key: cord-315321-kax4gqhc
authors: Laplana, M.; Yuguero, O.; Fibla, J.
title: Lack of protective effect of chloroquine derivatives on COVID-19 disease in a Spanish sample of chronically treated patients
date: 2020-09-09
journal: nan
DOI: 10.1101/2020.09.03.20158121
sha: 
doc_id: 315321
cord_uid: kax4gqhc

Background: The search for a SARS-CoV-2 treatment has emerged as a worldwide priority. We evaluated the role of chloroquine and its derivatives in COVID-19 in Spanish individuals. Methods: We performed a survey addressed to patients regularly taking chloroquine and its derivatives for the control of their autoimmune diseases. The survey was distributed with special attention to Spanish patient associations centred on autoimmune diseases and rheumatology and to the general population. A sample of untreated subjects was matched to the treated group according to sex, age range and incidence region. COVID-19 disease prevalence was compared between treated and untreated-matched control sample. Results: A total of 319 surveys of patients regularly taking chloroquine and its derivatives were recovered for further analysis. The prevalence of declared COVID-19 status in the treated group was 5.3% and the mean prevalence among the untreated-matched groups was 3.4%. A community exposition to COVID-19 was associated with a greater prevalence of COVID-19 in both, treated (17.0% vs. 3.2%; p-value<0.001) and untreated groups (13.4% vs. 1.1%; p-value=0.027). Conclusion: We did not find differences of reported COVID-19 cases between treated and untreated groups, indicating a lack of protection by regular administration of chloroquine and its derivative drugs on COVID-19 infection. Of relevance, data indicates that patients that regularly take chloroquine derivatives are exposed to SARS-CoV-2 infection and must take the same protection measures as the general population.

The search for an effective treatment against SARS-CoV-2 infection has emerged as a worldwide 42 priority. A promising strategy to fight the virus causing the newest pandemic is the identification of 43 already available drugs active against other diseases that can be effective also against this new 44 SARS-CoV-2 infection [1] . 45 We present here an evaluation of the potential protective effect of chloroquine and its derived drugs 46 on the prevalence of COVID-19 in patients undergoing an active treatment with these drugs. 47

Chloroquine and its derived drugs such as hydroxychloroquine have shown in vitro and in vivo 48 effectiveness against viral infections such as SARS [2], influenza A H5N1, and Zika. Chloroquine is 49 believed to interfere with both the entry and exit of viruses from cells hosts, as well as in the 50 manifestation of acute respiratory syndrome. The virus enters the cells through binding to the 51 angiotensin-converting enzyme 2 (ACE2). Chloroquine can reduce ACE2 glycosylation, thereby 52 preventing viruses from effectively binding to cells [3] . On the other hand, chloroquine accumulates 53 in lysosomes, increasing the endosome pH levels which interferes with the viral particle release 54 process [4] . In addition, chloroquine could block the production of proinflammatory cytokines, thus 55 preventing the pathway that subsequently leads to acute respiratory syndrome [ Chloroquine and its derivatives such as chloroquine phosphate or hydroxychloroquine are commonly 61 used in the treatment of autoimmune diseases. Not without serious side effects, the use of these drugs 62 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint 4 under medical prescription is widely spread. It has been proposed hydroxychloroquine as a 63

prophylaxis treatment against SARS-CoV-2 infection for exposed caregivers [9] . Our hypothesis is 64 that if chloroquine treatment is effective against SARS-CoV-2 infection, those patients following an 65 active chloroquine or derivative drug treatment would be protected against the infection or against 66 COVID-19 adverse effects. Thus, our study aims to test this hypothesis by evaluating the incidence 67 of COVID-19 disease in the population according to chloroquine treatment subgroups through a 68 survey. 69

A survey was designed to be conducted electronically via smartphone or personal computer, and 72 therefore it was also designed to ensure accessibility and simplicity to facilitate its completion. 73

Information about the project and a link to the URL of the survey were disseminated in the press and 74 via social media and email, with special attention to Spanish patient associations centered on 75 autoimmune diseases and rheumatology. The survey included demographic questions about gender, 76 age range, and province of residence, as well as questions pertaining to health-status outcomes such 77 as treatment, COVID-19 diagnosis and symptoms due to COVID-19 infection. In addition, questions 78 about infection diagnosis and symptoms in close relatives and friends were also included. Because 79 the first cases of COVID-19 were reported in Spain in March 2020, our survey collects cases that 80 occurred between March and May 2020. 81

Individuals of any age above legal age (18 years old) with residence in Spain were eligible for 82 inclusion in the study. Individuals undergoing a stable chloroquine or derived drug treatment before 83 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted September 9, 2020. Data on symptoms related to COVID-19 infection were collected and used to assign individuals as 99 suspected COVID-19 cases when reporting loss of taste or smell and/or three or more COVID-19 100 associated symptoms [11] . 101 A sample of untreated subjects was matched to the treated group according to sex, age range and 102 incidence region with e1071 R package. The matching process was repeated using a bootstrap 103 strategy and re-sampling of the untreated-matched dataset was repeated 1,000 times to obtain the 104 distribution and mean values of the descriptive statistics such as age range, gender, incidence region 105 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint 6 and declared COVID-19 prevalence. Comparisons between groups were performed by Fisher exact 106 test. P-value<0.05 was considered statistically significant. 107

