Universal mean-field upper bound for the generalization gap of deep neural networks.

Modern deep neural networks (DNNs) represent a formidable challenge for theorists: according to the commonly accepted probabilistic framework that describes their performance, these architectures should overfit due to the huge number of parameters to train, but in practice they do not. Here we employ results from replica mean field theory to compute the generalization gap of machine learning models with quenched features, in the teacher-student scenario and for regression problems with quadratic loss function. Notably, this framework includes the case of DNNs where the last layer is optimized given a specific realization of the remaining weights. We show how these results-combined with ideas from statistical learning theory-provide a stringent asymptotic upper bound on the generalization gap of fully trained DNN as a function of the size of the dataset P. In particular, in the limit of large P and N_{out} (where N_{out} is the size of the last layer) and N_{out}≪P, the generalization gap approaches zero faster than 2N_{out}/P, for any choice of both architecture and teacher function. Notably, this result greatly improves existing bounds from statistical learning theory. We test our predictions on a broad range of architectures, from toy fully connected neural networks with few hidden layers to state-of-the-art deep convolutional neural networks.

Personalized treatment planning in eye brachytherapy for ocular melanoma: Dosimetric analysis on ophthalmic structure at risk.

PURPOSE: To evaluate the impact on dose distribution to eye organs-at-risk (eOARs) of a computed tomography (CT)-based treatment planning in eye plaque brachytherapy (EPB) treatment. METHODS: We analyzed 19 ocular melanoma patients treated with ruthenium-106 plaques to a total dose of 100 Gy to tumor apex using conventional central-axis-point dose calculation. Treatments were re-planned using the Plaque Simulator (PS) software implementing two different strategies: a personalized CT-eye-model (CT-PS) and a standard-eye-model (SEM-PS) defined by Collaborative Ocular Melanoma Study. Dice coefficient and Hausdorff distance evaluated the concordance between eye-bulb-models. Mean doses (Dmean) to tumor and eOARs were extracted from Dose-Volume-Histograms and Retinal-Dose-Area-Histogram. Differences between planning approaches were tested by Wilcoxon signed-rank test. RESULTS: In the analyzed cohort, 8 patients (42%) had posterior tumor location, 8 (42%) anterior, and 3 (16%) equatorial. The SEM did not accurately described the real CT eye-bulb geometry (median Hausdorff distance 0.8 mm, range: (0.4-1.3) mm). Significant differences in fovea and macula Dmean values were found (p = 0.04) between CT-PS and SEM-PS schemes. No significant dosimetric differences were found for tumor and other eOARs. The planning scheme particularly affects the OARs closest to the tumor with a general tendency of SEM-PS to overestimate the doses to the OARs closest to the tumor. CONCLUSION: The dosimetric accuracy achievable with CT-PS EPB treatment planning may help to identify ocular melanoma patients who could benefit the most from a personalized eye dosimetry for an optimal outcome in terms of tumor coverage and eOARs sparing. Further research and larger studies are underway.

Radiation-induced acute dysphagia : Prospective observational study on 42 head and neck cancer patients.

PURPOSE: Acute toxicity in head and neck (H&N) cancer patients treated with definitive radiotherapy (RT) has a crucial role in compliance to treatments. The aim of this study was to correlate doses to swallowing-associated structures and acute dysphagia. METHODS: We prospectively analyzed 42 H&N cancer patients treated with RT. Dysphagia (grade ≥ 3) and indication for percutaneous endoscopic gastrostomy (PEG) insertion were classified as acute toxicity. Ten swallowing-related structures were considered for the dosimetric analysis. The correlation between clinical information and the dose absorbed by the contoured structures was analyzed. Multivariate logistic regression method using resampling methods (bootstrapping) was applied to select model order and parameters for normal tissue complication probability (NTCP) modelling. RESULTS: A strong multiple correlation between dosimetric parameters was found. A two-variable model was suggested as the optimal order by bootstrap method. The optimal model (Rs = 0.452, p < 0.001) includes V45 of the cervical esophagus (odds ratio [OR] = 1.016) and Dmean of the cricopharyngeal muscle (OR = 1.057). The model area under the curve was 0.82 (95% confidence interval 0.69-0.95). CONCLUSION: Our results suggested that the absorbed dose to the cricopharyngeal muscle and cervical esophagus might play a relevant role in the development of acute RT-related dysphagia.

