Immunotherapy with interleukin-2 (IL-2) has been limited by dose-dependent systemic toxicities secondary effects inducing a "vascular leak" syndrome. The purpose of our study was directly to observe and to quantitate changes in skin capillary permeability in response to microinjection of IL-2 by measuring transcapillary diffusion of sodium fluorescein. Twelve healthy volunteers were studied. IL-2 (2.5 microliters; 45,000 i.u.) was injected into the subepidermal skin layer of the distal tibial plateau by using a new microinjection technique. At the opposite leg, an equivalent amount of the solvent was injected to serve as the intraindividual control site. Three and 24 h after injection, Na-fluorescein was given intravenously, and transcapillary diffusion of the dye was simultaneously recorded with two different video microscopes. Perivascular fluorescent light intensities (FLI) corresponding to transcapillary diffusion of the dye were measured in arbitrary units (AU) by videodensitometry around the sites of microinjection during playback of the videotapes. Mean FLI values representing microvascular permeability 10 s after dye appearance were at 3 h, 1,504 +/- 592 AU for IL-2 and 983 +/- 652 AU for the solvent; and at 24 h, 2,450 +/- 447 AU for IL-2 and 658 +/- 329 AU for the solvent. At 3 and 24 h, the mean values after IL-2 application were significantly enhanced (p < 0.05-0.005) when compared with the mean values after injection of the solvent. The results document that IL-2, as compared with the solvent, significantly increases transcapillary diffusion of Na-fluorescein, reflecting capillary permeability in human skin. The increase in capillary permeability may explain the edema-promoting effect of IL-2 after systemic application.

For the first time measurements of lymph flow velocities in cutaneous microlymphatics of patients with lymphedema were performed and compared with healthy subjects. Flow velocity in single lymphatic skin capillaries was measured using fluorescence video microscopy after subepidermal microinjection of FITC-dextran 150,000 in 15 healthy volunteers and 16 patients with primary lymphedema. Initial filling of the lymphatic capillary network was fast with significantly higher mean velocities in patients with primary lymphedema than in healthy controls (890 +/- 43 vs. 550 +/- 390 microns/s, p < 0.05). The resting velocities were not significantly different between controls and patients (10.3 +/- 4.1 vs. 16.6 +/- 13.9 microns). In 12 out of the 16 lymphedema patients cutaneous backflow of the fluorescent contrast medium from deeper invisible lymphatics was observed. In 4 of these patients rhythmic reflux with a mean frequency of 1.4 +/- 0.5 cycles/min was measured by video densitometry in microlymphatics with a significantly (p < 0.01) enhanced diameter. Mean flow velocity (Vp) in these precollectors was significantly increased compared to the resting velocities (p < 0.01). On the basis of these results the hypothesis is advanced that rhythmic cutaneous backflow originates from intrinsic contractions of deeper lymph collector segments and is transmitted to the superficial microlymphatics through incompetent connecting channels. This newly recognized mechanism appears to be an important factor for the pathophysiology of lymphedema.

The purpose of this study was to investigate the previously unknown flow velocity in single lymphatic capillaries of humans in the supine position. Fifteen healthy subjects (10 women and 5 men; mean age 35.8 +/- 13.1 yr) were studied. Ten microliters of fluorescein isothiocyanate-dextran (150,000 mol wt) were injected into the subepidermal layer of the foot dorsum. The filling of the microlymphatics from the resulting depot was visualized by fluorescence video microscopy and stored on videotape. Flow velocity in the microlymphatics was determined on the video screen by direct measurement of the advancement of dyed lymph during a given time. The following median velocities were obtained: 0.51 mm/s (0.27 and 0.61 mm/s for lower and upper quartiles, respectively) for velocity during initial network filling and 9.7 microns/s (6.9 and 14.2 microns/s for lower and upper quartiles, respectively) for resting velocity at the end of the filling period. Mean lymphatic capillary diameter was 54.8 +/- 8.2 microns, and mean network extension was 8.3 +/- 3.2 mm. The high filling velocities are probably due to increased interstitial pressure and volume caused by dye microinjection, whereas the values measured during the end of network filling seem to approach resting flow velocities.

