Psychosurgery was introduced in the early 1900s before the emergence of psychopharmacology. In 1936, Moniz published his experiences of prefrontal leukotomies for severe depression, anxiety, and aggression, performed by freehand injections of absolute alcohol into the frontal lobes, resulting in substantial symptom improvement in 14 of 20 patients [19]. He reported additional cases, received the Nobel Prize in Medicine in 1949, and coined the term “psychosurgery.” [20] However, at the time, the standards for reporting adverse events were loose, and Moniz did not report even a single adverse event [19], making it difficult to accept the results from a modern medical perspective. Freeman and Watts [21] described their prefrontal lobotomy with the transorbital approach, which is used to surgically interrupt the frontal lobe white matter tracts. Transorbital leukotomy became widely practiced as Freeman traveled across the United States. Before the development of psychopharmacology, prefrontal lobotomies were widely performed across the United States when skepticism about the treatment of mental illness was prevalent. In a review of 10,365 prefrontal lobotomy operations performed between 1943 and 1954, Tooth and Newton confirmed a 70% rate of improvement but also found a mortality rate of 6%, a rate of post procedure new-onset epilepsy of 1%, and post-procedure marked disinhibition in 1.5% of patients. However, the occurrence of unpredictable side effects has criticized the ethics and scientific basis of treatment. Movement against lobotomy was observed. In 1950, some countries, including the Soviet Union, Japan, and Germany, banned lobotomy [22]. In 1967, Freeman was banned from performing further lobotomies after one of his patients experienced a fatal brain hemorrhage.

In the late 1940s, the advent of frame-based stereotactic procedures enabled the development of more precise and reliable neurosurgical lesions [23]. These procedures have become significantly used since the 1970s, allowing surgeons to accurately target specific brain areas using three-dimensional imaging and computer guidance [24]. The primary goal was to physically alter the brain regions and circuits associated with OCD by creating lesions. Lesion procedures for treatmentresistant OCD mainly target four areas: anterior capsulotomy (ALIC), cingulotomy (anterior cingulate cortex and cingulum fibers), subcaudate tractotomy (frontothalamic fibers), and limbic leukotomy (a combination of anterior cingulotomy and subcaudate tractotomy) [4,25]. The effects and side effects at each surgical site differed slightly; however, overall, they were similar [12], suggesting that these surgical sites are brain structures associated with the same CSCT circuit [26]. Owing to the successful use of DBS in the ALIC in the treatment of OCD, the internal capsule has recently become a popular target.

Capsulotomy for OCD during this period employed two surgical techniques: radiofrequency (RF) and radio-surgical ablation [27]. RF capsulotomy involves a standard stereotactic technique in which RF electrodes are introduced into the ventral ALIC for thermal ablation using fixed parameters [28,29]. Advantages include short operative time and a relatively uncomplicated procedure, in addition to low procedural cost; however, although stereotaxic techniques have been employed, RF ablation still carries unavoidable risks associated with craniotomy, including intracranial hemorrhage (ICH), damage to surrounding brain tissue, or procedural infections. Additionally, it cannot be confirmed whether the lesion has been created successfully during the procedure.

Radiosurgical ablation has been developed as a noninvasive alternative to RF capsulotomy. The first radiosurgical capsulotomy was performed in 1976 for treatment-resistant OCD and other anxiety disorders and has since been conducted widely [30]. In this study, lesions were created by directing radiation from a Cobalt-60 source to specific brain sites, guided by MR and computed tomography images. The main advantage of this procedure is its less invasive nature, as it does not require opening of the skull or causes minimal damage to the surrounding brain tissue, thus reducing immediate postoperative adverse events [24]. A nonrandomized comparison of 25 patients with treatment-refractory OCD at the Karolinska Institute who underwent either RF or radiosurgical capsulotomy between 1988 and 2000 suggested that RF capsulotomy is associated with greater adverse effects than is radiosurgical capsulotomy [28]. However, radiosurgical procedures require a longer time for the lesion to form, making it difficult to confirm the results after the procedure and require a longer time for symptom improvement. There were cases where reoperation was needed due to insufficient lesion formation. Additionally, potentially serious adverse events such as radiationinduced necrosis and cyst formation can be associated with radiation. A detailed summary of the history and clinical experience of radiosurgical capsulotomy has been provided in another review [24].

Recently, more precise and less invasive neuroablative surgical techique with MRI guidance have developed. In MRIguided neuroablation, MRI was performed to confirm the creation of lesions at the appropriate location and size. MRguided-LITT is a minimally invasive ablative technique. MRguided-LITT involves the insertion of an optical wire through a burr hole into the target structure and activating laser illumination at the tip to heat and ablate the tissue [31]. Compared to RF thermal lesions, LITT has the advantage of being able to clearly see the target, and obtain a lesion estimate through thermography. Currently, LITT is most widely applied to epilepsy and brain tumors. After the first case report in 2016 [32], two case series involving 18 patients have been reported (Table 1) [33-36].