Predictors of Asymptomatic Radiation-induced Abdominal Atherosclerosis.

AIMS: To identify predictors of asymptomatic radiation-induced abdominal atherosclerosis in patients treated with radiotherapy and evaluated by abdominal vascular ultrasonography. MATERIALS AND METHODS: Forty-two testicular classic seminoma patients (median age 34 years, range 16-56) undergoing radical inguinal orchiectomy were analysed. Twenty-six patients underwent post-surgery radiotherapy (median total dose 25 Gy, range 25-43), two of them also received chemotherapy (CHT) and 16 patients were treated with surgery alone or by surgery followed by CHT (control group). The presence of stenosis in an abdominal vessel and renal resistive index (RRI), evaluated by echo-colour Doppler (ECD), were considered as indicators of late vascular damage. Chi-square and Mann-Whitney tests were used to compare groups. For the radiotherapy group, near maximum (D2%) and mean dose (Dmean) metrics of critical structures (abdominal arteries and renal hila) were extracted from retrievable dose maps (18 of 26 radiotherapy patients). To evaluate clinical and dosimetric factors associated with vascular damage, univariate and multivariate analyses were carried out. The impact of dose to arteries, evaluated as separate subvolumes, was analysed comparing the stenotic arteries with normal ones by logistic regression. The area under the receiver operator characteristic curve (AUC) was used to evaluate the test accuracy. RESULTS: In the radiotherapy group there was a significantly different incidence of stenosis (31% versus 0%, P = 0.016) and a higher median average RRI (0.63 versus 0.60, P = 0.032) compared with the control group. The median time intervals between treatment and ECD were 64 months (range 12-120) and 48 months (range 12-168) in the radiotherapy and control groups (P = 0.399), respectively. A younger age at radiotherapy was the only clinical risk factor for stenosis (P = 0.006). Artery Dmean was significantly associated with stenosis (P = 0.008), with an odds ratio of 1.13 (95% confidence interval 1.01-1.26) and an AUC of 0.85 (95% confidence interval 0.77-0.91). Renal hilum D2% was correlated with RRI (Rs = 0.406, P = 0.02). CONCLUSIONS: Late vascular damage represents a potential effect of abdominal radiotherapy, even at a moderate dose. Younger age at irradiation, artery and renal hila dose metrics are associated with increased risk. Ultrasound-based follow-up may allow for non-invasive early detection of asymptomatic radiation-induced damage, helping to prevent severe vascular events.

Radiotherapy for the treatment of thyroid eye disease-a prospective comparison: Is orbital radiotherapy a suitable alternative to steroids?

BACKGROUND AND AIMS: To evaluate the efficacy of orbital radiotherapy (OR) for the treatment of thyroid eye disease (TED). METHODS: Thirty-five consecutive patients with active TED with contraindications to steroid therapy received a course of OR. Bilateral retrobulbar irradiation was performed with a total dose of 20 Gy. 7-points clinical activity score (7-CAS), ocular motility, visual acuity (VA), exophthalmos and eyelid retraction were prospectively evaluated at 3, 6 and 12 months and compared with baseline data. RESULTS: There was a statistically significant improvement in 7-CAS at 3, 6 and 12 months post-treatment (p < 0.05). Ocular motility disturbances improved at 6 and 12 months (p < 0.05). Visual acuity remained stable; there was no significant change in exophthalmos (mean 24 mm, SD 3 mm) or eyelid retraction (marginal reflex distance mean 6 mm, SD 1.5 mm) during the follow-up period. No side effects were registered. CONCLUSIONS: This study suggests that OR might be effective in reducing 7-CAS and ocular motility disturbances. No significant improvement in proptosis or eyelid retraction should be expected from this treatment. OR might be considered a suitable alternative treatment in TED for patients who cannot tolerate steroids.

