After the approval of institutional review board (IRB), the authors conducted a retrospective analysis of all patients with anterior circulation aneurysms (n=1,068) at our hospital between January 2015 and December 2018. All aneurysms were first treated with endovascular coiling if possible (n=543). Thus, the remaining aneurysms not suitable for coiling were subjected to surgical clipping (n=525). During the study period, surgical clipping of most anterior circulation aneurysms was performed with MPT. Although MPT was launched in January 2014, the cases of first year were not included in the population considering the skill and familiarity of operator. Patients who underwent wide craniotomy or craniectomy due to RIA with large volume ICH (n=23), performed parasagittal craniotomy for pericallosal aneurysm (n=18), had undergone a previous craniotomy (n=11), had a history of facial palsy (n=0), or failed to get a follow-up radiological examination and appointment at outpatient department (OPD) for over 6 months (n=15) were excluded from the study population. Therefore, 458 patients with 562 aneurysms were included in the MPT group.

We also conducted a retrospective review of patients between January 2012 and December 2013 to compare the outcomes of MPT and CPT (n=274). Population criteria for CPT were the same as those for MPT. Thus, 170 patients with 198 aneurysms were included in the CPT group. All surgical procedures in the two groups were performed by one senior neurosurgeon (C.Y. Lee). Both UIA (MPT: 313, CPT: 106) and RIA (MPT: 145, CPT: 64) were included in our study. For the UIAs, transfemoral cerebral angiography (TFCA) was performed to evaluate the number, location, size and shape of aneurysms. Scheduled surgery was performed based on the patient’s age and medical history after a preoperative risk assessment. For the RIAs, emergency surgery was performed within 1 day of the hospital visit in all cases after taking a computed tomographic angiography (CTA).

A curvilinear scalp incision was made from the height of lateral epicanthus to the ipsilateral midpupillary line. The skin incision was made just posterior to hairline, and it was located anterior to incision in CPT. the incision line was made such that the scar would be completely covered by patient’s hair after recovery. Thus, the length of the MPT incision was shorter than that of the CPT incision, especially in patients with a smaller head and narrower forehead. The patient’s head was only shaved up to 2–3 cm posterior to the incision line, in order to reduce the patient’s discomfort (Fig. 1).

Infero-lateral to superior temporal line (STL), the scalp flap was reflected through performing subfascial dissection between the deep temporal fascia and temporalis muscle, in order to minimize exposure of the superficial temporal fat pad, which contains the frontotemporal branch of the facial nerve. Supero-medial to STL, the galea aponeurotica, loose areolar tissue, and pericranium were left to the skull so that re-approximation could be performed by suturing with the temporalis fascia at the end of the surgery. An incision was made in the temporalis fascia along the STL, and a subperiosteal dissection of the temporal muscle was performed to expose the pterion using a dura elevator, rather than monopolar cautery.

The upper border of the craniotomy was within the inferior range of the STL. Craniotomy was performed in a downward curve towards the squamous temporal bone using the pterion as a distal landmark in MPT, unlike CPT, where the pterion is a central landmark. Therefore, MPT included the anterior part of the frontal bone inferior to the STL, the lateral part of the sphenoid bone, and only a minimal portion of the squamous temporal bone. The size of MPT was much reduced to 2.5×2 cm or 3×2 cm. Burr-hole was performed only once on the superoposterior side of the bone flap to minimize the temporal hollowing. The outer table of skull was preserved to the maximal extent during the sphenoid ridge drilling. These processes of craniotomy were performed only beneath the temporal muscle (Fig. 2).

When a semilunar dural incision was made on the base of the flattened sphenoid ridge, part of the inferior frontal gyrus, the anterior ascendant ramus of the Sylvian fissure, and part of the superior temporal gyrus were exposed (Fig. 3). Sufficient microsurgical field could be obtained by cerebrospinal fluid (CSF) drainage using an arachnoid dissection and frontal or temporal retraction using a Greenberg retractor (Fig. 4). In the RIA, an intraoperative ventriculostomy was performed at Paine’s point (the point of an isosceles triangle with the Sylvian fissure as the base and a side-length of 2.5 cm) to alleviate brain swelling (Fig. 5).

After aneurysmal clipping has been completed, the bone flap was fixed in anatomical alignment using cranial plate hardware consisting of a burr-hole cover and two straight bars, autologous bone chips and hydroxyapatite cements to fill the gap (Fig. 6). The incised temporalis fascia and the galea aponeurotica and pericranium, which had been left on the superomedial side of the STL, were reapproximated as tightly as possible using Vicryl absorbable suture. And then the craniotomy scar was completely concealed by the temporal muscle (Fig. 7).

To assess the feasibility of MPT in anterior circulation aneurysms, the sex, age, presence of hemorrhage, and operation time were compared with CPT. The size, location and direction of aneurysm, and the relationship between skull and aneurysm were examined by analyzing preoperative CTA and TFCA. Additionally, the applicability of MPT in the case of subarachnoid hemorrhage (SAH) due to RIA was investigated by examining Hunt and Hess (H-H) grade and Fisher’s scale.

To examine the microsurgical safety of MPT and CPT, postoperative radiologic data and medical records were analyzed. We confirmed the presence of postoperative ICH or ischemic lesions, and epidural (EDH) or subdural hematoma (SDH) on postoperative immediate and 8th day computed tomography (CT). In the RIA, the presence of vasospasm or delayed hydrocephalus was verified, and the presence of aneurysmal remnants was confirmed through TFCA. we checked the presence of surgical wound infection, cerebrospinal fluid (CSF) leakage, meningitis and brain abscess. Permanent neurological deficits were also evaluated with the modified Rankin scale (mRS) score.

To assess the functional and cosmetic efficiency of MPT and CPT, we analyzed temporomandibular dysfunction, facial nerve paralysis, facial asymmetry, and surgical aesthetic results. A survey research was carried out in all patients at about 6 months after surgery, and the questionnaire included the following 11 items. In order to verify postoperative temporomandibular joint dysfunction, the survey evaluated time durations for postoperative pain to disappear, progression to chronic pain, masticatory disability, and discomfort of maximal mouth opening. To identify the occurrence of facial nerve paralysis, we investigated drooping of the upper eyelid, duration of ptosis, progression to permanent ptosis, and unilateral facial weakness. In addition, the change of facial sense due to skin incision, and facial asymmetry due to temporal muscle atrophy were examined. The subjective satisfaction of patients was scored of 0–10 based on a combination of the above questions.

SPSS 20.0 for Windows (SPSS Inc., Chicago, IL, USA) was used for the statistical analyses in this study. The clinical outcomes and complications between MPT and CPT were respectively analyzed by examining several variables such as postoperative hemorrhage or infarction, surgical infection, vasospasm and hydrocephalus. The functional outcomes of MPT and CPT were based on a survey research (including 11 questionnaires) of temporomandibular dysfunction, facial nerve paralysis, facial asymmetry, and patient’s satisfaction in all patients. The Fisher’s exact test and the independent t-test were performed for the statistical significance between two groups. Results with p-values less than 0.05 were considered statistically significant.

We made effort to improve the MPT for aneurysmal clipping by modifying the surgical techniques such as decreasing the length of skin incision, preserving the temporal muscle, and reducing the craniotomy size. First, the shortened incision line was able to reduce skin numbness at the surgical site by about 7.4%. Of course, the length could be shortened further if a straight skin incision were made at the craniotomy location in the pterional area. Some studies have attempted this,16) but because the tiny scar over this area is on an exposed part of the face, it can be more stressful for the patient. Moreover, there is a risk of injury to the frontal branch of facial nerve, and if a problem arises, it could make conversion to CPT more challenging. Therefore, we decided to make the incision just posterior to the hairline so that the scar would be completely covered once the hair grows.

