Can't Catch Your Breath? Calcium Channel Blockers and Pulmonary Shunting - A Deep Dive

 

Periodically, a patient on the unit on a nicardipine or clevidipine infusion develops what appears to be unexplainable hypoxia. The situations are inevitably multifactorial, but often end with a workup for pneumonia, PE, or volume overload turning up dry and a question to me about whether it could have been the nicardipine all along. Despite the frequency in which continuous infusion CCBs are used in the NeuroICU this doesn't come all that often, so my answer often boils down to "well, theoretically" - vasodilators, including CCBs and nitroprusside¹ (see also here²), have been demonstrated to be capable of reversing hypoxia-induced pulmonary vasoconstriction, so it is perfectly within the realm of possibility that nicardipine could contribute to pulmonary shunting and thus new-onset hypoxia. The adverse effects are only briefly mentioned in the package inserts of either nicardipine or clevidipine, with major Phase III trials not reporting it at all. Because of the high frequency of use and low frequency of these events, however, I have always wondered exactly how big of a problem this really is, what patients are at the highest risk for it, and if there is anything we can do besides turning off the drip (which sometimes is not an option in our particularly fragile patients) and I decided to take a deep dive into the literature to see if I could find an answer.

Key Points Calcium channel blockers like nicardipine, clevidipine, and nimodipine can inhibit hypoxic pulmonary vasoconstriction, counteracting the physiologic response to poorly oxygenated lung units and leading to shunt physiology Clinically meaningful hypoxia with CCBs is rare, and generally coincides with some other pulmonary insult (atelectasis, pulmonary edema, pneumonia, etc) The effect is dose related, but it’s not clear if different calcium channel blockers cause shunting more than others. If shunting occurs with nicardipine, you could try clevidipine and vice versa, but an antihypertensive that does not cause pulmonary vasodilation (like labetalol) may be preferable Because the clinical effects of CCB-associated shunting is reliant on the underlying state of the lungs, you can re-trial a CCB in a patient who previously became hypoxic if their lungs are now otherwise healthy

 

First, let's describe the kind of patient this happens to. I think this is best done through some illustrative case reports, of which there are six currently published in the literature. 

Illustrative Case Reports

Case 1: Nimodipine

Our first case report³ comes from a familiar author to many of us in the critical care pharmacy space and describes an unfortunate case of a 63 year old man who was receiving nimodipine for a traumatic subarachnoid hemorrhage after an acute decompensation which led to the discovery of substantially expanded subarachnoid blood.  After the first dose of nimodipine 60mg, the patient's oxygen saturation declined to 68% with an accompanying decline in PaO2 from 114.0 to 32.9 mmHg. This resolved within 20 minutes after increasing FiO2 to 100% and PEEP to 7.5 mmHg, but after re-trialing the patient on nimodipine 3 days later, a similar episode occurred, with SpO2 dropping to 86% and PaO2 dropping from 89.5 to 58.7 mmHg. This also improved after increasing FiO2 to 100%.

Notably, the patient was already on 0.60 of FiO2 prior to the first episode and the patient ultimately developed ARDS the day after the second nimodipine trial, suggesting the patient had some element of pulmonary injury prior to nimodipine initiation (a pulmonary contusion was seen on the day 1 chest Xray). The patient's pulmonary status continued to decline (the worst P/F ratio reported was 91) and the patient died on day 12. The Naranjo score for nimodipine contributing to the hypoxemic events was 5 (probable) because of the repeated administration, but the degree to which nimodipine caused vs contributed to the patient's desaturations are unclear.

Case 2: Nimodipine #2

This case⁴ is an interesting one, although contains little detail - an 81 year old woman is brought to the ED hypotensive (70/45 mmHg) and hypoxemic (PaO2 42 mmHg). ED workup was largely unremarkable, but it was revealed that the patient takes warfarin, amiodarone, and nimodipine as an outpatient (curious choice of outpatient CCB, this is not explored further) and there is a concern for an acute CCB overdose. After administration of IV calcium gluconate, the patient's blood pressure rebounds to 125/80 mmHg and PaO2 improves to 113 mmHg (with the help of FiO2 of 50%, the delivery method of which is not explained).

