The Acute Respiratory Distress Syndrome and Treatment with Mesenchymal Stem Cells

Part 1: Acute Respiratory Distress Syndrome: An Overview

00:00:07.15 My name is Carolyn Calfee, and I'm an Associate Professor of Medicine and Anesthesia
00:00:11.26 here at the University of California, San Francisco.
00:00:14.21 And I’m going to talk with you today about mesenchymal stem, or stromal, cells and the
00:00:18.20 acute respiratory distress syndrome.
00:00:21.07 I'll give the first portion of the talk and then my colleague, Dr. Michael Matthay,
00:00:24.22 who's a Professor of Medicine and Anesthesia here at UCSF, also, will give the second portion.
00:00:31.04 This is a brief overview of what I'm going to cover today.
00:00:34.11 I'll start with talking about, what is ARDS, or the acute respiratory distress syndrome?
00:00:39.08 Its clinical presentation and pathophysiology.
00:00:42.22 And then I'll move on to discuss our current treatments for ARDS and why we're
00:00:46.17 in search of new ones.
00:00:48.21 And then I'll hand the pointer over to Dr. Matthay, who will talk about the work
00:00:52.20 that we and other colleagues have been doing to try to translate mesenchymal stem cells from
00:00:57.02 the preclinical side, using bench science and animal models, into human subjects.
00:01:03.10 I thought it would be useful to start with a clinical case of ARDS, to explain
00:01:09.18 what the syndrome is and give you a sense of what we're trying to treat.
00:01:13.17 So, this is a case of a patient that we took care of several years ago, a 42 year old man
00:01:19.17 who was previously in good health, until he presented to the emergency room with
00:01:23.25 fevers for five days, muscle aches and cough.
00:01:27.03 In addition, he had increasing shortness of breath for the past day.
00:01:31.07 He had no past medical history and was a nonsmoker.
00:01:35.26 This is his initial chest X-ray and, while this may look relatively normal when you compare
00:01:40.28 it to a normal chest X-ray, here you'll see that on his X-ray there are faint
00:01:46.16 interstitial opacities, as we term them, or faint whitish lines extending out from the center of
00:01:52.08 the chest out to the periphery.
00:01:56.08 Over time, as we followed his chest X-rays through his hospitalization, on the next day,
00:02:01.05 these opacities were increasing.
00:02:03.17 And increasing even further, as you can see here, progressing to what we call frank consolidation,
00:02:09.16 until the point where he required mechanical ventilator support, as you can see here.
00:02:15.20 And the diagnosis here is acute lung injury, also known as ARDS
00:02:19.24 or the acute respiratory distress syndrome,
00:02:22.02 due to H1N1, viral pneumonia.
00:02:25.06 So, this is a classic case of the acute respiratory distress syndrome.
00:02:30.07 And what this is is what we call non-cardiogenic pulmonary edema, that is, where the lungs
00:02:35.05 fill up with fluid in the alveolar space, or the air sacs of the lung, for reasons
00:02:39.22 other than heart failure.
00:02:42.13 It's diagnosed by three defining criteria, including the acute onset of bilateral opacities
00:02:47.07 on the chest X-ray, as you see here;
00:02:50.11 low oxygen levels in the blood;
00:02:52.20 and exclusion of heart failure as the primary cause of the pulmonary edema.
00:02:58.07 ARDS is a common syndrome that develops following certain clinical conditions.
00:03:03.12 These may be conditions that cause direct injury to the lungs, like pneumonia,
00:03:07.19 aspiration of gastric contents into the lungs, or near drowning.
00:03:11.13 Or conditions that cause what we call indirect lung injury, such as sepsis, which is a
00:03:15.16 severe infection in the bloodstream, severe trauma,
00:03:18.26 blood transfusion or other causes.
00:03:21.17 This is a common syndrome.
00:03:22.27 It affects about 200,000 people per year in the US alone,
00:03:27.05 and is associated with 75,000 deaths per year in the US,
00:03:30.22 more than breast cancer, colon cancer, or HIV.
00:03:34.17 And mortality -- I'll show you in the subsequent slides -- remains high, 30-40%.
00:03:39.17 Unfortunately, at this time, no specific therapies are available other than specific ventilator
00:03:44.24 and fluid management.
00:03:45.26 So, what I'm going to show you in the next couple of slides is our understanding of
00:03:50.17 the process by which normal lung, which you see here, with the alveolar air sacs nice and
00:03:56.02 filled with air, where you can imagine gas exchange occurs easily,
00:03:59.28 develops acute lung injury, which you can see here on the right,
00:04:03.15 where the air sacs of the lung are swollen,
00:04:05.03 filled with inflammatory, pink pulmonary edema fluid.
00:04:09.09 So, what I'm showing here is a schematic of what I just showed you in the previous slide.
00:04:13.27 And on the left side of your screen, you'll see a normal alveolus, or a normal air sac
00:04:18.07 in the lung.
00:04:19.17 What you see there is that the alveolus is filled with air, there's a nice thin lining
00:04:26.10 of surfactant, which helps to keep the alveolus open,
00:04:29.18 and you can see that the epithelial line -- the epithelial cell surface layer --
00:04:33.15 is thin and well... closely opposed to the capillary endothelium, here.
00:04:38.03 You can imagine that gas exchange occurs very easily across this barrier.
00:04:44.00 In contrast, this is the injured alveolus during the acute phase of acute lung injury.
00:04:48.12 And you'll notice a number of abnormalities.
00:04:50.06 So, for one, the alveolus is filled with this pink, pulmonary edema fluid, which has filled
00:04:56.02 with inflammatory cells elaborating inflammatory cytokines.
00:05:00.05 You can also see, here, that the lining of the alveolus, that epithelial barrier that's
00:05:04.21 normally very tight and helps to keep any fluid from crossing over into the alveolus,
00:05:09.13 is now necrotic, and frankly denuded in some places.
00:05:14.02 You can see that the border between the epithelial and the endothelial surface is swollen.
00:05:19.01 So, you can imagine how gas exchange is much more difficult there.
00:05:22.21 And you can see that the capillary endothelium itself is also involved.
