The application of substantial forces to body tissue leads to tissue destruction as a result of either mechanical limits being exceeded or complete ischemia in the affected organ segment. The ischemia can be caused by irreversible vascular destruction, by prolonged pressure that exceeds capillary perfusion pressure and the ischemic tolerance of the tissue, or by protracted angiospasm. On the other hand, a "properly dosed" and, if possible, cyclically acting force in the form of tissue expansion causes stimulation of the proliferation and metabolism of the cells . The problem in the use of soft tissue expansion, to bring about regenerative wound healing processes, clearly lies in the poor definition of the optimal application of force over time. Ilizarov has demonstrated that excessive or overly rapid expansion can also result in tissue degeneration . Thus, the surgical maxim of "tension-free" wound closure cannot be accepted in this apodictic form. The correct tension is the key to optimal healing results. However, this "correct tension" must be individually defined for each cell group - a research task that has only been solved to a limited extent to date .
Yet, there is a theoretical risk of ischemic tissue damage caused by high, non physiological forces acting on the blood vessels. The most meaningful parameters for evaluating tissue ischemia are, in addition to oxygen saturation and blood flow, the oxygen partial pressure, as determined in this study.
The minimum oxygen partial pressure required for cell survival by oxidation of cytochrome a3 is 0.01 mmHg in the mitochondria. This corresponds to a pO2 of 0.05 mmHg at the cell membrane . In healthy volunteers, the intracutaneously measured pO2 was 50 mmHg . These two sets of data span a wide range. Simultaneously, the pO2 content of the tissue exerts a decisive influence on wound healing. For example, a pO2 of 20 mmHg was measured as a limit for collagen synthesis . This increased with a higher oxygen supply, which was attributed to the strong pO2 dependency of the reaction kinetics of the key enzyme for collagen synthesis (propyl-4-hydroxylase) . As the ambient air already contains a pO2 of 50 mmHg, it was logical to assume a measuring range of 0 to 70 mmHg for testing human tissue. It is highly likely that tissue will die at a pO2 as low as 15 mmHg. This statement was confirmed by histographic measurements. Necrosis occurs at values under 10 mmHg, while tissue is not vitally threatened at values above 25 mmHg .
Additional factors affecting the supply of oxygen to the tissue include the hematocrit and blood pressure, but only when their values fall below physiological limits. In contrast, the flow properties of the blood exert a strong influence on tissue supply at normal blood pressure and Hb. Notably, it was possible to increase tissue oxygen partial pressure at normal hematocrit levels by increasing volume, e.g. by means of NaCl 0.9% infusions . Accordingly, collagen synthesis was elevated in patients who received infusions of NaCl and glucose or dextranes , during wound healing. These influences of Hb, blood pressure and volume were specified in standardized form by anesthesia in our experiment setup.
It has been shown that the pO2-measured in the subcutaneous tissue is heavily dependent on the depth of penetration and the diffusion of oxygen through the skin surface into the underlying tissue . A pO2 of 40-60 mmHg was expected at a tissue depth of 2-3 mm. Notably, the diffusion component of oxygen increases as supply from the blood circulation decreases. Several studies demonstrated that this effect could be eliminated by blocking diffusion with a so-called "Oxyblock foil" . When this diffusion was not blocked, a diffusion artifact of up to 5% was obtained in the tissue. When the blood supply was interrupted, the diffusion artifact exceeded 5% and the pO2 approached zero. If the measurement was taken just beneath the surface (to a depth of approximately 2 mm), the pO2 did not decrease to zero, even when the blood supply was interrupted .
In our experiment setup, an Oxyblock® foil could not be applied, due to deformation of the tissue by the tension. This meant that diffusion of oxygen through the skin was possible, which could lead to readings of > 0 mm Hg, even with complete "no flow anoxia". However, this effect disappeared completely with increasing puncture depth of the measuring probe. Consequently, we placed the measuring probe into the intermediate zone between the cutaneous and subcutaneous tissue, which corresponds to a depth of about 4 mm in the skin of the thigh, so that it could be assumed that the diffusion artifact was essentially eliminated (< 1%).
