The authors reply
Just because the TPG does not include a measure of flow in its calculation (i.e., stroke volume [SV]), this does not mean it is flow “independent.” In fact, TPG is impressively flow dependent—as cardiac output declines toward zero, so does the TPG. In 1971, Harvey et al (3) first witnessed a disproportionate rise in the pulmonary pulse pressure (PPP), and therefore a rise mean pulmonary artery pressure (MPAP) and TPG, as pulmonary artery occlusion pressure (PAOP) increased (importantly in the setting of a fixed SV). Analysis from our current Fluid and Catheter Treatment Trial cohort yields similar results to Harvey et al (3) (Fig. 1): a widening PPP as PAOP rises (note: SV, 78–82 mL; n = 23). Mathematical review of the determinants of the diastolic pressure decay constant (RC time; the product of resistance and compliance), which declines with increasing PAOP (4), also illustrates that these effects are not mediated by SV as SV is included in both terms and cancels out.
The mechanism by which elevated PAOP is associated with a lower PAC for a given pulmonary vascular resistance (PVR) is less well established. Because the majority of the compliance of the lung vasculature is peripherally located (i.e., small distal arteries and arterioles), one possibility is that pulmonary venous congestion may lower the blood storage capacity of these vessels, thereby decreasing compliance. Alternatively, and considering a scenario of equal PVR, higher MPAP compared with lower MPAP may still result in lower vascular compliance due to the nonlinear pressure-diameter relation of the pulmonary arteries. It is also important to remember that both CPA and PVR are estimated using assumptions that assuredly do not reflect all complex features of multifaceted physiologic systems (5); nonetheless, these simplified variables are associated with outcome across multiple clinical conditions (left heart failure, pulmonary arterial hypertension, and now ARDS), suggesting they are useful to model behavior of complex physiologic systems and suggest hypotheses to be tested in prospective trials.
We agree that a myriad of clinical variables affect pulmonary vascular physiology in the critically ill, among them positive end-expiratory pressure as Drs. Jha and Jha (1) suggest, and balancing oxygenation with pulmonary overdistention is a key concept in protecting the right ventricle (6). We demonstrate that different clinical variables are associated with different facets of the pulmonary vasculature-ventilator driving pressure with PVR and fluid balance with CPA. Optimally protecting the right ventricular (RV) in ARDS likely requires balancing multiple variables as Drs. Jha and Jha (1) note; the vasopressor profile is another such variable that should be further studied. In summary, we agree with Drs. Jha and Jha (1) that additional and prospective studies are needed to define which aspects of RV function should be targeted therapeutically.
All authors contributed in drafting the article and critical revisions. Dr. Metkus received funding from BestDoctors (consulting), EBIX/Oakstone (consulting), and McGraw-Hill (royalties); he disclosed that he received unrestricted research funds from Abbott to support an investigator initiated study; and he received support for article research from the National Institutes of Health (NIH). Dr. Mathai received funding from consulting for Actelion, Bayer, and United Therapeutics. Dr.