Abstract
Aim: To gain greater insight into the biological mechanisms occurring shortly after discontinuation of VEGFR TKIs treatment because of progressive disease (PD). Patients and Methods: Sixteen patients with PD during treatment with sorafenib or sunitinib were randomized to either directly stop the VEGFR TKI or to continue for another two weeks. At baseline (i.e. at the moment of PD) and after two weeks FDG-PET/CT, functional-MRI and blood biomarkers of disease were evaluated. Results: A statistically significant difference in median change from baseline to two weeks later in Ktrans and LDH levels was observed between patients who directly stopped versus those who continued treatment (1.6 s−1 versus −1.1s−1, p=0.03; −73.0 U/L versus 52.0 U/L, p=0.008; respectively). There were no further differences between groups. Conclusion: Two weeks after discontinuation of VEGFR TKIs in mRCC because of PD, a rise in Ktrans accompanied by a decrease in LDH indicates an increase in tumor vascularization. This implies that at the moment of PD the effect of VEGFR TKIs is not completely exhausted.
- Vascular endothelial growth factor receptor inhibitor
- progressive disease
- renal cell carcinoma
- treatment beyond progression
- re-vascularization
- tyrosine kinase inhibitor
In clear cell renal cell carcinoma (RCC) a loss of function of the Von Hippel-Lindau (VHL) gene product leads to accumulation of the hypoxia inducible factor 1α (HIF-1α), with subsequent up-regulation of HIF target genes, including vascular endothelial growth factor (VEGF) (1-3). These biological insights provided the rationale to target the VEGF pathway to treat metastatic RCC (mRCC). A number of VEGF receptor tyrosine kinase inhibitors (VEGFR TKIs) has been approved for the treatment of mRCC, such as sunitinib, sorafenib, and more recently pazopanib and axitinib.
Treatment with VEGFR TKIs is continued until progressive disease (PD) or intolerable toxicities occur. After discontinuation because of PD, a range of clinical sequelae is seen. Most patients will have an indolent course with slowly-progressing tumors, but in approximately 10% of all RCC patients accelerated tumor growth is described (4, 5). The biological mechanisms of this so-called flare-up syndrome are poorly understood.
It has been hypothesized that during therapy with VEGFR TKIs, due to VEGF blockage and tumor hypoxia, up-regulation of pro-angiogenic genes and proteins (such as VEGF, PDGF, fibroblast growth factor, angiopoietin and interleukin-8) may occur. When treatment is stopped rapid re-vascularization and cell proliferation can occur due to the abundance of angiogenic factors (6). Apart from these mechanisms, it is hypothesized that a more malignant tumor phenotype may develop during treatment due to selection of hypoxia-tolerant tumor cells and metastatic conditioning (7, 8).
Flare-up syndromes have a negative impact on the quality of life and can preclude the start of a next-line therapy due to rapid deterioration of the clinical condition. Insights in the biological mechanisms occurring after discontinuation of VEGFR TKIs due to progressive disease are necessary to determine the optimal management strategy for these patients.
To gain greater insight, we performed a randomized study in patients with mRCC at the moment of progression during treatment with VEGFR TKIs. We performed functional MR imaging, 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) and obtained biological markers of disease progression.
Patients and Methods
Study design. Patients with mRCC with PD, as assessed by RECIST during either sorafenib or sunitinib treatment, were eligible for this single centre randomized study. The clinical condition of the patients should permit continuation of the same VEGFR TKI treatment for two more weeks. Baseline measurements (i.e., at the moment of PD) were performed during active treatment and consisted of physical examination, assessment of adverse events according to the common terminology criteria for adverse events (CTCAE) version 3.0 (9) and measurement of biomarkers of disease including plasma VEGF, C-reactive protein (CRP), D-dimer, lactate dehydrogenase (LDH) and regulatory T-cells (Tregs). In all patients, FDG-PET/CT to assess tumor metabolism was mandatory. Functional MR imaging (fMRI) to assess changes in tumor perfusion was optional in patients with abdominal RCC lesions with a diameter of at least two centimeters and without contra-indications for MRI. After baseline measurements, patients were randomized in a 1:1 ratio to continue the current VEGFR TKI for two more weeks or discontinue immediately. After two weeks, baseline measurements were repeated, including a contrast-enhanced CT scan to evaluate changes in tumor size according to RECIST version 1.1 (10). The study was approved by the medical ethics committee. All patients provided their written informed consent before any study procedure.
