Supplementary Tables S1-S5 from Prognosis and Gene Expression Profiling of 20q13-Amplified Breast Cancers

Supplementary Figure S1 from Prognosis and Gene Expression Profiling of 20q13-Amplified Breast Cancers

Supplementary Data from Aldehyde Dehydrogenase 1–Positive Cancer Stem Cells Mediate Metastasis and Poor Clinical Outcome in Inflammatory Breast Cancer

Supplementary Figures 1-4 from Comprehensive Profiling of 8p11-12 Amplification in Breast Cancer

Supplementary Tables 1-7 from Comprehensive Profiling of 8p11-12 Amplification in Breast Cancer

BACKGROUNDThe present monocentric and prospective phase 1 study evaluated the safety of a metronomic chemotherapy in refractory tumors.PATIENTS AND METHODSPatients with advanced solid cancer refractory to standard therapy received a combination of low-dose vinorelbine, cyclophosphamide and interferon-alpha. A dose escalation model with 3 levels was planned. The primary end-point was safety and tolerability, secondary end-points were treatment continuation rate at 4 months, progression-free survival (PFS), overall survival (OS), radiological assessment (MRI) of anti-angiogenic effect.RESULTSThirty patients were enrolled. No dose-limiting toxicity was observed. All but two adverse events were toxicities of grade 1-2. Treatment continuation rate at 4 months was 6.67% (2 out of 30 patients). Median PFS and OS were 1.6 and 6.1 months. Exploratory MRI analyses related to anti-angiogenic effect did not show any relevant modification.CONCLUSIONThis combination of metronomic chemotherapy is well-tolerated and deserves to be deeply explored in refractory solid tumors.

Acute Myeloid Leukemia (AML) and myelodysplasia (MDS) with chromosome 3q abnormalities have a dismal outcome either untreated or with conventional treatments. Azacitidine (AZA) is now considered as the standard of care in high-risk MDS and oligoblastic AML patients. The objective of this study was to evaluate the impact of azacitine treatment in this cytogenetic subgroup. We report here a multicentre retrospective study of 157 patients treated with AZA for AML/MDS with chromosome 3q abnormalities and 27 patients with isolated EVI-1 overexpression. Median age was 65 years, 40 patients (25%) had inv(3)(q21q26.2) or t(3;3)(q21;q26.2), 36 patients (23%) had other balanced 3q26 rearrangements, 8 patients (5%) had balanced 3q21 rearrangements and 73 patients (46%) had other 3q abnormalities. The overall response rate was 50% (29% CR). Median overall survival was 10.6 months. By multivariate analysis, patients with lower bone marrow blast counts, higher platelet counts, non-complex cytogenetics, and absence of prior treatment with intensive chemotherapy had a better outcome. 27 patients were allo-transplanted and achieved a 21-month median OS. Balanced 3q21 translocations were associated with a better response rate and overall survival. Outcome of patients with isolated EVI-1 overexpression was comparable to that of patients with chromosome 3q lesions. Thus, AML/MDS patients with 3q abnormalities appear to be a heterogeneous group in their response to AZA, and AZA may represent a suitable option in particular as a bridge to allogeneic transplantation.

