Introduction

Back to December 2019, several cases with acute respiratory symptoms were successively reported from different hospitals in Wuhan, China [1]. Subsequently, a novel coronavirus was extracted from respiratory tract secretions and named as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [2] and the illness was officially termed as coronavirus disease 2019 [3]. Since then, it has been extensively spread to all over the world and led to more than 900 thousand death globally until September 21st, 2020 [4]. Although patients with COVID-19 seemed to show more pulmonary symptoms and signs than other systems, multiple organs (kidneys, hearts, gastrointestinal tracts, etc.) have been demonstrated involved as well [5,6,7,8]. Particularly, several researchers illustrated kidney involvement in COVID-19 patients. They noticed acute kidney injury (AKI) among COVID-19 patients was common, and AKI was proved to be an independent predictor for poor prognosis of COVID-19 patients [9, 10]. Histopathologically, researchers have verified the viral particles located in both renal glomerular and tubular by light microscopy and transmission electron microscopy [11]. Besides, Su et al. also provided a direct evidence of renal viral invasion from kidney samples of 26 deceased COVID-19 patients [12]. A review has recently summarized the potential mechanisms of kidney involvement and divided it into 3 aspects: cytokine production, organ–organ crosstalk and systemic influence [13].

So far, studies regarding kidney damage with novel coronavirus pneumonia have mainly focused on AKI abnormality due to the direct tubular infection and severity of the disease [9, 14,15,16,17]. However, the relationship between COVID-19 and CKD has not been paid much attention at the initial outbreak due to less urgent than AKI. Besides, for some patients, the situation of preexisting CKD history or COVID-19-associated kidney damage was unclear. In addition, investigations about the incidence of patients with known CKD history were extremely rare [18]. Hence, our study aimed to describe the features of COVID-19 patients with known CKD history and seek for prognosis-related factors if possible.

Methods

Diagnostic criteria of COVID-19

On the basis of the New Coronavirus Pneumonia Diagnosis and Treatment Program [19] (7th edition, in Chinese), issued by National Health Commission of People’s Republic of China on March 3rd, 2020, the diagnosis criteria of COVID-19 pneumonia were: (1) epidemiological history: exposure to confirmed patients or travel or dwell in anywhere reported confirmed cases, especially in Wuhan; (2) clinical manifestation: fever and/or other respiratory symptoms, implied signs of viral pneumonia from chest CT scan, and/or normal or reduced white blood cells (WBC) and lymphocytes; and (3) pathogenic and/or serological detection: nucleic tests of samples collected from swabs of patients’ nasopharynx are positive, and/or antibodies to SARS-CoV-2 (immunoglobulin M and immunoglobulin G) are positive.

Patients

The study consists of 20 cases with preexisting CKD prior to COVID-19 diagnosis between January 20th and March 1st, 2020 in Tongji Hospital in Wuhan. Patients with normal GFR (≥ 90 mL/min/1.73 m2) and kidney transplantation history were excluded. Diagnosis and classification of CKD and AKI was based on the 2012 clinical practice guideline formulated by the Kidney Disease: Improving Global Outcomes (KDIGO) organization [20]. According to the eGFR on admission, those patients were classified as G2, G3a, G3b, G4 and G5, respectively. All records were reviewed retrospectively, including clinical baseline data, laboratory findings and chest imaging changes for each case involved.

Chest CT evaluation

Non-enhanced chest CT scans were regularly operated using either one of the two CT scanners (uCT 780, United Imaging; Somatom Force, Siemens Healthcare), patients’ CT images were gathered under a single breath-hold status in the supine position. All CT images were reconstructed to 1.25-mm-thin slices.

The scans were conducted for each patient at different intervals. For each patient, the chest scan was evaluated for characteristics as follows: (1) presence of GGO, consolidation or interstitial abnormalities (including reticular appearance, fibrous strips and interlobular septal thickening); (2) the laterality of lesions involved (unilateral or bilateral); (3) the number of lobes affected; and (4) degree of involvement was subjectively scored from 0 to 5 for each lung lobe as no involvement, < 5%, 5–25%, 26–49%, 50–75% and >75% involvement, separately. Each lobe’s score was added up to obtain the final total scores, ranging from 0 to 25. Imaging evaluation was performed on a PACS system (Synapses, Fujifilm) by two radiologists with over 8-year experience, the final scores were determined by consensus.