FDG-PET/CT. Patient preparation and PET/CT acquisition and processing parameters were in strict accordance to the Dutch (NEDPAS) and European Association of Nuclear Medicine (EANM) standardization guidelines (11, 12). PET scans were acquired on a hybrid PET/CT scanner (Biograph Duo, Siemens Medical Solutions Inc., Malvern, Pa., USA). Up to five target lesions per patient were identified and for quantitative assessment of these lesions, regions of interest (ROI) around the whole tumor were obtained semi-automatically. FDG accumulation for each voxel within the ROI was registered in MBq/ml. This was corrected for the injected dose of FDG, patients' body weight and time elapsed between injection and scanning to obtain the standardized uptake value (SUV). Volumes of interest (VOIs) were delineated using 50% and 70% isocontour thresholds based on a fixed percentage of the maximum activity within the lesion. As such, the maximum SUV, SUV50 and SUV70 within the VOI could be determined. The total lesion glycolysis (TLG) per patient was calculated by summing the products of the mean SUV within the VOI and the volume of that VOI. The fractional change (Δ) in the mean SUV per patient and TLG between FDG-PET/CT at baseline and after two weeks was calculated and expressed in percentage.
Functional MR imaging. After conventional T1- and T2-weighted imaging diffusion weighted MRI (DWI), dynamic contrast-enhanced MRI (DCE-MRI) and T2* perfusion MRI were acquired. Sequence parameters are listed in Table I. Breath-hold DWI was obtained using three gradient factors (b=50, 300 and 600 s/mm2) after maximal expiration. After administration of 15 ml 0.5 M Gadolinium (Gd)-DTPA the DCE-MRI sequence was performed with a temporal resolution of 2 seconds during 5 min. A second bolus of 15 ml 0.5 M Gd-DTPA was administered iv followed by dynamic T2*-weighted echo-planar measurements during 90 sec in order to measure perfusion (13). For quantification, one tumor in the abdomen was selected and ROIs were drawn around the whole tumor. Tumor apparent diffusion coefficient (ADC) (s/mm2) maps and Tofts parameters were calculated. (Mevislab, Siemens Syngo VB15). The Tofts parameters include Ktrans, a volume transfer constant (s−1), ne which is the volume of the extravascular extracellular space (EES) per unit volume of tissue and kep the flux rate constant between EES and plasma (s-1) (kep=Ktrans /ne). Higher values of kep or Ktrans can indicate higher perfusion, higher permeability and/or a larger blood vessel surface area, globally denoting an increased tumor vascularization (14). Using T2* images, ROIs were drawn around the tumor and the aorta in order to calculate tumor blood volume (tBV)(AU) and tumor blood flow (tBF) (AU/s) (Neuro Perfusion Evaluation, Siemens Syngo VB17) (15).
Statistics. To analyze the change from baseline to two weeks later in CT, FDG-PET/CT and MRI parameters and in blood biomarkers, the Wilcoxon-signed rank test was performed. Differences in change of the parameters from baseline to two weeks between both groups were assessed using the Mann-Whitney test. The associations between parameter changes were tested using the Spearman's correlation test. All tests performed were two-sided and a p<0.05 was considered significant.
Results
Patients. Sixteen patients with mRCC were included in the study, performed in our University Medical Center between October 2008 and January 2012. Baseline characteristics are provided in Table II. Nine patients were randomized for direct discontinuation of VEGFR TKI, seven patients continued their VEGFR TKI beyond progression for another two weeks. The median age of all patients was 61.5 years (range=48-79 years). Before PD occurred patients were treated for a median of 86 weeks (range=10-219). Of the 10 patients using sutent, only 4 patients used sutent intermittently. In the context of this study patients re-started their sutent after the CT showed progression, so that at the moment of the first FDG-PET and MRI scan, these 4 patients used sutent for a minimum of one week. There were no significant differences in baseline characteristics between the groups. Two patients, both with pulmonary and pleural metastases, had a flare-up syndrome, consisting of development of dyspnea, cough and pleural effusion starting 5-10 days after discontinuation of sorafenib (Figure 1). Both patients had an increase in tumor size on CT and one of the patients also developed new subcutaneous metastases. FDG-uptake, MRI parameters and biomarkers of these two patients were not different from the parameters observed in the other seven patients who directly discontinued treatment.
