Introduction. - Factors contribution to children's distress when a parent is affected with a cancer are still insufficiently known, This study aimed at searching for associations between psychosocial distress in children living with a parent suffering from cancer, the severity of parental cancer, the levels of psychosocial distress in both parents and the openness to discuss cancer in the family.Methods. - Thirty families encompassing a parent treated for cancer and 54 children aged four to 16 were examined. Each parent's psychosocial distress was assessed by the General Health Questionnaire (GHQ-28) and the distress of the children living within the family by the Child Behavior Check List (CBCL) filled Out by both parents. Each parent's communication ability about cancer was assessed by the Openness to Discuss Cancer in the nuclear Family questionnaire (ODCF).Results. - No association was found between children's distress and objective cancer characteristics, Higher externalized disorders scores at CBCL (aggression) were found when the ill parent was the mother (P = 0.018). After controlling for cancer parent's gender, CBCL total score and internalized disorders (anxiety, depression) score were higher in families characterized by an "open" style of communication, defined on the Parental couple as a whole (respectively p = 0.007 and 0.024), such in effect being, present only when the ill parent was the mother (interaction effect: p < 0.001).Conclusion. - These results underline the importance of family characteristics for Understanding the suffering observed in children living with a parent affected with a cancer in comparison with objective cancer characteristics. (c) 2008 Elsevier Masson SAS.

Introduction. - Each cancer can have a psychological impact not only on the patient him-self/herself, but also on his/her spouse.Objective. - Our study concerned 30 couples encompassing a member treated for a cancer, non related to gender. It was aimed at determining the [inks between the levels of psychosocial distress measured in both members of each couple, patients' sociodemographic and clinical characteristics, as well as communication skills about cancer in both members of the couples.Methods. - Psychosocial distress and communication about cancer were measured by the general health questionnaire (GHQ-28) and the openness to discuss cancer in the nuclear family (ODCF), with an additional version adapted for the spouse on the occasion of this study.Results. - A positive correlation was found between the respective scores of the two members of the couples, for the GHQ-28 (r=0.53; p=0.005) as well as for the ODCF (r=0.44; p=0.024). GHQ-28 scores were not associated with the sociodemographic characteristics of the patients, nor with the stage of cancer, the number of months elapsed since the diagnosis of cancer, or the ODCF personal or spouse's score. On the other hand, when the communication within each couple was classified into concordant (insufficient or, on the contrary, open for both members) or discordant (insufficient for one of the two members and open for the other), and after controlling for gender, higher levels of psychosocial distress were found in patients (p=0.038) as well in Spouses (p=0.052) belonging to discordant compared with concordant couples.Conclusion. - These results suggest an effect of contamination or a mutual reinforcement of the distress of each member of such couples, as well as the presence of relatively similar styles of communication in the two partners of each couple. They also underline the possible adaptive function of a restricted style of communication about cancer, if such a restriction is shared by both the members of the couple, and incites particular attention to be paid to couples where one of the partners, but not the other, adopt an open style of communication about cancer. (c) L'Encephale, Paris, 2008.

9059 Background: Nails and skin toxicities (NST) of the hand occur in about 50% of patients (pts) treated with docetaxel (D). No clinical data exists on specific NST of the foot after D treatment. We prospectively investigated the efficacy and safety of an Elasto-Gel (Akromed, France) flexible frozen sock (FS) for the prevention of D-induced NST according to the results of a precedent study on the hand toxicities (JCO 2005). Methods: Cancer pts receiving D at 75 to 100 mg/m2 (one hour infusion q3w) alone or in combination were eligible for this matched case-control study. Each patient wore a FS for a total of 90 minutes (min) on the right foot (15 min before to 15 min after D-infusion). The left foot acted as control was not protected by FS. NST were assessed at each cycle using NCI-CTC version 3 criteria and documented by digital photography. Comfort in socks wearing was assessed using an ad-hoc scale. Wilcoxon matched-pairs ranks test as a non-parametric method was used to determine the magnitude of the difference in terms of nails and skin toxicities between two matched groups. Results: Forty-nine pts were included: median age 64 years, M/F: 37/12, ECOG performance status 0/1/2 (%): 46/39/15, type of cancer (%): prostate: 53; breast: 20; lung: 18; others: 8. Nails toxicities were significantly lower in the FS-protected foot compared with the control foot (P=0.002, Wilcoxon test). Grade (G) of nails toxicities were (FS vs. control %): G0 100 vs. 80, G1–3; 0 vs. 20. Skin toxicity was not significantly different (P=0.18) with a low incidence of disorder (%) G0: 94 vs. 90, G1: 2 vs. 4, G2: 0 vs. 2 (missing data for 2 pts). Median time until nails toxicities occurrence was not significantly different between feet: 97 days with FS vs. 74 days in control foot. Skin toxicity occurred in both sides after a median of 96 days. FS comfort satisfied 74% of patients. Conclusions: Frozen sock significantly reduced nails toxicities of the foot associated with docetaxel treatment. No significant financial relationships to disclose.

