Considerations for study design in the DAHANCA 35 trial of protons versus photons for head and neck cancer.

Proton radiotherapy offers a dosimetric advantage compared to photon therapy in sparing normal tissue, but the clinical evidence for toxicity reductions in the treatment of head and neck cancer is limited. The Danish Head and Neck Cancer Group (DAHANCA) has initiated the DAHANCA 35 randomised trial to clarify the value of proton therapy (NCT04607694). The DAHANCA 35 trial is performed in an enriched population of patients selected by an anticipated benefit of proton therapy to reduce the risk of late dysphagia or xerostomia based on normal tissue complication probability (NTCP) modelling. We present our considerations on the trial design and a test of the selection procedure conducted before initiating the randomised study.

NTCP model validation method for DAHANCA patient selection of protons versus photons in head and neck cancer radiotherapy.

Introduction: Prediction models using logistic regression may perform poorly in external patient cohorts. However, there is a need to standardize and validate models for clinical use. The purpose of this project was to describe a method for validation of external NTCP models used for patient selection in the randomized trial of protons versus photons in head and neck cancer radiotherapy, DAHANCA 35. Material and methods: Organs at risk of 588 patients treated primarily with IMRT in the randomized controlled DAHANCA19 trial were retrospectively contoured according to recent international recommendations. Dose metrics were extracted using MatLab and all clinical parameters were retrieved from the DAHANCA database. The model proposed by Christianen et al. to predict physician-rated dysphagia was validated through the closed testing, where change of the model intercept, slope and individual beta's were tested for significant prediction improvements. Results: Six months prevalence of dysphagia in the validation cohort was 33%. The closed testing procedure for physician-rated dysphagia showed that the Christianen et al. model needed an intercept refitting for the best match for the Danish patients. The intercept update increased the risk of dysphagia for the validation cohort by 7.9 ± 2.5% point. For the raw model performance, the Brier score (mean squared residual) was 0.467, which improved significantly with a new intercept to 0.415. Conclusions: The previously published Dutch dysphagia model needed an intercept update to match the Danish patient cohort. The implementation of a closed testing procedure on the current validation cohort allows quick and efficient validation of external NTCP models for patient selection in the future.

Dosimetric and geometric evaluation of the use of deformable image registration in adaptive intensity-modulated radiotherapy for head-and-neck cancer.

The aim of this study was to carry out geometric and dosimetric evaluation of the usefulness of a deformable image registration algorithm utilized for adaptive head-and-neck intensity-modulated radiotherapy. Data consisted of seven patients, each with a planning CT (pCT), a rescanning CT (ReCT) and a cone beam CT (CBCT). The CBCT was acquired on the same day (± 1 d) as the ReCT (i.e. at Fraction 17, 18, 23, 24 or 29). The ReCT served as ground truth. A deformed CT (dCT) with structures was created by deforming the pCT to the CBCT. The geometrical comparison was based on the volumes of the deformed, and the manually delineated structures on the ReCT. Likewise, the center of mass shift (CMS) and the Dice similarity coefficient were determined. The dosimetric comparison was performed by recalculating the initial treatment plan on the dCT and the ReCT. Dose-volume histogram (DVH) points and a range of conformity measures were used for the evaluation. We found a significant difference in the median volume of the dCT relative to that of the ReCT. Median CMS values were ∼ 2-5 mm, except for the spinal cord, where the median CMS was 8 mm. Dosimetric evaluation of target structures revealed small differences, while larger differences were observed for organs at risk. The deformed structures cannot fully replace manually delineated structures. Based on both geometrical and dosimetrical measures, there is a tendency for the dCT to overestimate the need for replanning, compared with the ReCT.

Expression of P-glycoprotein and multidrug resistance associated protein in Ehrlich ascites tumor cells after fractionated irradiation.