Effect of preoperative chemoradiotherapy on outcome of patients with locally advanced esophagogastric junction adenocarcinoma-a pilot study.

BACKGROUND: To date, few studies of preoperative chemotherapy or chemoradiotherapy (crt) in gastroesophageal junction (gej) cancer have been statistically powered; indeed, gej tumours have thus far been grouped with esophageal or gastric cancer in phase iii trials, thereby generating conflicting results. METHODS: We studied 41 patients affected by locally advanced Siewert type i and ii gej adenocarcinoma who were treated with a neoadjuvant crt regimen [folfox4 (leucovorin-5-fluorouracil-oxaliplatin) for 4 cycles, and concurrent computed tomography-based three-dimensional conformal radiotherapy delivered using 5 daily fractions of 1.8 Gy per week for a total dose of 45 Gy], followed by surgery. Completeness of tumour resection (performed approximately 6 weeks after completion of crt), clinical and pathologic response rates, and safety and outcome of the treatment were the main endpoints of the study. RESULTS: All 41 patients completed preoperative treatment. Combined therapy was well tolerated, with no treatment-related deaths. Dose reduction was necessary in 8 patients (19.5%). After crt, 78% of the patients showed a partial clinical response, 17% were stable, and 5% experienced disease progression. Pathology examination of surgical specimens demonstrated a 10% complete response rate. The median and mean survival times were 26 and 36 months respectively (95% confidence interval: 14 to 37 months and 30 to 41 months respectively). On multivariate analysis, TNM staging and clinical response were demonstrated to be the only independent variables related to long-term survival. CONCLUSIONS: In our experience, preoperative chemoradiotherapy with folfox4 is feasible in locally advanced gej adenocarcinoma, but shows mild efficacy, as suggested by the low rate of pathologic complete response.

Complete response to preoperative chemoradiation and survival in esophageal cancer: a pooled analysis of three single-institution phase II trials.

This pooled analysis was performed using individual patient data from three phase II trials that included on the whole 113 esophageal cancer treated preoperatively with chemoradiotherapy (CRT), in order to analyze the efficacy and survival outcomes according to the achievement of the pathologic complete response (pCR). Thirty-nine patients were treated with 5-fluorouracil/cisplatin and RT (40 Gy), 33 patients received paclitaxel/cisplatin weekly during weeks 1-6 with and RT (46 Gy), 41 patients were treated with induction bio-chemotherapy with cetuximab and FOLFOX-4 followed by concomitant cetuximab and RT of 50.4 Gy. One hundred and two out of 113 resected patients were included in the survival analysis. The median overall survival (OS) time for the whole population was 21.5 months. The 12, 24, and 36 months OS rates were 85.4, 45.2, and 33%, respectively. The difference in survival probability between patients with pCR and patients with partial response or stable disease after treatment was significant (P= 0.0002, hazard ratios = 0.21, 95% CI 0.18-0.60). On multivariate analysis, the pathologic response and histology were the only covariates independently associated with OS (P= 0.0157 and P= 0.0212, respectively). In our series, complete responder patients had a significant longer survival probability after treatment when compared to patients with partial response or stable disease.

HMGA proteins promote ATM expression and enhance cancer cell resistance to genotoxic agents.

DNA-damaging therapies represent a keystone in cancer treatment. Unfortunately, many tumors often relapse because of a group of cancer cells, which are resistant to conventional therapies. High-mobility group A (HMGA) proteins has a key role in cell transformation, and their overexpression is a common feature of human malignant neoplasias, representing a poor prognostic index often correlated to anti-cancer drug resistance. Our previous results demonstrated that HMGA1 is a substrate of ataxia-telangiectasia mutated (ATM), the main cellular sensor of genotoxic stress. Here we also report thatHMGA2, the other member of the HMGA family, is a novel substrate of ATM. Interestingly, we found that HMGA proteins positively regulate ATM gene expression. Moreover, induction of ATM kinase activity by DNA-damaging agents enhances HMGA-dependent transcriptional activation of ATM promoter, suggesting that ATM expression is modulated by a DNA-damage- and HMGA-dependent positive feedback loop. Finally, inhibition of HMGA expression in mouse embryonic fibroblasts and in cancer cells strongly reduces ATM protein levels, impairing the cellular DNA-damage response and enhancing the sensitivity to DNA-damaging agents. These findings indicate this novel HMGA-ATM pathway as a new potential target to improve the effectiveness of conventional anti-neoplastic treatments on the genotoxic-drug resistant cancer cells.