Second, we performed subfascial dissection between the deep temporal fascia and temporalis muscle, which can secure a broader space and prevent facial palsy. The integrity of the muscle fiber was maintained by making an incision in the temporalis fascia and minimizing the use of monopolar cautery. In addition, adequate tension was applied during closure by tightly reattaching the muscle to the galea aponeurotica and the pericranium, which were left on the superomedial side of the STL. Myocutaneous flap, another method to reflect the scalp flap and expose the skull, can completely avoid the exposure of facial nerve, but this requires a long incision of temporal muscle,19)23)30) Above all, it is unsuitable for ACoA aneurysms located in deep positions, since the bulky flap obscures the microsurgical field in MPT.

Third, the size of bone flap was considerably reduced as far as possible to minimize the exposure of cortical areas, which was irrelevant to the surgery. Since the range of surgical field was smaller and the time duration of surgery was reduced by approximately 55.8 minutes, we consider it can be a cause of lower incidence of surgical wound infection. Owing to the reduced physical burden of the surgeon, 18 patients were able to undergo bilateral MPT under one anesthesia. Above all, although the smaller operative window in MPT may seem restrictive, it was also well preserved like the surgical corridor of CPT. Figueiredo et al. have reported that when the pterion is used as a distal landmark for MPT, dissection is performed only up to the anterior ascendant ramus of the Sylvian fissure.9) Any additional opening provides no advantage for inspecting the basal cistern and the circle of Willis, and extensive dissection using the CPT method can rather promote chronic SDH. In other words, the authors inferred that MPT could be applied broadly to anterior circulation aneurysms.

Several reports with MPT for anterior circulation aneurysms included relatively small volume less than 50 cases,8)22)25)27) and we think that the evaluation of clinical and functional results maybe not fully investigated.2)6)11)16) By the way, Rychen et al. reported a systemic review of the literatures on minimally invasive alternative approaches to CPT,21) and they collectively reviewed approach-related complications in a total of 22 MPT articles. Postoperative hematoma and infarction occurred in 2.2% and 3.1% of all patients, and wound infection and CSF leak was reported in 1.2% and 0.3%. Painful mastication occurred in 7.5% of all patients, and permanent facial palsy and temporal hollowing was reported in 0.4% and 0.9%, respectively. The present study showed similar or slightly better results than this systemic review.

However, the incidence of facial palsy-induced ptosis, a longstanding problem of CPT, was somewhat high (7.0%) compared to other functional problems in our study. It has been reported that its incidence could be up to 30% in CPT.3) Considering that the ptosis was temporary and disappeared without any permanent remnants, it was estimated that excessive tension during temporalis muscle retraction was the cause of ptosis. As 1.6% of patients felt facial asymmetry even in MPT, this method could not essentially avoid the bony degeneration and muscle atrophy, so it might be difficult to completely prevent the temporal depression. Therefore, a small amount of hydroxyapatite cement was particularly used for cranial reconstruction, only in cases of women in whom the temporalis muscle and pterional portion of skull were thin. It is possible that excessive frontal retraction or rectus gyrus excision in the ACoA aneurysm located in the deep position may cause anosmia dysfunction,4) unfortunately, but we did not perform an evaluation of postoperative olfactory function. Our study has also some limitations in that it is a retrospective analysis that MTP and CPT were performed in different period, respectively. In addition, since our population was selected based on patients who had OPD follow-up for more than 6 months, we think it might be somewhat short to assess the facial asymmetry. Maybe, long term and large prospective study could be needed in future.

We minimized the injury of temporal muscle and the size of craniotomy for improving the pitfalls of CPT, and in our knowledge, we collected large population among several studies on MPT for aneurysmal clipping at a single institute. The present study showed that overall functional outcomes of MPT were superior to those of CPT, while the clinical outcomes of MPT including postoperative complications were similar to those of CPT. Since no perfect surgical technique exists that can be applied to all cases, it could be necessary that surgeons would know a variety of approaches. However, this requires a long time and considerable effort. Given that MPT is only a slight modification of CPT, which is the most familiar approach for neurosurgeons, it has the potential to become a safe and efficient alternative for surgical clipping in anterior circulation aneurysms.