### Marco Durante

1 Introduction

While the XX century is remembered for the first human space travel, the XXI century will be characterized by the colonization of the Solar System. Current exploration programs include missions to near-Earth asteroids, the moon (Moon Village), and Mars. Astronauts should be prepared to remain in space for longer periods, and the recent 1-year mission on the International Space Station (ISS) is considered a stepping stone to future missions beyond low- Earth Orbit (BLEO). The question is whether the extraterrestrial space is safe enough for humans to allow colonization. Traditionally, three main risks are identified in space:

i) Physiological problems caused by microgravity (or reduced gravity).
ii) Psychological and medical problems caused by isolation.
iii) Acute and late risks caused by exposure to radiation.

The physiological changes due to weightlessness have been extensively studied in LEO.

Bone loss, kidney stone formation, skeletal muscle mass reduction, cardiovascular alterations, impaired sensory-motor capabilities, immune system dysfunctions are among the consequences of prolonged stays in microgravity. The risks are very well characterized, and several countermeasures are available. None of these risk are rated 1 (the highest rank, corresponding to a risk that makes the mission impossible without mitigation) in the NASA Bioastronautics roadmap.

Isolation may lead to serious neurobehavioral problems caused by poor psychosocial adaptation. Several ground platforms are used to study these problems and develop countermeasures, such as the Concordia base in Antarctica and the Mars500 isolation facility in Russia. Isolation also brings the problem of autonomous medical care (AMC), i.e. the capability to handle sickness or accidents in complete isolation. This is a risk category 1 for the mission to Mars. Countermeasures for AMC risks are mostly technological, i.e. rely on the development of portable medical equipment and telemedicine.

Radiation risk is therefore now generally acknowledged as the main showstopper for the human colonization of the Solar system. In the past 20 years, space agencies have invested resources in research on space radiation research, both with experiments on ISS or BLEO on radiation dosimetry and with important ground-based radiobiology programs: the NASA Space Radiation Health Program and the ESA Investigation on Biological Effects of Radiation (IBER) program. Certainly the results of these experiments have substantially reduced the uncertainty compared to those present at the time of the NASA Roadmap. Yet, as it happens, they have also found new problems that beg for new questions. The most recent results will be summarized below.

2 Dosimetry

Recent flight experiments have now largely reduced our uncertainty on the radiation dose absorbed in space.

The MATROSHKA experiment (fig. 3) provided the first accurate map of the internal organ dose in LEO. Astronauts’ dose is always measured with badge dosimeters, providing the skin dose. Dose to the critical organs (e.g. bone marrow, lung, gonads) is calculated using transport codes, which are affected by large uncertainties. The anthropomorphic phantom MATROSHKA, a modified phantom used in radiotherapy for quality assurance, was filled with different passive and active dosimeters and was exposed to space radiation both outside and inside the ISS. The accurate dose maps registered with the phantom were used to tune the Monte Carlo and analytical codes used to calculate the effective dose in space. Fluence to dose conversion factors for different organs have been tabulated in the most recent report of the International Commission of Radiological Protection (ICRP) on exposure in space.

The second major improvement comes from the measurements of the Radiation Assessment Detector (RAD) instrument on the Mars Science Laboratory (MSL), carrying the rover Curiosity (fig. 4), during the cruise to Mars and on the planet’s surface. Measurements were accumulated around the 2012–2013 solar maximum activity, and represent the first accurate determinations of the physical dose and of the equivalent dose in BLEO. Even though the mission was around the solar maximum period, SPE only contributed 5% to the total dose during the journey, perhaps because the present solar maximum is relatively weak. During solar minimum the solar magnetic field is reduced and the GCR equivalent dose rate can be up to two times higher. However, the actual dose rate within the spacecraft will depend on the shielding. Based on the MSL data, we can calculate the expected doses in different Mars mission scenarios (table 1). It is interesting to see that most of the dose is incurred during cruise phase. The dose on the planet can be further reduced using bases with heavy shielding, exploiting in situ planetary materials.

