A Meta-Analysis of Rhesus Macaques (Macaca mulatta), Cynomolgus Macaques (Macaca fascicularis), African green monkeys (Chlorocebus aethiops), and Ferrets (Mustela putorius furo) as Large Animal Models for COVID-19

Abstract

Animal models are at the forefront of biomedical research for studies of viral transmission, vaccines, and pathogenesis, yet the need for an ideal large animal model for COVID-19 remains. We used a meta-analysis to evaluate published data relevant to this need. Our literature survey contained 22 studies with data relevant to the incidence of common COVID-19 symptoms in rhesus macaques (Macaca mulatta), cynomolgus macaques (Macaca fascicularis), African green monkeys (Chlorocebus aethiops), and ferrets (Mustela putorius furo). Rhesus macaques had leukocytosis on Day 1 after inoculation and pneumonia on Days 7 and 14 after inoculation in frequencies that were similar enough to humans to reject the null hypothesis of a Fisher exact test. However, the differences in overall presentation of disease were too different from that of humans to successfully identify any of these 4 species as an ideal large animal of COVID-19. The greatest limitation to the current study is a lack of standardization in experimentation and reporting. To expand our understanding of the pathology of COVID-19 and evaluate vaccine immunogenicity, we must extend the unprecedented collaboration that has arisen in the study of COVID-19 to include standardization of animal-based research in an effort to find the optimal animal model.

Human research of disease presents a number of ethical dilemmas, prompting scientists to use animal models in their research with the primary goal of enhancing the understanding of a human disease or phenomenon. Animal models have been instrumental to our understanding of pathologies, the assessment of novel vaccines, and the testing of acute therapies. Of the past 222 Nobel prizes awarded in the physiology and medicine categories since 1901, all but 36 have been a direct result of animal-based research.³¹

Insects, nematodes, fish, amphibians, and numerous mammals have enabled some of the most important advances in physiology and medicine since their introduction in disease research. Through genetic modification, surgical adaptation, xenografts, chemical induction, and infection models, these animals have been used to model human phenomena.³¹ However, although particular animal species are often chosen based on their ability to meet specific criteria in line with the research question, their size remains an important factor.^26,31

Small animals are often preferred in laboratory settings for their ease of use, shorter life cycle, easier handling and care, and short gestation.⁵ Rodents are the most commonly used animal for the study of human diseases for these very reasons, although they frequently fail to fully mimic the clinical signs and significant pathologic hallmarks of human diseases.^11,18 For this reason, some researchers use large animal models. Nonhuman primates (NHPs), in particular, have been extremely useful in reproducing the clinical signs of human diseases due to their close phylogenetic relationship to humans and resulting genetic, behavioral, and biochemical similarities.¹⁴

On March 11, 2020, the World Health Organization declared a SARS-CoV-2 pandemic. SARS-CoV-2 is a novel coronavirus causing symptoms similar to, but distinct from, those found in individuals infected with SARS-CoV, the coronavirus that caused the 2003 SARS pandemic. As of September 10, 2021, this coronavirus has infected 219 million individuals with the COVID-19 disease.¹⁰ Although vaccines have been developed and approved in record time, we still need to better understand the pathogenesis of the disease and the long-term implications of infections. To do this, and to increase our understanding of the immunogenicity of current vaccines, finding an animal that replicates the manifestation of COVID-19 in humans is imperative.

Much of the research on COVID-19 thus far has been aided by previous SARS research. In both SARS-CoV and SARS-CoV-2 studies, mice^33,45 and hamsters^19,34 were small animal models of choice. Large animals such as ferrets, cats, pigs, chickens, dogs, and nonhuman primates have also been tested for their reproducibility of COVID-19, with varying degrees of success.^27,41,49 While a perfect animal model of this viral infection is unlikely, the need remains to identify at least one large animal species as a frontrunner in reproducibility of the human clinical signs and significant pathologies of SARS-CoV-2 infection.

