# The Linear-Quadratic Model and Fractionation Biology

## Overview
The linear-quadratic (LQ) model is the most widely accepted mathematical framework used to describe the relationship between radiation dose, fractionation, and the resulting biological effect. It provides the theoretical foundation for understanding why fractionation improves therapeutic outcomes and allows for the comparison of different dose-fractionation schedules through the concept of biologically effective dose (BED). Despite its broad utility, the model has important limitations, especially at high doses per fraction, which are increasingly relevant in modern stereotactic treatments.

## Derivation of the LQ Model

### Cell Survival Curves
Cell survival following radiation exposure is typically plotted on a semi-logarithmic scale, with the logarithm of the surviving fraction on the y-axis and dose on the x-axis. The LQ model mathematically describes the survival curve as \( S = \exp(-\alpha d - \beta d^2) \), where \( S \) is the surviving fraction, \( d \) is the dose per fraction, \( \alpha \) is the linear coefficient, and \( \beta \) is the quadratic coefficient. This formulation captures an initial linear slope determined by the \( \alpha \) term and a quadratic bending component represented by \( \beta d^2 \). At low doses, cell kill is primarily driven by the linear \( \alpha \) component, whereas at higher doses, the quadratic \( \beta d^2 \) term increasingly contributes to cell death.

### Biological Interpretation
The linear \( \alpha \) component corresponds to lethal damage caused by single-track events, such as a single radiation track inducing double-strand breaks (DSBs) in DNA. The probability of such damage is proportional to the dose. In contrast, the quadratic \( \beta \) component represents lethal damage arising from the interaction of two independent sublethal events, with a probability proportional to the square of the dose. The alpha/beta ratio is defined as the dose at which the contributions of the linear and quadratic components to cell kill are equal.

### Alpha/Beta Ratios
Tissues and tumors differ in their alpha/beta ratios, which reflect their sensitivity to fraction size. High alpha/beta ratios, approximately 10 Gy, are typical of rapidly proliferating tumors and acute-responding normal tissues. Examples include squamous cell carcinomas, lymphomas, and most common epithelial cancers. Acute radiation effects such as mucositis, dermatitis, and enteritis are associated with these tissues, which tend to be relatively insensitive to changes in fraction size. Conversely, low alpha/beta ratios, ranging from about 1.5 to 5 Gy, characterize late-responding normal tissues and some slow-growing tumors. Examples include the spinal cord (~2 Gy), lung (~3 Gy), kidney (~2.5 Gy), and brain (~2 Gy). Late radiation effects such as fibrosis, myelopathy, and nephropathy occur in these tissues. Notably, prostate cancer has a particularly low alpha/beta ratio (~1.5 Gy), making it more sensitive to fraction size changes.

| Tissue / Tumor | Alpha/Beta Ratio (Gy) | Category | Clinical Significance |
|---|---|---|---|
| Squamous cell carcinoma | ~10 | High (acute-responding) | Relatively insensitive to fraction size |
| Lymphoma | ~10 | High (acute-responding) | Relatively insensitive to fraction size |
| Mucosa (acute effects) | ~10 | High (acute-responding) | Mucositis, dermatitis |
| Breast cancer | ~4 | Intermediate | Supports moderate hypofractionation |
| Lung (late effects) | ~3 | Low (late-responding) | Fibrosis; sensitive to fraction size |
| Kidney | ~2.5 | Low (late-responding) | Nephropathy risk |
| Spinal cord | ~2 | Low (late-responding) | Myelopathy risk; sensitive to fraction size |
| Brain | ~2 | Low (late-responding) | Necrosis risk |
| Prostate cancer | ~1.5 | Low (tumor) | Strongly favors hypofractionation |

## Biologically Effective Dose (BED)

### Formula
The biologically effective dose (BED) is calculated using the formula \( \text{BED} = nd \left[1 + \frac{d}{\alpha/\beta}\right] \), where \( n \) is the number of fractions, \( d \) is the dose per fraction, and \( nd \) is the total dose. BED enables comparison of different fractionation schemes for tissues with a specified alpha/beta ratio. It is also referred to as the biologically effective dose or sometimes the "extrapolated response dose."

### Equivalent Dose in 2-Gy Fractions (EQD2)
The equivalent dose in 2-Gy fractions (EQD2) converts any fractionation schedule to its equivalent in conventional 2 Gy fractions. It is calculated as \( \text{EQD2} = \frac{\text{BED}}{1 + \frac{2}{\alpha/\beta}} = \text{total dose} \times \frac{d + \alpha/\beta}{2 + \alpha/\beta} \). This conversion is essential for comparing hypofractionated, hyperfractionated, and conventional schedules. For example, a regimen of 20 Gy delivered in 5 fractions (4 Gy per fraction) with an alpha/beta of 10 Gy yields an EQD2 of \( 20 \times \frac{4 + 10}{2 + 10} = 23.3 \) Gy.

