An audit of Organ at Risk (OAR) constraints achieved for carcinoma of the nasopharynx treated with intensity modulated radiotherapyGreeshma KE, Geetha Muttath*, Joneetha Jones, Nabeel Yahia, Vinin N V, Arun P Narendran and Akhil P Suresh
Geetha Muttath, Department of Radiation Oncology, Malabar Cancer Centre, Thalassery Kannur, Kerala, India, Email: firstname.lastname@example.org
Received: 06-Jun-2021 Accepted: 21-Jun-2021 Published: 28-Jun-2021
Background: In Nasopharyngeal Carcinomas (NPC), due to its anatomical location and radio sensitivity, the primary treatment modality is radical radiotherapy. Intensity Modulated Radiotherapy (IMRT) provides excellent loco regional control and sparing of Organs at Risk (OARs) and it has become the technique of choice for radiotherapy of NPCs. Still late toxicities can occur in up to 40% of patients. The present study analyses the organ at risk doses achieved in patients with NPC treated with IMRT. Materials and Methods: A retrospective audit of NPC treated with IMRT from January 2013 to August 2018 was done. The prescription dose for PTV HR, PTV IR and PTV LR were 69.3 Gy, 59.4Gy and 54 Gy respectively in 33 fractions. Concurrent chemotherapy was added for patients with stage II and above with Cisplatin 100 mg /m2 every 21 days. OAR constraints were restricted to the tolerance doses as per the recommendations. The data was analysed for the degree of adherence to the recommended dose volume constraints for OARs and the correlation of achieved OAR doses against Gross Tumour Volume (GTV) of primary, nodes and total GTV was analysed using Pearson’s correlation coefficient. Results: Plans of 40 patients were analysed. Adequate target dose coverage (D95 for PTV HR, IR and LR) was achieved in the majority (93% of patients for PTV HD, 100% of patients for PTV ID and 98% of patients for PTV LD) of our patients. More than 80% of patients had met the dose constraints for brainstem, spinal cord, v69 of temporal lobe, v75 of mandible, eyes, optic chiasm and optic nerves. The achieved doses for parotids and temporal lobes in particular were higher. Significant positive correlation was noted for OARs close to the primary site against GTV primary and GTV total. Conclusions: Adherence to the recommended dose volume constraints were achieved for optic and neuronal structures close to the primary site as well as for mandible in majority of patients. But a higher priority needs to be given for parotids and temporal lobes during radiation treatment planning.
nasopharyngeal carcinoma, IMRT, OAR doses
Nasopharyngeal cancers comprise less than 1% of cancers in India . The primary treatment modality for Nasopharyngeal Carcinomas (NPC) is radical radiotherapy as the anatomic location of the cancer provides limited surgical access. Moreover, these are relatively radiosensitive cancers. Intensity Modulated Radiotherapy (IMRT) has evolved as the technique of choice when compared with conventional two-dimensional radiotherapy, as it provides superior normal tissue sparing without compromising on disease control [2,3].
The severity of adverse effects of radiotherapy is related to the dose to organs at risk (OARs) . Although OARs sparing has improved significantly with IMRT, in up to 40% of patients, grade 2-4 xerostomia and sensorineural hearing loss can still occur [5, 6]. Total radiation doses and fraction size contribute to the development of radiation toxicities.
QUANTEC (Quantitative Analyses of Normal Tissue Effects in the Clinic) guidelines are widely used to guide radiation tolerance limits to organs at risk. As the Gross Tumour Volume (GTV) increases the adherence to QUANTEC guidelines can become difficult. In this study we analysed the OAR doses achieved in patients undergoing IMRT for nasopharyngeal carcinoma, their degree of adherence to the recommended dose volume guidelines, and the relationship between OAR dose constraints achieved and GTV volume.
This is a retrospective analysis of patients with nasopharyngeal carcinoma treated at our hospital with Volumetric Modulated Arc Therapy (VMAT) between January 2013 and August 2019. All patients were planned for radical chemo radiotherapy and underwent contrast CT scan for treatment planning. The scans were acquired with slice thickness of 3 mm after immobilising the patient in the planned treatment position, on a dedicated CT simulator (Optima GE). Three Clinical Target Volumes (CTV) were defined as follows. High Risk CTV (CTV HR) including GTV primary and GTV nodes with 1 cm expansion (edited for the normal anatomical barriers), intermediate risk CTV (CTV IR) including GTV primary with 1.5 cm expansion (edited for natural anatomical barriers), the nasopharynx, posterior one third of the nasal cavity, posterior one third of the maxillary sinus, pterygoid fossae, parapharyngeal spaces, sphenoid sinus (the entire sinus was included only in patients with intracranial extension), foramen ovale, foramen rotundum, and clivus depending on the extent of tumour infiltration. The CTV IR was extended superiorly if there is intracranial extension, and included the cavernous sinuses and occasionally adjacent brain. The nodal regions with gross nodes were also included in the CTV IR. Low risk CTV (CTV LR) included retropharyngeal lymph nodes, bilateral cervical lymph nodes from level II to V in addition to the CTV IR. Three Planning Target Volumes (PTV) were defined. PTV HD (high dose PTV), PTV ID (intermediate dose PTV) and PTV LD (low dose PTV), by expanding the corresponding CTVs by 5 mm. Sample cases depicting the target volumes are shown in figure 1 and figure 2.