PURPOSE: To characterize irradiated murine tumor cells with respect to drug resistance, drug kinetics, and ATPase activity, and to evaluate the possible role of P-glycoprotein (PGP) and murine multidrug resistance associated protein (Mrp1) in the drug-resistant phenotype of these cells. METHODS AND MATERIALS: Sensitive Ehrlich ascites tumor cells (EHR2) were in vitro exposed to fractionated irradiation (60 Gy). Western blot analysis was performed for determination of PGP and Mrp1, reverse transcriptase-polymerase chain reaction (RT-PCR) for determination of mdr1a + b mRNA, and semiquantitative RT-PCR for Mrp1 mRNA. The clonogenic assay was applied to investigate sensitivity, whereas the steady-state drug accumulation of daunorubicin (DNR), 3H-vincristine (VCR), and 3H-etoposide (VP16) was measured by spectrofluorometry and scintillation counting, respectively. For determining of ATPase activity, the release of inorganic phosphate from ATP was quantified using a colorimetric method. RESULTS: Compared with EHR2, the irradiated cell line EHR2/irr showed increased expression of PGP (threefold), Mrp1 (eightfold), and Mrp1 mRNA (sixfold), and a slight reduction of mdr1b mRNA, whereas mdr1a was present in EHR2 but could not be detected in EHR2/irr. EHR2/irr developed sixfold resistance to VP16, twofold resistance to vincristine, but remained sensitive to DNR. Addition of the PGP inhibitor, verapamil (VER) or depletion of glutathione by buthionine sulfoximine (BSO) partly reversed the resistance in EHR2/irr. In EHR2/irr, the steady-state accumulation of 3H-VCR and 3H-VP16 was significantly decreased as compared with EHR2, whereas the accumulation of DNR was unchanged. The ATPase activity of plasma membrane vesicles prepared from EHR2/irr cells was similar to that of wild-type EHR2 cells. The ATPase activity was neither stimulated by vinblastine nor VER. CONCLUSION: Irradiation induced a multidrug-resistant phenotype in sensitive tumor cells. This phenotype was characterized by increased expression of Mrp1 mRNA, Mrp1, and PGP but decreased expression of mdr1a + b mRNA. The influence of irradiation on PGP and Mrp1 expression seemed to be different.

Characterisation of multidrug-resistant Ehrlich ascites tumour cells selected in vivo for resistance to etoposide.

An Ehrlich ascites tumour cell line (EHR2) was selected for resistance to etoposide (VP16) by in vivo exposure to this agent. The resulting cell line (EHR2/VP16) was 114.3-, 5.7-, and 4.0-fold resistant to VP16, daunorubicin, and vincristine, respectively. The amount of salt-extractable immunoreactive topoisomerase IIalpha and beta in EHR2/VP16 was reduced by 30-40% relative to that in EHR2. The multidrug resistance-associated protein (MRP) mRNA was increased 20-fold in EHR2/VP16 as compared with EHR2, whereas the expression of P-glycoprotein was unchanged. In EHR2/VP16, the steady-state accumulation of [(3)H]VP16 and daunorubicin was reduced by 64% and 17%, respectively, as compared with EHR2. Deprivation of energy by addition of sodium azide increased the accumulation of both drugs to the level of sensitive cells. When glycolysis was restored by the addition of glucose to EHR2/VP16 cells loaded with drug in the presence of sodium azide, extrusion of [(3)H]VP16 and daunorubicin was induced. Addition of verapamil (25 microM) decreased the efflux of daunorubicin to the level of sensitive cells, but had only a moderate effect on the efflux of [(3)H]VP16. The resistant cells showed moderate sensitisation to VP16 on treatment with verapamil, whereas cyclosporin A had no effect. Compared with that of sensitive cells, the ATPase activity of plasma membrane vesicles prepared from EHR2/VP16 cells was very low. Vanadate inhibited the ATPase activity of EHR2/VP16 microsomes with a K(i) value of 30 microM. ATPase activity was slightly stimulated by daunorubicin, whereas vinblastine, verapamil, and cyclosporin A had no effect. In conclusion, development of resistance to VP16 in EHR2 is accompanied by a significant reduction in topoisomerase II (alpha and beta) and by increased expression of MRP mRNA (20-fold). MRP displays several points of resemblance to P-glycoprotein in its mode of action: 1) like P-glycoprotein, MRP causes resistance to a range of hydrophobic drugs; 2) MRP decreases drug accumulation in the cells and this decrease is abolished by omission of energy; and 3) MRP increases efflux of drug from cells. However, compared with that of P-glycoprotein-positive cells, the ATPase activity of MRP-positive cells is found to be low and not able to be stimulated by verapamil.

Characterisation of non-P-glycoprotein multidrug-resistant Ehrlich ascites tumour cells selected for resistance to mitoxantrone.