### Practical BED Calculations
In standard whole breast radiotherapy, a schedule of 50 Gy in 25 fractions corresponds to a BED of 60 Gy for tissues with an alpha/beta of 10 and 83.3 Gy for an alpha/beta of 3. The FAST-Forward trial regimen of 26 Gy in 5 fractions results in a BED of 39.5 Gy for alpha/beta 10 and 71.1 Gy for alpha/beta 3. For prostate stereotactic body radiotherapy (SBRT), delivering 36.25 Gy in 5 fractions with an alpha/beta of 1.5 yields a BED of 211.6 Gy, whereas the standard prostate regimen of 78 Gy in 39 fractions corresponds to a BED of 182 Gy for the same alpha/beta.

| Regimen | Total Dose / Fractions | Dose/Fx | BED (α/β = 10) | BED (α/β = 3) | BED (α/β = 1.5) |
|---|---|---|---|---|---|
| Standard breast | 50 Gy / 25 fx | 2.0 Gy | 60.0 Gy | 83.3 Gy | — |
| FAST-Forward (breast) | 26 Gy / 5 fx | 5.2 Gy | 39.5 Gy | 71.1 Gy | — |
| Standard prostate | 78 Gy / 39 fx | 2.0 Gy | — | — | 182.0 Gy |
| Prostate SBRT | 36.25 Gy / 5 fx | 7.25 Gy | — | — | 211.6 Gy |
| Lung SBRT | 54 Gy / 3 fx | 18 Gy | 151.2 Gy | — | — |

## Fractionation Strategies

### Conventional Fractionation
Conventional fractionation typically involves delivering 1.8 to 2.0 Gy per fraction, five days per week. This approach exploits the "4 Rs" of radiobiology—repair, reassortment, repopulation, and reoxygenation—to maximize the therapeutic ratio. The total treatment time usually spans 5 to 7 weeks depending on the disease. Conventional fractionation remains the standard against which alternative schedules are compared.

### Hypofractionation
Hypofractionation involves fraction sizes greater than 2 Gy with fewer total fractions. Moderate hypofractionation uses doses between 2.5 and 4 Gy per fraction, such as 40 Gy in 15 fractions for breast cancer or 60 Gy in 20 fractions for prostate cancer. Ultra-hypofractionation or SBRT employs even larger doses per fraction, ranging from 6 to 34 Gy, exemplified by 54 Gy in 3 fractions for lung SBRT. The advantages of hypofractionation include shorter treatment times, improved patient convenience, and greater resource efficiency. However, it carries the disadvantage of increased sensitivity to fraction-size errors and a higher risk of late effects if dose constraints are not strictly observed. It is crucial that the late-tissue BED remains within tolerance limits.

### Hyperfractionation
Hyperfractionation uses fraction sizes smaller than 1.8 Gy, typically delivered twice daily with at least a 6-hour interval between fractions. This strategy exploits the greater sensitivity of late-responding tissues to fraction size, allowing delivery of a higher total BED to the tumor while maintaining acceptable late toxicity. An example is the regimen of 81.6 Gy in 68 fractions of 1.2 Gy twice daily for head and neck cancer, as studied in RTOG 9003. Despite its radiobiological rationale, hyperfractionation is logistically challenging and less commonly used in current practice.

### Accelerated Fractionation
Accelerated fractionation shortens the overall treatment time to counteract accelerated tumor repopulation. This is achieved by delivering multiple daily fractions or including weekend treatments. The CHART regimen, which delivers 54 Gy in 36 fractions over 12 consecutive days with three fractions per day, is an example used for non-small cell lung cancer (NSCLC). Accelerated fractionation is particularly relevant for rapidly proliferating tumors such as head and neck squamous cell carcinoma and NSCLC.

| Fractionation Strategy | Dose/Fraction | Schedule | Example Regimen | Key Advantage |
|---|---|---|---|---|
| Conventional | 1.8–2.0 Gy | 5 days/week, 5–7 weeks | 70 Gy / 35 fx (H&N) | Maximizes 4 Rs; well-tolerated |
| Moderate Hypofractionation | 2.5–4.0 Gy | 5 days/week, 3–4 weeks | 40 Gy / 15 fx (breast) | Shorter course; equivalent outcomes |
| Ultra-hypofractionation / SBRT | 6–34 Gy | 1–5 fractions | 54 Gy / 3 fx (lung) | Very short course; high local control |
| Hyperfractionation | < 1.8 Gy (BID) | Twice daily, 6–7 weeks | 81.6 Gy / 68 fx @ 1.2 Gy (H&N) | Higher tumor BED; spares late tissues |
| Accelerated | Standard or reduced | Shortened OTT | CHART: 54 Gy / 36 fx / 12 days | Counteracts tumor repopulation |

## Incomplete Repair and the Interfraction Interval
The LQ model assumes complete repair of sublethal damage between fractions, which requires a minimum interfraction interval of about 6 hours for conventional schedules and approximately 8 hours for central nervous system tissues. If repair is incomplete, a modified LQ formula incorporating the Lea-Catcheside time factor (G) is used to account for residual damage. Repair half-times are generally around 1 to 2 hours for most tissues but may be longer for the spinal cord, approximately 4 hours. This consideration is clinically important for twice-daily fractionation and continuous low-dose-rate (LDR) brachytherapy.

