Pain hypersensitivity, sensorimotor impairment, and decreased muscle force in a novel rat model of radiation‐induced peripheral neuropathy

Authors: Allison B. Vittert, Melissa Daniel, Shelby R. Svientek, Mary Jane Risch, Noah S. Nelson, Alexis Donneys, Amir Dehdashtian, Gina N. Sacks, Steven R. Buchman, Stephen W. P. Kemp

Abstract

Introduction

Radiation‐induced peripheral neuropathy is a rare, but serious complication often resulting in profound morbidity, life‐long disability, and chronic debilitating pain. Unfortunately, this type of peripheral neuropathy is usually progressive, and almost always irreversible. To date, a standardized rat model of radiation‐induced peripheral neuropathy has not been established. The purpose of the present study was to examine neuropathic pain, sensorimotor impairment, and muscle force parameters following the administration of a clinically relevant radiation dose in a rat model.

Methods

Ten rats were randomly assigned to one of two experimental (1) radiation and (2) sham‐radiated controls. Radiated animals were given a clinically relevant dose of 35 Gray (Gy) divided into five daily doses of 7 Gy/day. This regimen represents a human equivalent dose of 70 Gy, approximating the same dosage utilized for radiotherapy in oncologic patients. Sham‐radiated controls were anesthetized and placed in the radiation apparatus but were not given radiation. All animals were tested for baseline values in both sensorimotor and pain behavioral tests. Sensorimotor testing consisted of the evaluation of walking tracks with the calculation of the Sciatic Functional Index (SFI). Pain‐related behavioral measures consisted of mechanical allodynia (von Frey test), cold allodynia (Acetone test), and thermal allodynia (Hargreaves test). Animals were tested serially over an 8‐week period. At the study endpoint, electrophysiological and muscle force assessments were completed, and histomorphometric analysis was performed on all sciatic nerves.

Results

Animals that underwent radiation treatment displayed significantly greater pain hypersensitivity to mechanical stimulation as compared to sham radiated controls from weeks 4 to 8 of testing. SFI values indicated sensorimotor impairments in the overground gait of radiated animals as compared to non‐radiated animals. Furthermore, radiated animals displayed reduced twitch and tetanic muscle force when compared to sham radiated controls.

Conclusions

A clinically relevant human equivalent dose of fractionated 35 Gy in rats established significant pain hypersensitivity, impairments in sensorimotor locomotion, and decreased muscle force capacity. This novel rodent model of radiation‐induced peripheral neuropathy can be utilized to assess the potential efficacy of therapeutic treatments to either prevent or remediate this clinically debilitating condition.

INTRODUCTION

Radiation therapy (XRT) is one of the mainstays of cancer therapy as approximately 50% of cancer patients require it as part of their treatment regimen. ^1^ Of these patients, approximately 31% subsequently suffer from debilitating neuropathic pain for which there are no therapeutically efficacious treatment regimens. ^2^ XRT‐induced peripheral neuropathy results in progressive, intractable, and often irreversible nerve pain which can have a devastating effect on a patient's quality of life, and potentially reduce the overall benefit of XRT ^3^ on survivorship. Furthermore, a patient's ability to carry out daily activities of living (ADL) is hampered secondary to sensory deficits in light touch, pain, vibration, position, sensation, and temperature. ^4^

Among the tissues commonly affected by radiation, peripheral sensorimotor nerves are at particular risk given their reduced reparative capabilities following injury. ^5^ In contrast to acute traumatic nerve injury where disruption of the myelin and/or axon results in immediate deficits, radiation‐induced nerve injury can take months to years to appear following radiation exposure. ^6^ , ^7^ Specifically, radiation‐induced peripheral neuropathy (RIPN) involves the following sequential pathologic events independent of (1) chronic inflammation in a pre‐fibrotic, asymptomatic phase; (2) excess extracellular matrix deposition and fibroblast proliferation in an organized fibrotic phase; and (3) symptomatic fibro‐atrophic phase characterized by poor vascularity. ^8^ As a radiation injured nerve progresses along this sequence, demyelination and resultant conduction delay/block occur, thereby causing associated neurological deficits. ^1^ , ^7^