Overall, 2295 individuals completed the surveys between May and June 2020. From these, data 109 collection was complete, with all key data for the study, for 2161 individuals (94.2%). Of these 110 completed surveys, 11 entries did not meet eligibility criteria and were excluded from the analysis. 111

The final number of completed surveys used for the study was 2150. Among them, 319 (14.8%) were 112 from patients following an active chloroquine or derived drug treatment and have been included in 113 the treatment group, and 1831 (85.2%) have been included in the untreated group and serves as the 114 source to obtain the untreated-matched subgroups. We note that 94% of the treatment group 115 individuals were following a hydroxychloroquine treatment. 116

The main descriptive characteristics of both treated (n=319) and untreated-matched (n=319) 117 subgroups, arranged according to their declared COVID-19 status, are reported in Table 1 . 118

Distribution of declared COVID-19 status did not differ significantly within age group, gender and 119 incidence region (Table 1 ). In contrast, having a community exposition (defined as those individuals 120 declaring a COVID-19 positive case in a close family member or flatmate) was associated with a 121 greater prevalence of COVID-19 disease when compared with non-exposed individuals in both 122 treated (17.0%, 95%CI 12.9%-21.1% vs. 3.2%, 95%CI 1.3%-5.2%; p-value<0.001) and untreated 123 subgroups (13.4%, 95%CI 9.6%-17.1% vs. 1.1%, 95%CI 0%-2.2%; p-value=0.027) ( Table 1 and Fig  124   1 ). 125 126 127 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint 7 128

The prevalence of declared COVID-19 status in the treated group was 5.3% (95%CI 2.9-7.8) and the 129 mean prevalence among the untreated-matched groups was 3.4% (95%CI 1.4-5.4) . Furthermore, the 130 prevalence of suspected COVID-19 patients in treated subjects was of 18.8% (95%CI 14.5-23.1) and 131 the mean prevalence among the untreated-matched groups was 15.7% (95%CI 11.7-19.7) . These 132 figures are nearly similar to those recently found in the study of the seroprevalence of IgG antibodies 133 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint This research did not receive any specific grant from funding agencies in the public, commercial, or 177 not-for-profit sectors. 178 179 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint

COVID-19: Wait for a 181 novel drug or act with the age old drug -Do we have a choice?

In vitro inhibition of severe acute 184 respiratory syndrome coronavirus by chloroquine

New insights on the antiviral effects of 187 chloroquine against coronavirus: what to expect for COVID-19?

Targeting endosomal acidification by chloroquine analogs as a promising 190 strategy for the treatment of emerging viral diseases

Effects of chloroquine on viral 193 infections: an old drug against today's diseases?

Breakthrough: Chloroquine phosphate has shown apparent 196 efficacy in treatment of COVID-19 associated pneumonia in clinical studies

against SARS-CoV-2 in the Spanish population showing an estimated prevalence of 5% (95%CI 134 4.7%-5.4%), and a prevalence of suspected COVID-19 cases of nearly 20%. [11] 135 Discussion 136 Our results show no differences in COVID-19 prevalence among untreated and chronically treated 137 individuals with chloroquine or derivative drugs. Independently of the exposure, both groups showed 138 the same prevalence of COVID-19 disease or suspected COVID-19 disease according to symptoms. 139 We must note that we found a clear association between the COVID-19 disease prevalence and 140 exposure to a close family member or flatmate positive for COVID-19 in both, treated and untreated 141 subjects, that points to a lack of any protective effect on SARS-CoV-2 infection attributable to 142 chronic treatment with chloroquine or derivative drugs. 143We should mention some limitations of our study such as a limited power to detect small changes in 144 prevalence between treatment groups. However, the design of this study addressed the need to collect 145 and analyse data within a particularly short period of time due to the rapid onset and progression of 146 the pandemic, as well as the urgency of identifying and evaluating rapidly possible therapies, thus 147 partially compensating the reduced sample size. The prevalence of COVID-19 found in our study is 148 similar to the seroprevalence of IgG antibodies against SARS-CoV-2 in the Spanish population [11], 149 which is higher than the reported COVID-19 prevalence in the general population based on RNA's 150 virus detection [10] . This difference could be attributed to self-reported disease and the diagnosis of 151 COVID-19 by medical practitioners, which in many cases does not involve results of diagnostic tests 152 due to the lack of such tests. Finally, we could not eliminate completely the possibility of some bias 153 due to the intrinsic condition of the individuals within the treatment group that are undergoing 154 chloroquine or derivative drug treatment due to other diseases that alter their health status and may 155 have different comorbidities. However, our results are in line with a recent study conducting a 156 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint 9 randomized trial that reported no effect of hydroxychloroquine when used as a post exposure 157prophylaxis for COVID-19 [12] . 158All these data together point towards a lack of a protective effect of chloroquine or derivative drugs 159 as a prophylaxis for COVID-19, including prophylactic treatment before and after exposure. Of 160 relevance, data indicates that people that regularly take chloroquine derivatives are exposed to 

The authors declare that the research was conducted in the absence of any commercial or financial 174 relationships that could be construed as a potential conflict of interest. 175

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted September 9, 2020. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)The copyright holder for this preprint this version posted September 9, 2020. . https://doi.org/10.1101/2020.09.03.20158121 doi: medRxiv preprint