Fig. 1. (A) Axial section of planning CT scan image at the level of nasopharynx showing planning target volumes (PTV). (B) Axial section of planning CT scan image at the level of gross nodes showing PTVs. PTV HD- High Dose PTV, PTV ID -intermediate dose PTV, PTV LD-low dose PTV, GTV P- gross tumour volume primary, GTV N- gross tumour volume nodes
Fig. 2. (A) Sagital section of planning CT scan image showing Planning Target
Volumes (PTV). (B)Coronal section of the planning CT scan depicting PTVs.
PTV HD- high dose PTV, PTV ID -intermediate dose PTV, PTV LD-low dose
PTV, GTV P- gross tumour volume primary, GTV N- gross tumour volume
After delineation of target and OARs, treatment planning was done with VMAT, using the SIB (Simultaneous Integrated Boost) technique (Eclipse Version 13.6, Varian). The dose prescribed was 69.3 Gy to PTV HD, 59.4Gy to PTV ID and 54Gy to PTV LD given in 33 fractions. Concurrent chemotherapy was given with Inj. Cisplatin at a dose of 100 mg/m2 every 21days.
Dose volume coverage for PTVs were approved and accepted in accordance with the recommendations from ICRU (International Commission on Radiation Units & Measurements). Ideal coverage is the one where 95% of the PTV receives 100% of the prescribed dose (v95 100%). This is not achievable in all cases and hence a dose distribution where 95% of the PTV received 95% of the dose (v95 95%) was considered acceptable. Dose Volume Histograms (DVH) of each OAR was evaluated, in accordance with the QUANTEC, RTOG 0255 (Radiation Therapy Oncology Group), and RTOG 0615 recommendations, and the plans satisfying the criteria was approved for treatment. The demographic and clinical data including age, gender, TNM stage, and treatment details were collected from patient’s case records. Treatment plans and DVH parameters of PTVs and OARs were retrieved from the treatment planning system (Varian Eclipse).
The patients were categorised according to their total Gross Tumour Volume (GTV) into four groups <30 cc, 30 cc-60 cc, 60 cc-90 cc and >90 cc. The average dose received by OARs in the study population as a whole, and in the four groups were analysed. The degree of adherence to the recommended dose constraint guidelines were analysed according to whether the dose constraint achieved for each OAR is less than or equal to 100%, between 100% and 110% or more than 110% of the recommended dose.
Descriptive statistics were used for expressing the demographic and treatment details. Pearson's correlation coefficient test was used to find out the correlation between dose volume parameters of OARs with GTV primary, GTV nodes and GTV total in the 4 GTV groups and the study population as a whole. The data was analysed using the SPSS statistical software, version 20.0.
Between January 2013 and August 2019, a total of 40 patients with NPC were treated with radical radiotherapy at our centre. The median age was 51.5y (19y - 70y). Among the 40 patients, 22 (55%) patients had a clinical stage 4 disease (Table 1). All patients received concurrent chemotherapy with Cisplatin 100 mg/m2 every 21 days.
|Age||Median 51.5y(19y- 70y)|
|Age group||<40 yrs.||7||17.5|
|Dose prescription||69.3 Gy/33#||37||92.5|
Tab. 1. Demographic and treatment details of patients with nasopharyngeal carcinoma treated with concurrent chemo radiation
Among the 40 patients, 37 (92.5%) patients were prescribed a PTV Dose of 69.3 Gy/33# and remaining 3 patients (7.5%), 70 Gy/35#. Majority of patients achieved the planning goals of target coverage of at least v95 of 95% (%volume receiving 95% of the prescribed dose of 69.3Gy, 59.4Gy and 54Gy to PTV HD, ID and LD respectively). The target coverage of at least v95 of 95% was achieved for 93% of patients for PTV HD, 100% of patients for PTV ID and 98% of patients for PTV LD. Average v95 was 97.8% for PTV HD, 99% for PTV ID and 97% for PTV LD.
The Dose Maximum (Dmax) was 115% of the prescribed dose of 69.3Gy which occurred in one patient, but the v115 (%volume receiving 115% of the prescribed dose) was only <1% of the irradiated volume. All other patients had a Dmax less than 115% of the prescribed dose (108% to 114.7% of the prescribed dose of 69.3Gy). Only 4 patients had v110 ((% volume receiving 110% of the prescribed dose) >1% (Table 1).
