INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE ABILITY to measure lung filtration coefficient (K_fc) in humans would aid the clinician in the diagnosis and treatment of several pulmonary diseases, including acute respiratory distress syndrome. Until recently, isolated lung preparations were used to measure K_fc. An estimate of K_fc could be made by using animal preparations in which lung lymph was available, but these values could be influenced by problems such as the unknown fraction of lung lymph collected and microvascular pressure (Pmv) (4). Both of these methods are highly invasive, thus preventing their being extended to use in humans.

The gravimetric method performed on an isolated lung preparation has been considered the "gold standard" for measurement of K_fc. The technique entails monitoring the weight of an isolated lung preparation that is subjected to step increases of pulmonary venous pressure (Ppv). The lung gains weight rapidly during the first 1-3 min because of vascular filling. The slow weight gain that follows is normally attributed to fluid filtration across the microvascular barrier (5). The constant-slope method of calculating K_fc assumes that the slow weight gain is entirely caused by filtration and is linear with respect to time. Then K_fc can be calculated by dividing the slope of the weight gain by the Pmv increase and the dry lung weight (DLW) (5).

Harris et al. (8) have shown with the use of ⁵¹Cr-labeled red blood cells (RBC) that vascular volumes can increase for up to 30 min after the pressure elevation. Maron and Lane (17) have shown, with the use of indicator dilution methods, that blood volume (BV) may continue to increase for longer than the 1-3 min after Ppv increases if the isolated lung was not perfused for an extended period of time before the pressure elevation (17). Thus, slow vascular BV changes can adversely affect K_fc calculation based on gravimetric techniques.

Oppenheimer et al. (19) introduced an optical technique for measuring K_fc. Changes in optical-tracer concentrations in lung venous blood are measured after a step change in Ppv. The tracer concentrations increase after a pressure elevation, because fluid flows across the pulmonary capillary barrier more readily than proteins or large macromolecules can move across. K_fc can be calculated by determining the rate of change of optical-tracer concentrations. However, RBC strongly absorb and scatter light at the wavelengths used to measure the optical-tracer concentrations. Multiple wavelengths (19) have been used to correct for oxygen saturation levels or small hematocrit changes caused by the Fahraeus-Lindqvist effect (22) or by respiration (16). With the use of this optical system, Hancock et al. (7) found K_fc values measured optically were ~25% of those calculated by using weight changes, possibly because of slow vascular volume changes that were misinterpreted as filtration when using weight analysis.

Harris et al. (8) have optically measured K_fc in isolated canine lungs at low flows and small hematocrits by using a spectrophotometer. Lung venous blood was monitored with two wavelengths, so that optical-tracer concentration changes could be corrected for fluctuations in hematocrit. The K_fc values obtained by using optical techniques were significantly smaller than those obtained via the gravimetric method until the gravimetric values were corrected for BV changes. Unfortunately, RBC artifacts masked small changes in plasma optical-tracer concentrations, thus restricting this method to low hematocrits and low flow rates.

Earlier, we (15) showed that on-line separation of RBC from plasma before optical measurements allows for measurements of K_fc at physiological hematocrits and flow rates. A polysulfone filter cartridge provides on-line separation of plasma and RBC so that small concentration changes of an optical tracer can be measured by using a spectrophotometer. The K_fc values measured with this technique were within physiological ranges (15), but the values were not compared with simultaneous gravimetric measurements on the same isolated canine lung preparation. In another study, we used a commercially available filter from Cellco (model 4007-10, Germantown, MD) to measure the K_fc of an isolated canine lung by using normal hematocrits and flow rates. We showed that values obtained by using the optical technique compared favorably with those obtained via the blood-corrected (BC) gravimetric method (14).