00:05:26.05 So, the endothelial surface is abnormal and activated.
00:05:30.00 You can see that neutrophils are able to cross over from the endothelium into the interstitium.
00:05:35.08 And you can see that platelets are involved here as well, too.
00:05:37.14 So, this is a very complex process with compact... complex pathophysiology.
00:05:43.22 So, given that, what are our current treatment strategies for ARDS?
00:05:48.00 What can we do to try to treat the syndrome once it develops?
00:05:50.22 Well, I'm going to outline our current treatment strategies, which are focused primarily on
00:05:55.15 ventilation with lower tidal volumes, a fluid conservative management strategy, and then
00:06:01.18 I'll talk about the history of pharmacologic therapies and why we're in search of new therapies.
00:06:06.23 So, ventilation with lower tidal volumes is now the standard of care for ARDS patients.
00:06:13.03 These patients are placed on mechanical ventilators in the intensive care unit.
00:06:17.04 And while we initially thought, years ago, that ventilation with larger tidal volumes
00:06:21.14 was necessary to help open up the lungs and maintain normal oxygen levels,
00:06:26.05 we've now known for the past 14 years, since the publication of this landmark trial,
00:06:30.21 that ventilation with lower tidal volumes actually improves survival from this condition.
00:06:36.24 How does this work?
00:06:37.26 Well, you've seen a similar diagram to this before, just a couple of slides ago.
00:06:42.26 The injured alveolus is over here on the left.
00:06:45.21 And you can see that in the setting of high stretch or higher tidal volume ventilation,
00:06:49.22 that that actually worsens the injury, we now know.
00:06:52.22 In fact, there's more pulmonary edema, the migration of neutrophils into the
00:06:58.04 alveolar space is enhanced, and in addition alveolar fluid clearance, which is the ability of the lung
00:07:03.10 to actually remove the pulmonary edema fluid from the alveolus, is impaired.
00:07:09.11 In contrast, when lower stretch ventilation is used, alveolar fluid clearance is enhanced
00:07:14.26 and the epithelium is allowed to proliferate and heal again, thus enhancing the resolution
00:07:20.08 of ARDS.
00:07:23.11 In addition, we now know that fluid conservative therapy helps patients with ARDS to get
00:07:28.22 off the ventilator and out of the intensive care unit sooner.
00:07:32.10 This was another landmark clinical trial supported by the NIH and carried out by the ARDS Network,
00:07:38.05 which showed that patients who were treated with the fluid conservative strategy were
00:07:42.21 able to get off the ventilator sooner compared to patients
00:07:45.13 who were treated with the fluid liberal strategy.
00:07:48.08 Why does this happen?
00:07:49.24 Well, in part this goes back to what we know about the physiology of the lung.
00:07:55.01 Even under conditions of normal permeability, as pulmonary hydro... hydrostatic pressure increases,
00:08:00.21 which happens with increasing fluid in the intravascular space, the rate of lung
00:08:05.18 edema formation increases.
00:08:07.26 But under conditions of increased permeability, such as we see in ARDS,
00:08:11.23 as the pulmonary hydrostatic pressure increases,
00:08:14.16 the rate of lung edema formation increases dramatically.
00:08:17.28 And this is what we see in the case of ARDS.
00:08:20.18 And we think that by treating patients with fluid conservative therapy, that is, by decreasing
00:08:24.26 their pulmonary hydrostatic pressure,
00:08:27.02 we're actually decreasing the rate of lung edema formation.
00:08:30.18 So, these are supportive care practices.
00:08:34.09 What about specific pharmacologic therapies for ARDS?
00:08:37.08 Well, unfortunately, despite 25 years of attempts to find a successful therapy for ARDS,
00:08:44.15 none of these pharmacologic therapies have been successful.
00:08:47.28 This is just a partial list of some of the... the pharmacologic agents that have been tried
00:08:52.22 for ARDS, including, most recently, albuterol, omega-3 fatty acids, and statins,
00:08:59.00 all of which have failed to show a benefit.
00:09:04.24 During this time, ARDS mortality has improved significantly, largely as a result of ventilation
00:09:10.20 with lower tidal volumes and treatment with fluid conservative therapy,
00:09:13.19 as I've just shown you.
00:09:15.12 But as you can see from the figure, despite this improvement in mortality, mortality from
00:09:20.17 ARDS, even in randomized controlled trials, remains high, in the range of 20-25%.
00:09:26.09 And in unselected patients, mortality from ARDS remains 30-40%.
00:09:30.15 So, we still have a long way to go before we're effectively treating this condition.
00:09:35.06 And that's why we're continuing to search for novel therapies.
00:09:39.06 The therapy that we're going to focus on today is mesenchymal stem, or stromal, cells.
00:09:44.03 Why is it that we think that these cells might be beneficial for ARDS?
00:09:47.24 The original idea when mesenchymal stem cells were first studied for pulmonary disease,
00:09:53.08 now a couple of decades ago, was that the mesenchymal stem cells
00:09:57.02 would engraft in the lung and thereby facilitate repair.
00:10:00.09 However, now we know that it's very unlikely that these stem cells actually engraft in
00:10:04.22 the lung, but instead we think they act by exerting pleiotropic effects on both
00:10:09.20 lung repair and regeneration.
00:10:12.13 This is largely, though not entirely exerted through paracrine mechanisms.
00:10:17.01 The cells also have anti-inflammatory and antimicrobial properties.
00:10:21.01 And we think that it's this multiple pathway approach that may explain why mesenchymal
00:10:25.11 stem cells hold a lot of promise for potentially treating ARDS.
00:10:29.04 Now, at this point in the talk, I'm going to hand it over to Dr. Michael Matthay,
00:10:33.23 Professor of Medicine and Anesthesia at UCSF, who's going to talk about work that we and other
00:10:38.12 colleagues have been doing, both on the preclinical side and the early clinical side, to try to
00:10:43.12 translate mesenchymal stem cell therapy into clinical use.