BACKGROUND:We evaluated the possible use of prostate-specific antigen doubling time (PSA-DT) before chemotherapy initiation as a surrogate marker of survival in hormone-refractory prostate cancer (HRPC) patients. PATIENTS AND METHODS:Data from 250 consecutive metastatic HRPC patients treated with chemotherapy between February 2000 and November 2006 were retrospectively analysed. At least three PSA assays were required within 3 months before chemotherapy. PSA-DT was calculated as ln 2 divided by the slope of the log PSA line, and the difference between two log PSA levels was divided by the time interval. The primary endpoint was overall survival (OS). Survival rates according to PSA-DT were stratified on chemotherapy regimen. Multivariate Cox regression analysis was performed to isolate the impact of PSA-DT on OS, controlling for associate prognostic covariates. RESULTS:Patients received docetaxel- (82%) or mitoxantrone-based chemotherapy. The median PSA-DT was 45 days (range 4.7-1108 days). There were 174 deaths (70%). The median survival was 16.5 months (95% confidence interval [CI] = 12.5-20.5) and 26.4 months (95% CI = 20.3-32.4) for patients with a PSA-DT < 45 and > or =45 days, respectively. In the multivariate setting, the adjusted hazard ratio (HR) was 1.39 (95% CI = 1.03-1.89; P = 0.04), stratified by chemotherapy regimen. CONCLUSION:A short PSA-DT before onset of chemotherapy in HRPC patients was associated with an increased risk of death. This could be useful as a stratification parameter in trials with new drugs in a metastatic setting.

4634 Background: Docetaxel (DTX) based chemotherapy is considered the standard treatment for metastatic HRPC patients (pts), with a median survival advantage of 2 months, compared to mitoxantrone (MT). DTX advantage at long term is not defined and further evaluations are needed. Methods: We conducted a meta-analysis to assess the effect of DTX chemotherapy (CT) on overall survival (OS) compared to MT for metastatic HRPC pts. OS probabilities at 12, 18, 24, 30 and 36 months, were considered as the endpoints of interest. Subgroups analysis were performed to evaluate the impact on OS of DTX dose intensity (DI), between dose-dense (q1w) and dose-intense (q3w) schedules. Estimates of the effectiveness of CT were expressed as relative risk reduction (RRR), using a fixed effects model. Associated statistics with 95% confidence interval (CI) were calculated based on adjusted number of pts at risk and events (according to the extent of follow-up) using EasyMA software. Results: Three randomized controlled trials comparing DTX with MT-based CT in HRPC pts were selected (Tannock, Petrylak, NEJM 2004; Oudard, JCO 2005). A total of 1807 pts were included, 1092 allocated to DTX and 715 allocated to MT. The follow-up range was 13.9–58.5 months. No differences were found according to Gleason score, performance status and sites of metastasis. Overall analysis demonstrates a significant OS benefit for DTX at any time points, without any significant heterogeneity (Table). There was a significant difference in OS in favor of DTX dose-intense arm comparing to MT, with no OS benefit for DTX dose-dense schedule. The rank test for publication bias were not significant. Conclusions: This meta-analysis shows on a large data set that DTX-based CT significantly reduced the risk of death by 8–21%, a benefit persisting 3 years later after the start of CT. The benefit of DTX over MT was related to DTX schedule and dose-intense model. No significant financial relationships to disclose.

4641 Background: The PSA half-time (PSA HT) dynamics and corrected area under PSA curve (PSA c-AUC) during chemotherapy (CT) predict overall survival (OS) probability among metastatic HRPC patients...