An Ehrlich ascites tumour cell line (EHR2) was selected in vivo for resistance to mitoxantrone (MITOX). The resistant cell line (EHR2/MITOX) was 6123-, 33-, and 30-fold-resistant to mitoxantrone, daunorubicin, and etoposide, respectively, but retained sensitivity to vincristine. The resistant cells showed moderate sensitisation to mitoxantrone on treatment with verapamil or cyclosporin A. Compared with EHR2, the multidrug resistance-associated protein mRNA was increased 13-fold in EHR2/MITOX. Western blot analysis showed an unchanged, weak expression of P-glycoprotein. Topoisomerase IIalpha was reduced to one-third in EHR2/MITOX relative to EHR2 cells, whereas topoisomerase IIbeta was present in EHR2 but could not be detected in EHR2/MITOX. In the resistant subline, net accumulation of MITOX (120 min) and daunorubicin (60 min) was reduced by 43% and 27%, respectively, as compared with EHR2. The efflux of daunorubicin from preloaded EHR2/MITOX cells was significantly increased. EHR2/MITOX microsomes had a significant basal unstimulated ATPase activity. The apparent K(i) value for vanadate inhibition of the ATPase activity in EHR2/MITOX microsomes was not significantly different from the K(i) value for P-glycoprotein-positive cells. However, whereas verapamil (50 microM) inhibited the ATPase activity of EHR2/MITOX microsomes, it stimulated the ATPase activity of microsomes derived from P-glycoprotein-positive cells. In conclusion, the resistance in EHR2/MITOX was multifactorial and appeared to be associated with: 1) a quantitative reduction in topoisomerase IIalpha and beta protein; 2) reduced drug accumulation, probably as a result of increased expression of a novel transport protein with ATPase activity; and 3) increased expression of MRP mRNA.

P-glycoprotein expression in Ehrlich ascites tumour cells after in vitro and in vivo selection with daunorubicin.

Fluctuation analysis experiments were performed to assess whether selection or induction determines expression of P-glycoprotein and resistance in the murine Ehrlich ascites tumour cell line (EHR2) after exposure to daunorubicin. Thirteen expanded populations of EHR2 cells were exposed to daunorubicin 7.5 x 10(-9) M or 10(-8) M for 2 weeks. Surviving clones were scored and propagated. Only clones exposed to daunorubicin 7.5 x 10(-9) M could be expanded for investigation. Drug resistance was assessed by the tetrazolium dye (MTT) cytotoxicity assay. Western blot was used for determination of P-glycoprotein. Compared with EHR2, the variant cells were 2.5- to 5.2-fold resistant to daunorubicin (mean 3.6-fold). P-glycoprotein was significantly increased in 11 of 25 clones (44%). Analysis of variance supported the hypothesis that spontaneous mutations conferred drug resistance in EHR2 cells exposed to daunorubicin 7.5 x 10(-9) M. At this level (5 log cell killing) of drug exposure, the mutation rate was estimated at 4.1 x 10(-6) per cell generation. In contrast, induction seemed to determine resistance in EHR2 cells in vitro exposed to daunorubicin 10(-8) M. The revertant EHR2/0.8/R was treated in vivo with daunorubicin for 24 h. After treatment, P-glycoprotein increased in EHR2/0.8/R (sevenfold) and the cell line developed resistance to daunorubicin (12-fold), suggesting that in EHR2/0.8/R the mdr1 gene was activated by induction. In conclusion, our study demonstrates that P-glycoprotein expression and daunorubicin resistance are primarily acquired by selection of spontaneously arising mutants. However, under certain conditions the mdr1 gene may be activated by induction.

Cellular resistance to anthracyclines.

The antracyclines induce multiple intracellular effects; however, inhibition of the nuclear enzyme topoisomerase II (TOPO II) is the main mechanism of action. Resistance to anthracyclines in tumor cells is multifactorial. The main mechanisms are: (1) the classic multidrug resistance (MDR) phenotype, which is due to the presence of P-glycoprotein (PGP) in plasma membrane, that is, a "pump" that can extrude a wide range of anticancer drugs. Membrane-active drugs (e.g., verapamil) have been found in vitro to reverse this phenotype. Most clinical studies including chemosensitizers have, however, been disappointing. (2) Non-PGP-mediated MDR: this phenotype is characterized by expression of other proteins in the plasma membrane which are also able to extrude anticancer drugs. (3) Changes in the intracellular distribution of drug: this mechanism has been demonstrated in several cell lines, most often in combination with PGP or non-PGP-mediated resistance. (4) Glutathione transferases (GST) and detoxification mechanisms: these represent a multigene family of enzymes that conjugate glutathione to chemically reactive groups. Direct evidence for a causative role of GST in anthracycline resistance is missing. (5) Alterations in TOPO II (at-MDR): DNA topoisomerases are involved in several aspects of DNA metabolism, in particular genetic recombination, DNA transcription, and chromosome segregation. Low levels of expression or alterations in TOPO II are associated in vitro with resistance. (6) Increased DNA repair: in several cell lines, an increase in the efficacy of DNA repair has been associated with resistance to doxorubicin (DOX). So far, only classic MDR has been shown to contribute to resistance in clinical conditions, whereas evidence for the other mechanisms of resistance is still missing.

