Correlation Study of Chest CT Features of Severe/Critical type COVID-19 with Early Renal Damage and Clinical Prognosis


 Background: Among patients with confirmed severe/critical type COVID-19, we found that although the seurm creatinine (Cr) value is in normal range, patients might have occured early renal damage. For severe/critical type COVID-19 patients, whether some chest CT features can be used to predict the early renal damage or clinical prognosis.Methods: 162 patients with severe/critical type COVID-19 were reviewed retrospectively in 13 medical centers from China. According to the level of eGFR, 162 patients were divided into three groups, group A (eGFR < 60 ml/min/1.73m2), group B (60 ml/min/1.73m2 ≤ eGFR < 90 ml/min/1.73m2 group) and group C (eGFR ≥ 90 ml/min/1.73m2). All patients’ baseline clinical characteristics, laboratory data, CT features and clinical outcomes were collected and compared. The eGFR and CT features was assessed using univariate and multivariate Cox regression.Results: Baseline clinical characteristics showed that there were significant differences in age, hypertension, cough and fatigue among groups A, B and C. Laboratory data analysis revealed significant differences between the three groups of leukocyte count, platelet count, C-reactive protein, aspartate aminotransferase, creatine kinase. Chest CT features analysis indicated that crazy-paving pattern has significant statistical difference in groups A and B compared with group C. The eGFR of patients with crazy-paving pattern was significant lower than those without crazy-paving pattern (76.73 ± 30.50 vs. 101.69 ± 18.24 ml/min/1.73m2, p < 0.001), and eGFR (OR = 0.962, 95% CI = 0.940-0.985) was the independent risk factor of crazy-paving pattern. The eGFR (HR = 0.549, 95% CI = 0.331-0.909, p = 0.020) and crazy-paving pattern (HR = 2.996, 95% CI = 1.010-8.714, p = 0.048) were independent risk factors of mortality.Conclusions: In patients with severe/critical type COVID-19, the presence of crazy-paving pattern on chest CT are more likely occured the decline of eGFR and poor clinical prognosis. The crazy-paving pattern appeared could be used as an early warning indicator of renal damage and to guide clinicians to use drugs reasonably.

According to the diagnosis and treatment program of COVID-19 (Trial Seventh Edition) issued by the National Health Commission of the People's Republic of China, the clinical classi cation of COVID-19 include mild, moderate, severe, and critical types [5]. Patients with severe type need meet any of the followings: (I) severe respiratory distress (respiratory rate (RR) ≥ 30 breaths/min); (II) SpO2 < 93% at rest; (III) PaO2/FiO2 ≤ 300 mmHg; and additional supple patients' pulmonary imaging that the lesions progressed more than 50% within 24 ~ 48 hours should be managed as severe type. Critical type, one of the following occurred: (I) respiratory failure requiring mechanical assistance; (II) shock; and (III) Complicated with extra pulmonary organ failure, requiring intensive care unit (ICU) care. Among patients with con rmed severe/critical type COVID-19, we found that although the seurm creatinine (Cr) value is in normal range, patients might have occured early renal damage (namely 60 ml/min/1.73 m 2 ≤ estimated glomerular ltration rate (eGFR) < 90 ml/min/1.73 m 2 ). According to the National Kidney Foundation (NKF) Kidney Disease Outcome Quality Initiative (K/DOQI) proposed that eGFR could be used to detect early renal damage [6].
Relevant research proved that severe/critical type COVID-19 have some chest CT features [7,8]. For severe/critical type COVID-19 patients with an early decline in eGFR, while Cr value is in the normal range, whether some chest CT features appear can be used to indicate the possibility of kidney injury, so as to guide clinicians to further calculate eGFR and avoid the selection of COVID-19 drugs that aggravate kidney damage. In the present study, we explored the correlation between chest CT features and early renal damage, and the relationship between chest CT features and clinical prognosis.