7323 Background: Gefitinib (IressaTM) is approved in for use as monotherapy for the treatment of patients (pts) with locally advanced or metastatic NSCLC after failure of both platinum-based and docetaxel chemotherapy (CT). CT that include a taxane, vinorelbine or gemcitabine remains the alternative for the second- and third-line CT. The modality of combining these CT in a sequential manner and the place of target therapies could have a major impact on survival. Methods: The present analysis evaluates the efficacy of Gefitinib for metastatic NSCLC pts. The primary outcome measure was the impact of the moment of gefitinib initiation on Time from Primary Diagnosis (TPD). TPD was calculated from primary diagnosis until death or last follow-up. Cox hazard regression analysis was performed, stratified according the initial stage of disease (AJCC 1997) and baseline ECOG performance status (PS). Progression-free survival (PFS), overall survival (OS) and TPD were estimated by Kaplan-Meier method. Gefitinib was given at a dose of 250 mg daily until disease progression in an expanded access program. Results: A retrospective analysis was performed on 71 pts, treated with gefitinib from August 2001 to October 2004 in a single institution. Median age was 59 year (range 44–85), sex M/F: 52/19 pts, ECOG PS: 0–1/2–3: 20/44 pts. Pts distribution according to treatment line was:17 pts (24%) receiving gefitinib as second-line, 16 pts (23%) as third-line, 23 pts (32%) as fourth-line CT. There were 4 pts (6%) with partial response, and 15 pts (21%) with stable disease. The median PFS and OS were 2.6 months (95% CI,1.8–3.4 months) and 5 months (95% CI,3–7 months), respectively, independently of CT line number. The median TPD for our cohort was 18.5 months (95% CI, 15.7–21.2 months). A negative and strong association (Cox coefficient = -0.6) between TPD and the moment of gefitinib initiation was observed (P=0.006, two-sided), demonstrating a major impact of precocity of treatment initiation. Conclusions: Our analysis demonstrates that the prognosis in these patients depends on the precocity of gefitinib initiation. These data may be useful in the design of clinical trials to earlier introduce target therapy, combined with chemotherapy. No significant financial relationships to disclose.

Purpose Onycholysis and skin toxicity occur in approximately 30% of patients treated with docetaxel. We investigated the efficacy and safety of an Elasto-Gel (84400 APT Cedex, Akromed, France) frozen glove (FG) for the prevention of docetaxel-induced onycholysis and skin toxicity. Patients and Methods Patients receiving docetaxel 75 mg/m 2 alone or in combination chemotherapy were eligible for this case-control study. Each patient wore an FG for a total of 90 minutes on the right hand. The left hand was not protected and acted as the control. Onycholysis and skin toxicity were assessed at each cycle by National Cancer Institute Common Toxicity Criteria and documented by photography. Wilcoxon matched-pairs rank test was used. Results Between August 2002 and September 2003, 45 patients were evaluated. Onycholysis and skin toxicity were significantly lower in the FG-protected hand compared with the control hand (P = .0001). Onycholysis was grade (G) 0 in 89% v 49% and G1 to 2 in 11% v 51% for the FG-protected hand and the control hand, respectively. Skin toxicity was G0 in 73% v 41% and G1 to 2 in 27% v 59% for the FG-protected and the control hand, respectively. Median time to nail and skin toxicity occurrence was not significantly different between the FG-protected and the control hand, respectively (106 v 58 days for nail toxicity; 57 v 58 days for skin toxicity). Five patients (11%) experienced discomfort due to cold intolerance. Conclusion FG significantly reduces the nail and skin toxicity associated with docetaxel and provides a new tool in supportive care management to improve a patient's quality of life.

4709 Background: Predicting outcome for men with PC, treated with curative intent, remains imprecise and further evaluation of accepted and potential predictive factors is needed. Methods: The PSA exposure (PSA dynamics during lifetime) was estimated by corrected area under PSA curve [(PSA c-AUC), abstract 9579, ASCO 2004]. Based on their PSA c-AUC values (< 60 vs ≥ 60 ng/mL) and Gleason score (5–7 vs 8–10), patients (pts) were analyzed according to time from primary diagnosis (TPD) as major survival endpoint. TPD was calculated from primary diagnosis until death or last follow-up. A prognostic scoring system was created using the Kaplan-Meier test to classify pts according to risk of death. Our model identified four risk groups with predicted 3-, 5-, 7-, 10-, 13-, 15- and 18-year TPD percentages. Associated 95% confidence interval (CI) were calculated. Results: Characteristics of the 449 PC pts: median age at initial diagnosis was 64 years (range 41–82 years), median PSA follow-up duration of 2.7 years (range 0.1–16.1 years). Median number of PSA determinations was 15 (range 2–59). There were significant TPD differences (P = 0.00001) according to risk group (Table). The median TPD for the low PSA c-AUC level (< 60 ng/mL) groups was similar, independently of Gleason score [10.3 years (95% CI 8.7–11.9 years) vs 9.7 years (95% CI 7.1–13.1 years)]. Only 22% of pts in high risk group (Gleason > 8 and PSA c-AUC ≥ 60 ng/mL) are alive at 5 years, with a median TPD of 3 years (95% CI 2.3–3.8 years), two times lower than Gleason 5–7 and PSA c-AUC ≥ 60 ng/mL group [median of 7.3 years (95% CI 5.3–9.4 years), 67% alive]. Conclusions: By combining PSA c-AUC with Gleason score, we were able to predict TPD survival probabilities and to identify a high-risk group. A higher PSA exposure (≥ 60 ng/mL) during lifetime was correlated with a higher mortality. Consequently, these patients are candidates for early aggressive treatments in phase II-III clinical trials. No significant financial relationships to disclose.

