Discussion
The notable findings of this case series are as follows: (i) Tissue acquisition through ultrasound-guided postmortem biopsies was acceptable, particularly for the liver and the heart. A sufficient amount of lung tissue was obtained in five out of the seven patients. (ii) The universal radiological finding was the presence of GGO, particularly in the lower lung lobes. The pathologic counterpart of these images was the organizing phase of ARDS. All but one patient had hyaline membranes in their alveoli. (iii) This is the first study to report splenic postmortem biopsy results in COVID-19 patients. The common finding was necrosis and neutrophil infiltration in the red pulp.
Several studies and case reports have been published to understand histopathological changes and acquire possible ultrastructural clues underlying the widespread damage inflicted by SARS-CoV-2. As the lungs are the main target of the virus, pulmonary imaging and histopathology have attracted much attention. Three rough stages have been defined for COVID-19 [15]. The earliest phase is SARS-CoV-2 infection, characterized by fever, dyspnea, and other flu-like symptoms. The second stage involves viral pneumonia, which causes pulmonary inflammation and coagulopathy. Some cases progress to ARDS in this stage. The third and last stage is characterized by the relieving of symptoms and pulmonary fibrosis. These stages are not necessarily consecutive and can coexist in the same patient.
Pulmonary findings are histopathologically classified into three patterns [16-18]. The most common and earliest pattern is diffuse alveolar damage (DAD), which shows varying degrees of organization, desquamation, type-2 pneumocyte hyperplasia, and viral cytopathic changes. The second pattern (vascular) involves intra-alveolar fibrin deposition and/or microvascular thrombi. Fibrotic patterns include fibrotic DAD and/or interstitial fibrosis. A meta-analysis [19] evaluating autopsy studies revealed that pulmonary histopathologic examinations showed an epithelial pattern of injury in 85% of the included patients. When present, mononuclear cells were the predominant cell type infiltrating the lung interstitium. A vascular pattern was present in 59% of the patients, and a fibrotic DAD pattern was observed in 22% of the patients. Among the patients, 60% had more than one pattern concurrently. Our findings were generally in agreement with previous autopsy studies. The most common pulmonary injury pattern among our patients was epithelial DAD. It was present in four out of the five patients (80%), isolated in two patients, and alongside a vascular DAD pattern in the other two patients. Interstitial inflammation was noted in three patients (50%) and was mononuclear in all cases. One patient had a concomitant fibrotic and vascular pattern.
Only a few studies to date have reported postmortem biopsy findings in COVID-19 decedents [12-14]. In these studies, 18 COVID-19 decedents were examined through postmortem biopsies. One study [13] evaluated only lung tissue, and two [12, 14] studies also examined liver and heart tissues, in addition to lung tissue. Beigmohammadi et al. [12] performed postmortem biopsies with ultrasound guidance, whereas Tian et al. [21] performed blind biopsies based on reference points. Flickweert et al. [13] used ultrasound or CT guidance to perform postmortem biopsies. In these studies, the authors obtained sufficient lung tissue for histopathological examination. Conversely, sufficient lung tissue was not obtained from the two patients in the present study.
Similar to our study, Flikweert et al. and Tian et al. [13, 14] also included chest radiological findings in their reports. Regarding pulmonary histopathologic findings, diffuse alveolar damage was the predominant finding in all three studies.
Myocardial injury is prevalent in COVID-19 patients hospitalized in the ICU [20-22]. In autopsy studies, no specific pathological finding associated with COVID-19 was found in more than half of the patients. Most of the cardiac histopathological findings could be explained by the underlying cardiac disease of the patients, such as coronary artery disease and heart failure [23, 24]. In the other half, mild interstitial mononuclear cellular infiltration, lymphocytic myocarditis, and myocardial edema were reported [19, 25, 26]. One remarkable aspect was that pericardial effusion was rare. Postmortem biopsy studies have reported nonspecific histopathological findings that are mostly compatible with the underlying cardiac disease of the patients [12, 14]. In contrast to autopsy and postmortem biopsy studies, pericardial effusion was frequent in our series. Four out of the six patients (66.6%) had varying degrees of pericardial effusion. However, no specific COVID-19-associated cardiac histopathologic findings were observed in any of the examined specimens. The observed changes were in accordance with the underlying cardiac diseases of the decedents.
Autopsy studies performed on patients who died of COVID-19 showed mild steatosis, patchy hepatic necrosis, and mild sinusoidal dilation in the vast majority of patients [23, 27]. In their postmortem series, Beigmohammadi et al. [12] reported sinusoidal dilation, microvesicular and macrovesicular steatosis, and mild portal inflammation. Tian et al. [14] reported sinusoidal dilation and patchy hepatocyte necrosis. Our results showed patchy hepatocyte necrosis, mononuclear and/or neutrophil infiltration, and macrovesicular steatosis in the liver biopsy samples.
Autopsy studies have shown lymphocyte reduction and focal necrosis, infarction, and hemorrhage in the spleen [19, 28]. None of the available postmortem studies have obtained splenic biopsies [12, 14]. In our series, splenic tissue was available in three patients. Splenic tissue was not present in the four biopsy samples. All patients had focal necrosis and neutrophil infiltration in the red pulp.
Histopathologic changes in the liver, lung, heart, and spleen in the postmortem biopsy samples were generally similar to those obtained through autopsy. The tissue acquisition rates were quite high: 100% for liver, 88.8% for lung, and 88.8% for heart. Thus, we believe that postmortem biopsies can provide comparably good histopathological results compared with autopsy studies. Furthermore, in postmortem studies, pulmonary histopathologic findings were concordant with radiological images.
This study has several limitations. First, although our sample was not smaller than that of other postmortem biopsy studies, examining more cases can yield more generalizable results. Second, we did not use immunohistochemical staining. Nevertheless, H&E staining was largely sufficient for the purposes of the study. Third, we did not examine the presence of SARS-CoV-2 in the biopsied tissues. These data, along with the ACE2 expression level, can shed more light on the pathophysiological changes seen in COVID-19. Lastly, as organ involvement could be patchy and as there was no chance of seeing the organs macroscopically, as in autopsy, postmortem biopsy studies have inherently more limited tissue availability than autopsy studies. This might have affected our results to some extent.
In conclusion, this study performed a postmortem biopsy series on COVID-19 decedents by sampling lung, heart, liver, and spleen tissues. This work is the first postmortem biopsy study to examine splenic core needle biopsies. An adequate amount of tissue for histopathologic examination was obtained in most biopsies. When adequate tissue was sampled, the results were generally consistent with previous autopsy studies. The pulmonary radiological findings were correlated with the lung histopathologic examination results. A study design comparing postmortem biopsies with autopsy findings in the same patients can certainly determine the role of postmortem biopsies in COVID-19 decedents.
Acknowledgments: N/A
Conflict of Interest: The authors declare no conflicts of interest.
Financial Disclosure: The authors declare that this study did not receive financial support.