Diagnostic algorithms are the tools used by clinicians to rationally advance from the symptoms and signs causing the patient to seek medical help (usually grouped as syndromes) to the diagnosis of a specific disease that may benefit from guided treatment.23 “Disease” is possibly one of the most commonly used words in Medicine, and although the term is very familiar to us, its definition is a great challenge.24

From the semantic perspective, the dictionary of the Spanish National Royal Academy of Medicine defines disease as a “series of alterations, signs and symptoms that organize according to a concrete scheme in time and space in response to a concrete cause, and which manifest in a similar way in different individuals, thus allowing classification and identification of the different disorders”.25

From the medical and clinical epidemiological perspective, the relevant events or variables are only those which: (a) affect mortality (death); (b) allow the diagnosis of a disease; (c) have an economical impact (destitution); or (d) cause discomfort; (e) disability; and/or (f) dissatisfaction in the patient (the so-called “6Ds”: death, disease, destitution, discomfort, disability, dissatisfaction).26 Thus, from this perspective, a disease is defined as a series of signs and symptoms implying similar 6Ds. Evidently, there may be certain differences leading to the identification of different subtypes, classes or endophenotypes, but these cannot condition substantial variations in the 6Ds. As an example, up until only a few years ago a cancer could be classified simply on the basis of the histology and TNM classification. At present this approach is clearly insufficient, and the genetic, epigenetic and proteomic differences also must be addressed in order to subclassify most cancers and thus optimize the diagnosis, prognosis and treatment.27,28 The magnitude of the difference between two conditions required in order to consider them to be two different diseases or subtypes of one same disease remains to be established, though it undoubtedly will be based on the combination of clinical (6Ds), histological (e.g., light microscopy, confocal microscopy, etc.) and molecular factors (e.g., genomics, proteomics, metabolomics, etc.).

In 1935, Louis Hamman and Arnold Rich described a histological pattern (which later became known as the organizing phase of DAD) in four patients who died due to hypoxemia of unknown cause, characterized by the diffuse interstitial proliferation of fibroblasts.29,30 In 1986, Katzenstein et al.,31 with the purpose of reducing the confusion between the clinical course and the histological findings, proposed two terms: (a) acute interstitial pneumonia (AIP) referred to patients with an acute course and DAD (previously referred to as Hamman–Rich syndrome); and (b) usual interstitial pneumonia or idiopathic pulmonary fibrosis referred to patients with a chronic course and DAD.31,32

The consideration of ARDS adds a new confounding factor, since DAD is a diagnostic criterion of acute interstitial pneumonia, but also a hallmark of ARDS.33 In other words, if the pathologist observes findings consistent with DAD (Fig. 1) and the patient does not meet the clinical criteria of ARDS, the diagnosis should be acute interstitial pneumonia or usual interstitial pneumonia. In contrast, in the presence of clinical criteria of ARDS, the diagnosis should be ARDS with DAD.

Figure 1.

Diffuse alveolar damage (DAD). Case 1: A 42-year-old patient admitted to the intensive care unit (ICU) with septic shock associated to acute nosocomial pneumonia. Death occurred due to refractory hypoxemia after two days on mechanical ventilation. The postmortem study (Fig. a, hematoxylin–eosin [HE], ×5) showed a heterogeneous histological pattern with areas of DAD in the exudative phase, with the presence of hyaline membranes and interstitial and alveolar edema, vascular congestion of the interalveolar septae, and desquamation and hyperplasia of the type II pneumocytes. In other areas (Fig. b, HE, ×5) DAD in the organizing phase with lax interstitial and intraalveolar fibrosis was noted, together with hyaline membranes and the rest of the previously mentioned findings. Case 2: A 28-year-old patient admitted to the ICU due to ARDS secondary to respiratory sepsis attributed to H1N1 influenza. Death occurred due to refractory hypoxemia after 48 days on mechanical ventilation. The lung tissue micrograph (Fig. c, HE, ×5) obtained at autopsy showed evolved DAD in the fibrotic stage, with great distortion of the lung parenchyma and practically complete substitution of the alveolar spaces with fibrotic tissue.

(0.85MB).

