–78.0) in the non-SL group, population-wide studies show that currently, approximately respectively. The follow-up time did not differ between the 95% of all patients are treated with TT of at least one line two groups (P = 0.496). The median PFS was 8.0 months . The best clinical outcomes are achieved with sequential (95% CI 6.7–9.4), the median OS was 20.0 months (95% use of targeted drugs which is a mainstream in present and CI 17.5–24.7) and the median PPS was 7.7 months (95% CI near-future therapy of mRCC . However, about 50% of 1 3 91 Page 4 of 9 Medical Oncology (2018) 35:91 Table 1 Patients characteristics at the start of first-line TT (total N = 267) Variable The SL group (N = 152) The non-SL group (N = 115) P Age, years: median (range) 62 (25–83) 61 (22–85) 0.656 Male: n (%) 102 (67) 81 (70) 0.562 2 a b BMI [kg/m ]: median (range) 25.7 (17.1–48.8) 26.0 (16.8–39.6) 0.433 Time since diagnosis to first-line TT initiation < 1 year: 66 (43) 66 (57) 0.024 n (%) Karnofsky PS: n (%) < 0.001 100 79 (52) 23 (20) 80–90 72 (47) 84 (73) < 80 1 (< 1) 8 (7) Primary tumour site, right: n (%) 69 (45) 60 (52) 0.272 c d Fuhrman grade, 3–4: n (%) 49 (35) 54 (52) 0.008 Non-clear cell histology: n (%) 10 (7) 6 (5) 0.643 Sarcomatoid features: n (%) 8 (5) 7 (6) 0.772 Number of metastatic sites: median (range) < 0.001 Metastatic sites: n (%) Lung 112 (74) 86 (75) 0.839 Lymph nodes 71 (47) 65 (57) 0.112 Bone 41 (27) 46 (40) 0.025 Liver 26 (17) 33 (29) 0.024 Pancreas 14 (9) 11 (10) 0.922 Suprarenal gland 21 (14) 26 (23) 0.062 Brain 1 (< 1) 15 (13) < 0.001 Local recurrence 32 (21) 35 (30) 0.08 Contralateral kidney 13 (9) 7 (6) 0.449 Other soft tissues 30 (20) 38 (33) 0.013 Haemoglobin [g/dl]: median (range) 13.1 (9.6–19.1) 11.8 (8.9–17.4) < 0.001 Corrected calcium [mg/dl]: median (range) 9.5 (8.0–11.3) 9.6 (6.8–14.7) 0.043 e f Lactate dehydrogenase [U/l]: median (range) 177 (106–406) 184 (115–1185) 0.285 Albumin [g/dl]: median (range) 4.3 (2.9–5.6) 3.9 (2.3–5.9) < 0.001 WBC [× 10 /ml]: median (range) 7.6 (3.4–15.4) 7.8 (3.5–20.5) 0.28 Neutrophil count [× 10 /ml]: median (range) 4.8 (2.0–11.5) 5.1 (2.2–19.1) 0.09 Platelet count [× 10 /ml]: median (range) 250 (101–831) 299 (140–966) < 0.001 Lymphocyte count [× 10 /ml]: median (range) 1.6 (0.4–4.56) 1.6 (0.2–4.8) 0.168 IMDC risk group: n (%) < 0.001 Favourable 69 (46) 25 (22) Intermediate 75 (49) 66 (57) Poor 8 (5) 24 (21) Prior immunotherapy: n (%) 15 (10) 14 (12) 0.549 First-line TT therapy: n (%) 0.075 Sunitinib 114 (75) 83 (72) Pazopanib 23 (15) 27 (23) Sorafenib 15 (10) 5 (4) BMI body mass index, IMDC International Metastatic Renal Cell Carcinoma Database Consortium, KPS Karnofsky performance status, LDH lactate dehydrogenase, LLN lower limit of normal, SL second-line, TT targeted therapy, ULN upper limit of normal, WBC white blood count Number of evaluated patients: 143 Number of evaluated patients: 96 Number of evaluated patients: 139 Number of evaluated patients: 103 Number of evaluated patients: 148 Number of evaluated patients: 110 1 3 Medical Oncology (2018) 35:91 Page 5 of 9 91 metastatic sites, presence of bone, liver and brain metas- tases, haemoglobin, calcium, albumin and platelet count, which are widely recognised as independent RCC prog- nostic factors . Not surprisingly, it translated into more frequent assignment of patients in the non-SL group to the IMDC intermediate- and poor-risk groups than those in the SL group. Additionally, patients in the SL group had less frequently Fuhrman grade 3–4 histopathology which stays in accordance with previous report stating that patients with grade 1 tumour received second-line therapy more frequently than those with grade 2/3 tumours . Likewise in other reports, in patients ineligible for second-line treatment—the IMDC status at first-line treatment initiation is more often intermediate (~ 50) or poor (~ 40%), age is higher (age > 75 in ~ 40%), nephrectomy