Results

Clinical characteristics

Ten patients were male (50%) and ten were female (also 50%), with a median age of 68 years (range, 46–88 years). The average time from the onset of initial symptom to hospitalization was 13.05 ± 10.23 days. Only 2 patients (10%) had a history of lung diseases, while hypertension occurred in 13 cases (80%) and was the most common comorbidity. Diabetes mellitus and coronary heart disease were also present in partial cases and 1 case was previously diagnosed with kidney cancer. 9 patients (45%) were current smokers. The foremost symptoms were cough, fever and dyspnea, accounting for 95% (19/20), 85% (17/20) and 40% (8/20), individually. On admission, 6 patients were classified as G2 of CKD, 5 were as G3a, 2 were as G3b, 3 were G4 and 4 were G5, respectively.

Concerning the laboratory findings, most patients showed normal white blood cell count. Lymphocytes count decreased in 80% (16/20) of the cases and 89% (17/19) displayed an elevated d-dimer on admission. 6 patients (30%) presented a slight decrease of platelets counts while only 1 case exceeded the upper limiting value. Indications such as alanine aminotransferase, aspartate aminotransferase and total bilirubin with respect to liver function were normal overall. Serum albumin, however, revealed decreased in 60% (12/20) cases. Infection-related indications hsCRP, IL-6 and TNF-α escalated in the proportion of 83% (15/18), 94% (15/16) and 94% (15/16), separately. The median of blood urea nitrogen (BUN) was 7.25 mmol/L (range, 4.0–29.6 mmol/L) and serum creatinine (Scr) was 141.5 μmol/L (range, 74–1011 μmol/L). On admission, the majority of patients showed urine test abnormalities. Table 1 lists comprehensive information about clinical characteristics.

Table 1 Clinical characteristics and laboratory findings of 20 COVID-19 patients with known CKD history
Full size table

Image changes

Each of the 20 patients underwent chest CT examinations, ranging from 1 to 7 times. A total of 63 person-time examinations were performed. According to time (days) from the onset, the chest CT scans were classified into 4 stages: Stage 1 (5–8 days, n = 2); Stage 2 (9–13 days, n = 3); Stage 3 (14–28 days, n = 20); and Stage 4 (≥ 29 days, n = 38).

The chest CT imaging characteristics and total CT scores are shown in Table 2. All 20 patients showed bilateral involvement. GGO and consolidation were the most universal CT manifestations. GGO was observed in 1 patient (50.0%) of stage 1, 1 patient (33.3%) of stage 2, 3 patients (15.0%) of stage 3 and 11 patients (28.9%) of stage 4, while consolidation was exclusively seen in stage 3 and 4 with 6 patients (30.0%) and 3 patients (7.9%) involved respectively. GGO with consolidation was seen in 1 patient (50.0%), 2 patients (66.7%), 5 patients (25.0%) and 13 patients (34.2%) of stage 1, 2, 3 and 4 separately. The mean CT involvement score was 12.0 ± 4.3 (mean ± standard derivation) for stage 3, which was the highest among all four stages.

Table 2 Chest CT features of 20 patients with total 63 scans among 4 stages divided by the intervals from the onset of initial symptom
Full size table

From the day of admission, follow-up CT scans were performed in 17 (85.0%) patients, ranging from 4 to 86 days. 15 (75.0%) patients showed improved, 1 (0.5%) deteriorated and 1 (0.5%) stayed unchanged. Among these 15 patients who exhibited improvement, 3 patients improved within 13–15 days after admission, 2 within 16–20 days, 8 within 21–30 days and 2 within 31–35 days, respectively. 3 cases (15%) did not perform follow-up CT due to critical illness or death. The typical manifestations in chest CT images of 1 patient discharged and 1 died during hospitalization are shown in Figs. 1 and 2, respectively.