Role of FK506-binding protein 51 in the control of apoptosis of irradiated melanoma cells.

FK506-binding protein 51 (FKBP51) is an immunophilin with isomerase activity, which performs important biological functions in the cell. It has recently been involved in the apoptosis resistance of malignant melanoma. The aim of this study was to investigate the possible role of FKBP51 in the control of response to ionizing radiation (Rx) in malignant melanoma. FKBP51-silenced cells showed reduced clonogenic potential after irradiation compared with non-silenced cells. After Rx, we observed apoptosis in FKBP51-silenced cells and autophagy in non-silenced cells. The FKBP51-controlled radioresistance mechanism involves NF-kappaB. FKBP51 was required for the activation of Rx-induced NF-kappaB, which in turn inhibited apoptosis by stimulating X-linked inhibitor of apoptosis protein and promoting authophagy-mediated Bax degradation. Using a tumor-xenograft mouse model, the in vivo pretreatment of tumors with FKBP51-siRNA provoked massive apoptosis after irradiation. Immunohistochemical analysis of 10 normal skin samples and 80 malignant cutaneous melanomas showed that FKBP51 is a marker of melanocyte malignancy, correlating with vertical growth phase and lesion thickness. Finally, we provide evidence that FKBP51 targeting radiosensitizes cancer stem/initiating cells. In conclusion, our study identifies a possible molecular target for radiosensitizing therapeutic strategies against malignant melanoma.

Early clinical and neuroradiological worsening after radiotherapy and concomitant temozolomide in patients with glioblastoma: tumour progression or radionecrosis?

OBJECTIVES: This study investigates the diagnosis and management of patients with resected brain glioblastomas who presented early clinical and neuroradiological worsening after the completion of the Stupp protocol. Its aim is to discuss the occurrence of early radionecrosis. METHODS: Fifty patients with brain glioblastoma treated by surgical resection and Stupp protocol were reviewed; 15 among them (30%) had early clinical and neuroradiological worsening at the 6-month follow-up. The MR spectroscopy and surgical findings of these patients are reviewed. RESULTS: MR spectroscopy was in favour of tumour recurrence in 14 among 15 patients and showed radionecrosis in one. Among 10 patients who were reoperated on, 7 had histologically verified tumour recurrence or regrowth, whereas in 3 histopathology showed necrosis without evidence of tumour. The 7 patients with tumour progression had prevalence of focal neuroradiological signs (6/7) and a survival of 7.5-12 months (median survival 10 months). The 4 patients with early radionecrosis (including one patient who was not reoperated on) had clinical worsening with mental deterioration, confusion and ataxia, and MR spectroscopy positive for tumour recurrence in 3. Three were alive 24-30 months after the end of the radiotherapy, whereas one died at 40 months. CONCLUSION: Early radionecrosis after the Stupp protocol is not a rare event due to the radiosensitization effect of temozolomide. This phenomenon may predict a durable response to radiotherapy. MR spectroscopy may simulate tumour recurrence. A correct diagnosis is necessary to avoid useless reoperations and incorrect withdrawal of temozolomide.

Neo-adjuvant treatment of rectal cancer with capecitabine and oxaliplatin in combination with radiotherapy: a phase II study.

