The same 115 patients got two scans, and the newer one declared most of the "ineligible" group eligible for treatment after all.
That is the finding of a retrospective study published online June 12 in Neuroendocrinology ↗. All 115 had well-differentiated neuroendocrine tumors of the gut and pancreas, the slow-growing cancers that often need a specialized therapy because they spread quietly and resist ordinary chemo. The two scans disagreed on who qualified for that therapy. The disagreement ran almost entirely in one direction.
The therapy, and the gate in front of it
The treatment is peptide receptor radionuclide therapy, or PRRT. The peptide is a lab-made copy of the hormone somatostatin, re-engineered to carry a radioactive atom of lutetium-177. Neuroendocrine tumors tend to stud their surface with somatostatin receptors, so the peptide homes to the tumor and irradiates it from the inside. Sold as Lutathera, it is the same class of "radioactive peptide" the platform covered on June 4, when a single-center study ↗ found it worked in tumor subtypes the big trials skipped.
To qualify, a patient's tumor has to light up brightly enough on a scan that detects those same receptors. The brightness gets graded on the Krenning score, a 0 to 4 scale. The pivotal trials set the bar there: NETTER-1 used Krenning of at least 2, the stricter NETTER-2 used at least 3.
The catch is that there is more than one scan, and they do not light tumors up equally.
A brighter scan, a higher score
The older method is [111In]octreotide scintigraphy, a planar gamma-camera image built on a radiolabeled version of octreotide ↗, the somatostatin-mimicking drug. The newer method is [68Ga]DOTATOC PET, a positron scan that is simply more sensitive. Point both at the same tumor and the PET image reads brighter, which pushes the Krenning score up.
The numbers are stark. Median Krenning score was 3 on PET versus 2 on the planar scan. Eligibility under the NETTER-1 bar rose from 59.1 percent of patients on planar imaging to 81.7 percent on PET. Under the stricter NETTER-2 bar it rose from 48.7 percent to 76.5 percent. Among patients already eligible on the old scan, the PET almost always agreed: 95.6 percent stayed eligible under NETTER-1, 98.2 percent under NETTER-2. The movement was nearly one-way. The newer scan rarely knocked anyone out, it mostly pulled people in. More than half of the patients ruled ineligible on planar imaging, 61.7 percent under NETTER-1 and 55.9 percent under NETTER-2, were reclassified as eligible once the PET was read.
Why the scan is the story
A patient does not choose their Krenning score. The scanner in the building does. If a center still runs planar octreotide scintigraphy, a patient whose tumor would have qualified on PET can be scored out of a treatment that might have helped, not because the disease changed but because the camera did. The eligibility thresholds were written on trials, and the trials assumed an imaging standard. When the imaging gets sharper than the standard, the threshold quietly drifts.
The authors offer one way to pin it down. Quantitative PET, rather than the eyeball Krenning grade, separated the high-uptake tumors cleanly: a maximum standardized uptake value of 9.2 discriminated Krenning of at least 3 with an area under the curve of 0.93, where 1.0 is perfect. Metastatic disease lit up brightest, with nearly ten times the odds of strong uptake (odds ratio 9.5).
This is one retrospective series at one center, and Krenning remains a semi-quantitative read rather than a hard cutoff. But the direction is hard to argue with. A peptide therapy is only as available as the scan that gates it, and the gate just moved.