Comment on Hadadi et al., Does recombinant human erythropoietin administration in critically ill COVID-19 patients have miraculous therapeutic effects? J. Med. Virol. 2020; 92:915-918.

To the Editor,

We have read with great interest the report of Hadadi et al. on the possible therapeutic effect of epoetin which they found in a critically ill patient suffering from COVID‐19. They also proposed plausible mechanisms that could have contributed to the evolution of their patient. Encouraged by the results reported by Hadadi et al. and the lack of validated therapeutic options for the treatment of SARS‐CoV‐2 infection that was becoming pandemic and affected seriously our population, some of us contacted a manufacturer of epoetin alfa (recombinant human erythropoietin) to perform an exploratory study that included a control group. The design was an open study (blind for laboratory determinations) with the addition of epoetin alfa (Hemax; Biosidus S.A.) at 4000 IU each other day by subcutaneous injection for 10 days (five doses) to standard care in comparison to only standard care. Responses included clinical evolution and variation of the viral load by quantitative reverse‐transcription polymerase chain reaction. After consultation with the regulatory medicines agency of Argentina (National Administration of Drugs, Foods and Medical Devices [ANMAT]), which did not take intervention as the trial, was not intended for registration, and approval by the Institutional Review Board of the Hospital Sirio‐Libanés de Buenos Aires, we recruited six in‐patients hospitalized with confirmed SARS‐CoV‐2 infection between June and July 2020. Exclusion criteria included hypersensitivity to epoetin, high hematocrit level, noncontrolled hypertension, and thrombotic antecedent. Though based on the report by Hadadi et al., we selected not critically ill patients since the evidence basis to include so risky patients seemed not enough at the time the study was designed. Table 1 summarizes their clinical characteristics. Two of the cases were staff (an MD and a nurse) at the hospital and probably acquired the infection while performing their duties in the care of patients.

Table 1

Effect of epoetin alfa on the virological evolution of COVID‐19 patients

Patient and treatment: standard care (S) or standard care plus epoetin (S + E)	Age (years)/sex (M/F)	Initial viral load (N1)	Final viral load (N1)	Duration of hospitalization (days)
1 (S + E)	40/M	1.18E+07	ND	20
2 (S + E)	34/F	3.61E+05	ND	14
3 (S + E)	49/F	5.14E+05	1.42E+04a	19
4 (S)	29/F	ND	ND	14
5 (S)	31/M	6.83E+08	ND	16
6 (S)	30/F	ND	ND	12

Note: No evident difference was detected between treatment with or without the addition of epoetin to conventional treatment.

Abbreviations: F, female; M, male; ND, not detected.

In the previous two assessments, viral load was undetectable.

The diagnosis was performed according to Lu et al. Viral load was measured at the Virology Laboratory of the Hospital de Niños de Buenos Aires, one of the reference centers in the city. In brief, viral RNA automated extraction was performed by QIAsymphony DSP Virus/Pathogen Kit (Qiagen) in a QIAsymphony SP Sample Preparation equipment (Qiagen) according to the manufacturer's instructions. For the calibration curve, the 2019‐nCoV_N Positive Control plasmid (Integrated DNA Technologies Inc.) was used.

Consistent with the clinical picture of most COVID‐19 patients, all cases presented evidence of inflammatory activity at hospitalization as well as increased ferritin levels, without difference between the patients who received epoetin (163.4 ± 48.0 ng/ml, as mean ± SD) and patients under standard care (193.6 ± 121.8 ng/ml). None of the cases presented anemia, with normal hemoglobin values (14.4 ± 0.95 g/dl for standard treatment and 14.6 ± 0.85 g/dl for the epoetin one). No difference was found in the neutrophil‐to‐lymphocyte ratio (p = .70 with Mann–Whitney U test), with values between 0.893 and 3.125. While hospitalized, O2 saturation varied between 92% and 98%.

None of the patients died during hospitalization and all cases recovered between 12 and 20 days. All were discharged without medication. None required mechanical ventilation nor became septic. No difference was found in the clinical parameters nor in the virological evaluation between cases treated with epoetin and those who received only standard care. As expected, due to the short period of treatment and low dose, no serious adverse events attributable to epoetin were detected.

In comparison with the case reported by Hadadi et al., we consider that the main differences are that our patients were not so critically ill as their patients and the presence of a simultaneous group of patients who only received standard care. A high proportion of patients infected with SARS‐CoV‐2 are asymptomatic or present only limited symptoms and recover spontaneously, though this characterization was not evident at the beginning of the pandemics. On the contrary, patients with specific risk factors, such as diabetes, obesity, or being elderly, mainly if requiring mechanical ventilation, show a high mortality rate, though not all cases die. The patient of Hadadi et al. presented several of those factors, and we cannot explain his clinical course, though we believe that spontaneous recovery cannot be excluded, as the authors also have speculated. Such differences in the clinical course are just becoming to be understood.

An answer to the question posed by Hadadi et al. cannot be completely provided, and further experimental and basic studies are required to test whether epoetin possesses additional, still unknown activities to improve the treatment of COVID‐19. Meanwhile, we propose that though interesting, the current information does not support treatment with epoetin to fight SARS‐CoV‐2 infection unless other factors, such as renal anemia, are present.

CONFLICT OF INTERESTS

Lucía N. Cano and Roberto A. Diez are staff at the Clinical Research Department of Biosidus S.A. The only financial support of Biosidus was the provision of Hemax® and to cover the additional costs required for the assessment of viral loads and transportation to the Laboratorio de Virología. The remaining authors declare that there are no conflict of interests.