Imaging results. The imaging results are summarized in Table III. Baseline parameter values were not significantly different between patients who directly stopped and patients who continued therapy. All patients underwent CT and FDG-PET scanning. All but one patient had FDG avid lesions. A total of 56 lesions was evaluated with FDG-PET, of which 10 were located in bone, 11 in the lungs, 13 in lymph nodes and 8 in the liver. There were no significant differences in the change from baseline to two weeks later in values between both groups, as assessed by RECIST and FDG-PET, indicating that discontinuation of VEGFR TKI did not significantly affect tumor growth or tumor metabolism in the study period. Out of the 16 patients, 11 underwent two consecutive MRIs as per protocol. Two patients had contra-indications for MRI (claustrophobia, pacemaker), two patients did not have an abdominal mass larger than 2 cm and one patient experienced tinnitus after the first MRI and refused to undergo a second MRI. Of the 11 patients who underwent MRI, six were randomized to stop directly and five to continue therapy for two weeks. Patients who stopped therapy directly had a median increase from baseline to two weeks later in Ktrans of 1.6 s−1 ±1.9, whereas the patients who continued had a median decrease of -1.1s−1 ±15.2 (p=0.03). The median change in kep in the patients who stopped versus those who continued was 3.0 s−1 ±5.1 and -3.3 s−1 ±20.9 respectively (p=0.08). There were no significant differences in changes in ADC, tBV and tBF between the groups.
Pharmacodynamic results. Patients who stopped therapy had a median decrease in LDH (−73.0 U/L; range −372 to 66), whereas patients who continued therapy had a median increase in LDH (52.0 U/L; range -13 to 86) (p=0.008). Interestingly, seven out of nine patients who stopped therapy, showed a decrease in LDH, whereas in six out of seven patients who continued therapy the LDH increased. Hemoglobin levels decreased in both groups, but mainly in the group who stopped (−0.9 mmol/L; range −1.5 to −0.1 versus -0.1 mmol/L; range −0.6 to 0.5, p=0.003). The change in VEGF was −2.35 ng/ml (−0.66 to 0.23) in the group who stopped, versus 0.11 ng/ml (−0.38 to 0.88) in the group who continued. This difference just failed to reach statistical significance (p=0.065). Moreover, no significant differences were observed for changes in Tregs, or the acute-phase proteins CRP and D-dimer between both groups.
Discussion
In this prospective study, in which VEGFR-TKI-treated mRCC patients were randomized at the moment of PD to stop therapy immediately or to continue for two more weeks, we aimed to gain greater insight regarding the effects of discontinuing VEGFR TKI. Within two weeks after discontinuation, a rise in Ktrans and a decrease in LDH were observed, indicating an increase in tumor vascularization, contrasting to the decline in Ktrans and increase in LDH found in patients continuing therapy beyond progression. This decrease in LDH is in contrast to the general idea under which a rising LDH is used as a (albeit non-specific) marker of progression. However, the production of LDH is stimulated by HIF-1α, that is up-regulated in hypoxic conditions. In our study, the decrease in LDH in patients who stopped taking the VEGFR TKI is accompanied by an increase in Ktrans and kep. This points to an increased perfusion, a higher permeability and a larger blood vessel surface area, denoting an increased tumor vascularization. Our results are in line with the observation that after initiating treatment with sunitinib a temporally increase in LDH is observed, which could be explained by the tissue hypoxia induced by the inhibition of VEGFR due to sunitinib (16). The change in VEGF is also consistent with other studies, that have shown a rise in VEGF during treatment with sunitinib and a decrease after discontinuation (17, 18).
Currently, PD is determined by radiological assessment according to the Response Evaluation Criteria In Solid Tumors (RECIST) (10). However, the adequacy of these RECIST guidelines to evaluate treatment response of VEGFR TKIs has been questioned, since the anti-angiogenic effect leads to necrosis and cavitation with, frequently, only minimal decrease in tumor size. Using the size-based RECIST criteria often underestimates the effect of VEGFR TKIs (6, 19) and therefore a patient can be labeled as PD too early.