RIPN can be associated with a vast array of sensorimotor deficits over the entire body, ranging from breast cancer‐associated brachial plexopathy ^4^ to optic nerve neuropathy and associated blindness. ^9^ Although the degree of deficit often correlates with the volume of the area treated and the cumulative radiation exposure, ^1^ deficits can still result from minor amounts of exposure and complete characterization of the cause–effect relationship is still limited in the literature. ^7^ Most notably, chronic pain can be a significant complication following radiotherapy with a prevalence near 90% in certain cancers. ^10^ Despite this frequency, surprisingly little is known as to the pathophysiology, prevalence, and adequate treatment modalities. ^4^ , ^10^

To properly assess these problematic gaps in the literature, a reliable animal model of radiation‐induced neuropathy is required. Although several animal models have been previously utilized that have investigated the effect of radiation on overground locomotion, muscle force, and histology, there is limited data on pain behaviors in these animals. ^11^ , ^12^ , ^13^ In this paper, we propose a novel rat model of RIPN following clinically relevant human equivalent doses (HEDR) of radiation to the hindlimb. The primary purpose of this study was to evaluate the resultant motor deficits and neuropathic pain behaviors following irradiation of the sciatic nerve in this novel rat model.

METHODS

Animals and experimental groups

Ten adult male Lewis rats weighing approximately 300–400 g (Charles River) were acclimated and maintained in a specific pathogen‐free (SPF) environment on a 12 h dark cycle. Rats were randomly assigned to one of two experimental groups (n = 5/group): (1) Control group—sham radiation and (2) Experimental group—radiation.

Radiation protocol

All procedures were approved by the University of Michigan Committee on the Use and Care of Animals (IACUC) in accordance with the NIH Guide for the Care and Use of Laboratory Animals. The University of Michigan NIH/NCI radiation core assisted with the appropriate XRT dosing and delivery. Prior to and following radiation therapy, all rats were fed a standard chow diet with access to water ad lib. Following initiation and maintenance of inhalational anesthesia (2% isoflurane in 98% oxygen; VetOne), the right thigh was irradiated in a custom fabricated radiation apparatus to maximize sciatic nerve exposure utilizing a Philips RT250 orthovoltage unit (250 kV, 15 mA; Kimtron, Oxford, CT) fractionating the dose at approximately 3.72 Gy/min (see Figure 1 for detailed experimental radiation set‐up). Appropriate lead shielding was utilized to prevent radiation exposure to adjacent anatomical structures. Over a 5‐day period, rats were exposed to a total of 35 Gy of fractionated radiation at 7 Gy/day. Previous studies in our laboratory have established fractionated exposures of 35 Gy over a 5‐day period as equivalent to therapeutic and standard human exposures of 70 Gy ^14^ , ^15^ for oncologic applications. Control group animals were similarly anesthetized and placed in the radiation apparatus for the same duration as experimental animals without irradiation taking place.

Functional assessments

Walking track analysis

To determine the recovery of overground locomotion, we utilized the Sciatic Functional Index (SFI). ^16^ , ^17^ Rat hindpaws were dipped in black ink and allowed to voluntarily ambulate along an 8.2 × 36 cm walking track lined with a white sheet of paper. Measurements for the following parameters were recorded using the resultant (1) experimental print length (EPL); (2) normal print length (NPL); (3) experimental toe spread (ETS); (4) normal toe spread (NTS); (5) experimental intermediary toe spread (EIT); and (6) normal intermediary toe spread (NIT). The SFI was calculated according to the following equation based on the work of Bain et al. ^17^ : SFI = −38.3 × [(EPL−NPL)/NPL] + 109.5 × [(ETS−NTS)/NTS] + 13.3*[(EIT−NIT)/NIT]−8.8.