The need for a large animal model to study COVID-19 does not imply a replacement for murine models, but rather an adjunct. The closer phylogenetic relationship of humans to NHPs makes them excellent candidates for the study of this disease. Vaccine trials have already shown that the responses of NHPs are closer to those of humans than are those of mice.²³ This difference may be due to species differences in IgG antibody and T helper type 1 cell responses that influence virus-immune system interactions, which make small animal models problematic for studying SARS-CoV-2 infection and vaccine performance in humans.¹⁵ NHPs have potential high value as a model due to their homology to the human angiotensin‐converting enzyme‐2, which is the SARS-CoV-2 binding site.^23,28 After the outbreak, the World Health Organization (WHO) formed the WHO COVID-19 modelling ad-hoc expert grouping. The working group identified various NHP models, including rhesus macaques, cynomolgus macaques and African green monkeys, in addition to ferrets as being susceptible to SARS Co-V-2 isolates that would result in reproducible infection, with mild to moderate disease.⁵² Therefore, the present article is focused on summarizing the results of multiple studies on rhesus macaque, cynomolgus macaque, African green monkey, and ferret infection with SARS-CoV-2. To highlight the species that best replicate the human clinical and laboratory findings of COVID-19, we synthesized the results of 22 animal studies to provide a comprehensive analysis of what is known about their infections to date.

Materials and Methods

Search strategy and data extraction.

We conducted a literature search in May 2020 and again in June 2021 using PubMed,³² BIOSIS,³ Web of Science,⁴⁸ Scopus,³⁹ and EMBASE¹² databases with the aim of identifying animal experiments modeling COVID-19 since November 2019. Following PRISMA-ScR guidelines, the databases were searched using the following MeSH terms: SARS-CoV-2, COVID-19, coronavirus, coronavirus-19, animal model, monkey, rhesus, cynomolgus, macaque, African green, ferret, Chlorocebus aethiops, Macaca fascicularis, Mustela putorius furo, Macaca mulatta, nonhuman primate, clinical symptoms, CBC, disease progression, pathogenesis, and symptoms. A full breakdown of the search strategy is shown in Table 1.

Table 1.

Search strategy of each database.

Database	Search Terms	Filters
PubMed	(“COVID-19” OR “SARS-CoV-2” OR “coronavirus” OR “coronavirus-19”) AND (“monkey” OR “Nonhuman primates” OR “rhesus” OR “cynomolgus” OR “macaque” OR “ferret” OR “African green” OR “Chlorocebus aethiops” OR “Macaca fascicularis” OR “Mustela putorius furo” OR “Macaca mulatta” OR nonhuman primate) AND (“Pathogenesis” OR “symptoms” OR “clinical symptoms” OR “CBC” OR “disease progression”)	2019/11/1 to 2021/06/24; English; Medline; Other animals
Web of Science	(“COVID-19” OR “SARS-CoV-2” OR “coronavirus” OR “coronavirus-19”) AND (“monkey” OR “Nonhuman primates” OR “rhesus” OR “cynomolgus” OR “macaque” OR “ferret” OR “African green” OR “Chlorocebus aethiops” OR “Macaca fascicularis” OR “Mustela putorius furo” OR “Macaca mulatta” OR nonhuman primate) AND (“Pathogenesis” OR “symptoms” OR “clinical symptoms” OR “CBC” OR “disease progression”)	2019 to 2021; English
BIOSIS	(COVID-19 OR SARS-CoV-2 OR coronavirus OR coronavirus-19) AND (monkey OR Nonhuman primates OR rhesus OR cynomolgus OR macaque OR ferret OR African green OR Chlorocebus aethiops OR Macaca fascicularis OR Mustela putorius furo OR Macaca mulatta OR nonhuman primate) AND (Pathogenesis OR symptoms OR clinical symptoms OR CBC OR disease progression)	2019/11/01 to 2021/06/24; English; Exclude letters, meetings, books
EMBASE	(“COVID-19” OR “SARS-CoV-2” OR “coronavirus” OR “coronavirus-19”) AND (“monkey” OR “Nonhuman primates” OR “rhesus” OR “cynomolgus” OR “macaque” OR “ferret” OR “African green” OR “Chlorocebus aethiops” OR “Macaca fascicularis” OR “Mustela putorius furo” OR “Macaca mulatta” OR nonhuman primate) AND (“Pathogenesis” OR “symptoms” OR “clinical symptoms” OR “CBC” OR “disease progression”)	2019 to 2021
Scopus	(TITLE-ABS-KEY (covid-19 OR sars-cov-2 OR coronavirus OR coronavirus-19) AND TITLE-ABS-KEY (monkey OR {Nonhuman primates} OR rhesus OR cynomolgus OR macaque OR ferret OR {African green} OR {Chlorocebus aethiops} OR {Macaca fascicularis} OR {Mustela putorius furo} OR {Macaca mulatta} OR {nonhuman primate}) AND TITLE-ABS-KEY (pathogenesis OR symptoms OR {clinical symptoms} OR cbc OR {disease progression})) AND PUBYEAR > 2018 AND (LIMIT-TO (PUBYEAR, 2021) OR LIMIT-TO (PUBYEAR, 2020) OR LIMIT-TO (PUBYEAR, 2019)) AND (LIMIT-TO (DOCTYPE, “ar”)) AND (LIMIT-TO (LANGUAGE, “English”))	2019/2020/2021; Selected document type – Article; English