Fig. 1
figure 1

Chest CT images of a 60-year-old woman, who has been diagnosed with chronic glomerulonephritis for 1 year. a 6 days after onset: consolidation was seen in the lower lobe of the right lung, the edge of the lesion was blurred with “halo sign”. b 15 days after onset, also 5 days after admission: the consolidation in a became stiffer and denser, with sharper edge. c 29 days after onset, also 19 days after admission: consolidation was still observable, but the lesion area was reduced. d 37 days after onset, also 27 days after admission: the density of the lesion decreased, and appeared as GGO accompanied with consolidation. e 42 days after onset, also 32 days after admission: the area of the lesion was increased, but the density was lower, like sugar melted. f 48 days after onset, also 38 days after admission: the area of lesion was further increased, but the density was even lower

Full size image
Fig. 2
figure 2

Chest CT images of an 83-year-old man, who has been diagnosed with chronic renal failure for 40 years. a 11 days after onset, also 4 days after admission: diffused GGO with thickened interlobular septa. b 16 days after onset, also 9 days after admission: GGO increased significantly with patchy consolidation and interlobular septal thickening. c 77 days after onset, also 70 days after admission: the area of GGO was further increased with patchy consolidation, bilateral pleural effusion was observed. The patient died 79 days after hospitalization

Full size image

Treatments and outcomes

Drugs usage before this-time hospitalization included angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs), immunosuppressive (IS) agents, antibiotics and antivirals. Among these patients, antibiotics (75%, 15/20) and antivirals (60%, 12/20) use were quite general.

During hospitalization, various therapy strategies were disposed individually. Concisely, none of these cases accepted extracorporeal membrane oxygenation (ECMO). Only 1 case performed invasive mechanical ventilation for 7 days, while non-invasive mechanical ventilation was conducted for 2 patients, one for 1 day and the other for 4 days. 4 patients received continuous renal replacement therapy (CRRT) including 2 have accepted continuous hemodialysis before. Medications of systematic steroids and intravenous immunoglobulin (IVIG) were given to 13 (65%) and 11 (55%) patients, separately. 80% of these patients were prescribed for at least one kind of antiviral agents. The details about treatments reviewed in Table 3.

Table 3 Treatment of cases classified by the CKD stage on admission before and after hospitalization
Full size table

The clinical outcomes are revealed in Table 4. Severity of COVID-19 was defined as the occurrence of life-threatening complications or entrance to intensive care unit (ICU). As the table showed, of the 20 patients, 16 were convalescent and discharged. 12 patients were discharged with stable renal function while the other 4 had a deteriorating renal function. 4 patients died during the study period and all of them were severe on admission or developed as severe cases during hospitalization. A total of 7 patients occurred AKI with 5 cases developing at stage 1, 1 at stage 2 and 1 at stage 3. Among those 7 cases, 2 passed away, 2 had declined renal function and 3 patients’ renal function returned back to admission levels when discharged.

Table 4 Clinical outcomes among patients of various CKD stages on admission
Full size table

For these 4 departed patients, all of them were elderly and severe cases. 2 patients were at G4, 1 at G5 and 1 at G3a on admission. Half of them developed as AKI (one as stage 1 and the other as stage 2). 75% of them had the history of hypertension. Decreased lymphocytes and platelets occurred in 3 patients and all of them displayed increased d-dimer. The data about infection-associated indications (hsCRP, IL-6, TNF-α) of these 4 cases were incomplete. However, 2 patients with complete data presented a milder rise than those who were alive.

