Nonparametric estimation of the survival function for ordered multivariate failure time data: a comparative study

In longitudinal studies of disease, patients may experience several events through a follow-up period. In these studies, the sequentially ordered events are often of interest and lead to problems that have received much attention recently. Issues of interest include the estimation of bivariate survival, marginal distributions and the conditional distribution of gap times. In this work we consider the estimation of the survival function conditional to a previous event. Different nonparametric approaches will be considered for estimating these quantities, all based on the Kaplan-Meier estimator of the survival function. We explore the finite sample behavior of the estimators through simulations. The different methods proposed in this article are applied to a data set from a German Breast Cancer Study. The methods are used to obtain predictors for the conditional survival probabilities as well as to study the influence of recurrence in overall survival.

Estimação da função de sobrevivência em dados de eventos recorrentes com censura pela direita

Dissertação de mestrado em Estatística

Natural history of surgically treated high-risk prostate cancer

BACKGROUND No data exist on the patterns of biochemical recurrence (BCR) and their effect on survival in patients with high-risk prostate cancer (PCa) treated with surgery. The aim of our investigation was to evaluate the natural history of PCa in patients treated with radical prostatectomy (RP) alone. MATERIALS AND METHODS Overall, 2,065 patients with high-risk PCa treated with RP at 7 tertiary referral centers between 1991 and 2011 were identified. First, we calculated the probability of experiencing BCR after surgery. Particularly, we relied on conditional survival estimates for BCR after RP. Competing-risks regression analyses were then used to evaluate the effect of time to BCR on the risk of cancer-specific mortality (CSM). RESULTS Median follow-up was 70 months. Overall, the 5-year BCR-free survival rate was 55.2%. Given the BCR-free survivorship at 1, 2, 3, 4, and 5 years, the BCR-free survival rates improved by+7.6%,+4.1%,+4.8%,+3.2%, and+3.7%, respectively. Overall, the 10-year CSM rate was 14.8%. When patients were stratified according to time to BCR, patients experiencing BCR within 36 months from surgery had higher 10-year CSM rates compared with those experiencing late BCR (19.1% vs. 4.4%; P<0.001). At multivariate analyses, time to BCR represented an independent predictor of CSM (P<0.001). CONCLUSIONS Increasing time from surgery is associated with a reduction of the risk of subsequent BCR. Additionally, time to BCR represents a predictor of CSM in these patients. These results might help provide clinicians with better follow-up strategies and more aggressive treatments for early BCR.

Future long-horizon performance measurement conditional on past survival

This paper examines the measurement of long-horizon abnormal performance when stock selection is conditional on an extended period of past survival. Filtering on survival results in a sample driven towards more-established, frequently traded stocks and this has implications for the choice of benchmark used in performance measurement (especially in the presence of the well-documented size effect). A simulation study is conducted to document the properties of commonly employed performance measures conditional on past survival. The results suggest that the popular index benchmarks used in long-horizon event studies are severely biased and yield test statistics that are badly misspecified. In contrast, a matched-stock benchmark based on size and industry performs consistently well. Also, an eligible-stock index designed to mitigate the influence of the size effect proves effective.

Survival Analysis Employed in Predicting Corporate Failure: A Forecasting Model Proposal

In face of the current economic and financial environment, predicting corporate bankruptcy is arguably a phenomenon of increasing interest to investors, creditors, borrowing firms, and governments alike. Within the strand of literature focused on bankruptcy forecasting we can find diverse types of research employing a wide variety of techniques, but only a few researchers have used survival analysis for the examination of this issue. We propose a model for the prediction of corporate bankruptcy based on survival analysis, a technique which stands on its own merits. In this research, the hazard rate is the probability of ‘‘bankruptcy’’ as of time t, conditional upon having survived until time t. Many hazard models are applied in a context where the running of time naturally affects the hazard rate. The model employed in this paper uses the time of survival or the hazard risk as dependent variable, considering the unsuccessful companies as censured observations.

