Towards economy of effort in quantitative ultrastructural pathology: Efficient sampling schemes for studying experimental carcinogenesis

Abstract

Hamster cheek pouch mucosa was painted with DMBA in order to compare samples of epithelium at different pathological stages of carcinogenesis. Stereological methods were applied to ultrathin sections of tissue to estimate one ultrastructural index of change, the surface ratio of lamina densa compared with the overlying plasma membrane of cells in stratum basale. Analysis of variance techniques were then employed to isolate and quantify the contributions which different levels of sampling (animals, tissue blocks, microscopic fields) made to the total observed variation in this surface ratio. Economical sampling schemes for future use were calculated from the sampling variances by taking into account the relative costs at each sampling level. Though illustrated by means of the hamster cheek pouch-DMBA model, our results are pertinent to many other experimental models for quantitative histopathology.

Inter‐animal variation and its influence on the overall precision of morphometric estimates based on nested sampling designs

“…With these techniques, it is possible to quantify the importance of biological variation as a contributor to total observed variance. In those systems for which information is currently available, biological variation has proved to be either of no consequence (Gundersen & Plsterby, 1980) or a major determinant of overall variance (Shay, 1975;Gundersen & Plsterby, 1980, 1981Mayhew et al, 1982). In this report, we adduce further evidence to support the contention that biological variation can be the decisive factor influencing the overall precision of morphometric estimates obtained from nested sampling designs.…”

“…For instance, in the hamster cheek pouch model we have found that biological variation accounts for only about 30? (, of overall variance when the nested design described in this study (example D) is applied to hyperplasia and dysplasia groups (Mayhew et al, 1982). In part this dilution of the influence of biological variation reflects an appreciable intermicrograph variation which may be partially attributable to too high a final print magnification.…”

“…Until comparatively recently (Shay, 1975;Gundersen & Osterby, 1980, 1981Mayhew et al, 1982), few morphometric studies have provided any firm quantitative indication of the importance of between-animals variation versus within-animals variation in nested sampling designs. The present results lend support to a growing body of evidence which suggests that biological variation may overshadow all sources of within-animals variation which characterize experiments based on such designs.…”

“…The present aim has not been to determine optimal sampling designs and so the systems described in this study should be regarded as exploratory. In fact, the original sampling scheme for analysing surface ratios in neoplastic cheek pouch epithelium-four tissue blocks and sixteen fields per hamster-could be improved by selecting only two blocks and twelve fields per animal (Mayhew et al, 1982).…”

A quantitative investigation of immunocytochemically stained blood vessels in normal, benign, premalignant and malignant human oral cheek epithelium

“…For the purpose of our investigation, blood vessels were defined as those structures lined by endothelial cells which might contain erythrocytes. A stratified systematic random sampling procedure was adopted [19,26]. Table 1 summarises the sampling strategy adopted in this study.…”