The introduction of reduced intensity/toxicity conditioning regimens has allowed allogeneic hematopoietic cell transplantation to be performed in patients who were previously considered too old or otherwise unfit. Although it led to a reduction in non-relapse mortality, disease control remains a major challenge. We studied the outcome of 165 patients with acute myeloid leukemia (n=124) or myelodysplastic syndrome (n=41) transplanted after conditioning with fludarabine (30 mg/m2/day for 5 days), intravenous busulfan (either 260 mg/m2: reduced intensity conditioning, or 390–520 mg/m2: reduced toxicity conditioning), and rabbit anti-thymoglobulin (2.5 mg/kg/day for 2 days). The median age of the patients at transplantation was 56.8 years. The 2-year relapse incidence was 29% (23% versus 39% for patients transplanted in first complete remission and those transplanted beyond first complete remission, respectively; P=0.008). The 2-year progression-free survival rate was 57% (95% CI: 49.9–65). It was higher in the groups with favorable or intermediate cytogenetics than in the group with unfavorable cytogenetics (72.7%, 60.5%, and 45.7%, respectively; P=0.03). The cumulative incidence of grades 2–4 and 3–4 acute graft-versus-host disease at day 100 was 19.3% and 7.9%, respectively. The cumulative incidence of chronic graft-versus-host disease at 1 year was 21.6% (severe forms: 7.8%). Non-relapse mortality at 1 year reached 11%. The 2-year overall survival rate was 61.8% (95% CI: 54.8–69.7). Unfavorable karyotype and disease status beyond first complete remission were associated with a poorer survival. This well-tolerated conditioning platform can lead to long-term disease control and offers possibilities of modulation according to disease stage or further development.

BACKGROUND:Breast cancer (BC) is the second most common cause of brain metastases (BM). Optimal management of BM from BC is still debated. In an attempt to provide appropriate treatment and to assist with optimal patient selection, several specific prognostic classifications for BM from BC have been established. We evaluated the prognostic value and validity of the 6 proposed scoring systems in an independent population of BC patients with BM.METHODS:We retrospectively reviewed all consecutive BC patients referred to our institution for newly diagnosed BM between October 1995 and July 2011 (n = 149). Each of the 6 scores proposed for BM from BC (Sperduto, Niwinska, Park, Nieder, Le Scodan, and Claude) was applied to this population. The discriminative ability of each score was assessed using the Brier score and the C-index. Individual prognostic values of clinical and histological factors were analyzed using uni- and multivariate analyses.RESULTS:Median overall survival was 15.1 months (95% CI,11.5-18.7). Sperduto-GPA (P < .001), Nieder (P < .001), Park (P < .001), Claude (P < .001), Niwinska (P < .001), and Le Scodan (P = .034) scores all showed significant prognostic value. The Nieder score showed the best discriminative ability (C-index, 0.672; Brier score error reduction, 16.1%).CONCLUSION:The majority of prognostic scores were relevant for patients with BM from BC in our independent population, and the Nieder score seems to present the best predictive value but showed a relatively low positive predictive value. Thus, these results remain insufficient and challenge the routine use of these scoring systems.

8568 Background: Patients with refractory/relapsed Hodgkin Lymphoma and Non Hodgkin Lymphoma (R/R HL-NHL) have poor prognosis. Allogeneic stem cell transplantation (allo-HSCT) represents a potentially curative strategy. The increased use of Reduced Intensity conditioning (RIC) regimens has decreased the non-relapse mortality (NRM) previously reported with myeloablative conditioning. However, Chronic Graft versus Host disease (cGVHD) is still matter of concern. The aim of this study is to evaluate the long-term survival of a cohort of R/R HL-NHL. Methods: From 9/1999 to 10/2013, 314 patients from 2 associated institutions received allo-HSCT. Diagnoses were aggressive-NHL in 157 patients, indolent-NHL in 59 and HL in 98. 148 patients relapsed after autologous transplant (auto-HSCT). Out of the 166 remaining patients, 109 received a tandem auto-allo-HSCT and 57 benefit of allo-HSCT without prior auto-HSCT. At transplant, 172 were on complete remission, 86 were on partial remission and 56 were not in response. Donors were HLA identical sibling in 54%, unrelated in 21% and haploidentical in 21%. Conditioning regimen consisted in RIC in 68% of cases. Results: With a median follow-up of 29 months [2-160] the 5-year overall survival (OS), the 5-years progression free survival (PFS) and the five years incidence of relapse were 53.9% [47-61], 47.2% [41-54] and 26%[21-32]. The 1-years NRM was 18% [14-23]. Incidence of acute and chronic GVHD was 28% (9% grade 3-4) and 35% (severe NIH cGVHD: 14%). The median duration of immunosuppressive therapy was 7 months [2-117] and it was interrupted in 72% at 1 year for patients who did not relapse. In multivariate analysis, disease status at transplant only influences OS. Age, comorbidities, prior auto-ASCT, donor and conditioning regimen do not have any significant prognostic impact. Conclusions: Allo-ASCT is an option for patients with R/RHL-NHL with a 5-years OS of 54% and PFS of 47%. Taking into account population characteristics and period of treatment, toxicity is acceptable with only 14% of severe NIH cGVHD. Most of patient (72%) could stop immunosuppressive therapy demonstrating the absence of GVHD affecting their quality of life.