Hirshowitz and coworkers reported that the normalized tension decreased to its final value of 0.95 (corresponds to 95% of the initial tension) after an average of 50 seconds and then remained constant for the next 400 seconds [15, 16]. When measuring skin expansion in patients, it was possible to dispense with time-consuming allowances for creep behavior (0.03% per hour  and hysteresis. In our test series, the skin was initially loaded with 0.5 kg. The amount of time until readings were taken then amounted to more than 50 seconds. All other weights were additionally attached without removing any load from the skin in the interim. Under the assumption that the creep behavior decreases exponentially, we can assume that our measurements were essentially performed in the end state. Consequently, the viscous component of the viscoelastic expansion was ignored in our analysis and the analysis was limited to the sole elastic relationship.
The skin expansions exhibited individual progressions of the test curves that were highly differentiated in quantitative terms. A correlation with the gender of the patients could not be found, nor was there any correlation with the age of patients. The slope of 0.0049 (1/year) at best suggested a decline in maximum expansion with increasing age, although the degree of confidence of 42% provided no significance. In a study by Vogel et al., significant age dependency of the elastic skin properties was reported in vitro . The modulus of elasticity, maximum tension and maximum expansion increased until termination of growth and then gradually decreased with advancing age. The viscous properties, represented by relaxation, creep and hysteresis, continually decreased with advancing age.
In our study the mean expansion at the skin surface was 24%. Of this expansion, 63% (corresponding to 15% expansion) occurred after application of 5 N of force, while 88% (corresponding to 21% expansion)had already occurred after application of 15 N of force. This led to the conclusion that it is not necessary to stretch the skin to its tear limit in order to obtain sufficient tissue by utilizing viscoelasticity. In conclusion, the elastic fibers become effective in the tissue within the range of smaller expansion. This permits atraumatic expansion of the skin with small amounts of tension.
To the best of our knowledge, our study is the first to present this kind of in vivo experimental setup. The indicated movement of the wound edges represents a combination of the mobilization and displacement (viscosity) of the subdermal tissue layers with actual skin expansion. The mobilization of subdermal tissue is probably caused by elastic shear forces, which the skin exerts on the deeper tissue layers. Quantification of the individual components was not possible, as the total area of the expanded and mobilized skin was not known. In fact, a precise quantification would require an analysis of the geometrical skin conditions at the respective measurement site, which would involve recording the three-dimensional expansion profile at the surface.
With regard to pO2 measurements we observed strong inter-individual fluctuations among the initial values. An age dependency of the pO2 values could not be verified. On the anesthesia side, it could be assumed that consistent standard conditions were applied during the individual measurements on one patient. The conditions for individual patients were also standardized and therefore approximately equal. Therefore, inter-individual fluctuations due to changes in anesthesia conditions (e.g., by volume substitution or changes in the oxygen content of the inhaled anesthesia gas, FiO2) can be excluded.
Notably, there are various studies, demonstrating the latter also exhibited strong inter-individual fluctuations, as evidenced by the high standard deviation of the mean values provided (up to 20 mm Hg) [19, 20]. The values, measured by various authors in the subcutaneous fatty tissue of healthy volunteers, averaged 50 mmHg . Other authors indicated values of 64 ± 20 mmHg . If we compare these values with the mean value we obtained, 30.6 ± 25.6 mmHg, which was measured while introducing tension at the wound edge, our value is significantly lower than those stated in the literature. Replacement of the measuring probes led to new initial values that were not comparable with the previously determined values . This is explained by the fact that the oxygen supply is an inhomogeneous function of the location, and that it is subject to fluctuations of up to 1 mmHg per μm of tissue, which corresponds to the functional units of a terminal arteriole with its capillaries, which supply 0.04 - 0.27 mm2 of skin. In our measurements, however, the fluctuations can only be attributed to a functional combination of several terminal arterioles, as the length of the measuring window of the Licox probe is 5 mm. Finally, the average normalized pO2 decreased in linear fashion with the tension at the wound edge and still comprised 80% of its initial value at a maximum introduced force of 30 N. This linear relationship between the pO2 and force is not consistent with the data in the literature [20, 23].
In our opinion, the expansion of the skin results in deformation of the vascular tree. Its morphological structure is such that when the collagen fibers become parallelized, and therefore the elasticity and tear limit is reached, the expansion reserve of the capillaries is not yet exceeded. Only with further expansion and destruction of the collagen fibers does irreversible vascular damage does occur. Our measuring range was significantly below the range of parallelization of collagen fibers. The decrease in the pO2 within the range of lower tension is therefore best understood as a vasogenic reaction to the deformation of the capillary tree, as the experiment setup could not completely rule out vegetative influences on the tissue blood supply.