3 Biological effects

The health risks associated to the exposure to space radiation have been discussed in several reports and publications, and can be essentially divided into four groups:

i) Cancer
ii) Tissue degenerative late effects
iv) Hereditary effects

Radiation carcinogenesis has been traditionally considered the main health risk associated to radiation exposure. Dose limits for workers and astronauts are based on the cancer risk, and therefore major research efforts have been dedicated to reducing the uncertainty on cancer risk. Most of the uncertainty shown in fig. 1 is caused by the radiation quality (fig. 5). To convert physical dose (in Gy) into equivalent dose (in Sv) ICRP recommends the use of an LET-dependent quality factor, but these values were notoriously affected by large uncertainties, because of the lack of epidemiological data. The organ dose equivalent $H_{T}$ is calculated as

(1) $H_{T} = \frac{1}{m} \int_{m} \mathrm{d}m \int Q ( L ) F_{T} ( L ) L \mathrm{d}L$ ,

where $L$ is the linear energy transfer (LET=$\mathrm{d}E/\mathrm{d}x$, normally expressed in keV/μm), $m$ the organ mass, $F_{T}$ the fluence of particles through the organ $T$ (in particles/cm2), and $Q$ the LET-dependent quality factor (dimensionless).

While radiation protection on Earth has the solid epidemiological base of the A-bomb survivors’ database, no epidemiological data are available for exposure to energetic protons and high-LET heavy ions. To simulate this radiation, both NASA and ESA supported radiobiological research at high-energy accelerators: the NASA Space Laboratory (NSRL) at the Brookhaven National Laboratory (BNL) in Upton, Long Island (NY, USA) and the SIS18 synchrotron of the GSI Helmholtz Center for Heavy Ion Research in Darmstadt (Hessen, Germany). Animal experiments at NSRL have shown that the quality factor for heavy ions is strongly dependent on the tumor (fig. 6): a high quality factor is found for some solid tumors (e.g. liver cancer) but values around unity are measured for liquid cancers (leukemia), much lower than those found for fission-spectrum neutrons. These data have certainly reduced the uncertainty on the quality factor for carcinogenesis, thus improving the estimates of the equivalent does for long-term missions (table 1)

However, accelerator-based studies have also found significant noncancer effects induced by energetic charged particles. A major concern comes from the effects on the central nervous system (CNS), because even low doses (down to 5 cGy) of energetic heavy ions induce significant impairment in spatial, episodic and recognition memory in mice, associated to deficits in executive function and reduced rates of fear extinction and elevated anxiety. This was somehow surprising, because the brain is generally considered a radioresistant organ, but is consistent with observed reductions in dendritic complexity, spine density (fig. 7) and altered spine morphology along mice medial prefrontal cortical neurons. The low-dose effects are caused by the high density of energy deposition along the ion tracks, which can now be visualized in the neural tissues using immunohistochemistry (fig. 8). CNS risk is therefore considered a major concern, and its uncertainty is higher than cancer.

A second tissue late effect has gained large importance in radiation protection: cardiovascular disease. In fact, epidemiological evidence (A-bomb survivors, Mayak workers, and radiotherapy patients) clearly show that radiation can damage the heart even at doses below 1 Gy, even if the data do not exclude a possible threshold at low doses. Even if the excess relative risk per unit dose is small, cardiovascular morbidity is the main cause of death in adults, and therefore even a small increase in risk corresponds to a high mortality. Experiments at accelerators point to an increased effectiveness of heavy ions in the induction of vascular endothelial cell dysfunction, possibly associated to an increased cardiovascular mortality among Apollo astronauts.

Acute effects are possible in case of major SPE, but the solar protons can be effectively shielded. The risk is limited to exposure during EVA, e.g. on the moon surface. Uncertainty is related to the threshold doses for prodromal syndromes under reduced gravity. The main countermeasure is reliable space weather forecast ability. Today, SPE forecasts are still inadequate to provide advance warning with sufficient credibility to lead operators to initiate protective measures, but substantial improvements in understanding solar energetic particle acceleration and propagation have been accomplished in the past few years.

Hereditary effects are generally considered of minor importance, especially because there is no epidemiological evidence of transgenerational effects in humans, not even from the large A-bomb cohort. Nevertheless, the relative effectiveness of protons and heavy ions for hereditary effects is still unknown and this gap in knowledge should be filled at least with accelerator-based experiments.

4 Countermeasures

A fundamental tenet of radiation protection is that there are three means to reduce exposure to ionizing radiation: increasing the distance from the radiation source, reducing the exposure time, and by shielding. Distance is not an issue in space, GCR being isotropic. Time in space should be increased rather than decreased according to the plans of exploration and colonization, although reduction of the transit time to the planet, where heavy shielding can be more easily achieved, may contribute to reducing radiation exposure.

