key: cord-324823-bw2x9h45
authors: Little, Mark P.; Zhang, Wei; van Dusen, Roy; Hamada, Nobuyuki
title: Pneumonia after bacterial or viral infection preceded or followed by radiation exposure - a reanalysis of older radiobiological data and implications for low dose radiotherapy for COVID-19 pneumonia
date: 2020-10-01
journal: Int J Radiat Oncol Biol Phys
DOI: 10.1016/j.ijrobp.2020.09.052
sha: 
doc_id: 324823
cord_uid: bw2x9h45

Purpose Currently, there are about 15 ongoing clinical studies on low dose radiotherapy (LDRT) for COVID-19 pneumonia. One of the underlying assumptions is that irradiation of 0.5–1.5 Gy is effective at ameliorating viral pneumonia. We aimed to reanalyze all available experimental radiobiological data to assess evidence for such amelioration. Methods and Materials With standard statistical survival models, and based on a systematic literature review, we re-analyzed thirteen radiobiological animal datasets published in 1937-1973 in which animals (guinea pigs/dogs/cats/rats/mice) received radiation before or after bacterial/viral inoculation, and assessing various health endpoints (mortality/pneumonia morbidity). In most datasets absorbed doses did not exceed 7 Gy. Results For six studies evaluating post-inoculation radiation exposure (more relevant to LDRT for COVID-19 pneumonia) the results are heterogeneous, with one study showing a significant increase (p<0.001) and another showing a significant decrease (p<0.001) in mortality associated with radiation exposure. Among the remaining four studies, mortality risk was non-significantly increased in two studies and non-significantly decreased in two others (p>0.05). For pre-inoculation exposure the results are also heterogeneous, with six (of eight) datasets showing a significant increase (p<0.01) in mortality risk associated with radiation exposure and the other two showing a significant decrease (p<0.05) in mortality or pneumonitis morbidity risk. Conclusions These data do not provide support for reductions in morbidity or mortality associated with post-infection radiation exposure. For pre-infection radiation exposure the inconsistency of direction of effect is difficult to interpret. One must be cautious about adducing evidence from such published reports of old animal datasets.

Low dose radiotherapy (LDRT) for Coronavirus Disease 2019 (COVID- 19) pneumonia was proposed in early April 2020 (1, 2) . At least 15 clinical studies are currently ongoing in 9 countries (3). The rationale for clinical benefit, in other words the effectiveness of irradiation at the level of 0.5-1.5 Gy in treating viral pneumonia, largely relies on early human case studies or animal studies mostly obtained in the pre-antibiotic era, when a number of attempts were made to treat various non-cancer diseases with ionizing radiation, including virally-or bacteriallyassociated pneumonia. An influential paper underlying a number of proposals made for use of LDRT to treat COVID-19 pneumonia (1,2) was Calabrese and Dhawan in 2013 (4) who reviewed 17 papers describing various relatively small case series, describing outcomes from low dose radiotherapy with X-rays (LDRT) for pneumonia. Their sampling frameworks are unknown, and therefore they are subject to ascertainment bias and are largely uninterpretable.

Calabrese and Dhawan (4) also identified four radiobiological animal studies of post-inoculation LDRT, all from experiments done in the 1940s, namely Fried (5) using a guinea pig model, Lieberman et al. (6) using a canine model, Baylin et al (7) using a cat model, and Dubin et al. (8) using a murine model, the first two of these for bacterially-induced pneumonia and the last two for virally-induced pneumonia. However, Calabrese and Dhawan (4) did not consider four other radiobiological studies relating to post-inoculation RT, nor results of eight others relating to preinoculation RT, and did not attempt any statistical reanalysis of these old data.

The aim of the present paper is to look at the totality of published radiobiological data relating to radiation exposure before or after inoculation with a viral or bacterial agent likely to result in pneumonia. Because of the age of the data being considered there are shortcomings in the original statistical analysis that was performed -indeed in all but a few cases (9,10) there J o u r n a l P r e -p r o o f was no formal statistical analysis in the original reports. It is the purpose of this paper to report reanalysis of the data abstracted from the original publications so far as that is achievable, using standard statistical survival models in order to assess modification of pneumonia morbidity or mortality risk by radiation exposure before or after inoculation.

