Sokal Index (CML)

Baseline chronic-phase CML prognostic index (Blood 1984). Enter age, palpable spleen size (cm below the costal margin), platelet count in ×10⁹/L, and peripheral blood blast percentage.

Age (years)

Spleen below costal margin (cm)

Use 0 if the spleen is not palpable.

Platelet count (×10⁹/L)

Peripheral blasts (%)

Disclaimer: For education only. The Sokal index reflects a pre-TKI population; modern CML care depends on molecular monitoring, comorbidities, and current guidelines. Do not use this tool alone to direct therapy.

Sokal Index (CML) — Formula Explained

Background

The Sokal index is a baseline prognostic score for chronic-phase CML derived from age, spleen size, platelet count, and peripheral blood blast percentage. It was developed in patients treated in the pre–tyrosine kinase inhibitor era and remains a historical reference; many guidelines also emphasize newer scores (e.g., ELTS, EUTOS) for contemporary risk discussion.

Primary reference: Sokal JE, Cox EB, Baccarani M, et al. Prognostic discrimination in "good-risk" chronic granulocytic leukemia. Blood. 1984;63(4):789-799.

Score

Sokal score = exp(L)
L = 0.0116 × (age − 43.4) + 0.0345 × (spleen cm − 7.51) + 0.188 × [(platelets ÷ 700)² − 0.563] + 0.0887 × (peripheral blasts % − 2.10)

Inputs

Age: years
Spleen: centimeters of palpable spleen below the left costal margin (use 0 if not palpable)
Platelets: ×10⁹/L (10⁹/L); numerically the same as ×10³/µL
Blasts: myeloblasts in peripheral blood (%)

Risk groups (original Sokal cutoffs)

Sokal score	Group
< 0.8	Low risk
0.8 – 1.2	Intermediate risk
> 1.2	High risk

Boundaries: scores exactly 0.8 and 1.2 are classified as intermediate (inclusive), matching common clinical calculator conventions.

Limitations

Derived before widespread TKI therapy; does not replace molecular response monitoring or modern scores.
Spleen and blast values must reflect the same clinical encounter and laboratory used for the original model.
Not validated for blast crisis or non–chronic-phase presentations in the same way as for chronic-phase CML.

What the Sokal index measures

The Sokal index (often called the Sokal score) is a baseline prognostic index for patients with chronic-phase chronic myeloid leukemia (CML). It distills four readily available clinical and laboratory features into one number. That number was originally used to separate patients into low, intermediate, and high risk groups for outcomes in an era when therapy relied heavily on nonspecific agents such as interferon and conventional chemotherapy rather than modern targeted drugs.

Although treatment of CML has been transformed by tyrosine kinase inhibitors (TKIs), the Sokal index remains a familiar historical benchmark. It still appears in educational materials, older trial reports, and discussions of how baseline tumor burden and host factors relate to early response. It should be interpreted as one piece of context, not as a standalone directive for therapy in contemporary practice.

Intended population and timing

The model was developed for adults with chronic-phase CML at diagnosis or early in the chronic phase, using variables measured at a comparable baseline visit. It is not designed for accelerated phase, blast crisis, or other acute leukemia presentations. Applying the same arithmetic in those settings can produce misleading reassurance or alarm because the underlying biology and the validation cohort differ sharply from chronic-phase disease.

When using any baseline score, clinicians should ensure that all inputs reflect the same time point—for example, the peripheral smear differential and platelet count from the same laboratory draw, and the physical examination finding for splenic size from the same encounter (or clearly documented closest approximation).

Input variables in plain language

Age

Age in years enters the formula as a continuous variable centered near middle adulthood in the original cohort. Older age is associated with a higher linear predictor and therefore a higher Sokal score, reflecting the model’s finding that increasing age carried prognostic weight after accounting for disease extent markers.

Spleen size

The index uses palpable splenic size below the left costal margin, measured in centimeters. If the spleen is not felt below the costal margin, this value is zero. Consistency matters: different examiners, patient habitus, and inspiratory effort can change the measured distance slightly; in borderline cases, imaging may clarify organ size, but the original score was tied to clinical palpation as used in the derivation study.