Figure 1 shows the PRISMA-ScR flow diagram for the selection process. The literature search yielded 755 papers, which were first filtered to remove 236 duplicates. The remaining 519 papers were filtered on the basis of 8 criteria. Reports were removed if they 1) were performed in vitro or used another species (n = 254); 2) used a nonmucosal route of inoculation (n = 1);²⁴ 3) treated or vaccinated all animals (n = 7); 4) did not inoculate the animals (n = 1); 5) focused on a pathology other than SARS-CoV-2 (n = 11); 6) could not be obtained (n = 4); 7) were a review, summary, meta-analysis of non-relevant animal models, or other commentary (n = 197); or 8) had a different scope, such as a bioinformatics approach (n = 9).

Figure 1.

PRISMA flow diagram.

Data was then extracted from the 22 remaining studies for the incidence of fever, cough, malaise, and pneumonia, and for measurements of lymphocyte, leukocyte, blood urea nitrogen (BUN), creatinine, alanine transaminase (ALT), aspartate transaminase (AST), and albumin. Subsets of data from experimental groups within the 22 papers were excluded if the animals were juvenile (n = 1), the titer of viral inoculation was less than 0.1% compared with all other articles (n = 1), the article did not investigate any symptoms of interest (n = 9), or the data were uninterpretable (n = 2). Data presented only in graphical form without the inclusion of raw values were interpolated via WebPlotDigitizer and excluded from analysis if this was not possible.

The final dataset included 22 papers which were stratified by animal model and days post-inoculation (dpi.), focusing on 1 dpi (or 2 dpi if 1 was unavailable), 7 dpi (± 1 d), and 14 dpi (± 1 d). The strata included inoculation titer (6.9 · 10⁴ – 1 · 10⁷ PFU) (Table 2) and different or combination mucosal routes, which varied so greatly that power was too low for stratification.

Table 2.

Basic characteristics of included studies.