According to Katzenstein et al.,34 DAD is a specific reaction of the lungs to a range of aggressive factors. The common feature is endothelial and alveolar damage giving rise to the exudation of fluids and cells, with progression in some cases to extensive interstitial lung fibrosis. The most characteristic pathological finding of the acute phase of DAD is the presence of hyaline membranes, which under the light microscope appear as a homogeneous and eosinophilic material extending over the internal surface of the alveoli (Fig. 1a and b). These hyaline membranes are composed of the necrotic remains of epithelial cells and serum proteins that have penetrated from the bloodstream into the alveolar space secondary to an increase in permeability of the alveolar-capillary barrier. One of the most widely debated issues, and for which no firm answers have been obtained, is the need or not to include the presence of hyaline membranes in the diagnosis of DAD. At present, the studies involving the largest number of patients, as well as the only study demonstrating evolutive differences between patients with ARDS according to the presence or absence of DAD, have considered the presence of hyaline membranes to be essential for the diagnosis of DAD.22,35–38

The morphological features of DAD depend on the time elapsed from risk factor exposure to the moment of the histological study (Fig. 1).34,36,39 Schematically, three consecutive phases can be observed: an early or exudative phase (Fig. 1a) characterized by an intraalveolar exudate, hyaline membranes and a cellular infiltrate mainly composed of lymphocytes, plasma cells and macrophages; an intermediate or proliferative phase characterized by hyperplasia, atypia and mitosis of type II pneumocytes, with thrombosis of the small pulmonary arteries; and a final fibrotic or organizing phase (Fig. 1b) characterized by thickening of the alveolar-capillary membrane, the proliferation of fibroblasts particularly at interstitial level, and fibrosis. In the more severe cases, fibrosis can progress over weeks, restructuring the entire lung parenchyma and generating a characteristic “honeycomb lung” image, with practically total substitution of the alveolar space (Fig. 1c). It is important to mention that the different phases can overlap; as a result, different evolutive phases can be found in one same patient (Fig. 1).36

The evolution of the concept of ARDS since the classical term idiopathic pulmonary anasarca postulated by Laënnec in the year 1821 to the recent definition of Berlin33,40 has had a considerable impact on the designing of research and on the measurement of the variables implicated in the 6Ds.40,41 Of the many proposed definitions,21,33,42–48 only 5 have received the necessary acceptance to be regarded as genuine definitions of ARDS.21,33,42–44

The first definition of ARDS was proposed by Ashbaugh et al.21 in the year 1967, and although the lung findings in their 7 postmortem studies were described, no histological pattern was included in the definition.

The second definition of ARDS was proposed by Murray et al.44 in the year 1988, and considered three aspects: (a) the chronology of the event (acute or chronic); (b) a composite severity scale (chest X-rays, PaO2/FiO2 ratio, PEEP and lung compliance); and (c) the origin of the lung damage (caused by aspiration pneumonia, lipid embolism, drug use, inhalation of toxic agents or infection; or associated to sepsis, multiple blood transfusions, acute pancreatitis or disseminated intravascular coagulation). This definition likewise contemplated no histological pattern.

The third definition of ARDS was the result of two meetings (15 May 1992 in Miami, USA, and 25 October 1992 in Barcelona, Spain) auspiced by the American Thoracic Society and the European Society of Intensive Care Medicine.42 This definition considers ARDS (it includes the term acute lung injury for patients meeting the criteria of ARDS but with a PaO2/FiO2 ratio between 200 and 300) to be an inflammatory syndrome with increased lung permeability, associated to a range of clinical, radiological and physiological anomalies that cannot be explained, but which may coexist with increased left atrial pressure or pulmonary capillary hypertension. The definition also mentions that the lung histological findings may include damage to the endothelial and/or epithelial barrier, as well as a humoral and cellular inflammatory response. However, no mention is made of DAD, and no histological criterion of any kind is included in the definition.

The fourth definition of ARDS was proposed by Ferguson et al.43 in the year 2005, using the Delphi method. Although the general criteria submitted to the participants before the start of the study mentioned that ARDS is the “reflection” of DAD, neither the rest of the manuscript nor the final criteria made any reference to DAD.

The last definition of ARDS is called the Berlin definition, in acknowledgment of the city where the 2012 consensus meeting was held. In the preliminary version of the definition, the panel of experts agreed that ARDS is an acute, diffuse inflammatory type of lung lesion characterized by an increase in vascular permeability and the lose of lung aeration. The clinical hallmarks are hypoxemia and bilateral opacities associated to an increase in pulmonary shunting and physiological dead space. The characteristic morphological finding of the acute phase is composite DAD, e.g., with edema, inflammation, hyaline membranes or hemorrhage. However, the final version of the definition includes no reference to DAD, on the grounds that there is a poor correlation between ARDS and the finding of DAD in postmortem studies.49 In addition, it was underscored that not all the members of the panel of experts agreed with the idea of DAD as the only histological correlate of ARDS, and considered that pneumonia and non-cardiogenic lung edema could also be implicated.50

The present study received no type of financial support.

The authors declare that they have no conflicts of interest.