Von Frey filament test

Nerve injuries frequently cause mechanical allodynia, and in rats, and this behavior manifests commonly as a painful withdrawal response to a non‐noxious mechanical stimulus. ^18^ , ^19^ To determine if radiation led to mechanical allodynia in this study, the von Frey test was used. ^18^ , ^20^ As the sciatic nerve provides innervation to the plantar surface of the paw via the tibial nerve, the central portion of the plantar paw was the target of assessment. Rats were singularly placed in a plastic cage on the testing platform and allowed to acclimate to the experimental environment for 15 min. The filaments (range 0.008–100 g) were applied to the central paw until an appropriate bend was observed in the filament. Testing was initiated utilizing the 6 g filament, and if a response was noted, lighter filaments were subsequently used until no response occurred. If no response was noted following the application of the 6 g filament, heaver‐weighted filaments were used. The lowest filament to evoke a response was then recorded as the value for that trial. Following a 5‐min inter‐trial interval, subsequent trials were initiated at the filament considered one grade lighter than the result obtained for the preceding trial. A total of three trials were conducted for each rat, and results were averaged for each week's assessment.

Hargreaves test

In addition to mechanical allodynia, nerve injuries typically cause thermal allodynia which is a painful withdrawal response to a non‐noxious thermal stimulus. ^21^ To assess heat allodynia, we used the Hargreaves test from IITC Life Science Technology. ^22^ Rats were singularly placed in a clear plastic enclosure with a glass surface, with a radiant heat source (adjustable beam intensity in 1% increments of room temperature to a maximum of 250°C) positioned below the location of the rat's hindpaw. Following an acclimation period, the heat source was activated, automatically initiating an electronic timer. When the rat withdrew its hindpaw from the surface, the heat source was immediately stopped, and the duration of time elapsed from the initiation of heat to paw withdrawal was recorded.

Acetone test

To assess cold allodynia, animals were placed in a similar experimental setup to that described for the Hargreaves test. Following acclimation, 50 μL of acetone was applied to the mid‐plantar surface of the hindpaw, and subsequent pain‐related behaviors were recorded over a 2 min testing period. ^23^ Pain‐related behaviors that were documented paw withdrawal, paw licking, paw shaking, and constant biting of the affected paw.

Electrophysiologic testing

Following the conclusion of the eighth week of behavioral testing, in situ endpoint assessments were performed under inhalational anesthesia as previously described. ^24^ , ^25^ A surgeon who was blinded to the experimental groups made incisions over both the thigh and lower limb to isolate the sciatic nerve and extensor digitorum longus (EDL) muscle, respectively. Nerve conduction studies were subsequently performed using a hook electrode (Harvard Apparatus) placed on the proximal sciatic nerve, bipolar surface probe electrodes (Neurosign) on the distal EDL, and a ground electrode inserted at the base of the tail. Conduction studies were performed using a series of bipolar electrical stimulations (1 pulse/s, 0.1 ms duration) to the proximal sciatic nerve, increasing in amplitude until the maximal compound muscle action potential (CMAP) was recorded distally at the EDL muscle.

Muscle force testing

Following completion of the electrophysiologic assessment, both maximal twitch (F t) and tetanic (F o) forces were measured at the distal EDL muscle. Following the dissection of the distal EDL tendon and transection at the level of the distal interphalangeal joints, the tendon was then secured to a stainless steel S‐hook after placing the animals in a custom‐designed force measurement jig (Red Rock Inc., St. Louis, MO). The S‐hook was then connected to a 2 N thin film load cell (S100, Strain Measurement Devices Inc., Meriden, CT) to measure muscle forces generated by the EDL muscle. Cathodic, monophasic electrical impulses (delay = 200 ms, duration = 300 ms, frequency = 40–80 Hz, and amplitude = 0–3 V) were generated using a single‐channel isolated pulse stimulator (Model 2100, AM Systems Inc., Carlsborg, WA) and delivered to the proximal sciatic nerve via a hook electrode (Harvard Apparatus). Recorded passive and total force measurements were then utilized to calculate the active force tracings using MATLAB software (The Mathworks Inc., Natick, MA). Isometric twitch contractions were obtained and were then utilized to calculate the optimal muscle length (L o) and stimulus amplitude (A o) for maximal isometric force production. Using these values, single twitch contractions were recorded, and F t was calculated. Maximal F o was then determined by delivering 300 ms bursts of increasing stimulation frequency (5–200 Hz) to the sciatic nerve to define a force‐frequency curve, whereupon F o was calculated from the active force plateau.