BACKGROUND: Preoperative chemoradiation is now standard treatment for stages II-III rectal cancer. Capecitabine (CAP) and oxaliplatin (OX) are synergistic with radiotherapy (RT) and active in colorectal neoplasms. PATIENTS AND METHODS: Two cycles of CAP 825 mg/m(2) b.i.d. (days 1-14) and OX 50 mg/m(2) (days 1 and 8) every 3 weeks were given concomitantly with pelvic conformal RT (45 Gy). Patients with a > or =T3 and/or node-positive rectal tumour were eligible. The pathologic tumour response was defined according to the tumour regression grade (TRG) scale. RESULTS: Forty-six patients were enrolled. Gastrointestinal adverse events were mostly G1-G2; only two patients experienced G3 vomiting and diarrhoea and six patients had G1 peripheral neuropathy. Haematological toxicity was rare. G2 proctitis and anal pain occurred in two patients. Pathological complete response (TRG1) was observed in nine patients (20.9%; 95% CI 8.7%-33.1%); TRG2 in 19 patients (44.2%); TRG3 in 12 patients (27.9%); and TRG4 in three patients (7%). Overall, nine patients recurred: five with distant metastases, one with local recurrence, and three with both local recurrence and distant metastases. CONCLUSIONS: CAP-OX-RT as preoperative treatment for rectal cancer induces a remarkable rate of complete or near-complete pathologically documented response and is well tolerated.

Involvement of H4(D10S170) protein in ATM-dependent response to DNA damage.

H4(D10S170) gene has been identified upon its frequent rearrangement with RET in papillary thyroid tumours (RET/PTC1). The kinase ataxia telangectasia mutated (ATM) phosphorylates a limited number of downstream protein targets in response to DNA damage. We investigated the potential role of H4(D10S170) in DNA damage signaling pathways. We found that in cells treated with etoposide or ionizing radiation (IR), H4(D10S170) underwent ATM-mediated phosphorylation at Thr 434, stabilizing nuclear H4. In ataxia telangectasia cells (A-T), endogenous H4(D10S170) was localized to cytoplasm and was excluded from the nucleus. Moreover, H4(D10S170) was not phosphorylated in ATM-deficient lymphoblasts after ionizing irradiation. Inhibition of ATM kinase interfered with H4(D10S170) apoptotic activity, and expression of H4 with threonine 434 mutated in Alanine, H4(T434A), protected the cells from genotoxic stress-induced apoptosis. Most importantly, after exposure to IR we found that silencing of H4(D10S170) in mammalian cells increased cell survival, as shown by clonogenic assay, allows for DNA synthesis as evaluated by bromodeoxyuridine incorporation and permits cells to progress into mitosis as demonstrated by phosphorylation on Histone H3. Our results suggest that H4(D10S170) is involved in cellular response to DNA damage ATM-mediated, and that the impairment of H4(D10S170) gene function might have a role in thyroid carcinogenesis.

Comparison of control of Listeria by nitric oxide redox chemistry from murine macrophages and NO donors: insights into listeriocidal activity of oxidative and nitrosative stress.

The physiological function of nitric oxide (NO) in the defense against pathogens is multifaceted. The exact chemistry by which NO combats intracellular pathogens such as Listeria monocytogenes is yet unresolved. We examined the effects of NO exposure, either delivered by NO donors or generated in situ within ANA-1 murine macrophages, on L. monocytogenes growth. Production of NO by the two NONOate compounds PAPA/NO (NH2(C3H6)(N[N(O)NO]C3H7) and DEA/NO (Na(C2H5)2N[N(O)NO]) resulted in L. monocytogenes cytostasis with minimal cytotoxicity. Reactive oxygen species generated from xanthine oxidase/hypoxanthine were neither bactericidal nor cytostatic and did not alter the action of NO. L. monocytogenes growth was also suppressed upon internalization into ANA-1 murine macrophages primed with interferon-gamma (INF-gamma) + tumor necrosis factor-alpha (TNF-alpha or INF-gamma + lipid polysaccharide (LPS). Growth suppression correlated with nitrite formation and nitrosation of 2,3-diaminonaphthalene elicited by stimulated murine macrophages. This nitrosative chemistry was not dependent upon nor mediated by interaction with reactive oxygen species (ROS), but resulted solely from NO and intermediates related to nitrosative stress. The role of nitrosation in controlling L. monocytogenes was further examined by monitoring the effects of exposure to NO on an important virulence factor, Listeriolysin O, which was inhibited under nitrosative conditions. These results suggest that nitrosative stress mediated by macrophages is an important component of the immunological arsenal in controlling L. monocytogenes infections.