To define the prognostic value of isolated tumor cells (ITC), micrometastases (pN1mi) and macrometastases in early stage breast cancer (ESBC). We conducted a retrospective multicenter cohort study at 13 French sites. All the eligible patients who underwent SLNB from January 1999 to December 2008 were identified, and appropriate data were extracted from medical records and analyzed. Among 8001 patients, including 70% node-negative (n = 5588), 4% ITC (n = 305), 10% pN1mi (n = 794) and 16% macrometastases (n = 1314) with a median follow-up of 61.3 months, overall survival (OS) and recurrence-free survival (RFS) rates at 84 months were not statistically different in ITC or pN1mi compared to tumor-free nodes. Axillary recurrence (AR) was significantly more frequent in ITC (1.7%) and pN1mi (1.5%) compared to negative nodes (0.6%). Survival and AR rates of single macrometastases were not different from those of ITC or pN1mi. In case of 2 macrometastases or more, survival rates decreased and recurrence rates increased significantly. Micrometastases and ITC do not have a negative prognostic value. Single macrometastases might have an intermediate prognostic value while 2 macrometastases or more are associated with poorer prognosis.

Initial clinical trials demonstrated that lenalidomide monotherapy has a significant activity against some subtypes of lymphoma, but in diffuse large B-cell lymphoma (DLBCL) its activity is limited. The combination of lenalidomide with rituximab may be a promising therapeutic strategy. We retrospectively analyzed clinical outcomes in 17 patients with relapsed/refractory (R/R) DLBCL treated with lenalidomide, 25 mg/day for 21/28 days and rituximab, 375 mg/m(2) on day 7 of every lenalidomide cycle, for a maximum of 12 months. The overall response rate (ORR) was 41.2% with 35.3% complete response (CR), while median response duration was 26.5 months at a median follow-up of 24.9 months. Two patients with CR relapsed after 4 and 27 months of CR, and another four are actually in CR at + 13, + 23, + 24 and + 29 months. The estimated 24-month overall survival (OS) was 45% and progression-free survival (PFS) was 38%. Adverse events were manageable and mostly included thrombocytopenia and neutropenia. Lenalidomide-rituximab is active in R/R DLBCL with an important percentage of continuous CR.

7039 Background: In patients lacking HLA identical (HLA-id) donors, haploidentical (Haplo) donors is increasingly considered as a valid alternative stem cell source.The introduction of post-infusion Cyclophosphamide (Cy) as prophylaxis for immunological complications using T-cell replete graft is one of the most usual strategy. However, efficacy is still a matter of debate. Here we reviewed the outcome of 28 consecutives Haplo-HSCT patients with advanced NHL that we compared to an historical control cohort of patients transplanted from HLA-id donors. Methods: 28 Haplo-HSCT were performed between 04/06/2010 and 27/09/2013. Diagnoses were aggressive NHL in 22 patients and low-grade NHL in 6. Median age was 48 years (19-67). At transplant, 64% were in complete remission, 29% in partial remission and 7% in stable disease. These 28 patients were fully matched to 61 HLA-id patients. Conditioning regimen varied according to donor source. Haplo-HSCT patients received a NMAC associating of Fludarabine (150mg/m² total dose over 5 days), pre-infusion Cy (29 mg/kg total dose over two days), Low dose TBI 2 gray and post infusion Cy (100mg/kg total dose over 2 days). The HLA-id cohort reduced intensity conditionning consisted in Fludarabine (150mg/m² total dose over 5 days), intravenous Busulfan (6.4 mg/kg total dose over 2 days) and antithymocyte globulin (Thymoglobulin; 5 mg/kg total dose over 2 days). Results: With a median follow up of 15 months (2-70), the 2-year PFS and OS were 67% and 62% respectively in the Haplo-HSCT group. Incidence of day-100 acute grade 3-4 GVHD was 7%. The 24-months estimated incidence of chronic GVHD (cGVHD) and extensive cGVHD, was 7% and 4%. The 2-year NRM was 18%. The 2-year cumulative incidence of relapse was 20%. There was no statistical difference found when compared to the HLA-id group of patients for all outcomes. Conclusions: These results suggest that NMAC HSCT from Haploidentical donors achieves promising anti-tumor effect with an acceptable NRM and a low incidence of GVHD. Haplo-HSCT should represent a valuable alternative for patients with advanced NHL without available HLA-id donors.