Le manque de lits d’accueil en unités fixes ou mobiles de soins palliatifs oblige les équipes soignantes oncologiques à développer un concept proposé par Calman Hine en 1995, consistant en l’intégration de soins palliatifs de qualité au sein même des services curatifs de cancérologie. Dans ce but, un service d’oncologie et une unité de soins palliatifs ont créé une réunion régulière de présentation de dossiers de patients en échec de traitements curatifs. Ces échanges d’expertises permettent une formation pratique des équipes ainsi qu’un échange d’informations. La connaissance des dossiers des patients transférés permet d’éviter le sentiment de rupture que ceux-ci peuvent éprouver et avancer dans l’idée de continuité de soins. Une consultation de soins de suite et d’accompagnement a également été créée afin de préparer au mieux le patient et sa famille à ce transfert, tout en retardant au plus l’hospitalisation en gérant les situations au domicile. L’analyse de cinq mois de réunions donne une approche des manques et des espérances d’une telle démarche.

9586 Background: The concentration of serum hemoglobin (Hb) during rHuEpo treatment can be used to predict Hb response probability among patients (pts) treated by chemotherapy (CT). We were able to determine whether Hb dynamics over time correlate with rHuEpo efficacy defining a new predictive factor: Hb c-AUC. Methods: Our retrospective analysis validated the Hb c-AUC concept, defined as total Area Under serum Hb Curve (Hb AUC) per unit of time. Pts had regular Hb evaluation during rHuEpo treatment until blood cell transfusion, if applicable. The trapezoidal and Simpson's rules were used to calculate the Hb AUC. Response was defined as an Hb increase of at least 1 g/dL over the treatment. Based on their Hb c-AUC values, pts were analyzed according to maximal Hb response. Cox regression and log-rank tests were used after adjustment according to the Hb baseline level. Results: A total of 91 pts were analyzed: 46 men and 45 women, with a median age 61 years; 81 pts with solid tumors and 10 pts with hematological malignancies were treated with CT. Twenty-seven pts presented bone metastases. Median rHuEpo treatment duration and Hb c-AUC value were 76 days (8–422 days) and 10,3 g/dL (6.7–13.8 g/dL), respectively. Hb response was observed in 67% of pts. There was a significant difference in Hb response in favor of the high (> 10 g/dL) vs. low (≤ 10 g/dL) Hb c-AUC level group (73% vs. 41%, p=0.002), respectively. Predictive variables for maximal Hb response in univariate analysis were lymphocyte baseline level (p=0.04) and Hb c-AUC (p=0.0001). Only Hb c-AUC variable remained significant in multivariate analysis. Transfusional requirements were reduced by 30% in favor of high Hb c-AUC group (20% vs. 50%). Conclusions: Since only a proportion of cancer anemic pts benefit from rHuEpo, a reliable means of predicting potential responders would be useful. This analysis suggests that response to rHuEpo can be earlier predicted by Hb dynamics over time: Hb c-AUC. No significant financial relationships to disclose.

This study investigated the multidrug resistance, proliferation and apoptosis expression in renal cell carcinomas compared to adjacent normal kidney (ANK) tissues. Multidrug resistance (MDR1), multidrug resistance-associated protein (MRP), glutathione-S-transferase-pi (GST-pi), Topoisomerase-II alpha (TOPO-IIalpha), thymidylate synthase (TS), thymidine kinase (TK), Ki67, BAX and BCL-2 genes were analysed in a series of 30 renal cell carcinomas (RCC) and 16 biopsies from adjacent normal kidney (ANK) tissue using reverse-transcription-PCR (rt-PCR). The mean MDR1 expression was significantly lower in RCC than that of ANK (0.4 +/- 0.2 sd versus 0.75 +/- 0.19, p = 0.0008). The expression of MRP, GST-pi and TOPO-IIalpha was not significantly different in RCC as compared with ANK. The mean TK expression in RCC was significantly higher than in ANK (0.31 +/- 0.15 versus 0.09 +/- 0.08, p = 0.002). The TS and Ki67 expression in RCC was significantly higher than in ANK (87.5%, IC95% 71-100% versus 0%, p = 0.001; 56% IC95% 32-81% versus 0%, p = 0.004, respectively). BAX and BCL-2 expression in RCC was significantly higher than that of ANK (0.51 +/- 0.08 versus 0.18 +/- 0.12, p = 0.0001; 0.73 +/- 0.16 versus 0.5 +/- 0.22, p = 0.01, respectively). No significant correlation was found between MDR1, MRP, GST-pi, TOPO-IIalpha, TS, TK and BAX expression with the grade and the clinical stage in RCC.