The investigators aim to develop an advanced imaging platform, such as dynamic nuclear polarization (DNP) 13C-MRI, MR fingerprinting (MRF) and MR Relaxometry, which combines with traditional anatomical contrast CT, MRI and PET, and integrate blood/urine metabolomics methods. A comprehensive strategy to thoroughly analyze the immune activation of spleen pattern, microstructure, cell density, red blood cell iron content, immune cell glycolysis and metabolic flow rate.

The investigators plan a 3-year trial with a randomized, two-groups allocation observational study design that will enroll 90 subjects with newly diagnosed or recurrent gynecological cancer from Linkou Chang Gung Memorial Hospital (CGMH) receiving ICI therapy for this prospective study. Standard cares, eg. MRI/CT/PET, are priorly required for the first-line screening method, with tumor slides and routine blood work at the initial visit. After being randomly assigned, 30 eligible subjects will be assigned to the next-generation imaging group, and will receive two integrated examinations between the baseline and 2 weeks of immunotherapy-the integrated examination of DNP and metabolomics.

Subjects in the next-generation imaging group will receive splenic DNP imaging, followed by non-contrast-enhancing MRF at the lesion and MR Relaxometry, which provide quantitative measurements of metabolism occurring within the spleen and within the cancer cells, respectively. DNP is injected with 13C contrast agent hyperpolarized 13C pyruvate (HP [1-13C] Pyruvate) intravenously in the arm, and the study drug is injected at a dose of 0.43 ml/kg, which is labeled with the non-radioactive isotope 13C at the C1 position, hyperpolarized 13C signals are obtained through MR spectrum acquisition, and then MRF sequences are performed to obtain multi-parameter tissue characteristics. The general imaging group will receive the same MR strategies but without spleen DNP scan.

All MRI sequences will be performed on a 3T clinical scanner (Discovery MR750w,GE Healthcare, Milwaukee, USA) with a flexible surface 13C/1H multinuclear transmit-receive coil (RAPID Biomedical, Bavaria, Germany) to cover the splenic region.

DNP-MRI:

To comply with the Good Manufacturing Practice (GMP) of pharmaceuticals, [1-13C]pyruvate for human will be prepared in the ISO 8 clean room and ISO 5 laminar flow in the Department of Nuclear Medicine, Chang Gung Memorial Hospital at Linkou, Taiwan. The premixed [1-13C]pyruvic acid/12.5 mM Electronic Paramagnetic Agent (Electronic Paramagnetic Agent; EPA; AH111501, GE Healthcare, Oslo, Norway) will be packed into a sterile pharmacy Kit (Sterile Fluid Path; SFP, GE Research Circle Technology) and polarized in the polarizer (SPINlab, GE Research Circle Technology) at 0.7 - 0.8 K and a magnetic field of 5 T for at least 2 - 3 hrs. After being quickly dissolved in heated water, it will pass through an EPA filter to remove excess EPA and be mixed with neutralization medium (NaOH/TRIS/EDTA) in the receiver, and then automatically detected by optical measurement in the QC machine (GE GE Research Circle Technology), including pyruvate concentration, EPA concentration, pH, temperature, volume, and polarization level, while simultaneously drawing a solution containing hyperpolarized [1-13C]pyruvate through a 0.2 μm terminal filter (ZenPure, Manassas, Virginia) into an administration syringe (Medrad, Warrendale, Pennsylvania) to ensure sterility. After the responsible physician checks that the QC parameters meet the criteria, the patient will be injected intravenously at a dose of 0.43 mL/Kg at a flow rate of 5 mL/s, and then flushed with 20 mL of physiological saline. Terminal filter integrity (Threshold = 50 psi) will be checked immediately after dispensing.

During 13C-MRI scanning, a set of image detections will be performed first for positioning, and then the 13C signal will be calibrated, including gradient shimming, 13C center frequency confirmation (calculated from the 1H center frequency) and emission gain optimization (using the Bloch-Sieger sequence ), after a 30-second delay from injection, the following rapid 13C sequence will be performed, which will take approximately 2-3 minutes:

1. Pulse acquisition of 13C spectra: flip angle = 10°; TR = 2000 ms; slice thickness = 30 mm; read bandwidth = 5000 Hz; spectral collection points 2048; repetition = 14; acquisition time = 2000 ms; time-resolved Rate = 2000 ms.
2. 13C metabolite-specific imaging of pyruvate, lactate, and alanine (spectral-spatial RF pulses followed by helical readout): TR = 250 ms; TE = N/A; in-plane spatial resolution = 10 x 10 mm^2; matrix size = 128 x 128 mm^2; field of view = 22 x 22 cm^2; slice thickness = 20 mm; flip angle = 15°/30°/30° (pyruvate/lactate/alanine); acquisition time = 250 ms; time resolution = 1000 ms.
3. 13C chemical shift imaging: TR = 80ms; TE = N/A; in-plane spatial resolution = 25 x 25 mm^2; matrix size = 8 x 8 mm^2; field of view = 20 x 20 cm^2; slice thickness = 20 mm ; Flip angle = 15°; Acquisition time = 80 ms.