While the brief case report is lacking in detail, it is an interesting concept nonetheless - should the patient have been experiencing shunting from her nimodipine intoxication, calcium appeared to reverse the process.

Case 3: Nimodipine #3

The third case⁵ is yet another strike against nimodipine. A 62 year old man experienced an aSAH from a ruptured AComm aneurysm and experienced worsening hypoxia ultimately leading to intubation. On hospital day 4, the patient's oxygenation worsens, with a PaO2 of 63 mmHg at an FiO2 of 40%. A chest CT demonstrates posterior atelectasis out of proportion to the P/F ratio, and a TTE demonstrated normal cardiac function without noted anatomic shunting after an agitated saline study. The patient's PEEP was increased to 7.5 mmHg and a follow up gas demonstrated improvement. The patient was extubated the next day but experienced a similar worsening which led to re-intubation which is when the relationship between desaturation and nimodipine administration was noted.

 

Like the Devlin case report, this led to a high Naranjo score (7!) and follows a similar pattern of nimodipine likely worsening some degree of underlying pulmonary dysfunction. The pattern of desaturation was attenuated after the nimodipine was fractionated to 30mg q2 however, suggesting the peak effect of nimodipine may have played a role as well.

Case 4: Nicardipine

This French case report⁶ (lovingly translated by Google) describes a 55 year old man admitted for C6-7 fracture that was complicated by Pseudomonal pneumonia leading to ARDS. After the patient was successfully treated with lung protective ventilation and underwent placement of a tracheostomy with significant improvements in his oxygenation, he experienced significant hypertension after waking up from sedation which was treated with a nicardipine infusion. Soon after nicardipine initiation, the patient experienced an acute desaturation (PaO2 111 -> 60.2 mmHg) with no obvious cause. Nicardipine was switched to urapidil (a short acting IV alpha antagonist used widely outside the US) and FiO2 was increased from 45% to 80% which resolved the hypoxic event. Seven days later, a similar hypertensive episode occurred and the patient received a single nicardipine IV 5mg dose which led to a similar hypoxic event.

Case 5: Nicardipine #2

The second report⁷ involving nicardipine leaves the neuro population and describes a 63 year old woman who underwent bilateral orthotopic lung transplant. The patient did well post-operatively but did experience moderate (grade 2) graft dysfunction with a P/F ratio of 215 24 hours post-op. The patient was extubated and did well on HFNC but did become hypertensive, prompting initiation of a nicardipine infusion. The patient experienced an acute desaturation with PaO2 dropping from 93->69 which resolved after the discontinuation of nicardipine.
Similar to other cases, this patient had an underlying reason to experience alveolar hypo-oxygenation of which nicardipine likely worsened some degree of shunting. This article also has an excellent discussion and physiology-based explanation as to why nicardipine likely contributed - highly suggested reading!

Case 6: Clevidipine

The last illustrative case report⁸ involves a 16 year old boy who underwent complex AVM resection and was started on clevidipine post-operatively to reduce bleeding risk. The patient remained intubated post-operatively because a repeat surgery was planned, but 15 hours after initiation of clevidipine, the patient, who was already on an FiO2 of 60%, experienced a gradual decline in his oxygen saturation which was confirmed with a PaO2 of 33 (down from 77) 19 hours after starting clevidipine. An extensive diagnostic and therapeutic workup for the hypoxemia was negative and epoprostenol, esmolol, and paralysis with cisatracurium did not resolve the hypoxia. After consulting with their clinical pharmacist, clevidipine was identified as a potential cause, was discontinued, and the hypoxemia subsequently improved.

What is interesting about this case is that unlike the other cases, there is not an obvious underlying cause of clevidipine "tipping them over the edge", apart from likely volume resuscitation in the OR (there was only mild atelectasis on chest Xray). The dose of clevidipine here (16mg/hr) is relatively higher than the that of the nicardipine infusions used in the other cases, so there is likely an interplay between the dose and the sensitivity of the underlying substrate to shunting.

