Early results of COVID-19 pneumonia cases treated with Ultra-Low doses of Radiotherapy (ULTRA-COVID study)

Introduction: Since the outbreak of COVID-19 pandemic, healthcare system has focused its effort to nd a treatment to avoid the fatal outcome of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). Benets and risks of systemic treatments are unclear. Radiotherapy could play a role in reducing the inammatory response in the lungs and relieve life-threatening symptoms Methods: We designed a prospective study of Ultra-Low-Doses of Therapy with Radiation Applied to COVID-19 (ULTRA-COVID) for patients that are no candidates for invasive mechanical ventilation and show no improvement with medical therapy. (ClinicalTrials.gov Identier: NCT04394182) Results: We present the preliminary results of two patients diagnosed with COVID-19 pneumonia treated with ULTRA-COVID. Signicant clinical response is accompanied by lower radiological one, both have happened, achieving hospital discharge after 1 radiotherapy session over a period of 8 and 14 days, respectively. Conclusion: Preliminary clinical and radiological results suggest a potential benet of treating COVID-19 pneumonia with ultra-LDRT.


Introduction
The onslaught of coronavirus disease 2019 (COVID-19) has challenged the healthcare´s infrastructures worldwide. In this evolving situation with unanswered questions about proper clinical management and therapeutic approach, the healthcare system has struggled with a raise of critically ill patients (1).
Although most part of COVID-19 patients will be asymptomatic, complications such as severe pneumonia, respiratory failure, or acute respiratory distress syndrome (ARDS) can occur, leading to fatal consequences. Most of the time, these cases require of intensive care unit (ICU) admission and invasive mechanical ventilator (IMV) (2).
In these critically ill patients, the host response against the virus appears to be mediated by a 'cytokine storm or release syndrome (CRS)', leading to a macrophage-mediated in ammatory mechanism (in ammatory M1-phenotype) (3) and ARDS, in the form of a bilateral pneumonitis. In terms to prevent progression to the critical state, it has also been proposed that the CRS can be safely treated by a single course of ultra-low-dose of radiotherapy (Ultra-LDRT)< 1 Gy (2,4) which could potentially alleviate symptoms of respiratory distress quickly, helping to reduce mortality without signi cant long-term sequelae (5).
As Confucius said: "Study the past if you would de ne the future", indeed the past may hold the key. The use of a single ultra-LDRT to treat pneumonia was reported in the early 1900s in 15 studies, involving 863 patients. It showed 80-85% rates of in ammatory relief success, and a mortality decrease from 30% to 10% (6). Besides, the use of RT in non-malignant in ammatory conditions (7) and its anti-in ammatory properties such as decreasing levels of pro-in ammatory cells including cytokines and macrophages (polarizing them toward an anti-in ammatory M2-like phenotype) have been considered in our study (8)(9)(10).
Therefore, we hypothesized that Ultra-LDRT applied to COVID-19 patients (ULTRA-COVID) could play a role in reducing the lungs in ammatory response, counteracting the CRS, reducing the risk of requiring IMV and relieve life-threatening symptoms (4).
Given the need to rapidly communicate information on the global clinical effort against COVID-19 we would like to share this report that describes the eligible criteria, clinical course, treatment and evolution of our rst two COVID-19 cases treated with ultra-LDRT.