Study	Date of publication	Article	Animal	No. of animals	Inoculation titer (PFU)
Corbett and colleagues.7	July 28, 2020	Evaluation of the mRNA-1273 Vaccine against SARS-CoV-2 in Nonhuman Primates	Rhesus macaque	8	7.6 × 105
Williamson and colleagues.50	June 9, 2020	Clinical benefit of remdesivir in rhesus macaques infected with SARS-CoV-2	Rhesus macaque	6	1.8 × 106
Munster and colleagues.29	May 12, 2020	Respiratory disease in rhesus macaques inoculated with SARS-CoV-2	Rhesus macaque	8	1.7 × 106
Hoang and colleagues.17	November 10, 2020	Baricitinib treatment resolves lower-airway macrophage inflammation and neutrophil recruitment in SARS-CoV-2-infected rhesus macaques.	Rhesus macaque	3	1.1 × 106
Shan and colleagues.40	July 7, 2020	Infection with novel coronavirus (SARS-CoV-2) causes pneumonia in Rhesus macaques.	Rhesus macaque	6	4.8 × 106
Deng and colleagues.9	September 2, 2020	Ocular conjunctival inoculation of SARS-CoV-2 can cause mild COVID-19 in rhesus macaques.	Rhesus macaque	5	6.9 × 105
Jiao and colleagues.20	December 9, 2020	The Gastrointestinal Tract Is an Alternative Route for SARS-CoV-2 Infection in a Nonhuman Primate Model. Gastroenterology.	Rhesus macaque	11	6.9 × 106
Fahlberg and colleagues.13	November 27, 2020	Cellular events of acute, resolving or progressive COVID-19 in SARS-CoV-2 infected nonhuman primates.	Rhesus macaque	2	3.61 × 106
Blair and colleagues.4	November 7, 2020	Acute Respiratory Distress in Aged, SARS-CoV-2-Infected African Green Monkeys but Not Rhesus Macaques.	Rhesus macaque	2	3.61 × 106
Jiao and colleagues.21	April 24, 2020	The olfactory route is a potential way for SARS- CoV-2 to invade the central nervous system of rhesus monkeys.	Rhesus macaque	5	1 × 107
Singh and colleagues.42	January 18, 2021	Responses to acute infection with SARS-CoV-2 in the lungs of rhesus macaques, baboons and marmosets	Rhesus macaque	8	1.05 × 106
Sokol and colleagues.43	March 8, 2021	SARS-CoV-2 Infection in Nonhuman Primates Alters the Composition and Functional Activity of the Gut Microbiota	Rhesus macaque	2	6.9 × 106
			Cynomolgus macaque	2	6.9 × 106
Salguero and colleagues.37	February 24, 2020	Comparison of rhesus and cynomolgus macaques as an infection model for COVID-19.	Rhesus macaque	6	5 × 106
			Cynomolgus macaque	6	5 × 106
Chandrashekar and colleagues.6	December 29, 2020	Neutralizing antibody-dependent and -independent immune responses against SARS-CoV-2 in cynomolgus macaques.	Cynomolgus macaque	3	1.5 × 106
Rockx and colleagues.35	April 17, 2020	Comparative pathogenesis of COVID-19, MERS, and SARS in a nonhuman primate model	Cynomolgus macaque	4	6.9 × 104
Li and colleagues.25	June 17, 2021	Cell-mimicking nanodecoys neutralize SARS-CoV-2 and mitigate lung injury in a nonhuman primate model of COVID-19	Cynomolgus macaque	3	1.1 × 105
Woolsey and colleagues.51	November 24, 2020	Establishment of an African green monkey model for COVID-19 and protection against reinfection	African Green monkey	6	4.6 × 105
Speranza and colleagues.44	January 11, 2021	Single-cell RNA sequencing reveals SARS-CoV-2 infection dynamics in lungs of African green monkeys.	African Green monkey	8	1.8 × 106
Hartman and colleagues.16	September 18, 2020	SARS-CoV-2 infection of African green monkeys results in mild respiratory disease discernible by PET/CT imaging and shedding of infectious virus from both respiratory and gastrointestinal tracts	African Green monkey	2	2.5 × 106
Cross and colleagues.8	August 18, 2020	Intranasal exposure of African green monkeys to SARS-CoV-2 results in acute phase pneumonia with shedding and lung injury still present in the early convalescence phase.	African Green monkey	6	3 × 106
Bao and colleagues.2	May 20, 2021	Sequential infection with H1N1 and SARS-CoV-2 aggravated COVID-19 pathogenesis in a mammalian model, and covaccination as an effective method of prevention of COVID-19 and influenza	Ferret	6	6.9 × 105
Ryan and colleagues.36	January 4, 2021	Dose-dependent response to infection with SARS-CoV-2 in the ferret model and evidence of protective immunity	Ferret	6	3.45 × 106

Data charting was performed independently and included, for each paper, authors, publication year, article title, animal species, number of animals assessed, inoculation titer, and symptom incidence. If an age was not given, it was assumed that the animals were sexually mature.

Statistical analysis.

Meta-analysis of incidence of clinical characteristics of interest (r) was carried out using the Stuart-Ord (inverse double arcsine square root) method in a DerSimonian-Laird (random effects) model using StatsDirect Statistical Software (V. 3.3.5; StatsDirect, England). The Incidence (r) with accompanying 95% confidence intervals (95% CI) was used to report individual and summary effect measures. GraphPad Prism (V. 9.2.0; GraphPad Software, San Diego, California USA) was used to conduct Fisher exact test of independence to evaluate whether the proportion of clinical characteristics in the animal model differed from the reported proportion of clinical characteristics in humans. Bonferroni correction was applied (⟨/n) for multiple comparisons. The null hypothesis, which was that the frequency of clinical symptoms would be the same among species, was rejected if any of the tests reached the tail probability ⟨ (0.05).