Alloxan has been used in many studies to alter permeability in isolated canine lungs because it increases permeability while minimally affecting surface area (2, 11, 25). Harris et al. (10) and Zelter et al. (26) used alloxan injury (62.5 mg/kg) to increase lung microvascular permeability which was measured by using [¹⁴C]urea permeability-surface (PS_U) area product in canine preparations. Olson et al. (18) used alloxan to induce injury in canine lungs to show that the PS_U to 1,4-[¹⁴C]butanediol permeability-surface (PS_B) area product ratio (PSR) was sensitive to changes in permeability that were independent of surface area changes (18). To our knowledge, the optical method of measuring K_fc has not been applied to lungs injured with alloxan or with any other agent that causes lung injury.

In this study, K_fc was measured in isolated canine lung preparations by using gravimetric, BC gravimetric, and optical methods. The permeability of the lungs was altered with the use of alloxan to determine the sensitivity of the optical technique to changes in the permeability of the capillary barrier. Changes in permeability, independent of surface area, were determined with multiple-indicator dilution (MID) studies by using two barrier-limited tracers, [¹⁴C]urea and 1,4-[¹⁴C]butanediol.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Optical K_fc measurement theory. The basic principle employed by the optical method for calculating K_fc is that, when Ppv is increased, fluid crosses the microvascular barrier, leaving large solutes in the plasma to become slightly more concentrated. This small transient concentration change can then be used to calculate the product of the reflection coefficient, for the optical tracer (_f) and K_fc (8)

(1)

where pa is arterial plasma flow (in ml/min), Cpv is change in plasma concentration of nondiffusing tracer (in mg/dl), Cpv is initial plasma concentration of nondiffusing tracer (in mg/dl), m_dlw is blood-free dry lung weight (BFDLW in g), and Pmv is change in microvascular pressure (in cmH₂O).

(2)

Gravimetric K_fc measurement theory. The gravimetric technique depends on measuring the weight gain of the lung caused by fluid filtration. The Ppv is increased in a stepwise fashion, and the weight gain is monitored. The initial rapid weight gain is normally attributed to BV increases, but the slower secondary weight gain is assumed to be caused by filtration. This rate of weight gain is then divided by the change in pressure and DLW to determine the K_fc.

(3)

where  is density of fluid (in g/ml) and dw/dt is rate of weight gain (in g/min).

MID theory. The MID measurements of permeability-surface area product (PS) involved the bolus injection of several radioactive tracers with different diffusing characteristics near the inlet of the lung and the collection of samples downstream of the lung. Two separate sets of injections were prepared, with different barrier-limited diffusing markers. One was [¹⁴C]urea, which is sensitive to PS (10, 26), and the other was 1,4-[¹⁴C]butanediol, which is sensitive to perfused surface area (18). The vascular, or reference, tracers used in this study were ⁹⁹Tc-labeled RBC and ¹²⁵I-labeled albumin. Forty samples were collected downstream of the lung by using a revolving collection wheel which turned at a rate of one well/s. The gamma isotope activity of each sample and injectate was measured in a Packard Auto-Gamma Scintillation Spectrometer (model 5921), and the beta activity was measured in a liquid-scintillation counter (Beckman LS 3150T). The radioactive counts of the isotopes in each sample were then converted to tracer concentrations and were plotted with respect to time. These tracer concentration-time profiles were then used in the equations described by Harris and Roselli (9) and Haselton et al. (12) to calculate the PS. The integral extraction (E_i) for urea (U) was computed by using the equation

(4)

where t_peak is time of peak of reference curve, C_R is concentration of reference tracer normalized to injectate concentration, and C_D is concentration of diffusing tracer normalized to injectate concentration.