Part 2: Mesenchymal Stem Cell Therapy for ARDS

00:00:07.08 My name is Michael Matthay.
00:00:09.04 I'm a professor of Medicine and Anesthesia at UCSF.
00:00:12.20 I work with Dr. Calfee.
00:00:14.11 And I will now discuss the details of our preclinical work that has led to our efforts
00:00:20.02 to use mesenchymal stem cells for... as a novel treatment for the
00:00:24.05 acute respiratory distress syndrome.
00:00:28.04 As Dr. Calfee explained, mesenchymal stem cells, or mesenchymal stromal cells,
00:00:34.13 have a history of being potentially capable of reducing inflammation and enhancing repair
00:00:41.03 in the lung.
00:00:42.17 And they work by multiple mechanisms, which is why we thought that they may have relevance
00:00:47.22 for patients with ARDS.
00:00:49.28 I'll show you data from our experiments in mice that suggested they might be effective.
00:00:57.01 And then additional experiments we did with the ex vivo perfused human lung to test
00:01:01.20 their potential efficacy.
00:01:03.22 And then, finally, some studies we did in a severe lung injury model in sheep over a
00:01:09.00 24-hour period of time.
00:01:11.13 And then I'll describe to you the work we're doing now to test and translate these cells
00:01:16.19 for therapy in patients with ARDS.
00:01:19.02 Now, this slide you've already seen in terms of the injury to the lung that occurs in patients
00:01:26.11 with ARDS, where there's an accumulation of protein-rich edema fluid in the air spaces.
00:01:32.08 That is accompanied by injury to the endothelium and the epithelium of the lung associated
00:01:37.12 with a large number of inflammatory cells.
00:01:40.00 Now, on the left side of this slide, you can see the mechanisms by which we anticipate
00:01:46.13 mesenchymal stem cells might be effective in reducing the injury and also enhancing
00:01:51.26 or refair... repair.
00:01:53.26 We believe they have anti-inflammatory properties.
00:01:56.22 They can restore endothelial and epithelial barrier integrity.
00:02:00.12 They can enhance the clearance of the edema fluid from the lung.
00:02:04.17 They also have antimicrobial properties, as I will show you.
00:02:08.01 They also decrease the death of some of the endogenous cells in the lung.
00:02:13.13 And they work by both cell contact dependent and independent mechanisms.
00:02:18.04 Now, the early work on mesenchymal stem cells in lung injury was prompted, really, by decades
00:02:24.15 of work that go back to the original discovery of the mesenchymal stromal cells
00:02:29.22 by Dr. Friedenstein in 1968.
00:02:32.24 As Dr. Calfee mentioned, initial work had suggested that perhaps the mesenchymal stem
00:02:38.12 cells could engraft in the lung and actually generate new alveolar epithelium.
00:02:42.20 However, subsequent studies showed that that was probably not the main mechanism but rather
00:02:48.08 the release of paracrine... paracrine molecules that enhanced repair.
00:02:53.09 The initial mouse studies that I will show you were made possible in part by the availability
00:02:58.09 of a NIH repository, directed by Dr. Prockop, that supplied the bone-marrow derived mouse
00:03:05.13 and human mesenchymal stem cells for our experimental studies.
00:03:10.17 Our first experiments were published in 2007.
00:03:14.08 And in these experiments we tested the ability of mesenchymal stem cells to work in a severe
00:03:20.06 lung injury model in mice, in which a high dose of endotoxin was given into the air spaces,
00:03:26.28 and then was followed four hours later by either a treatment with just a placebo
00:03:33.02 -- PBS, phosphate-buffered saline --
00:03:35.16 or actually the mesenchymal stem cells themselves.
00:03:39.01 And as you can see on this figure, survival was markedly improved with the MSC treatment.
00:03:46.22 This figure shows some of the histology from those lungs, in which you can see, in this panel,
00:03:51.28 the control PBS-treated mice had considerable edema and inflammation versus the MSC-treated mice,
00:04:00.10 over here, in which you can see the air spaces were mostly clear of fluid and
00:04:04.26 the injury was markedly reduced.
00:04:07.08 This was associated with an increase in anti-inflammatory cytokines
00:04:11.14 in the mesenchymal stem cell treated mice
00:04:13.21 as well as a reduction in the pro-inflammatory cytokines.
00:04:17.20 Now, we went on, then, to test the potential efficacy of mesenchymal stem cells in a
00:04:24.05 more clinically relevant mouse model, in which we gave live bacteria into the airspace of the lung.
00:04:30.15 And so this shows the evidence that the mesenchymal stem cells markedly increased survival,
00:04:35.13 even in the presence of a severe injury from live E. coli.
00:04:39.06 This slide again shows you survival on the y-axis and the time in hours on the x-axis,
00:04:45.21 and you can see that the mice that were treated with mesenchymal stem cells had a much longer
00:04:53.17 and better survival compared to the two control groups treated with either 3T3 fibroblasts
00:04:59.08 or PBS.
00:05:00.27 These experiments were very important because they modeled more closely what happens to
00:05:06.06 patients.
00:05:07.12 Now, these experiments turned out to show a surprising and interesting result.
00:05:13.14 In addition to the improved survival, we found that the number of bacteria present in the
00:05:19.00 lungs at the end of the experiment was markedly reduced in the MSC-treated mice.
00:05:26.04 We actually had been concerned that perhaps the MSC treated mice might have a worse host
00:05:32.00 defense against bacteria because the MSCs are immunomodulatory.
00:05:36.11 And in fact we found the opposite.
00:05:38.16 Interestingly, this was explained in part by the release of antimicrobial peptides
00:05:45.15 by the MSCs themselves, which resulted in better bacterial killing in the MSC-treated mice.
00:05:53.26 This slide shows the results of additional experiments we did to test the efficacy of
00:05:59.17 mesenchymal stem cells in another model of infection in the mice.
00:06:03.28 In this model, the infection was introduced directly into the abdomen, into the peritoneum,
00:06:09.02 to simulate what may happen to patients when they have a perforated acute appendicitis
00:06:14.01 or another acute abdominal emergency from infection.
00:06:17.14 And in these experiments, you can see in green, here, that the mesenchymal stem cell-treated mice
00:06:22.25 had a much better survival compared to the controls with 3T3 fibroblasts or PBS.
00:06:28.25 Now, note that in these experiments we gave human mesenchymal stem cells.
00:06:34.14 And this is an interesting point because it turns out we learned that either mouse or
00:06:40.01 human bone marrow-derived MSCs were equally effective in treating the lung injury that
00:06:46.12 we produced in the mice.
00:06:47.27 Further, it's important to point out that in some human studies in which mesenchymal
00:06:53.28 stem cells have been given to patients for other inf... conditions besides ARDS,
00:06:59.22 we have learned that these cells are very well tolerated, immunologically, because they do not express
