Interim FDG PET/CT in DLBCL: a borderline Deauville score obtained under bad conditions
A 58-year-old man with stage III diffuse large B-cell lymphoma is referred for interim FDG PET/CT after two cycles of R-CHOP. He is not diabetic but ate a full breakfast at 07:00 and, because a relative told him it would protect his kidneys, took 8 units of rapid-acting insulin at 08:30 without eating again. He arrives at 09:00; point-of-care glucose is 238 mg/dL. Injection is at 09:40 in a room kept at 19 C, and he waits alone in a cold hallway for 60 minutes. Imaging shows a 3.2 cm residual mediastinal mass with SUVmax 4.2; liver SUVmean 2.5; mediastinal blood pool SUVmax 1.9. There is also intense symmetric uptake in supraclavicular, paravertebral and perinephric fat. The oncologist calls asking whether to switch to salvage chemotherapy tomorrow.
1. Was this scan adequately performed? What do the glucose and insulin do to interpretation, and what would you change?
Read the worked answer
No. The study is technically inadequate and should not drive a treatment change.
- Fasting and glucose. Oncologic FDG protocols require a 4-6 h fast with a documented glucose, ideally below about 150-200 mg/dL. At 238 mg/dL, glucose competes with FDG at GLUT transporters, lowering tumor uptake and compressing the lesion-to-background ratio.
- Insulin. Eight units of rapid-acting insulin at 08:30, with injection at 09:40, is only about 70 minutes - far short of the 4-hour interval usually recommended after short-acting insulin. Insulin drives glucose and FDG into skeletal muscle, producing diffuse muscular uptake and a falsely low tumor SUV.
- Why it matters here. SUVmax 4.2 against a liver of 2.5 sits exactly in the borderline range where a one-point Deauville difference changes management, and none of these numbers can be trusted.
What to do instead: do not cancel chemotherapy on this scan. Rebook within 24-48 h with an afternoon appointment, a 6 h fast, no insulin on the day (or a dose given at least 4 h before injection with a light meal), a warm waiting room, a blanket, and a quiet uptake room. Document the glucose value and insulin timing in the report so the next reader knows the study was repeated for cause.
2. How do you classify the fat uptake, and how do you decide whether the mediastinal activity is residual lymphoma?
Read the worked answer
The fat uptake is brown adipose tissue. Its features are bilateral symmetry, distribution in fat rather than in lymph node chains (supraclavicular, paravertebral, perinephric), and provocation by a cold room plus shivering in a fasting, sympathetic-driven patient. Lymph nodes would be discrete, rounded, and would follow the CT anatomy of a nodal station.
Practical handling: warm the uptake room to roughly 22-24 C, provide blankets and loose clothing, and repeat the scan after warming if the fat uptake is obscuring a region of interest. Some protocols use a beta-blocker or a benzodiazepine, but warming is the intervention with the best risk-benefit profile.
The mediastinal mass is a separate question. Its uptake is focal, conforms to the CT mass, and exceeds the liver (4.2 vs 2.5), which is Deauville 4. Brown fat neither explains it nor explains it away. Given the failed preparation, the honest interpretation is 'borderline Deauville 4, technically compromised' - neither positive nor negative can be defended without a repeat. Do not let the striking fat uptake distract from the fact that the actual decision hinges on the mass.
3. State the Deauville score for this case and what you would tell the oncologist to do.
Read the worked answer
Score: Deauville 4 (uptake greater than liver, but less than roughly twice liver). For reference: 1 = no uptake; 2 = at or below mediastinal blood pool; 3 = above blood pool but at or below liver; 4 = above liver; 5 = markedly above liver (about 2-3x) or new lesions.
In DLBCL, Deauville 4 is conventionally read as positive - the Hodgkin-specific allowance for score 4 with uptake no more than twice liver does not transfer. Taken literally, this score could justify salvage therapy tomorrow.
Why that would be the wrong decision:
- Interim PET response-adaptation has weaker evidence in DLBCL than in Hodgkin lymphoma; a single borderline focus is a poor basis for changing a curative regimen.
- The preparation was flawed (hyperglycemia, recent insulin, cold room), so the SUV is not reproducible.
- A 3.2 cm residual mass with mild uptake after two cycles is common and frequently inflammatory.
Recommendation: repeat FDG PET/CT with correct preparation within a week before altering therapy. If the repeat is Deauville 1-3, complete planned R-CHOP. If 4-5 persists, restage, consider biopsy of the residual mass if accessible, and take the case to tumor board. Never let the number 4.2 alone make the decision.