Overall, there are a number of very convincing cases that calcium channel blockers can cause clinically significant pulmonary shunting.

What is Shunting?

Shunting is likely best (or at least most comprehensively) explained in the marvelous Radermacher⁹ article on gas exchange in AJRCCM's ARDS series. In short, all blood flowing through the pulmonary circulation either passes through ventilated lung or through non-ventilated pulmonary tissue where the blood does not come into contact with alveolar gas. The blood that does not participate in gas exchange is "shunted" through the lungs, either through normal anatomic structures (bronchial and Thebesian veins) or through physiologic structures, meaning the blood passes through alveoli but the alveoli are either not aerated or are not capable of participating in gas exchange. This shunted blood flow cannot be oxygenated and will contribute to subsequent arterial hypoxemia. The total volume of pulmonary blood flow could then be modelled as total pulmonary blood flow (Qt) is equal to pulmonary shunt/venous admixture (Qva) plus pulmonary alveolar capillary blood flow (Qc). In other words, the blood that gets oxygenated in the alveoli and the blood that is shunted around the alveoli make the sum of all the blood that flow through the lungs. The amount of blood flow made up by the shunt can be quantified (Qva/Qt) by dividing the difference of ideal capillary O2 (estimated at 100% in ARDS patients, lower with normal lung) and arterial oxygen concentration by the difference in ideal capillary O2 and mixed venous oxygen concentration. Mathematically, that would be written as Qva/Qt = (CcO2 - CaO2)/(CcO2 - CvaO2). In practice this requires right heart catheterization to measure pulmonary arterial blood gases but conceptually, the amount of shunt can be thought of as how much of the "ideal" oxygen makes it to arterial supply.

One caveat to this estimation is that shunt can either be from hypoventilated lung areas (e.g. atelectasis or pulmonary edema) or entirely nonventilated lung areas. Physiologically, this is crucial especially for the concept of whether calcium channel blockers can measurably worsen shunting - nonventilated lung areas will be entirely nonresponsive to increases in FiO2 while hypoventilated areas could have shunting partially overcome by increasing FiO2. This relationship is nicely demonstrated in the classic Benatar¹⁰ paper describing shunt lines - with significant degrees of nonventilated shunt, PaO2 will only marginally respond to increases in FiO2 because the increased oxygen delivery is simply not exposed to poorly oxygenated pulmonary blood. In this diagram, the diagonal lines represent the fraction of pulmonary blood flow made up by shunt (Qva/Qt). The X axis is the FiO2 delivered to the patient and the Y axis is the expected change in PaO2 with changing FiO2.

This becomes particularly problematic with CCBs when considering the pulmonary circulation's ability to vasoconstrict in the setting of hypoxia. Unlike the systemic circulation which *vasodilates* in the presence of hypoxemia to increase blood flow (and hopefully delivery of oxygen), the pulmonary circulation does the opposite - when pulmonary vessels are exposed to low oxygen concentrations, calcium-dependent depolarization occurs, constricting the vessel, directing blood flow *away* from the hypo-oxygenated alveoli towards better oxygenated alveoli (where vessels are not constricted). The exact mechanism of "hypoxic pulmonary vasoconstriction" (lovingly abbreviated as HPV in most articles) isn't exactly clear, with some authors¹¹ heavily attributing the process to calcium release (both intracellular and extracellular) and some authors¹² suggesting calcium plays a minor role at best. The general consensus seems to be that calcium likely plays a role, but it remains unanswered exactly how big a role and exactly how it does so.

The Physiologic Basis for CCB Shunting

While the 1980s were not the best decade for mongrel dogs, they were remarkable for the volume of research conducted on lung physiology. With the introduction of dihydropyridine calcium channel blockers, there was an especially heavy focus on how these agents might improve (or worsen!) pulmonary function in patients with pulmonary hypertension.

Long story short, the data are quite mixed on exactly what CCBs do to pulmonary hemodynamics and oxygenation. Studies look at a few key variables - pulmonary vascular resistance, pulmonary arterial pressure, and arterial oxygenation. The available studies vary in the conditions in which patients were tested, the underlying lung health, and the species the experiments were conducted (see above re: an unfortunate decade for dogs).

