Atypical glandular cells of undetermined significance favor endometrial origin. Criteria for separating low grade endometrial adenocarcinoma from benign endometrial lesions

OBJECTIVE: To evaluate cytologic criteria for separating atypical glandular cells of undetermined significance favor endometrial origin (AGUS-EM), on Papanicolaou-stained (Pap) smears into favor benign and favor malignant categories. STUDY DESIGN: All patients who had a Pap smear diagnosis of AGUS-EM, not further qualified, followed by tissue follow-up were identified from the surgical pathology and cytopathology files from January 1992 through December 1996. The Pap smears were scored blindly for the presence or absence of 40 cytologic criteria, and univariate analysis was performed to determine which criteria were most indicative of malignancy by tissue follow-up. RESULTS: The presence of an atrophic smear, nuclear size greater than twice that of an intermediate cell nucleus and absence of clusters with irregular borders were highly indicative of adenocarcinoma, although other criteria were also helpful in identifying malignancy. CONCLUSION: There are no combinations of cytologic criteria that definitely separate AGUS-EM cases into those with benign or malignant findings on follow-up. However, some isolated criteria were useful in the differential diagnosis in a [table: see text] significant number of cases.     

Self-modeling for two-dimensional response curves

Two-dimensional response curves are an important experimental outcome in speech kinematics and other areas of research. These parameterized curves are usually obtained by recording the two-dimensional location of an object over time. In this setting, time is the independent variable and the x and y locations on specified coordinate axes define the multivariate response. Collections of such parameterized curves can be obtained either from one subject or from a number of different subjects, each producing one or several realizations of the response curve. When only one dependent variable is observed over time and no parametric model is specified, self-modeling regression (SEMOR) is an attractive modeling approach. SEMOR assumes that each of a collection of curves differs from a smooth, average shape function by some simple parametric transformation of the coordinate axes (usually linear). We will describe the extension of SEMOR to two-dimensional parameterized curves using affine transformations of a two-dimensional, time-parameterized shape function.     

Metachronous serous endometrial intraepithelial carcinoma and serous peritoneal carcinoma: analysis of probable independent lesions

Background

Uterine serous endometrial intraepithelial carcinoma (SEIC) is an immediate precursor of invasive carcinoma. The majority of stage IA SEICs are curable, but those with latent peritoneal metastasis and/or capillary lymphatics invasion may have poor prognoses Careful pathologic staging is thus needed to predict the risk of recurrence and to determine postoperative therapeutic strategies.

Case Presentation

A 71-year-old woman was hospitalized for the treatment of peritoneal carcinoma. She had undergone total hysterectomy and bilateral salpingo-oophorectomy due to SEIC (stage IA) at age 63 years, and had received medical check-ups every year since. Elevated serum CA125 (184 U/mL) was detected for the first time 8 years after surgery. A thorough workup revealed no potential primary lesion other than that in the peritoneum. Tumor reduction surgery was performed. Histologic analysis of the peritoneal lesion was high-grade serous carcinoma. The peritoneal carcinoma was diffusely immunostained for p53; thus, possible recurrence of SEIC was suspected. Tumor DNAs were microdissected from the uterine and peritoneal lesions and p53 mutation analysis was done. SEIC and peritoneal carcinomas had distinct p53 mutations that were mutually exclusive.

Conclusions

The present case raised a concern about the difficulty of histologic staging for SEICs. Although SEICs confined to the uterine endometrium in most cases predict a good prognosis, microscopic metastasis to the peritoneum may not be detectable at hysterectomy. If secondary malignancies of a serous phenotype develop years later, comprehensive reexamination of SEIC is mandated, with the help of DNA analysis.

     

Dose-response curves for radish seedling phototropism

Radish seedlings (Raphanus sativus L.) were grown for 4 days in complete darkness, or in white light, or for 3 days in darkness followed by 1 day of red light. Phototropic dose-response curves for the seedlings grown in these three ways were determined with 460-nm light. The dark-grown and red light-treated seedlings responded with positive curvatures of no more than 10° to energy doses in the first positive range and with larger positive curvatures in the second positive dose range. No indifferent or negative curvature was seen with the light intensity used. White light-grown seedlings did not respond to first positive energy doses, but responded as strongly to second positive doses as the other types of seedlings.     