Methods
After approval by an Ethics Committee, a prospective study was designed and initiated at La Milagrosa Hospital (Madrid, Spain) to treat COVID-19 patients with Ultra-LDRT. The main purpose of the study was to analyze the e cacy of LDRT, as an anti-in ammatory intention, in patients with SARS-Cov-2 pneumonia with a poor response to medical treatment and would otherwise have no other treatment except IMV, to which they were no candidates. Given the extremely unusual situation and poor prognosis of the disease if left untreated the study has been designed without a control arm, to evaluate security and e ciency of the treatment.
We reviewed the medical records of these patients to evaluate biographical data and medical history.
Based on that, Charlson Comorbidity Index (CCI) (>or< 6 score) (11) was calculated for each patient. Previously to the treatment, diagnosis of COVID-19 was proven by polymerase-chain-reaction (PCR), as well as blood gas analysis was performed to calculate the Pa02 / Fi02 (>or< 300 mmHg). We measured the oxygen saturation status (>or< 93%) and the ventilatory support with oxygen therapy (from less to more: Nasal Cannula-NC-; Ventimask -VMK-and VMK with reservoir). Blood analysis was obtained to assess in ammatory and immunological parameters, such as lymphocytes, IL-6 , D-dimer, ferritin, LDH, C Reactive Protein (CRP) and brinogen (12,13).
All patients underwent a baseline thoracic computed tomography (CT) scan, in which we assessed the radiological involvement through the Total Severity Score (TSS) (14). This score values ranged from 0 to 20 according to the sum of the percentage of involvement of each 5 lung lobes, which were scored from 0 to 4 points. The same senior thoracic radiologist also estimated the lung involvement qualitatively (subjectively) as mild (TSS 0-5), moderate (TSS 6-15), or severe (TSS >15). A worsening of TSS during the hospital stay or score at admission > 5 was considered as inclusion criteria.
All the patients were provided and signed a written informed consent.

Treatment protocol:
Ultra-LDRT was administered for 6-MV photon beams by a Tomotherapy Hi-Art Accuray® under institutional safety procedures.
The simulation images were acquired by megavoltage CT (MVCT) in the Tomotherapy®. Immobilization was done in supine position with thorax board with arms support. Three radiopaque-marks were placed on the patient skin. The contouring was made in Pinnacle® station and dosimetry in Tomotherapy Hi-Art Planning Station®. The planning target volume (PTV) was de ned as both whole lungs extended 1cm isotropically. No dose constraints were applied to surrounding organs. Regarding the target coverage, the 90% of PTV should receive 100% of the prescription dose and the maximum hotspots dose should be <110%.
Veri cation imaging was carried out using a MVCT limited to the central third of the thorax, to correct for any error. Total single dose administered was 0.8 Gy in a 3-minutes session.

Response evaluation:
The radiological response, assessed by TSS change, was evaluated from a thoracic CT scan 7 days and 4 weeks after the treatment. Radiological improvement was de ned as mild (TSS decrease <3 points), moderate (TSS decrease 3-5 points), or high (TSS decrease >5 points) from the baseline CT.
The clinical response was evaluated by pulse-oximetry, blood gas analysis and labs at days 2, 5, 7, and at 4 weeks after Ultra-LDRT. Two months later, oxygen status and pulse-oximetry were evaluated again.
Sat02>93%, descent of oxygen therapy support, Pa02 / Fi02 > 300 mmHg and the achievement of normal range value in one or more of the in ammatory and immunological parameters, was considered as clinical improvement.
Toxicity was assessed according to the NIH Common Terminology Criteria for Adverse Events (CTCAE v5.0) scale (16).

Cases report
After establishing our protocol, 4 patients with COVID-19 pneumonia were candidates for LDRT. One refused to participate and another died before receiving the treatment. The other two participants met the study criteria and are discussed below.
Patients´ clinical characteristics are summarized in Table 1.

Patient 1
An 80-year-old-man presented to the emergency department with a 3-day history of dyspnea, cough and chest pain. He showed 70% 02-Sat and tachypnea. Pulmonary auscultation revealed crackles predominantly in bilateral lower two-thirds. During hospitalization his evolution was torpid with a 87% 02-Sat needing of increased ventilatory support (50% reservoir, 15L of ow). The baseline CT showed bilateral pneumonia and extensive bilateral ground-glass opacities corresponding to an acute in ammatory stage. (Figure 1.A)