MRgFUS uses multiple acoustic energy sources within a specially designed helmet to create a lesion under real-time MRI guidance (Figure 1) [37]. By contrast with LITT, in MRgFUS craniotomy is not required. It also allows for ongoing patient condition monitoring and early identification of intra-operative side effects [38]. The results of the two prospective clinical trials have been reported [35,36]. MRgFUS has been widely applied for the treatment of various movement disorders (essential tremors, etc.). Both techniques have shown an efficacy comparable to that of traditional surgical methods but with their minimally invasive nature and real-time monitoring capabilities. In a cost-effectiveness analysis that considered effectiveness, side effects, and follow-up treatment process, MRgFUS was reported to be more advantageous than RF capsulotomy [39].

Throughout the past few decades, various studies have employed diverse selection criteria when selecting OCD patients for neurosurgical interventions. In recent capsulotomy clinical trials, symptom severity, chronicity, and treatment resistance have been used as criteria for patient selection (Table 2) [35,41,47-52]. Treatment resistance is typically diagnosed when a patient does not respond to a certain number of psychopharmacological and psychotherapeutic treatments, including electroconvulsive therapy [2]. However, this criterion related to treatment resistance poses a complex challenge, given the constant evolution of medical and non-invasive treatment methods. For example, recent guidelines recommend non-invasive neuromodulation techniques, including deep repetitive transcranial magnetic stimulation (deep rTMS), for the treatment of OCD [2]. Therefore, before considering neurosurgical interventions, it is important to refer or consult with a specialist, such as an OCD expert team or psychiatrist experienced in treating refractory OCD. These specialists can provide a comprehensive evaluation, suggest alternative treatment options, or recommend adjunctive therapies to improve outcomes. Additional general criteria for the presurgical evaluation of psychiatric patients, as well as ethical considerations, were described in a multinational consensus statement [53].

Adherence to treatment is another important issue. It is crucial to assess if patients have adhered to their psychiatrist’s treatment plan correctly and engaged actively in cognitive-behavioral therapy. Certain individuals may sometimes envision that neurosurgical interventions will miraculously alleviate their symptoms, regardless of their personal efforts [54]. Since maintaining medical and behavioral treatment after neurosurgical intervention is required until the optimal surgical effect appears, evaluate adherence to treatment is required not only defining the treatment resistant, but also deciding psychosurgery.

Establishing a scientific and grounded plan to evaluate treatment effectiveness is essential for the development of modern neuroablation. Older literature lacked objective outcome measures, necessitating the use of replicable tools for assessing effectiveness [42]. Recent clinical trials use tools that can evaluate not only the symptoms OCD, such as Y-BOCS [55], but also overall clinical improvement (clinical global improvement) to confirm treatment response (Table 2). Nevertheless, the scales mentioned may be insufficient in capturing the life changes experienced by patients who undergo neurosurgery.

According to a recent qualitative study of capsulotomy experiences [54], factors such as quality of life, personal values, and social functioning are important in the decision-making process and assessment of treatment effects. In particular, for individuals with OCD, even small improvements in symptoms can lead to significant enhancements in their perceived quality of life owing to the substantial subjective distress caused by the symptoms [54,56]. Conversely, there have been reports of significant improvement according to Y-BOCS scores, while the quality of life remained unchanged [57]. Therefore, when assessing treatment effects, it may be necessary to use various scales that reflect mood, impulsivity, suicidality, cognitive function, and so on [9,58]. The same considerations apply to the observation of adverse effects.

The timing of assessing these effects also requires further discussion. In the case of gamma capsulotomy, lesions gradually appear, and improvements in OCD symptoms are reported to occur slowly [24]. However, in studies such as that of Kim et al. [35] involving MRgFUS found that symptom improvement could take up to two years, indicating delayed progress. Moreover, studies observing electrophysiological changes have shown that alterations in electrophysiological connectivity occur gradually over a span of six months [59]. Considering these findings, the evaluation of surgical outcomes should be conducted after a minimum period of at least 12 months following surgery.

The development of an MR-guided procedure and diffusion tensor imaging (DTI) imaging enables a more precise and planned execution of capsulotomy. ALIC is a particularly salient structure for psychiatric surgery because it is a point of convergence for key white matter fibers connecting the prefrontal and anterior cingulate cortices with the hippocampus, amygdala, and thalamus, which together form a core limbic network in the human brain [60]. Therefore, when performing a capsulotomy, the effects and side effects can vary depending on the fiber tract affected by the lesion. Primate fiber-tracking studies and human DTI techniques have revealed how fiber tracts originating from the prefrontal cortex are positioned within the ALIC [61,62]. Additionally, the ventral portion of the ALIC is related to responder status, suggesting that lesions in this region are more likely to produce a clinical response [63]. However, recent research results on treatment location and outcomes are not the same [64], which can be attributed to two factors. First, the fiber tract in the ALIC can exhibit inter-individual variability in neural fiber pathways. Second, each patient may have different abnormalities in neural circuits [65]. Further studies on neurocircuit-based patient selection [10] and use of tractography to guide patient-specific anterior capsulotomy for OCD are needed [66].