Background: Toxicity, which is a key parameter in the evaluation of cancer treatments, can be underestimated by clinicians. We investigated differences between patients and physicians in reporting adverse events of androgen deprivation therapy (ADT) with or without docetaxel in a multicentre phase III trial in non-castrate metastatic prostate cancer.Methods: The 385 patients included were invited to complete a 26-symptom questionnaire 3 and 6 months after the start of treatment, among which eighteen symptoms were also assessed by physicians, reported in medical records and graded using the Common Toxicity Criteria of the National Cancer Institute. Positive and negative agreements as well as Kappa concordance coefficients were computed.Findings: Data were available for 220 and 165 patients at 3 and 6 months respectively. Physicians systematically under-reported patients' symptoms. Positive agreement rates (at respectively 3 and 6 months) for the five most commonly reported symptoms were: 61.0% and 64.3% hot flushes, 50.0% and 43.6% fatigue, 29.4% and 31.1% sexual dysfunction, 24.4% and 14.4% weigh gain/loss, 16.7% and 19.3% for joint/muscle pain. For symptoms most frequently reported as disturbing or very disturbing by patients, the clinicians' failure to report them ranged from 50.8% (hot flushes) to 89.5% (joint/muscle pain) at 3 months, and from 48.2% (hot flushes) to 88.4% (joint/muscle pain) at 6 months.Interpretation: Physicians often failed to report treatment-related symptoms, even the most common and disturbing ones. Patients' self-evaluation of toxicity should be used in clinical trials to improve the process of drug assessment in oncology.Funding: French Health Ministry and Institut National du Cancer (PHRC), Sanofi-Aventis, Astra-Zeneca, and Amgen. (C) 2013 Elsevier Ltd. All rights reserved.

510 Background: Metastatic renal cell carcinoma (RCC) is still associated with a poor prognosis with an overall survival (OS) of 37 months, although some patients are long survivors despite a disseminated disease. The purpose of this study was to determine whether the presence of at least one glandular metastasis (GM), defined by metastasis in pancreas, thyroid, breast or adrenal contralateral, affect survival outcome in patients with metastatic clear cell RCC (cc-RCC). Methods: In this retrospective analysis, we included patients treated for GM and non GM (NGM) metastatic RCC at the Institut Paoli Calmettes (IPC), Marseille, France between January 2004 and May 2013. Two more centers provided data for GM cc-RCC. GM were authentified using CTscan and/or by histopathological analysis obtained at biopsy or at surgery. Patients were analyzed from the time of metastatic disease to last follow up or death. The purpose of this study was to analyze patients outcomes based on the presence or absence of GM. We analysed survival using a log rank test from the time of the diagnosis of GM to the death or last follow-up visit. Results: Among 180 patients analysed, 59 had GM. Median age 65.2 vs 61.8 years old, male 55% vs 72%, 90% vs 78% prior nephrectomy in GM and NGM group. GM sites were pancreas for 45.5%, adrenal 44.5%, thyroid: 10%, breast10%. GM was the only site of metastasis for 10% of pts. For GM group, 55.5% had histopathology proven RCC metastasis. The median follow up was 72 months and the median survival is 114 months (IC95% [62.9-NR.]) in GM group vs 33.2 months (IC95% [23-41.9]; p<0.0001, HR 2.7 [1.7-4.3]) in non GM group. Conclusions: This retrospective study in a rare metastatic localization finds a longer survival in patients with GM. Despite some obvious limitations, this study questions GM as a surrogate for a better prognosis in RCC. Prospective follow-up and the collection for biological and genomic markers are needed to confirm and extend our present findings.

