Prof. Luciano Gattinoni describes some pathophysiological and clinical aspects of COVID-19 patients



We met Professor Luciano Gattinoni, one of the leading experts of anesthesia and resuscitation, former Scientific Director of the Polyclinic of Milan, emeritus Professor of the Universities of Milan and Gottingen, Germany, where he now lives, and we asked him some questions about the infection from SARS-CoV-2.r
Hello Professor, what do you think is the pathophysiology of lung damage caused by COVID-19?

I think the biggest problems happen on a vascular level. In fact, several colleagues are informing me of the presence of many endothelial cells in the blood. Why do I say that it is a vascular problem, beyond the autopsy reliefs? The reason is very simple. These patients, in the initial phase of the disease, are characterized by an extraordinary dissociation between the amount of gas inside the lungs, i.e. lung elasticity and arterial hypoxemia. Normally, in acute respiratory insufficiencies that we classify as ARDS, low pulmonary compliance is associated with hypoxia. It is very rare, a condition that personally, in 40 years of activity, I had hardly ever seen, a dissociation of this kind. Patients with almost normal compliance who present with a monstrous hypoxemia, with PaO2 values ​​of 35-40 mmHg and a saturation of 65-70%. So what is the only explanation for such a phenomenon? The gas exchange is mathematics, it is not supposition. Basically we in these patients have a huge alteration of the ventilation / perfusion ratio. In these patients the regulation of pulmonary flow is lost, the flow becomes more gravity dependent. Blood flow and ventilation therefore follow two different paths. the flow goes on one side and the ventilation goes on the other. We have an increase in dead space, for example.r
At this stage, the lung does not yet have collapsed areas and in imaging we see some ground glass aspects, especially of the mantle. They are mainly mantellar because it is at the mantle level that the stress and strain are most concentrated during lung ventilation. Below the pleura, between the pleura and the alveoli, there are differences in elasticity, so we have a concentration of the pressures and pulmonary stretch in those areas, which are then the ones that first highlight the damage that is known to the radiological vision. This is the ground glass vision, the effect of a very localized and also very modest edema.
How does the patient respond to hypoxia? When the levels of 40-50 mmHg are reached, our respiratory centers begin to increase the tidal volume. So imagine breathing twice as normal, increasing your tidal volume by 30-40%. This also occurs in the absence of dyspnea, because since compliance is very good, when the patient increases the tidal volume it is true that he makes a certain muscular effort, but it is also true that the muscular sense corresponds to the sense of filling that he expects, so he has no wheezing. In fact, when a sportsman runs, he has no dyspnea. Dyspnea occurs when a certain muscular effort is made to forfeit air, the brain awaits a certain sensation of filling, and instead the air does not arrive.r
By making these efforts, the more time passes, with the increase of these tidal volumes, obviously the patient must also increase the negative pressure inside the pleura. The increase in negative pressure inside the pleura only attracts more blood centrally. The amount of central blood increases, filtration increases in an inflamed lung, edema increases. Edema causes the lung to become like a wet sponge. The lung becomes heavier and tends to collapse on itself. Then the lung imaging shows a greater ground glass, you can see completely white areas. The white areas are noticed at the bases. This is the effect of crushing the lung. At this point the evolution of the disease leads to an aggravation of the patient's condition. Even if the resuscitator manages to balance the PaO2, increasing it, these patients are characterized by a greatly increased respiratory drive, which I believe depends precisely on the viral infection, but we still do not know by what precise mechanism.
The progress of the disease is therefore due, in addition to the severity of the infection, also to these mechanical factors. When the patient has severe dyspnea, with a CT picture showing severe lung impairment, then only then does the patient return to the so-called ARDS. The problem in my opinion lies in treating a patient from the beginning who presents with severe hypoxemia as if he had an ARDS. Here a notable mistake is made, with consequences that are also important.r
In a patient with ARDS, the parameters in ICU must be kept stable and wait for the storm to pass. I believe that the ICU mortality of these patients should not be higher than 15-20%. It is currently much higher, I fear we are around 40%, because it is not possible to treat patients as they should. They are patients who are hospitalized in intensive care late, perhaps after 4 days of non-invasive ventilation, CPAP helmet, all the various remedies that are told around. They are patients on whom drugs are attempted without precise scientific evidence. Medicine is not based on anecdotal (especially on the one that runs on the Internet), but on well-founded evidence. I know for sure one thing, that until the mechanisms are known, therapy cannot be adequate. Without scientific evidence, the administration of any drug does not make much sense.r
With a view to limited resources, their correct attribution plays a fundamental role. So it is necessary not to hospitalize the patient who does not need it, but also and above all to hospitalize what he needs, with the correct timing. Which do you think is the correct criterion for hospitalization from the territory? For example: tout court oximetry, walking test oximetry, ultrasound - with a view to being able to do it?

