The week in imaging: five advances that change management
The last week of July packed an unusual run of protocol-changing news: two FDA decisions, two diagnostic accuracy studies and one national dose survey. The common thread is imaging shifting from descriptive to decisive — increasingly it is what determines who is eligible for treatment. Here are the five items from Diagnostic Imaging’s weekly roundup, with the technical context that does not fit in a headline.
Pluvicto moves into hormone-sensitive disease
The FDA expanded the approval of lutetium Lu 177 vipivotide tetraxetan (Pluvicto, Novartis) for use in combination with an androgen receptor pathway inhibitor (ARPI) in PSMA-positive metastatic androgen pathway modulation-naive/sensitive prostate cancer — the setting clinicians know as hormone-sensitive disease. The decision came on July 31, 2026.
The acronym soup deserves translation. Pluvicto is a radioligand: a molecule that binds prostate-specific membrane antigen (PSMA), densely expressed on the surface of prostate tumor cells, carrying lutetium-177, a short-range beta emitter. The radiation dose is delivered cell by cell, wherever the target sits. ARPIs are the abiraterone, enzalutamide, apalutamide and darolutamide class.
The basis is the phase III PSMAddition trial, which showed a 28% reduction in the risk of progression or death with the combination versus standard of care, with an updated analysis raising that to 33% alongside a favorable overall survival trend. The point that matters most to imaging departments is different: by moving the radioligand into an earlier line, the approval may nearly double the number of eligible patients — and every one of them needs a PSMA PET/CT to document positivity before treatment.
The regulatory decision is oncologic, but the demand lands in nuclear medicine. It is the same structural pull behind the race for production and distribution capacity, visible in the multibillion-dollar Curium–Lantheus negotiation.
CaRi-Heart: FDA authorizes coronary inflammation measurement on CCTA
On July 29, the FDA granted De Novo classification to CaRi-Heart, from Caristo Diagnostics — the first technology authorized in the United States to quantify coronary inflammation from a routine coronary CT angiography (CCTA) exam.
The principle is elegant. Arterial wall inflammation alters the composition of the perivascular fat surrounding the vessel, and that change translates into a measurable attenuation shift on CT. The tool reads that signal and returns, alongside the conventional report, a personalized 10-year cardiovascular mortality risk estimate.
The clinically decisive detail: the method flags elevated risk even in patients with low or zero coronary artery calcium (CAC) scoring. That is precisely the subgroup that currently slips through stratification — inflamed, non-calcified plaque does not register on a calcium score, and the patient is cleared as low risk. Paired with the already-cleared CaRi-Plaque, the combination delivers inflammation and quantitative plaque assessment from a single acquisition. U.S. commercial launch is expected in the fourth quarter of 2026.
Radiomics predicts pleural invasion in lung adenocarcinoma
A combined radiomic and deep learning model applied to intratumoral CT threshold segmentation reached an area under the curve (AUC) of 82.3% and sensitivity of 95.4% for predicting visceral pleural invasion in lung adenocarcinomas smaller than 30 mm — in an external test set, not just the training cohort.
Why it matters: visceral pleural invasion is an upstaging factor in the TNM system. A small nodule that invades the pleura is no longer treated as minimal disease. That information directly shapes the choice between sublobar resection and lobectomy, and the case for adjuvant therapy. Today it usually arrives from pathology, after surgery. Moving it to the preoperative setting with 95.4% sensitivity changes the surgical planning conversation.
The usual methodological caveat applies: external validation is a good sign, but threshold segmentation is sensitive to acquisition protocol and reconstruction kernel. Cross-vendor reproducibility remains radiomics’ practical bottleneck.
Biparametric MRI matches multiparametric for extraprostatic extension
A comparative study found that biparametric MRI (bpMRI) offers detection comparable to multiparametric MRI (mpMRI) for assessing extraprostatic extension in prostate cancer.
The difference between the two is contrast. Multiparametric combines T2, diffusion-weighted imaging and a dynamic gadolinium-enhanced sequence; biparametric drops the dynamic phase and, with it, the IV access, the contrast cost and a good chunk of table time. If local staging performance holds, the throughput gain is substantial — in a fully booked service, shortening a prostate protocol by 10 to 15 minutes means more patients per day without buying a scanner. It fits the broader trend of MRI taking over local staging, which we covered when reporting that MRI could replace the digital rectal exam in prostate staging.
CT dose: the national picture that resets reference levels
Closing the week, the national CT dose survey published in Radiology: compared with 2014 benchmarks, CTDIvol achievable doses fell 9.3% across 10 common adult exam categories, while diagnostic reference levels dropped 21.8%. The dataset is robust — 5,234,285 exams in 2,802,416 adults during 2025.
We covered that study in depth when it landed, including variation by exam type and what changes in each department’s benchmarks: see our analysis of the decade-long drop in CT radiation dose.
The thread connecting all five
Step back and a pattern appears. In four of the five cases, imaging is not merely describing anatomy — it is selecting who gets treated, quantifying risk invisible to the previous method, or delivering upfront what only pathology used to provide. PSMA PET becomes the gateway to radioligand therapy; CCTA becomes an inflammatory stratification tool; chest CT starts estimating pleural invasion; MRI takes over local staging.
For service managers that carries a concrete budget consequence. Demand grows for PET/CT and for post-processing capacity, and so does the requirement for protocol standardization — quantitative models only hold up if acquisition is reproducible. For radiologists, the message is that structured reporting stops being an aesthetic preference and becomes an input to treatment eligibility.
As always, the limits deserve saying out loud: regulatory authorization is not the same as long-term outcome evidence, and accuracy in an external test set is not the same as demonstrated clinical benefit in routine practice. Over the coming months, real-world data on Pluvicto in hormone-sensitive disease and early experience with CaRi-Heart in unselected populations will show how much of this promise converts into outcomes.
Sources: Diagnostic Imaging — Weekly Scan (July 26 to August 1): FDA expands approval of Pluvicto in combination with ARPI and FDA grants De Novo for AI-based coronary inflammation quantification.