Platelet count

Platelets should be entered as the count in ×10⁹/L (10⁹ per liter), which is numerically the same as the value often reported as ×10³/µL on North American reports (for example, 350 × 10⁹/L corresponds to 350,000/µL expressed in conventional “K” units). The formula squares a normalized platelet term, so very high platelet counts can increase the score substantially.

Peripheral blood blasts

The model uses the percentage of blasts on the peripheral blood differential—specifically myeloblasts in the CML context. Even small deviations from the reference anchor in the equation can move the linear predictor because the blast term is scaled by a coefficient. Laboratories and manual differentials vary in precision; repeat sampling or expert review may be appropriate when blast estimates are critical to management.

How the score is calculated

First, a linear predictor (sometimes denoted L or the log-risk component) is computed from the four variables. Each variable is compared with a constant derived from the original statistical model, then multiplied by a weight. The terms are added together.

Conceptually:

One term depends on age relative to a reference age.
One term depends on spleen size relative to a reference splenic extent.
One term depends on platelet count through a quadratic normalization (platelets divided by 700, then squared, then adjusted by a constant).
One term depends on peripheral blast percentage relative to a small reference percentage.

The Sokal score itself is the exponential of that sum: mathematically, score = exp(linear predictor). Because of the exponential step, modest increases in the linear predictor can translate into larger increases in the final score. When reviewing a calculator output, seeing both the final score and the linear predictor (or its individual terms) helps explain why a patient falls into a given risk band.

Risk groups and how they were used historically

After computing the Sokal score, patients are grouped by cut points that became standard in the literature:

Low risk: score below 0.8
Intermediate risk: score from 0.8 through 1.2 (inclusive at both boundaries in common calculator implementations)
High risk: score above 1.2

In the pre-TKI datasets that motivated the index, these bands summarized differences in survival and failure patterns at the population level. They were never intended to override bedside judgment, comorbidity assessment, or later findings such as cytogenetic complexity and molecular response depth, which dominate modern CML management.

Relationship to modern CML care

Contemporary guidelines emphasize molecular monitoring (for example, BCR::ABL1 transcript levels), treatment adherence, mutation testing when response is suboptimal, and TKI selection based on efficacy, toxicity, comorbidities, and patient goals. Several newer baseline scores—such as the ELTS score and EUTOS long-term survival score—were developed or recalibrated with TKI-treated cohorts and may track long-term outcomes more closely than Sokal in current practice.

That does not make the Sokal index useless for learning or for reading older literature, but it does mean the score should be framed honestly: it is primarily a historical prognostic stratifier with fixed coefficients from a specific treatment era. Decisions about drug choice, intensity of monitoring, or transplant referral should follow current disease-specific standards and multidisciplinary input.

Limitations and pitfalls

Treatment era mismatch: Coefficients reflect outcomes under older therapeutic approaches; absolute prognostic expectations change with TKI therapy.
Data quality: Misentered platelet units (for example, using cells/µL without converting to 10⁹/L) will distort the quadratic platelet term.
Exam variability: Splenic palpation is subjective; borderline measurements can shift the score.
Phase of disease: Not validated for accelerated or blast phases in the same way as for chronic phase.
Populations: Original cohort demographics may not match every contemporary patient mix; calibration can differ by region, age distribution, and access to care.

Using this calculator responsibly

This tool is for education and clinical reasoning support. It does not establish a diagnosis, stage disease by itself, or replace consultation with hematology/oncology specialists, molecular pathology, or institutional pathways. Any unexpected blast fraction, rapidly changing counts, or symptomatic splenomegaly warrants urgent specialist evaluation rather than reliance on a prognostic index alone.

CalcMD.

Sokal Index for Chronic Myelogenous Leukemia (CML)

Compute the Sokal index for chronic-phase CML from age, spleen size (cm below the costal margin), platelets (×10⁹/L), and peripheral blasts (%). Explains low, intermediate, and high risk bands and how the score relates to modern TKI-era care.