Hemodynamic deterioration in chronic venous disease J o h n F. Welkie, M D , A n t h o n y J. Comerota, M D , Mira L. Katz, M L A , RVT, Samuel C. Aldridge, M D , R o b b P. Kerr, BA, RVT, and John V. White, M D ,
Philadelphia, Pa. Clinical deterioration of patients with chronic venous disease (CVD) has been well described and a standardized classification has been proposed. The progressive hemodynamic deterioration producing these clinical findings is less well appreciated. This study examines and correlates venous hemodynamics with clinical severity in patients with CVD. Two hundred seventy-four extremities from 149 patients with varying degrees of CVD and 56 extren~'ties from 28 symptom-free volunteers were evaluated clinically and hemodynamically. Each limb was assessed for fimctional venous volume, degree of valvular insufficiency, efficiency of the calf muscle pump, and noninvasive estimate of ambulatory venous pressure. In addition, exercise venous pressures were recorded in 56 extremities from 36 patients and 9 extremities from 6 volunteers. As CVD progresses from class 0 to class 2, venous volume expands, valvular function deteriorates, the calf muscle pump becomes inefficient, and ambulatory venous hypertension develops. However, once extremities develop brawny edema or hyperpigmentation, further deterioration of limb hemodynamics does not occur. Patients with deep venous obstruction have more severe val~atlar insufficiency, calf muscle pump dysfunction, and ambulatory venous hypertension than have patients without evidence of obstruction. Residual volume fraction offers a reliable noninvasive estimate of ambulatory venous pressure (r = 0.76), although its correlation was significantly better for patients without venous obstruction (r = 0.86) than for those with obstruction (r = 0.40; p < 0.05). Deterioration in venous hemodynamics parallels clinical severity through class 2. Once brawny edema and hyperpigmenration occur, ulceration develops without additional deterioration of venous hemodynamics. (J VASC SURG 1992;16:733-40.)
Insight into the basic pathophysiology of chronic venous insufficiency began with the description of venous valves by Fabricius 1 in the early 1600s and the clinical observations and speculations of Gay2 during the 1800s. More recently, the progressive signs and symptoms of chronic venous disease (CVD) have been attributed to ambulatory venous hypertension. 3'* Although the exact relationship between ambulatory venous hypertension and CVD has been questioned, venous pressure measurements remain the hemodynamic gold standard for the evaluation of patients with CVD. s7 From the Vascular Laboratory,Department of Surgery,Temple UniversityHospital, Philadelphia. Supported in part by US Public Health Servicesgrant RR00349, GeneralClinicalResearchCentersBranch,NationalInstitutesof Health. Presented at the Fourth AnnualMeetingof the AmericanVenous Forum, Coronado, Calif., Feb. 26-28, 1992. Reprint requests: Anthony J. Comerota, MD, Department of Surgery, Temple UniversityHospital, Broad and Ontario St., Philadelphia, PA 19140. 24/1/40619
Two major factors contributing to exercise venous hypertension are venous outflow obstruction and valvular insufficiency. The importance of a wellfunctioning calf muscle pump is also frequently discussed but has not been studied thoroughly. In addition, the interrelationship of each of these factors with clinical deterioration has not been elucidated. Recent technical advances now permit a more complete and accurate evaluation of venous hemodynamics, assessing both obstruction and insufficiency.8 It has been suggested that ambulatory venous pressure (AVP) can be calculated from lower extremity volume changes that occur with changes in position and after exercise? The purposes of this study are to examine hemodynamic alterations associated with clinically progressive CVD; to evaluate whether there is a characteristic insult or specific hemodynamic deterioration responsible for causing some patients with edematous, pigmented extremities to have ulcers; to evaluate the impact of chronic venous obstruction on 733
lournal of VASCULAR SURGERY
734 Wdkie et al.
Table I. Patient demographics and results of venous duplex imaging History
Class
No. EXT
No. patients
M/F (no. EXT)
0
94
66
50/44
1
109
77
49/60
2
67
49
37/30
Mean age (yr)
Age range (yr)
35.5 (p < 0.01") 48.2 (NS*) 52.4
20-76
History of DVT n
%
(+) VDI
of D VT or (+) VDI
n
n
0
% 0
% 0
19-77
9
8.3
9
8.3
11
10.1
23-80
22
32.8
19
28.4
24
35.8
(p = o.oolf)
22-77
23
38.3
19
31.7
24
40.0
(p = o.ooost)
(NS*) 3
60
49
37/23
49.1
EXT, Extremities; DVT, deep venous thrombosis; VD/, venous duplex imaging; NS, not statistically significant (p > 0.05). *Wilcoxon rank sum test for nonparametric distributions. tPearson's ×2 test (compared with class 1).
limb hemodynamics; and to determine whether AVP can be predicted accurately by a noninvasive method. MATERIAL
AND
METHODS
S u b j e c t s . Two hundred seventy-four lower extremities from 149 patients with varying degrees of CVD were evaluated prospectively during an 18month period from July 1990 to December 1991. In addition, 56 limbs from 28 symptom-free volunteers were evaluated similarly. A medical history was obtained and each extremity was examined for signs of CVD. Extremities were classified according to the guidelines established by the Ad Hoc Committee For Reporting Standards of the Society for Vascular Surgery and the International Society for Cardiovascular Surgery.I° Because this classification is based on clinical signs and symptoms, it includes patients with primary valvular insufficiency and those with the postthrombotic syndrome. Each patient gave informed consent before evaluation. The number of extremities evaluated in each class of CVD, as well as patient demographics, are listed in Table I. Class 0 represents extremities without clinical evidence of CVD. This group consisted of 56 extremities from 28 symptom-free volunteers and 38 asymptomatic extremities from 38 patients included in the study, whereas 74% (111/149) of patients had bilateral disease. The mean age for the volunteers was 30.9 years and the mean age for patients with unilateral CVD was 42.5 years. The latter value was not significantly different from those with bilateral disease (mean age 50.7 years) (Table 1I). Class 1 represents mild disease, with extremities typically showing mild swelling or dilated subcutaneous veins. This group included 109 extremities from 7 7 patients. Class 2 signifies moderately severe disease, with the presence of significant edema or
hyperpigmentation. In this group there were 67 extremities from 49 patients. Finally, extremities with the most severe manifestations of CVD were placed into class 3 and included 60 extremities with active or healed venous ulcers from 49 patients. Air plethysmography. All extremities were evaluated by air plethysmography. The air plethysmograph (ACI Medical Inc., Sun Valley, Calif.) is a pneumatic volume plethysmograph consisting of a polyurethane cuff that surrounds the leg from the knee to the ankle. The cuffis connected to a pressure transducer and the signal is amplified and then processed by an analog chart recorder. Calibration of the instnmaent to each patient's extremity enables the detection of volume changes that occur in response to postural change and exercise. Subsequent calculations are then made from these volume changes. Details of this technique and interpretation of the criteria have been published previously.9,u Measured or calculated values include the functional venous volume of the extremity (VV), venous filling time 90 (VFT90), venous filling index (VFI), ejection fraction, and residual volume fraction. The VV is the increase in leg volume (in milliliters) that occurs when the patient stands from the supine position with the leg elevated. This parameter varies according to patient and extremity size; therefore differences in VV alone may not be particularly instructive when comparing patient populations. The VFT90 is the amount oftime required to fill 90% of the W . Because of the exponential nature of the venous refilling curve, it is easier and more accurate to identify a point on the curve (90% of the refill time) as opposed to the point when complete filling occurs. The VF is defined as 90% VV/VFT90 and is
Volume 16 Number 5 November 1992
independent of patient and extremity size. This value represents the average rate of venous filling, which is a result of arterial inflow and venous valvular reflux. As such, it provides an overall assessment of the degree of valvular insufficiency. The ejection fraction is the percent of the VV that is expelled from the calf by one tiptoe exercise and assesses the efficiency of the calf muscle pump. Finally, the residual volume fraction represents the percent of the VV remaining in the calf after 10 tiptoe exercises. This value has been shown to correlate closely with AVP. 8'9'12 Venous pressure measurements. Venous pressures were recorded in 56 extremities from 36 patients and 9 extremities from 6 symptom-free volunteers. A 23-gauge butterfly needle was inserted into a dorsal foot vein and connected to a pressure transducer, amplifier, and strip chart recorder. Pressures were recorded with the patient in the supine and standing positions and after exercise. The AVP was defined as the lowest mean pressure recorded after 10 tiptoe exercises. Venous duplex imaging. The 274 extremities from patients with lower extremity complaints were evaluated by venous duplex imaging (Biosound 2000 I I S A ; Biosound Inc., Indianapolis, Ind.). The common femoral, superficial femoral, popliteal, anterior tibial, and posterior tibial veins were examined with patients in the reverse Trendelenburg position. Veins were evaluated for acute thrombosis, as well as for evidence of chronic obstruction (thickened walls or recanalization). No attempt was made to identify or quantitate valvular insufficiency with venous duplex imaging. Statistical analysis. Statistical analysis was performed with the CLINFO II program on a MAcroVMS computer (Digital Equipment Corp. Maynard, Mass.). Mean values for the various parameters were compared with the Wilcoxon rank sum test for nonparametric distributions (Tables I through IV). Pearson's X2 test was used to compare the incidences of deep venous thrombosis between classes 1, 2, and 3. The correlation of residual volume fraction with AVP for extremities with and without obstruction were compared with Fisher's transformed Z test. RESULTS The results of the study are summarized in Tables 1 through IV and Figs. 1 through 3. Patient demographics and results of venous duplex imaging are shown in Table I. A minority of extremities in classes 1 through 3 had a history of or showed duplex imaging evidence of previous deep venous thrombo-
Hemodynamic deterioration in venous disease 735
sis, consistent with other reports.~S,~4The percentage was significantly higher for classes 2 (p = 0.001) and 3 (p --- 0.0005) compared with class 1. Most of the patients evaluated in class 1 had uncomplicated varicose veins without previous deep venous thrombosis. Patients with postphlebitic extremities infrequently seek medical attention until significant signs or symptoms develop, and therefore the opportunity to evaluate these patients is often missed when they are clinically in class 1. The mean age of patients with class 1, 2, or 3 extremities was similar, whereas patients in class 0 extremities were younger (p < 0.01). Class 0 consisted of two subsets: extremities from symptom-free volunteers and asymptomatic extremities from patients with contralateral C VD. The younger age of the volunteers reduced the overall mean age for class 0. However, as expected, the mean age of the second subset of extremities was not different from that of the remaining patients in classes 1, 2, and 3. The two subsets of extremities constituting class 0 were compared with each other, and no hemodynamic differences were observed (Table II). Significant differences were observed in all of the hemodynamic parameters between classes 0, 1, and 2, with the exception of VV between classes 0 and 1 and ejection fraction between classes 1 and 2 (Table III). The VV increased steadily from class 0 to class 2 with a significant increase between classes 1 and 2. Empirically, this appears logical because factors that tend to increase venous capacitance (i.e., chronic outflow obstruction and valvular insufficiency) would tend to become more prominent in the latter stages of CVD. This also fits the clinical picture of the swollen extremity seen in classes 2 and 3. The VFT90 decreased from class 0 to class 2 (p < 0.0001), whereas the VFI increased during the same clinical interval (p < 0.0001). These findings suggest progressive valvular dysfunction that parallels clinical deterioration to the point of swelling and hyperpigmentation. The values for VFT90 and VFI also may reflect to some degree the increased arterial inflow that has been demonstrated in extremities with venous ulcers or pigment c h a n g e s , is,16 Calf ejection fraction significantly decreased between classes 0 and 1 but did not diminish further despite progressive clinical deterioraton. This suggests that calf muscle pump dysfunction occurs early in patients with C VD. Increases in residual volume fraction paralleled the measured ambulatory venous hypertension that occurred as extremities deteriorated from class 0 to class 2. The mean values for these parameters
736
~ourn£ of vASCULAR SURGERY
Welkie et al.
Table II. Patient demographics and hemodynamic data for the two subgroups of class 0
Class
No. EXT
No. patients
Mean age (yr)
Age range (yr)
M/F (no. EXT)
VV (ml)
VFT90 (sec)
VFI (ml/sec)
0 (volunteers) 0 (patients)
56 38
28 38
30.9 42.5
24-62 20-76
28/28 22/16
130.5 + 4.7 125.2 + 7.5
79.1 _+ 6.4 87:1 ± 8.2
1.48 _+ 0.15 1.57 ± 0.18
NS
NS
NS
p Valuer
< 0.01
EXT, Extremities; EF, ejection fraction; RVF, residual volume fraction; NS, not statistically significant (p > 0.05). *n = 9 and 6 for the volunteer and patient groups, respectively. tWilcoxon rank sum test for nonparametric distributions.
Table III. Air plethysmographic and venous pressure data for the various classes of CVD VV (ml)
VFT90 (sec)
VFI (ml/sec)
EF (%)
R VF (%)
A VP* (mm Hg)
128.4 ± 4.2 (NSt) 142.7 + 6.7 (p < O.O01t) 195.1 -+ 10.4 (NSt) 187.2 +- 10.3
83.9 _+ 5.1 (p < O.O01t) 67.3 ± 4.6 (p < 0.0001t) 29.4 + 2.1 (NSt) 26.0 ± 2.2
1.52 -+ 0.11 (p < 0.001I-) 2.74 +- 0.26 ~ < 0.0001") 7.86 _+ 0.80 (NS1-) 8.28 +_ 0.73
65.6 + 3.2 (p < O.O01t) 51.7 +_ 1.6 (NSt) 48.5 _+ 2.3 (NSt) 45.2 -4- 2.1
16.1 + 2.3 (p < O.O01t) 37.6 _+ 2.8 (p < 0.02t) 46.2 -+ 2.7 (NSt) 49.0 -- 2.3
16.5 ± 2.1 (p < 0.02t) 32.9 -+ 4.1 ~ < 0.021-) 46.6 ---4.7 (NSt) 53.6 +- 4.6
Class 0 (n = 94) 1 (n = 109) 2 (n = 67) 3 (n = 60)
Values are mean ± SEM.
EF, Ejection fraction; RVF, residual volume fraction; NS, not statistically significant (p < 0.05). *n = 15, 15, 12, and 23 for classes 0 to 3, respectively. tWilcoxon rank sum test for nonparametric distributions.
Table IV. Air plethysmographic and venous pressure data for symptomatic classes of CVD,
categorized according to presence ( + ) or absence ( - ) of a history or duplex evidence of previous deep venous thrombosis Class
No. EXT
VV (ml)
VFT90 (sec)
VFI (ml/sec)
EF (%)
R VF (%)
1 ( - ) DVT
98
1 (+) DVT 2 (-) DVT
11 43
2 (+) DVT 3 (-) DVT
24 36
3 (+) DVT
24
145.9 + 7.0 (NS*) 117.7 _+ 21.6 195.4 _-2 12.0 (NS*) 196.4 + 19.6 187.9 +_ 11.0 (NS*) 187.4 -+ 21.0
69.8 + 4.9 (NS*) 44.3 _+ 6.1 32.5 -+ 2.5 (p < 0.01") 23.8 + 3.8 26.9 + 2.6 (NS*) 24.6 +-- 4.2
2.80 + 0.28 (NS*) 2.66 + 0.52 6.87 + 0.82 (p < 0.02*) 9.63 - 0.80 8.09 + 0.89 (NS*) 8.56 -+ 1.16
52.0 + 1.8 (NS*) 48.6 + 6.5 49.2 +- 2.7 (NS*) 47.2 + 4.3 48.6 _ 2.5 (p < 0.05*) 38.5 -+ 3.4
36.4 + 2.7 (NS*) 48.5 +_ 7.8 46.9 -+ 3.5 (NS*) 45.0 + 4.5 48.7 + 3.4 (NS*) 49.5 ± 4.4
A VP (mm Hg) 32.0 + 4.3 (n = 13) (NS*) 39.0 (n = 2) 44.9 -+- 8.1 (n = 7) (NS*) 49.0 - 4.4 (n = 5) 49,1 -+ 5.2 (n = 13) (NS*) 59.5 _+ 7.5 (n = 10)
Hemodynamic data are mean _+ SEM. EXT, Extremities; EF, ejection fraction; RVF, residual volume fraction; DVT, deep venous thrombosis; NS, not statistically significant
~o > 0.05). *Wilcoxon rank sum test for nonparametric distributions.
correlated well for each class of CVD. Overall, residual volume fraction provided a satisfactory estimation of AVP (r = 0.76), which agrees with previous reports (Fig. 1). 8'9"12 Extremities with duplex imaging evidence of obstruction showed a significantly poorer correlation between residual volume fraction and AVP (r = 0.40) compared with those extremities with no evidence of obstruction (r = 0.86); (p < 0.05)
(Figs. 2 and 3). Hemodynamic parameters were compared between extremities with and without chronic venous obstruction for each class of CVD, and there was a trend toward worse hem0dynamics in those with obstruction. Because of the smaller sample size for obstructed extremities, the differences became statistically significant only with respect to VFT90 and VFI in class 2 and ejection fraction in class 3.
Volume 16 Number 5 November 1992
Hemodynamic deterioration in venous disease 7 3 7
EF (%)
R VF (%)
A VP* (mm Hg)
63.8 -+ 4.7 69.6 _+ 4.3
13.9 -+ 2.8 18.1 -_+ 3.2
16.3 _+ 2.6 16.7 +_ 4.1
NS
NS
NS
Surprisingly, there were no significant hemodynamic differences between the edematous, hyperpigmerited limbs (class 2) and those with venous ulcers (class 3) (Table III). Therefore no additional hemodynamic deterioration occurred after an extremity reached clinical class 2. DISCUSSION
Progressive hemodynamic deterioration that parallels increasingly severe CVD is intuitively acceptable and has been recognized previously. 3,9a7 However, these studies did not simultaneously evaluate multiple hemodynamic parameters in patients with progressive CVD. Patients with venous ulceration may have combinations of poor valvular function, deep venous obstruction, diminishcd calf muscle pump fimction, and ambulatory venous hypertension. 3,s7,1saT,18 A previously unanswered question was, "What hemodynamic changes occur in a patient with a swollen, pigmented limb to cause skin breakdown?" We were surprised to find no differences in venous macrohemodynamics between the latter two clinical groups. A possible explanation for progressive clinical deterioration might be ongoing changes at the microcirculatory level. The arteriolarvenom reflux is lost in class 2 and 3 extremities, producing capillary hypertension. ~9,2° Altered capillary permeability, progressive pericapillary fibrin deposition, and diminished regional fibrinolysis lead to lipodermatosclerosis, dermal hypoxia, and subsequent ulceration. 18,21"23White blood cells have also been implicated in the development of venous ulcers. Leukocytes become trapped in extremities with advanced venom disease and may occlude or further alter the permeability of capillaries and small lymphatic channels. Activation of these leukocytes may cause direct tissue injury after the release of various proteolytic enzymes, superoxide radicals, and chemot a c t i c s u b s t a n c e s . 2426
Significant hemodynamic deterioration occurred as patients progressed from being symptom free to the development of severe swelling and hyperpigmentation (class 2). These progressive changes oc-
curred with VFT90, VFI, residual volume fraction, and AVP. Although the ejection fraction was significantly worse compared wtih asymptomatic limbs, the major dysfunction of the calf muscle pump occurs early in the course of CVD. Despite significant changes in all other hemodynarnic variables, additional calf muscle pump failure did not occur unless the patient had obstruction of the deep venous system. It appears that normal function of the calf muscle pump is dependent on venous valvular integrity and a nonobstructed deep venous system. Any abnormality that occurs, either obstruction or valvular insufficiency, adversely affects the efficiency of the calfmmcle pump early in the clinical course of CVD. However, it is evident that the calf muscle pump is capable of maintaining its ejection capability despite deterioration in other hemodynamic parameters and perhaps is responsible for halting clinical progression in some patients. Obstruction of the deep venous system leads to more severe hemodynamic dysfunction per clinical dass of disease and was significantly worse for patients with class 2 or 3 disease. Obstruction was associated with more valvular dysfunction in patients with class 2 disease and worse calf muscle pump function in patients with venous ulceration. This corroborates the observations of Shull et al.,27 who demonstrated that for comparable degrees of valvular insufficiency, those postthrombotic limbs with phlebographically proved venom obstruction had higher AVPs compared with those without obstruction. These observations may have implications for the treatment of patients with acute deep venous thrombosis. Patients in this analysis did not undergo specific hemodynamic evaluation of their obstructive component. The standard method of assessing venous obstruction is by maximal venous outflow, 8,28which has been shown to be insensitive. 29Maximum venous outflow techniques are fundamentally inconsistent and theoretically incapable of offering an appropriate assessment of patients with CVD, because these methods are performed with patients at rest, whereas the underlying hemodynamic abnormality surfaces during exercise. 4 Therefore hemodynamic evaluation of venous obstruction in patients with CVD should occur during exercise. Raju a° has proposed a technique correlating arm-leg pressure differences. Nicolaides s has calculated venous outflow resistance by simultaneously measuring venous volume and pressure changes over time with the air plethysmograph. Both methods have merit and appear to improve on the more commonly used techniques; however, each
738
Journal of VASCULAR SURGERY
Welkie et al.
751
[] o./.I
0 r = 0.76
0
~ T
•
•
•
I Class0 o Class i
•
Class 2 [] Class 3
•
-25 0
20
40
60
BO
1O0
A V P (ram Hg) Fig. 1. Correlation of residual volume fraction (RVF) with AVP for all subjects. Solid boxes represent class 0, empty circles represent class 1, solid circles represent class 2, and empty boxes represent class 3. Correlation coefficient is 0.76.
100
75 RVF
S
~B 00_
50
~
[]
0 0© 25
-o.~#
I .,na~m 0 I/~'m •
r=086 i
o ~iass 1
•
•
| m2 ~
[]
class
o
_
•
Class2 Class3
t 0
20
40 AVP
60
80
1O0
( r a m Hg)
Fig. 2. Correlation of residual volume fraction (RVF) with AVP for extremities without evidence of deep venous obstruction. Solid boxes represent class 0, empty circles represent class 1, solid circles represent class 2, and empty boxes represent class 3. Correlation coefficient improves to 0.86. needs to be corroborated by other investigators. Our experience with these techniques is too preliminary to offer conclusions. Therefore we relied on visual demonstration of luminal compromise by venous duplex imaging to identify patients with venous
obstruction. All patients in this series who gave a history of deep venous thrombosis had acute, symptomatic deep venous thrombosis. Although deep venous thrombosis does not always result in chronic obstruction, 72% of patients who gave this history
Volume 16 Number 5 November 1992
Hemodynamic deterioration in venous disease 739
100
[]
75
RVF
: o j
50
[]
(%) 25
o 0
r = 0.40
•
o Class 1 • Class 2 [] Class 3 -25 0
20
40
AVP
60
80
100
(ram Hg)
Fig. 3. Correlation of residual volume fraction (RVF) with AVP for extremities with deep venous obstruction. Empty circles represent class 1, solid circles represent class 2, and empty boxes represent class 3. Correlation coefficient fails to 0.40.
had duplex imaging evidence of obstruction. Conversely, 89% of extremities that showed obstruction on duplex imaging had a history of prior deep venous thrombosis. The correlation of residual volume fraction with AVP (r = 0.76) in our patients was somewhat worse than that found by Christopoulos et al.9 and Nicolaides et al. (r = 0.83)8,:z This may be caused in part by the fact that in most of our patients the air plethysmography tests and venous pressure measurements were performed separately although on the same day. We have observed variation in patient performance during the tiptoe portion of both tests, even when performed within minutes of each other. The diminished correlation may be the result of accurately measuring the two parameters at separate points in time, as opposed to a truly poor correlation of simultaneously obtained values. The correlation improved when the tests were performed simultaneously, although only a small number of our patients were evaluated in this fashion. In addition, a poor correlation existed between residual volume fraction and AVP in extremities with venous obstruction (r = 0.40). Obstruction changes the pressure/volume relationship during exercise, requiring higher venous pressures to maintain venous return. The correlation of residual volume fraction and AVP was very good in patients without obstruc-
tion (r = 0.86) and was significantly better than in patients with obstruction (p < 0.05). Clinicians and investigators now have the opportunity to evaluate patients more completely, identify the specific disease, and quantitate its contribution to the patients' underlying hemodynamic dysfunction. All patients treated for CVD should have complete hemodynamic assessment before and after any operative or nonoperative therapy, to determine objectively the therapeutic benefit. Integration of these data with accurate anatomic delineation of the venous system should enable clinicians to predict the therapeutic merit of any planned intervention for patients with CVD. REFERENCES
1. Fabricius HA. Anatomici Patavini de Venarum Ostiolis (valvesof veins). LaurentiusPasquatus, 1603; Padua, 45-56. In: LaufmanH, ed. The veins.Austin:Silvergirl,1986:14-25. 2. Gay J. Varicose disorders of the lower extremities: the Lettsomianlecture of 1863. In: LaufmanI{, ed. The veins. Austin: Silvergirl, 1986:122-31. 3. NicolaidesAN, ZukowskiAJ. The value of dynamicvenous pressure measurements.World J Surg 1986;10:919-25. 4. SumnerDS. Hemodynamicsand pathophisiologyof venous disease. In: Rutherford RB, ed. Vascular surgery. 3rd ed. Philadelphia: WB Saunders, 1989;1483-504. 5. RajuS, FredericksR. Hemodynamicbasisofstasisulceration: a hypothesis.5 VASeSURG1991;13:491-5. 6. SchanzerH, ConversePierce E. Pathophysiologicevaluation
740
7. 8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
j'ournalof VASCULAR SURGERY
Welkie et al.
of chronic venous insufficiency with anabulatory venous pressures. Angiolo~, 1982;33:183-91. Welkie JF, Comerota AJ, Kerr RP, et al. The hemodvnamics of venous ulceration. Ann Vasc Surg 1992;6:1-4. Nicolaides AN. Diagnostic evaluation of patients with chronic venous insufficientT. In: Rutherford RB, ed. Vascular surgery.. 3rd ed. Philadelphia: WB Saunders, 1989:1583-601. Christopoulos D, Nicolaldes AN, Szendro G, et al. Airplethysmography and the effect of elastic compression on venous hemodynamics of the legs. J VAS("SUF,Cd1987;5:14859. Porter JM, Rutherford RB, Clagett GP, et al. Reporting standards in venous disease. J VAS(" SURG 1988;8:172-81. Katz ML, Comerota AJ, Kerr R. Air-plcthysmography (APG'rM): a new technique to evaluate patients with chronic venous insufficiency. J Vasc Technol 1991 ; 15:23-7. Nicolaides AN, Christopouk)s D, Vasdekis S. Progress in the investigation of chronic venous insufficiency. Ann Vase Surg 1989;3:278-92. Raju S, Fredericks R. Valve reconstruction procedure for nonobstructive venous insufficiency: rationale, techniques, and results in 107 procedures with 2- to 8-year follow-up. J VASCStJV,G 1988;7:301-10. Hanrahan LH, Araki CT, Rodriguez AA, et al. Distribution of valvular incompetence in patients with venous stasis ulceration. J VAS("St;kc; 1990;13:805-12. Christopoulos D, Nicolaides AN, Szendro G. Venous reflux: quantitation and correlation with severity of chronic venous disease. Br J Surg 1988;75:352-6. Christopoulos DC, Nicotaides AN, Bclcaro G, et al. Venous hypertensive microangiopathy in relation to clinical severity and effect of elastic compression. J Dermatol Surg Oncol 1991;17:809-13. McEnroe CS, O'Donnell TF Jr; Mackey VC. Correlation of clinical findings with venous hemodynamics in 386 patients with chronic venous insufficiency. Am J Surg 1988;156:14853. Browse NL, Burnand KG. The cause of venous ulceration. Lancet 1982;2:243-5.
19. Belcaro G, Grigg M, Vasdekis S, et al. Evaluation of the effect of elastic compression in patients with postphlebitic limbs by laser Doppler flowmetrv. Phlebology 1988;41:797-802. 20. Belcaro G, Grigg M, Rulo A, et al. Blood flow in peri malleolar skin in relation to posture in patients with venous hypertension. Ann Vase Surg 1989;1:5-7. 21. Burnand KG, Whimster I, Naidoo A, et al. Peticapillary fibrin in the ulcer bearing skin of the leg: the cause of lipodermatosclerosis and venous ulceration. Br Med J 1982;285:1071-5. 22. Partsch H. Hyperemic hypoxia in venous ulcers. Br J Dermatol 1984;110:249-53. 23. Franzeck UK, Bollinger A, ltirch R, et al. Transcutaneous oxygen tensions and capillary"morphologic characteristics and density in patients with chronic venous incompetence. Circulation 1984;70:806-10. 24. Thomas PRS, Nash GB, Dormandy JA. White cell accumulation in the dependent legs of patients with venous h xq~ertcnsion:a t~ssible mechanism for trophic changes in the skin. Br Med J 1988;296:1693-5. 25. Bollinger A, Isenring G, Franzeck UK. Lymphatic microangiopathy: a complication of severe chronic venous incompetence. Lymphology 1982;15:60-5. 26. Coleridge-Smith PD, Thomas P, Scurr ]It, et al. Causes of venous ulceration: a new h~othesis. Br Med J 1988;296: 1726-32. 27. Shull KC, Nicolaides AN', Fernandes e Fernandes J, et al. Significance of [n)pliteal reflux in relation to ambulatory venous pressure and ulceration. Arch Surg 1979; 114:1304-6. 28. Barne,~ R, Collicott PE, Moz~erski DJ, et al. Noninvasive quantitation of maximum venous outflow in acute thrombophlebitis. SurgeD, 1972;72:791-9. 29. Comerota AJ, Katz ML, Grossi RI, et al. The comparative value of noninvasive tests for diagnosis and surveillance of deep venous thrombosis. J VASCSURG 1988;7:40-9. 30. Raju S. New approaches to the diagaaosis and treatment of venous obstruction. I VASe SUt
Philadelphia, Pa. Clinical deterioration of patients with chronic venous disease (CVD) has been well described and a standardized classification has been proposed. The progressive hemodynamic deterioration producing these clinical findings is less well appreciated. This study examines and correlates venous hemodynamics with clinical severity in patients with CVD. Two hundred seventy-four extremities from 149 patients with varying degrees of CVD and 56 extren~'ties from 28 symptom-free volunteers were evaluated clinically and hemodynamically. Each limb was assessed for fimctional venous volume, degree of valvular insufficiency, efficiency of the calf muscle pump, and noninvasive estimate of ambulatory venous pressure. In addition, exercise venous pressures were recorded in 56 extremities from 36 patients and 9 extremities from 6 volunteers. As CVD progresses from class 0 to class 2, venous volume expands, valvular function deteriorates, the calf muscle pump becomes inefficient, and ambulatory venous hypertension develops. However, once extremities develop brawny edema or hyperpigmentation, further deterioration of limb hemodynamics does not occur. Patients with deep venous obstruction have more severe val~atlar insufficiency, calf muscle pump dysfunction, and ambulatory venous hypertension than have patients without evidence of obstruction. Residual volume fraction offers a reliable noninvasive estimate of ambulatory venous pressure (r = 0.76), although its correlation was significantly better for patients without venous obstruction (r = 0.86) than for those with obstruction (r = 0.40; p < 0.05). Deterioration in venous hemodynamics parallels clinical severity through class 2. Once brawny edema and hyperpigmenration occur, ulceration develops without additional deterioration of venous hemodynamics. (J VASC SURG 1992;16:733-40.)
Insight into the basic pathophysiology of chronic venous insufficiency began with the description of venous valves by Fabricius 1 in the early 1600s and the clinical observations and speculations of Gay2 during the 1800s. More recently, the progressive signs and symptoms of chronic venous disease (CVD) have been attributed to ambulatory venous hypertension. 3'* Although the exact relationship between ambulatory venous hypertension and CVD has been questioned, venous pressure measurements remain the hemodynamic gold standard for the evaluation of patients with CVD. s7 From the Vascular Laboratory,Department of Surgery,Temple UniversityHospital, Philadelphia. Supported in part by US Public Health Servicesgrant RR00349, GeneralClinicalResearchCentersBranch,NationalInstitutesof Health. Presented at the Fourth AnnualMeetingof the AmericanVenous Forum, Coronado, Calif., Feb. 26-28, 1992. Reprint requests: Anthony J. Comerota, MD, Department of Surgery, Temple UniversityHospital, Broad and Ontario St., Philadelphia, PA 19140. 24/1/40619
Two major factors contributing to exercise venous hypertension are venous outflow obstruction and valvular insufficiency. The importance of a wellfunctioning calf muscle pump is also frequently discussed but has not been studied thoroughly. In addition, the interrelationship of each of these factors with clinical deterioration has not been elucidated. Recent technical advances now permit a more complete and accurate evaluation of venous hemodynamics, assessing both obstruction and insufficiency.8 It has been suggested that ambulatory venous pressure (AVP) can be calculated from lower extremity volume changes that occur with changes in position and after exercise? The purposes of this study are to examine hemodynamic alterations associated with clinically progressive CVD; to evaluate whether there is a characteristic insult or specific hemodynamic deterioration responsible for causing some patients with edematous, pigmented extremities to have ulcers; to evaluate the impact of chronic venous obstruction on 733
lournal of VASCULAR SURGERY
734 Wdkie et al.
Table I. Patient demographics and results of venous duplex imaging History
Class
No. EXT
No. patients
M/F (no. EXT)
0
94
66
50/44
1
109
77
49/60
2
67
49
37/30
Mean age (yr)
Age range (yr)
35.5 (p < 0.01") 48.2 (NS*) 52.4
20-76
History of DVT n
%
(+) VDI
of D VT or (+) VDI
n
n
0
% 0
% 0
19-77
9
8.3
9
8.3
11
10.1
23-80
22
32.8
19
28.4
24
35.8
(p = o.oolf)
22-77
23
38.3
19
31.7
24
40.0
(p = o.ooost)
(NS*) 3
60
49
37/23
49.1
EXT, Extremities; DVT, deep venous thrombosis; VD/, venous duplex imaging; NS, not statistically significant (p > 0.05). *Wilcoxon rank sum test for nonparametric distributions. tPearson's ×2 test (compared with class 1).
limb hemodynamics; and to determine whether AVP can be predicted accurately by a noninvasive method. MATERIAL
AND
METHODS
S u b j e c t s . Two hundred seventy-four lower extremities from 149 patients with varying degrees of CVD were evaluated prospectively during an 18month period from July 1990 to December 1991. In addition, 56 limbs from 28 symptom-free volunteers were evaluated similarly. A medical history was obtained and each extremity was examined for signs of CVD. Extremities were classified according to the guidelines established by the Ad Hoc Committee For Reporting Standards of the Society for Vascular Surgery and the International Society for Cardiovascular Surgery.I° Because this classification is based on clinical signs and symptoms, it includes patients with primary valvular insufficiency and those with the postthrombotic syndrome. Each patient gave informed consent before evaluation. The number of extremities evaluated in each class of CVD, as well as patient demographics, are listed in Table I. Class 0 represents extremities without clinical evidence of CVD. This group consisted of 56 extremities from 28 symptom-free volunteers and 38 asymptomatic extremities from 38 patients included in the study, whereas 74% (111/149) of patients had bilateral disease. The mean age for the volunteers was 30.9 years and the mean age for patients with unilateral CVD was 42.5 years. The latter value was not significantly different from those with bilateral disease (mean age 50.7 years) (Table 1I). Class 1 represents mild disease, with extremities typically showing mild swelling or dilated subcutaneous veins. This group included 109 extremities from 7 7 patients. Class 2 signifies moderately severe disease, with the presence of significant edema or
hyperpigmentation. In this group there were 67 extremities from 49 patients. Finally, extremities with the most severe manifestations of CVD were placed into class 3 and included 60 extremities with active or healed venous ulcers from 49 patients. Air plethysmography. All extremities were evaluated by air plethysmography. The air plethysmograph (ACI Medical Inc., Sun Valley, Calif.) is a pneumatic volume plethysmograph consisting of a polyurethane cuff that surrounds the leg from the knee to the ankle. The cuffis connected to a pressure transducer and the signal is amplified and then processed by an analog chart recorder. Calibration of the instnmaent to each patient's extremity enables the detection of volume changes that occur in response to postural change and exercise. Subsequent calculations are then made from these volume changes. Details of this technique and interpretation of the criteria have been published previously.9,u Measured or calculated values include the functional venous volume of the extremity (VV), venous filling time 90 (VFT90), venous filling index (VFI), ejection fraction, and residual volume fraction. The VV is the increase in leg volume (in milliliters) that occurs when the patient stands from the supine position with the leg elevated. This parameter varies according to patient and extremity size; therefore differences in VV alone may not be particularly instructive when comparing patient populations. The VFT90 is the amount oftime required to fill 90% of the W . Because of the exponential nature of the venous refilling curve, it is easier and more accurate to identify a point on the curve (90% of the refill time) as opposed to the point when complete filling occurs. The VF is defined as 90% VV/VFT90 and is
Volume 16 Number 5 November 1992
independent of patient and extremity size. This value represents the average rate of venous filling, which is a result of arterial inflow and venous valvular reflux. As such, it provides an overall assessment of the degree of valvular insufficiency. The ejection fraction is the percent of the VV that is expelled from the calf by one tiptoe exercise and assesses the efficiency of the calf muscle pump. Finally, the residual volume fraction represents the percent of the VV remaining in the calf after 10 tiptoe exercises. This value has been shown to correlate closely with AVP. 8'9'12 Venous pressure measurements. Venous pressures were recorded in 56 extremities from 36 patients and 9 extremities from 6 symptom-free volunteers. A 23-gauge butterfly needle was inserted into a dorsal foot vein and connected to a pressure transducer, amplifier, and strip chart recorder. Pressures were recorded with the patient in the supine and standing positions and after exercise. The AVP was defined as the lowest mean pressure recorded after 10 tiptoe exercises. Venous duplex imaging. The 274 extremities from patients with lower extremity complaints were evaluated by venous duplex imaging (Biosound 2000 I I S A ; Biosound Inc., Indianapolis, Ind.). The common femoral, superficial femoral, popliteal, anterior tibial, and posterior tibial veins were examined with patients in the reverse Trendelenburg position. Veins were evaluated for acute thrombosis, as well as for evidence of chronic obstruction (thickened walls or recanalization). No attempt was made to identify or quantitate valvular insufficiency with venous duplex imaging. Statistical analysis. Statistical analysis was performed with the CLINFO II program on a MAcroVMS computer (Digital Equipment Corp. Maynard, Mass.). Mean values for the various parameters were compared with the Wilcoxon rank sum test for nonparametric distributions (Tables I through IV). Pearson's X2 test was used to compare the incidences of deep venous thrombosis between classes 1, 2, and 3. The correlation of residual volume fraction with AVP for extremities with and without obstruction were compared with Fisher's transformed Z test. RESULTS The results of the study are summarized in Tables 1 through IV and Figs. 1 through 3. Patient demographics and results of venous duplex imaging are shown in Table I. A minority of extremities in classes 1 through 3 had a history of or showed duplex imaging evidence of previous deep venous thrombo-
Hemodynamic deterioration in venous disease 735
sis, consistent with other reports.~S,~4The percentage was significantly higher for classes 2 (p = 0.001) and 3 (p --- 0.0005) compared with class 1. Most of the patients evaluated in class 1 had uncomplicated varicose veins without previous deep venous thrombosis. Patients with postphlebitic extremities infrequently seek medical attention until significant signs or symptoms develop, and therefore the opportunity to evaluate these patients is often missed when they are clinically in class 1. The mean age of patients with class 1, 2, or 3 extremities was similar, whereas patients in class 0 extremities were younger (p < 0.01). Class 0 consisted of two subsets: extremities from symptom-free volunteers and asymptomatic extremities from patients with contralateral C VD. The younger age of the volunteers reduced the overall mean age for class 0. However, as expected, the mean age of the second subset of extremities was not different from that of the remaining patients in classes 1, 2, and 3. The two subsets of extremities constituting class 0 were compared with each other, and no hemodynamic differences were observed (Table II). Significant differences were observed in all of the hemodynamic parameters between classes 0, 1, and 2, with the exception of VV between classes 0 and 1 and ejection fraction between classes 1 and 2 (Table III). The VV increased steadily from class 0 to class 2 with a significant increase between classes 1 and 2. Empirically, this appears logical because factors that tend to increase venous capacitance (i.e., chronic outflow obstruction and valvular insufficiency) would tend to become more prominent in the latter stages of CVD. This also fits the clinical picture of the swollen extremity seen in classes 2 and 3. The VFT90 decreased from class 0 to class 2 (p < 0.0001), whereas the VFI increased during the same clinical interval (p < 0.0001). These findings suggest progressive valvular dysfunction that parallels clinical deterioration to the point of swelling and hyperpigmentation. The values for VFT90 and VFI also may reflect to some degree the increased arterial inflow that has been demonstrated in extremities with venous ulcers or pigment c h a n g e s , is,16 Calf ejection fraction significantly decreased between classes 0 and 1 but did not diminish further despite progressive clinical deterioraton. This suggests that calf muscle pump dysfunction occurs early in patients with C VD. Increases in residual volume fraction paralleled the measured ambulatory venous hypertension that occurred as extremities deteriorated from class 0 to class 2. The mean values for these parameters
736
~ourn£ of vASCULAR SURGERY
Welkie et al.
Table II. Patient demographics and hemodynamic data for the two subgroups of class 0
Class
No. EXT
No. patients
Mean age (yr)
Age range (yr)
M/F (no. EXT)
VV (ml)
VFT90 (sec)
VFI (ml/sec)
0 (volunteers) 0 (patients)
56 38
28 38
30.9 42.5
24-62 20-76
28/28 22/16
130.5 + 4.7 125.2 + 7.5
79.1 _+ 6.4 87:1 ± 8.2
1.48 _+ 0.15 1.57 ± 0.18
NS
NS
NS
p Valuer
< 0.01
EXT, Extremities; EF, ejection fraction; RVF, residual volume fraction; NS, not statistically significant (p > 0.05). *n = 9 and 6 for the volunteer and patient groups, respectively. tWilcoxon rank sum test for nonparametric distributions.
Table III. Air plethysmographic and venous pressure data for the various classes of CVD VV (ml)
VFT90 (sec)
VFI (ml/sec)
EF (%)
R VF (%)
A VP* (mm Hg)
128.4 ± 4.2 (NSt) 142.7 + 6.7 (p < O.O01t) 195.1 -+ 10.4 (NSt) 187.2 +- 10.3
83.9 _+ 5.1 (p < O.O01t) 67.3 ± 4.6 (p < 0.0001t) 29.4 + 2.1 (NSt) 26.0 ± 2.2
1.52 -+ 0.11 (p < 0.001I-) 2.74 +- 0.26 ~ < 0.0001") 7.86 _+ 0.80 (NS1-) 8.28 +_ 0.73
65.6 + 3.2 (p < O.O01t) 51.7 +_ 1.6 (NSt) 48.5 _+ 2.3 (NSt) 45.2 -4- 2.1
16.1 + 2.3 (p < O.O01t) 37.6 _+ 2.8 (p < 0.02t) 46.2 -+ 2.7 (NSt) 49.0 -- 2.3
16.5 ± 2.1 (p < 0.02t) 32.9 -+ 4.1 ~ < 0.021-) 46.6 ---4.7 (NSt) 53.6 +- 4.6
Class 0 (n = 94) 1 (n = 109) 2 (n = 67) 3 (n = 60)
Values are mean ± SEM.
EF, Ejection fraction; RVF, residual volume fraction; NS, not statistically significant (p < 0.05). *n = 15, 15, 12, and 23 for classes 0 to 3, respectively. tWilcoxon rank sum test for nonparametric distributions.
Table IV. Air plethysmographic and venous pressure data for symptomatic classes of CVD,
categorized according to presence ( + ) or absence ( - ) of a history or duplex evidence of previous deep venous thrombosis Class
No. EXT
VV (ml)
VFT90 (sec)
VFI (ml/sec)
EF (%)
R VF (%)
1 ( - ) DVT
98
1 (+) DVT 2 (-) DVT
11 43
2 (+) DVT 3 (-) DVT
24 36
3 (+) DVT
24
145.9 + 7.0 (NS*) 117.7 _+ 21.6 195.4 _-2 12.0 (NS*) 196.4 + 19.6 187.9 +_ 11.0 (NS*) 187.4 -+ 21.0
69.8 + 4.9 (NS*) 44.3 _+ 6.1 32.5 -+ 2.5 (p < 0.01") 23.8 + 3.8 26.9 + 2.6 (NS*) 24.6 +-- 4.2
2.80 + 0.28 (NS*) 2.66 + 0.52 6.87 + 0.82 (p < 0.02*) 9.63 - 0.80 8.09 + 0.89 (NS*) 8.56 -+ 1.16
52.0 + 1.8 (NS*) 48.6 + 6.5 49.2 +- 2.7 (NS*) 47.2 + 4.3 48.6 _ 2.5 (p < 0.05*) 38.5 -+ 3.4
36.4 + 2.7 (NS*) 48.5 +_ 7.8 46.9 -+ 3.5 (NS*) 45.0 + 4.5 48.7 + 3.4 (NS*) 49.5 ± 4.4
A VP (mm Hg) 32.0 + 4.3 (n = 13) (NS*) 39.0 (n = 2) 44.9 -+- 8.1 (n = 7) (NS*) 49.0 - 4.4 (n = 5) 49,1 -+ 5.2 (n = 13) (NS*) 59.5 _+ 7.5 (n = 10)
Hemodynamic data are mean _+ SEM. EXT, Extremities; EF, ejection fraction; RVF, residual volume fraction; DVT, deep venous thrombosis; NS, not statistically significant
~o > 0.05). *Wilcoxon rank sum test for nonparametric distributions.
correlated well for each class of CVD. Overall, residual volume fraction provided a satisfactory estimation of AVP (r = 0.76), which agrees with previous reports (Fig. 1). 8'9"12 Extremities with duplex imaging evidence of obstruction showed a significantly poorer correlation between residual volume fraction and AVP (r = 0.40) compared with those extremities with no evidence of obstruction (r = 0.86); (p < 0.05)
(Figs. 2 and 3). Hemodynamic parameters were compared between extremities with and without chronic venous obstruction for each class of CVD, and there was a trend toward worse hem0dynamics in those with obstruction. Because of the smaller sample size for obstructed extremities, the differences became statistically significant only with respect to VFT90 and VFI in class 2 and ejection fraction in class 3.
Volume 16 Number 5 November 1992
Hemodynamic deterioration in venous disease 7 3 7
EF (%)
R VF (%)
A VP* (mm Hg)
63.8 -+ 4.7 69.6 _+ 4.3
13.9 -+ 2.8 18.1 -_+ 3.2
16.3 _+ 2.6 16.7 +_ 4.1
NS
NS
NS
Surprisingly, there were no significant hemodynamic differences between the edematous, hyperpigmerited limbs (class 2) and those with venous ulcers (class 3) (Table III). Therefore no additional hemodynamic deterioration occurred after an extremity reached clinical class 2. DISCUSSION
Progressive hemodynamic deterioration that parallels increasingly severe CVD is intuitively acceptable and has been recognized previously. 3,9a7 However, these studies did not simultaneously evaluate multiple hemodynamic parameters in patients with progressive CVD. Patients with venous ulceration may have combinations of poor valvular function, deep venous obstruction, diminishcd calf muscle pump fimction, and ambulatory venous hypertension. 3,s7,1saT,18 A previously unanswered question was, "What hemodynamic changes occur in a patient with a swollen, pigmented limb to cause skin breakdown?" We were surprised to find no differences in venous macrohemodynamics between the latter two clinical groups. A possible explanation for progressive clinical deterioration might be ongoing changes at the microcirculatory level. The arteriolarvenom reflux is lost in class 2 and 3 extremities, producing capillary hypertension. ~9,2° Altered capillary permeability, progressive pericapillary fibrin deposition, and diminished regional fibrinolysis lead to lipodermatosclerosis, dermal hypoxia, and subsequent ulceration. 18,21"23White blood cells have also been implicated in the development of venous ulcers. Leukocytes become trapped in extremities with advanced venom disease and may occlude or further alter the permeability of capillaries and small lymphatic channels. Activation of these leukocytes may cause direct tissue injury after the release of various proteolytic enzymes, superoxide radicals, and chemot a c t i c s u b s t a n c e s . 2426
Significant hemodynamic deterioration occurred as patients progressed from being symptom free to the development of severe swelling and hyperpigmentation (class 2). These progressive changes oc-
curred with VFT90, VFI, residual volume fraction, and AVP. Although the ejection fraction was significantly worse compared wtih asymptomatic limbs, the major dysfunction of the calf muscle pump occurs early in the course of CVD. Despite significant changes in all other hemodynarnic variables, additional calf muscle pump failure did not occur unless the patient had obstruction of the deep venous system. It appears that normal function of the calf muscle pump is dependent on venous valvular integrity and a nonobstructed deep venous system. Any abnormality that occurs, either obstruction or valvular insufficiency, adversely affects the efficiency of the calfmmcle pump early in the clinical course of CVD. However, it is evident that the calf muscle pump is capable of maintaining its ejection capability despite deterioration in other hemodynamic parameters and perhaps is responsible for halting clinical progression in some patients. Obstruction of the deep venous system leads to more severe hemodynamic dysfunction per clinical dass of disease and was significantly worse for patients with class 2 or 3 disease. Obstruction was associated with more valvular dysfunction in patients with class 2 disease and worse calf muscle pump function in patients with venous ulceration. This corroborates the observations of Shull et al.,27 who demonstrated that for comparable degrees of valvular insufficiency, those postthrombotic limbs with phlebographically proved venom obstruction had higher AVPs compared with those without obstruction. These observations may have implications for the treatment of patients with acute deep venous thrombosis. Patients in this analysis did not undergo specific hemodynamic evaluation of their obstructive component. The standard method of assessing venous obstruction is by maximal venous outflow, 8,28which has been shown to be insensitive. 29Maximum venous outflow techniques are fundamentally inconsistent and theoretically incapable of offering an appropriate assessment of patients with CVD, because these methods are performed with patients at rest, whereas the underlying hemodynamic abnormality surfaces during exercise. 4 Therefore hemodynamic evaluation of venous obstruction in patients with CVD should occur during exercise. Raju a° has proposed a technique correlating arm-leg pressure differences. Nicolaides s has calculated venous outflow resistance by simultaneously measuring venous volume and pressure changes over time with the air plethysmograph. Both methods have merit and appear to improve on the more commonly used techniques; however, each
738
Journal of VASCULAR SURGERY
Welkie et al.
751
[] o./.I
0 r = 0.76
0
~ T
•
•
•
I Class0 o Class i
•
Class 2 [] Class 3
•
-25 0
20
40
60
BO
1O0
A V P (ram Hg) Fig. 1. Correlation of residual volume fraction (RVF) with AVP for all subjects. Solid boxes represent class 0, empty circles represent class 1, solid circles represent class 2, and empty boxes represent class 3. Correlation coefficient is 0.76.
100
75 RVF
S
~B 00_
50
~
[]
0 0© 25
-o.~#
I .,na~m 0 I/~'m •
r=086 i
o ~iass 1
•
•
| m2 ~
[]
class
o
_
•
Class2 Class3
t 0
20
40 AVP
60
80
1O0
( r a m Hg)
Fig. 2. Correlation of residual volume fraction (RVF) with AVP for extremities without evidence of deep venous obstruction. Solid boxes represent class 0, empty circles represent class 1, solid circles represent class 2, and empty boxes represent class 3. Correlation coefficient improves to 0.86. needs to be corroborated by other investigators. Our experience with these techniques is too preliminary to offer conclusions. Therefore we relied on visual demonstration of luminal compromise by venous duplex imaging to identify patients with venous
obstruction. All patients in this series who gave a history of deep venous thrombosis had acute, symptomatic deep venous thrombosis. Although deep venous thrombosis does not always result in chronic obstruction, 72% of patients who gave this history
Volume 16 Number 5 November 1992
Hemodynamic deterioration in venous disease 739
100
[]
75
RVF
: o j
50
[]
(%) 25
o 0
r = 0.40
•
o Class 1 • Class 2 [] Class 3 -25 0
20
40
AVP
60
80
100
(ram Hg)
Fig. 3. Correlation of residual volume fraction (RVF) with AVP for extremities with deep venous obstruction. Empty circles represent class 1, solid circles represent class 2, and empty boxes represent class 3. Correlation coefficient fails to 0.40.
had duplex imaging evidence of obstruction. Conversely, 89% of extremities that showed obstruction on duplex imaging had a history of prior deep venous thrombosis. The correlation of residual volume fraction with AVP (r = 0.76) in our patients was somewhat worse than that found by Christopoulos et al.9 and Nicolaides et al. (r = 0.83)8,:z This may be caused in part by the fact that in most of our patients the air plethysmography tests and venous pressure measurements were performed separately although on the same day. We have observed variation in patient performance during the tiptoe portion of both tests, even when performed within minutes of each other. The diminished correlation may be the result of accurately measuring the two parameters at separate points in time, as opposed to a truly poor correlation of simultaneously obtained values. The correlation improved when the tests were performed simultaneously, although only a small number of our patients were evaluated in this fashion. In addition, a poor correlation existed between residual volume fraction and AVP in extremities with venous obstruction (r = 0.40). Obstruction changes the pressure/volume relationship during exercise, requiring higher venous pressures to maintain venous return. The correlation of residual volume fraction and AVP was very good in patients without obstruc-
tion (r = 0.86) and was significantly better than in patients with obstruction (p < 0.05). Clinicians and investigators now have the opportunity to evaluate patients more completely, identify the specific disease, and quantitate its contribution to the patients' underlying hemodynamic dysfunction. All patients treated for CVD should have complete hemodynamic assessment before and after any operative or nonoperative therapy, to determine objectively the therapeutic benefit. Integration of these data with accurate anatomic delineation of the venous system should enable clinicians to predict the therapeutic merit of any planned intervention for patients with CVD. REFERENCES
1. Fabricius HA. Anatomici Patavini de Venarum Ostiolis (valvesof veins). LaurentiusPasquatus, 1603; Padua, 45-56. In: LaufmanH, ed. The veins.Austin:Silvergirl,1986:14-25. 2. Gay J. Varicose disorders of the lower extremities: the Lettsomianlecture of 1863. In: LaufmanI{, ed. The veins. Austin: Silvergirl, 1986:122-31. 3. NicolaidesAN, ZukowskiAJ. The value of dynamicvenous pressure measurements.World J Surg 1986;10:919-25. 4. SumnerDS. Hemodynamicsand pathophisiologyof venous disease. In: Rutherford RB, ed. Vascular surgery. 3rd ed. Philadelphia: WB Saunders, 1989;1483-504. 5. RajuS, FredericksR. Hemodynamicbasisofstasisulceration: a hypothesis.5 VASeSURG1991;13:491-5. 6. SchanzerH, ConversePierce E. Pathophysiologicevaluation
740
7. 8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
j'ournalof VASCULAR SURGERY
Welkie et al.
of chronic venous insufficiency with anabulatory venous pressures. Angiolo~, 1982;33:183-91. Welkie JF, Comerota AJ, Kerr RP, et al. The hemodvnamics of venous ulceration. Ann Vasc Surg 1992;6:1-4. Nicolaides AN. Diagnostic evaluation of patients with chronic venous insufficientT. In: Rutherford RB, ed. Vascular surgery.. 3rd ed. Philadelphia: WB Saunders, 1989:1583-601. Christopoulos D, Nicolaldes AN, Szendro G, et al. Airplethysmography and the effect of elastic compression on venous hemodynamics of the legs. J VAS("SUF,Cd1987;5:14859. Porter JM, Rutherford RB, Clagett GP, et al. Reporting standards in venous disease. J VAS(" SURG 1988;8:172-81. Katz ML, Comerota AJ, Kerr R. Air-plcthysmography (APG'rM): a new technique to evaluate patients with chronic venous insufficiency. J Vasc Technol 1991 ; 15:23-7. Nicolaides AN, Christopouk)s D, Vasdekis S. Progress in the investigation of chronic venous insufficiency. Ann Vase Surg 1989;3:278-92. Raju S, Fredericks R. Valve reconstruction procedure for nonobstructive venous insufficiency: rationale, techniques, and results in 107 procedures with 2- to 8-year follow-up. J VASCStJV,G 1988;7:301-10. Hanrahan LH, Araki CT, Rodriguez AA, et al. Distribution of valvular incompetence in patients with venous stasis ulceration. J VAS("St;kc; 1990;13:805-12. Christopoulos D, Nicolaides AN, Szendro G. Venous reflux: quantitation and correlation with severity of chronic venous disease. Br J Surg 1988;75:352-6. Christopoulos DC, Nicotaides AN, Bclcaro G, et al. Venous hypertensive microangiopathy in relation to clinical severity and effect of elastic compression. J Dermatol Surg Oncol 1991;17:809-13. McEnroe CS, O'Donnell TF Jr; Mackey VC. Correlation of clinical findings with venous hemodynamics in 386 patients with chronic venous insufficiency. Am J Surg 1988;156:14853. Browse NL, Burnand KG. The cause of venous ulceration. Lancet 1982;2:243-5.
19. Belcaro G, Grigg M, Vasdekis S, et al. Evaluation of the effect of elastic compression in patients with postphlebitic limbs by laser Doppler flowmetrv. Phlebology 1988;41:797-802. 20. Belcaro G, Grigg M, Rulo A, et al. Blood flow in peri malleolar skin in relation to posture in patients with venous hypertension. Ann Vase Surg 1989;1:5-7. 21. Burnand KG, Whimster I, Naidoo A, et al. Peticapillary fibrin in the ulcer bearing skin of the leg: the cause of lipodermatosclerosis and venous ulceration. Br Med J 1982;285:1071-5. 22. Partsch H. Hyperemic hypoxia in venous ulcers. Br J Dermatol 1984;110:249-53. 23. Franzeck UK, Bollinger A, ltirch R, et al. Transcutaneous oxygen tensions and capillary"morphologic characteristics and density in patients with chronic venous incompetence. Circulation 1984;70:806-10. 24. Thomas PRS, Nash GB, Dormandy JA. White cell accumulation in the dependent legs of patients with venous h xq~ertcnsion:a t~ssible mechanism for trophic changes in the skin. Br Med J 1988;296:1693-5. 25. Bollinger A, Isenring G, Franzeck UK. Lymphatic microangiopathy: a complication of severe chronic venous incompetence. Lymphology 1982;15:60-5. 26. Coleridge-Smith PD, Thomas P, Scurr ]It, et al. Causes of venous ulceration: a new h~othesis. Br Med J 1988;296: 1726-32. 27. Shull KC, Nicolaides AN', Fernandes e Fernandes J, et al. Significance of [n)pliteal reflux in relation to ambulatory venous pressure and ulceration. Arch Surg 1979; 114:1304-6. 28. Barne,~ R, Collicott PE, Moz~erski DJ, et al. Noninvasive quantitation of maximum venous outflow in acute thrombophlebitis. SurgeD, 1972;72:791-9. 29. Comerota AJ, Katz ML, Grossi RI, et al. The comparative value of noninvasive tests for diagnosis and surveillance of deep venous thrombosis. J VASCSURG 1988;7:40-9. 30. Raju S. New approaches to the diagaaosis and treatment of venous obstruction. I VASe SUt