Results

The analysis was performed on 22 studies, outlined in Tables 2 and 3, that measured at least one symptom of interest in either rhesus macaques, cynomolgus macaques, African green monkeys, or ferrets. A random-effects meta-analysis determined the overall incidence of symptoms in each animal species by dpi across all studies.

Table 3.

Rhesus macaque, cynomolgus macaque, African green monkey and ferret study information.

		Day 1		Day 7		Day 14
Animal model	Symptoms	No. studies	No. animals	No. studies	No. animals	No. studies	No. animals
Rhesus macaque	Fever	7	41	8	34	5	254
	Malaise	7	41	5	22	4	17
	Pneumonia	7	40	6	26	6	26
	Lymphopenia	7	30	—	—	—	—
	Leukocytosis	6	29	—	—	—	—
	High ALT	5	23	2	10	2	10
	High AST	2	10	2	10	2	10
	Low Albumin	2	10	2	6	2	6
	High BUN	—	—	2	6	—	—
	High creatinine	2	10	2	6	—	—
Cynomolgus macaque	Fever	4	16	4	16	3	13
	Malaise	3	20	—	—	—	—
African green monkey	Fever	2	12	—	—	—	—
	Malaise	3	20	2	10	—	—
Ferret	Fever	3	18	2	6	2	4

Humans presented with fever in 80% of cases.⁵³ At 1 dpi, the incidences of fever were 10% in rhesus macaques, 5% in cynomolgus macaques, 4% in African green monkeys, and 4% in ferrets (Table 4). At 7 dpi, the incidences of fever were 5% in rhesus macaques, 5% in cynomolgus macaques, and 7% in ferrets (Table 4). At 14 dpi, the incidences of fever were 4% in rhesus macaques, 5% in cynomolgus macaques, and 9% in ferrets (Table 4).

Table 4.

Incidence of clinical symptoms in rhesus macaques, cynomolgus macaques, African green monkeys and ferrets via random-effects meta-analysis.

		Day 1		Day 7		Day 14
Animal model	Symptoms	I 2	R (95% CI)	I 2	R (95% CI)	I 2	R (95% CI)
Rhesus macaque	Fever	8.5%	0.105 (0.030, 0.216)	0%	0.048 (0.004, 0.138)	0%	0.041 (0.000, 0.145)
	Malaise	84.9%	0.209 (0.008, 0.575)	76.4%	0.263 (0.016, 0.661)	0%	0.072 (0.004, 0.209)
	Pneumonia	67.2%	0.542 (0.248, 0.821)	49.3%	0.779 (0.537, 0.948)	49.3%	0.779 (0.537, 0.948)
	Lymphopenia	0%	0.263(0.128, 0.426)	b		b
	Leukocytosis	15.1%	0.132 (0.024, 0.308)	b		a
	High ALT	0%	0.960 (0.773, 0.991)	0%	0.960 (0.773, 0.991)	65.8%	0.833 (0.280, 0.971)
	High AST	62.9%	0.379 (0.009, 0.888)	5.5%	0.766 (0.470, 0.963)	74.3%	0.427 (0.001, 0.972)
	Low albumin	a		0%	0.634 (0.274, 0.924)	67.2%	0.600 (0.045, 0.9998)
	High BUN	b		b		a
	High creatinine	b		b		a
Cynomolgus macaque	Fever	0%	0.052 (0.000, 0.198)	0%	0.052 (0.000, 0.198)	0%	0.048 (0.001, 0.213)
	Malaise	57.6%	0.128 (0.002, 0.406)	a		a
African green monkey	Fever	0%	0.037 (0.006, 0.202)	a		a
	Malaise	57.6%	0.13 (0.0, 0.41)	35%	0.13 (0.0, 045)	a
Ferret	Fever	0%	0.037 (0.001, 0.163)	0%	0.067 (0.012, 0.349)	0%	0.092 (0.017, 0.461)

^a Not tested.

^b Incidence of zero.