Phase response curves for social entrainment

Summary Phase shifts in free-running activity rhythms of male golden hamsters,Mesocricetus auratus, often occur when they establish a new territory and home after a cage change. Similar shifts also often occur after pairs of animals interact with each other for half an hour. When these events take place during the middle of the hamsters' subjective day, they produce phase advances: when late in the subjective night, they produce phase delays. Repeated social interactions at the same time of day can entrain activity rhythms in a way consistent with the shape of the phase response curves. Not all individuals become entrained, as is predictable from the modest amplitude of the phase response curve. The effects of social interactions and of other disturbances may be mediated through an oscillator phased by general arousal. The present findings have implications for the interpretation of drug-induced changes in biological rhythms.     

Regression models for convex ROC curves

The performance of a diagnostic test is summarized by its receiver operating characteristic (ROC) curve. Under quite natural assumptions about the latent variable underlying the test, the ROC curve is convex. Empirical data on a test's performance often comes in the form of observed true positive and false positive relative frequencies under varying conditions. This paper describes a family of regression models for analyzing such data. The underlying ROC curves are specified by a quality parameter delta and a shape parameter mu and are guaranteed to be convex provided delta > 1. Both the position along the ROC curve and the quality parameter delta are modeled linearly with covariates at the level of the individual. The shape parameter mu enters the model through the link functions log(p mu) - log(1 - p mu) of a binomial regression and is estimated either by search or from an appropriate constructed variate. One simple application is to the meta-analysis of independent studies of the same diagnostic test, illustrated on some data of Moses, Shapiro, and Littenberg (1993). A second application, to so-called vigilance data, is given, where ROC curves differ across subjects and modeling of the position along the ROC curve is of primary interest.     

Potential of the learning vector quantizer in the cell classification of endometrial lesions in postmenopausal women

OBJECTIVE: To investigate the potential of artificial neural networks for cell identification in endometrial lesions from postmenopausal women. STUDY DESIGN: The study was performed on cytologic material obtained by the Gynoscann endometrial cell samplerfrom 12 cases of atrophic endometrium, 48 cases of hyperplasia without cytologic atypia (18 cases of simple hyperplasia and 30 cases of complex hyperplasia), 12 cases of hyperplasia with cytologic atypia (complex atypical hyperplasia) and 48 cases of adenocarcinoma (30 cases of well-differentiated, 12 cases of moderately differentiated and 6 cases of poorly differentiated carcinoma). From each case approximately 100 cells were examined using a custom image analysis system. A learning vector quantizer (LVQ) identified the collected data. RESULTS: Investigation of cells from Endometrial Alterations with LVQ proved that according to the nuclear characteristics, as expressed by morphometric and textural measures, the endometrial cells from postmenopausal women may be identified as belonging to one of thefollowing three groups: atrophy, hyperplasia without cytologic atypia (simple and complex hyperplasia) and malignant neoplastic lesions (atypical complex and adenocarcinoma). CONCLUSION: The role of nuclear morphologic features in the cytologic diagnosis of endometrial alterations was confirmed. The overlap in thefeature space observed indicates that cell characteristics do not form strictly separate clusters. Thatfact explains the difficulty that morphologists have with the reproducible identification of cells from endometrial lesions in postmenopausal women. Application of LVQ offers a good classification at the cell level and promises to be a powerful toolfor classification on the individual patient level andfor the clarification of the natural history of endometrial pathology.     