(5)

Protocol. The system shown in Fig. 1 was used to make the optical and gravimetric measurements of K_fc. The filter fibers (Cellco 4007-10) were prepared for blood contact by infusing 5,000 units of heparin. These filter cartridges were different from the experimental cartridges used in our previous paper (15), but they have been characterized by Klaesner et al. (13) and are commercially available from Cellco. The lung-perfusion pump (Millipore peristaltic pump) circulated 1,000-1,200 ml/min through the lungs, with the flow rate measured by an electromagnetic flow probe (Carolina Biological Supply). A second pump (Cole Parmer model 7523-20) pulled 200-300 ml/min of the lung perfusate from the venous outflow through the filter cartridge. The resistor on the outflow of the filter was adjusted until we achieved an adequate filtrate flow rate that did not cause significant hemolysis or bubble formation. Normally, filtrate flow rate was ~15 ml/min, with a transmembrane pressure of ~150 cmH₂O. Hydraulic capacitors, constructed from disposable syringes partially filled with air, were used to dampen oscillations caused by the pumps. RBC and filtrate were returned to the main reservoir. The perfusate temperature was maintained between 37° and 39°C by pumping warm water through the jacketed beaker used as the main reservoir.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Isolated canine lung perfusion system. Conc. blood, concentrated blood; EB-albumin, Evans Blue albumin.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Eight isolated canine lung preparations were used in this study. Each lung was subjected to three to seven step increases in Ppv, averaging 8.79 ± 0.93 (mean ± SD) cmH₂O. The baseline and elevated Ppv and Ppa are listed in Table 1 for pre- and posttreatment with alloxan. The optical, gravimetric, and RBC-corrected gravimetric K_fc values (in ml · min¹ · cmH₂O¹ · 100 g DLW¹) were measured for each run, and the average values are listed in Table 2. Although the average value of _fK_fc increased after alloxan infusion, the difference from baseline was not significant as shown by two-way ANOVA, with the Student-Newman-Keuls test for multiple comparisons. The gravimetric and BC values measured before and after alloxan infusion were statistically different. Time-course analysis was not done on these experiments, because this analysis was performed in an earlier paper, and it was shown that any changes in K_fc caused by time were small compared with those measured before and after alloxan infusion in this study (14).

In this study, the BC gravimetric and the optical technique both gave results that were not statistically different in the uninjured lung, similar to the results we presented earlier (14). However, after injury, the optically measured _fK_fc values were significantly different than the BC gravimetric K_fc values. Assuming that the difference between these two values was caused by a change in _f, we computed _f by taking the ratio of the optical _fK_fc product to the BC K_fc. These values are listed in Table 3. With the exception of results for animal FA9, the results suggest that _f decreases when K_fc increases, and the product of _fK_fc may not change significantly with injury.

View this table: [in this window] [in a new window]   Table 3.   Values of f calculated by optical fKfc/blood-corrected Kfc

The ratios of post- to preinjury K_fc values (optical is _fK_fc) were calculated for the eight studies and are also listed in Table 4. The average ratios for the gravimetric and the BC gravimetric techniques were not statistically different from each other, whereas the ratio for the optical technique was significantly different from that for the other two techniques. BV increases account for ~40% of the preinjury gravimetric value of K_fc and for ~30% of the postinjury gravimetric value.

The extent of injury to the lung caused by alloxan was measured by using MID techniques. PS_U, PS_B, and PSR for the eight dogs used in this study are listed in Table 5. After alloxan infusion, the decreases in PS_B and the increases in PS_U were not statistically significant, but the PSR increased significantly, indicating an increase in permeability.

Table 6 lists the pulmonary vascular resistance (PVR) values calculated for the eight studies. The average baseline PVR value, 0.60 ± 0.15 cmH₂O · s · ml¹, was statistically smaller than the average postinjury PVR value, 0.75 ± 0.29 cmH₂O · s · ml¹, indicating that the alloxan injury caused an increase in vascular resistance.

Vascular compliance values were calculated for the eight studies by dividing the initial weight change by the pressure step and the BFDLW. These data are shown in Table 7. The preinjury compliance value of 15.57 ± 5.48 g · cmH₂O¹ · 100 g DLW¹ is statistically greater than the value calculated after alloxan injury, 13.45 ± 5.49 g · cmH₂O¹ · 100 g DLW¹. This indicates that the lung vasculature became stiffer after the alloxan injury. This small change is consistent with the BV increase being smaller after alloxan injury.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Several studies have been devoted to the comparison of different methods for measuring lung K_fc. Gravimetric methods are limited to isolated lung preparations and have been shown to be influenced significantly by vascular stress relaxation that accompanies vascular pressure elevations (8, 14). The optical technique can be used in vivo (15) and is unaffected by BV changes that occur during a measurement (6-8, 13-15, 19, 20, 22). One drawback to the optical method is that it requires precise plasma-concentration measurements, because only small changes occur during a filtration measurement. Significant artifacts can be introduced by RBC, which absorb and scatter light. In addition, absorption can be influenced by the extent of oxyhemoglobin saturation (19). These artifacts can be eliminated by removing RBC with a filter or by centrifugation before making the optical measurements (14, 15).

The optical method has only been used to estimate K_fc in normal lungs. In the present study, simultaneous gravitational and optical measurements were used to estimate lung K_fc and albumin _f after inducing lung injury. Alloxan was used to damage the lungs, and damage was confirmed by a significant increase in the ratio of PS_U to PS_B, a significant increase in gravimetric and BC gravimetric K_fc, a decrease in albumin _f, an increase in PVR, and a decrease in pulmonary vascular compliance.

In the use of the optical method, we encountered a potential difficulty which occurs when comparing filtration before and after lung injury. The optical method measures the product of K_fc and _f for the molecular tracer used in the measurement. Although the K_fc increased by a factor of three in this study, the _f for albumin decreased, so the product did not increase significantly. This is a serious limitation, because the detection of lung injury is the principal reason for making filtration measurements.

This limitation can be overcome by making an independent measure of either K_fc (as was done in this study) or _f. Such methods will normally apply to isolated organ preparations. A much better approach would be to select a tracer that is truly intravascular and thus always has a _f of unity. Albumin, with a _f below unity in normal lungs, is not a good choice. Labeled RBC would seem to be an ideal choice, but the Fahraeus-Lindqvist effect severely limits the interpretation of filtration experiments with RBC (16, 22). Very large macromolecules, such as blue dextran (mol wt 2,000,000) would be expected to remain in the plasma during a postinjury measurement. This would allow the optical method to be used to accurately estimate K_fc during lung injury without a separate measure of _f.

The values of _f listed in Table 3 were calculated by dividing the optical _fK_fc values by the BC K_fc values. Both of these values have some error associated with them, so that when the two values were used to calculate _f, the error could be compounded. Thus, some of the calculated _f baseline values were larger than unity.

Results from this study confirm earlier reports (8, 14) that suggest that vascular relaxation occurs during gravimetric filtration measurements. The resulting BV accumulation causes lung K_fc to be overestimated by ~40%. In the present study, however, we found that the rate of intravascular volume change is similar after alloxan injury (30%).

Thus, although the absolute value for K_fc is overestimated both before and after injury, the postinjury-to-baseline ratio is nearly identical to the actual BC ratio. This would imply that percent changes in gravimetric measurements already reported in the literature after alloxan injury are valid, without the need to correct for stress relaxation. It remains to be seen whether this is valid for injurious substances other than alloxan. Lungs treated with mediators that cause simultaneous changes in permeability and vascular tone may exhibit different rates of vascular filling postinjury than preinjury.

In summary, care should be taken in performing and interpreting both optical and gravimetric lung filtration studies. Intravascular stress relaxation during a measurement can significantly contribute to gravimetric K_fc. However, the percent change after lung injury measured with this method will be valid if vascular accumulation is similar before and after lung injury. The optical method is independent of vascular stress relaxation but measures the product of _f, K_fc, _f. Therefore, if a measure of K_fc is desired in an injured lung, without an independent measure of _f, macromolecules much larger than albumin should be used as the intravascular tracer.