Influx of daunorubicin in multidrug resistant Ehrlich ascites tumour cells: correlation to expression of P-glycoprotein and efflux. Influence of verapamil.

Classic multidrug resistance is characterized by a decrease in the intracellular concentration of drugs in resistant cells as compared to sensitive cells. This is correlated with the presence of P-glycoprotein in the membrane. P-glycoprotein is responsible for an active efflux of drug. In this study we investigated the correlation between P-glycoprotein and influx of daunorubicin. Four Ehrlich ascites tumour cell lines selected in vivo for resistance to daunorubicin were investigated. The sublines EHR2/0.1, EHR2/0.2, passage no. 12 of EHR2/0.8, EHR2/0.4, and passage no. 72 of EHR2/0.8 were 6-, 6-, 5-, 33-, and 35-fold resistant to daunorubicin, respectively. All sublines overexpressed P-glycoprotein as determined with Western blot. Influx was measured over 40 sec. In glucose-enriched medium influx was significantly decreased in all but one of the resistant sublines. A correlation between P-glycoprotein, degrees of resistance, and influx was demonstrated in four sublines. Comparing influx experiments with efflux experiments (Nielsen et al., Biochem Pharmacol 1994, 47, 2125-2135) we found a linear relationship between influx and efflux in the resistant sublines (r = 0.97). Verapamil (5.5 microM, 11.0 microM) increased influx significantly in all resistant sublines, whereas the drug had no effect on sensitive cells. Verapamil (3.3 microM) increased influx in the EHR2/0.8 (passage no. 72) subline to the level of sensitive cells. Comparing this result with efflux experiments, verapamil was found to increase influx preferentially. Depletion of energy (medium without glucose including Na(+)-azide) increased influx in all resistant sublines. In EHR2/0.4 and EHR2/0.8 (passage no. 72) the influx, however, was still significantly decreased after depletion of energy. In these cells further addition of verapamil increased influx to the level of EHR2. These data were consistent with the hypothesis that P-glycoprotein effluxes drug directly from the plasma membrane.

Kinetics of daunorubicin transport in Ehrlich ascites tumor cells with different expression of P-glycoprotein.

The classical multidrug resistance (MDR) phenotype is characterized by a decrease in the intracellular drug concentration in resistant cells as compared to sensitive cells. P-glycoprotein (P-gp) is thought to be responsible for an active efflux of lipophilic drugs. Four Ehrlich ascites tumor cell lines selected in vivo for resistance to daunorubicin (DNR) and their sensitive counterpart were investigated. The resistant sublines EHR2/0.1, EHR2/0.2, EHR2/0.4, and EHR2/0.8 were developed by treatment of tumor bearing mice with DNR 0.1, 0.2, 0.4, and 0.8 mg/kg x 4 weekly, respectively. One passage from EHR2/0.1, EHR2/0.2, and EHR2/0.4 and two passages from EHR2/0.8 were investigated. Western blot analysis showed significantly different amounts of P-gp (a 6-fold variation). Efflux of DNR in a drug free medium was investigated with and without presence of verapamil (VER). Efflux from sensitive and resistant cells was described by mono- and bi-exponential kinetics, respectively. In all cases but one, a correlation between resistance, expression of P-gp, P-gp mediated efflux capacity and effect of VER was established. In passage No. 12 of EHR2/0.8, however, a high expression of P-gp was found in spite of a low degree of resistance and a low efflux capacity. In this subline the effect of VER did not correlate to the expression of P-gp. Active efflux seemed to be saturable and was suggested to constitute the major route of efflux in MDR cells. A dose-response relationship was established for the effect of VER on efflux. In conclusion, the results support that P-gp acts as a drug efflux pump. No simple correlation, however, could be established between P-gp and drug transport in all the investigated cell lines. Other factors which might influence transmembranous transportation of DNR are suggested. The active efflux capacity of the cell lines seemed to determine the degree of resistance and the sensitivity to circumvention by VER.