Nonparametric estimation of time-dependent ROC curves conditional on a continuous covariate

The receiver-operating characteristic (ROC) curve is the most widely used measure for evaluating the performance of a diagnostic biomarker when predicting a binary disease outcome. The ROC curve displays the true positive rate (or sensitivity) and the false positive rate (or 1-specificity) for different cut-off values used to classify an individual as healthy or diseased. In time-to-event studies, however, the disease status (e.g. death or alive) of an individual is not a fixed characteristic, and it varies along the study. In such cases, when evaluating the performance of the biomarker, several issues should be taken into account: first, the time-dependent nature of the disease status; and second, the presence of incomplete data (e.g. censored data typically present in survival studies). Accordingly, to assess the discrimination power of continuous biomarkers for time-dependent disease outcomes, time-dependent extensions of true positive rate, false positive rate, and ROC curve have been recently proposed. In this work, we present new nonparametric estimators of the cumulative/dynamic time-dependent ROC curve that allow accounting for the possible modifying effect of current or past covariate measures on the discriminatory power of the biomarker. The proposed estimators can accommodate right-censored data, as well as covariate-dependent censoring. The behavior of the estimators proposed in this study will be explored through simulations and illustrated using data from a cohort of patients who suffered from acute coronary syndrome.

In vivo conditional Pax4 overexpression in mature islet β-cells prevents stress-induced hyperglycemia in mice.

OBJECTIVE To establish the role of the transcription factor Pax4 in pancreatic islet expansion and survival in response to physiological stress and its impact on glucose metabolism, we generated transgenic mice conditionally and selectively overexpressing Pax4 or a diabetes-linked mutant variant (Pax4R129 W) in β-cells. RESEARCH DESIGN AND METHODS Glucose homeostasis and β-cell death and proliferation were assessed in Pax4- or Pax4R129 W-overexpressing transgenic animals challenged with or without streptozotocin. Isolated transgenic islets were also exposed to cytokines, and apoptosis was evaluated by DNA fragmentation or cytochrome C release. The expression profiles of proliferation and apoptotic genes and β-cell markers were studied by immunohistochemistry and quantitative RT-PCR. RESULTS Pax4 but not Pax4R129 W protected animals against streptozotocin-induced hyperglycemia and isolated islets from cytokine-mediated β-cell apoptosis. Cytochrome C release was abrogated in Pax4 islets treated with cytokines. Interleukin-1β transcript levels were suppressed in Pax4 islets, whereas they were increased along with NOS2 in Pax4R129 W islets. Bcl-2, Cdk4, and c-myc expression levels were increased in Pax4 islets while MafA, insulin, and GLUT2 transcript levels were suppressed in both animal models. Long-term Pax4 expression promoted proliferation of a Pdx1-positive cell subpopulation while impeding insulin secretion. Suppression of Pax4 rescued this defect with a concomitant increase in pancreatic insulin content. CONCLUSIONS Pax4 protects adult islets from stress-induced apoptosis by suppressing selective nuclear factor-κB target genes while increasing Bcl-2 levels. Furthermore, it promotes dedifferentiation and proliferation of β-cells through MafA repression, with a concomitant increase in Cdk4 and c-myc expression.

Destruction of conditional insulinoma cell lines in NOD mice: a role for autoimmunity.

AIMS/HYPOTHESIS: betaTC-tet (H2(k)) is a conditional insulinoma cell line derived from transgenic mice expressing a tetracycline-regulated oncogene. Transgenic expression of several proteins implicated in the apoptotic pathways increase the resistance of betaTC-tet cells in vitro. We tested in vivo the sensitivity of the cells to rejection and the protective effect of genetic alterations in NOD mice. METHODS: betaTC-tet cells and genetically engineered lines expressing Bcl-2 (CDM3D), a dominant negative mutant of MyD88 or SOCS-1 were transplanted in diabetic female NOD mice or in male NOD mice with diabetes induced by high-dose streptozotocin. Survival of functional cell grafts in NOD-scid mice was also analyzed after transfer of splenocytes from diabetic NOD mice. Autoreactive T-cell hybridomas and splenocytes from diabetic NOD mice were stimulated by betaTC-tet cells. RESULTS: betaTC-tet cells and genetically engineered cell lines were all similarly rejected in diabetic NOD mice and in NOD-scid mice after splenocyte transfer. In 3- to 6-week-old male NOD mice treated with high-dose streptozotocin, the cells temporarily survived, in contrast with C57BL/6 mice treated with high-dose streptozotocin (indefinite survival) and untreated 3- to 6-week-old male NOD mice (rejection). The protective effect of high-dose streptozotocin was lost in older male NOD mice. betaTC-tet cells did not stimulate autoreactive T-cell hybridomas, but induced IL-2 secretion by splenocytes from diabetic NOD mice. CONCLUSION/INTERPRETATION: The autoimmune process seems to play an important role in the destruction of betaTC-tet cells in NOD mice. Genetic manipulations intended at increasing the resistance of beta cells were inefficient. Similar approaches should be tested in vivo as well as in vitro. High dose streptozotocin influences immune rejection and should be used with caution.