Pulmonary Vascular Resistance

Most studies evaluating the effect of calcium channel blockers objectively report PVR values through right heart catheterization, and the overall conclusion is that (unsurprisingly), calcium channel blockers lower PVR. This is not without caveats, though - first, most studies (in humans) evaluate a single dose of sublingual nifedipine. While this is generally ok to test the theory if the drug class itself affects pulmonary circulation, it is hard to extrapolate continuous infusions like nicardipine or clevidipine particularly because nifedipine (especially sublingually) will increase cardiac output, affecting the interpretation of many of these values. The second caveat is the effect on PVR is highly dependent on the quality of the oxygen delivered to the patient (or animal). That is, the drug response is nearly always different when the patient is breathing room air, supplemental oxygen, or when they are hypoxic or breathing hypoxic air (FiO2 <0.21).

In humans, this is perhaps best illustrated by Simonneau et al² who administered 20mg of sublingual nifedipine to 13 patients with pulmonary hypertension due to COPD in acute respiratory failure who had a PA line placed to estimate accurate pulmonary hemodynamics. Group 1 (7 patients) had standard pulmonary hemodynamics measured before and 1, 2, and 4 hours after their single dose of nifedipine. PVR decreased from 6.2 mmHg/L/min to a nadir of 3.9 at 1 hour, recovering to 4.9 at 4 hours. This response was also proportional to the degree of hypoxemia the patients experienced - the more severe the baseline hypoxemia, the greater the reduction in PVR.

In group 2, however, they compared hemodynamics in patients on room air vs those on supplemental oxygen - in patients supplemented with oxygen, nifedipine only modestly lowered PVR from baseline (from 4.2 mmHg/L/min to 3.6 mmHg/L/min) which was not significant. This suggests that when the patients' lungs were appropriately oxygenated, hypoxic pulmonary vasoconstriction was near-maximally suppressed and there was little additional room for nifedipine to vasodilate.

This general observation was confirmed in a more extreme way Naejie et al¹ - 8 healthy volunteers underwent PA line placement and also received 20mg of sublingual nifedipine. Pulmonary hemodynamics were then measured under two scenarios - breathing room air (FiO2 0.21) and breathing hypoxic air (FiO2 0.125). PVR was significantly lower than baseline after nifedipine in patients breathing FiO2 0.125 air (92 to 75 dynes per sec per cm5), but was unchanged when breathing room air (43 to 48 dynes per sec per cm5). This has also been confirmed in animal models, with Archer et al¹³ finding similar results in anesthetized dogs undergoing hypoxic challenges, Nakazawa et al¹⁴ finding similar findings in anesthetized dogs even with cardiac output kept the same, and Johnston et al¹⁵ confirming the findings in an experimental model of pulmonary edema in dogs. Casthely et al¹⁶ also found similar results in anesthetized dogs undergoing intentional hypotension. The additional point of information from these animal studies is the effect is clearly dose related,¹⁴ with higher doses of sublingual or intravenous calcium channel blocker (with a mix of nifedipine, nisoldipine, and nicardipine) resulting in lower PVR in the presence of hypoxia.

Pulmonary Artery Pressure

Effects on pulmonary arterial pressure are less well-described and appear more dependent on changes in cardiac output than the actual vascular effect of the drug. Studies either suggest a mild increase in PAP likely secondary to increases in CO from sublingual nifedipine (Simonneau et al²) or no change, likely from some equalization of vasodilation and increased CO (Naeije et al¹, Melot et al¹⁷, Boldt et al¹⁸).

Arterial Oxygenation

Perhaps most importantly, nearly all studies evaluating the effects of CCBs on pulmonary hemodynamics report the effect of the changes on arterial oxygenation. Unfortunately, the effect is the least clear cut - studies report decreases in PaO2, increases in PaO2, and a neutral effect on PaO2, all depending on the drug, dose, pulmonary condition, and species(!).