00:07:06.09 MHC class... type 1 and type 2 antigens, so they are not rejected.
00:07:11.06 Now, it turned out, in these last series of mouse experiments, that the antibacterial
00:07:18.25 effect was observed, just like we had found in the pneumonia model with E. coli.
00:07:23.25 And we discovered an additional mechanism that could be important and have
00:07:28.01 clinical relevance for why the bacteria were cleared better in the me...
00:07:33.05 MSC-treated mice.
00:07:35.03 And we found in these experiments that the monocytes in the mice were able to phagocytize
00:07:42.23 the bacteria more rapidly compared to the two control groups.
00:07:47.22 There was no difference in neutrophil phagocytosis.
00:07:50.11 But, in fact, the... the data surprisingly showed an enhanced phagocytosis of the bacteria
00:07:57.06 by the monocytes.
00:08:00.15 In these same experiments, we also discovered another pathway that could be important
00:08:04.16 in explaining the potential beneficial effect of MSCs for treating lung injury.
00:08:10.24 In studies we did in these mice of the spleen, we found that the MSCs upregulated expression
00:08:16.25 of two markers of the alternatively activated macrophages, CD163 and CD206.
00:08:25.27 These are markers of an anti-inflammatory phenotype and indicate, probably, that the
00:08:32.02 MSCs are able to enhance the resolution of inflammation.
00:08:37.11 So, this slide summarizes the main results of the M...
00:08:41.21 MSC studies in mice.
00:08:43.15 First, treatment with MSCs improves survival and reduces lung injury in both endotoxin
00:08:48.27 and the live bacterial models of lung injury I've shown you, both associated with reduction
00:08:54.12 in the pro-inflammatory responses to endotoxin while increasing anti-inflammatory cytokines.
00:09:01.02 And the antibacterial effect of MSCs in mice appears to be mediated by antimicrobial peptides
00:09:06.08 and also enhanced monocyte phagocytosis.
00:09:10.12 This slide shows you our ex vivo perfused human lung preparation.
00:09:15.01 This is the pre... preparation that we developed and have used to test the potential efficacy
00:09:21.01 of the mesenchymal stem cells with injury in the human lung that resembles what we see
00:09:26.07 in patients with ARDS.
00:09:27.28 Now, these lungs are procured from brain-dead subjects, in whom the families of the brain-dead subject
00:09:35.19 have given permission for the lungs to be used for, first, if possible, lung transplantation,
00:09:41.22 and if not for research.
00:09:43.13 So, we receive the lungs to study in the laboratory if they are not suitable for lung transplantation.
00:09:50.06 This slide shows you the basic setup, which is that the human lung arrives in the laboratory,
00:09:56.06 we cannulate the trachea and inflate it with continuous positive airway pressure of
00:10:01.17 10 centimeters of water, and give the lung 95% oxygen and 5% carbon dioxide, and then we
00:10:09.10 also cannulate the pulmonary artery, and perfuse it with a physiologic solution to slowly
00:10:16.04 warm the lung from 4 degrees up to 37 degrees.
00:10:20.00 And then the perfusate comes out through the left atrium, into what's shown us the perfusate reservoir,
00:10:25.02 and then is recycled by a peristaltic pump, so that we essentially have a
00:10:30.19 perfused human lung preparation with lung inflation, so that we can actually study these lungs
00:10:37.24 under pathologic conditions for a period of at least 4 hours, and even longer if needed.
00:10:44.21 This slide shows you how we adapt this lung preparation to study and test the potential
00:10:50.28 efficacy of MSCs for... for lung injury.
00:10:56.03 There are three lobes to the right lung, shown here: the right upper lobe, the right middle lobe,
00:11:00.19 and the right lower lobe.
00:11:02.10 We first examine and test the ability of the right upper lobe to remove alveolar edema fluid,
00:11:08.09 as I will explain to you on the next slide.
00:11:11.05 If we find that alveolar fluid clearance is normal, then we go ahead and induce injury
00:11:16.11 into the right middle lobe with high dose endotoxin,
00:11:20.20 similar to what we did in the mouse experiments.
00:11:23.27 And then after the introduction of endotoxin, we treat the lungs with the mesenchymal stem
00:11:28.18 cells themselves or the conditioned media.
00:11:31.21 And, as I will explain, some other experimental interventions to try to understand how
00:11:37.11 the MSCs work.
00:11:39.07 Now, I mentioned that I would explain this measurement of alveolar fluid clearance.
00:11:43.28 Dr. Calfee already referred to it in her part of the discussion.
00:11:48.15 Alveolar fluid clearance is the capacity of the lung to remove the edema fluid from the
00:11:53.16 air spaces of the lung.
00:11:55.05 And so this is a diagram to try to illustrate how we measure this experimentally.
00:11:59.26 We instill into the right upper lobe of the human lung a 5% albumin solution, which is
00:12:06.08 similar to the edema fluid that patients would develop.
00:12:09.17 And then we measure the concentration of albumin in that airspace compartment.
00:12:15.00 If it rises by approximately 20% per hour, then that means the salt and water is being
00:12:21.08 removed across the alveolar epithelium normally, by a sodium and chloride-dependent transport system.
00:12:28.20 So, this slide shows the effect of using the
00:12:32.18 mesenchymal stem cells or mesenchymal stem cell-conditioned media
00:12:36.24 on lung fluid and protein balance in the endotoxin- or LPS-injured lung.
00:12:43.03 On the panel A, which shows endothelial permeability, which is the leakiness of the microcirculation
00:12:51.20 of the lung, you can see with LPS that the permeability increased threefold,
00:12:57.02 whereas treatment with the MSCs or the conditioned media of MSCs returned permeability to normal
00:13:03.18 when they were given one hour after the injury.
00:13:05.24 And these experiments were done over four hours.
00:13:08.22 In these experiments, the MCSs were given directly into the bronchus, into the distal
00:13:13.19 air spaces.
00:13:15.00 In the same experiments, we also found that... as shown in panel B, that the quantity of
00:13:20.18 pulmonary edema, which increased in the LPS-treated lungs, was returned to normal,
00:13:26.17 and even to a below-normal level, by the conditioned media.
00:13:30.17 So, these were very favorable effects that suggested that the MSCs might be effective
00:13:36.10 in human lung injury.
00:13:39.00 In these same experiments, we explored possible mechanisms by which the MSCs worked.
00:13:44.13 In order to do this, we reduced the ability of the MSCs to release a paracrine growth factor
00:13:53.15 known as keratinocyte growth factor.
00:13:57.12 There is considerable work from our laboratory and other investigators that suggested that
00:14:01.00 this growth factor could reduce lung injuries.
00:14:03.22 So, we pretreated some of the MSCs with an siRNA strategy that reduced KGF production.
00:14:11.16 And this slide shows, on the y-axis, alveolar fluid clearance, the ability of the lung to
