How this instrument works
This calculator scores two prognostic indices for diffuse large B-cell lymphoma (DLBCL) from the same five clinical inputs: age, serum LDH expressed as a ratio to the upper limit of normal (LDH ÷ ULN), whether Ann Arbor stage is III or IV, whether ECOG performance status is 2 or worse, and whether more than one extranodal site is involved. The primary output is the NCCN-IPI (National Comprehensive Cancer Network International Prognostic Index), published by Zhou and colleagues in Blood in 2014. The classic IPI — the original International Prognostic Index from 1993 — is calculated alongside it for reference, since it remains widely cited in older literature and trial eligibility criteria.
The two indices score the same five factors differently. The classic IPI uses simple binary cutoffs: one point if age is over 60, one point if LDH is above the upper limit of normal, with no further grading. The NCCN-IPI instead grades both age and LDH ratio into multiple tiers — age contributes 0, 1, 2, or 3 points depending on whether it falls at 40 or under, 41 to 60, 61 to 75, or over 75, and LDH ratio contributes 0, 1, or 2 points at ratios of 1 or under, 1 to 3, or over 3. Stage, ECOG, and extranodal involvement each still contribute a single point on both indices. This finer grading is why the NCCN-IPI is the primary output here: in the head-to-head comparisons published alongside its derivation, it separated patients into risk groups with more distinct survival outcomes than the classic IPI achieved, particularly at the low-risk and high-risk extremes.
Both indices sort patients into four risk groups with published 5-year overall survival figures from their original validation cohorts. For the NCCN-IPI: low risk (score 0-1) at 96%, low-intermediate (2-3) at 82%, high-intermediate (4-5) at 64%, and high risk (6 or more) at 33%. For the classic IPI: low risk (0-1) at 73%, low-intermediate (2) at 51%, high-intermediate (3) at 43%, and high risk (4-5) at 26%. These percentages describe group-level outcomes in the original rituximab-era study cohorts, not a guaranteed outcome for any one patient.
This is a clinical and oncology reference tool intended to support staging discussions between hematology-oncology providers, not a self-diagnosis or self-prognosis instrument for patients. Neither index captures every factor known to affect DLBCL outcomes — cell-of-origin subtype (germinal-center versus activated B-cell), double-hit or triple-hit cytogenetics, and the specific treatment regimen and response all matter clinically and are not part of either score.
- Enter the patient's age in years — both indices weight age differently, so this single field drives two separate point calculations.
- Enter serum LDH as a ratio to the upper limit of normal (LDH ÷ ULN); a value of 2 means LDH twice the lab's ULN, 0.8 means LDH below normal.
- Set 'Ann Arbor stage III or IV' to Yes for advanced-stage disease, No for stage I or II.
- Set 'ECOG performance status ≥ 2' to Yes if performance status is 2 or worse.
- Set 'More than 1 extranodal site involved' to Yes if disease involves more than one extranodal site.
- Read the NCCN-IPI total and risk group first (the primary result), then compare against the classic IPI total and risk group shown alongside it.
Worked example — three DLBCL risk profiles
A 30-year-old with LDH 0.8 times the upper limit of normal, limited-stage disease, preserved performance status, and no extranodal involvement scores 0 age points and 0 LDH points on the NCCN-IPI, for a total of 0 (low-risk group). The classic IPI also totals 0 (low-risk group) — the two indices agree at this end of the spectrum.
A 65-year-old with LDH twice the upper limit of normal and advanced-stage disease, but preserved performance status and no extranodal involvement, scores 2 NCCN-IPI age points and 1 NCCN-IPI LDH point, plus 1 for stage, for a total of 4 (high-intermediate group). The classic IPI scores 1 for age over 60, 1 for elevated LDH, plus 1 for stage, for a total of 3 (also high-intermediate) — both indices land in the same group here.
A 70-year-old with LDH four times the upper limit of normal, advanced-stage disease, ECOG performance status 2 or worse, and more than one extranodal site scores 2 NCCN-IPI age points, 2 NCCN-IPI LDH points, plus 1 each for stage, ECOG, and extranodal involvement, for a total of 7 of 8 (high-risk group). The classic IPI scores 1 for age, 1 for LDH, plus the same three single points, for a total of 5 of 5 (also high-risk) — both indices agree again at this extreme.
Questions
Why does this calculator show two different risk scores?
Both the NCCN-IPI and the classic IPI are calculated from the same five inputs because they weight those inputs differently. The NCCN-IPI is shown as the primary result because, in comparisons against the classic IPI, its graded scoring of age and LDH gave better separation between risk groups. The classic IPI is included alongside it because it remains widely referenced in older literature, trial eligibility criteria, and prior treatment records.
What's the actual difference between the NCCN-IPI and the classic IPI?
The classic IPI (1993) uses simple binary cutoffs: one point each if age is over 60 and if LDH is above the upper limit of normal. The NCCN-IPI (Zhou et al., 2014) grades both factors into multiple tiers instead — age into four bands and LDH ratio into three — while scoring stage, ECOG performance status, and extranodal involvement the same way on both indices.
What do the 5-year survival percentages next to each risk group mean?
They are the 5-year overall survival rates reported for each risk group in the original validation cohorts: 96/82/64/33% for the NCCN-IPI's low, low-intermediate, high-intermediate, and high-risk groups, and 73/51/43/26% for the same four classic IPI groups. These are outcomes observed across a group of patients in those studies, not a specific forecast for any individual.
Is this tool meant for patients to use for self-diagnosis?
No. It is a clinical and oncology reference tool intended to support staging conversations among hematology-oncology providers who already have a confirmed DLBCL diagnosis. It performs the published index arithmetic on the values entered; it does not replace pathology review, imaging, or an oncologist's overall assessment.
What counts as an extranodal site for this calculator?
An extranodal site is any location of lymphoma involvement outside the lymph nodes and spleen — for example the bone marrow, liver, gastrointestinal tract, or central nervous system. Both indices only add a point when more than one such site is involved, not for a single extranodal site.
What other factors affect DLBCL prognosis that these indices don't capture?
Neither index accounts for cell-of-origin subtype (germinal-center B-cell versus activated B-cell), double-hit or triple-hit cytogenetics (MYC rearrangements with BCL2 and/or BCL6), the specific chemoimmunotherapy regimen used, or how a patient responds on interim or end-of-treatment imaging — all of which carry independent prognostic weight in current practice.
References
- Zhou Z, Sehn LH, Rademaker AW, et al. — NCCN-IPI for DLBCL, Blood. 2014;123(6):837-842
- International Non-Hodgkin's Lymphoma Prognostic Factors Project — NEJM. 1993;329(14):987-994
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.