The problem in cosmic rays shielding derives from the physics interaction of charged particles with matter. Cosmic radiation is very energetic and produces showers of light fragments and neutrons by nuclear fragmentation when hitting the shields. Both electromagnetic energy loss (Bethe-Bloch formula) and nuclear fragmentation cross-sections per unit mass decrease by increasing the target atomic number $A$. Therefore, light, highly hydrogenated materials are more effective per unit mass in decreasing the dose than heavy, high-$Z$ materials such as Al (common structural material in the spacecraft) or Pb (generally used for shielding radiation sources on Earth). This prediction has been confirmed in many accelerator-based tests and in tests on the ISS using polyethylene and water.

Shielding calculations are performed using analytical (such as the NASA HZETRN code) or Monte Carlo (GEANT4, FLUKA, PHITS) codes. The prediction of the codes have some substantial differences, due to the different physics models used and to the lack of reliable cross-sections for the fragmentation of several ions at high energy. Predictions of different codes for the dose equivalent in free space as a function of different thickness of Al and polyethylene are shown in fig. 9. Apart from clear significant inter-code differences, the plot shows a peculiar minimum around 20 g/cm2 Al-shielding, and the equivalent dose tends to increase for thicker shields. This paradoxical result is caused by the generation of neutrons, whose high quality factor eventually increases the dose equivalent behind the shield. This effect is not seen when light shielding materials are used, yet all codes predict saturation for high thickness, suggesting that there will be no gain in heavier shelters.

The situation will be similar for a planetary base (fig. 10) where, however, the thickness of the shield can be increased $>100$ g/cm2 using in situ resources (fig. 11) or placing the bases in deep, underground caves. As water is an effective shielding material, one interesting option would be to cover an inflatable base with a shell of ice, which could be extracted from Mars (fig. 12). The Mars ice home (or igloo house) would be nearly transparent, hence allowing natural light inside compared to the dark sub-surface modules (fig. 11) or the deep caves.

Another approach to improve shielding would be the use of specially designed spacesuits filled with highly hydrogenated materials. This can be water, or more sophisticated light fibers, which can be modeled on every crewmember individual body and made thicker in correspondence of the most sensitive organs (e.g. breast or gonads). This is the concept of the AstroRad space vest (fig. 13), developed by the Stemrad company and that will be tested on a phantom in the next Orion mission on the moon.

Biomedical countermeasures would be ideal for reducing the effects of space radiation. Unfortunately, at present radioprotectors have low effectiveness, or are too toxic for use in space. The only radioprotective drug used in clinics to mitigate the side effects of radiotherapy in the head-and-neck region is amifostine (marketed by Clinigen as Ethyol) an organic thiophosphate effective in free radical scavenging. However, amifostine has severe side effects, including vomiting and diarrhea, preventing its use in astronauts. Dietary antioxidants are beneficial in reducing longterm radiation effects, but their effectiveness is limited. Research in the field is very active, and several new drugs and molecules have been tested. New molecules selected in US within the homeland security program are also under test for protection of radiotherapy patients and possibly astronauts.

5 Dose limits

The significant advancements in knowledge of the past years should be reflected in a revision of the dose limits for the astronauts. This actually happened at NASA, where a careful gender- and age-specific risk model is used. The model calculates the risk of exposure-induced death (REID) and its associated uncertainty based on the calculated equivalent organ doses. The dose limits are set to maintain the individual astronaut’s REID <3% within a 95% confidence interval. The most recent version of the model includes the cardiovascular disease mortality in addition to cancer. The model is complex and highly dependent on the inputs from scientific experiments. The other space agencies prefer more pragmatic approaches for dose limits, often based on extrapolation of the maximum annual dose for radiation workers on Earth to a full career. A summary of the dose limits is provided in table 2. Certainly these limits should be made more uniform among the different agencies. There is a general consensus on the ALARA (as-low-as-reasonably-achievable) principle also for space activities, but an agreement on the dose limits would be desirable. The current level of knowledge seems to be sufficient to grant a sciencebased limit for interplanetary missions, and this value is needed for the design of countermeasures, such as the shield thickness.

6 Conclusions

We learned a lot in the past decade about the radiation risk in space. Flight measurements have largely increased our knowledge of the exposure in LEO and BLEO. Acceleratorbased experiments have reduced the uncertainty on the biological effects of cosmic rays, but they also unexpectedly showed that noncancer endpoints (CNS and cardiovascular) could be even more harmful than carcinogenesis for crews of interplanetary missions. The development of countermeasures, both physical and biomedical, remains indispensable for a safe human space exploration.