00:14:16.18 remove edema fluid, first under control conditions, and you can see that was approximately 20%.
00:14:22.27 With the LPS, you can see that the... the... the clearance went down to approximately zero.
00:14:29.16 It essentially eliminated the ability of the lung to remove fluid.
00:14:34.08 Whereas when the lungs were treated with the conditioned media, it came up to a normal level.
00:14:39.12 When we removed the KGF from the MSCs, you can see that the beneficial effect was largely
00:14:45.24 abrogated, by about 70-80%.
00:14:48.23 And if we gave, actually, KGF itself, as shown in the yellow panel, the alveolar fluid clearance
00:14:54.04 came back to normal.
00:14:55.09 So, this identified one paracrine factor that was important in explaining the beneficial effect
00:15:00.20 of MSCs.
00:15:01.28 Now, at this point in time, we began to think that it might be possible to translate this
00:15:08.10 work to patients with lung injury.
00:15:11.07 But we did not have a source of clinical grade mesenchymal stem cells until we fortunately
00:15:17.21 found a program funded by the NIH at the University of Minnesota,
00:15:23.24 the Production Assistance for Cellular Therapies Program.
00:15:27.21 And from this program we were able to obtain clinical grade human bone marrow-derived MSCs,
00:15:33.21 all allogeneic, for all of our subsequent preclinical studies.
00:15:37.22 So, I'll show you now our work that tested the ability of these cells to treat lung injury
00:15:44.20 in the human lung.
00:15:45.26 So, the overall hypothesis is shown on this slide, namely that the clinical grade cryopreserved
00:15:52.21 allogeneic human mesenchymal stem cells would decrease lung injury in the perfused human lung,
00:15:57.26 much like I've shown you already, but this now is with the cell that we potentially
00:16:02.14 would use in patients.
00:16:04.15 And we tested whether the cells would restore alveolar fluid clearance after entry with endotoxin.
00:16:10.21 And we also wanted to test if giving the cells by the intravenous route, in the perfusate,
00:16:15.05 would be as effective as intratracheal installation.
00:16:18.18 And we also wanted to test that the MSCs would home to the site of injury in the airspace
00:16:23.28 of the lung.
00:16:24.28 So, this slide shows you alveolar fluid clearance, again, on the y-axis.
00:16:30.16 And indicates that the cells we obtained from the University of Minnesota, the allogeneic
00:16:35.11 human bone marrow-derived MSCs, were effective in restoring alveolar fluid clearance when
00:16:41.28 they were given intratracheally or intravenously.
00:16:45.21 And you can see the control cells, normal human lung fibroblasts, had no effect.
00:16:50.05 So, this was encouraging, and in fact we also found in other studies that the cells,
00:16:56.13 even when they were injected intravenously, did migrate to the air spaces of the lung from
00:17:01.21 the circulation.
00:17:04.22 We also measured in these experiments the ability of the cells to reduce pro-inflammatory
00:17:09.25 cytokines.
00:17:11.20 And in this slide you can see that the levels of interleukin-1 beta and the levels of interleukin-8
00:17:17.09 were markedly reduced in the endotoxin-injured lungs, whether we gave the cells
00:17:22.03 by the IT or the IV route.
00:17:25.04 So, then we went ahead and tested the ability of the MSCs to work in the human lung using
00:17:31.27 live bacteria, as we had done in the mouse studies.
00:17:36.05 And again, the hypothesis was that the cells might work to reduce lung injury, even in
00:17:41.24 the presence of high quantities of live bacteria.
00:17:45.04 So, in these experiments, we again gave the cells intratracheal or intravenous to test
00:17:51.00 their respective efficacy.
00:17:53.10 And again... again, this slide shows alveolar fluid clearance on the y-axis.
00:17:57.17 And you can see that the E. coli reduced alveolar fluid clearance to about 80% of normal,
00:18:04.25 whereas intratracheal or intravenous MSCs
00:18:07.26 restored alveolar fluid clearance to about two-thirds of normal,
00:18:11.27 and the normal human lung fibroblasts had a modest effect, but not as potent as
00:18:15.28 the MSCs.
00:18:18.08 We then studied, also, how the human MSCs worked in terms of reducing inflammation in
00:18:26.22 the presence of live bacteria.
00:18:29.02 And we found that, as shown on this slide, that the number of neutrophils, as graphed
00:18:33.09 on the y-axis, was markedly increased, as would be expected with a high dose of E. coli bacteria,
00:18:39.26 whereas the intratracheal and intravenous MSCs reduce the number of neutrophils dramatically.
00:18:45.13 Now, it's important to know that in this perfused human lung preparation, we perfuse the lungs
00:18:50.25 with a 5% albumin solution, but we also added fresh blood, approximately mLs,
00:18:57.15 from one of us in the laboratory.
00:18:59.04 So, the perfused human lung contained red blood cells, white blood cells, neutrophils,
00:19:04.10 monocytes, and platelets.
00:19:06.04 And that's why there is a vigorous response in terms of neutrophil recruitment with the
00:19:10.14 presence of the bacteria, and yet a remarkable decrease in the number of neutrophils
00:19:15.25 with the MSCs.
00:19:17.22 This slide shows the histology and pathology from several of the human lung studies
00:19:22.23 that we have done.
00:19:23.28 You can see in the panel in... marked "control right upper lung" that the lung actually appears
00:19:29.07 completely normal, similar to what Dr....
00:19:32.02 Dr. Calfee showed you in her slides with the diagram of the uninjured lung.
00:19:36.20 The air spaces are not injured in any way; they're air-filled.
00:19:41.15 In contrast, in the E. coli pneumonia experiments, you can see that the lungs have
00:19:47.05 inflammatory cells and excess fluid.
00:19:50.04 When the lungs were treated with intratracheal normal human lung fibroblasts as a control
00:19:55.23 there was no... no effect whatsoever in terms of changing the degree of injury.
00:20:01.01 In contrast, all three of the lower panels with intratracheal or two different doses
00:20:06.07 of intravenous MSCs showed that the degree of injury is markedly reduced and virtually
00:20:11.11 returned to normal after four hours.
00:20:14.19 Now, we also needed to study, as we did in the mice, the antibacterial effects of the
00:20:20.20 mesenchymal stem cells.
00:20:22.22 And on this slide, you can see the... that in fact either intratracheal or intravenous MSCs
00:20:29.04 reduce the absolute bacterial counts in the injured lung.
00:20:32.22 E. coli... the number of E. coli in the alveolar fluid is shown in the y-axis, and you can
00:20:37.28 see with intratracheal MSCs it was reduced dramatically.
00:20:42.21 And then there was a dose-dependent decrease in the number of bacteria with intravenous MSCs,
00:20:47.27 similar to the findings in the mouse.
00:20:51.08 We also found in the human lung preparation that part of the mechanism for this beneficial effect,
00:20:57.02 as had been the case in the mice, was enhanced phagocytosis of the bacteria
