Efficacy and predictors of pain relief

Overall, CK successfully provided acute pain relief, with its efficacy gradually lessening over time. The initial pain relief rate after the first treatments was 86.3%, which is similar to the 83% rate reported in a recent meta-analysis by Spina et al. (2021) on GK for TN in patients with MS.^{Reference Spina, Nocera, Boari, Iannaccone and Mortini11} Initial pain relief occurred at a median of 14 days, which is faster than typically observed in GK cohorts.^{Reference Xu, Mathieu and Heroux8,Reference Alvarez-Pinzon, Wolf and Swedberg12} This rapid response may be attributed to the distinctive non-isocentric delivery properties of the CK technique compared to the isocentric approach of the GK. However, as this is the first study exclusively dedicated to CK in MS patients, further validation is necessary to confirm this unprecedented finding. The sustained pain relief rates (86.0% at 6 months, 52.2% at 12 months, 35.0% at 36 months and 24.2% at 60 months) were comparable to those reported in larger studies,^{Reference Xu, Mathieu and Heroux8– Reference Conti, Pontoriero and Iati10,Reference Franzini, Tropeano and Olei14} which documented pain relief ranging from 54% to 85% at 12 months, 35% to 65% at 24 months, 55% to 57% at 36 months, 24% to 44% at 48 months and 52% to 57% at 60 months.^{Reference Xu, Mathieu and Heroux8– Reference Conti, Pontoriero and Iati10,Reference Franzini, Tropeano and Olei14}

In an effort to optimize candidate selection for radiosurgery, we further assessed whether specific clinical and radiological factors contributed to persistent pain relief. Interestingly, the presence of a cisternal vascular compression was identified as a predictor. Although this finding has not been documented in MS patients, it has been observed in non-MS cohorts.^{Reference Brisman, Khandji and Mooij15,Reference Erbay, Bhadelia and Riesenburger16} This favorable response in non-MS, classical TN patients with a neurovascular conflict may be attributed to the underlying pathophysiology of the disorder. In this regard, in classical TN, the commonly accepted mechanism for the painful symptoms is that mechanical compression by surrounding vessels results in focal demyelination.^{Reference Antonini, Di Pasquale and Cruccu17,Reference Maarbjerg, Wolfram, Gozalov, Olesen and Bendtsen18} To support this, recent neuroimaging studies reported white matter abnormalities in the cisternal segment of the trigeminal nerve.^{Reference Fujiwara, Sasaki and Wada19,Reference Leal, Roch, Hermier, Souza, Cristino-Filho and Sindou20} This demyelination process affects thickly myelinated sensory fibers, likely leading to ectopic excitation, high-frequency discharges and ephaptic transmission to nociceptive fibers, generating painful signals.^{Reference Devor, Amir and Rappaport21,Reference Devor, Govrin-Lippmann and Rappaport22} Although the mechanism of action of radiosurgery in classical TN is not entirely clear, it is recognized that radiation, which targets the cisternal segment of the nerve where the aforementioned pathological processes occur, may directly interrupt the transmission of nociceptive impulses through a local inhibitory effect.^{Reference Hodaie, Chen, Quan and Laperriere23,Reference Kondziolka, Lacomis and Niranjan24}

To date, the pathophysiology of TN in patients with MS remains incompletely understood, but most authors agree that demyelinating plaques are the main contributors to the painful symptoms. Indeed, in their neuroimaging study assessing microstructural abnormalities in MS patients, Chen et al. (2016) reported white matter lesions in the pontine tract of the trigeminal nerve.^{Reference Chen, DeSouza, Hayes, Davis, O’Connor and Hodaie25} Although this explanation is compelling, it is important to note that in many MS cases, vascular compression coexists with a plaque.^{Reference Antonini, Di Pasquale and Cruccu17,Reference Broggi, Ferroli and Franzini26,Reference Athanasiou, Patel, Renowden and Coakham27} Just like in classical TN cases, this vascular compression may also contribute to the painful symptoms. Interestingly, an elegant neuroimaging study by Truini et al. (2016) reported that demyelination and vascular compression could act synergistically to elicit pain, a theory they coined the “double-crush mechanism.” The authors suggested that these two processes could give rise to distinct types of lesions affecting the trigemino-thalamic tract, an inflammatory lesion of the central myelin and a mechanical injury affecting the peripheral myelin sheaths.^{Reference Pichiecchio, Bergamaschi, Tavazzi, Romani, Todeschini and Bastianello28,Reference Truini, Prosperini and Calistri29} More advanced neuroimaging techniques that highlight microarchitectural features would aid in understanding the extent to which this double mechanism contributes to the pathophysiology.