Statistical analysis

Descriptive statistics and graphical analysis were utilized to analyze the experimental data. A comparison of means between experimental and control groups was performed using an unpaired t‐test for each experimental parameter. Statistical significance was set at p ≤ .05. All statistical analyses were performed via SPSS (IBM SPSS Statistics 25, Chicago, Illinois, USA).

RESULTS

All rats involved in this study survived to endpoint evaluation 8 weeks after radiation/control therapy. All experimental animals healed well following radiation therapy without the development of open wounds and/or infection. No episodes of self‐mutilation or autotomy were noted post‐treatment. For the following described evaluations, all were conducted by a single evaluator blinded to the experimental groups.

Walking track analysis

Overground locomotor recovery was assessed with the SFI which ranges from a score of 0 (normal) to −100 (severely impaired). ^17^ Radiation‐treated animals demonstrated greater sensorimotor impairment compared to control animals [t (16)=4.23, p < .05]. Animals that were administered radiation had significantly lower SFI scores compared to controls starting at Week 4 through Week 8 of the study (Figure 2).

Electrophysiologic testing

CMAPS were measured at the EDL muscle following ipsilateral stimulation of the proximal sciatic nerve. When comparing CMAPs at the study endpoint between the experimental radiation group (13.3 ± 1.62) and the control group (12.82 ± 0.92), no significant differences were seen (Figure 3A). Although trending toward a decreasing muscle mass, no significant differences were seen in EDL muscle weights between the two groups (experimental: 176.4 ± 7.48; 206.4 ± 19.78, Figure 3B).

Muscle force testing

Following stimulation of the proximal sciatic nerve, both the twitch and tetanic muscle force were measured at the distal EDL muscle. The evaluation demonstrated a significant reduction of twitch force, t (8) = 3.36, p < .01 (experimental: 401 ± 65.01; 642 ± 29.67, Figure 3C). Similar results were seen when tetanic force was measured. The evaluation demonstrated a significant reduction of tetanic force, t (8)=4.69, p < .005 (experimental: 961.60 ± 198.40; 1901 ± 27.03, Figure 3D).

Mechanical allodynia (von Frey filament testing)

The von Frey test measures mechanical allodynia which is a painful response to a non‐noxious substance and is shown in Figure 4. Upon assessment, von Frey testing revealed a significantly lower withdrawal threshold in animals receiving radiation starting at Week 4 [t (8) = 3.85, p < .05], Week 5 [t (8) = 4.72, p < .05], Week 6 [t (8) = 3.65, p < .05], Week 7 [t (8) = 6.84, p < .05], and Week 8 [t (8) = 4.83, p < .05].

Temperature allodynia

Animals were tested on both cold allodynia (Acetone test) and thermal allodynia (Hargreaves test). For both tests, there were no significant differences between either of the two groups (Figure 5A,B).

DISCUSSION

Major findings

The major findings of this study were (1) overground locomotion as measured by the SFI was impaired in the radiation group starting in Week 4 through Week 8; (2) animals who underwent radiation displayed significant decreases in both muscle twitch and tetanic forces when compared to the sham group; (3) the radiation group displayed significantly more mechanical allodynia when compared to the sham controls from Week 4 through Week 8; and (4) there were no significant differences between the groups in response to either cold or heat.

Walking track analysis reveals significant differences in overground locomotion between control and radiation‐treated animals

Radiation‐induced toxicities often result in major impairments to cancer survivors' quality of life. ^26^ , ^27^ For example, sensory neuron dysfunction following radiation treatment can lead to impairments in normal gait. ^28^ , ^29^ In one study assessing the effects of palliative radiotherapy for brain metastases, patients reported a significant deterioration in motor function, and weakness of the legs as soon as 3 months after treatment. ^30^ These deficits are likely attributed to damage of multiple sensory modalities including in the brain, spinal cord, peripheral nerves, and muscle. ^31^ Medical management of this complication involves intensive exercise rehabilitation, often lasting a lifetime. However, the insidious progression of radiation‐induced damage ultimately leads to continued weakness and motor dysfunction which is currently unavoidable. ^28^