DWI:

The clinical MR study will continue after 13C DNP with the original scanner and will use single-echo planar technique with fat suppression (repetition time ms/echo time ms: 3300/79; number of signal averages: 4; slice thickness: 4 mm ; gap; 1 mm; matrix: 128x128 mm^2; field of view 20 x 20 cm^2). The highest b-value of 1000 sec/mm^2 has been chosen to optimize the signal-to-noise ratio. Diffusion-weighted gradients are applied orthogonally for slice selection, phase encoding and readout direction, and the slope will be calculated from the logarithmic decay curve of signal intensity versus b-value (Syngo, Siemens, Erlangen, Germany).

MR Fingerprinting (MRF):

MAGiC (MAGnetic resonance image Compilation) will use 2D fast spin echo based multi-saturation delayed multi-echo acquisition. Steady-state free precession (SSFP) is used for MRF acquisition with acquisition traces interleaved using 89 under sampled golden angle helices, sampling bandwidth = ±250 kHz, TR = 9 ms, TE = 2.2 ms, NEX = 1 and 979 frames . B0 and B1 are not included in the sequence but slice outlines are included to improve T2 accuracy, other imaging parameters used in MAGiC and MRF scans are: FOV = 22 x 22 mm^2; matrix = 256 x 256; slice thickness = 4 mm, 1 mm gap; 20 pieces. Time intervals of scan are approximately 4 minutes and 3.6 minutes, respectively.

MR Relaxometry:

The iterative decomposition of water and fat with an echo asymmetry at least-square estimation-iron quantification (IDEAL-IQ) sequence has the following parameters: TR 10 ms; TE 4.7 ms; echo number 6, ranging from 1.1 ms to 6.38 ms FOV 35-40 cm; matrix size 128 × 128; pixel bandwidth 325 Hz; flip angle 6; slice thickness 10 mm; space between slices 5 mm. The scans are acquired during a breath-hold lasting less than 30 seconds. R2* is calculated using the following formula: R2*IDEAL = -1.276 + 1.2366 * R2*GRE. R2* is then converted to iron concentration (mg/g) using the widely used formula proposed by Wood et al.: Iron concentration = 0.202 + 0.0254 x R2*

While there is no accepted normal range for spleen iron, 1.98 mg/g is suggested as an upper limit, with 2.74 mg/g reported to be pathological. Splenic iron varies by age and sex. Spleen iron content is higher in men than in women (men: 0.96 ± 0.34 mg/g, women: 0.87 ± 0.29 mg/g). In women, menopause is associated with 0.12 mg/g higher spleen iron. It must be noted that the GRE echo time of abdominal MRI is insufficient to quantify iron when iron exceeds the threshold (approximately 4 mg/g)

Metabolomics:

Plasma, urine and/or tissue samples collected from patients will be further analyzed by 1H high-resolution NMR (Bruker 600MHz NMR spectrometer) and LC-TOFMS (Waters Q-TOFMS) located at the Clinical Metabolomics Center, Chang Gung Memorial Hospital. Standard metabolomics software - GeneSpring MS, Markerlyn XS and Metaboanalyst, will also be used for multivariate data analysis and presentation. The METLIN, HMDB, and KEGG databases will be searched for an accurate set of features showing significant differences between the groups of before and after immunotherapies. Compound predictions will be performed using metabolite databases and formula generation software. Non-commercial and commercial applications such as MetaboAnalyst for metabolomics studies are also available. Laboratory examinations include routine inflammatory indices, such as erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), procalcitonin, plasma antioxidant index (TAC), cellular antioxidant index (GPX), DNA damage index (8-OHdG ) and lipid oxidation and inflammation index (MPO); conventional metabolic indexes, such as LDH, lactate and pyruvate; and immune function-related indexes, such as T Cell & B Cell, T Cell Subset tests, HLA-B27, IL-6, a1-globulin and a2-globulin. If the patient has undergone slides for histopathological diagnosis, tissue LDH and MCT protein expression will also be determined.

Inclusion Criteria:

1. Newly diagnosed or recurrent gynecological cancer confirmed by histology.
2. Age ≥ 20 years old.
3. Expected to receive immunotherapy.

Exclusion Criteria:

1. Suffering from other malignancies (excluding non-melanoma skin cancer).
2. History of splenic abnormalities (such as splenic damage, cirrhosis-related splenomegaly or primary/metastatic splenic tumors).
3. Insufficient function of bone marrow, liver and kidney.
4. Contraindications to MRI studies (e.g. claustrophobia, cardiac pacemaker, metal implants in the pelvis).
5. Uncontrolled concurrent diseases, including but not limited to kidney stones, persistent or active infection, symptomatic heart failure, unstable angina, cardiac arrhythmias, or mental illness/social situations that limit compliance with research requirements.
6. Pregnant or lactating women.

Gynecological cancer