## Limitations of the LQ Model

### High Dose Per Fraction
At doses per fraction exceeding 8 to 10 Gy, the LQ model may overestimate cell kill because the survival curve tends to straighten out, becoming more linear at high doses. Alternative models have been proposed to address this, including the Universal Survival Curve (USC), which transitions from LQ behavior at low doses to linear at high doses; the Linear-Quadratic-Linear (LQL) model, which adds a linear component at high doses; and generalized LQ models that incorporate repair kinetics. Despite these theoretical concerns, many institutions continue to use the LQ/BED formalism for SBRT dose comparisons. Clinical outcomes with SBRT have empirically validated certain dose-fractionation schedules.

### Heterogeneity of Alpha/Beta
The alpha/beta ratio represents an average value for a given tissue, but individual patient variation exists. Tumor alpha/beta ratios may vary depending on factors such as hypoxia, cell cycle phase, and molecular subtype. Population-based alpha/beta estimates often have wide confidence intervals, limiting precision.

### Does Not Account for Repopulation
The basic LQ formula does not include a time factor to account for tumor repopulation during treatment. A modified BED formula incorporates repopulation as follows: \( \text{BED} = nd \left[1 + \frac{d}{\alpha/\beta}\right] - \frac{0.693}{\alpha} \frac{(T - T_k)}{T_{\text{pot}}} \), where \( T \) is the overall treatment time, \( T_k \) is the time at which accelerated repopulation begins (approximately 28 days for head and neck squamous cell carcinoma), and \( T_{\text{pot}} \) is the potential doubling time. This modification highlights the clinical importance of minimizing treatment breaks, especially for rapidly proliferating tumors.

### Does Not Model Vascular/Stromal Effects
At very high single doses, additional mechanisms of cell kill may occur, such as endothelial cell apoptosis, vascular disruption, and immune activation. These effects are not captured by the LQ model and may explain why SBRT appears more effective than predicted by LQ-based BED calculations for certain tumors.

<image>A graph showing cell survival curves on a semi-log plot (log surviving fraction on y-axis, dose in Gy on x-axis). Two curves are shown: one for a high alpha/beta tissue (alpha/beta = 10 Gy, straighter curve) and one for a low alpha/beta tissue (alpha/beta = 3 Gy, more curved with a prominent shoulder). The alpha and beta components are visually separated, and the dose where alpha*d = beta*d^2 (i.e., d = alpha/beta) is marked on each curve.</image>

<image>A comparative bar chart showing BED values for different fractionation schedules applied to the same clinical scenario (e.g., prostate cancer). Schedules include: 78 Gy/39 fractions, 60 Gy/20 fractions, 36.25 Gy/5 fractions (SBRT). BED is calculated and displayed for both tumor (alpha/beta = 1.5 Gy) and rectum (alpha/beta = 3 Gy), demonstrating the therapeutic window achievable with hypofractionation for low alpha/beta tumors.</image>

<image>A schematic diagram illustrating the therapeutic ratio concept. Two curves (tumor control probability and normal tissue complication probability) are plotted against dose. The separation between the curves represents the therapeutic window. Three panels show how the therapeutic ratio changes with (A) conventional fractionation, (B) hypofractionation in a low alpha/beta tumor (window widens), and (C) hypofractionation in a high alpha/beta tumor (window narrows or disappears).</image>

## Key Clinical Pearls
The most important clinical application of the LQ model is in comparing fractionation schedules; therefore, it is essential to calculate BED or EQD2 when evaluating any non-standard fractionation scheme. Prostate cancer’s notably low alpha/beta ratio (~1.5 Gy) means that hypofractionation delivers a higher tumor BED relative to late-tissue BED compared to conventional fractionation, providing the radiobiological rationale for prostate SBRT. Treatment interruptions are costly, particularly for head and neck squamous cell carcinoma, where each day of prolongation beyond the planned treatment time reduces local control by approximately 1.4% due to accelerated repopulation. When using BED to evaluate SBRT schedules with doses greater than 8 Gy per fraction, one must recognize that the LQ model may overestimate the biologically effective dose; thus, clinical trial data should guide dose selection over BED calculations alone. Finally, it is critical always to calculate BED for both the tumor and the relevant dose-limiting organs at risk when designing or evaluating a fractionation schedule.

## References
- Fowler JF. "The linear-quadratic formula and progress in fractionated radiotherapy." *Br J Radiol*. 1989;62(740):679-694.
- Fowler JF. "21 years of biologically effective dose." *Br J Radiol*. 2010;83(991):554-568.
- Brenner DJ, Hall EJ. "Fractionation and protraction for radiotherapy of prostate carcinoma." *Int J Radiat Oncol Biol Phys*. 1999;43(5):1095-1101.
- Park C et al. "Universal survival curve and single fraction equivalent dose: useful tools in understanding potency of ablative radiotherapy." *Int J Radiat Oncol Biol Phys*. 2008;70(3):847-852.
- Withers HR et al. "Treatment volume and tissue tolerance." *Int J Radiat Oncol Biol Phys*. 1988;14(4):751-759.