Quantitative Thresholds & Decision Cutoffs
The numbers that change management, with their evidence and their caveats
Quick Answer
Most nuclear-medicine decisions turn on a small number of quantitative cutoffs — but almost none are absolute. They are protocol-, software-, tracer-, and population-dependent, and several (TID, MFR, drainage T½) require lab-specific normal values. This page collects the thresholds that actually change management, each with what it decides, its evidence, and the caveat that keeps it from being applied blindly. Read the number and the caveat together — the caveat is the expertise.
Theranostics — Selection
| Parameter | Cutoff | Decides | Caveat |
|---|---|---|---|
| Krenning score | ≥ 3 (uptake ≥ liver) | SSTR-avidity sufficient for ¹⁷⁷Lu-DOTATATE | Visual, reference-organ-dependent; higher-grade disease needs paired FDG |
| PSMA positivity | Uptake > liver at ≥ 1 site, no dominant FDG-discordant disease | Eligibility for ¹⁷⁷Lu-PSMA-617 (VISION-type) | Agent- and criteria-specific; FDG/PSMA discordance predicts poorer benefit |
| Dual-tracer discordance | FDG-positive / target-negative | Dedifferentiation → poorer PRRT/RLT benefit | Qualitative; changes plan more than any single SUV |
The selection principle is that uptake intensity is a surrogate for target expression, which predicts response — so these are imaging eligibility decisions, not purely histologic ones.
Oncology — Response
| Framework | Cutoff | Decides | Caveat |
|---|---|---|---|
| Deauville | 1–3 vs 4–5 (vs blood pool/liver) | Complete metabolic response vs residual | Interim de-escalation may require 1–2; protocol-specific |
| PERCIST | ≥ 30% SUL-peak decrease (PMR) | Metabolic response of hottest lesion | Same scanner/uptake-time; SUL (lean mass); liver reference |
| RECIST 1.1 | ≥ 30% ↓ (PR) / ≥ 20% (+≥5 mm) or new lesion (PD) | Anatomic response | Size lags biology; misleading for GIST/immunotherapy |
| Choi (GIST) | ≥ 10% size or ≥ 15% CT-density ↓ | GIST response to imatinib | Metabolic/FDG change precedes even Choi |
| Seminoma residual | > 3 cm, imaged ≥ 6 weeks post-chemo | FDG worth performing (viable vs fibrosis) | Not valid for NSGCT (teratoma FDG-negative) |
Cardiac — Perfusion, Flow, Function
| Parameter | Cutoff | Decides | Caveat |
|---|---|---|---|
| Ischemic burden | commonly > 10% ischemic myocardium | flagged as high-risk / prompts discussion | ISCHEMIA reframed this — burden alone does not mandate revascularization |
| Myocardial flow reserve (MFR) | < 2.0 abnormal (broadly) | balanced/microvascular disease; prognosis | Tracer/software/lab-specific — validate locally |
| Transient ischemic dilation (TID) | ~ 1.2 upper-limit-of-normal | high-risk marker | Protocol/software/normal-database-dependent |
| Summed scores | SSS 0–3 normal; 4–8 mild; 9–13 moderate; ≥ 14 severe | severity/extent | Software normal-database-dependent |
Amyloid (ATTR)
| Parameter | Cutoff | Decides | Caveat |
|---|---|---|---|
| Perugini grade | 2–3 (+ negative monoclonal screen) | ATTR without biopsy (> 99% specific) | AL can be avid — monoclonal screen is mandatory |
| H/CL ratio | ≥ 1.5 at 1 h (~1.3 at 3 h) | supports ATTR | SPECT must confirm myocardial (not blood-pool/rib) uptake |
Thyroid
| Parameter | Cutoff | Decides | Caveat |
|---|---|---|---|
| Ablation activity | 30 mCi (1.1 GBq) non-inferior to 100 mCi | low/intermediate-risk remnant ablation | HiLo/ESTIMABL1; higher activities for high-risk/structural disease |
| TSH stimulation | > 30 mIU/L | adequate uptake for RAI | Withdrawal or rhTSH; exclude iodine load |
| Radioiodine age | avoided < 5 y; low/calculated 5–10 y | pediatric therapy safety | Specialized-center dosimetry |
Renal & GI
| Parameter | Cutoff | Decides | Caveat |
|---|---|---|---|
| Diuretic drainage T½ | < 10 min non-obstructed; > 20 min obstructed; 10–20 equivocal | obstruction assessment | Only valid with hydration/empty bladder/adequate function; protocol (F+20 etc.) matters |
| Gastric retention | > 60% at 2 h and/or > 10% at 4 h | gastroparesis | Standardized meal + 4-h imaging required |
| Gallbladder EF (CCK) | < 38% | biliary dyskinesia | Standardize CCK infusion (rapid infusion → false-low) |
Pulmonary / Radioembolization
| Parameter | Cutoff | Decides | Caveat |
|---|---|---|---|
| Lung dose (Y-90) | ≤ 30 Gy single / ≤ 50 Gy cumulative | deliverable ⁹⁰Y activity | Derived from MAA lung-shunt fraction |
| Tumor-absorbed dose (glass, HCC) | ~ ≥ 120 Gy target | response likelihood | DOSISPHERE-01; personalized dosimetry |
Radiation Safety & Regulatory (US NRC)
These are US NRC values (board-relevant); ICRP recommendations differ (e.g. ICRP occupational effective dose ~20 mSv/yr averaged).