Patient 2
A 65-years-old-woman debuted with dry cough. A week after, she reported to the emergency department with persistent cough, fever, asthenia and dysgeusia, hence she was admitted. During hospitalization, radiological study showed pneumomediastinum, making her not a candidate to IMV. After 5 weeks of admission and several desaturation episodes her respiratory status evolved until support with VMK 40%. The CT scan ruled out the possibility of Pulmonary Embolism and showed a moderate pneumonia, bronchiectasis, and subpleural bands suggesting an advanced in ammation phase (Figure 1.D).
The medical therapy administrated to both patients consisted of lopinavir/ritonavir, hydroxychloroquine, azithromycin, piperazillin/tazobactam, prophylactic doses of low-molecular-weight-heparins (LMWHs), corticosteroids (methyprednisolone 250mg x 3 boluses) and Tocilizumab (single dose). Despite this pharmacotherapy, prone position and the oxygen support, the respiratory status and high in ammatory parameters of both patients kept worsening. At this point, their enrollment in the ULTRA-COVID study was decided and a single ultra-LDRT was administered on April 23rd, 2020.

Clinical Status
Respiratory status improved rapidly in both patients. Patient 1 showed an improvement on his 02-Sat and PaFi02 (>300) two days after the treatment. Supplemental oxygen with 2L NC was discontinued at day ve and he was discharged on day eight after ULTRA-COVID with 95% 02-Sat values while he was breathing ambient air, keeping this breath status 1 and 2 months later. Patient 2 showed a slower recovery, achieving less need of oxygen support at 2, 5 and 7 days after the treatment, intermittently requiring 2L NC 1 month after and, nally, two months later no longer need of it. PaFi02 passed 300mmHg at day 5. She was discharged 14 days after ultra-LDRT.
Subjectively, both patients reported improvement of symptoms such as less asthenia and dyspnea. After 48h the in ammatory parameters showed a decrease, however they uctuated or stabilized on subsequent controls. Above all, we highlight the decrease in IL-6 (CRS). The viral status results by PCR 1 month after were positive in patient 1 and negative in patient 2. Immune state showed immunity measured as positive IgG by rapid test.
No signi cant or acute adverse events were observed, even after two months of follow-up period.
Radiological status ULTRA-COVID intervention showed an improvement on the TSS score in both patients after the rst scan, with a higher improvement on patient 1 ( Table 2). The rst CT-scan on patient 1 showed brotic subpleural bands and bronchial and vascular retraction was also present (Figure1.B). The second scan-at 4 weeks showed fewer consolidation areas, although moderate lung involvement persisted. (Figure 1.C) Patient 2 rst scan showed the resolution of the pneumomediastinum, brosis and loss of volume were present, most accused during the monthly control, as well as fewer consolidation areas. Globally, we observed an improvement from moderate to mild-moderate pattern. (Figure 1.E and 1.F).