Discussion

This is a descriptive study about clinical and radiological manifestations of 20 COVID-19 patients with CKD history. The incidence of pre-existent CKD in COVID-19 patients was diverse among different studies and a systematic meta-analysis of 22 articles revealed it as 5.4% [21]. Similarly, AKI prevalence among COVID-19 patients ranged from approximately 5.0 to 50.0% [9, 22,23,24,25]. AKI occurred on CKD could largely delay the recovery of kidney function as well as augment the mortality risk, which was reviewed by He et al. [26]. It was obviously improper to assess the fatality rate in our study because of the limited samples. Officially, it presented a death rate of 5.4% in 85,297 cases in China up to September 21st, 2020 [27]. A series of researches reported gender disparity in the epidemiology of COVID-19—men were more likely to be infected than women [28,29,30]. Nevertheless, no distinct sex discrepancy was noticed in our study. The possible reason may be attributed to small samples and the higher incidence of CKD in women essentially [31]. A large scale of multicenter research revealed that the median age was 47 years among infected patients (range 35–58 years) [25], which was much younger than ours. This could be attributed to the chronic-condition population that we concerned about.

Consistent with many other studies, we detected lymphopenia in most patients. This has already been one of the alternative diagnosis markers in the update edition of our national COVID-19 diagnosis and treatment program. We also observed elevated inflammatory factors such as IL-6 and TNF-α in the majority of infected patients. One study also demonstrated that the decrease of T cell count accompanied with a rapid rise of cytokines such as IL-2, IL-4, IL-10 and TNF-α [32]. In fact, CKD patients usually remained B and T cell phagocytic dysfunction and persistent inflammatory activation [33]. Hence, COVID-19 patients with CKD may had a superimposed negative effect on immune system.

The major chest CT manifestations of CKD patients attacked by COVID-19 were GGO and consolidation, the same as other studies among general population [34,35,36]. Pan and his colleagues classified the pulmonary CT changes of COVID-19 patients into 4 stages: 0–4 days as early stage, 5–8 days as progressive stage, 9–13 days as peak stage and absorption stage (≥14 days) [36]. However, our study showed that the lung involvement peaked at 14–28 days and started to absorb at 21–30 days. The possible reason may be the occurrence of inflammation storm of these CKD patients. Besides, lower metabolism rate may be another cause due to the decrease eGFRs and the old ages. Of course, the influence of small samples to the result could not be ignored as well.

In terms of treatments, no effective measures have been explored until now to cure this infection. In spite of ACE2 as the significant portal of the coronavirus entry to organs, whether ACEI/ARBs could be a promising application for COVID-19 management is still of controversy [37]. Kaletra (lopinavir/ritonavir) had been expected to work before, nevertheless, World Health Organization (WHO) put an end to the clinical trials of this drug according to earlier experiment results [38].

Prognostic analysis from other researches illustrated that CKD raised the severe or mortality risk of infected patients [39,40,41]. In our study, 20% (4/20) and near half (45%, 9/20) of the patients ended up with death or progressed to severe cases, respectively. A single-center study showed a 11.6% mortality and 58.1% severity rate of CKD patients with viral infection [40], while another meta-analysis of 5 involved researches demonstrated a higher fatality and severity rate as 53.33% and 83.93%, separately [42]. In spite of these statistical discrepancies, CKD patients infected with COVID-19 may acquire poorer outcomes. 4 deceased patients in our study were almost at advanced ages with underlying diseases (mainly hypertension). All of them had poor renal function on admission and one of them endured persistent hemodialysis before this infection. The influence of CKD stage on COVID-19 prognosis has recently articled by Alireza and Silvia. They both pointed out higher CKD stage was not associated with the worse outcomes [40, 43]. On the contrary, a study demonstrated the level of eGFR on admission was in negative correlation with the fatality [10]. Further investigations are required to be carried out.

Several limitations were supposed to be mentioned in our study. First of all, the samples in our study came from single-center with relatively less cases, which may exert bias on the calculation of prevalence of some events (e.g., mortality rate, AKI incidence, severity proportion). Second, we classified CKD stages based on the eGFR on admission. Owing to the different durations between initial symptoms and admission among infected patients, whether COVID-19 infection affected the serum creatinine was unclear actually. Hence, the CKD stage of patients involved may be of inaccuracy. In addition, we only collected the laboratory findings on admission and some were incomplete, thus, it would be difficult to evaluate the dynamic changes of these indications for each case.

Conclusions

In conclusion, our study revealed that COVID-19 patients with pre-existent CKD inclined to progressed to critically ill cases and the mortality was higher than the general population.