| Parameter | Value (NRC) | Note |
|---|---|---|
| Occupational TEDE | 50 mSv/yr (5 rem) | Lens 150 mSv; skin/extremity 500 mSv |
| Declared-pregnant worker | 5 mSv (0.5 rem) over gestation | ~0.5 mSv/month, ALARA |
| Public | 1 mSv/yr (100 mrem) | ≤ 2 mrem/hr in unrestricted areas |
| Patient release (10 CFR 35.75) | TEDE to others ≤ 5 mSv | Common surrogates: ≤ 33 mCi I-131 or ≤ 7 mR/hr at 1 m; written instructions if > 1 mSv |
| Mo-99 breakthrough | ≤ 0.15 µCi Mo-99 / mCi Tc-99m | At administration (0.15 kBq/MBq) |
| Aluminum (Tc eluate) | ≤ 10 µg/mL | Column breakthrough |
| Radiochemical purity | typically ≥ 90–95% | Agent-specific |
| Written directive | any therapy dose; I-131 > 30 µCi | Required before administration |
| Medical event | dose differs by ≥ 0.5 Sv EDE (or ≥ 0.5 Sv organ) and ≥ 20%, or wrong patient/drug/route/site | Reportable |
Physics — Half-Lives (Reference)
| Nuclide | t½ | Nuclide | t½ |
|---|---|---|---|
| Tc-99m | 6.0 h | Ga-68 | 68 min |
| F-18 | 110 min | Rb-82 | 76 s |
| I-123 | 13.2 h | N-13 | 10 min |
| I-131 | 8.0 d | C-11 | 20 min |
| In-111 | 2.8 d | Lu-177 | 6.65 d |
| Tl-201 | 73 h | Y-90 | 64.1 h |
| Mo-99 | 66 h | Ac-225 | 10.0 d |
| Cu-64 | 12.7 h | Ra-223 | 11.4 d |
Effective half-life: 1/t½(eff) = 1/t½(physical) + 1/t½(biological) — the number that actually governs dose and release timing.
How to Use These
Every cutoff answers a specific decision, and almost every one carries a caveat that determines whether it applies: Krenning ≥ 3 gates PRRT but needs paired FDG in higher grade; Deauville 1–3 is a complete metabolic response except where an interim de-escalation protocol demands 1–2; > 10% ischemic myocardium is high-risk but ISCHEMIA showed burden alone does not mandate revascularization; H/CL ≥ 1.5 supports ATTR only with a negative monoclonal screen and SPECT confirmation.
Several thresholds are not portable — MFR, TID, drainage T½, and summed-score bands depend on tracer, software, and the lab's normal database; PERCIST demands the same scanner/uptake-time. Quoting a universal cutoff without local validation is a classic error.
The regulatory numbers are jurisdiction-specific: the values above are US NRC; ICRP differs (occupational ~20 mSv/yr averaged). Patient-release, written-directive, and medical-event definitions are the ones most often tested and most often misremembered — and the Mo-99 (0.15) / aluminum (10) breakthrough limits anchor generator QC.
Related Pages
- Reference: Oncology response criteria (RECIST/PERCIST), PSMA-PET reporting (PROMISE/PSMA-RADS), Lugano/Deauville, ATA thyroid risk.
- Physics/safety: Physics of nuclear medicine, radiation biology & protection, radiation safety & patient release.
Figure / Diagram Suggestions
- A one-page threshold-with-caveat matrix by domain.
- A decision-cutoff → management flow for the top theranostic/cardiac numbers.
Self-Check
Q1. A referring oncologist quotes "MFR < 2.0 = disease" from a paper using a different scanner. What is your caveat?
Answer: MFR thresholds are tracer-, software-, and lab-specific; the ~2.0 figure is a broad guide that must be validated against local normal values before being applied as a hard cutoff.
Q2. Grade-3 PYP uptake is present. Which single result determines whether this is diagnostic of ATTR, and what confirms the uptake is myocardial?
Answer: A negative monoclonal-protein screen (AL can be avid); SPECT confirms myocardial rather than blood-pool/rib uptake.
Q3. What is the US NRC patient-release limit, and two common surrogate criteria used to satisfy it?
Answer: TEDE to any other individual ≤ 5 mSv; surrogates include ≤ 33 mCi I-131 retained activity or ≤ 7 mR/hr at 1 m, with written instructions if the projected dose > 1 mSv.
Q4. Why is "> 10% ischemic myocardium = revascularize" an oversimplification?
Answer: ISCHEMIA showed that in stable CAD, ischemia burden alone did not identify a survival benefit from routine invasive management — the number flags risk and prompts discussion, it does not mandate revascularization.
Evidence & sources
https://nuc.med/thresholds-decision-cutoffs/ Report an issue