We aimed to capture all radiobiological datasets relating to moderate or LDRT whether given before or after viral or bacterial inoculation leading to pneumonia. We searched literature by means of a PubMed search (using terms ((radiation OR radiotherapy) AND pneumonia AND viral AND animal) OR ((radiation OR radiotherapy) AND pneumonia AND bacterial AND animal)) conducted on 2020-8-8, which returned 184 articles. We also searched for citations of the articles of Fried (5), Lieberman et al. (6) , Baylin et al (7) , Dubin et al. (8) , and an authoritative contemporary review (of 1951) by Taliaferro and Taliaferro (11) on the same date.

We did not restrict by date or language of the publication. We selected from these searches all relevant articles with information on radiobiological animal experiments in which there was any type of ionizing radiation exposure with determination of mortality or morbidity from bacterially-or virally-induced pneumonia. The datasets used are listed in Table 1 . It should be noted that the datasets we used include three of the four cited by Calabrese and Dhawan (4), but did not include the paper of Fried (5) which we judged did not contain any quantitatively useful information. In Appendix A we provide details of the process used to abstract data from the publications that we identified as being potentially informative. The data was abstracted independently three times by XXX, YYY and ZZZ. We convert the given free-in-air dose in J o u r n a l P r e -p r o o f legacy units radiation absorbed dose (rad), roentgen (R) or rep in all studies to absorbed dose in Gray (Gy) via the scaling 1 R/rep =0.00877 Gy. 1 rad = 0.01 Gy (12) .

Details of the statistical models fitted are given in Table 1 , and some further details on adjustments used are also given in the summary Table 2 . Mortality and morbidity risks in the radiobiological cohorts of Lieberman et al. (6) and Dubin et al. (8) were assessed using a Cox proportional hazards models (13) , with time after radiation exposure, if that followed the inoculation, or time after bacterial or viral inoculation, if that followed the radiation exposure, as timescale, in which the relative risk (RR) (=hazard ratio) of death for animal i at time a after inoculation was given by a linear model in dose:

or alternatively using a log-linear model in dose:

where i D is the total dose (in Gy), α is the excess relative risk coefficient (ERR) per unit dose (Gy). For most of the other datasets a linear logistic model is fitted to the data (generally on number of animals that died in each group):

In some cases the more standard loglinear logistic model is fitted to the data (generally on number of animals that died in each group):

For the data of Fried (14) , numbering only 7 animals and using as outcome improvement in pneumonia in relation to unirradiated controls, an exact logistic model was used (15), as non-J o u r n a l P r e -p r o o f exact methods did not converge. It is well known that the excess odds ratio (EOR) approximates to the excess relative risk (16) . All confidence intervals (CIs) and two-sided p-values are profilepartial-likelihood based (17) . In the murine dataset of Dubin et al. (8) in various subgroups risks were assessed in relation to radiation dose administered after inoculation or dose before inoculation. In the murine dataset of Quilligan et al (18) pre-inoculation dose was given to all animals; there is ambiguity in Quilligan et al (18) as to whether radiation exposure may also have been given post-inoculation. There is some uncertainty associated with the number of mice in the first of the control groups in this dataset, so a range is employed, spanning the plausible range of 6-10 mice, with 8 as the central estimate ( Table 2 ). The model was stratified by the three experiments reported in the data of Dubin et al. (8) and by the three groups used by

Lieberman et al. (6) . Tables B2 and B3 and Figure 1 and 2 show the risks in relation to dose for these two datasets. In various other datasets adjustment was made for various other covariates, as detailed in Tables B1-B14. In the fits to the pneumonia intensity data of Baylin et al. (7) we used either loglinear logistic regression (as described above) comparing each pneumonia intensity group and those with greater intensity vs every group with reduced intensity and because of the small number of animals (22) we also employed exact logistic methods (15); we also used ordinal regression with log-linear link (19) fitting to all the ordered intensity groups. In fits of the days of infection data of Baylin et al. (7) we used a linear regression model, estimating the CIs via the bias-corrected advanced method (20) . All models were fitted via Epicure (21), R (22) or LogXact (15). Two-sided levels of statistical significance are reported in all cases, with a conventional threshold for type I error of 2-sided p<0.05 used to assign statistical significance.

All statistical analyses were independently performed by XXX and YYY to check for J o u r n a l P r e -p r o o f concordance. All datasets and analysis files are available in online supporting information (Appendix C).

We present results of analyses of risk in relation to whether radiation exposure occurred after inoculation or before inoculation. The results are given in summary form in Table 2 , which also provides summary details of the models used and assumptions made in fitting, and in more detail in Appendix B Tables B1-B14. Table 2, Table   B2 ), as also shown by Figure 1 . However, Table B2 shows that this is largely driven by a single group, group 3, as also shown by Figure 2 . The three groups in the study of Lieberman et al (6) , were treated with slightly different X-ray energies, 80 kVp, 135 kVp and 200 kVp, respectively, and mean doses also slightly differed, 0.947 Gy, 1.639 Gy and 2.027 Gy, respectively (Table   B2 ).