Principles of chemoradiation: theoretical and practical considerations.

Chemotherapy agents known to enhance the effects of radiation in preclinical studies have been used concurrently with radiotherapy in numerous clinical trials with the prospect of further enhancing radiation-induced local tumor control. While some success in several tumor histologies has been achieved using this approach, a major concern has been enhancement in normal tissue toxicity. This brief review addresses both theoretical and practical issues with respect to chemoradiation clinical trials. Recommendations for clinical trials are provided that, if implemented, can increase our understanding of basic mechanisms (in patients) and provide a more rational approach for future trials.

Effects of amygdaloid lesions, hippocampal lesions, and buspirone on black-white exploration and food carrying in rats.

Unlesioned rats exploring a black-white two compartment box spent most of the time in the covered, black half of the box and only little time in the uncovered, white half (67 s/5 min). Large radio-frequency lesions of the amygdala or hippocampus did not alter this pattern of exploration, but rats with hippocampus lesions were more active than the other two groups of rats. Treatment with the 5-HT1A receptor agonist buspirone (0.1 mg/kg, s.c.) increased the time that unlesioned rats spent in the uncovered compartment (103 s), an effect that was less pronounced in hippocampus-lesioned rats and completely abolished by amygdala lesions. In a food transport test, unlesioned rats that traveled from a home cage to an exposed food source consumed small and medium-sized pellets immediately at the food source. Larger pellets, however, were carried back to the home cage for consumption. Rats with amygdala lesions ate fewer pellets at the food source and tended to carry more pellets back to the home cage for consumption than unlesioned rats. Rats with hippocampus lesions carried fewer pellets back to the home cage and ate more pellets at the food source. Buspirone (0.5-1.5 mg/kg, s.c.) reduced the carrying of large food items to the home cage and increased consumption of these pellets at the food source in all groups of rats. These results suggest that neither the amygdala nor the hippocampus play an important role in controlling exploratory behavior in a black-white compartment box, but that the amygdala may have some role in mediating the effect of buspirone to increase exploration of the white/open compartment. Further, the amygdala and hippocampus have opposing influences on the transport of food items to a shelter, the amygdala suppressing food carrying, and the hippocampus enhancing it. Neither structure is essential for the effect of buspirone to reduce food carrying. The hypothesis that limbic structures mediate 'fear/anxiety' responses is discussed critically.

Superoxide modulates the oxidation and nitrosation of thiols by nitric oxide-derived reactive intermediates. Chemical aspects involved in the balance between oxidative and nitrosative stress.

Thiol-containing proteins are key to numerous cellular processes, and their functions can be modified by thiol nitrosation or oxidation. Nitrosation reactions are quenched by O-2, while the oxidation chemistry mediated by peroxynitrite is quenched by excess flux of either NO or O-2. A solution of glutathione (GSH), a model thiol-containing tripeptide, exclusively yielded S-nitrosoglutathione when exposed to the NO donor, Et2NN(O)NONa. However, when xanthine oxidase was added to the same mixture, the yield of S-nitrosoglutathione dramatically decreased as the activity of xanthine oxidase increased, such that there was a 95% reduction in nitrosation when the fluxes of NO and O-2 were nearly equivalent. The presence of superoxide dismutase reversed O-2-mediated inhibition, while catalase had no effect. Increasing the flux of O-2 yielded oxidized glutathione (GSSG), peaking when the flux of NO and O-2 were approximately equivalent. The results suggest that oxidation and nitrosation of thiols by superoxide and NO are determined by their relative fluxes and may have physiological significance.

Nitric oxide and some nitric oxide donor compounds enhance the cytotoxicity of cisplatin.

