Predicting the location of the preoptic and anterior hypothalamic region by visualizing the thermoregulatory center on fMRI in craniopharyngioma using cold and warm stimuli

Participant characteristics

From July 2019 to June 2020, 9 HVs and 32 patients with primary CPs participated in this study. The baseline HV characteristics are shown in Table 1. The average patient age was 27.5 (±2.1) years, ranging between 25 and 32 years. The detailed demographic and clinical characteristics of the 32 patients with CPs are summarized in Table 2. In total, 30 patients (93.8%) had vision impairment. Among them, 6 patients had visual acuity loss only, 7 patients had visual field defects only, and 17 patients had both visual acuity loss and visual field defects. Additionally, 21 patients (65.6%) had a headache, 20 patients (62.5%) suffered from endocrine defects, and 4 patients (12.5%) had nausea. All patients underwent surgery with the EEA for CPs in our institution. Overall, 26 patients (81.3%) were diagnosed with adamantinomatous CP, and 6 patients (18.7%) were diagnosed with papillary CP.

fMRI results of hypothalamic activity

In all HVs, the significant functional hypothalamic BOLD response area was located in the anterior hypothalamus, and both the cold and warm stimuli produced a significantly positive BOLD signal at the same location. Therefore, the region was speculated to be the POAH (Figure 1A). The time course of activation was similar for both warm and cold stimuli. No significant difference in the BOLD percentage signal change was found between the cold and warm stimuli (P=0.994) (Figure 1B). Based on the BOLD response of the activated position, the location of the POAH was confirmed.

In all patients with CPs, both the cold and warm stimuli also generated a significantly positive BOLD signal, and the significant single functional hypothalamic BOLD signal response around the tumor was distinguished (Figure 1C). Based on the location of the BOLD responses, the POAH was found to be pushed to the anterosuperior (n=5, 15.6%), posterosuperior (n=11, 34.4%), left (n=4, 12.5%), and right (n=12, 37.5%) areas of the tumor (Table 3). The time course of activation was also the same for both warm and cold stimuli. No significant difference in the BOLD percentage change was found between cold and warm stimuli (P=0.895) (Figure 1D).

We compared the beta scores of warm and cold stimuli between patients and HVs. The beta score is the average strength of the BOLD signal caused by stimulation for a period of time. The BOLD activity of the patients was found to be significantly higher than that of the HVs with both cold (P=0.03) and warm (P=0.03) stimuli (Figure 1E).

Surgical treatment

Different surgical strategies were applied based on the preoperative location of the POAH. The CPC approach was applied in 30 patients, and the EETLT approach combined with the CPC approach was applied in 2 patients. Among all 32 patients, gross total resection (GTR) of the tumor was achieved in 28 patients (87.5%), and near-total resection (NTR) was achieved in 4 patients (12.5%). No significant difference of the rate of GTR was identified among the 4 groups classified by the preoperative location of the POAH (left: 75% [3/4]; right: 91.7% [11/12]; posterosuperior: 90.9% [10/11]; anterosuperior: 80% [4/5]). In our group of cases, the complete pituitary stalk was preserved in 9 patients (28.1%), part of the pituitary stalk was preserved in 14 patients (43.8%), and the pituitary stalk was not preserved in 9 patients (28.1%).

Complications

Only one patient developed central hyperthermia on the first day after the operation (the highest body temperature was 39.5° C). During the follow-up period, new-onset diabetes insipidus occurred in 6 (18.8%) patients, and new-onset hypopituitarism occurred in 10 (31.3%) patients. Lower incidence rates of new-onset diabetes insipidus and hypopituitarism were found in this study compared to a case summary without the preoperative location of the POAH previously published by our institute (new-onset diabetes insipidus: 18.8% vs 26.4%; new-onset hypopituitarism: 31.3% vs 49.5%) [11].

Illustrative cases

Case 19

A 48-year-old male patient presented with a 3-month history of progressive blurring of vision and bilateral visual field defects. Preoperative task-related fMRI located the POAH to the left of the tumor. During the operation, the surgeon found that the tumor was not tightly adhered to the surrounding tissues and dissected the tumor via careful microdissection between the tumor and the POAH. Postoperative MRI showed that GTR was achieved. The patient did not present a significant increase in perioperative body temperature. The follow-up period was eight months, and the patient presented no new-onset diabetes insipidus or hypopituitarism (Figure 2).

Case 3

A 35-year-old female patient visited the outpatient clinic for a 6-month history of amenorrhea and a 3-month history of progressive blurring of vision and bilateral visual field defects. Preoperative task-related fMRI located the POAH to the right of the tumor. During the operation, the tumor adhered very tightly to the POAH, and the part of the tumor that adhered tightly to the hypothalamus was not forcibly removed. Postoperative coronal postcontrast T1-weighted MRI showed that the retained tumor closely adhered to the POAH. No significant increase in body temperature was observed during the perioperative period. The follow-up period was six months, and the patient presented no new-onset diabetes insipidus or hypopituitarism (Figure 3).

Case 12

A 57-year-old male patient presented with a 6-month history of visual impairment and headache. Preoperative task-related fMRI located the POAH to the posterosuperior of the tumor. Based on the POAH location identified before surgery, the surgeon removed the tumor through the CPC approach combined with the EETLT approach. Postoperative MR images showed that GTR was achieved. The patient did not present central hyperthermia, diabetes insipidus or hypopituitarism during the follow-up period (9 months) (Figure 4).

Participants

Nine healthy adult volunteers (HVs) participated in this study. The individuals with the following conditions were excluded: 1. a history of neurodegenerative, neurodevelopmental, or psychiatric diseases; 2. substance use disorders involving alcohol or heroin; 3. an inability to complete the MRI examinations; and 4. pre-processing issues (i.e., head motion).