There are few indications of trend of degree of pneumonia infection with dose in the feline data of Baylin et al. (7) , whether using logistic, exact logistic or ordinal models ( Table 2, Table   B3 ). However, Table B4 indicates that there is a significant decreasing trend of days of acute Figure 3 ); results did not appreciably vary with the type of model used (linear logistic, log-linear logistic) or whether or not adjustment was made for the virus concentration (Table B5 ).

There is a highly significant increasing trend (p<0.001) of mortality risk with dose after endemic coccobacillus infection in the rat data of Bond et al. (24) , whether adjusting for likelihood of infection or not ( as also shown by Figure 3 .

There is no significant trend (p>0.4) with post-inoculation dose in the data of Dubin et al. (8) , whether using linear or log-linear models, which is confirmed also by Figure 1 (Table B7) .

However, there is a borderline significant decreasing trend (p=0.029) of mortality with preinoculation dose in this dataset, with EOR/Gy = -0.62 (95% CI -0.90, -0.09) ( Table 2, Table B7) again confirmed by Figure 1 .

In the murine data of Beutler and Gezon (25) there are highly significant (all p<0.005) increasing trends of mortality with post-inoculation dose, whether in relation to mouse-adapted or eggadapted PR8 influenza A virus and irrespective of the type of statistical model (linear logistic, loglinear logistic) used; for example with a linear logistic model the EOR per Gy is 0.23 (95% CI 0.08, 0.43) ( Table 2, Table B8) , and as also shown in Figure 4 . The morbidity trends exhibit more heterogeneity, so that for the mouse-adapted virus the trends are generally negative, so that J o u r n a l P r e -p r o o f for example with a linear logistic model the EOR per Gy is -0.19 (95% CI -0.19, -0.18) (Table   B8 ); however, for the egg-adapted virus the trends are generally positive with dose, so that for example with a linear logistic model the EOR per Gy is 1.02 (95% CI 0.39, 2.17) (Table B8) , as also shown in Figure 4 .

In the Swiss mice data of Hale and Stoner (26) there is no overall trend (p>0.2) of mortality with radiation dose given before inoculation with type III pneumococcus. However, if attention is restricted to the animals that received inoculation there is a highly significant increasing trend with dose (p<0.001), so that the EOR per Gy is 1.40 (95% CI 0.39, 5.47) ( Table   2 , Table B9 ). The same researchers went on to study a wider range of infective agents in the same strain of mice, and observed a generally highly significant (p<0.005) increase in mortality risk associated with radiation before inoculation with influenza virus, pneumococcus type III bacterial infection or Trichinella spiralis larval infection (27), whether adjusted or not for type of first immunizing infection, so that for example without such adjustment and excluding the Trichinella spiralis challenge infections the EOR/Gy = 1.71 (95% CI 0.97, 3.02) ( Table B10,   Table 2 ).

There is a highly significant (p<0.001) increase in mortality risk in the C57BL mouse data of Quilligan et al (18) associated with post-influenza-inoculation radiation dose with EOR/Gy ranging from 3.73 (95% CI 0.42, 85.85) to 4.75 (95% CI 0.56, 108.10) depending on how many mice are assumed to be in the first control group (Table B11, Table 2 ).

Pneumonitis morbidity and mortality were significantly decreased (p<0.001) after 3.5 Gy whole body air-dose exposure in adult male albino CF-1 mice in the data of Berlin (9) , so that for pneumonitis morbidity the EOR/Gy = -0.24 (95% CI -0.28, -0.17) and for pneumonitis mortality the EOR/Gy = -0.21 (95% CI -0.26, -0.14) (Table B12, Table 2 ). In contrast Table 2 (see also J o u r n a l P r e -p r o o f Table B13 ) and Figure B2 show reanalysis of slightly later data of Berlin and Cochran (10), which exhibits slightly heterogeneous results, with one set of experiments (given in Table III of Berlin and Cochran (10)) indicating a highly significant increase (p≤0.02) in influenza mortality, whether or not adjusted for mode of administration of virus, but a different experimental set (reported in Table II of the paper) showing no significant effect (p>0.1) of radiation exposure on influenza morbidity or mortality. These experiments use a similar murine system, also given 3.5

Gy whole body air-dose exposure, as in the earlier paper of Berlin (9). Lundgren et al (28) used a novel type of radiation exposure, aerosolized 144 CeO 2 , which delivers localized β dose to the lungs of C57BL/6J mice. There was a small but highly significant increase in mortality risk associated with radiation exposure, with EOR/Gy = 0.007 (95% CI 0.002, 0.017, p=0.002) ( Table 2, Table B14 ).