was less often performed (~ 60%), but metastases are more often found in liver (~ 20%), bones (~ 30%), skin/soft-tissue (~ 30%) and central nervous system (13%) . Overall, five factors were recognised as independently influencing the probability of receiving second-line treat- ment: platelet count, LDH and albumin levels, total num- ber of metastatic sites and the presence of brain metastases. Within these, brain metastases status had the largest impact on the calculated probability. For example, a hypothetical patient with platelet count of 200 000/ml, LDH level of 100 U/I, albumin concentration of 4 g/dl and metastases to two organs other than brain has the probability of 80% to receive second-line therapy, but only 25% if brain is within the two organs affected by the metastatic process. Notably, the choice of first-line agent may not be predictive for receiv - ing second-line treatment as it was reported previously by Leavy et al. . The internal validation confirmed the appropriate con- struction of our model because all variables of the regression formula in the first bootstrap procedure and four (platelet count, LDH and albumin levels, total number of metastatic Fig. 1 The Kaplan–Meier curves for a progression-free survival sites) in the second bootstrap procedure remained signifi- (PFS), b overall survival (OS), and c post-progression survival (PPS) cant. The brain metastases status did not reach the planned stratified by second-line targeted therapy (TT) status 50% frequency of entry probably due to statistical uncer- tainty caused by small proportion of patients who had brain patients will not receive second-line treatment and, there- metastases (6% in the analysed cohort). fore, their survival benefit will strictly depend on first-line The median follow-up time in our study was about was treatment efficacy. Thus, the proper identification of patients almost six years which is one of the longest reported in the ineligible for subsequent therapy becomes essential in a con- literature [24, 25]. Such long follow-up increases the reli- struction of a long-term treatment plan. Herein, we aimed ability of the research results because our study captures to develop a calculator that could predict the probability of more patients who might not be recognised as receiving second-line treatment based on patient characteristics pre- second-line in a case of long duration of first-line treat- sent at first-line therapy initiation. ment and short follow-up period. The median OS of In our study, the proportion of patients not receiving sec- 30.4 months for the SL-group is very close to 29.5 months ond-line therapy was 43% and was similar to those reported reported recently for sunitinib-everolimus sequential treat- previously [6–10]. Patients in the SL and non-SL groups ment which actually was a common therapeutic strategy differed in numerous baseline features, including perfor - in our patients . The median OS of 7.4 months in the mance status, diagnosis-to-treatment interval, number of non-SL group echoes the median OS of patients assigned 1 3 91 Page 6 of 9 Medical Oncology (2018) 35:91 Table 2 Results of univariable Variable Univariable Multivariable and multivariable binominal logistic regression with second- OR (95% CI) P OR (95% CI) P line targeted therapy status as a Age 1.00 (0.976–1.025) 0.986 dependent variable Gender Male 1 Female 1.168 (0.691–1.973) 0.562 BMI 1.030 (0.976–1.087) 0.281 Time since first-line TT initiation ≥ 1 year 1 < 1 year 0.570 (0.349–0.929) 0.024 KPS ≥ 80% 1 < 80% 0.089 (0.011–0.719) 0.023 Primary tumour site Right 1 Left 1.312 (0.807–2.133) 0.273 Fuhrman grade 1–2 1 3–4 0.494 (0.294–0.831) 0.008 Histology Clear-cell 1 Other 1.279 (0.451–3.628) 0.643 Sarcomatoid features No 1 Yes 0.857 (0.302–2.436) 0.772 No. of metastatic sites 0.600 (0.483–0.743) < 0.001 0.740 (0.575–0.953) 0.020 Lung metastases No 1 Yes 0.944 (0.542–1.643) 0.839 Lymph nodes metastases No 1 Yes 0.674 (0.414–1.098) 0.113 Bone metastases No 1 Yes 0.554 (0.330–0.929) 0.025 Liver metastases No 1 Yes 0.513 (0.286–0.920) 0.025 Pancreas metastases No 1 Yes 0.959 (0.418–2.199) 0.922 Suprarenal gland metastases No 1 Yes 0.549 (0.291–1.035) 0.064 Brain metastases No 1 Yes 0.044 (0.006–0.340) 0.003 0.084 (0.010–0.707) 0.023 Local recurrence No 1 Yes 0.610 (0.349–1.063) 0.081 1 3 Medical Oncology (2018) 35:91 Page 7 of 9 91 Table 2 (continued) Variable Univariable Multivariable OR (95% CI) P OR (95% CI) P Contralateral kidney metastases No 1 Yes 1.443 (0.557–3.741) 0.451 Other soft tissues metastases No 1 Yes 0.498 (0.285–0.870) 0.014 Haemoglobin [g/dl] 1.277 (1.114–1.465) < 0.001 Corrected calcium [mg/dl] 0.737 (0.520–1.045) 0.087 Lactate dehydrogenase [× 10 U/l] 0.952 (0.918–0.986) 0.007 0.952 (0.910–0.997) 0.035 Albumin [g/dl] 3.379 (2.049–5.572) < 0.001 1.924 (1.057–3.503) 0.032 WBC [× 10 /ml] 0.919 (0.830–1.017) 0.101 Neutrophil count [× 10 /ml] 0.863 (0.763–0.977) 0.020 Platelet count [× 10 /ml] 0.967 (0.948–0.987) 0.002 0.971 (0.947–0.997) 0.027 Lymphocyte count [× 10 /ml] 1.326 (0.953–1.843) 0.094 Prior immunotherapy No 1 Yes 0.790 (0.365–1.710) 0.549 First-line TT therapy Sunitinib 1 Pazopanib 2.184 (0.764–6.247) 0.145 Sorafenib 0.620 (0.332–1.157) 0.133 BMI body mass index, CI confidence interval, KPS Karnofsky performance status, LDH lactate dehydro- genase, LLN lower limit of normal, OR odds ratio, SL second-line, TT targeted therapy, ULN upper limit of normal, WBC white blood count Number of evaluated patients: 239 Number of evaluated patients: 242 Number of evaluated patients: 258 Number of evaluated patients: 263 Table 3 The results of bootstrap Variable Frequency of Entry OR (95% CI) P procedures for multivariable (%) binominal logistic regression with second-line targeted Brain metastases 45 0.084 (0.026–0.274) < 0.001 therapy status as dependent No. of metastatic sites 51 0.740 (0.568–0.965) 0.026 variable Platelet count [× 10 /ml] 56 0.997 (0.994–0.999) 0.042 Lactate dehydrogenase [× 10 U/l] 52 0.995 (0.991–0.999) 0.020 Albumin [g/dl] 52 1.924 (1.014–3.650) 0.045 CI confidence interval, OR odds ratio to the IMDC poor-risk group in other populations stud- treatment . However, this parameter will not be known ies . Interestingly, PFS and PPS were also shorter in at the start of first-line treatment. patients not receiving second-line therapy, which may sup- Nowadays, everolimus, axitinib, nivolumab and cabo- port the thesis that first-line PFS may act as a surrogate zantinib are used extensively in patients who progressed on end-point for overall OS . What is more, Eggers et al. prior antiangiogenic TKI therapy. Currently with multiple reported that early progression, defined as progression treatment options, including immunotherapy, reimbursed within 6 months since the start of first-line therapy, was in selected countries, optimal choice and sequencing is associated with lower probability of having second-line more and more challenging . The Bayesian fixed- effects network meta-analysis model comparing PFS and 1 3 91 Page 8 of 9 Medical Oncology (2018) 35:91 OS of cabozantinib versus everolimus, nivolumab, axi- Until today, the MSKCC score and first-line treatment tinib, sorafenib and best supportive care (BSC) showed type were considered as predictive factors of receiving sec- that cabozantinib was superior to all its comparators with ond-line therapy. No second-line treatment-oriented nomo- a higher probability of longer PFS and OS during 3 years, grams or prediction scales are available. We have evaluated but in the Gompertz model nivolumab was preferred after 17 clinical and biochemical parameters that are widely eval- 24 months . These trials are expected to determine the uated at RCC first line treatment initiation and defined these shift of everolimus to the third-line and