Beyond potentially unveiling the drivers of painful stimuli, the double crush mechanism theory may also explain why MS patients with neurovascular conflicts benefit from more sustained pain relief following radiosurgery. Indeed, according to this theory, these patients also exhibit pathological processes resulting from the vascular compression itself, making radiosurgery potentially more effective for them through local peripheral mechanisms unrelated to demyelinating plaques.^{Reference Hodaie, Chen, Quan and Laperriere23,Reference Kondziolka, Lacomis and Niranjan24} Conversely, in cases where a demyelinating plaque in the brainstem is the sole contributor to the symptoms, radiosurgery – which does not target the central myelin – has minimal impact on the primary pain generator, potentially resulting in less effective pain relief. To our knowledge, no study identified vascular compression as a predictor of better efficacy of radiosurgery for TN in MS. This novel finding is clinically pertinent as it may help streamline the best candidates for radiosurgery, though further validation is needed.

Active smokers did not exhibit better pain relief than nonsmokers (p = 0.098), contrasting with Weller et al. (2014), who hypothesized that smokers experience more beneficial effects due to a more potent radiation effect on the demyelination process. Smoking may exacerbate radiosurgically induced demyelination, potentially leading to quicker nerve dysfunction, interruption of painful stimuli and better therapeutic effect.^{Reference Weller, Marshall and Lovato9} Although a statistical trend was observed, our results were not significant, likely due to the limited statistical power of our sample size.

In classical and idiopathic TN cases, cumulative evidence suggests that radiosurgically induced hypesthesia could predict efficacy.^{Reference Tuleasca, Regis and Sahgal30} Indeed, hypesthesia reflects significant radiosurgically induced axonal degeneration, a potent injury likely sufficient to interrupt peripheral painful signals.^{Reference Gorgulho31} However, this predictive factor was not observed in the current study or in other MS studies.^{Reference Spina, Nocera, Boari, Iannaccone and Mortini11} This discrepancy may be attributed to the key role of demyelinating plaques in the pathophysiology. Because the pain in MS patients originates (at least in part) from these central plaques, which are located further along in the pain pathway than the trigeminal nerve (i.e., closer to the thalamus), it is not surprising that the extent of peripheral radiosurgical injury, reflected by sensory changes, may not be related to the efficacy of radiosurgery.

Safety

Our findings indicate that radiosurgery in MS patients with TN is a safe technique. The rate of hypesthesia after the first treatments was 15.7%, which is slightly higher than the 10% rate reported in the largest cohort study to date.^{Reference Xu, Mathieu and Heroux8,Reference Spina, Nocera, Boari, Iannaccone and Mortini11} After repeat treatments, hypesthesia was observed in 41.2% of patients. This increase is likely due to pre-existing nerve damage from the initial radiosurgical treatment, the cumulative effects of the natural progression of MS and the heightened susceptibility of MS patients to neurotoxicity.^{Reference Miller, Lachance and Lucchinetti32} Thankfully, radiosurgically induced facial numbness was mild and not bothersome in most cases, reinforcing the safety of the technique.

Other treatment-related complications included decreased corneal blink reflex (7.8%), dry eye/keratitis (2%) and transient masticatory weakness (2%), though these occurred infrequently. Surprisingly, decreased corneal blink reflex was, to our knowledge, never reported in previous studies,^{Reference Xu, Mathieu and Heroux8–Reference Alvarez-Pinzon, Wolf and Swedberg12,Reference Leduc, Mathieu, Adam, Ferreira and Iorio-Morin33} which may be due to it either not being systematically examined or not being classified as a complication because of its typically benign nature. It is also possible that clinicians may consider it merely a form of hypesthesia in the first trigeminal division rather than a separate ophthalmological issue. However, we believe that recognizing it as a distinct complication is important, as untreated corneal reflex deficits could potentially lead to more severe issues, such as dry eye and keratitis.^{Reference Hoşal, Örnek, Zilelioğlu and Elhan34} In our study, the incidence of such more severe ophthalmological complications (i.e., dry eye/keratitis) fell within the range reported in the literature (0%–4.1%).^{Reference Xu, Mathieu and Heroux8–Reference Alvarez-Pinzon, Wolf and Swedberg12,Reference Leduc, Mathieu, Adam, Ferreira and Iorio-Morin33,Reference Helis, McTyre and Munley35} Finally, one case of transient masticatory weakness was observed, a complication not documented in other studies of MS patients but noted in 1.5% of idiopathic TN cases .^{Reference Xu, Mathieu and Heroux8–Reference Alvarez-Pinzon, Wolf and Swedberg12,Reference Leduc, Mathieu, Adam, Ferreira and Iorio-Morin33,Reference Helis, McTyre and Munley35,Reference Mohammed, Hung, Eluvathingal Muttikkal, Bliley, Xu and Sheehan36}

Limitations and future implications

This retrospective study contains intrinsic limitations. First, because of its retrospective design, one can expect the influence of recall bias. Second, the radiological assessment was conducted by a single evaluator. Third, pertinent and potentially predictive neuroimaging data, such as the degree of vascular compression (displacement/distortion) and the size of demyelinating plaques, were not analyzed. Fourth, important radiosurgical data were not assessed in detail such as the variable target delineation associated with the non-isocentric CK technique. Fifth, using spherical volumes for CK treatments could limit the generalization of our results, as cylindrical volumes are more commonly used. Sixth, the relatively small sample size of this study limited the power of our analysis, prompting a large multicentric CK cohort study to provide more robust results.