Alveolization: does “A” stand for appropriate morphometry?

To the Editor:

In a recent issue of the European Respiratory Journal, Hind and Maden 1 reported a series of well-designed experiments conducted to further enlighten the role of retinoic acid in alveolar regeneration. The ability of retinoic acid to rescue emphysematous lungs, which was initially reported by Massaro and Massaro 2, has attracted great attention. However, the potential of retinoic acid is still controversially discussed 3. Unquestionably, the study of Hind and Maden 1 added important new data to our knowledge of the role of retinoic acid in alveolization, as was emphasised in an accompanying editorial 4. In their editorial, Torday and Rehan 4 ask “Does “A” stand for alveolization?”. Can this question be answered on the basis of the data presented?

In the context of lung development, alveolization is recognised as the process of subdivision of lung saccules by outgrowing secondary septa, which results in an increased number of smaller subunits, now termed alveoli 5. As a consequence, total alveolar surface area is considerably enlarged. Hence, alveolization is defined by changes in at least three morphologic characters: alveolar size (volume), total alveolar surface area and number of alveoli. Unfortunately, Hind and Maden 1, in pursuing to demonstrate that retinoic acid induces alveolar regeneration in the adult mouse, used but one single parameter to assess alveolization: the mean chord length (L_m), also called mean linear intercept length.

The L_m is a measure of the average distance between two intercepts of a test-line, arbitrarily superimposed on parenchymal tissue, with the alveolar walls. No distinction is made between test lines running within alveoli and those crossing alveolar ducts. Hence, changes in L_m may reflect changes in alveoli or in alveolar ducts. In turn, if alterations are present in both alveoli and alveolar ducts, but are of opposite sign, this may not result in any differences in L_m at all. Therefore, determining L_m provides us with an indicator of changes in airspace (alveoli plus alveolar duct) size, but it cannot be used to assess changes in alveoli with sufficient certainty, as has been emphasised by others 6.

The enthusiastic reader might argue that the authors also presented data on the total alveolar surface area (S_a). However, as was stated in the Material and methods section, this parameter was derived by calculation from L_m and lung volume according to Weibel 7. Hence, this calculation is based on the inverse relationship between the surface-to-volume ratio (S/V) of an object and L_m according to the formula S/V=4/L_m. Thus, the total alveolar surface area of the lung is calculated as S_a=4×V_lung/L_m. As the surface-to-volume ratio strongly depends on the shape of the object measured, the alveolar surface area calculated from L_m may be affected by changes in shape; as was convincingly demonstrated (fig. 4c and d of 1), there were considerable alterations in alveoli from control as compared with dexamethasone-treated animals. Even if we accept the assumption to be true that alveoli are of equivalent shape irrespective of treatment with disulphiram, dexamethasone or/and retinol, one problem still needs to be addressed. Measurements of L_m were done on sections of lungs embedded into paraffin, whereas organ volume was determined prior to embedding. Paraffin embedment introduces considerable tissue shrinkage, which affects L_m measurements 8, and no details are given how measurements were corrected for tissue shrinkage.

E.R. Weibel and his co-workers and colleagues have greatly advanced the field of quantitative morphology during the last decades. Today, a whole range of stereological tools is available that can readily be applied to the quantification of morphologic characters related to volume and size, surface area and numbers 9–11, and new promising tools are still emerging 12. Quantitative morphology does no longer need to make assumptions about the shape or other features of structures to be analysed. Massaro and Massaro 2, for example, have already demonstrated how the tools of design-based stereology can be implemented into the investigation of alveolization (including the effects of retinoic acid) as well as alveolar regeneration, and how these tools enormously increase the impact of such studies. It is a pity that Hind and Maden 1, as well as others 3, wasted so much of the inherent power of their beautifully designed experiments by relying on a single (biologically ill-defined) morphometric measurement, the mean linear intercept length. A variety of versatile design-based stereological tools are available today. Researchers in respiratory medicine should no longer hesitate to make their choice to get better answers to questions like “does “A” stand for alveolization?”.