subsequent lines that impact first-line treatment survival and, therefore, sec- of treatment if positive in future in selected countries with ond line treatment enrolment. The presence of brain metas- more robust resources allocated to healthcare system. Unfor- tases, number of metastatic sites abnormal platelet count tunately, it is very unlikely that prospective trials compar- and lactate dehydrogenase level are found in patents that are ing head to head the activity of axitinib, cabozantinib, len- at a risk of nor eligibility of second line treatment. Normal vatinib and nivolumab will be conducted. At this point of albumin concentration is associated with increased prob- time, clinicians still lack biomarkers and recommendation ability or sequential treatment. Based on identified factors, on the optimal sequence of treatment in individual cases. We multi-factorial model and on-line calculator was built for believe that selected clinical variables can help physicians treatment prediction. The MRCCSECLINE calculator may to make decisions in the future  and personalised deci- become a practical tool to identify mRCC patients, who are sions could be supported with calculator developed within unlikely to receive second-line treatment, and, therefore, to this project. help determine optimal first lime treatment to obtain best The limitations of the study include its retrospective response and treatment safety. design and lack of external validation in another cancer cen- Funding This study was founded by grant No 347/WIM - Military tre. Nevertheless, the proposed model was successfully vali- Institute of Medicine statutory founding (2015–2017). dated using two internal bootstrap procedures and has shown good statistical performance. Similar models, including Compliance with ethical standards nomograms and calculators are being developed in the field of medical oncology practice including advanced oesoph- Conflict of interest AMC, CS, LB and RS received lecture, travel and agogastric adenocarcinoma nomogram for patients under- accommodations expenses from Pfizer, and Novartis. going first-line combination chemotherapy , advanced Ethical approval The institutional ethics committee approved the study urothelial carcinoma patients to estimate the activity of sec- (Agreement No 48/WIM/2014). ond-line therapy  or advanced luminal subtype breast cancer patients to estimate PFS after first-line therapy . Open Access This article is distributed under the terms of the Crea- Medical calculators incorporating prognostic factors may tive Commons Attribution 4.0 International License (http://creat iveco facilitate the evaluation of outcomes across different groups mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- of patients before treatment enrolment. We believe that the tion, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the MRSCCSECLINE calculator should contribute to informed, Creative Commons license, and indicate if changes were made. evidence-based clinical decision making and optimise medi- cal practice as well as future trial recruitment and design. References Conclusions 1. Escudier B, Porta C, Schmidinger M, Rioux-Leclercq N, Bex A, Khoo V, et al. Renal cell carcinoma: ESMO Clinical Practice The MRCCSECLINE calculator developed in our study Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2016;27(suppl 5):v58-v68. may be a useful tool for clinicians to identify those mRCC 2. Czarnecka AM, Sobczuk P, Korniluk J, Spychalska M, Bogusz patients, who are unlikely to receive second-line treatment, K, Owczarek A, et al. Long-term response to sunitinib: everoli- and subsequently, to help determine the most optimal, long- mus treatment in metastatic clear cell renal cell carcinoma. Future term treatment plan at the beginning of systemic TT. How- Oncol. 2017;13(1):31–49. 