BACKGROUND:Early chemotherapy might improve the overall outcomes of patients with metastatic non-castrate (ie, hormone-sensitive) prostate cancer. We investigated the effects of the addition of docetaxel to androgen-deprivation therapy (ADT) for patients with metastatic non-castrate prostate cancer. METHODS:In this randomised, open-label, phase 3 study, we enrolled patients in 29 centres in France and one in Belgium. Eligible patients were older than 18 years and had histologically confirmed adenocarcinoma of the prostate and radiologically proven metastatic disease; a Karnofsky score of at least 70%; a life expectancy of at least 3 months; and adequate hepatic, haematological, and renal function. They were randomly assigned to receive to ADT (orchiectomy or luteinising hormone-releasing hormone agonists, alone or combined with non-steroidal antiandrogens) alone or in combination with docetaxel (75 mg/m(2) intravenously on the first day of each 21-day cycle; up to nine cycles). Patients were randomised in a 1:1 ratio, with dynamic minimisation to minimise imbalances in previous systemic treatment with ADT, chemotherapy for local disease or isolated rising concentration of serum prostate-specific antigen, and Glass risk groups. Patients, physicians, and data analysts were not masked to treatment allocation. The primary endpoint was overall survival. Efficacy analyses were done by intention to treat. This trial is registered with ClinicalTrials.gov, number NCT00104715. FINDINGS:Between Oct 18, 2004, and Dec 31, 2008, 192 patients were randomly allocated to receive ADT plus docetaxel and 193 to receive ADT alone. Median follow-up was 50 months (IQR 39-63). Median overall survival was 58·9 months (95% CI 50·8-69·1) in the group given ADT plus docetaxel and 54·2 months (42·2-not reached) in that given ADT alone (hazard ratio 1·01, 95% CI 0·75-1·36). 72 serious adverse events were reported in the group given ADT plus docetaxel, of which the most frequent were neutropenia (40 [21%]), febrile neutropenia (six [3%]), abnormal liver function tests (three [2%]), and neutropenia with infection (two [1%]). Four treatment-related deaths occurred in the ADT plus docetaxel group (two of which were neutropenia-related), after which the data monitoring committee recommended treatment with granulocyte colony-stimulating factor. After this recommendation, no further treatment-related deaths occurred. No serious adverse events were reported in the ADT alone group. INTERPRETATION:Docetaxel should not be used as part of first-line treatment for patients with non-castrate metastatic prostate cancer. FUNDING:French Health Ministry and Institut National du Cancer (PHRC), Sanofi-Aventis, AstraZeneca, and Amgen.

BACKGROUND:A subgroup of T1N0M0 breast cancer (BC) carries a high potential of relapse, and thus may require adjuvant systemic therapy (AST).PATIENTS AND METHODS:Retrospective analysis of all patients with T1 BC, who underwent surgery from January 1999 to December 2009 at 13 French sites. AST was not standardized.RESULTS:Among 8100 women operated, 5423 had T1 tumors (708 T1a, 2208 T1b and 2508 T1c 11-15 mm). T1a differed significantly from T1b tumors with respect to several parameters (lower age, more frequent negative hormonal status and positive HER2 status, less frequent lymphovascular invasion), exhibiting a mix of favorable and poor prognosis factors. Overall survival was not different between T1a, b or c tumors but recurrence-free survival was significantly higher in T1b than in T1a tumors (P = 0.001). In multivariate analysis, tumor grade, hormone therapy and lymphovascular invasion were independent prognostic factors.CONCLUSION:Relatively poor outcome of patients with T1a tumors might be explained by a high frequency of risk factors in this subgroup (frequent negative hormone receptors and HER2 overexpression) and by a less frequent administration of AST (endocrine treatment and chemotherapy). Tumor size might not be the main determinant of prognosis in T1 BC.

