Bibliographic Reference
Traina, T. A., Dugan, U., Higgins, B., Kolinsky, K., Theodoulou, M., Hudis, C. A., & Norton, L. (2010). Optimizing chemotherapy dose and schedule by Norton-Simon mathematical modeling. Breast Disease, 31(1), 7–18. https://doi.org/10.3233/BD-2009-0290
Core Argument
Norton-Simon growth kinetic modeling — derived from Gompertzian growth — can be applied to preclinical xenograft data to predict optimal chemotherapy schedules, identifying the time point where drug effect peaks rather than defaulting to maximum-tolerated-dose regimens. The method was validated by predicting from MX-1 xenograft data that a 7-days-on/7-days-off capecitabine schedule would outperform the standard 14-days-on/7-days-off schedule, a prediction subsequently confirmed in KPL-4 xenograft experiments.
Methods
Norton-Simon mathematical modeling applied to tumor-volume data from breast cancer xenograft models (MX-1 and KPL-4 cell lines) treated with capecitabine. The model incorporates Gompertzian growth kinetics — where the relative growth rate falls exponentially as tumor size increases — and posits that the rate of tumor regression under therapy is proportional to the Gompertzian-predicted unperturbed growth rate. Drug effect as a function of time D(t) was fitted using NONMEM V software with nonlinear mixed-effects population modeling. Two schedules were compared: the conventional 14 days treatment with 7 days rest (14-7), and the model-predicted optimal schedule of 7 days treatment with 7 days rest (7-7).
Key Findings
- The Norton-Simon model, fitted to MX-1 xenograft tumor-volume data, revealed that capecitabine’s drug effect D(t) reaches a maximum at approximately 8–10 days after the start of therapy, after which it declines despite continued dosing — a pattern consistent with tachyphylaxis. This suggested that the second week of drug administration in the conventional 14-7 schedule may contribute more to toxicity than to anticancer effect.
- The model-predicted optimal schedule (7-7, biweekly capecitabine) was tested in KPL-4 xenograft experiments: it permitted safe delivery of higher daily doses (MTD of 700 mg/kg/day vs. 400 mg/kg/day on 14-7), achieved greater tumor growth inhibition, produced significant tumor regressions (which the 14-7 schedule did not), and prolonged survival compared to the conventional 14-7 schedule.
- At the higher doses made possible by the 7-7 schedule, a dose-response relationship was observed: 700 mg/kg/day had significantly higher antitumor activity than 500 mg/kg/day (P<.001) or 600 mg/kg/day (P=.009), supporting the hypothesis that the 7-7 schedule shifts the therapeutic index by reducing toxicity while preserving or enhancing efficacy.
Concepts Introduced or Used
- Norton-Simon model: mathematical framework for chemotherapy scheduling based on Gompertzian growth kinetics
- Gompertzian growth: relative growth rate falls exponentially as population size increases, approaching a plateau
- Dose density: concept that administering chemotherapy at the maximal feasible rate (minimum safe interval between doses) improves outcomes
- Tachyphylaxis: waning of medication effect with continuous exposure; cited as biochemical rationale for why more drug is not always more effective
- Drug effect function D(t): time-varying measure of drug activity; peaks and then declines despite continued dosing, and when D(t) follows this pattern the optimal strategy is to dose to the point of maximal impact and then discontinue
- therapy-resistance — the paper addresses practical optimization of cytotoxic chemotherapy rather than resistance mechanisms directly, but the Norton-Simon framework informs dose-dense strategies relevant to preventing resistant clone emergence
Entities Referenced
- Cell lines: MX-1 (human breast cancer xenograft), KPL-4 (human breast cancer xenograft)
- Drugs: Capecitabine (Xeloda), 5-fluorouracil, cyclophosphamide, doxorubicin, paclitaxel, docetaxel, ixabepilone, lapatinib
- Software: NONMEM V (nonlinear mixed-effects modeling)
- Clinical trials referenced: CALGB 9741 (dose-dense breast cancer trial), phase I trial of capecitabine 7-7 by Traina et al. 2008
Limitations (as stated by authors)
- The study was performed in xenograft models; the authors state that the method “warrants further investigation and application in clinical drug development” and that clinical translation is ongoing (phase I completed, phase II and III trials planned).
- The method was demonstrated on capecitabine specifically; application to other agents was in early stages at the time of publication, with phase I/II studies ongoing for a new cytotoxic agent using the same modeling approach.
Relevance to Clonal Evolution
This paper does not directly address clonal evolution or tumor heterogeneity — it models tumors as single Gompertzian populations. However, its demonstration that Norton-Simon scheduling improves outcomes establishes the baseline that the Castorina et al. (2009) two-population extension critiques: when clonal heterogeneity is present, the single-population schedule becomes suboptimal. Together, these papers show that growth kinetics matter for treatment strategy, and that ignoring clonal diversity in growth models has concrete therapeutic consequences.