Humans presented with malaise in 46% of cases.⁵³ At 1 dpi, the incidences of malaise were 21% in rhesus macaques, 13% in cynomolgus macaques, and 13% in African green monkeys (Table 4). At 7 dpi, the incidences of malaise were 26% in rhesus macaques and 13% in African green monkeys (Table 4). At 14 dpi, the incidence of malaise was 7% in rhesus macaques (Table 4).

Humans presented with bilateral pneumonia in 76% of cases.⁵³ At 1 dpi, the incidences of pneumonia as diagnosed by radiographic lung lesions were 54% in rhesus macaques and 4% in African green monkeys (Table 4). At 7 dpi, the incidence of pneumonia was 78% in rhesus macaques (Table 4). At 14 dpi, the incidence of pneumonia was 78% in rhesus macaques (Table 4).

Humans presented with lymphopenia in 56% of cases.⁵³ At 1, 7 and 14 dpi, the incidences of lymphopenia in rhesus macaques were 26%, 0%, and 0%, respectively (Table 4).

Humans presented with leukocytosis in 13% of cases.⁵³ At 1 and 7 dpi, the incidences of leukocytosis in rhesus macaques were 13% and 0%, respectively, with no leukocytosis at 14 dpi; none of the other species showed leukocytosis after inoculations (Table 4).

Humans presented with abnormal liver function in 29% of cases.⁵³ At 1 dpi, the incidences of high ALT and AST levels in rhesus macaques were 96% and 38%, respectively (Table 4). At 7 dpi, the incidences of high ALT, high AST, and low albumin levels in rhesus macaques were 96%, 77%, and 63%, respectively (Table 4). At 14 dpi, the incidences of high ALT, high AST, and low albumin levels in rhesus macaques were 83%, 43%, and 60%, respectively (Table 4).

Humans presented with abnormal renal function in 25% of cases.⁵³ At 1 and 7 dpi, the incidences of high BUN and creatinine levels in rhesus macaques were all 0% (Table 4).

Fisher exact tests were performed to compare human and animal incidence for each symptom on each day, with the objective of accepting the null hypothesis that the incidence rates would be equal (Table 5). In most cases, the null hypothesis was rejected, as the incidence rates of the symptoms in the animal models were significantly different from those of the humans. In 3 cases, rhesus macaques modeled the human incidence of a symptom: at 1 dpi, leukocytosis rates were comparable to humans (P = 0.6026), and at 7 and 14 dpi, pneumonia incidence also rejected the null hypothesis (P = 0.0919) (Table 5).

Table 5.

Fisher exact test of rhesus macaque, cynomolgus macaque, African green monkey and ferret symptom incidence compared with humans.

		Day 1	Day 7	Day 14
Animal model	Symptoms	P value	P value	P value
Rhesus macaque	Fever	<0.0001	<0.0001	<0.0001
	Malaise	<0.0001	<0.0001	<0.0001
	Pneumonia	<0.0001	0.0919	0.0919
	Lymphopenia	<0.0001	<0.0001	<0.0001
	Leukocytosis	0.6026	<0.0001	a
	Liver failure	0.0021	<0.0001	<0.0001
	Renal failure	b	b	a
	Fever	<0.0001	<0.0001	<0.0001
	Malaise	<0.0001	a	a
	Pneumonia	<0.0001	a	a
Cynomolgus macaque	Fever	<0.0001	<0.0001	a
	Malaise	<0.0001	<0.0001	<0.0001
African green monkey	Fever	<0.0001	<0.0001	<0.0001
	Malaise	<0.0001	<0.0001	<0.0001
Ferret	Fever	<0.0001	0.0919	0.0919

^a Not tested.

^b Incidence of zero could not statistically be compared with human incidence.

Discussion

In this meta-analysis, 22 studies of SARS-CoV-2 animals were compared with the human presentation of the disease. In both clinical and symptomatic comparisons, rhesus macaques, cynomolgus macaques, African green monkeys, and ferrets all failed to fully mimic the severity of the presentation of COVID-19 in humans.