Our study aimed to recapitulate motor deficits displayed by patients following radiation therapy. To the best of our knowledge, there are no current rodent models of radiation‐induced peripheral neuropathy that mimic the clinical scenario. We therefore introduced a clinically relevant fractionated human equivalent radiation dose to the hindlimb of the rat and evaluated it. Our results show that radiated animals displayed a significant decrease in SFI values starting from Week 4 through Week 8 of our study (Figure 4). These results are similar to the clinical scenario, where the effects of radiation are not immediate. In this situation, patients who undergo radiation therapy experience delayed symptoms following treatment due to the progressive nature of radiation‐induced fibrotic damage to nerves. ^32^ One example of this is radiation‐induced lumbosacral radiculoplexopathy, where patients typically present lower motor neuron dysfunction following a significant delay from the stoppage of radiation treatment. ^33^

Electrophysiologic analysis demonstrates similar motor recruitment between control and radiated animals

Interestingly, our study found no significant differences in either CMAPs or final muscle weights (Figure 3). Our results are in contrast to that of Liao et al., ^34^ who previously described the progression of changes in CMAPs every 2 months following radiation. In that study, the CMAPs of 5 of the study's 42 rats displayed no changes in their CMAPs following radiation. However, half of the rats in the study demonstrated significantly reduced CMAPs and latencies were delayed at 2 months following the initial radiation. One important difference between that study and the current one is the dose of radiation. Liao et al. (2016) used a radiation dose of 40 Gy, whereas the present experiment used 35 Gy. Furthermore, their dosing was not fractionated like in the current study consistent with current normal clinical practice. Indeed, fractionation was specifically introduced to reduce the concomitant side effects of radiotherapy. It is possible that either our sample size or radiation exposure was not powerful enough to observe changes in these muscle parameters. In a more recent randomized control trial evaluating the treatment of radiation‐induced plexopathy in the lumbosacral region, CMAPs were only slightly reduced after the initial radiation and did not significantly change 18 months post‐radiation. ^35^

In terms of muscle weight, the radiation treatment group demonstrated a trend toward a decrease in muscle mass. This result was expected given that radiation damages the satellite cells responsible for muscle regeneration and causes hypertrophy after damage. ^36^ Further, Jurdana et al. ^36^ showed that a dose of 45 Gy was not sufficient to cause muscle necrosis in an otherwise healthy cancer patient, whereas a dose of only 20 Gy was sufficient for muscle necrosis in a debilitated cancer survivor. It is therefore possible that if the radiation dose was increased or if the experimental cohort was observed for a longer period of time that the muscle mass difference would have reached statistical significance.

Muscle force testing reveals significant force deficits in radiation‐treated animals

Muscle weakness and atrophy have been shown to occur as early as 3 months following exposure to radiation or as late as 45 years following initial radiation exposure. ^37^ , ^38^ Specifically in irradiated head and neck cancer (HNC) patients, nerve injury results in a severe disabling condition named Dropped Head Syndrome due to weakness of extensor muscles and would require a multimodal therapy consisting of surgery and physical therapy. ^38^ To recapitulate these clinical findings of extensor muscle weakness, we exposed the rat hindlimb to a HEDR similar to that experienced of HNC patients. In our study, we observed a ~38% decrease in EDL twitch force in our irradiated animals. This is in accordance with another group who exposed the mice hindlimb to a total of 24.6 Gy of radiation and described a 27% decrease in the fast twitch of the extensor digitorum longus muscle. ^39^ In another experiment assessing maximum strain and ultimate tensile strength of muscle after a single dose of 0, 10, or 20 Gy exposure, rats with 10 and 20 Gy were significantly weaker than the rats without irradiation. ^40^ Thus, we believe we have created the murine model of radiation‐induced nerve injury that appropriately simulates muscle injury found clinically.