I would have to make a measurement that I don't think anyone detects, or measure the inspiratory effort, the negativity of the pleural pressure that the patient presents at that moment. If the patient has a very negative pleural pressure, which exceeds 12-15 cmH2O, we know for sure, we have very extensive experimental evidence on the matter, that a healthy lung is destroyed in 24-48 hours. If a patient who has dyspnoea arrives at the observation, what I would do immediately is measure the pressure with an esophageal balloon, or I would focus on the aspect of respiratory mechanics: on the contraction of the sternocleidomastoid, on the diaphragmatic and intercostal movements, on the movements of the abdomen , or I would evaluate variations of the central venous pressure, which go hand in hand with the variations of the pleural pressure, I would use the ultrasound to analyze the displacements of the diaphragm. We have several systems that quantify respiratory effort.r

A patient who does not present respiratory effort reassures us. It may get worse, of course, but we will notice if respiratory effort increases. At this point, when I evaluate the presence of excessive respiratory effort, I can try to use a CPAP helmet. Let's imagine putting the CPAP at 15 cmH2O. For some patients, it is the right therapy, because the respiratory effort decreases, therefore the vicious circle that leads to aggravation is interrupted. But for other patients not, because the respiratory effort may not decrease and remain very high. Furthermore, with a pressure of 15 cmH2O, there is the risk of causing a crisis in the right ventricle. Positive pressure in fact involves hemodynamics problems. Patients may become hypotheses, noradrenaline is administered, with the risk that after 24 hours they show renal failure. At that point, of course, you need intubation, a ventilator, a doctor and two nurses to take care of it.r
Several hospitals found themselves faced with a disproportionate demand for intensive care in the face of reduced supply. Of necessity, several patients were assisted with non-invasive methods (CPAP, NIV), more manageable and less expensive in terms of means and resources. What do you think about it?

I struggle to answer this question. In this situation what you do is fine, I feel like saying. CPAP helmet, 10-15 cmH2O of PEEP and let's see how it goes ... But this is not medicine, this is war medicine. In which scores can also be used to decide which patients are worth visiting and which are not. Medicine is another thing.r
In several hospitals, the increase in intensive care places has become necessary, without a corresponding increase in intensivist doctors and critically ill nurses. In a recent TV interview, you have restored the dignity of the profession by calling it 'paranoid attention to detail by people who know how to do it'. Placing multiple tubes, in the face of the obvious lack of attention to detail that derives from them, alone is not a parameter capable of increasing survival. With a view to working 'as best as possible' and not acting as a chisel as is usually done in intensive care, what 'details' do you consider to be priorities in COVID patients?r

My colleagues are making huge efforts to handle this. In Italy, but not only.r
I have colleagues who in 40 years of experience have never found themselves facing scenarios like the ones we are experiencing today. We have no drugs, the beds are missing, the opportunity and the time to observe are lacking. Intensive care requires 1 nurse per 2 beds. If where 10 patients were treated today 40 are treated, with the same number of doctors and nurses, that is no longer intensive care. 1 doctor for every 10 patients is not intensive care. Therapy is not done with a bed and a respirator. We also need those who can dedicate themselves to patients by dedicating the right time, who can do it. These are not things that can be improvised in two weeks.r
Intensive care has capable doctors and nurses, accustomed to critical situations, to patients who do not respond to treatment, to grueling shifts. I believe intensive care is doing well to support the weight of this emergency.r
The real problem, in my opinion, is outside intensive care, in the so-called sub-intensive where, by necessity, doctors who usually do something else are working. We have doctors who are now knowing what a CPAP is, who are using it without knowing its mechanisms. Imagine what could happen if they told me tomorrow to go to the operating room to operate a femur fracture. We are not talking about medicine, I repeat.r
I understand that in these early stages of emergency systems other than ordinary had to be found. But now, hopefully as soon as possible, we will have a long period in which the number of patients will decrease and we can go back to treating them. Mortality will decrease, but we will have to have data, accumulate experiences and errors on how to treat these patients.r
I believe, I repeat, that the fundamental mistake today is to treat these patients as if they all had severe ARDS. As they cross the hospital threshold, they put on a CPAP helmet, 15 cmH2O of PEEP and see what happens, limiting themselves to assessing oxygenation alone. Oxygenation is the last sign of an upstream process that needs to be known. If these patients are not known, they are not treated properly. Now there is no data, but soon, I hope, we will have it.r
In some hospitals the technique of the prone position of patients in spontaneous breathing and / or assisted with CPAP is spreading and the saturation often rises considerably. What do you think? Can it be really advantageous or can it mask the severity of the picture and dangerously delay the moment in which intubation occurs?r