A total of thirty-two patients with primary CPs who underwent surgical treatment in the Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, were enrolled in our study. The inclusion criteria for the patient group were as follows: 18-70 years of age; confirmation that the tumor was in the suprasellar and/or third ventricle regions; and histologically proven CPs. The exclusion criteria were as follows: 1. a history of stroke, cerebral trauma, brain surgery or brain radiotherapy, or other intracranial abnormalities (i.e., an arachnoid cyst); 2. an inability to complete the MRI examinations; and 3. pre-processing issues (i.e., head motion).

This study was approved by the Medical Ethics Committee of Beijing Tiantan Hospital. Written informed consent was obtained from the legal representatives of all the patients and all the HVs in accordance with the Declaration of Helsinki.

Experimental procedure

The hypothalamus is a small area in the brain involved in releasing hormones and regulating body temperature [23]. The POAH of the hypothalamus is the center of feedforward regulation of reflex thermoregulation [24], contains numerous thermosensitive neurons and receives feedforward thermosensory signals from cutaneous thermoreceptors [25]. Primary cold and warm cutaneous thermoreceptors synapse on somatosensory dorsal horn (DH) neurons in the spinal cord and send projections to the pontine lateral parabrachial nucleus (LPB) [26–28]. The cutaneous thermoreceptor signaling from the LPB to the POAH drives feedforward thermoregulatory responses (Figure 5) [29, 30].

An ice-water mixture at 0° C (cold stimulus) and a warm compress bag at 60° C (warm stimulus) were used to stimulate cold and warm cutaneous thermoreceptors. A balanced-design stimulus was administered during the fMRI scan by an independent person at the State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences, according to task prompts. Task-related fMRI included ten cold blocks (ice-water mixture at 0° C) and ten warm blocks (warm compress bag at 60° C) interleaved with rest blocks. The cold and warm blocks lasted 12 s, and the rest block lasted 8 s. A countdown prompted the operator to start and end the stimulus five seconds before the end of each block (Figure 6). The participant was in the supine position in the scanner in the 64-channel head coil.

MRI data acquisition and analysis

All subjects were scanned on a 3.0-T Siemens scanner (Siemens Healthineers, Erlange, Germany) at the Beijing MRI Center for Brain Research, Institute of Biophysics, Chinese Academy of Sciences, with a 64-channel head coil. The T1-weighted images were obtained with a 3D sagittal magnetization prepared rapid acquisition gradient echo (MPRAGE) sequence (resolution = 0.8 mm isotropic voxels; field of view = 256 × 240 mm²; slices = 224; flip angle = 8°; TI/TR/TE = 1000/2400/2.22 ms; data matrix= 320 × 300; bandwidth = 220 Hz/px; iPAT = 2). A gradient echo planar imaging sequence was used to acquire the task-related fMRI data covering the whole thalamus (resolution = 1.4 mm isotropic voxels; field of view = 192 × 192 mm2; slices = 50; flip angle = 90°, TR/TE = 2000/31 ms; iPAT = 2).

BOLD fMRI data were pre-processed using the Statistical Parametric Mapping toolbox (SPM12, http://www.fil.ion.ucl.ac.uk/spm) in the MATLAB environment (MATHWORKS, CA, United States). The pre-processing procedure included slice timing, motion correction, co-registering fMRI images to T1-weighted images, and smoothing. Then, a general linear model was used to analyze the activation strength of each stimulus.

In HVs, based on T1-weighted images, the activated functional BOLD response in the hypothalamus was used as the region of interest (ROI) to observe the functional POAH response to the cold and warm stimuli. In patients, the activated functional BOLD response around the tumor was used as the ROI to locate the POAH and observe the functional POAH response to the thermal stimuli.

Surgical strategies based on the location of the POAH

The endoscopic endonasal approach (EEA) was applied in all patients. Based on the location of the activated POAH, we divided the POAH position into four types according to the location of the POAH relative to the tumor: left, right, posterosuperior and anterosuperior. The tumor was first debulked to depress the tumor capsule. Once the tumor was debulked and decompressed, the surgeon dissected the extracapsular tumor away from the hypothalamus via careful microdissection based on the preoperative location of the POAH on fMRI. We sharply dissected the tumor from the hypothalamus under direct vision of the interface between the tumor and hypothalamus to avoid hypothalamic injury. For large tumors involving the third ventricle, if the POAH was located anterosuperior to the tumor, the endoscopic endonasal trans-lamina terminalis (EETLT) approach was not used to remove the tumor to protect the POAH, and if the POAH was not located anterosuperior to the tumor, the EETLT approach combined with or without the chiasm-pituitary corridor (CPC) approach could be considered for tumor removal.

Gross total resection (GTR) was defined as resection without visible remnant tumor according to an intraoperative assessment and postoperative MRI. The perioperative body temperature, extent of resection and endocrine outcomes were collected to evaluate the significance of this method.

Statistical analysis

Uncorrected P values <0.05 were deemed significant for all tests. The threshold was increased to distinguish the single activated area around the tumor to locate the POAH in patients with CPs. All fMRI results are reported as the percentage of BOLD signal changes relative to the BOLD signal at the beginning of each stimulus. For presentation purposes, data were pooled into "second" data points, and the start of each stimulus was considered the baseline BOLD response. Unpaired samples were analyzed using the Mann-Whitney U test or the Kruskal-Wallis test, while paired samples were assessed using the Wilcoxon t test. Differences correlating to an error of probability with P< 0.05 were considered statistically significant. Two-way ANOVA was used to determine whether significant differences existed between cold and warm stimuli and in BOLD values between the HVs and patients.