Direct increases in PaO2 are the least commonly described. Boldt et al¹⁸ report the effect of several vasoactive medications, including both intravenous nifedipine 0.7 mcg/kg/min and nimodipine 1 mcg/kg/min (they have all the cool drugs in Germany) in 50 men undergoing a CABG (10 received nimodipine and 10 received nifedipine). The authors report significant improvements in arterial oxygenation with both nimodipine (368 to 425 mmHg) and nifedipine (382 to 424 mmHg), but it should be noted that these patients had no underlying pulmonary dysfunction and were ventilated with aggressive oxygenation (FiO2 of 1.0 and PEEP of 5 mmHg throughout the procedure), so changes in PaO2 likely reflect an oxygen rich environment supported by a mild increase in cardiac output from the CCBs and these findings would be difficult to generalize to an ICU population.

Neutral effects on PaO2 are also relatively uncommon in the experimental realm - Naeije et al¹ did not observe lower PaO2 in healthy volunteers breathing FiO2 of 0.125 who received nifedipine vs control (41 vs 41 mmHg), but these patients had no underlying pulmonary pathology and intact compensatory mechanisms - the lower PVR was compensated by increased cardiac output. Nakazawa et al¹⁴ also reported a neutral effect of nicardipine on arterial oxygen in their population of anesthetized dogs, even when cardiac output was kept constant. This is in the context of only the left lower lobe of the lung being exposed to nicardipine, however, so sufficient lung area may not have been involved to affect systemic oxygenation.

A potential decrease in PaO2 appears to be more expected especially in the setting of underlying respiratory dysfunction. Worsening oxygenation was noted in Simonneau et al² in patients with COPD not receiving oxygen (42 to 38 mmHg, p < 0.05) but not when supplemental oxygen was delivered (227 to 224 mmHg, p >0.05) and oxygenation worsened in a dose-related fashion in Archer et al's¹³ group of anesthetized dogs with nisoldipine specifically, but only when the dogs were intentionally made hypoxic (36 mmHg at baseline to 35 with low dose, 32 with medium dose, and 28 mmHg with high dose nisoldipine). Casthely et al¹⁶ and Johnston et al¹⁵ also noted worsening oxygenation in their animal models, but interesting Johnston et al, which used a oleic acid model of pulmonary edema, noted that while PaO2 was lower (from a baseline of 177 torr to 67 torr, although this was up form when the initial oleic acid injury occurred), total *delivery* of oxygen increased, likely secondary to improvements in CO.

Overall, the effect here appears to be more complex are hard to predict. In patients who are already hypoxic, it appears a calcium channel blocker has the potential to worsen the underlying hypoxia. In patients with pristine lung function without risk factors for hypoxia, the evidence suggests that a CCB is going to induce de novo hypoxemia.

Shunt

To round out the discussion, it's important to recognize whether new hypoxia observed in patients on a CCB is directly attributable to an actual worsening of right to left shunt. Quantifying shunt is challenging, however, and thus few studies actually report what percent of pulmonary blood flow is made up of shunt. In patients with underlying *pathologic* shunt, it does appear that CCBs have the potential to improve right to left shunt - Melot et al¹⁷ report on two patients with underlying PAH with one patient who had an average V/Q ratio of 2.6 (ideal is ~0.8) and 20% of blood flow being made up of shunt. While sublingual nifedipine did shift more lung units towards worsening V/Q mismatch (18.9% of lung units had a V/Q ratio <0.2 vs a baseline of 11.9%), the overall distribution shifted closer to an ideal ratio, with overall shunt decreasing to 10%. Boldt et al¹⁸ reported no effect of nimodipine or nifedipine on shunt percent (Qs/Qt of 15% at baseline to 16.2% with nimodipine and 15.7% to 17.9% with nifedipine, p>0.05), but as noted above, these patients were hyperoxygenated which may have attenuated the ability to detect a change.

Quantification of worsening shunt was observed in two of the animal model experiments. Both Casthely et al¹⁶ and Nakazawa et al¹⁴ reported significant increases in total pulmonary blood flow being made up by shunt, with Casthely revealing a doubling (9.7% -> 20.5%) after administration of nifedipine to dogs and Nakazawa similarly reporting a doubling (10% -> 20.4%) with nicardipine, but notably only in dogs who were made intentionally hypoxic and with constant cardiac output. At a fixed cardiac output, shunt did increase, but only mildly (9.7% -> 13.1%). Additionally, when kept at a constant cardiac output, the effect was also dose dependent, with higher doses (6 mcg/kg/min) leading to a larger shunt percentage than lower doses (1 and 3 mcg/kg/min).

While data are more limited on the quantification of shunt percentage, it stands to reason that this acts similarly to what was found with arterial oxygenation. In patients with derecruited lung units contributing to hypoxia, a calcium channel blocker likely will worsen underlying shunt which is what is driving worsening hypoxia. In patients with fully recruited lungs, however, a calcium channel blocker is unlikely to worsen what little shunt already exists.

What to Do if it Happens

There are a few insights from the available data which can help to develop an action plan for an acutely hypoxic patient on a calcium channel blocker. I listed a few of the things I gleaned from my reading which will hopefully help you if you're faced with a similar situation:

1. Calcium channel blockers are unlikely to cause hypoxia on their own with absolutely no inciting cause. Make sure you do your due dilligence evaluating the standard causes of hypoxia before chalking the hypoxia up to the calcium channel blocker, because even if stopping the nicardipine or clevidipine improves things, there may still be something underlying that you may have missed.
2. Reasonable lung recruitment maneuvers may be enough to reverse the hypoxia. If shunt is being caused by atelectasis, recruiting additional lung units may be enough to abolish the hypoxic pulmonary vasoconstriction being inhibited by the calcium channel blocker and reverse the process.
3. The effect is dose related, but there isn't clear evidence that the effect is larger with one calcium channel blocker vs another (apart from one study that suggested it was particularly notable with nisoldipine, but who uses that?). If your patient needs a calcium channel blocker (i.e. nimodipine in aSAH), try lower doses or fractionated doses so the patient sees a lower peak concentration. For continuous infusions, there may be a continuous infusion rate that achieves SBP goals but doesn't affect PVR, especially if combined with other agents. Also, because of the relation to peak effect, amlodipine is unlikely to be an issue here.
4. While not discussed at length here, this effect is not universal to all blood pressure agents. While it is also observed with nitroprusside (which should essentially never be used in the NeuroICU), labetalol does not affect that pulmonary circulation to a significant degree and is a potential alternative continuous infusion if a CCB can't be used.
5. You can absolutely rechallenge someone with a CCB if they experienced hypoxia previously. Because the effects depend on the underlying state of the lungs, if someone is otherwise improved (say extubated after their prior event with a CCB) and needs tight blood pressure control, you could trial a CCB again.

Conclusion

Calcium channel blockers can cause shunting and hypoxia, but it's more complicated than it may seem at face value. While calcium channel blockers are active in the pulmonary vasculature and can inhibit hypoxic vasoconstriction which can worsen shunting, they will not cause substantial shunting and hypoxia without underlying vasoconstriction already happening.

 Andrew Webb, PharmD, BCCCP

Clinical Pharmacist, Neurocritical Care

Massachusetts General Hospital

Ajwebb@mgh.harvard.edu

@AJWPharm

References

1.            Naeije R, Mélot C, Mols P, Hallemans R. Effects of vasodilators on hypoxic pulmonary vasoconstriction in normal man. Chest. 1982;82(4):404-410. doi:10.1378/chest.82.4.404

2.            Simonneau G, Escourrou P, Duroux P, Lockhart A. Inhibition of Hypoxic Pulmonary Vasoconstriction by Nifedipine. New England Journal of Medicine. 1981;304(26):1582-1585. doi:10.1056/NEJM198106253042606

3.            Devlin JW, Coplin WM, Murry KR, Rengachary SS, Wilson RF. Nimodipine-induced acute hypoxemia: case report. Neurosurgery. 2000;47(5):1243-1246; discussion 1246-1247. doi:10.1097/00006123-200011000-00048

4.            Gerloni R, Copetti R. Easily reversible hypoxemia and hypotension induced by nimodipine. Eur J Emerg Med. 2004;11(5):295-297. doi:10.1097/00063110-200410000-00012

5.            Baker M, Bastin MT, Cook AM, Fraser J, Hessel E. Hypoxemia associated with nimodipine in a patient with an aneurysmal subarachnoid hemorrhage. Am J Health Syst Pharm. 2015;72(1):39-43. doi:10.2146/ajhp140196

6.            Cotte J, D’Aranda E, Esnault P, Bordes J, Meaudre E. Hypoxie sous nicardipine : rôle de la vasoconstriction pulmonaire hypoxique. Revue de Pneumologie Clinique. 2012;68(3):221-224. doi:10.1016/j.pneumo.2011.08.003

7.            Mishra A, Reed RM, Eberlein M. Severe, Rapidly Reversible Hypoxemia in the Early Period after Bilateral Lung Transplantation. Ann Am Thorac Soc. 2016;13(6):979-985. doi:10.1513/AnnalsATS.201602-107CC

8.            Short JH, Fatemi P, Ruoss S, Angelotti T. Clevidipine-Induced Extreme Hypoxemia in a Neurosurgical Patient: A Case Report. A A Pract. 2020;14(2):60-62. doi:10.1213/XAA.0000000000001146

9.            Radermacher P, Maggiore SM, Mercat A. Fifty Years of Research in ARDS.Gas Exchange in Acute Respiratory Distress Syndrome. Am J Respir Crit Care Med. 2017;196(8):964-984. doi:10.1164/rccm.201610-2156SO

10.          Benatar SR, Hewlett AM, Nunn JF. THE USE OF ISO-SHUNT LINES FOR CONTROL OF OXYGEN THERAPY. British Journal of Anaesthesia. 1973;45(7):711-718. doi:10.1093/bja/45.7.711

11.          Weigand L, Foxson J, Wang J, Shimoda LA, Sylvester JT. Inhibition of hypoxic pulmonary vasoconstriction by antagonists of store-operated Ca2+ and nonselective cation channels. Am J Physiol Lung Cell Mol Physiol. 2005;289(1):L5-L13. doi:10.1152/ajplung.00044.2005

12.          Robertson TP, Hague D, Aaronson PI, Ward JPT. Voltage-independent calcium entry in hypoxic pulmonary vasoconstriction of intrapulmonary arteries of the rat. J Physiol. 2000;525(Pt 3):669-680. doi:10.1111/j.1469-7793.2000.t01-1-00669.x

13.          Archer SL, Yankovich RD, Chesler E, Weir EK. Comparative effects of nisoldipine, nifedipine and bepridil on experimental pulmonary hypertension. J Pharmacol Exp Ther. 1985;233(1):12-17. PMID: 3872360

14.          Nakazawa K, Amaha K. Effect of nicardipine hydrochloride on regional hypoxic pulmonary vasoconstriction. Br J Anaesth. 1988;60(5):547-554. doi:10.1093/bja/60.5.547

15.          Johnston WE, Vinten-Johansen J, Tommasi E. Effect of nifedipine on oxygen delivery in canine asymmetric oleic acid lung injury. Crit Care Med. 1990;18(7):738-743. doi:10.1097/00003246-199007000-00012

16.          Casthely PA, Villanueva R, Rabinowitz L, Gandhi P, Litwak B, Fyman PN. Intrapulmonary shunting during deliberate hypotension with nifedipine, diltiazem and labetalol in dogs. Can Anaesth Soc J. 1985;32(2):119-123. doi:10.1007/BF03010034

17.          Mélot C, Naeije R, Mols P, Vandenbossche JL, Denolin H. Effects of nifedipine on ventilation/perfusion matching in primary pulmonary hypertension. Chest. 1983;83(2):203-207. doi:10.1378/chest.83.2.203

18.          Boldt J, Von Bormann B, Kling D, Ratthey K, Hempelmann G. Influence of nimodipine and nifedipine on intrapulmonary shunting--a comparison to other vasoactive drugs. Intensive Care Med. 1987;13(1):52-56. doi:10.1007/BF00263558