Despite the paucity of data, rhesus macaques seem a frontrunner of the four species with significantly similar (P > 0.002) presentations of leukocytosis at 1dpi and pneumonia at 7 and 14 dpi (Table 5). The ideal animal model should demonstrate fidelity to the human disease process and is a hybrid of homologous and analogous animal models. Homology refers to morphologic identity of corresponding parts with structural similarity descending from common form, thus genetic similarity. Analogous models refer to the quality of the resemblance or similarity in function or appearance.⁴⁶ At this time, none of the animals tested were an ideal model of COVID-19.

While limited, the data indicate species differences in COVID-19 disease symptomology. These studies, and those before them dealing with SARS,³⁰ found that the virus has a faster life cycle in animals. Their subdued disease presentation points to a different genetic susceptibility and reaction to the virus. Until an ideal model is found, future research will likely require a combination of both small and large animals based on the advantages and disadvantages of the individual species and the questions being asked.⁴²

One limitation of the present study is the focus on clinical characteristics, as many humans are asymptomatic when infected with SARS-CoV-2. Although the species evaluated in this study demonstrated some fidelity to the human disease, data were not available to be able to calculate the percentage of asymptomatic animals. Although the laboratory findings make this review more robust, future studies should focus on expanding the molecular variables that can be used for diagnosis so that they can be targets in scoping reviews such as this.

A further limitation is the small sample size of the current review. A greater number of studies might reveal an ideal animal model of COVID-19. However, we excluded numerous studies from this review because viral titers were too low, the inoculation route was intravenous or otherwise nonmucosal, or raw data were unavailable. Future international guidance on experimental methods and requirements for consistency in reporting might allow such studies to be included and thereby increase the power of the overall analysis. In multiple instances, a symptom was tested in only 2 studies or was not studied at all (Table 3). These gaps in knowledge weaken the power of this analysis and which prevent us from highlighting any one animal as the ideal model for COVID-19.

Standardization in reporting has been an issue for some time now, with few advances.¹ The ARRIVE (Animals in Research: Reporting In Vivo Experiments) guidelines were published in 2010 with multiple goals, one of which was to maximize the utility of the information gained from every animal experiment.²² Two years later, evidence indicated that endorsement of the guidelines was insufficient to promote complete adherence.¹ This problem is apparent in the papers excluded from the present review, as they lack key parameters of methodology and prevent both reproducibility of experiments and the use of their data for systematic review or meta-analysis.

A meta-analysis of animal model experimentation using stratification of inoculum titer was the original intent of this study, and would be the best way to determine the best animal model for COVID-19. Unfortunately, lack of standardization in reporting standards has meant that a meta-analysis of viral experiments is not possible due to variations in assays, units, and vague statistical comparisons, which are only a 3 of the inherent challenges. Even with regard to clinical symptoms, misleading interpretation of data can occur if the standards are unclear. D-dimer, for example, is a prominent prognostic indicator in many diseases, but could not be singled out with regard to COVID-19 due to a lack of standardization in its reporting.⁴⁷ If measurements are made in non-standard units, conversion should become customary.

The aim of this study was to identify an ideal large animal model for COVID-19. Based on the available and included studies, we conclude that neither rhesus macaques, cynomolgus macaques, African green monkeys, nor ferrets can be declared an ideal model of the disease. ACE2 is well documented to be the entry receptor of COVID-19.³⁸ Thus, homology in the ACE2 receptor between species could be a predictor of SARS-CoV2 fidelity. Based on ACE2 nucleotide sequence, 97.0% and 96.8% of bases are identical across the full length of the gene sequence for cynomolgus macaques and rhesus macaques, respectively.²⁸ This translates to 95.2% and 94.9% of amino acid residues that are identical across the full length of the protein sequence. In ferrets, the full length of the protein sequence is 85.9% identical to humans.²⁸ Thus, one could hypothesize that based on homology to ACE2, cynomolgus macaques would be the best model. However, this conclusion is not consistent with the clinical characteristics from the meta-analysis, which show that rhesus macaques are marginally better. This perhaps suggests that at this point in time, examining each component of a disease process in many species might be more appropriate.²⁶

The greatest obstacle to identifying the ideal model is the lack of standardization for experimentation and reporting. Based on the information already published and the current data, we can only conclude that the absence of standardization and reporting prevents effective collation and comparison of data. We therefore recommend that standard practices be adopted to facilitate the registration of data for maximization of animal use and data comparison.