The laser-Doppler technique was used to assess local muscle and skin blood fluxes at the lower limb in 20 healthy volunteers. After puncturing the anterior tibial muscle with a steel cannula, a single-fibre probe with a diameter of 0.5 mm was inserted into the muscle. Simultaneously, the skin blood flux was measured at calf and foot. The muscle blood flux at rest was 3.5 to 4 times higher than the skin blood flux at calf or foot. The spatial variability of the muscle blood flux at three different sites of measurement was considerable and tended to be higher than in the skin of the calf. After an arterial occlusion lasting 3 min, peak flux was reached in the muscle after 18.7 +/- 9.8 s, in the skin of the calf after 16.8 +/- 9.3 s, and in the skin of the foot after 22.9 +/- 14.6 s (NS). The relative flux increase during reactive hyperaemia was significantly lower in the muscle (2.7 +/- 1.3) than in the skin of the calf (3.9 +/- 1.9; p<0.05) or the foot (5.1 +/- 3.5; p<0.005). The reproducibility of reactive hyperaemia response in muscle was excellent with unchanged probe position, but exhibited a marked variability on different days. The laser-Doppler technique provides the possibility for simultaneous measurement of flow dynamics in muscle and skin with a high temporal resolution. Methodological problems include differences in probe geometry of the single-fibre compared to standard probes and differences in optical properties of the tissues. Direct comparison of flux values may, therefore, be subject to criticism, but not the comparative analysis of relative flux changes. The influence of tissue trauma on muscle blood flux has to be considered for the analysis of flux data.

A newly designed triple probe is introduced for measurements of transcutaneous oxygen tension, laser Doppler flowmetry (LDF) and microangiodynamics of skin capillaries by dynamic video microscopy with and without fluorochromes. The performance of the triple probe was checked in 9 healthy volunteers (6 women, 3 men; mean age: 34 years) and 9 patients (5 women, 4 men; mean age: 67 years) with peripheral arterial occlusive disease (PAOD). The mean Doppler ankle/arm pressure ratio was 0.54 +/- 0.30. Six patients suffered from severe claudication, 2 from rest pain and 1 patient had toe and forefoot necrosis. The foot dorsum was selected as measuring site. After recording baseline values of skin surface PO2 (ssPO(2)) at 37 degrees C, LDF and capillary images, a suprasystolic compression at the ankle level was performed for 4 min. Thirty seconds before cuff opening 0.2 ml/l 1 blood volume of 20% sodium fluorescein was injected in an antecubital vein. Sodium fluorescein arrival times, filling times and maximum fluorescent light intensity times were measured, and ssPO(2) and LDF were recorded continuously during postocclusive reactive hyperemia (PORH). The results indicate an adequate function of the triple probe. The mean resting ssPO(2) was 2.0 +/- 1.9 mm Hg in PAOD patients and 4.0 +/- 3.9 mm Hg in controls (p = 0.185). Maximum ssPO(2) during PORH was significantly reduced (p = 0.008) in patients (3.1 +/- 2.1 mm Hg) compared to healthy subjects (11.8 +/- 7.7 mm Hg). Resting LDF values were 6.5 +/- 6.4 perfusion units (PU) in PAOD patients versus 10.3 +/- 8.2 AU in controls (p = 0.295). Peak LDF during PORH was significantly reduced (p = 0.005) in patients (19.5 +/- 6.4 PU) versus healthy subjects (33.8 +/- 11.5 PU). The mean sodium fluorescein appearance time and maximum fluorescent tight intensity time during PORH were significantly delayed (p = 0.006; p = 0.003) in PAOD patients (11.1 +/- 9.0 s; 330.6 +/- 69.4 s) versus healthy subjects (1.6 +/- 0.9 s; 220.7 +/- 60.6 s), respectively. Microvascular flow distribution during PORH estimated by the mean sodium fluorescein filling time was not different in both groups. The main advantage of the combined device is the simultaneous performance of all three techniques at the almost identical sensing site under the same measuring conditions.