Radiographic Inversion of Pulmonary Blood Flow in Acute Asthma
Bahman Joorabchi, M.D., M.Ed.
Emad Hammoude, M.D.
Mahmood A. Khalid, M.D.
Summary:
Chest radiographs are obtained routinely in children hospitalized with asthma but are considered to be of low utility. We describe a previously unreported radiographic inversion of pulmonary artery flow patterns in the chest films of a majority of children and young adults hospitalized for asthma. Seventy-five hospital records of 65 patients aged 2 to 25 years ( mean 9.2) admitted for asthma during the calendar year 1991 were reviewed . As compared with the normal pattern, patients with inverted vascularity were younger ( mean age 6.75 years), tended to be admitted through the emergency department, and had significantly more signs of labored breathing, such as tachypnea, retractions, nasal flaring, and tachycardia. We propose that the exaggerated respiratory effort inthese children raises the transmural pulmonary artery pressure, impending pulmonary blood flow. This results in dissension of the right ventricle, Which then compresses the left ventricle in the confines of the pericardium by posterior displacement of the interventricular septum. The resultant loss of left-sided compliance reases the left atrial and pulmonary venous pressures, leading to inversion of vascular markings through the same mechanism as seen in left-heart failure. We conclude that attention to this easily recognized sign will detect presence of the hemodynamically significant respiratory obstruction in asthmatic patients.
Introduction:
Chest radiographs are commonly obtained during the initial assessment of patients with severe asthma. There is however, no agreement as to their utility. Studies of children admitted for intractable asthma and of those treated in the emergency room have concluded that routine radiographs are not justified because of their low Yield of abnormal findings. This is probable true if one is searching for gross abnormalities such as pneumothorax, aterlectasis, and pneumonia.
Review of chest radiographs , as part of another study, revealed a high incidence of inversion of pulmonary vascular markings in plain upright films of children with asthma. The purpose of this study was to determine the incidence of this finding in the chest films of children and young adults hospitalized for asthma and to correlate it with their clinical findings.
Methods
The medical records of children and young adults hospitalized for asthma during the calendar year 1991 were reviewed. Hospitalized patients were selected, since there would be more data available for over a longer period of time . The following data were extracted age, sex, date, time of initial assessment, route of admission, vital signs and mental status, presence or absence of retrations , nasal flaring, and cyanosis, quality of wheezing and air exchange, nature and duration of treatment, and when available results of arterial blood gases.
The initial chest radiographs were blindly reviewed by one of us (M.K.) Noting date and time of the study. The films were divided into three groups, according to whether the distribution of vascular markings was normal ( more prominent in the lower lobes), even, or inverted ( more prominent in the upper lobes). The decision for grading of vascular distribution was based on :
Results
During the calendar year 1991, 65 patients aged 2 to 25 years ( mean age 9.2 years) were admitted to North Oakland Medical Center with the diagnosis of acute-severe or life-threatening asthma. Six patients had two hospitalizations each, and two were hospitalized three times. Thus, there were 75 charts for review. Every patient had upright posterior-anterior (PA) and lateral chest films that were obtained within 2 hours of the time of initial physical assessment, as recorded either in the emergency department or, for those admitted directly, in the ward.
The patient population of the study was divided into three groups according to the distribution of pulmonary vascular pattern on the upright chest radiographs. Group 1, consisting of 11 patients ( 15%) had a normal pattern with a more prominent vascularity in the lower lobes. In group 2, with 36 patients (48%), the vascular pattern was evenly distributed. The 28 patients ( 37%) in group 3 had an inverted distribution, with the vascularity being more prominent in the upper portions of the lungs. Thus, a significant majority of these patients ( 85%) had recognizable changes in the distribution of their pulmonary blood flow.
Figure 1 shows an example of normal pulmonary vascular distribution ( group 1) in a 17-year-old woman admitted through a physicians office for acute asthma. On admission, the respiratory rate was 22 breaths/min and heart rate 90 beats/min, there was bilateral wheezing but no retrations or flaring.
Figure 2 is a film of a 2-year-old boy admitted through the emergency room for life-threatening asthma with a respiratory rate of 44 breaths/min, and heart rate of 160 beats/min, he had flaring , retractions, and poor air entry. The radiograph shows obvious signs of hyperaeration. The upper-lobe vessels are more prominent than the lower ones, the reverse of normal (group 3).
By alternately focusing on the upper and lower segments exclusively, e.g., by covering and uncovering them in turn, one is better able to judge the vascular distribution This is demonstrated in Figure 3 showing inversion of the vascular markings.
The inversion is readily detectable in lateral projections , as is seen in Figure 4.
Figure 5 shows two films from the same patient, a 9-year-old boy, taken during two different hospitalizations 9 months apart. The left film shows inversion of the markings, and was taken in the emergency room. The respiration were 32 breaths/min and the heart rate 116 beats/min. There were marked retractions, flaring and poor air exchange. He stayed in the hospital for 4 days and required 26 inhalation therapies. The film on the right shows an even distribution of the markings ( group 2) and was taken en route to the ward from the physician’s office. His respiration were 24 breaths/min and heart rate was 96 beats/min. There were mild retractions, and no flaring and fair air exchange. He stayed for 2 days and received 12 respiratory therapy sessions.
On initial assessment, only two patients were noted to be clinically cyanotic and mentally obtunded. One of these had to be intubated. A third was noted to be agitated. All three belonged to group 3. There were no differences in systolic blood pressures. All patients had wheezing on initial examination. Table 1 shows the distinguishing clinical features of the three groups. As can be seen, a significant majority of this group of patients ( 85%) had recognizable disturbance of their pulmonary blood-flow pattern. The patients with an inverted pattern ( group 3) were younger ( but did not differ in sex distribution). There was a strong correlation between inversion of pulmonary blood flow pattern on chest radiographs and signs of labored respiration. The patients showing an even distribution ( group 2) occupied an intermediate position.
In order to control for the effect of age in the vascular distribution-possibly due to higher chest compliance in the young patients were devided into two groups : those 6 years and younger were compared with those 7 and older. In the younger age group, the same relationships existed those with inverted pulmonary vascular markings had significantly more retraction, flaring, tachypnea, and tachycardia than those with a normal distribution. In older group, the differences in these signs of respiratory obstruction were also significant, except for tachypnea.
The patients in group 3 tended to be admitted through the emergency room ( 71%, as compared to 27% for group 1, P= 0.03), received more inhalation therapy during their stays, but their lengths of stay were not longer ( mena of 2.8 days vs 2.5 for group 1). There were no differences among the groups on terms of proportions receiving oxygen therapy, steroids, or theophylline. Arterial blood gases were available in 20 patients.
These were evenly distributed among the three groups. Comparison of the blood gases in this relatively small subgroup showed that the group 3 patients with inverted pulmonary blood flow were significantly more hypoxic than those with normal patterns ( mean Po2 64 vs 77 mm Hg, P=0.01). There was a tendency toward hypocapnia ( mean Pco2= 29 mm Hg) and respiratory alkalosis ( mean pH of 7.45) in group 3 patients, but the difference was not statistically significant. It should be noted that there was a wider distribution of Pco2 in the group 3 patients ( 18 to 64 mm Hg ) than in the others ( 31 to 37): some were hypocapnic while others had CO2 retention.
There were no major abnormalities noted in any of the chest films. Minor abnormalities, such as hyperinflation, patchy infiltrates, and mild chronic changes were noted in 27 of the 75 examinations ( 36%). Group 3 patients tended to have more of these ( 54% as compared to 27% for group1), but the difference did not reach statistical significance. The inversion of markings was more noticeable on the left side in a quarter of patients in group 3.
All eight patients with multiple hospitalizations had at least some disturbance of the vascular pattern. Of the six patients with two hospitalizations each, two had the same vascular pattern both times ( one pair inverted, the other even). The other four had different patterns in different hospitalizations. The two patients with three admissions each had inverted vascularity once and an even pattern twice.
In summary , the presence of inverted pulmonary vascular markings was an indication of significant respiratory obstruction. Even distribution of markings suggested an embarrassment of lesser degree. The inversion of markings tended to occur in the young, this reflected the severity at the onset of the illness, but did not predict a lengthy or complicated hospital course.
Discussion
Inversion of vascular markings in upright chest radiographs (“cephalization”) is a well-known phenomenon in acute and chronic left heart failure. Normally, in the upright position, the vessels in the lower lobes of the lungs, both arteries and veins, are more prominent than those in the upper zones. This is largely because of the gravitational hydrostatic pressure difference that directs a substantially larger volume of the blood to the lower lobes. In left heart failure, the increased left artial and pulmonary venous pressure places the lower lobes at higher risk for pulmonary edema. The selective vasoconstriction of lower regions redirects the blood to the upper zones. The wries advanced to explain this inversion of the flow included reflex pulmonary vasoconstriction from stretch receptors in the lower lobes veins, or from greater alveolar hypoxia in the lower lobes and the higher compression of the vessels from the dependent interstitial edema.
Cardiovascular effects of bronchial obstruction have been controversial . Studies in infants and young children with bronchiolitis have variously concluded that there is and there is not pulmonary hypertension in this disease. In children and adults with asthma, pulmonary hypertension in this disease. In children and adults with asthma, pulmonary hypertension is reported only rarely. Pulmonary artery pressure was found to be “normal” in adult asthmatics with electrocardiographic changes of right-sided overload ) P pulmonale). The confusion probably stems from the fact that pulmonary artery pressures were measured against the atmospheric pressure. Yet if the transmural pulmonary artery pressure- i.e., the gradient between pulmonary artery and intrathoracic cavity-is measured, the effective resistance to pulmonary artery flow would in deed be high. Intrathoracic ( esophageal) pressures in asthmatic children reach very high negative values ( up to -40 cm H2O at peak inspiration, with the mean tidal pressures of -7 to -28) and remain essentially negative throughout the respiratory cycle.
Indeed, transmural pulmonary artery and right-artial pressures ( absolute minus intrathoracic pressure) have been shown to be markedly increased with heightened respiratory effort in asthma. Such an increase in right-sided pressures will tend to distend the right ventricle, flattening and displacing the intervetricular septum posteriorly. The competition for space within the confines of the pericardium would compress the left ventricle, reducing its volume and increasing the left-artial filling pressure. These changes have been demonstrated in asthmatic patients, in cases of right ventricular overload from other causes, in acute respiratory failure, and during the Mueller maneuver ( forced inspiration against a closed glottis) in normal subjects. The resulting increase in the pulmonary venous pressure sets off the same chain of events in the distribution of pulmonary blood flow described in left-heart failure. There is, of course, no heart failure, per se, in that the pulmonary venous distention is caused by left ventricular compression and not through pump failure. The fact that, in some of the patients, the inversion of the flow was more noticeable on the left side than the right, is probably due to the leftward direction of the main pulmonary artery and is in deeping with previous observations. The changes in pulmonary hemodynamics with respiration would be expected to be cyclic, being most prominent during inspiration, but the overall effect would persist- especially with tachypnea. In protracted cases, other factors may intervene, notably the development of interstitial edema. This edema would overcome the tethering action of the ling parenchyma and compress the pulmonary vessels- both arteries and veins- again, to a greater extent in the lower lobes. In extreme cases , the development of significant hypoxia may effect left ventricular function directly and increase left-sided filling pressures even further. This was not seen in our patients: there were no signs of hear failure, and patients tended to recover quickly.
Conclusion
Close observation of the distribution of vascular markings in the plain upright chest radiographs of children and young adults with asthma, obtained at the height of their symptoms, may provide an index of severity of the episode. We are not suggesting that chest films be taken routinely in all asthmatic patients. Other less expensive clinical tests, such as measurement of paradoxical pulse and peak flow rates, should be obtained more regularly. However, if chest films are obtained for other indications this useful sign should not be overlooked. Further, presence of inverted markings , indirect evidence of pulmonary venous hypertension, has implications as to the benefits of sitting or semi-reclining positions and avoidance of overhydration. Finally, even though this study was limited to asthmatic patients, the same sign may be detectable in patients with other disorders characterized by an increased respiratory effort.
Acknowledgments
We wish to extend a strong note of appreciation to Dr. Victor Maldonado, pediatric radiologist at the St. Joseph Mercy Hospital, for invaluable review, critique, and guidance.