00:21:02.22 by monocytes and alveolar macrophages.
00:21:06.13 Now, this slide is designed to show you various paracrine pathways that seem to explain
00:21:14.21 some of the beneficial effects of mesenchymal stem cells.
00:21:17.22 I've already told you about KGF.
00:21:20.14 And there are other paracrine molecules we and other investigators have studied, including,
00:21:24.19 for example, angiopoietin 1, which is released in high quantities by the MSCs.
00:21:29.08 There are other pathways that are important, for example,
00:21:32.12 the release of interleukin-1 receptor antagonists,
00:21:35.19 prostaglandins, and also, as I mentioned, antimicrobial peptides.
00:21:40.03 So, it seems that one major mechanism for benefit is by the release of paracrine molecules
00:21:45.17 that reduce injury.
00:21:47.07 Now, it also turns out that there are other studies from colleagues in the field
00:21:53.01 that suggest that some of the beneficial effect may also occur by what we call
00:21:57.13 cell contact dependent mechanisms.
00:21:59.27 So, in this slide, I'm showing you work from Dr. Bhattacharya's group at the...
00:22:04.20 at Columbia University,
00:22:07.04 where he found that the mesenchymal stem cells, when given directly into the airspace
00:22:10.27 of the lung, attach to the alveolar epithelium by Connexin 43 dependent pathways, and actually
00:22:17.22 delivered mitochondria from the MSCs to the injured alveolar epithelium.
00:22:23.04 And the injured alveolar epithelium appears to have injured mitochondria do not...
00:22:28.21 that do not perform normally.
00:22:30.23 And the conclusion from his studies was that the mitochondrial transfer
00:22:35.05 rescues the alveolar epithelium and allows it to function better
00:22:39.25 by restoring ATP levels to a more normal level, allowing the cells to work normally...
00:22:45.15 more normally for fluid transport, surfactant release, and other barrier properties.
00:22:50.04 So, this slide is intended to show you that there are probably two major pathways of benefit:
00:22:55.20 those that I described in more detail, the paracrine pathways, with release of molecules
00:23:00.21 such as KGF, and then the cell contact dependent pathways.
00:23:05.13 Now, our job at this point was to further test, in a preclinical model, whether these
00:23:11.28 cells would actually be safe when we gave them to patients.
00:23:16.28 And in order to accomplish this, we formed a collaboration with the University of Texas,
00:23:22.04 Dr. Traber's group, and tested the mesenchymal stem cells in a well-developed model of
00:23:28.12 severe lung injury in sheep.
00:23:30.11 This is a 24 hour model in which the sheep were given hot cottonwood smoke when they
00:23:35.15 were anesthetized and then had high doses of Pseudomonas aeruginosa,
00:23:40.24 a common clinical pathogen, instilled into both lungs.
00:23:45.17 And then the sheep develops severe lung injury.
00:23:48.15 They require mechanical ventilation and care, just like a patient in the intensive care unit.
00:23:55.11 And our first purpose here was to be sure that there were no safety issues when we gave
00:23:59.24 these mesenchymal stem cells, again, the clinical-grade cells from the University of Minnesota,
00:24:04.27 to the sheep.
00:24:06.11 This figure shows you cardiac output on the y-axis over time, over 24 hours, and you can
00:24:12.21 see, in yellow, the control sheep treated with Plasmalyte A alone,
00:24:17.16 versus a lower dose and a higher dose of MSCs --
00:24:20.22 lower dose in red and the higher dose in white.
00:24:24.12 And this was five million cells per kilogram for the lower dose... five million cells per kilogram
00:24:30.08 compared to ten million cells per kilogram, the doses we anticipated we might
00:24:34.16 use clinically.
00:24:35.27 And what you see here is that cardiac output was not different over the 24 hours in the...
00:24:40.15 among the three groups.
00:24:42.02 Similarly, we wanted to particularly test if there would be any adverse effects on the
00:24:46.18 pulmonary circulation from giving these mesenchymal stem cells.
00:24:51.03 We know from experimental studies that the MS CS largely collect and are trapped in the
00:24:57.04 pulmonary circulation for several hours.
00:24:59.17 They are slowly released from the lung over time and their actual half-life in the body
00:25:05.22 is probably limited to no more than 7 to 14 days.
00:25:09.28 Now, in these experiments, we measured the pulmonary arterial pressure and we calculated
00:25:14.18 pulmonary vascular resistance.
00:25:16.18 And again, over 24 hours, you can see there was no difference among the three groups.
00:25:20.20 So, there were no safety issues, no indication that if we gave the MSCs into the
00:25:25.24 acutely injured lung that the... that the sheep would become hemodynamically unstable.
00:25:31.24 What about parameters of efficacy?
00:25:34.18 Well, the most important one we follow clinically in patients is oxygenation.
00:25:39.06 And this slide shows on the y-axis the PaO2/FiO2 ratio.
00:25:45.27 This is a way in which we measure the degree of hypoxemia patients have.
00:25:50.19 And what you can see is that all three groups had a dramatic decrease in their oxygenation
00:25:56.25 over the first 12 hours.
00:25:59.09 And this... you can focus right in here, at 12 hours, and see that all three groups have
00:26:05.16 a PF ratio in the range of 100 to 200, which... which is very similar to what we see in patients
00:26:11.11 with severe respiratory failure from ARDS.
00:26:15.14 But note that the sheep that were treated with either the lower dose or the higher dose
00:26:19.23 of hMSCs had a marked improvement in their PF ratio at the end of 24 hours.
00:26:26.06 And... whereas the control group continued to decline and ended up at a much lower level.
00:26:31.25 So, this was encouraging because it indicated
00:26:34.12 the mesenchymal stem cells could improve oxygenation.
00:26:38.12 In this model, the mesenchymal stem cells were given intravenously over 1 hour,
00:26:43.08 1 hour after the development of the lung injury.
00:26:47.19 And finally, after the sheep were sacrificed, we measured the quantity of pulmonary edema
00:26:53.03 in the lung in all three groups.
00:26:55.27 And this is expressed, as it was in those studies I showed you before in the human lung,
00:27:00.08 on the y-axis as the wet-to-dry ratio.
00:27:03.17 And please note that the sheep that were treated with the higher dose MSCs, in the white bar,
00:27:09.10 had a significantly reduced quantity of pulmonary edema in the lung.
00:27:13.12 So, for this reason, our target dose for the clinical studies is to give patients the
00:27:20.15 higher dose of 10 million cells per kilogram, because it appears,
00:27:24.03 on the basis of this large animal sheep model,
00:27:26.25 that it might be more effective than even the 5 million cells.
00:27:30.14 Both of them improved oxygenation, but the higher dose reduced the quantity of pulmonary
00:27:36.13 edema in 24 hours.
00:27:38.16 So, the conclusions from the sheep studies are summarized here.