Background With the improvement of cancer therapy, long term survivors are more exposed to the risk of secondary myeloid neoplasm including myelodysplasia and acute leukemias. The WHO 2008 classification individualize the therapy related acute myeloid leukemia (AML) as a specific entity and highlight the role of chemotherapy/radiotherapy in the pathogenesis of the disease. There is a demonstrated link between specific therapeutic agents, recurrent genetic lesions (such as t(15;17), CBF AML, monosomy 7, …), and outcome. However, most of the published series include multiple types of primary cancer and treatments. This heterogeneity may represent a problem as only limited data are available in patients with specific cancer subtypes, such as the most frequent one, breast cancer. This may be important at a public health level but also to homogenize age and types of prior treatment. Moreover, the complexity of t-AML could not be resumed to the association between treatment and karyotype, as some patients developed AML without chemotherapy, suggesting a potential predisposition to AML. In the present report, we focused on AML arising after breast cancer (BC) and describe the characteristics and outcome of this population. Patients and Methods This is a retrospective multicenter study. Patients were included if they had the diagnosis of breast cancer preceding diagnosis of acute myeloid leukemia whatever may be the interval between the 2 cancers and whatever treatment was administered for BC. All patients with AML were treated with induction chemotherapy. Clinical and biological data for both cancers were collected. Patient’s characteristics and results were compared with age, cytogenetic risk, and optionally sex matched (if possible) de novo AML with a 2/1 ratio. Results 408 patients were analyzed, including 136 AML associated with BC and 272 de novo AML. The median age at diagnosis of BC was 50 years. 47% of patients had invasive ductal carcinoma. Treatment of BC included chemotherapy in 81% of cases, radiotherapy in 91% of cases, and surgery in 99% of cases. The median time between BC and AML was 2.8 years. Median WBC was 3.4G/l and median platelets count was 49G/l. For AML-BC, Cytogenetics were abnormal in 82% of cases including 12% Complex Karyotype, 12% t(15;17), 17% CBF, and 20% MLL translocations. The recent introduction of taxanes in the treatment of BC did not seem to change the frequencies of these aberrations. With the exception of sex ratio, there was no significant difference of baseline characteristics as compared to control group. Regarding induction chemotherapy, CR rate was 81% and 8-week mortality 11.6%. In the control cohort, CR rate was 83% (p=NS) and 8 week mortality 7% (p=NS). Allogeneic transplantation was performed in 21% and 17% of patients respectively. In cytogenetic adjusted survival analysis, median overall survival and relapse free survival were similar between AML-BC and de novo AML for favorable risk (OS and RFS not reached for both groups; p=0.06 and p=0.3 respectively), and unfavorable risk (OS 12m vs. 13m p=NS, RFS 8m vs. 9m p=NS) groups. Interestingly, there was a difference for intermediate cytogenetics group, with median OS (21 months vs. 38 months p=0.01) and median RFS (14 vs . 25 months p=0.04). Difference was also confirmed for cumulative incidence of relapse (1 year probability 41% vs. 22%, p=0.04). Frequency of FLT3, NPM1, and CEBPA mutations were only available in a subset of patients. Only 1/12 pts in the intermediate AML-BC group was FLT3 mutated and 1/12 pt had isolated NPM1 mutation. In the de novo cohort, 11/28 pts had FLT3 mutations and 8/21 pts had isolated NPM1 mutation. Conclusion Our data showed that in AML arising after BC, the prior BC do not appear to impact the outcome in favorable and unfavorable cytogenetic risk groups. However, this is different for intermediate risk cytogenetics, our data suggesting a poorer outcome of AML-BC and potentially a different mutational profile. Regarding AML susceptibility, a matched pair analysis comparing the AML-BC and BC without AML will also be presented at the meeting. Disclosures: No relevant conflicts of interest to declare.

