Prediction of tumor origin in cancers of unknown primary origin with cytology-based deep learning.

Cancer of unknown primary (CUP) site poses diagnostic challenges due to its elusive nature. Many cases of CUP manifest as pleural and peritoneal serous effusions. Leveraging cytological images from 57,220 cases at four tertiary hospitals, we developed a deep-learning method for tumor origin differentiation using cytological histology (TORCH) that can identify malignancy and predict tumor origin in both hydrothorax and ascites. We examined its performance on three internal (n = 12,799) and two external (n = 14,538) testing sets. In both internal and external testing sets, TORCH achieved area under the receiver operating curve values ranging from 0.953 to 0.991 for cancer diagnosis and 0.953 to 0.979 for tumor origin localization. TORCH accurately predicted primary tumor origins, with a top-1 accuracy of 82.6% and top-3 accuracy of 98.9%. Compared with results derived from pathologists, TORCH showed better prediction efficacy (1.677 versus 1.265, P < 0.001), enhancing junior pathologists' diagnostic scores significantly (1.326 versus 1.101, P < 0.001). Patients with CUP whose initial treatment protocol was concordant with TORCH-predicted origins had better overall survival than those who were administrated discordant treatment (27 versus 17 months, P = 0.006). Our study underscores the potential of TORCH as a valuable ancillary tool in clinical practice, although further validation in randomized trials is warranted.

Rational Ligand Design Enables pH Control over Aqueous Iron Magnetostructural Dynamics and Relaxometric Properties.

Complexes of Fe3+ engage in rich aqueous solution speciation chemistry in which discrete molecules can react with solvent water to form multinuclear µ-oxo and µ-hydroxide bridged species. Here we demonstrate how pH- and concentration-dependent equilibration between monomeric and µ-oxo-bridged dimeric Fe3+ complexes can be controlled through judicious ligand design. We purposed this chemistry to develop a first-in-class Fe3+-based MR imaging probe, Fe-PyCy2AI, that undergoes relaxivity change via pH-mediated control of monomer vs dimer speciation. The monomeric complex exists in a S = 5/2 configuration capable of inducing efficient T1-relaxation, whereas the antiferromagnetically coupled dimeric complex is a much weaker relaxation agent. The mechanisms underpinning the pH dependence on relaxivity were interrogated by using a combination of pH potentiometry, 1H and 17O relaxometry, electronic absorption spectroscopy, bulk magnetic susceptibility, electron paramagnetic resonance spectroscopy, and X-ray crystallography measurements. Taken together, the data demonstrate that PyCy2AI forms a ternary complex with high-spin Fe3+ and a rapidly exchanging water coligand, [Fe(PyCy2AI)(H2O)]+ (ML), which can deprotonate to form the high-spin complex [Fe(PyCy2AI)(OH)] (ML(OH)). Under titration conditions of 7 mM Fe complex, water coligand deprotonation occurs with an apparent pKa 6.46. Complex ML(OH) dimerizes to form the antiferromagnetically coupled dimeric complex [(Fe(PyCy2AI))2O] ((ML)2O) with an association constant (Ka) of 5.3 ± 2.2 mM-1. The relaxivity of the monomeric complexes are between 7- and 18-fold greater than the antiferromagnetically coupled dimer at applied field strengths ranging between 1.4 and 11.7 T. ML(OH) and (ML)2O interconvert rapidly within the pH 6.0-7.4 range that is relevant to human pathophysiology, resulting in substantial observed relaxivity change. Controlling Fe3+ µ-oxo bridging interactions through rational ligand design and in response to local chemical environment offers a robust mechanism for biochemically responsive MR signal modulation.