Conditional deletion of ferritin h in mice reduces B and T lymphocyte populations.

The immune system and iron availability are intimately linked as appropriate iron supply is needed for cell proliferation, while excess iron, as observed in hemochromatosis, may reduce subsets of lymphocytes. We have tested the effects of a ferritin H gene deletion on lymphocytes. Mx-Cre mediated conditional deletion of ferritin H in bone marrow reduced the number of mature B cells and peripheral T cells in all lymphoid organs. FACS analysis showed an increase in the labile iron pool, enhanced reactive oxygen species formation and mitochondrial depolarization. The findings were confirmed by a B-cell specific deletion using Fth(lox/lox) ; CD19-Cre mice. Mature B cells were strongly under-represented in bone marrow and spleen of the deleted mice, whereas pre-B and immature B cells were not affected. Bone marrow B cells showed increased proliferation as judged by the number of cells in S and G2/M phase as well as BrdU incorporation. Upon in vitro culture with B-cell activating factor of the tumor necrosis factor family (BAFF), ferritin H-deleted spleen B cells showed lower survival rates than wild type cells. This was partially reversed with iron-chelator deferiprone. The loss of T cells was also confirmed by a T cell-specific deletion in Fth(lox/lox) ;CD4-Cre mice. Our data show that ferritin H is required for B and T cell survival by actively reducing the labile iron pool. They further suggest that natural B and T cell maturation is influenced by intracellular iron levels and possibly deregulated in iron excess or deprivation.

Neural crest cell survival is dependent on Rho kinase and is required for development of the mid face in mouse embryos.

Neural crest cells (NCC) give rise to much of the tissue that forms the vertebrate head and face, including cartilage and bone, cranial ganglia and teeth. In this study we show that conditional expression of a dominant-negative (DN) form of Rho kinase (Rock) in mouse NCC results in severe hypoplasia of the frontonasal processes and first pharyngeal arch, ultimately resulting in reduction of the maxilla and nasal bones and severe craniofacial clefting affecting the nose, palate and lip. These defects resemble frontonasal dysplasia in humans. Disruption of the actin cytoskeleton, which leads to abnormalities in cell-matrix attachment, is seen in the RockDN;Wnt1-cre mutant embryos. This leads to elevated cell death, resulting in NCC deficiency and hypoplastic NCC-derived craniofacial structures. Rock is thus essential for survival of NCC that form the craniofacial region. We propose that reduced NCC numbers in the frontonasal processes and first pharyngeal arch, resulting from exacerbated cell death, may be the common mechanism underlying frontonasal dysplasia.

Intravitreal chemotherapy for vitreous disease in retinoblastoma revisited: from prohibition to conditional indications.

BACKGROUND: Tumour control of vitreous seeds remains challenging owing to their resistance to radiation and systemic chemotherapy. OBJECTIVE: To describe the short-term efficacy of intravitreal melphalan for vitreous disease in retinoblastoma using a new injection technique and dose. METHODS: This study is a retrospective non-comparative review of 23 consecutive heavily pretreated patients (23 eyes) with active vitreous seeding and eligible for intravitreous chemotherapy (IViC). They received a total of 122 intravitreal injections of melphalan (20-30 μg) given every 7-10 days. The ocular status was objectively monitored under anaesthesia with fundus photography. RESULTS: All patients are alive without evidence of extraocular spread (95% CI 82.19% to 100%). Concomitant treatments, including other chemotherapeutic modalities, were used until complete sterilisation of the retinal seeding source and subretinal seeds. Globe retention was achieved in 87% (20/23) of cases. All retained eyes were in complete remission after a median follow-up period of 22 months (range 9-31 months). The Kaplan-Meier estimate of ocular survival rates at 2 years was 84.14% (95% CI 62.48% to 95.28%). A localised peripheral salt-and-pepper retinopathy was noted in 10 eyes (43%) at the site of injection. CONCLUSIONS: This study reports the first clinically documented case series of patients with retinoblastoma treated with IViC. Despite a possible confounding effect of concomitant chemotherapy prescription using other routes of administration in four of the successfully treated eyes (20%), IViC achieved an unprecedented success rate of tumour control in the presence of vitreous seeding. Of note, none of the treated eyes required external beam irradiation to control the vitreous seeding. Further studies are required to assess IViC retinal toxicity and to better delineate its role in the management of retinoblastoma.

TAK1 is required for survival of mouse fibroblasts treated with TRAIL, and does so by NF-kappaB dependent induction of cFLIPL.

Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) is known as a "death ligand"-a member of the TNF superfamily that binds to receptors bearing death domains. As well as causing apoptosis of certain types of tumor cells, TRAIL can activate both NF-kappaB and JNK signalling pathways. To determine the role of TGF-beta-Activated Kinase-1 (TAK1) in TRAIL signalling, we analyzed the effects of adding TRAIL to mouse embryonic fibroblasts (MEFs) derived from TAK1 conditional knockout mice. TAK1-/- MEFs were significantly more sensitive to killing by TRAIL than wild-type MEFs, and failed to activate NF-kappaB or JNK. Overexpression of IKK2-EE, a constitutive activator of NF-kappaB, protected TAK1-/- MEFs against TRAIL killing, suggesting that TAK1 activation of NF-kappaB is critical for the viability of cells treated with TRAIL. Consistent with this model, TRAIL failed to induce the survival genes cIAP2 and cFlipL in the absence of TAK1, whereas activation of NF-kappaB by IKK2-EE restored the levels of both proteins. Moreover, ectopic expression of cFlipL, but not cIAP2, in TAK1-/- MEFs strongly inhibited TRAIL-induced cell death. These results indicate that cells that survive TRAIL treatment may do so by activation of a TAK1-NF-kappaB pathway that drives expression of cFlipL, and suggest that TAK1 may be a good target for overcoming TRAIL resistance.

Hematopoietic stem cell function and survival depend on c-Myc and N-Myc activity.

Myc activity is emerging as a key element in acquisition and maintenance of stem cell properties. We have previously shown that c-Myc deficiency results in accumulation of defective hematopoietic stem cells (HSCs) due to niche-dependent differentiation defects. Here we report that immature HSCs coexpress c-myc and N-myc mRNA at similar levels. Although conditional deletion of N-myc in the bone marrow does not affect hematopoiesis, combined deficiency of c-Myc and N-Myc (dKO) results in pancytopenia and rapid lethality. Interestingly, proliferation of HSCs depends on both myc genes during homeostasis, but is c-Myc/N-Myc independent during bone marrow repair after injury. Strikingly, while most dKO hematopoietic cells undergo apoptosis, only self-renewing HSCs accumulate the cytotoxic molecule Granzyme B, normally employed by the innate immune system, thereby revealing an unexpected mechanism of stem cell apoptosis. Collectively, Myc activity (c-Myc and N-Myc) controls crucial aspects of HSC function including proliferation, differentiation, and survival.

Empirical study of correlated survival times for recurrent events with proportional hazards margins and the effect of correlation and censoring.

Background: In longitudinal studies where subjects experience recurrent incidents over a period of time, such as respiratory infections, fever or diarrhea, statistical methods are required to take into account the within-subject correlation. Methods: For repeated events data with censored failure, the independent increment (AG), marginal (WLW) and conditional (PWP) models are three multiple failure models that generalize Cox"s proportional hazard model. In this paper, we revise the efficiency, accuracy and robustness of all three models under simulated scenarios with varying degrees of within-subject correlation, censoring levels, maximum number of possible recurrences and sample size. We also study the methods performance on a real dataset from a cohort study with bronchial obstruction. Results: We find substantial differences between methods and there is not an optimal method. AG and PWP seem to be preferable to WLW for low correlation levels but the situation reverts for high correlations. Conclusions: All methods are stable in front of censoring, worsen with increasing recurrence levels and share a bias problem which, among other consequences, makes asymptotic normal confidence intervals not fully reliable, although they are well developed theoretically.