Discussion
Experimental therapies have been used since the outbreak of the COVID-19 pandemic and expert societies guided to enroll patients in clinical studies when possible, as no standard treatment exists so far. Although other drugs have been tested, the only potential supportive treatment for COVID-19 appears to be an IL-6 inhibitor, tocilizumab (17).
RT is a cost-effective non-toxic treatment available in most hospitals (18). The radiobiological basis is not well known, but literature available suggests that the ultra-LDRT would modulate the overreacted immune-landscape (8)(9)(10). The design of clinical trials to investigate RT effectiveness in COVID-19 scenario has been justi ed by the RT community and ongoing trials such as the one in Atlanta (RESCUE1-19), Italy (COLOR-19), 2 in Spain (IPACOVID and LOWRAD-Cov19), Iran (NCT04390412), Ohio (VENTED), New Delhi (NCT04394793) and Massachusetts (NCT04393948) exist (19). Despite these ongoing trials, the patient recruitment is still challenging.
With ULTRA-COVID study we wanted to analyze the anti-in ammatory effect of ultra-LDRT and its potential use to mitigate this pandemic. We discuss the potential activity of ultra-LDRT by presenting clinical recovery of two patients with SARS-Cov-2 pneumonia with a poor response to standard medical therapy and would otherwise have no other treatment than IMV, for which they were not candidates. To our knowledge, these are the rst results of using a single 0.8 Gy of helicoidal-RT for COVID-19 pneumonia treatment, with a follow-up of 2 moths, and adding lung CT scan and immunity status assess.
We agree that the timing of the irradiation in relation to disease progression is likely to in uence treatment outcomes. However, it is still unclear the best timing for ultra-LDRT administration. It appears that is imperative to distinguish between the dominant viral phase versus the host in ammatorycytokines-mediated response. Literature supports that at chronic stages of disease low-dose irradiation might not be as e cient as in the early progressive stage (20). In fact, in our study, the faster recovery and the greater radiological improvement in patient 1 who presented with an early stage of the disease, as opposed to patient 2, whose disease stage was advanced , suggests that the earlier we apply the ultra-LDRT in the acute in ammation host phase, the better the global response.
We defend that ultra-LDRT target is the cytokine storm and the treatment key is anticipating the need of ICU and IMV, because not all critical cases are candidates for it (19,20). Between 7 to 12 days from the onset of the symptoms appears to be the most effective moment to apply LDRT (3).
The use of RT could stop the progression to an advance stage of the disease; and this could help curb down the death rates, now 3.4% on average, and higher among elder and ICU patients (21). The optimal dose of RT is unknown with evidence supporting 0.5-1Gy (2,4,5). The average prescription doses of the ongoing trials are 0.5 and 0.7 Gy in a single dose and; optionally, an additional 0.5 Gy fraction 48h later. We decided to escalate dose and apply 0.8 Gy in a single session given the unawareness about the virusŕ esponse and because this is the minimum effective dose of the technique used (helicoidal Tomotherapy).
We believe that the strengths of our study are that we have been rigorous in our inclusion criteria. In addition, patients have been followed by CT scan and, although signi cant clinical response is accompanied by lower radiological one, both have happened. Finally, we have performed viral and immunity status control by PCR and rapid antibodies tests, respectively.
Although LDRT may increase the effectiveness of anti-viral immune responses, it does not decrease the viability of virus directly (20). It could be the reason why the PCR of patient 1 was positive 4 weeks later.
The main general concern about the RT use in non-neoplasic disease is its carcinogenesis risk. Evidence available supports this risk is quite scarce when low-doses are used (0.1-4/1000-<1/10000) (22)(23)(24). Given this, the old age of patients and the life-threatening condition they face, we believe it could be assumed. Nevertheless, a bigger study sample and a longer follow-up period is necessary to ratify this.
Our results should be understood as preliminary and more de nitive results are expected from this prospective study as well as from other ongoing clinical trials. A bigger study will need to face the possible bias, such as the lack of study recruitment due to a decrease in COVID-19 cases when the trial started.
We also consider the possible lead time bias as a factor in our study, due to the difference disease phase in which patients were included. Another survival factor that could interfere with our results is the drug response, above all, tocilizumab.
Despite these limitations, the preliminary results on our rst two patients suggest a good response to RT and encourage us to continue.
Researchers at Emory University have also shared the preliminary results of day-7 analysis of their phase I/II trial (RESCUE1-19). Five patients were treated with a single dose of 1.5Gy. They have also supported that LDRT appears to be safe without any acute toxicities noted and have shown early promise of e cacy. (27) This report highlights the importance of identifying those patients and could bene t from LDRT in order to decrease the worsening of the disease, especially in those patients who cannot bene t from IMV.
Further studies could demonstrate the effectiveness of LDRT, considering it as an alternative cotreatment with pharmacotherapy (or even exclusive). This could be of a great importance in countries without or limited access to expensive drugs or ICUs where it can, even become a standard of care in SARS-CoV-2 management. Additionally, we support that it could have an impact on other possible viral pneumonia epidemics in which CRS has been observed (28).

Conclusion
Preliminary clinical and radiological results suggest a potential bene t of treating SARS-Cov-2 pneumonia with ultra-LDRT during the acute in ammatory phase with a positive impact on the disease´s evolution.

Declarations
Ethic statements: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Approved by the HM Hospitales Group Ethics Committee (CEIm) on April 21st. CEIm code 20.4.1597-GHM Consent statement: Patients signed consent was obtained after verbal information on the procedure, objective and secondary effects of the treatment. Figure 1 CT scan evaluation