00:27:42.13 There were no safety concerns from administrating human MSCs at these doses, given one hour
00:27:48.03 after the injury.
00:27:50.04 There was evidence of efficacy, with significantly improved oxygenation
00:27:53.26 with both the lower and the higher dose MSC treatments.
00:27:57.24 Pulmonary edema was less with the higher dose MSCs.
00:28:01.00 Thus, as I explained, the clinical trial for ARDS...
00:28:04.01 ARDS is designed to test the safety and efficacy of the higher dose of MSCs.
00:28:10.15 So, now, how do we proceed with testing the allogeneic human MSCs in patients with ARDS?
00:28:18.19 Well, the translation objective all along has been to potentially test these
00:28:23.21 bone-marrow derived allogeneic cells in a clinical trial of severe ARDS.
00:28:29.04 We fortunately received funding for this trial from the NIH,
00:28:32.15 through the National Heart Lung and Blood Institute, by a U01 mechanism.
00:28:36.27 Our goal was to focus the trial on safety and respiratory efficacy endpoints, such as
00:28:42.14 oxygenation and pulmonary dead space fraction.
00:28:46.05 And we wanted to design the trial, also, to test potential biologic pathways of benefit
00:28:52.01 by sampling the plasma and the airspace of the lung to see if, for example,
00:28:56.06 anti-inflammatory and pro-inflammatory cytokines were changed,
00:29:00.03 as we found in the preclinical animal studies.
00:29:04.09 This slide shows our plans for the Phase 1 and Phase 2 trial of human mesenchymal stem cells
00:29:10.06 for severe ARDS.
00:29:12.18 The target population will be those patients with severe ARDS as defined by a PF ratio
00:29:18.17 less than 200.
00:29:20.13 The rationale here is to treat patients with a higher predicted mortality in hopes that
00:29:25.02 we can reduce their mortality.
00:29:26.26 For the Phase 1 trial, we will test the safety of giving three different doses of hMSCs:
00:29:34.11 1, 5, or 10 million cells per kilogram.
00:29:38.02 If we are successful in demonstrating safety with the
00:29:41.11 highest targeted dose of 10 million cells per kilogram,
00:29:44.18 then we will proceed to the Phase 2 trial.
00:29:47.16 In the Phase 2 trial, we would treat 60 patients.
00:29:50.21 A primary goal of the Phase 2 trial is still safety, but we will also be looking at specific
00:29:55.27 efficacy endpoints.
00:29:57.20 In the Phase 2 trial, there would be a two-to-one randomization, in which we would treat
00:30:02.05 40 patients with the MSCs and 20 patients with placebo, and it would be double-blind.
00:30:09.14 The next slide shows that, in the Phase 2 trial, safety, as I indicated, is still the
00:30:14.18 primary objective, although we will have secondary endpoints designed to test for potential efficacy.
00:30:21.09 And we'll accomplish this by measuring oxygenation, pulmonary dead space, the ventilator-free days,
00:30:27.17 mortality, systemic organ failure, as well as biologic markers that may indicate
00:30:33.10 that the MSCs are effective in patients.
00:30:36.07 If all of this goes well, and the Phase 2 trial shows efficacy, then potentially
00:30:41.09 we could go on to do a large, pivotal Phase 3 trial that would be designed to test the efficacy
00:30:47.23 in a much larger number of patients.
00:30:50.19 The next slide just summarizes much of what I have shown you, a roadmap for translation
00:30:58.05 of human mesenchymal stem or stromal cells to clinical trials for ARDS.
00:31:01.25 The work began in 2006 as I showed you with mouse experiments, and then included the
00:31:08.09 ex vivo perfused human lung experiments, some rat studies I did not show you,
00:31:12.24 the sheep studies that I have shown you, and then considerable work to prepare a IND,
00:31:18.09 which the FDA approved and gave us our IND in January of 2013.
00:31:24.13 The... the mechanism for support for the trial, as I mentioned, is the NIH U01 grant.
00:31:31.05 And we have now started the Phase 1 trial and in fact have enrolled four patients
00:31:35.11 so far and anticipate needing to enroll nine patients in the Phase 1 to establish
00:31:42.08 safety of the doses we proposed, and then to move on to the Phase 2 trial.
00:31:46.19 So, in conclusion, first, preclinical studies support the potential value of human MSCs
00:31:53.24 to provide a novel effective therapy for severe ARDS.
00:31:57.11 Future stat... strategies may not be confined to the MSCs themselves, since paracrine products
00:32:04.06 are important, as I've showed you, such as KGF, emphasizing the need to understand the
00:32:09.17 cell contact dependent and independent mechanisms in all of the studies that we do,
00:32:14.15 and others are carrying out.
00:32:17.00 The other caveat here is that although animal models are essential to test new therapies
00:32:21.22 for lung injury, it is difficult to model ARDS in the animal setting because it is
00:32:27.21 a complex syndrome that includes multiple comorbidities.
00:32:32.15 Patients often have liver disease, diabetes, variable degrees of alcohol exposure,
00:32:37.19 cigarette smoke exposure.
00:32:39.23 And then there's several process of care variables that can influence the clinical outcomes,
00:32:44.06 including how the patient is mechanically ventilated, fluid therapy,
00:32:49.07 and vasopressors and sedatives.
00:32:50.07 Dr. Calfee discussed in particular the importance of mechanical ventilation and fluid therapy,
00:32:55.18 and we're of course designing our current trial to standardize those in the patients
00:33:00.11 in the trial.
00:33:01.15 Thus, we believe that human trials are needed to test safety and efficacy, and to learn
00:33:06.20 more about the mechanisms of how MSCs work in patients, findings that might be of value,
00:33:12.27 potentially, and even considering MSCs for the treatment of other lung diseases.
00:33:17.22 I want to acknowledge the contributions of several investigators who have made this work possible,
00:33:23.06 in particular Dr. Jae Woo Lee, an Associate Professor at UCSF who carried out
00:33:27.17 the ex vivo human lung experiments.
00:33:29.21 Dr. McKenna at the University of Minnesota.
00:33:31.26 He and his group has... have provided the clinical grade MSCs.
00:33:35.13 Dr. Gupta, who is now at The Scripps Institute, did the... some of the initial work in mice.
00:33:41.07 Anna Krasnodembskaya also did some of the critical mouse work.
00:33:46.02 Dr. Traber was a crucial collaborator at the University of Texas.
00:33:51.00 And at UCSF, Dr. Kathleen Liu, Carolyn Calfee, who you've already heard today,
00:33:56.03 and Dr. Jenny Wilson have made major contributions to organizing this trial and carrying it out.
00:34:02.12 And of course we're grateful to the National Heart Lung and Blood Institute for supporting
00:34:06.02 the work.
00:34:07.10 And most importantly, perhaps, we are grateful to the patients who consent to be part of
00:34:11.26 our clinical trial... trials, and our observational studies.
00:34:15.24 And of course the nursing and respiratory staff in the intensive care units at UCSF.
00:34:21.01 Thank you.

Videos in this Talk

Part 1: Acute Respiratory Distress Syndrome: An Overview
Audience:
- Researcher
Part 2: Mesenchymal Stem Cell Therapy for ARDS
Audience:
- Researcher

Speakers: Carolyn Calfee, Michael Matthay

Total Duration: 0:46:10

Recorded: November 2013

All Talks in Human Disease

Talk Overview

Acute respiratory distress syndrome is a life threatening condition with few effective treatment options. Preliminary studies using mesenchymal stem cells, or stromal cells, to treat ARDS have shown promise with decreased levels of bacteria in the lungs, reduced pulmonary edema and improved oxygenation. In Part 1, Dr. Calfee begins by explaining that acute respiratory distress syndrome (ARDS) is pulmonary edema, or fluid in the lungs, not due to heart failure. It is a condition that affects 200,000 people/year in the USA with a 30-40% mortality rate. During ARDS, there are many cellular changes with complex pathophysiology making it extremely difficult to treat. Currently, patients are treated by ventilation with low tidal volume and fluid conservative therapy as many pharmacological interventions have failed. Mesenchymal stem cells (MSC), however, may hold promise as a treatment.

In Part 2, Dr. Matthay provides the rationale behind treating ARDS patients with MSC. Initial studies in a mouse model of ARDS, showed that treatment with MSCs increased levels of anti-inflammatory cytokines and antimicrobial peptides in the lung, and increased phagocytosis of bacteria by monocytes. Further studies in ex vivo perfused human lungs and in sheep with severe lung injury showed that treatment with MSCs improved oxygenation and reduced pulmonary edema. An NIH/NHLBI supported phase 1 clinical trial for safety has been completed and a randomized, blinded phase 2 trial has now been initiated to test the safety and efficacy of MSC treatment in human patients with ARDS.

Speaker Bio

Carolyn Calfee

Dr. Carolyn Calfee is an Associate Professor of Medicine and Anesthesia at the University of California, San Francisco and a member of the UCSF Cancer Center and the Center for Tobacco Control Research and Education at UCSF. Calfee’s clinical and research focus is the diagnosis and treatment of acute lung injury and acute respiratory distress… Continue Reading

Michael Matthay

Dr. Michael Matthay was an undergraduate student at Harvard University and a medical student at the University of Pennsylvania School of Medicine. He completed his training in Internal Medicine at the University of Colorado and in Pulmonary/Critical Care Medicine at UCSF. Currently, Matthay is a Professor of Medicine and Anesthesia at the University of California,… Continue Reading