After discontinuation of VEGFR TKI, CT and FDG-PET/CT suggest no clinical significant difference in tumor growth and metabolic activity, respectively. Of course, the period of two weeks is short and the group of patients is small. Nagengast et al. studied mouse xenograft models of tumors after discontinuing treatment with sunitinib for one week and found intra-tumoral differences, rapid tumor revascularization and re-growth of tumors with VEGF-PET, but not with FDG-PET (20). Furthermore, the similar tumor metabolism on FDG-PET in our study may be explained by the fact that the larger supply of FDG to the tumor due to the increasing tumor vascularization is balanced by a decrease in glucose demand due to less hypoxic conditions in the group of patients that discontinued VEGFR TKI.
The mechanisms leading to drug resistance and progressive disease are not fully-elucidated. Inadequate target inhibition either due to enhanced receptor signaling or decreased plasma levels of VEGFR TKIs might play a role in resistance development. The latter may occur in case of an altered pharmacodynamic effect of a TKI due to polymorphisms in specific genes encoding for metabolizing enzymes, efflux transporters and drug targets, as observed for sunitinib (21). It is also hypothesized that, due to VEGF blockage and tumor hypoxia, up-regulation of pro-angiogenic factors might take place when developing resistance to treatment with VEGFR TKIs. When treatment is stopped, rapid re-vascularization and cell proliferation occur due to the abundance of angiogenic factors (6). Griffioen et al. studied tumor nephrectomy tissues of patients pre-treated with sunitinib and observed an increase in the number of proliferating endothelial that was positively correlated with the time interval between treatment stop and cytoreductive surgery (22). In an animal study with two VEGFR inhibitors, complete tumor re-vascularization was seen at day 7 after cessation of therapy. Endothelial sprouting was already observed after one day (23). Another animal study showed an increase in tumor volume 9 days after discontinuation of sunitinib (24). The observed increase in Ktrans and kep in our patients after discontinuation are consistent with these observations. In our study we showed that although all patients had PD, only patients who discontinued had an increased perfusion and permeability, implying that at the moment of PD the effect of the VEGFR TKI was not completely exhausted. However, there was no difference between patients with and without a flare-up syndrome, but the number of patients examined is too small to detect such differences.
Apart from these mechanisms, the development of a more malignant tumor phenotype during treatment may occur, with accelerated infiltrative growth and dissemination due to selection of hypoxia-tolerant tumor cells and metastatic conditioning. Ebos et al. reported that mice receiving sunitinib for only seven days either before or after injection of tumor cells suffered from an accelerated tumor growth after discontinuation of sunitinib, resulting in a shorter survival (7). The authors suggested that short anti-angiogenic treatment resulted in ‘metastatic conditioning', which might be the effect of up-regulation of multiple circulating pro-angiogenic cytokines and growth factors, the mobilization of bone marrow-derived cells facilitating an enhanced ‘pre-metastatic niche’ including circulating endothelial and myeloid progenitors, or the target promiscuity, all together creating a more favorable metastatic environment (8).
Flare-up syndromes have a negative impact on the quality of life and can preclude the start of a next-line therapy due to rapid deterioration of the clinical condition. Insights in the biological mechanisms occurring after discontinuation of VEGFR TKIs due to PD are necessary to determine the optimal management strategy for these patients. The results of our study imply that at the moment of PD the effect of the VEGFR TKI is not completely exhausted. Therefore, a dose-escalation of the VEGFR TKI at the moment of PD, instead of discontinuation, might lead to clinical benefit as shown in a study with sorafenib (25). Alternatively, introducing a VEGF-antibody or switch to another VEGFR TKI might prevent rapid re-vascularization.
Within two weeks after discontinuation of VEGFR TKI in patients with PD, the rise in Ktrans and the decrease in LDH indicate a higher tumor perfusion or increased permeability of tumor blood vessels, reflecting an increase in tumor vascularization. Although all patients had PD, only those who stopped showed increased perfusion and permeability, implying that at the moment of PD the vascular effects of the VEGFR TKI are not completely exhausted.
- Received June 20, 2015.
- Revision received July 15, 2015.
- Accepted July 22, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved