A partitioning deletion/substitution/addition algorithm for creating survival risk groups.

Accurately assessing a patient's risk of a given event is essential in making informed treatment decisions. One approach is to stratify patients into two or more distinct risk groups with respect to a specific outcome using both clinical and demographic variables. Outcomes may be categorical or continuous in nature; important examples in cancer studies might include level of toxicity or time to recurrence. Recursive partitioning methods are ideal for building such risk groups. Two such methods are Classification and Regression Trees (CART) and a more recent competitor known as the partitioning Deletion/Substitution/Addition (partDSA) algorithm, both of which also utilize loss functions (e.g., squared error for a continuous outcome) as the basis for building, selecting, and assessing predictors but differ in the manner by which regression trees are constructed. Recently, we have shown that partDSA often outperforms CART in so-called "full data" settings (e.g., uncensored outcomes). However, when confronted with censored outcome data, the loss functions used by both procedures must be modified. There have been several attempts to adapt CART for right-censored data. This article describes two such extensions for partDSA that make use of observed data loss functions constructed using inverse probability of censoring weights. Such loss functions are consistent estimates of their uncensored counterparts provided that the corresponding censoring model is correctly specified. The relative performance of these new methods is evaluated via simulation studies and illustrated through an analysis of clinical trial data on brain cancer patients. The implementation of partDSA for uncensored and right-censored outcomes is publicly available in the R package, partDSA.

Lifetime history of indoor tanning in young people: a retrospective assessment of initiation, persistence, and correlates.

BACKGROUND: Despite educational and public health campaigns to convey the risks of indoor tanning, many individuals around the world continue to engage in this behavior. Few descriptive studies of indoor tanning have collected information pertaining to the lifetime history of indoor tanning, thereby limiting our ability to understand indoor tanning patterns and potentially target interventions for individuals who not only initiate, but continue to persistently engage in indoor tanning. METHODS: In-person interviews elicited detailed retrospective information on lifetime history of indoor tanning among white individuals (n = 401) under age 40 seen by a dermatologist for a minor benign skin condition. These individuals were controls in a case-control study of early-onset basal cell carcinoma. Outcomes of interest included ever indoor tanning in both males and females, as well as persistent indoor tanning in females - defined as females over age 31 who tanned indoors at least once in the last three or all four of four specified age periods (ages 11-15, 16-20, 21-30 and 31 or older). Multivariate logistic regression was used to identify sociodemographic and lifestyle correlates of ever and persistent indoor tanning in females. RESULTS: Approximately three-quarters (73.3%) of females and 38.3% of males ever tanned indoors, with a median age of initiation of 17.0 and 21.5, respectively. Among indoor tanners, 39.3% of females and 21.7% of males reported being burned while indoor tanning. Female ever indoor tanners were younger, had darker color eyes, and sunbathed more frequently than females who never tanned indoors. Using unique lifetime exposure data, 24.7% of female indoor tanners 31 and older persistently tanned indoors starting as teenagers. Female persistent indoor tanners drank significantly more alcohol, were less educated, had skin that tanned with prolonged sun exposure, and sunbathed outdoors more frequently than non-persistent tanners. CONCLUSIONS: Indoor tanning was strikingly common in this population, especially among females. Persistent indoor tanners had other high-risk behaviors (alcohol, sunbathing), suggesting that multi-faceted behavioral interventions aimed at health promotion/disease prevention may be needed in this population.

Evaluation of prognostic and predictive value of microtubule associated protein tau in two independent cohorts.

INTRODUCTION: Microtubule associated proteins (MAPs) endogenously regulate microtubule stabilization and have been reported as prognostic and predictive markers for taxane response. The microtubule stabilizer, MAP-tau, has shown conflicting results. We quantitatively assessed MAP-tau expression in two independent breast cancer cohorts to determine prognostic and predictive value of this biomarker. METHODS: MAP-tau expression was evaluated in the retrospective Yale University breast cancer cohort (n = 651) using tissue microarrays and also in the TAX 307 cohort, a clinical trial randomized for TAC versus FAC chemotherapy (n = 140), using conventional whole tissue sections. Expression was measured using the AQUA method for quantitative immunofluorescence. Scores were correlated with clinicopathologic variables, survival, and response to therapy. RESULTS: Assessment of the Yale cohort using Cox univariate analysis indicated an improved overall survival (OS) in tumors with a positive correlation between high MAP-tau expression and overall survival (OS) (HR = 0.691, 95% CI = 0.489-0.974; P = 0.004). Kaplan Meier analysis showed 10-year survival for 65% of patients with high MAP-tau expression compared to 52% with low expression (P = .006). In TAX 307, high expression was associated with significantly longer median time to tumor progression (TTP) regardless of treatment arm (33.0 versus 23.4 months, P = 0.010) with mean TTP of 31.2 months. Response rates did not differ by MAP-tau expression (P = 0.518) or by treatment arm (P = 0.584). CONCLUSIONS: Quantitative measurement of MAP-tau expression has prognostic value in both cohorts, with high expression associated with longer TTP and OS. Differences by treatment arm or response rate in low versus high MAP-tau groups were not observed, indicating that MAP-tau is not associated with response to taxanes and is not a useful predictive marker for taxane-based chemotherapy.

partDSA: deletion/substitution/addition algorithm for partitioning the covariate space in prediction.

MOTIVATION: Until now, much of the focus in cancer has been on biomarker discovery and generating lists of univariately significant genes, as well as epidemiological and clinical measures. These approaches, although significant on their own, are not effective for elucidating the synergistic qualities of the numerous components in complex diseases. These components do not act one at a time, but rather in concert with numerous others. A compelling need exists to develop analytically sound and computationally advanced methods that elucidate a more biologically meaningful understanding of the mechanisms of cancer initiation and progression by taking these interactions into account. RESULTS: We propose a novel algorithm, partDSA, for prediction when several variables jointly affect the outcome. In such settings, piecewise constant estimation provides an intuitive approach by elucidating interactions and correlation patterns in addition to main effects. As well as generating 'and' statements similar to previously described methods, partDSA explores and chooses the best among all possible 'or' statements. The immediate benefit of partDSA is the ability to build a parsimonious model with 'and' and 'or' conjunctions that account for the observed biological phenomena. Importantly, partDSA is capable of handling categorical and continuous explanatory variables and outcomes. We evaluate the effectiveness of partDSA in comparison to several adaptive algorithms in simulations; additionally, we perform several data analyses with publicly available data and introduce the implementation of partDSA as an R package. AVAILABILITY: http://cran.r-project.org/web/packages/partDSA/index.html CONTACT: annette.molinaro@yale.edu SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The effect of aging on the skeletal response to intermittent treatment with parathyroid hormone.

Little is known about the modifying effects of age on the skeletal response to intermittent treatment with PTH. We therefore compared the response of 63 aged (18 month old) and 61 young-adult (3 month old) C57BL/6 mice to 4 wk of daily sc injections of either vehicle or h(1-34)PTH at a dose of 95 ng/g body weight. The increase in total body bone mineral density (BMD), compared with vehicle-treated animals, was similar in aged and young-adult mice (+5.6 vs. +6.3%). Aged animals demonstrated a greater increase in spinal BMD than their younger counterparts (+12.0 vs. +5.1%, P = 0.01; absolute increment: 57 x 10(-4) vs. 28 x 10(-4) g/cm(2)). Microcomputed tomography analyses in a subset of the vertebrae showed a trend toward higher L5 trabecular bone volume fraction in the PTH-treated aged animals (+40.2 vs. +19.6%). Vertebral histomorphometry demonstrated a greater PTH-induced increase in osteoblast number in aged vs. young-adult animals (694 vs. 396 cells/mm(2)). In contrast, in the femur the PTH-induced increase in BMD tended to be greater in the young-adult than the aged animals, although this did not reach statistical significance (8.1 vs. 4.2%). The numbers of osteoblast progenitors and mineralizing colonies in cultured marrow were unaffected by PTH treatment in either group. We conclude that aging differentially impacts the regional skeletal response to PTH such that the increase in BMD in the spine is augmented, whereas that in the femur is unaffected. Effects on osteoblast progenitor recruitment do not seem to be the basis for these changes.