3. Tannir NM, Figlin RA, Gore ME, Michaelson MD, Motzer ever, independent validation of the calculator in prospective RJ, Porta C, et al. Long-term response to sunitinib treatment trials and additional studies to identify other tumour-specific in metastatic renal cell carcinoma: a pooled analysis of clinical prognostic factors for all therapies are needed in the future. trials. Clin Genitourin Cancer. 2017. https ://doi.org/10.1016/j. Medical calculator potentially facilitates evidence-based clgc.2017.06.005. 4. Bielecka ZF, Czarnecka AM, Solarek W, Kornakiewicz A, Szc- treatment decisions, individualised risk assessment and zylik C. Mechanisms of acquired resistance to tyrosine kinase helps to select suitable agents or BSC for second-line treat- inhibitors in clear-cell renal cell carcinoma (ccRCC). Curr Signal ment of mRCC. Transduct Ther. 2014;8(3):218 – 28. 1 3 Medical Oncology (2018) 35:91 Page 9 of 9 91 5. Buczek M, Escudier B, Bartnik E, Szczylik C, Czarnecka AM. Carcinoma Sorafenib (EU-ARCCS) expanded-access study: a Resistance to tyrosine kinase inhibitors in clear cell renal cell large open-label study in diverse community settings. Ann Oncol. carcinoma: from the patient’s bed to molecular mechanisms. Bio- 2011;22:1812–23. chem Biophys Acta. 2014;1845(1):31–41. 20. Steyerberg EW, Lingsma HF. Predicting citations: Validating pre- 6. Eggers H, Ivanyi P, Hornig M, Grunwald V. Predictive factors for diction models. BMJ. 2008;336(7648):789. second-line therapy in metastatic renal cell carcinoma: a retro- 21. Steyerberg EW, Vergouwe Y. Towards better clinical prediction spective analysis. J Kidney Cancer VHL. 2017;4(1):8–15. models: seven steps for development and an ABCD for validation. 7. Zanwar S, Joshi A, Noronha V, Patil VM, Sable N, Popat P, et al. Eur Heart J. 2014;35(29):1925–31. Patterns of care and outcomes for second-line targeted therapy in 22. Hirsch BR, Burke JM, Agrawal M, Hauke RJ, Hutson TE, Doshi metastatic renal cell carcinomas: a registry based analysis. Indian G, et al. Sequential therapy in metastatic renal cell carcinoma. J J Cancer. 2016;53(4):579 – 82. Kidney Cancer VHL. 2016;3(1):23–35. 8. Lakomy R, Poprach A, Bortlicek Z, Melichar B, Chloupkova R, 23. Heng DY, Xie W, Regan MM, Harshman LC, Bjarnason GA, Vyzula R, et al. Utilization and efficacy of second-line targeted Vaishampayan UN, et al. External validation and comparison with therapy in metastatic renal cell carcinoma: data from a national other models of the International Metastatic Renal-Cell Carci- registry. BMC Cancer. 2017;17(1):880. noma Database Consortium prognostic model: a population-based 9. Beisland C, Johannesen TB, Klepp O, Axcrona U, Torgersen KM, study. Lancet Oncol. 2013;14(2):141–8. Kowalski J, et al. Overall survival in renal cell carcinoma after 24. Porta C, Gore ME, Rini BI, Escudier B, Hariharan S, Charles introduction of targeted therapies: a Norwegian population-based LP, et al. Long-term safety of sunitinib in metastatic renal cell study. OncoTargets Therapy. 2017;10:371 – 85. carcinoma. Eur Urol. 2016;69(2):345 – 51. 10. Levy A, Menard J, Albiges L, Loriot Y, Di Palma M, Fizazi K, 25. Bergmann L, Kube U, Doehn C, Steiner T, Goebell PJ, Kindler M, et al. Second line treatment of metastatic renal cell carcinoma: the et al. Everolimus in metastatic renal cell carcinoma after failure Institut Gustave Roussy experience with targeted therapies in 251 of initial anti-VEGF therapy: final results of a noninterventional consecutive patients. Eur J Cancer. 2013;49(8):1898 – 904. study. BMC Cancer. 2015;15(1):303. 11. De Groot S, Blommestein HM, Redekop WK, Sleijfer S, Kieme- 26. Knox JJ, Barrios CH, Kim TM, Cosgriff T, Srimuninnimit V, ney L, Oosterwijk E, et al. Potential health gains for patients with Pittman K, et al. Final overall survival analysis for the phase II metastatic renal cell carcinoma in daily clinical practice: a real- RECORD-3 study of first-line everolimus followed by sunitinib world cost-effectiveness analysis of sequential first- and second- versus first-line sunitinib followed by everolimus in metastatic line treatments. PLoS ONE. 2017;12(5):e0177364. RCC. Ann Oncol. 