We have re-analyzed thirteen radiobiological animal datasets, dating from the late 1930s to the early 1970s, in which bacterial or viral agents were administered to induce pneumonia in animals that were also exposed to varying fractionated doses of radiation before or after inoculation. The statistical analysis in the original papers was limited, indeed in all but two cases (9,10) there was no formal statistical analysis in the publications. We therefore judged it necessary to statistically reanalyze the data from the original publications with standard statistical models.

For the six studies that evaluated post-inoculation radiation exposure (which is more (6); the groups were treated with slightly different X-ray energies, and mean doses also slightly differed (see Results and Table B2 ). It is possible that these variations in mean dose and radiation energy may have some bearing on the differences observed ( Table 1 ). It is possible that the range of doses used, and the variable degree of fractionation employed may also be factors, although there does not appear to be an obvious pattern linking these to the direction or strength of effect, as shown by Tables 1 and 2. More recently, Hasegawa et al (29) showed that irradiation before influenza vaccination and a J o u r n a l P r e -p r o o f succeeding lethal challenge influenza inoculation exacerbates mortality from influenza, but that vaccination prior to irradiation confers protection against a subsequent challenge influenza inoculation. Dadachova et al (30) reported that targeted radionuclide immunotherapy induces

Streptococcus pneumoniae killing in vivo, thereby alleviating bacterial pneumonia. As can be inferred, both these publications (29, 30) address somewhat different questions to those of this paper.

Calabrese and Dhawan (4) reviewed three of the studies we consider here (6) (7) (8) , and a fourth (5) which we judged did not contain any quantitatively useful information, and stated that "these studies constitute the entire set of animal model studies assessing the capacity of X-rays to affect pneumonia-induced clinical symptoms and mortality. Each study demonstrated some measure of support for the hypothesis that X-ray treatment could reduce the effects of the pneumonia induced by bacteria or viruses." Manifestly this is not the entirety of the literature relating to post-inoculation radiation exposure (Tables 1, 2) , and a review of our results (Table 2) demonstrates that there is little evidence overall of reduction of morbidity or mortality with increasing radiation dose. Calabrese and Dhawan (4) also reviewed 17 papers (all published before 1945) describing 15 or 16 mostly relatively small case series (the uncertainty reflecting whether or not an otherwise unpublished case series mentioned by a discussant in a paper of Powell (31) was included) and concluded that "X-ray therapy was successful in decreasing the mortality rate in untreated patients from about 30 percent to 5 to 10 percent". However, as noted in the Introduction, without a well-defined sampling framework such data are likely subject to ascertainment bias and are effectively uninterpretable. The radiation doses employed in some of the case series are also unknown. Similar conclusions have been arrived at by others in discussing this paper (32, 33) .

The major strength of the present analysis is that we use standard statistical models to assess the totality of published radiobiological data on LDRT given either before or after virallyor bacterially-induced pneumonia. Moreover, unlike the review of Calabrese and Dhawan (4) In this appendix we outline the steps taken to transcribe data from the 14 papers relating to radiobiological animal experiments produced by our literature search.

Data was transcribed from 

Data was transcribed from Tables 1-3 (pp.97-98) into a table giving dog number, group number, blood culture (coding "…." = -1, "+"=1, "++"=2, "+++"=3, "++++"=4, "0"=0) pneumonia type ("I"=1,"III"=3), radiation energy (keV)(from table title), radiation 1 st day (Gy converting 1 r = 0.00877 Gy), radiation 2 nd day (Gy converting 1 r = 0.00877 Gy), radiation 3 rd day (Gy converting 1 r = 0.00877 Gy), radiation total (Gy converting 1 r = 0.00877 Gy), radiation total excluding day of death (Gy converting 1 r = 0.00877 Gy), death day, dead (coding 1=dead, 0=no death). For those animals that are indicated as having recovered (animal 519 in Table 2 , animals 316, 383, 385, 481, 588 in Table 3 ) a large value (9999) was used as the end of follow-up time J o u r n a l P r e -p r o o f (Death day), with death coded as 0 for these 6 animals, so that the records for these animals only contribute to the denominator in the Cox model fit.