Patient population
In this study, records for 162 patients (105 males and 57 females, median age 55.66 ± 14.75 years, range from 21 to 91 years, with severe/critical type COVID-19 patients were reviewed retrospectively for the period from 15 January 2020 to 20 February 2020 in 13 medical centers from China. According to the level of eGFR (6), 162 patients were divided into three groups, namely eGFR < 60 ml/min/1.73m 2 group (Group A), 60 ml/min/1.73m 2 ≤ eGFR < 90 ml/min/1.73m 2 group (Group B) and eGFR ≥ 90 ml/min/1.73m 2 group (Group C). All institutional review boards approved this study and waived written informed consents.
All patients' medical history, laboratorial data and CT images were collected and reviewed by two radiologists with 15 and 10 years experience. The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation was used to calculate the eGFR value [9]. All baseline data were collected on the rst day in-hospital. The baseline clinical data include: age, sex, contact history (travel or residence history in Wuhan and the local community with con rmed patient), respiratory rate (RR), fever, cough, myalgia, fatigue, headache, nausea, diarrhoea, abdominal pain, dyspnea, comorbidities (cardiovascular disease, diabetes, hypertension, chronic obstructive pulmonary disease (COPD), chronic liver disease, chronic kidney disease and malignancy). The baseline laboratorial data include: leukocyte, neutrophil, lymphocyte, haemoglobin, platelet, prothrombin time, activated partial thromboplastin time, creatinine (Cr), eGFR, C-reactive protein (CRP), albumin (ALB), alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatine kinase (CK) and lactate dehydrogenase (LDH). The blood gas analyses include: SpO 2 , PaO 2 and FiO 2 . Chest CT changes: CT images rapid progression (> 50%) within 24 ~ 48 hours.
The exclusion criteria were as follows: (i) pregnant women or children; (ii) merely underwent chest radiography; (iii) larger CT artifacts on image.

CT images acquistion
All patients were supine position, held their breath, and scanned from the apex to the bottom of the lung. A Siemens Emotion 16 scanner CT (Siemens Healthineers; Erlangen, Germany) was applied to scan 18 patients from Yichang or Wuhan, China, using 5 mm slice thickness. A GE Discovery CT750 HD (GE Healthcare Milwaukee, Wis, USA) was adopted to scan 60 patients from Wenzhou, Xiaogan or Haikou, China, using 5 mm slice thickness. A Siemens second generation 64-slice dual-source CT scanner (SOMATOM De nition Flash, Siemens Healthcare, Erlangen, Germany) was used to scan 40 patients from Urumqi, Huangshi or Wuhan, China, using 5 mm slice thickness. A Philips Ingenuity core 128 spiral CT scanner (Philips Medical Systems, Best, the Netherlands) was used to scan 17 patients from Xiangyang or Xuzhou, China, using 1.5 mm slice thickness. A Siemens Emotion 16 VC20B 16-slice spiral CT scanner (Siemens Healthcare GmbH, Erlangen, Germany) was used to scan 27 patients from Huanggang or Jingzhou, China, using 1.5 mm slice thickness. All scans were underwent without contrast agent.

Statistical Analysis
Regarding measurement data: (I) those with a normal distribution are expressed as the mean ± SD; and (II) those with a non-normal distribution are expressed as the median (interquartile range) [M (IQR)]. Qualitative data were expressed as the number of cases and the percentage [n (%)]. between groups were analyzed using Student's t-test. Qualitative data were analyzed using the chi-square (χ 2 ) or Fisher's exact test. Intergroup comparisons were determined with Bonferroni correction, p < 0.05/3 = 0.0167 was considered statistically signi cant. Logistic regression analysis was used to estimate the signi cant variables (p < 0.05). Receiver operating characteristic (ROC) curve analysis was performed to determine the cut-off value. Patient's outcome was assessed using Kaplan-Meier survival analysis, and the in uence of eGFR and CT features on patient's outcomes was calculated using Cox proportional hazards modelMultivariate Cox analysis was used to determine the independent predictors of prognosis. The reported p values were two-sided, and a p < 0.05 was considered statistically signi cant. All the analyses were performed using SPSS software (version 25.0).