Log-normal variation belts for growth curves

Prediction (confidence) or tolerance belts compound the uncertainty of sample estimates with the estimated extent of individual variation. The latter is therefore better described by variation belts, in which sample estimates are simply substituted for population parameters. Variation belts can provide valuable graphical indications concerning the goodness of fit of postulated error models. While multiplicative least-squares (MLS) methods appear appropriate in principle for biological growth, they are unsatisfactory in practice when logarithmically transformed data are heteroscedastic. Heteroscedastic multiplicative error models can be fitted by iteratively reweighted multiplicative least squares (IRMLS), but unacceptable negative or infinite residual variance estimates and unreasonably wide variation belts are occasionally obtained. These difficulties can be prevented by constrained iteratively reweighted multiplicative least squares (CIRMLS). Examples are presented concerning the metabolic allometry of white rats, the somatic growth of male elephant seals, and the growth of an experimental population of Paramecium caudatum.     

Calculation of melting curves for DNA

A method is reported for calculating the melting curve of a DNA molecule of random base sequence, including in the formalism the dependence of the free energy of base pair formation on the size of a denatured section. Some explicit results are shown for a “typical” base sequence, in particular the probability of helix formation at individual base pairs in several different regions of the molecule and the amount of melting from the end of the chain. Particular attention is drawn to the variation of local melting behavior from one region of the molecule to another. It is found that sections rich in AT melt at relatively low temperatures with a fairly broad transition curve, whereas regions rich in GC pairs melt at higher temperatures (as expected) with a very abrupt, local transition curve. To account qualitatively for the results one may divide melting into two kinds of processes: (a) the nucleation and growth of denatured regions, and (b) the merging together of two denatured sections at the expense of the intervening helix. The first of these processes dominates in the first stages of melting, and leads to rather broad local melting curves, whereas the second process predominates in the later stages, and occurs, in a particular part of the molecule, over a very narrow temperature range. It is estimated that the average length of a helix plus adjacent coil section at the midpoint of the transition is approximately 600 base pairs. Since transition curves which measure the local melting behavior reflect local compositions fluctuations, these curves contain information about the broad outlines of base sequence in the molecule. Some suggestions are made concerning experiments by which this potential information source could be exploited. In particular, it is pointed out that one might hope to map AT or GC rich regions at particular genetic loci in a biologically active DNA molecule. Values of the relevant parameters found earlier for the transition of homopolymers produce melting curves for a DNA of random base sequence which are in good agreement with the experimental transition curve for T2 phage DNA. Hence the present theoretical picture of the melting of polynucleotides is at least internally self-consistent.     

The role of liquid-based cytology associated with curettage in the investigation of endometrial lesions from postmenopausal women

OBJECTIVE: This study investigates the role of liquid-based cytology with ThinPrep technique, in the detection of endometrial lesions, using direct endometrial sampling from postmenopausal women with the Endogyn endometrial device. METHODS: It was performed on 491 postmenopausal women referred to our clinic for abnormal bleeding or other symptoms and/or a thickness of endometrium >5 mm on ultrasound. Endometrial sampling, dilatation and curettage (D&C) and hysterectomy were performed on all patients. For the diagnosis, the WHO classification scheme was used. RESULTS: According to our findings a sensitivity of 98.08%, specificity of 100%, positive predictive value of 100%, negative predictive value of 100% and overall accuracy of 98.98% were observed in both endometrial sampling and in D&C. CONCLUSIONS: Endometrial sampling is complementary to D&C for the diagnosis of endometrial lesions and it is necessary for it to be performed before D&C and/or hysterectomy.     

Diagnostic usefulness of endometrial aspiration cytology for endometrial cancer cases with normal curettage findings

