Membrane Diffusion and Capillary Blood Volume in Chronic Thromboembolic Pulmonary Hypertension

Membrane Diffusion and Capillary Blood Volume in Chronic Thromboembolic Pulmonary Hypertension* Robert]. Bernstein, MD, FCCP; Robert L. Ford; Jack L. Clausen, MD; and Kenneth M. Moser, MD, FCCP

A reduced diffusing capacity for carbon monoxide (Dco) is common among patients with chronic thromboembolic pulmonary hypertension (CTEPH) and often persists for more than a year following successful pulmonary thromboendarterectomy (PTE). To determine the relative contribution the pulmonary membrane diffusing capacity (DM) and pulmonary capillary blood volume (Vc) make to the reduction in Dco, we measured both in 29 patients with CTEPH before and approximately 3 weeks after PTE. Mean preoperative DM was reduced in patients with CTEPH (28 mL min- 1 mm Hg- 1 vs 43 mL min- 1 mm Hg- 1 in control subjects; p0.05) despite substantial reduction in mean pulmonary artery pressure and increase in cardiac output after surgery. We conclude that the low Dco observed in patients with CTEPH before and after PTE is principally caused by a reduced DM and to a lesser extent by a low VC· The mechanisms responsible remain speculative but may reflect pathophysiologic changes in the pulmonary microcirculation caused by chronic pulmonary hypertension that did not improve in the postoperative period studied. (CHEST 1996; 110:1430-36) Key words: chronic thromboembolism; diffusing capacity Abbreviations: ATS=Ameiican Thoracic Society; CTEPH=chronic thromboembolic pulmonary hypertension; Dco=diffusing capacity for carbon monoxide; D~1 =pulmonary membrane diffusing capacity; Flo 2 =fraction of inspired oxygen; PTE=pulmonary thromboendmterectomy; VA=alveolar volume; Vc=pulmonmy capillary blood volume

chronic thromboembolic pulmonary hypertension (CTEPH) is an uncommon complication of pulmonary embolism characterized by widespread occlusion of the central pulmonary arteries by organized thrombus. Successful surgical pulmonary thromboendarterectomy (PTE) affords prompt reduction in pulmonary vascular resistance and improves cardiac output, while resolution of hypoxemia and restoration of exercise tolerance usually occur more gradually following surgery.l Recent investigations into the physiologic basis for this hypoxemia have found a moderate ventilation perfusion (V/Q ) inequality and reduced mixed venous oxygenation to be the predominant causes preoperatively and intrapulmonary shunt the dominant postoperative cause. 3 A r educed diffusing capacity for carbon monoxide (Dco) is commonly observed in patients with CTEPH. 1 *From the Division of Pulmona1y and Ciitical Care Medicine, Department of Medicine, University of California San Diego. Suprorted in pmt by UCSD-NHLBl Training Grant (HL-01022) ana UCSD-NHLBI SCOR (HL-23584), ana by a Will Rogers Memorial Fund Fellowship (Dr. Bernstein). Manusclipt received October 13, 1995; revision accepted May 30, 1996. Re7Jrint requests: Dr. Clausen, UCSD Medical Center, MC 8378, 200 W. Arbor St, San Diego, CA 92103 1430

However, this reduction has been documented to persist for more than a year following successful PTE despite clinical, radiographic, and hemodynamic evidence of functional improvement. 3 The pathophysiologic basis for this observation remains unknown. Fixed reductions of pulmonary membrane diffusing capacity (DM ) or pulmonary capillary blood volume (Vc), or opposite and off-setting changes in both could be responsible for the failure of the Dco to increase following PTE. To our knowledge, measurements of D~ and Vc have not been reported previously in this unique patient population either before or after PTE. Accordingly, in the present study, we measured DM and V c prospectively in a group of patients with CTEPH to determine their relative contributions to the Dco before and after PTE. MATERIALS AND METHODS

Patients

Forty-five consecutive patients with CTEPH were enrolled to undergo testing approximately l week before and 3 weeks (mean, 21 days; range, 10 to 104 days) after PTE. Timing of the postoperative study was determined by the patient's postoperative progress; Clinical lnvestigabons

it was performed within 48 h of the patient's planned hospital discharge. Patients were excluded from the study if they were unable to complete the testing protocol for any reason (11 patients ); such reasons included the following: the Dco system was not functional in the narrow window of time available preoperatively; patients were unable to breath-hold; patients were in clinically unstable conditions for pulmonary function testing; and postoperative death occurred before measurements could be made. Additional ctitetia for exclusion included significant obstructive lung disease as manifested by an FEV/FVC ratio less than 0.70 (3 patients) or restrictive lung disease as indicated by a FVC below the predicted lower limits of normal (predicted value minus 1.65 times the standard error of the estimate; 2 patients). Thus, atotal of29 patients comprised the study group. All subjects gave informed consent. Pulmonary Function Tests

Each p atient underwent spirometry using a volume displacement spirometer (Horizon; Sensormedics; Yorba Linda, Calif) following American Thoracic Society (ATS ) guidelines 4 and lung volume measurements using a variable pressure body plethysmograph5 (Warren E. Collins; Braintree, Mass ). We used the prediction equations of Morris et al6 for spirometty and Goldman and Becklake7 for lung volumes. Measurement of Pulmonary Diffusing Capacity

Single-breath Dco was measured following the recommendations of the ATS 8 using a pneumotachograph-based system with automated control of washout and sample collection volumes (thermal conductivity measurements of helium, infrared analyzer for measurements of carbon monoxide; E1ich Jaeger Inc; Hoechberg, Germany). The Dco was perform ed in duplicate with each of two different inspired gas mixtures containing 0.3% CO, 10% He, 21% Oz, and 68.7% N2, and 0.3% CO, 8% He, and 91.7% 02. Oxygen concentrations were measured vvith a paramagnetic analyzer. The subjects initially washed out nitrogen by breathing 100% 02 for 5 min. Expired Oz concentration exceeded 97.4% after C02 absorption, thus assuring an adequate alveolar Poz in all subjects. Dco was subsequently measured at fraction of inspired oxygen (Floz) of0.917. The subjects th en breatl1ed room air for 10 min and the Dco was measured at Floz of 0.21. At each inspired 0 2 concentration, the mean of all technically acceptable Dco values, as defined b y ATS standards,8 was taken as the reported value. The washout volume of gas expired p1ior to sampling was standardized to 1 L to ensure reproducible measurement of DM and Vc 9 The measured h elium concentration at the higher Floz was corrected for the influence of 0 2 on the thermal conductivity of the sampled gas mixture bymeasuring the helium concentrations of known dilutions of helium in various oxygen and nitrogen gas mixtures, and subsequently developing the following correction equation: Hec= HeM - [(OzETX0.0274)- (0 2ET 2x0.00008)-0.544] where Hec is the corrected h elium concentration expressed as % for FIOz>0.21, H e~·! is the measured h elium concentration, and 02ET is the end-tidal Oz concentration expressed as %. The output of the pneumotachograph used for measu ring inspiratory volume was corrected for changes in 02 viscosity at Floz of 0.917. The inspiratory volume was corrected for ambient temperature pressure--dry (ATPD ) to standard temperature and pressure-dry (STPD) conditions for the calculation of Dco and was corrected for APTD to body temperature pressure saturated (BTPS) conditions for the reported values of alveolar volume (VA) . Anatomic dead space was assumed to be 150 mL. Apparatus dead space for the inspiratory circuit was 155 mL. Both anatomic and apparatus dead space were adjusted for ATPD to BTPS changes and were then used to correct the computed VA. The dead space of the sample collection bag was 25 mL, which was used to correct expired CO and He concentra-

Table !-Pulmonary Function, Hemodynamics, and Resting PaO
Pulmonary function FVC, %pred FEV1, %pred FEV1/FVC,% Dco, %pred VA, L VA, % pred Hemodynamics PAM, mm Hg CO, Umin PVR, dynes-s-cm- 5 PaOz/floz

Before

After

89::!::17 91::!::18 77::!::4 69::!::13 5.28::!::1.31 84

71::!::15 75::!::18 80::!::5 56::!::8 4.27::!::1.02 68

49::!::14 3.8::!:: 1.2 973::!::581 359::!::70

26::!::9 5.9::!::1.4 242::!::142 307::!::51

*All data are expressed as mean ::!:l SD. PAM=pulmonary artery mean pressure; CO=cardiac output; PVR=pulmonary vascular resistance. p<0.001 by paired t test for all of the parameters listed above except for FEViFVC, for which p=0.025. tions. For measurement of PAOz, the sample collection dead space was assumed to be 20 mL. For the calculations of VA, expired helium measurements were corrected for C02 adsorbed prior to helium determinations (assuming a C0 2 value of 5%). The oxygen meter used for measuring fractional alveolar oxygen (F AOz) was linearized for measurements at higher Oz concentrations. Breathhold duration was measured according to the method ofJones and Meade. 10 Subjects were instructed to breath-hold against a closed glottis (instead of sustaining maximal inspiration for the duration of breath-hold). CO back pressure was measured in all subjects at the beginning and the end of testing with both inspired gas mixtures. This was accomplished by rebreathing in a closed system that included aC0 2 absorber and measuring the equilibrated CO concentrations for the Fio 2 of 0.917 gas mixture and by analysis of end-expired CO concentrations for the Flo2 of 0.21 mixture. 11 The Dco was corrected for CO back pressure by linear interpolation b etween the s tarting and finishing expired CO fractions with subtraction of this value from the measured alveolar CO concentration 11 ,1 2 Determination of DM and Vc

DM and Vc were determined using the te chnique of Roughton and Forster. 13 DM and Vc were calculated by applying the following equation: 1/Dco= 1/D~-1 +1/8 x 1N c where 8 is the reaction rate of CO vvith hemoglobin at standard conditions. 8 was obtained using the follov,~ng equation: 118= (0.33+0.0057xPc0 2)x14.6/Hgb as derived by Roughton and Forster,l3 and as used by others ,14-17 assuming a ratio of erythrocyte membrane to interior permeability, A., of co. The hemoglobin used was the patient's hemoglobin on the day of testing. Dco values used to compute Vc and DM were not adjusted for the patient's hemoglobin because the adjustment for hemoglobin was performed as part of the calculation of liE>. Pc02 indicates mean capillary Po2 and was estimated to be equal to the measured end-tidal Po2 minus 10 mm Hg, after Morrison et al 1 5 DM and Vc were estimated by east l squares linear regression using two or more values of Dco obtained at each of tlw two oxygen concentrations. Values ofDM and Vc were accepted only if the correlation coefficient of 1/Dco vs 1/E> was greater than 0.95. 17•18 Nonnal values for DM and Vc were derived from the prediction CHEST 1 110 I 6 I DECEMBER, 1996

1431

Table 2-DM and Vc Before and After Thromboendarterectomy DM, mUminlmm Hg, Value (%Pred)

Statistical Analysis

Vc,mL, Value (%Pred)

x

Subject

Before

After

Before

After

1

23 (48) 23(42) 31 (65) 16 (35) 34(66) 22 (42) 23(42) 19 (40) 40(77) 37(68) 37(66) 55 (118) 28(54) 35(93) 38 (80) 37 (71) 24 (47) 18 (36) 19 (38) 24(50) 19 (36) 25(45) 25(67) 28 (51) 25 (55) 29(72) 37(67) 27(54) 30(58) 28±1.6 58±3.5

24 (50) 26(48) 21 (45) 12 (26) 24(48) 23(45) 20(37) 24(50) 20(38) 20(37) 29 (51) 26 (55) 12 (23) 22 (58) 29(62) 21 (40) 14 (28) 18 (35) 14 (29) 19 (41) 17 (32) 17 (34) 18 (47) 35 (64) 16 (35) 15 (37) 32(58) 22(44) 22 (43) 21±1.1* 42±2.0*

54 (61 ) 54 (62) 93 (118) 37 (43) 59(69) 35(43) 72(82) 40 (53) 79(86) 45 (51) 51 (57) 71 (78) 58 (70) 35(48) 50(55) 53(55) 43 (55) 52(65) 50(63) 69(88) 70(83) 53(58) 44 (67) 89(99) 56(70) 48 (61) 71 (73) 76 (86) 55(68) 57±2.8 68±3.2

53(60) 54(62) 67 (84) 32 (37) 52 (61) 29(36) 34(39) 24 (31) 90(99) 48 (55) 52(58) 54(60) 70(84) 24(34) 46 (50) 52(54) 74(94) 65 (81) 68 (84) 46(59) 80(95) 62(69) 52 (84) 42(47) 47 (58) 66(84) 48 (49) 76 (86) 53(64) 54±3.0 1 64±3.7 1

2

3 4 5 6

7 8 9 10 ll 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Mean±SEM %pred±SEM

*p<0.001 vs preoperative values. Ip>0.05 vs preoperative values. equations of Georges and colleagues 16 and were also measured in 16 healthy age-matched volunteers who had no evidence of respiratory disease and who had normal results of spirometry and Dcos.

10~----------------------------------.

C)

:I:

•

r = 0.47 p=0.01

5

•

0

•

••

~

-5

-10 •

t5

-15

•

~

~ -20

•

-25

-30 '

-3

-2.5

-2

-1.5 -1 DettaVA(L)

-0.5

0

0.5

FIGURE l. Correlation of the change in pulmonary membrane diffusing capacity (delta DM) with the change in VA (delta VA) following PTE.

1432

RESULTS

The mean age of the patients was 41 years; the duration of symptoms prior to PTE ranged from 6 weeks to 7 years with an average of 3.5 years. There were 9 current smokers, 9 ex-smokers, and 11 nonsmokers. Pulmonary function data, hemodynamics, and resting PaO;z/Floz ratios are shown in Table 1. Pulmonary hypertension was present preoperatively in all subjects, with a mean pulmonary vascular resistance of 973 dynes·s·cm- 5 . With respect to their preoperative and postoperative hemodynamics, the patients studied were not different from our larger series. The PaOz/ Flo2 ratio measured at rest was reduced in most patients. Preoperative DM and V c

Mean preoperative DM was 28 mL min- 1 mm Hg- 1 (58% of predicted) and Vc was 57 mL (68% of predicted; Table 2), whereas in the healthy volunteer control group, DM was 43 mL min- 1 mm Hg- 1 (90% of predicted) and Vc was 67 mL (83% of predicted). The difference between patient and control group values was statistically significant for Vc (p=0.044) and more so for DM (p=0.0002). Eighteen patients (62%) had a preoperative DM that was below the predicted lower limit of normal; 13 (45%) had a preoperative Vc below the predicted lower limit of normal. In five subjects, both DM and Vc values were higher than the lower limits of normal. There were no significant correlations between either DM or Vc and duration of symptoms, preoperative hemodynamics, or resting Pa02/Floz ratios . Changes in V c and DM Following Thromboendarterectomy

E

·e §:

Comparisons of mean values between patients and control subjects were made by a Student's unpaired t test, and analysis of data before and after PTE was done by a paired t test. Differences between patient subgroups were tested by 2 analysis. Linear regression methods were used to compare preoperative and postoperative DM and Vc with VA, hemodynamics, and PaOdFlo2 ratios. All p values were two-tailed with statistical significance accepted at the p<0.05 leveL

The mean DM dropped to 21 mL min- 1 mm Hg- 1 (42% of predicted) after surgery (Table 2), a statistically significant (p<0.001) 25% decline from the mean preoperative DM. Twenty-eight of 29 patients (97%) had a post-PTE DM that was below the predicted lower limits of normal. There was a linear relationship between the postoperative decrease in DM and the decrement in VA associated with surgery (r=0.47, p=0.01; Fig 1). However, the postoperative change in VA only partly contributed to the variance in the postPTE change in DM (?=0.22). Clinical Investigations

Mean Vc did not change significantly following PTE (54 mL post-PTE vs 57 mL pre-PTE; p>0.05) despite a substantial reduction in mean pulmonary artery pressure and pulmonary vascular resistance and an increase in cardiac output after surgery (Tables 1 and 2). Eighteen subjects (62%) failed to exhibit an improvement in Vc following successful PTE, and only 14 (48%) had a postoperative Vc that was equal to or greater than the predicted lower limit of normal (Table 2). The changes in DM and Vc with surgery did not correlate significantly with the changes in mean pulmonary artery pressure, pulmonary vascular resistance, cardiac output, or resting Pao21Fio2 ratios. The incidence of clinically evident postoperative reperfusion pulmonary edema did not correlate with the magnitude of change in postoperative DM and VC· DISCUSSION

This study reveals that the low Dco observed among patients with CTEPH before and after PTE is chiefly caused by a reduced DM and to a lesser extent by a low VC· DM declines further following PTE, and Vc is unimproved despite the restoration of pulmonary artery blood flow to lung segments successfully opened by PTE. These changes in Dco subdivisions failed to correlate with either the improvement in hemodynamics or the change in the resting Pa02fFio2 ratio in the early postoperative period. Before considering the possible effects of CTEPH on DM and Vc, we must first address other variables that may have influenced our findings independent of the underlying pulmonary hypertensive process. These include the patient's age, height, and smoking status, and the changes in lung volume associated with PTE. Although both DM and Vc decrease with advancing age, 16.1 8 this effect does not become appreciated until after 40 years for DM and 60 years for Vc.l 6 Notably, less than half of our patients were older than 40 years, and only 3 were older than 60 years. Further, the normal reference equations from which we derived percent predicted values are age and height corrected,16 and our findings remain unaffected when expressed as a percentage of predicted. Although there is general agreement that smoking impairs the transfer of CO from the alveolus to circulating hemoglobin, 11 .19-22 there is no consensus on the effect smoking has on the measurement of DM and VC· Van Ganse et al 19 and, more recently, Mahajan et al20 reported that DM was not affected by smoking status, whereas Vc was significantly reduced among current smokers. Sansores and colleagues21 also found that Vc was lowered by the short-term effects of cigarette smoking. In contrast, Frans et al 22 observed that with correction of the measured Dco for CO back pressure, Vc did not differ between smokers and nonsmokers.

Instead, a reduced DM was implicated as the cause of the low Dco observed in smokers, and subclinical emphysema was invoked as an explanation for the low DM. Finally, a recent study of 37 subjects, 8 of whom were current smokers, found no significant difference in DM or Vc between smokers and nonsmokers. 12 In the present study, we corrected the Dco for the presence of CO back pressure, and current smokers (all of whom abstained from smoking for several days prior to testing) constituted less than one third of the total patient cohort. Finally, our observations regarding the changes in DM and Vc with PTE remain valid even if absolute values in some patients have been influenced by cigarette smoking in the past. It is possible that the decrease in lung volume noted in virtually all patients in the postoperative period following PTE is responsible for the observed changes in DM and VC· Cotes et al9 demonstrated that DM varied directly with VA, whereas Vc was inversely related to lung volume in a group of four normal subjects. However, a recent, more extensive analysis of these relationships in 37 normal subject,s.-by Starn et al 12 has revealed a more complex picture. Although there was a positive linear correlation of mean DM with VA, individual responses of DM to changing VA were not uniform. Further, in some individuals, the relationship between DM and VA was curvilinear with increasing DM as VA was decreased. Both the mean and individual relationship between Vc and VA could best be described by a nonlinear second-order polynomial with decreases in Vc at both high and low lung volumes and a maximum Vc between these extremes. In the present study, we found only a weak correlation between the change in DM and the change in VA with surgery (r2=0.22) and no correlation between the change in Vc with the change in VA (r2=0.00). It can be concluded that, although the postoperative decline in DM may be secondary to the observed decreases in VA , other factors may be contributing. The mean Vc obtained in our control subjects was lower than furedicted by the equations of Georges and colleagues. 6 However, our control group mean Vc of 67 mL is similar to values obtained in a number of other recent studies 12 ,1S,lB,20·21 ·23 in healthy normal subjects (means of Vc values ranged from approximately 50 mL21 to 71.62 m08 ). Although our procedure of measuring DM and Vc closely paralleled that of Georges et al, 16 our use of the ATS recommendations resulted in some differences (eg, washout volume of 1.0 L vs 0.75 L used by Georges et al; conditions of breath-hold were not described by Georges et al) . Hence, we cannot exclude differences in methods that may have led to this apparent discrepancy in results in normal subjects. To our knowledge, Vc and DM measurements have CHEST I 110 I 6 I DECEMBER, 1996

1433

not been reported previously for patients with CTEPH. Although disease duration may have influenced our findings, there was no correlation between DM or Vc and duration of symptoms. For patients with more acute pulmonary embolism, reports of Vc and DM findings vary. Nadel et al24 reported an average Vc of 50% in 3 patients with pulmonary emboli; DM was not reported. Daum25 noted decreased Vc values and normal OM measured by the steady-state technique in 31 patients with pulmonary emboli, 13 of whom had long duration of the disease and repeated emboli. The Dco and Vc values were more reduced in the patients with histories suggestive of chronic emboli than in the few patients with a single episode of pulmonary embolus. In a randomized trial of 40 patients with pulmonary emboli treated with either heparin followed by oral anticoagulants or thrombolytics followed by heparin and oral anticoagulants, Sharma et al 26 noted that 2 weeks and 1 year after the initiation of therapy, the mean Vc was 64% and 60% of predicted, respectively, in 21 patients treated with anticoagulants alone; in contrast, Vc was not reduced in those who had received thrombolytics followed by anticoagulants (96% and 104% at 2 weeks and 1 year, respectively) . In both groups, the DM values did not differ from predicted values. In this article, no mention of CO back pressure correction was made, which if present, would lead to a systematic underestimation ofVc and overestimation of DM. 22 In a more recent study, Fennerty and colleagues23 found no significant difference in Vc or OM between 14 age- and height-matched control subjects and 14 subjects with acute submassive pulmonary embolism studied shortly after diagnosis and at 3 months after the acute episode. The lower Dco values observed initially in the patients were attributed to lower VA findings than observed in the control subjects. The marginally low DM values that were observed initially in the patients were significantly larger when restudied after 3 months of anticoagulation therapy. The mechanisms underlying the reduction we observed in preoperative OM and Vc remain speculative. DM is compromised by processes that limit the surface area available for diffusion of gases and!or increase the thickness of the blood gas barrierP Chest CT scans done in 12 of our patients preoperatively disclosed no evidence of interstitial lung disease as a cause for reductions in OM. Decreased perfusion oflung distal to occlusive organized thrombi, as documented by both perfusion lung scans and pulmonary angiography in patients with CTEPH, may reduce the effective surface available for CO exchange and thus reduce the DM. Were this the case, however, we would expect to see decreases in DM with more acute but massive 1434

pulmonary emboli, changes that have not been reported in the limited studies available from the literature. Decreases in OM, however, could result from pathologic changes in the smaller pulmonary arteries secondary to chronic pulmonary hypertension. Medial hypertrophy, intimal fibrosis, organized thrombi, and plexiform lesions have been described in small pulmonary arteries (<200 pm) in patients with CTEPH27,28 and could result in thickening of the alveolar-capillary membrane or reductions in surface area available for diffusion with a resultant reduction in OM. The further decline in OM following surgery remains unexplained. Atelectasis, pain, and weakness contribute to the postoperative decrease in VA; but as noted earlier, this reduction in VA only partially correlates with the decrement in OM observed following PTE. However, because of the complex relationship between VA and OM and the variable relationship in different subjects as observed by Starn et al,12 the observed changes in VA may be more responsible for decrements in DM than the correlations would suggest. Alternatively, persistent reperfusion pulmonary edema, a syndrome of acute lung injury confined to lung zones subjected to endarterectomy,1 may play a role by damaging the alveolar capillary membrane, resulting in a reduced surface area and widened blood gas barrier for gas diffusion. However, only 6 of our patients developed clinically and radiographically manifest reperfusion lung injury, while 28 had a significant drop in their OM following PTE. Additional studies are needed to explore this concept further. It is also possible that changes in the small vessels from chronic pulmonary hypertension result in reductions in the compliance of these vessels so capillary surface area does not increase with increases in distal transmural pressures but may still fall if lung volumes decrease. The reduced Vc we observed before surgery is likely due to the extensive proximal pulmonary artery obstruction by large thromboemboli and pathologic changes in the more distal pulmonary vessels secondary to chronic pulmonary hypertension. 27·28 However, Vc was not as severely impaired as one might anticipate given the severe degree of compromised pulmonary arterial blood flow observed in these patients. 1· 3 In fact, 55% of this cohort had a preoperative Vc that was at least as large as the predicted lower limit of normal. This may be a reflection of bronchial collateral and recanalized blood flow with maintenance of transmural vascular pressure in regions of lung completely or partially obstructed by large vessel thromboemboli. As noted previously,2•3 because CO transfer is diffusion limited rather than perfusion limited, even very limited capillary blood flow via bronchial or recanalized Clinical Investigations

pulmonary arterial sources may maintain Vc close to normal. It is surprising that Vc remained unaffected by the improvement in pulmonary arterial blood flow and hemodynamics afforded by PTE. There are a number of possible scenarios that could account for this observation. Although perioperative in situ thrombosis or embolism could be responsible, postoperative clinical analysis, hemodynamic changes, lung perfusion imaging, and postoperative pulmonary angiography indicate that this is a very rare occurrence. 29 Another possibility is that, following PTE, there is a shift of blood supply from the high transmural pressure bronchial circulation to the lower transmural pressure, higher-flow pulmonary circulation with no net change, or even a decrease, in VC· We also have described a redistribution of pulmonary blood flow away from nonoperated-on lung regions (vascular "steal") that occurs postoperatively and persists for many months. 29 This redistribution, perhaps combined with a reduction in the extensive bronchial arterial collateral blood flow present preoperatively, may play a role in the persistence of the reduction ofVc at the time of the postoperative studies. Finally, abnormalities in small pulmonary arteries ( <200 pm) have been described in CTEPH in lung regions that are not blocked by large-vessel thrombi.27·28·30 These microscopic findings are indistinguishable from those observed in patients with congenital cardiac septal defects 31 and primary pulmonary hypertension.32 Whether these small-vessel alterations are associated with changes in the ultrastructure of the lung which could impact on gas diffusion is currently under study. In conclusion, we have found that a significantly reduced DM is primarily responsible for the low Dco noted before and shortly after endarterectomy for CTEPH, and Vc does not improve in the early postoperative period despite restoration of pulmonary flow to lung segments subjected to PTE. The mechanisms responsible remain speculative. Further investigation into the lung abnormality of CTEPH may reveal the pathophysiologic basis for our observations as may long-term, serial measurement of DM and Vc. REFERENCES

1 Moser KM , Auger WR, Fedullo PF. Chronic, major vessel thromboembolic pulmonary hypertension. Circulation 1990; 81:1735-43 2 Kapitan KS , Buchbinder M, Wagner PD, eta!. Mechanisms of hypoxemia in chronic thromboembolic pulmonary hypertension. Am Rev Respir Dis 1989; 139:1149-54 3 Kapitan KS , Clausen JL, Moser KM. Gas exchange in chronic thromboembolism after pulmonary thromboendarterectomy. Chest 1990; 98:14-9 4 Gardener RM , Hankinson JL, Clausen JL, eta!. Standardization

of spirometry-1987 update. Am Rev Respir Dis 1987; 136:1285-98 5 Zarins LP, Clausen JL. Body plethysmography. In: Clausen JL, ed. Pulmonary function testing: guidelines and controversies. New York: Academic Press, 1982; 141-53 6 Morris JF, Koski A, Johnson LC. Spirometric standards for healthy nonsmoking adults. Am Rev Respir Dis 1971; 103:57-67 7 Goldman HI, Becklake MR. Respiratory function tests: nonnal values at median altitudes and the prediction of normal results. Am Rev Tuberc 1959; 79:457-67 8 Single breath carbon monoxide diffusing capacity (transfer factor): recommendations for a standard technique. Am Rev Respir Dis 1987; 136:1299-1307 9 Cotes JE, Meade F , Saunders FJ. Effect of volume inspired and manner of sampling the alveolar gas upon components of the transfer factor (diffusing capacity of the lung) by the single breath method. Proc Physiol Soc 1965; 16-17 July:73P-74P 10 Jones RS, Meade F. A theoretical and experimental analysis of anomalies in the estimation of pulmonary diffusing capacity by the single breath method. Q J Exp Physiol1961; 46:131-43 11 Cotes JE. Lung function: assessment and application in medicine. 4th ed. Oxford, UK: Blackwell, 1979 12 Starn H, Kreuzer FJA, Versprille A. Effect of lung volume and positional changes on pulmonary diffusing capacity and its components. J Appl Physiol1991; 71:1477-88 13 Roughton RFW, Forster RE. Relative importance of diffusion and chemical reaction rates in detennining rate of exchange of gases in the human lung, with special reference to true diffusing capacity of pulmonary membrane and volume of blood in the lung capillaries. J Appl Physiol 1957; 2:290-302 14 Crapo RO, Morris AH, Gardner RM. Reference values for pulmonary tissue volume, membrane diffusing capacity, and pulmonary capillary blood volume. Bull Eur Physiopath Respir 1982; 18:893-99 15 Morrison NJ, Abboud RT, Muller NL, eta!. Pulmonary capillary blood volume in emphysema. Am Rev Respir Dis 1990; 141:53-61 16 Georges R, Saumon G, Loiseau A. The relationship of age to pulmonary membrane conductance and capillary blood volume. Am Rev Respir Dis 1978; 117:1069-78 17 Lewis BM, LinT, Noe FE, eta!. The measurement of pulmonary capillary blood volume and pulmonary membrane diffusing capacity in nonnal subjects: the effects of exercise and position. J Clin Invest 1958; 37:1061-70 18 Chang SC, Chang HI, Liu SY, et a!. Effects of body position and age on membrane diffusing capacity and pulmonary capillary blood volume. Chest 1992; 102:139-42 19 Van Ganse WF, Ferris BG, Cotes JE. Cigarette smoking and pulmonary diffusing capacity (transfer factor). Am Rev Respir Dis 1972; 105:30-41 20 Mahajan KK, Mahajan SK, Mishra N. Effect of exercise on lung transfer factor for CO in smokers. Respiration 1991; 58:167-70 21 Sansores RH, Pare PD, Abboud RT. Acute effect of cigarette smoking on the carbon monoxide diffusing capacity of the lung. Am Rev Respir Dis 1992; 146:951-58 22 Frans A, Stanescu DC, Veriter C, eta!. Smoking and pulmonary diffusing capacity. Scand J Respir Dis 1975; 56:165-83 23 Fenne1ty AG, Gunawardena KA, Smith AP. The transfer factor and its subdivisions in patients with pulmonary emboli. Eur Respir J 1988; 1:98-101 24 Nadel JA, Gold WM , Burgess JH. Early diagnosis of chronic pulmonary vascular obstruction. Am J Med 1968; 44:16-25 25 Daum S. Oxygen transferring factor (diffusing capacity) of the lungs in pulmonary embolism. Bull. Physiopath Resp 1970; 6:135-45 26 Sharma GVRK, Burleson VA, Sasalmra AA. Effect of thrombolytic therapy on pulmonary capillary blood volume in patients with pulmonary embolism. N Eng! J Med 1980; 303:842-45 CHEST I 110 I 6 I DECEMBER, 1996

1435

27 Anderson E, Simon G, Reid L. Primary and thromboembolic pulmonary hypertension: a quantitative pathologic study. J Pathol 1973; 110:273-93 28 Moser KM , Bloor CM. Pulmonary vascular lesions occurring in patients with chronic major vessel thromboembolic pulmonary hypertension. Chest 1993; 103:685-92 29 Olman MA, Auger WR, Fedullo PF, et a!. Pulmonary vascular steal in chronic thromboembolic pulmonary hypertension. Chest 1990; 98:1430-34 30 Moser KM, Daily PO, Peterson, KL. Unresolved chronic throm-

1436

boembolism. In: Will J, Dawson C, Weir EK, et a!, eds. The pulmonary circulation in health and disease. New York: Harcourt Brace Jovanovich, 1987; 533-54 31 Heath D, Edwards J. The pathology of hypertensive pulmonary vascular disease: a description of six grades of structural changes in the pulmonary arteries with special reference to congenital cardiac septal defects. Circulation 1958; 18:533-44 32 Wagenvoort CA, Wagenvomt N. Primary pulmonary hypertension: a pathologic study of the lung vessels in 156 clinically diagnosed cases. Circulation 1970; 42:1163-84

Clinical Investigations