Radiated animals demonstrate chronic neuropathic pain

The incidence of chronic pain in the cancer survivor population is dramatically increasing secondary to the availability of improved treatment modalities. ^41^ In fact, studies have described a phenomenon called central nervous system sensitization where cancer survivors suffer from debilitating pain hypersensitivity notably toward tactile stimulation due to radiochemotherapy. ^42^ Specifically, in breast and colon cancer patients, available evidence has described diminished thresholds to mechanical stimulation at either the affected side, contralateral side, and/or distal sites post‐intervention, while other studies illustrated local altered sensorium suggesting injury to the peripheral nervous system. ^43^ In addition to decreased tolerance toward mechanical pressure, patients also report increased sensitivities to light, sound, heat, cold, and electrical stimuli. ^44^ Furthermore, to the best of our knowledge, there are few effective strategies to remediate this pathology nor available animal models to test for potential therapeutics for this ailing growing population. Thus, it is crucial to address this unmet clinical need by creating a murine model of radiation‐induced neuropathic pain to study the disease progression with the goal of developing potential therapeutics that hold the prospect of significantly mitigating this devastating morbidity.

There is a multitude of animal models of neuropathic pain that primarily have outcome measures that focus on thermal and/or mechanical hypersensitivity through cancer infiltration, direct nerve damage, and pharmacologic manipulation. ^45^ However, studies do not consistently include assessment of chronic inflammatory and neuropathic pain which is a proxy for ongoing nociception and would accurately represent the pain progression of cancer survivors. ^46^ To the best of our knowledge, our murine model of radiation‐induced neuropathic pain is the first of its kind to characterize the behavioral pain response in detail that exhibits similar behaviors observed clinically. Comparable to human observations in the literature, rodents in our study exposed to radiation demonstrated a statistically significant reduction in tolerance of tactile stimulation compared to the sham controls from Week 4 through Week 8. This prolonged reaction to mechanical stimuli may be suggestive of the chronic inflammation cancer survivors also suffer from. One difference that we observed in our study compared to human behaviors is that we found no statistical differences between the sham and irradiated groups for thermal allodynia. Overall, irradiated animals demonstrated a trend toward decreased tolerance to cold stimuli. Our laboratories are actively assessing this phenomenon in a new cohort of rats.

Limitations

We acknowledge certain limitations of the current experimental studies. Although the primary focus of the paper was on behavioral (sensorimotor overground locomotion, pain) and electrophysiological (CMAPS, muscle force) measures, there are additional assessments that would have made the results more comprehensive. Previous studies have evaluated nerve histomorphometry/histology following radiation in rats with several showing axonal degeneration of the sciatic nerve. ^11^ , ^12^ , ^34^ , ^47^ , ^48^ Gorgulu et al. (2003) further demonstrated extensive fibrosis of the sciatic nerve using histological assessment of trichrome staining. ^49^ Ongoing studies in our laboratory are not only assessing nerve histology/histomorphometrical parameters following radiation of the sciatic nerve but we are also actively investigating the effect of radiation on several muscle parameters such as fibrosis, central nuclei counts, and alterations in neuromuscular junction number and function. ^50^ , ^51^ Our lab is also actively assessing changes in gene expression in individual cells following radiation of the sciatic nerve utilizing single‐cell mRNA sequencing. ^52^ To the best of our knowledge, this assessment has not been previously reported in the literature utilizing a rodent model of radiation‐induced peripheral neuropathy.

CONCLUSION

Despite the wide array of studies conducted in rodents to evaluate functional deficits following peripheral nerve injury, ^53^ relatively little has been published on the complete functional and pain assessment following radiation treatment. As detailed in the previous section, the discrete evaluation of neuropathic pain following radiation injury is largely non‐existent. This paper comprises the first comprehensive battery of pain tests in a clinically relevant rat model of radiation injury. In our model, rats exposed to a clinically relevant fractionated HEDR of radiation displayed significant deficits in both muscle twitch and tetanic force. Furthermore, these animals displayed chronic neuropathic pain through the assessment of mechanical allodynia. The results of this study are significant because it reliably produces clinically relevant deficits that are also seen in patients who have undergone radiation treatment. This model can be used in the future for investigation into the therapeutic efficacy of potential treatment modalities for this debilitating condition.

FUNDING INFORMATION

Research reported in this publication was supported by a grant from the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under Award Number P30 AR069620.

CONFLICT OF INTEREST STATEMENT

None of the authors has any conflict of interest to disclose.