At the Policlinico di Milano we were among the first to introduce the prone position. In the literature, we discovered it later, there was only an anecdotal article from 1972 in which a doctor reported having put patients on their stomachs and having noticed an improvement in oxygenation.r
Looking at the pulmonary CT of patients with acute respiratory failure in the supine position, we saw that all the lesions were substantially dependent, i.e. in the areas of the lung closest to the vertebrae, while at the top, towards the sternum, the lung did not seem so compromised. Radiography alone, however, told us that the whole lung was involved. Then the most natural thing seemed to us to put the patients on their stomach. We noticed that oxygenation was rising. Why did oxygenation rise?r
We thought that the increase in oxygenation was motivated by the flow shift in that part of the lung with lower densities observable by CT, what we called baby lung. By subjecting the patients in a prone position to CT we were surprised to see an increase in density towards the sternum. Thus was born the model, still valid, of the lung similar to a sponge.r
Imagine a dry sponge, which has pores of a certain size. So let's imagine putting this sponge in the water. Let's remove it from the water and look at it resting on the edge of the bathtub. It will be seen that the pores at the bottom decrease in volume because they are crushed by the weight of the overlying sponge. By placing the patient in a prone position, it is as if the lung / sponge forced the air to come out from the back. By releasing the posterior parts of the lung that were previously squeezed and compressing the anterior ones a bit, since perfusion remains uniform, unlike what we previously thought, then in this way there is a greater portion of 'open' lung, which increases oxygenation. This is the main mechanism, then there are 2-3 other minor mechanisms.




Patients with early COVID-19 do not have edema, that is, a lung that looks like a large sponge full of water. In the initial phase there is an altered ventilation / perfusion ratio. So, by placing these patients in a prone position, we get benefits thanks to the first mechanism I described. Since we have a perfusion that tends to go gravitationally, when we put patients on their stomachs, we put the non-dependent part of the lung in more favorable conditions of perfusion, i.e. the greater mass of the lung that is upward. So with a different mechanism, the end result is the same. Obviously if the situation worsens and the edema begins to increase, when the lung becomes a nice sponge to be clear, the prone position gives us benefits through the second mechanism.
What is being seen now, even with CT, is the consideration for this. There are two stages of the disease. In a first phase patients have 'L-light' lungs (high compliance, low V / Q): in these patients the lung does not weigh so much so I do not have the sponge effect. In a second phase the lungs become 'H-heavy' (more similar to the ARDS framework) and will have a redistribution of densities as in the classic ARDS. The very first data we have go in this direction.
When to extubate a COVID-19 patient in your opinion? What is the most correct setting in the face of limited resources? The maintenance in intensive care or the transition to pneumology / rehabilitation?

Extubation will only be possible when, with a decrease in support, these patients will not start with a terrifying respiratory drive. If, on the other hand, they manage to breathe proportionally, they can obviously be weaned.
The average length of stay in an intensive care unit is about ten days both for patients who are performing well and showing signs of recovery, and for patients who deteriorate to death. Then we have a third group of patients, about 60%, who have an uncertain trend, for which we cannot predict an outcome. Here it takes much longer. In this pathology the thing you need to have is patience. An improvement in the gas exchange, which can also occur quickly, must not correspond to a fan tinkering. In fact, the patient's condition has certainly not changed. You have to wait for the disease to run its course and be patient. One of the most difficult things in ICU is doing nothing, but that's what needs to be done in these patients.
Patients that I would all keep in intensive care, because in the Italian reality it is the resuscitators who have the greatest skills in the management of patients with severe lung infection.