Multiple sclerosis, brain radiotherapy, and risk of neurotoxicity: The Mayo Clinic experience

The published literature to date regarding the effect of EBRT has shown substantial toxicity resulting from demyelinating injury occurring approximately 2 months after treatment in most cases and delayed injury occurring after 1 year or more in 2 cases. The patients reported in the literature have been selected on the basis of toxicity; the lack of a series of consecutive patients prevents making a definitive estimate of the relative risk of toxicity after irradiation when prospectively considering treatment for patients. More recent case reports of treatment of patients with conformal, modern radiotherapy have not confirmed the toxicity initially suspected in publications before 2000.

The present study is the largest modern series of patients treated with therapeutic irradiation. The 15 cases in the present series were derived from approximately 12,000 patients seen for MS or other demyelinating diseases at Mayo Clinic during the time period of the study and, given the infrequency of treatment, it is unlikely that prospective data will ever be available on toxicity risks for such patients. The results of our investigation support the impression given in prior, isolated case studies that MS patients are at higher risk for neurotoxicity after EBRT treatment than patients without a history of MS. Patients in our population who received bilateral, central brain radiotherapy with opposed fields appeared to be at the highest risk for subsequent neurotoxicity, although the very small number of patients treated only to the periphery of the brain prevents any definitive conclusions. The constellation of symptoms observed in patients with toxicity indicates lesions involving the central white matter and/or brainstem, a finding corroborated by results of the univariate analysis showing an association between the frequency of neurotoxicity and irradiation of these areas and the temporal lobes.

The present study was designed to assess the rates of severe and fatal injury occurring as sequelae of radiotherapy in patients with definite MS as defined by established criteria. This study was not able to assess the rate of transient, mild flares of MS resulting from treatment, if such events do occur more frequently during and after treatment. It was not able to investigate whether CNS EBRT changed the natural history of MS in long-term survivors. The limited statistical power of the study, its retrospective nature, the various types of MS presentations, and the competing symptoms from brain involvement by tumor precluded any realistic possibility of reliably assessing these interesting and concerning issues. Although the symptoms in any given patient in the series could be attributed, in theory, to an acute worsening of MS from causes unrelated to radiotherapy, this appears highly unlikely given the frequency of neurotoxicity seen in the series as a whole in comparison to the gradual decline seen on average in groups of patients with MS.

Toxicity from treatment in the present study appeared to be secondary to lesions occurring within CNS tissue directly exposed to radiotherapy, as assessed by clinical examination, radiologic studies, and autopsy. Although this would appear to be a self-evident conclusion, radiotherapy toxicity has been noted to manifest beyond the treated region in organs outside the CNS. One example of this is the bilateral effects in radiation pneumonitis acute respiratory distress syndrome which can progress beyond the radiotherapy portals, presumably from an effect induced by inflammatory cytokines or other biochemical mediator. In our study, radiotherapy to the brain did not appear to induce a generalized worsening of MS outside the treated field such that it resulted in Grade 3 or greater toxicity.

Despite the high frequency of neurotoxicity in our study, half of the patients with toxicity were treated before 1986 with radiation techniques and doses not typically used at present. Extrapolation of the study results to the present era is problematic in this regard. Only 1 case of toxicity occurred in patients treated after 1995, and in this case the reviewers were divided in attributing the patient’s neurologic injury to treatment effect. Modern radiotherapy, using conformal techniques to spare uninvolved areas of brain tissue from unnecessary treatment, may decrease the risk of neurotoxicity in MS patients.

Reduction in the amount of white matter receiving radiation would appear to decrease the risk of neurotoxicity in MS patients, but the extent of this decrease remains undefined, and some risk will likely persist even with small conformal fields. It is not possible to quantify the risk associated with three-dimensional EBRT on the basis of the current series with its small sample size. Patients with toxicity were treated with fields resulting in a homogeneous dose distribution extending from the periphery of brain along the axis treated. With the narrow dose range used, no analyzable dose–response data are available for doses lower than the traditional range of therapeutic and palliative brain treatment.

Detailed anatomic and pathologic information regarding the CNS lesions causing neurotoxicity was available in only 1 case from autopsy data. In this patient, brain toxicity correlated with large areas of demyelination in the central white matter. It appears logical to suggest that the basic mechanism of injury in patients with treatment toxicity is related to acceleration of the process of focal demyelination seen in MS. Subacute radiation effects can lead to temporary demyelination and/or can disrupt the blood–brain barrier. Three of 6 cases of brain toxicity, as well as many of the previously published reports of toxicity in MS patients receiving EBRT, occurred within a window of time in which subacute EBRT injury may develop. We hypothesize that MS patients either (1) lack the ability to repair radiation-induced demyelination of the CNS that patients without preexisting MS would be able to withstand, or (2) that radiation-induced changes in the blood–brain barrier may allow immune-mediated effects on the irradiated brain, both of which could lead to further injury. Alternatively, the defect causing MS may lead to early delayed radiation toxicity becoming static instead of reversible, and injury that would be recoverable in healthy persons then leads to permanent radiation toxicity. This is conjecture, however. In the present study, it was difficult to retrospectively assess the etiology of neurotoxicity and attribute it to demyelination vs. other forms of brain toxicity. Also, given the limitations of magnetic resonance imaging before the 1990s, it is possible that unrecognized tumor recurrence may have led to symptoms attributed to treatment toxicity. However, the long survivals noted after the rapid occurrence of disabling symptoms in 4 of 6 cases suggest that tumor progression was not the underlying cause.

Of 5 patients treated with EBRT without a prior diagnosis of definitive MS, 3 received a subsequent diagnosis of MS primarily on the basis of occurrence of symptoms attributable to treatment-related injury. It is unclear whether clinically apparent MS would have developed subsequently in these patients in the absence of radiotherapy. The frequency of neurotoxicity derived from patients in Group A, who had evidence of MS at the time of EBRT, best estimates the risk of neurotoxicity.

Conclusion

In summary, brain radiotherapy in patients with MS is associated with an increased risk of neurotoxicity. This risk is most pronounced for treatment with parallel-opposed radiotherapy fields encompassing the temporal lobes, brainstem, and central white matter of the brain. Of note, the results of the present study report the frequency of toxicity historically in a single institution using radiotherapy doses and treatment techniques that typically would not be used today. Given the advances in treatment planning and conformal radiotherapy that have occurred in recent years, it is unclear whether the results of our study are applicable when modern 3D treatment planning is used to treat brain malignancies. Brain radiotherapy in the setting of MS should be carefully considered, weighing the expected benefits of treatment against the risks of tumor progression without radiotherapy. For situations in which EBRT is indicated but other options exist, such as radiosurgery in the treatment of isolated brain metastases, consideration could be given to alternative treatment. The ultimate risk of neurotoxicity for any treatment of the CNS with ionizing radiation in patients with MS remains to be defined.