Results
We collected 1,177 patients con rmed with COVID-19 from 13 medical centers, exclude the mild type (n = 38) and moderate type (n = 977) of COVID-19 patients, and only retain the severe/critical type (n = 162) of COVID-19 patients, including group A (n = 26), group B (n = 37) and group C (n = 99). 162 patients who  Table 1 showed that all patients' baseline clinical characteristics, there were signi cant differences in age, hypertension, cough and fatigue among the groups (all p < 0.05). After intergroup comparisons, the results showed that the age was older in group A than group C (63.19 ± 17.04 vs. 53.12 ± 12.89, p < 0.0167), the incidence of hypertension was higher in group A than group C (65% vs. 33%, p < 0.0167), and the clinical manifestation of fatigue was more in group C than group B (47% vs. 14%, p < 0.0167).   Chest CT showed abnormalities in all the 162 patients at baseline, 158 (97.5%) patients had multiple lesions, 150 (92.6%) patients had irregular shape of lesion, 94 (58%) patients had crazy-paving pattern, 132 (81%) patients had interstitial changes and 10 (6%) patients had pleural effusion (Fig. 1, 2). No signi cant differences in lesion-involved lung segment numbers among the three groups, and the average number of involved lung segments in each group was > 15 (Fig. 3). Compared with the group C, group A or B were more likely to appear crazy-paving pattern ( Table 4 revealed that the risk factors related to crazy-paving pattern identi ed by logistic regression results. Univariate logistic regression analysis indicated that eGFR, platelet count, LDH were risk factors of crazy-paving pattern (all p < 0.05). The factors with p < 0.10 in univariate logistic regression analysis were selected to multivariate logistic regression analysis, which indicated that eGFR (OR = 0.962, 95% CI = 0.940-0.985, p = 0.001) was independent risk factor of crazy-paving pattern ( Table 4). The cut-off level of eGFR was determined as 85.74 ml/min/1.73 m 2 based on ROC curve analysis (Area Under Curve (AUC) = 0.763, 95% CI 0.689-0.838) (Fig. 4).  Table 5 demonstrated that the clinical outcomes of groups A, B and C. In the incidence of mortality, enter ICU and adopt mechanical ventilation, group A or group B were signi cantly higher than group C (all p < 0.0167). Although there were no statistical difference between groups A and B in the incidence of mortality, group A seemed to have a higher mortality trend than group B. We believe that further expansion of the sample size will be statistically signi cant. Furthermore, univariate COX regression analysis indicated that age, eGFR, lymphocyte count and crazy-paving pattern were risk factors of mortality (all p < 0.05). The factors with p < 0.10 in univariate COX regression analysis were selected to multivariate COX regression analysis, which indicated that eGFR (HR = 0.549, 95% CI = 0.331-0.909, p = 0.020) and crazy-paving pattern (HR = 2.996, 95% CI = 1.010-8.714, p = 0.048) were independent risk factors of mortality (Table 6, Fig. 5).