We constructed a 

We constructed a table from Tables II and III Table II and Table III .

J o u r n a l P r e -p r o o f

We used a combination of the data in Table 1 (p.28) and Table 2 (p.29) to create a table with fields for experiment number, series, sex, age start weeks (minimum), age start weeks (maximum), infection status [coded as likely unexposed to infection, probably exposed to infection, highly probably exposed to infection -so that Experiments 1 and 2 were both coded as probably exposed to infection, and all except the last group of Experiment 3 were coded as highly probably exposed to infection, the last group of Experiment 3 being coded as likely unexposed to infection], minimum radiation dose (in Gy, using 1 r =0.00877 Gy), maximum radiation dose (in Gy, using 1 r =0.00877 Gy), deaths infected, deaths total, survived infected, survived total. These numbers were used to compute entries for numbers of deaths among infected and uninfected animals, which formed the basis of the analysis. The radiation dose used was the mean of the minimum and maximum doses.

Dubin et al ( Table. We continued in this way for all elements of this Table. All information about doses at various times before or after inoculation came from the text description above each row in the Table. There was information on 252

individual mice in this data.

From for the first group, increasing by 1 for each successive group), the radiation dose (Gy, using 1 r = 0.00877 Gy), the number of mice that died and the number exposed.

Likewise we constructed a similar table for morbidity from mouse-adapted virus, based on Table III (p.233) with records containing for each group of animals the experiment number, the suspension group, the negative log viral dilution, the group number (an arbitrary number, which we took to start at 1 for the first group, increasing by 1 for each successive group), the radiation dose (Gy, using 1 r = 0.00877 Gy), the number of mice with morbidity and the number in each treatment group.

From J o u r n a l P r e -p r o o f 

Observed Odds ratio = 1

Is low dose radiation therapy a potential treatment for covid-19 pneumonia?

Covid-19 tragic pandemic: Concerns over unintentional "directed accelerated evolution" of novel coronavirus (sars-cov-2) and introducing a modified treatment method for ards

Gov 18 studies found for covid-19 radiation

How radiotherapy was historically used to treat pneumonia: Could it be useful today?

The roentgen treatment of experimental pneumonia in the guinea-pig

Roentgen therapy of experimental lobar pneumonia in dogs

The effect of roentgen therapy on experimental virus pneumonia; on feline virus pneumonia

The effect of roentgen therapy on experimental virus pneumonia; on pneumonia produced in white mice by swine influenza virus

Sparing effect of x-rays for mice inoculated intranasally with egg-adapted influenza virus, cam strain

Delay of fatal pneumonia in x-irradiated mice inoculated with mouseadapted influenza virus, pr8 strain

Effect of x-rays on immunity; a review

unit)). Wikipedia

Regression models and life-tables

Die artefizielle Pneumonie und ihre Bestrahlung. Experimentalbeitrag zur Frage der Wirkung der Röntgenstrahlen auf Entzündungsgewebe

Statistical methods in cancer research. Volume II -The design and analysis of cohort studies

Generalized linear models

Continuous cobalt-60 irradiation and immunity to influenza virus

Regression models for ordinal data

Better bootstrap confidence intervals

Risk Sciences International

Project version 3.6.1. R: A language and environment for statistical computing

Effect of roentgen therapy on mouse encephalitis

The effects of X irradiation on a naturally occurring endemic infection

The effect of total body X irradiation on the susceptibility of mice to influenza A virus infection

The effect of cobalt-60 gamma radiation on passive immunity

Effects of ionizing radiation on immunity

Effects of inhaled 144 CeO 2 on influenza virus infection in mice

Protection against influenza virus infection by nasal vaccination in advance of sublethal irradiation

Feasibility of radioimmunotherapy of experimental pneumococcal infection

Low dose radiation therapy for covid-19 pneumonia: Is there any supportive evidence?

Is there any supportive evidence for low dose radiotherapy for covid-19 pneumonia?

Low-dose radiation therapy for covid-19 pneumopathy: What is the evidence?

Flying by the seat of our pants: Is low dose radiation therapy for covid-19 an option?

Lack of supporting data make the risks of a clinical trial of radiation therapy as a treatment for covid-19 pneumonia unacceptable

United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR)

Systematic review and meta-analysis of circulatory disease from exposure to low-level ionizing radiation and estimates of potential population mortality risks

 Tables and Figures of results  Table B1 . Improvement in pneumonia morbidity in guinea pigs associated with radiation exposure post inoculation to Staphylococcus aureus haemolyticus in the data of Fried (14)