A major emphasis in cancer therapy research is finding mechanisms to enhance the effectiveness of clinically used chemotherapeutic agents. In this report, we show the effects of direct NO exposure or NO delivery agents such as NONOate NO donors, DEA/NO ((C2H5)2N[N(O)NO]-Na+) and PAPA/ NO (NH2(C3H6)(N[N(O)NO]C3H7)), or S-nitrosothiol NO donors (GSNO, S-nitrosoglutathione, and SNAP, S-nitroso-N-acetylpenicillamine) on the cytotoxicity of cisplatin with Chinese hamster V79 lung fibroblast cells. Cells pretreated with bolus NO or NO delivered from NONOate NO donors were markedly sensitized to subsequent cisplatin treatment, whereas S-nitrosothiol NO donors exerted little effect. The enhancement in cisplatin cytotoxicity from pretreatment with DEA/NO and PAPA/ NO persisted for approximately 180 and 240 min, respectively; thereafter cytotoxicity returned to a level consistent with cisplatin treatment alone. Pretreatment of cells with GSNO or SNAP did not enhance cisplatin cytotoxity. To discern why there were differential effects among the different NO donors, formation of NO over the time course of the experiment was assessed by the nitrosation of 2,3-diaminonaphthylene. Bolus NO, DEA/NO, and PAPA/NO produced more reactive nitrogen oxide species (RNOS) than did treatment with GSNO or SNAP. Previously reported electrochemical studies revealed that temporal NO concentrations measured from DEA/NO and PAPA/NO (1 mM) were greater than 5 microM. It appears that the flux of NO, as well as the amount of RNOS, is important in the NO-mediated enhancement of cisplatin cytotoxicity. Our results demonstrate the importance of NO delivery systems in the enhancement of cisplatin cytotoxicity and may provide insights into strategies for participation of NO donors and nitric oxide synthase with cisplatin therapy.

Nitric oxide enhancement of melphalan-induced cytotoxicity.

The effects of the diatomic radical, nitric oxide (NO), on melphalan-induced cytotoxicity in Chinese hamster V79 and human MCF-7 breast cancer cells were studied using clonogenic assays. NO delivered by the NO-releasing agent (C2H5)2N[N(O)NO]- Na+ (DEA/NO; 1 mM) resulted in enhancement of melphalan-mediated toxicity in Chinese hamster V79 lung fibroblasts and human breast cancer (MCF-7) cells by 3.6- and 4.3-fold, respectively, at the IC50 level. Nitrite/nitrate and diethylamine, the ultimate end products of DEA/NO decomposition, had little effect on melphalan cytotoxicity, which suggests that NO was responsible for the sensitization. Whereas maximal sensitization of melphalan cytotoxicity by DEA/NO was observed for simultaneous exposure of DEA/NO and melphalan, cells pretreated with DEA/NO were sensitized to melphalan for several hours after NO exposure. Reversing the order of treatment also resulted in a time-dependent enhancement in melphalan cytotoxicity. To explore possible mechanisms of NO enhancement of melphalan cytotoxicity, the effects of DEA/NO on three factors that might influence melphalan toxicity were examined, namely NO-mediated cell cycle perturbations, intracellular glutathione (GSH) levels and melphalan uptake. NO pretreatment resulted in a delayed entry into S phase and a G2/M block for both V79 and MCF-7 cells; however, cell cycle redistribution for V79 cells occurred after the cells returned to a level of cell survival, consistent with treatment with melphalan alone. After 15 min exposure of V79 cells to DEA/NO (1 mM), GSH levels were reduced to 40% of control values; however, GSH levels recovered fully after 1 h and were elevated 2 h after DEA/NO incubation. In contrast, DEA/NO (1 mM) incubation did not reduce GSH levels significantly in MCF-7 cells (approximately 10%). Melphalan uptake was increased by 33% after DEA/NO exposure in V79 cells. From these results enhancement of melphalan cytotoxicity mediated by NO appears to be complex and may involve several pathways, including possibly alteration of the repair of melphalan-induced lesions. Our observations may give insights for improving tumour kill with melphalan using either exogenous or possibly endogenous sources of NO.