Discussion
Obtained from the latest data of World Health Organization (WHO), the con rmed COVID-19 cases has achieved 23788899 (until 24:00 of 26 August 2020) in the world already greatly exceeded the overall reported cases of SARS-CoV in 2003 (8422). The widespread spread of COVID-19 has seriously affected global public health. Relevant study reports COVID-19 could lead to kidney damage and recommend close monitoring the kidney functions [10,11]. In the present study, we found that group A and group B were signi cantly different from group C in crazy-paving pattern and mortality. This means that in patients with severe/critical type COVID-19, when eGFR declined to < 90 ml/min/1.73 m 2 , we should pay attention to the appearance of crazy-paving pattern on chest CT. When crazy-paving pattern appears, it indicates that patients will have more poorer clinical prognosis (include mortality, enter ICU and adopt mechanical ventilation). And we found that the lesion-involved lung segment numbers in severe/critical type COVID-19 were more than 15. At the same time, it also reminds that for severe/critical type COVID-19, when Cr value is at the upper limited in normal range, the crazy-paving pattern appear on the chest CT, we need to calculate eGFR value. According to the level of eGFR to detect early kidney damage, so as to guide clinicians to avoid using anti-COVID-19 drugs that affect kidney function.
In our study, we found 32% patients with severe/critical type COVID-19 occured eGFR < 90 mL/min/1.73 m 2 , and these patients are more likely to appear crazy-paving pattern on chest CT.
Moreover, by multivariate logistic regression analysis results, we proved that only eGFR was independent risk factor of crazy-paving pattern in severe/critical type COVID-19. And the presence of crazy-paving pattern means that patients with severe/critical type COVID-19 are more likely associated with eGFR declined, especially in the patients with normal Cr value but occured the decline of eGFR. Moreover, by multivariate COX regression analysis, we found that eGFR and crazy-paving pattern were independent risk factors of mortality. In patients with severe/critical type COVID-19 were more likely mortality when they have decreased eGFR and occured crazy-paving pattern.
At present, the renal function evaluation of patients with COVID-19 usually adopt serum Cr test, and when the Cr value > 110 µmol/L, patients could be considered as renal insu ciency or renal failure [12,13].
According to the CKD clinical practice guidelines, eGFR in 60 ~ 89 mL/min/1.73 m 2 was considered to have mild kidney damage [6]. Therefore, the eGFR conversion of Cr can early detect the kidney damage, especially for the severe/critical type COVID- 19  kidney damage, such as chloroquine phosphate, ribavirin and so on. At the same time, in the guidelines regarding the recommendation to use the Chinese traditional medicine "Qingfei Paidu Decoction" in the treatment of COVID-19 [5]. However, the "Asarum" component in the decoction has been clearly classi ed as aristolochic acid in Chinese Pharmacopoeia [14], which is induced nephrotoxicity, so the dosage can be removed or reduced. Therefore, crazy-paving pattern could be used as an effective early warning indicator to guide medication.
Since 1972, Jelliffe rst proposed that eGFR can be used to evaluate renal function, the index has been used up to now [15]. Based on Cr value to calculate eGFR is widely used to evaluate renal dysfunction in the early stage. At present, renal function can be categorized into ve stages based on the level of eGFR which may induce kidney injury and eGFR decline; In addition, the eGFR decline may also be secondary to in ammation, sepsis, shock or insu cient blood volume in the course of severe type COVID-19 [16,17].
The crazy-paving pattern can be de ned as diffuse or scattered ground-glass attenuation superimposed on a network of interlobular septal thickeding and intralobular lines [18]. In 1958, Rosen SH et al. rst described crazy-paving appearance and proved that it can appear in pulmonary alveolar proteinosis (PAP) [19]. After that, crazy-paving pattern was also con rmed to be present in pneumocystis jirovecii pneumonia (PCP), cryptogenic organizing pneumonia (COP), sarcoidosis, bronchioloalveolar carcinoma, adult respiratory distress syndrome (ARDS), etc [20][21][22]. Frazier et al. found that crazy-paving pattern was likely associated with an interstitial in ammatory cellular in ltration or brosis [23]. Johkoh et al.
proposed that crazy-paving pattern represents a slight increase in the severity of the pathologic process at the borders of unit structures [24]. In recent COVID-19 studies, we found that crazy-paving pattern can also appear in COVID-19 pneumonia. Li  Furthermore, through chest CT features analysis of groups A, B and C, we found lung segment involved numbers of three groups were all above 15, which further clari es the accompanying conditions when the crazy-paving pattern appears.
There are several limitations to our investigation. First, the sample size of this study is relatively small, and the conclusions of this research need to be further studied in a larger data set. Second, due to this study adopted patients' baseline laboratory results, there is still a lack of timeline about eGFR results for patients after onset of illness. Finally, it is uncertain whether the eGFR decline of patients with severe/critical COVID-19 is caused by CKD or AKI. These problems will be further demonstrated in future study.

Conclusion
In patients with severe/critical type COVID-19, eGFR declined has some correlation with chest CT feature (crazy-paving pattern), the presence of crazy-paving pattern are more likely occured the decline of eGFR and poor clinical prognosis. The crazy-paving pattern could be used as an early warning indicator of renal damage in severe/critical type COVID-19 and helps to guide clinicians to avoid using anti-COVID-19 drugs that affect kidney function. Figure 1 Relevant CT manifestations of COVID-19 A: Chest CT images showed crazy-paving pattern with interstitial change (arrows). B: Chest CT images showed crazy-paving pattern with multiple and irregular shape (arrows). C: Chest CT images showed GGO with consolidation and pleural effusion (arrows).
Page 19/21  The relationship between eGFR and crazy-paving pattern analyzed by ROC curve Figure 5 Kaplan-Meier curves for in-hospital mortality of patients with severe/critical type COVID-19 A. Subgroup by eGFR. B. Chest CT feature with or without crazy-paving pattern.