26 Background: Patients with upfront metastases at the time of prostate cancer (PC) diagnosis are less frequent than in the past in Western countries, but still represent 5-10% of all patients and almost one half of PC patients will eventually die of the disease. Prognostic factors (lymph node metastases/appendicular vs axial bone disease, performance status > 1, Gleason score > 8 and PSA > 65 ng/ml) have been proposed (Glass et al., 2003), leading to the definition of three subgroups with good, intermediate and poor prognosis. However, the current natural history of metastatic prostate cancer has not been well described. Methods: Patients with hormone sensitive metastatic PC were randomized to receive continuous androgen deprivation therapy (ADT) plus docetaxel (75 mg/m²/21d up to 9 cycles) and prednisone or ADT alone. Glass risk groups were used as stratification factors. Results: From October 2004 to December 2008, 385 pts were included. They were distributed into good (50%), intermediate (29%), and poor (21%) prognosis groups. The median follow up was 50 months [95% CI: 49 - 54]. The primary endpoint analysis showed no difference in overall survival (OS) (HR: 1.01 [95%CI: 0.75-1.36]) between the 2 arms but a significant PFS improvement was observed in the docetaxel group (HR: 0.75 [0.59-0.94] p=0.0147) (Gravis, ESMO 2012). The median OS in the ADT alone arm was 54 months [42-NR]. It was 69 [95% CI: 60.9-NR], 47 [95% CI: 37.7-NR, HR = 1.6] and 37 [95% CI: 28.5-58.9, HR = 2.12] months respectively in the good, intermediate, and poor prognosis groups (p=0.001) in the whole cohort. No interaction between prognosis groups and treatment was found. A detailed analysis using the Cox model will be presented. Conclusions: At the present time, median life expectancy of patients with metastatic PC seems to exceed 4 years. Subgroups with favorable or unfavorable outcome can be identified. Clinical trial information: NCT00055731.

BACKGROUND CONTEXT:The incidence of metastatic spinal cord compression (MSCC) is increasing, paralleling increasing life expectancy of patients. However, management of MSCC and relevance of scoring systems remain controversial. PURPOSE:The aims of our study were to analyze the feasibility and outcomes of spinal surgery, to identify prognostic factors for survival, and to assess the accuracy of scoring systems in patients with malignancies associated with MSCC. STUDY DESIGN:Retrospective analysis of all patients with MSCC operated in our institution. METHODS:Outcomes of surgery, prognostic factors for survival, and relevance of Tomita and Tokuhashi scores were investigated. RESULTS:One hundred forty-eight patients were included: 66% were hyperalgic (pain score >6) and Frankel score (FS) was decreased in 49%. Seventy-three percent of patients had laminectomy with spinal fixation. After surgery, pain decreased in 75% of cases, FS was improved in 31%, and 92% of patients were ambulatory. Postoperative complication rate was 16%. Median overall survival (OS) was 8.9 months (95% confidence interval, 4.4-13). Only Tokuhashi score was relevant, but predictive accuracy of survival was just 51%. In univariate analyses, hyperalgia (p=.001), primary tumor site, extrabone metastases (p<.001), Karnofsky performance status (KPS) less than 70 (p<.001), poor American Society of Anesthesiologist (ASA) score (p<.001) or FS (p=.01), and absence of postoperative chemotherapy (p<.001) were associated with shorter OS. In multivariate analysis, only extrabone metastases (p=.004), KPS (p=.001), and ASA score (p=.007) remained significantly associated with OS. CONCLUSIONS:Surgery for MSCC is associated with limited morbidity, improved autonomy, and pain relief. Usual scores do not seem relevant, whereas ASA score, KPS, and extrabone metastases are significantly associated with OS.