2017;28(6):1339–45. 12. Arranz Arija J, Pérez-Valderrama B, Rodriguez Sanchez A, 27. Negrier S, Bushmakin AG, Cappelleri JC, Korytowsky B, Sandin Puertas Alvarez J, Pinto Marin A, Maximiano Alonso C, et al. R, Charbonneau C, et al. Assessment of progression-free survival SPAZO2 (SOGUG): comparative effectiveness of everolimus as a surrogate end-point for overall survival in patients with meta- vs axitinib as second-line after first-line pazopanib in metastatic static renal cell carcinoma. Eur J Cancer. 2014;50(10):1766–71. renal carcinoma. Ann Oncol. 2017;28(suppl_5):v295-v329. 28. Jain RK, Gandhi S, George S. Second-line systemic therapy 13. Buti S, Leonetti A, Dallatomasina A, Bersanelli M. Everolimus in in metastatic renal-cell carcinoma: a review. Urol Oncol. the management of metastatic renal cell carcinoma: an evidence- 2017;35(11):640–6. based review of its place in therapy. Core Evid. 2016;11:23–36. 29. Amzal B, Fu S, Meng J, Lister J, Karcher H. Cabozantinib versus 14. Motzer RJ, Hutson TE, Tomczak P, Michaelson MD, Bukowski everolimus, nivolumab, axitinib, sorafenib and best supportive RM, Rixe O, et al. Sunitinib versus interferon alfa in metastatic care: a network meta-analysis of progression-free survival and renal-cell carcinoma. N Engl J Med. 2007;356(2):115 – 24. overall survival in second line treatment of advanced renal cell 15. Sternberg CN, Hawkins RE, Wagstaff J, Salman P, Mardiak J, carcinoma. PLoS ONE. 2017;12(9):e0184423. Barrios CH, et al. A randomised, double-blind phase III study of 30. Grande E, Martinez-Saez O, Gajate-Borau P, Alonso-Gordoa T. pazopanib in patients with advanced and/or metastatic renal cell Translating new data to the daily practice in second line treatment carcinoma: final overall survival results and safety update. Eur J of renal cell carcinoma: the role of tumor growth rate. World J Cancer. 2013;49(6):1287–96. Clin Oncol. 2017;8(2):100–5. 16. Czarnecka A, Sobczuk P, Bogusz K, Spychalska M, Szczylik C. 31. Custodio A, Carmona-Bayonas A, Jimenez-Fonseca P, Sanchez Survival, safety and treatment response duration in “real world” ML, Viudez A, Hernandez R, et al. Nomogram-based predic- patients with metastatic clear cell renal cancer—an update from tion of survival in patients with advanced oesophagogastric clinical practice. BJU Int. 2013;112(s3):1–17. adenocarcinoma receiving first-line chemotherapy: a multi- 17. Escudier B, Eisen T, Stadler WM, Szczylik C, Oudard S, Stae- center prospective study in the era of trastuzumab. Br J Cancer. hler M, et al. Sorafenib for treatment of renal cell carcinoma: 2017;116(12):1526–35. final efficacy and safety results of the phase III treatment 32. Pond GR, Agarwal N, Bellmunt J, Choueiri TK, Qu A, Fougeray approaches in renal cancer global evaluation trial. J Clin Oncol. R, et al. A nomogram including baseline prognostic factors to esti- 2009;27(20):3312–8. mate the activity of second-line therapy for advanced urothelial 18. Rini BI, Escudier B, Tomczak P, Kaprin A, Szczylik C, Hutson carcinoma. BJU Int. 2014;113(5b):E137–43. TE, et al. Comparative effectiveness of axitinib versus sorafenib 33. Carbognin L, Sperduti I, Ciccarese M, Fabi A, Petrucelli L, Vari S, in advanced renal cell carcinoma (AXIS): a randomised phase 3 et al. Prognostic model for advanced breast carcinoma with lumi- trial. Lancet. 2012;378(9807):1931–9. nal subtype and impact of hormonal maintenance: Implications for 19. Beck J, Procopio G, Bajetta E, Keilholz U, Negrier S, Szczy- post-progression and conditional survival. Breast. 2016;29:24–30. lik C, et al. Final results of the European Advanced Renal Cell 1 3
Medical Oncology – Springer Journals
Published: May 8, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera