Pulmonary hypertension research in the laboratory

John Gordon, MD, pediatric critical care specialist, Children's Hospital of Wisconsin, associate professor, Pediatrics (Critical Care), Medical College of Wisconsin; and Robert Jaquiss, MD, pediatric cardiothoracic surgeon, Children's Hospital of Wisconsin, assistant professor, Surgery (Cardiothoracic), Medical College of Wisconsin.

Pulmonary hypertension (PH) can be broadly classified as primary or secondary. Primary PH whether idiopathic, thromboembolic or veno-occlusive more commonly is seen in adults while secondary PH is more common in children. Typical causes of secondary PH include: 1) persistent pulmonary hypertension of the newborn (PPHN), which may be idiopathic or due to conditions such as meconium aspiration or congenital diaphragmatic hernia; 2) chronic pulmonary over-perfusion such as occurs with large atrial, ventricular or atrioventricular septal defects; and 3) chronic hypoxia such as occurs in infants and children with chronic lung disease or severe obstructive apnea. Regardless of the etiology, PH is a major cause of morbidity and mortality in neonatal and pediatric medicine. The goal of basic laboratory research is to identify strategies that may help prevent or treat clinical PH. To this end, laboratory investigators seek to identify the mechanisms underlying control of the normal pulmonary circulation, determine how these normal control mechanisms are altered during the development of PH, and test potential interventions or therapies that can prevent or reverse PH.

There are several investigators studying different aspects of secondary PH at the Medical College of Wisconsin. For example, Ganesh Konduri, MD, neonatologist at Children's Hospital of Wisconsin and associate professor of Pediatrics at the Medical College, studies factors responsible for normal and abnormal control of pulmonary vascular resistance in the fetus and newborn. Elizabeth Jacobs, MD, pulmonologist at Froedtert and Medical College Clinics and professor of Medicine at the Medical College, studies the effects of post-natal pulmonary over-perfusion on vascular reactivity. John Gordon, MD, studies chronic hypoxia-induced PH and recently Robert Jaquiss, MD, has begun studies of the relative effects of hypoxia and increased pressure on the development of chronic hypoxia-induced PH.

So how does one begin to investigate a clinical problem like PH in the laboratory? First, an animal model of human disease must be developed. During the 1960s through the 1980s, it became apparent that infants and children with a variety of airway and chronic lung diseases (obstructive apnea, bronchopulmonary dysplasia) are at risk for developing PH. A common noxious stimuli in all of these patients was sustained or repetitive hypoxia. It had been known since the 1940s and 1950s that acute hypoxia is one of the most potent causes of acute pulmonary vasoconstriction and increased pulmonary vascular resistance (PVR). These observations led to studies of the effects of chronic hypoxia on the pulmonary circulation during the 1970s and 1980s. Some of the early studies by pioneers in the field demonstrated that chronic hypoxia caused many of the classic features of PH seen in children with chronic lung disease or recurrent apnea, including increased PVR and remodeling of the vascular bed. While chronic hypoxia-induced PH in animals does not fully mimic human disease, it allows investigation of mechanisms underlying physiologic, pharmacologic and morphologic consequences of PH and yields clues about how to prevent or treat PH in infants or children.

To illustrate this, consider some of the recent and ongoing studies conducted by Medical College investigators using a newborn piglet model of chronic hypoxia-induced PH. Newborn piglets are placed in a chamber with either 10 percent O2 (to cause hypoxia-induced PH) or room air (to serve as normal controls) for between three days and three weeks. The hypoxia makes them a little lethargic on the first day, but by the second day they are generally up and playing with each other and with toys provided by the veterinary medical unit staff. After the requisite time in the chamber, the piglets are studied using different experimental preparations. The preparation that most closely resembles clinical medicine is an intact animal. After anesthesia, arterial and pulmonary artery catheters are placed. Then, in-vivo hemodynamics are measured allowing comparison of baseline PVR and responses to constrictor and dilator stimuli in PH and control piglets. Using this type of preparation, K. Jane Lee, MD, pediatric critical care fellow at Children's Hospital and instructor of Pediatrics at the Medical College, recently found that hypercapnic acidosis caused a significant increase in hypoxic PVR in both control and PH piglets. Moreover, this increase in PVR was not blocked by correcting the acidosis with bicarbonate. These data suggest that vasodilator therapy may be beneficial when using permissive hypercapnia ventilator strategies, particularly in patients with pre-existent PH. The potentially beneficial effect of vasodilator therapy in PH even in the absence of acidosis are illustrated in Figure 1 from Gordon's laboratory. Baseline and hypoxic PVR were much higher in the PH piglets compared to controls, but inhaled nitric oxide substantially reduced PVR in both.

Although intact animal studies provide confirmation of PH in-vivo and have been widely used to test potential therapies (such as nitric oxide and prostacyclin) in pre-clinical trials; it is difficult to elucidate the mechanisms underlying development of PH in intact animal preparations. Therefore, progressively more "reductionist" preparations have been developed. The most "physiological" of these is the isolated lung preparation. Animals are anesthetized deeply and the lungs are removed after euthanasia. The pulmonary artery and left atrium are cannulated and the lungs are pump-perfused with blood or a physiological saline solution (PSS). The lungs can be ventilated with any gas mixture desired (no oxygen, 20 percent CO2) and any number of different vasodilators or vasoconstrictors can be administered in this preparation without worrying about systemic effects or cardiac output. Thus, using isolated lungs, Michele VanderHeyden, MD, former pediatric critical care fellow at Children's Hospital and current assistant professor at the University of Massachussets, was able to determine that alkalosis-induced pulmonary vasodilation in piglets was caused by the combination of increased endogenous nitric oxide, increased endogenous prostacyclin, and increased calcium-activated potassium channel activity.

But even whole lungs are relatively complex. It is hard to tell whether responses are due to lung tissue, smooth muscle or endothelial cells. It is hard to know which size vessels are most important in any given response. It virtually is impossible to be sure what is happening at a cellular or molecular level within the vascular smooth muscle and endothelium. Therefore, more reductionist preparations including isolated vessels, fresh or cultured cells, fragments of cell membranes, and various sub-cellular components are studied using physiological, pharmacological and molecular biology techniques. Isolated vessels are an attractive preparation because one can collect many from a single animal. After anesthesia and euthanasia, the lungs are removed and vessels are dissected. They then can be cut into rings that are suspended from pressure transducers to measure the effects of different stimuli or chemicals on their tension. By gently removing the endothelium from the rings, one can distinguish between stimuli that require endothelium (acetylcholine) to cause vasodilation and those that are endothelium-independent. This kind of preparation has been widely used at the Medical College by many investigators. As an alternative to vessel rings, short segments of arteries or veins can be tied to tiny cannulae and their side-branches ligated under a dissecting microscope. A microscope, camera and video screen then are used to visualize the vessel and constantly measure its diameter. Increases and decreases in diameter then reflect dilation or constriction, respectively, in response to different stimuli. Using this technique, Jane Madden, PhD, professor of Neurology and Physiology at the Medical College has been able to demonstrate dilation and constriction under conditions of flow and no-flow in vessels under 0.2 mm in diameter.

Another investigator Nancy Rusch, PhD, professor of Pharmacology at the Medical College, studies potassium and calcium ion channels using patch clamp and molecular biology techniques. Patch clamping involves attaching a very fine pipette to a cell membrane from either an intact or disrupted cell then measuring the activity of selected ion channels in response to different stimuli. Studies from her lab and elsewhere have shown that activating and inactivating potassium channels can cause smooth muscle to dilate or contract, respectively. Rusch also measures the quantity and type of calcium and potassium channel protein and mRNA present in cells from animals exposed to different stimuli. For example, in a collaborative study with Gordon, she recently found that protein from calcium channels are increased in small pulmonary arteries from piglets with chronic hypoxia-induced PH (Figure 2). Since calcium influx through these channels is essential to vasoconstriction, these data suggest that one of the mechanisms involved in the development of PH is an increase in the calcium channels.

In addition to measures of hemodynamics in intact animals and various pharmacologic and physiologic responses in reductionist preparations, several imaging techniques are used in the laboratory. For example, Jaquiss recently has initiated a protocol in which he bands the left pulmonary artery of newborn piglets before they enter the hypoxia chamber. The purpose of this study is to evaluate the relative contribution of hypoxia that affects both lungs and high pressure, which only the unbanded lung to the development of hypoxia-induced PH. The first step then, is to confirm that flow is reduced to the banded lung. This is done by Christopher Dawson, PhD, professor of Physiology at the Medical College and Marquette University, and his colleagues who perform a perfusion scan to measure relative flow to each lung (Figure 3). Subsequently, the piglets are placed in hypoxia and hemodynamics, calcium channels and other measures are made after three weeks in the banded and unbanded lungs.

In addition to physiologic and pharmacologic changes, chronic hypoxia-induced PH results in vascular remodeling. This can be visualized morphologically as an increase in the amount of muscle in small and large pulmonary arteries as shown in Figure 4.

However, such pictures say little of the functional consequences of remodeling. Dawson and his group have developed a micro-CT imaging system that allows assessment of the effects of chronic hypoxia on distensibility of the pulmonary vasculature. By measuring diameters of large and small arteries at different pressures, one can better appreciate the functional consequences of chronic hypoxia. The CT image of a left lower lobe is shown in Figure 5.

It has been impossible to touch on all of the studies or mention all of the investigators contributing to our many protocols investigating chronic hypoxia-induced PH at Children's Hospital and the Medical College. Nor have we discussed exciting work in the areas of PPHN and high flow-induced PH conducted by Konduri and Jacobs and their colleagues. However, we hope this overview provides some insight into the breadth of laboratory work on PH. We invite any colleagues who are interested in further discussion, wish to collaborate in future studies or would like to observe some of these experiments to contact us.