Dose-effect relationships for the submandibular salivary glands and implications for their sparing by intensity modulated radiotherapy

Abstract

Purpose: Submandibular salivary glands (SMGs) dysfunction contributes to xerostomia after radiotherapy (RT) of head-and-neck (HN) cancer. We assessed SMG dose-response relationships and their implications for sparing these glands by intensity-modulated radiotherapy (IMRT).

Methods and materials: A total of 148 HN cancer patients underwent unstimulated and stimulated SMG salivary flow rate measurements selectively from Wharton's duct orifices, before RT and periodically through 24 months after RT. Correlations of flow rates and mean SMG doses were modeled throughout all time points. IMRT replanning in 8 patients whose contralateral level I was not a target incorporated the results in a new cost function aiming to spare contralateral SMGs.

Results: Stimulated SMG flow rates decreased exponentially by (1.2%)(Gy) as mean doses increased up to 39 Gy threshold, and then plateaued near zero. At mean doses < or =39 Gy, but not higher, flow rates recovered over time at 2.2%/month. Similarly, the unstimulated salivary flow rates decreased exponentially by (3%)(Gy) as mean dose increased and recovered over time if mean dose was <39 Gy. IMRT replanning reduced mean contralateral SMG dose by average 12 Gy, achieving < or =39 Gy in 5 of 8 patients, without target underdosing, increasing the mean doses to the parotid glands and swallowing structures by average 2-3 Gy.

Conclusions: SMG salivary flow rates depended on mean dose with recovery over time up to a threshold of 39 Gy. Substantial SMG dose reduction to below this threshold and without target underdosing is feasible in some patients, at the expense of modestly higher doses to some other organs.

Fig. 1

Selective collection of submandibular/sublingual saliva from Wharton’s duct orifices.

Fig. 2

Plots of SMG saliva flow rates vs. mean SMG doses at various post-RT time points (1, 3,6,12,18, and 24 months). A. Stimulated, B. Unstimulated salivary flow rates.

Fig. 2

Plots of SMG saliva flow rates vs. mean SMG doses at various post-RT time points (1, 3,6,12,18, and 24 months). A. Stimulated, B. Unstimulated salivary flow rates.

Fig. 3

Plots of the ratios of SMG saliva flow rates relative to pre-therapy base-line flow rates, vs. mean SMG doses, at various post-RT time points (1.3.6.12, 18, and 24 months). Note the logarithmic scale of the flow rate ratios; the horizontal line at 0 represents values near base-line. A. Stimulated, B. Unstimulated salivary flow ratios.

Fig. 3

Plots of the ratios of SMG saliva flow rates relative to pre-therapy base-line flow rates, vs. mean SMG doses, at various post-RT time points (1.3.6.12, 18, and 24 months). Note the logarithmic scale of the flow rate ratios; the horizontal line at 0 represents values near base-line. A. Stimulated, B. Unstimulated salivary flow ratios.

Fig. 4

Mean SMG doses vs. grade IV toxicity at 12 months (salivary flow rate <25% of baseline pre-RT). A. Stimulated, B. Unstimulated. The dots are the average observed toxicities of patients grouped in mean dose clusters at 10 Gy intervals. The bars represent 95% C.I.

Fig. 4

Mean SMG doses vs. grade IV toxicity at 12 months (salivary flow rate <25% of baseline pre-RT). A. Stimulated, B. Unstimulated. The dots are the average observed toxicities of patients grouped in mean dose clusters at 10 Gy intervals. The bars represent 95% C.I.

Fig. 5

Comparison of dose distributions in the original plan (A), and re-planning (B) containing a cost function to reduce mean contralateral (Lt) SMG dose to<39 Gy. Note that the contralateral jugulodigastric (subdigastric, JD) lymph node lies immediately posterior to the contralateral SMG; no PTV under-dosage was therefore allowed while sparing the gland. The ipsilateral (Rt) JD node is involved with gross metastasis.

Fig. 5

Comparison of dose distributions in the original plan (A), and re-planning (B) containing a cost function to reduce mean contralateral (Lt) SMG dose to<39 Gy. Note that the contralateral jugulodigastric (subdigastric, JD) lymph node lies immediately posterior to the contralateral SMG; no PTV under-dosage was therefore allowed while sparing the gland. The ipsilateral (Rt) JD node is involved with gross metastasis.