- Design
- Single-centre retrospective cohort study with ROC analysis against pathological findings
- Population
- 66 patients with oesophagogastric adenocarcinoma who had contemporaneous CECT and FDG PET-CT before and after neoadjuvant chemotherapy
- Primary outcome
- Accuracy for pathological T and N stage, pathological treatment response, and detection of distant metastases
- Effect
- ypT AUC 0.65 (CECT) vs 0.62 (PET-CT), p=0.76; ypN 0.52 vs 0.53; response 0.51 vs 0.72, p=0.041; metastases detected 4/8 vs 8/8, p=0.076
FDG PET-CT is established for staging oesophagogastric adenocarcinoma. Whether it adds anything over contrast-enhanced CT when restaging after neoadjuvant chemotherapy has been assumed rather than measured.
This single-centre cohort identified 105 patients, 92 of whom had surgical resection. Sixty-six had both scans performed contemporaneously — same scanner, same hospital visit — which is the comparison that counts. Response on CT was assessed by change in oesophageal thickness, on PET by change in SUVmax, with ROC analysis against pathological findings.
For pathological T stage the two were comparable (AUC 0.65 versus 0.62, p=0.76), and for nodal stage both were essentially uninformative (0.52 versus 0.53, p=0.92). PET-CT separated on response: AUC 0.72 against 0.51 for CT (p=0.041). And of eight patients found to have metastatic disease at restaging, PET-CT identified all eight and CT identified four (p=0.076).
The practical shape of that is a targeting rule, not a blanket one. If the question is what stage the tumour has come down to, neither modality answers it well and the cheaper one is not worse. If the question is whether the tumour responded, or whether there is occult metastatic disease that should stop an oesophagectomy, PET-CT is the one that answers. In a system where PET capacity is scarce and paid for out of pocket, that distinction decides who gets the slot.
- Do not order PET-CT to sharpen ypT or ypN staging — it does not
- Order it where occult metastatic disease would change the decision to operate
- Response assessment by SUVmax change outperformed wall thickness change on CT
- Eight metastases found, four missed by CT — small numbers, but the direction matters before a major resection
- Cost and access are a legitimate part of this decision; the authors say so explicitly
Why it matters
It identifies which question PET-CT actually answers after chemotherapy, in a setting where capacity has to be rationed.
Don't overread it
The metastasis difference rests on eight events and was not statistically significant.
The statistics, in plain English
An AUC of 0.5 is a coin toss; 0.52 and 0.53 for nodal stage means neither modality tells you anything useful about the nodes. The response comparison, 0.72 against 0.51, is a real separation but still far from a reliable test. The metastasis finding — 8 of 8 versus 4 of 8 — did not reach statistical significance (p=0.076) because eight events is too few to prove anything; it is suggestive, not established.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for radiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free