OBJECTIVE: To determine whether endometrial aspiration cytology is useful for endometrial cancer cases with normal endometrial curettage findings. STUDY DESIGN: Eleven cases in which endometrial cancer could be detected by endometrial aspiration cytology but not endometrial curettage were classified into 2 groups by cancer locus, on the endometrial surface (A) or in the myometrium (B). A clinicopathologic and cytologic analysis was performed to compare the 2 groups. RESULTS: Five cases had cancer lesions localized at the fundus and one at the isthmus (group A). The other 5 had lesions localized in the myometrium (group B). The myometrium invasion was beyond half the myometrium in group B and within half in group A. It required > 2 cytologic examinations for a definitive diagnosis in 33.3% of group A and 80.0% of group B. The endometrial cytology differed clearly between the groups: large clusters of malignant cells with a dirty background (group A) vs. small clusters with a clean background (group B). The log-rank test revealed that group B had significantly poorer prognoses than did group A despite nearly the same rate of stage I/II cases in the 2 groups (p = 0.004). CONCLUSION: Endometrial aspiration cytology was useful for endometrial cancer cases with normal curettage findings as part of early detection. However, the cytologic diagnosis did not indicate good prognoses in the cases of cancer localized in the myometrium.     

Spatial Stochastic Models for Cancer Initiation and Progression

The multistage carcinogenesis hypothesis has been formulated by a number of authors as a stochastic process. However, most previous models assumed “perfect mixing” in the population of cells, and included no information about spatial locations. In this work, we studied the role of spatial dynamics in carcinogenesis. We formulated a 1D spatial generalization of a constant population (Moran) birth–death process, and described the dynamics analytically. We found that in the spatial model, the probability of fixation of advantageous and disadvantageous mutants is lower, and the rate of generation of double-hit mutants (the so-called tunneling rate) is higher, compared to those for the space-free model. This means that the results previously obtained for space-free models give an underestimation for rates of cancer initiation in the case where the first event is the generation of a double-hit mutant, e.g. the inactivation of a tumor-suppressor gene.     

Biodiversity curves for the Ordovician of Baltoscandia

Biodiversity curves for the Ordovician of Baltoscandia show a substantial increase in taxonomical diversity through the period, as seen also in global data sets. A database of 10,340 records of first and last appearances of species at different localities in the region has been analysed using simple species counts, and partly validated with resampling methods. While the biodiversity curve for all fossil groups combined is probably reasonably accurate except for an unknown scaling constant, taxonomical or geographical subsets may not be sufficiently well sampled to allow precise measurement of their species counts through time. The analysis shows a major diversification event commencing in the middle Arenig (Ibex-Whiterock boundary), and more limited diversification events in the Llanvirn, Caradoc and Ashgill. The Scandinavian (Norwegian and Swedish) biodiversity curves are broadly correlated with major changes in sea level, with low biodiversity at highstands and high biodiversity at lowstands, although this pattern is not clear for all fossil groups. In the Arenig, graptolites and trilobites appear to have higher diversities at high sea levels, while the brachiopods and ostracodes show higher diversities at low sea level. As a consequence, the Arenig diversification is delayed for the latter two groups until the upper end of the interval.     

Uncertainty calculations for theoretical flight power curves

The comparison of theoretical and experimental estimates of the mechanical power requirement for flight is currently impossible owing to the absence of complete experimental data based on mechanical power, as opposed to measurements of metabolic rates. Nevertheless, comparisons of measured and predicted characteristic speeds, and inferred power curves are frequently made, despite the total absence of uncertainty estimates of the theoretically predicted quantities. Here the method for correct calculation of uncertainty estimates in mechanical power models is outlined in detail, and analytical and numerical results are derived for realistic examples. The sensitivity of the calculated variations in power requirement varies greatly among the independent variables, and the practical and theoretical consequences of this variation are discussed. Pending the arrival of appropriate experimental measurements, it is now possible, in principle, to make quantitative comparisons with theoretical predictions.     

Receiver operating characteristic curves for Poisson signals

Receiver operating characteristic (ROC) curves for the detection of Poisson distributed signals were evaluated by computer simulation and found to be a set of nearly collinear points when plotted on probability axes. Approximate expressions were derived for the slope and intercept (d) of the line containing the points. The predictive value of these expressions was verified by measuring slopé and d of ROC curves produced from computer simulation of Poisson signals. Estimates of slope and d from empirical ROC curves were distributed according to the Gaussian probability density with variance approximately equal to 3/n, where n is the number of stimulus repetitions used to determine the ROC.