In order to find out the degree of adherence of dose volume constraints achieved for each OAR to the recommended dose volume parameter, and to analyse the excess dose received by each OAR, the dose received by each OAR was classified into whether the achieved dose is </= 100%, 101%-110% or >110% of the recommended dose. Table 2 shows the OAR doses achieved as a percentage of the recommended dose.
|OAR||DoseVolume Constraint||≤ 100% n(%)||101%-110% n(%)||>110% n(%)|
|Brain stem||#Dmax 54Gy||32 (80)||5 (12.5)||3 (7.5)|
|Spinal cord||#Dmax 45Gy||38 (95)||1 (2.5)||1 (2.5)|
|Temporal lobe(r)||#Dmax 68Gy||14 (35)||18 (45)||8 (20)|
|*V69<1cc||34 (85) (V69<1cc)||6 (15)(V69>1cc)|
|Temporal lobe (l)||#Dmax 68Gy||12 (30)||16 (40)||12 (40)|
|*V69<1cc||26 (65) (V69<1cc)||14(35)(V69>1cc)|
|Mandible||#Dmax 70Gy||4 (10)||32(80)||4(10)|
|Lens(r)||#Dmax 7Gy||10 (25)||2(5)||28(70)|
|Lens(l)||#Dmax 7Gy||6 (15)||5(12.5)||29(72.5)|
|Eye(r)||**Dmean 35Gy||38 (95)||1(2.5)||1(2.5)|
|Eye(l)||**Dmean 35Gy||38 (95)||1(2.5)||1(2.5)|
|Parotid(r)||**Dmean 25Gy||3 (7.5)||2 (5)||35(87.5)|
|##V30- 50%||21 (52.5)||4 (10)||15(37.5)|
|Parotid(l)||**Dmean 25Gy||3 (7.5)||1 (2.5)||36(90)|
|##V30-50%||11 (27.5)||3 (7.5)||26(65)|
|Optic chiasm||#Dmax 55Gy||34 (85)||6 (15)||0|
|Larynx||#Dmax 44Gy||20 (50)||9 (22.5)||11(27.5)|
|Oral cavity||$$D1cc(Gy)<70||18 (45)(D1<70Gy)||22 (55)(D1>70 Gy)|
|Pituitary||#Dmax 45Gy||6 (15)||2 (5)||32 (80)|
|Optic nerve(r)||#Dmax 55Gy||35 (87.5)||1 (2.5)||4 (10)|
|Optic nerve(l)||#Dmax 55Gy||36 (90)||4 (10)||0|
|Constrictors||**Dmean 50Gy||2 (5)||1 (2.5)||37 (92.5)|
|Middle ear(r)||**Dmean 45Gy||18 (45)||4 (10)||18 (45)|
|Middle ear(l)||**Dmean 45Gy||21(52.5)||4 (10)||15 (37.5)|
|Inner ear(r)||**Dmean 4Gy||18 (45)||4 (10)||18 (45)|
|Inner Ear(L)||**Dmean 45Gy||20 (50)||4 (10)||16 (40)|
|Esophagus||###V45<33%||9 (22.5)||0||31 (77.5)|
|Brainstem PRV#$||#Dmax 60Gy||12 (30)||18(45)||10 (25)|
|Spinal cord PRV #$||#Dmax 50Gy||11(27.5)||23 (57.5)||6 (15)|
#Dmax- Dose maximum, *V69-volume receiving 69Gy, $V75-volume receiving 75Gy, **Dmean- dose mean, ##V30-volume receiving 30Gy, $$D1cc (Gy)-dose to 1cc volume, ###V45-volume receiving 45Gy, #$ PRV- planning organ at risk volume
Tab. 2. Dosimetric parameters and adherence of OAR dose constraints against recommended dose in patients with nasopharyngeal carcinomas
More than 80% of patients had achieved a dose of less than or equal to the recommended dose constraints for brainstem, spinal cord, temporal lobe, mandible, eyes, optic chiasm and optic nerves.
In more than 70% of the patients, the lens received 110% of the recommended dose of 700cGy. Dose to parotids was more than 110% of the recommended dose of Dmean (dose mean) of 25Gy, in more than 85% of the patients. In more than half of the patients, the v30 to 50% of the parotids was more than 110%. Similarly, the doses were more than 110% of the recommended dose in the majority of patients for oral cavity, pituitary, pharyngeal constrictors and esophagus.
The relationship of DVH parameters of OARs with the GTVs including GTV primary (GTV P), GTV nodes (GTV N) and total GTV (GTV T) were analysed. Patients were grouped into 4 groups based on the total GTV (GTV T) volume into Group 1 <30cc, Group 2 30-60cc, Group 3 60-90cc and Group 4 >90cc. The mean doses achieved for OARs in the study population as a whole, and in the four GTV groups are detailed in table 3.
|Mean OAR doses(Gy)|
|OAR||DoseVolume constraint||Total# (n= 40)||Group1# (n=11)||Group2# (n=13)||Group3# (n=12)||Group4# (n=4)|
|Brain stem||#*Dmax 54 Gy||51.86||50.95||50.74||51.76||58.34|
|Spinal cord||#*Dmax 45 Gy||41.96||42.01||40.34||44.22||40.24|
|#*Dmax 68 Gy||70.62||70.58||69.63||71.1||72.48|
|#*Dmax 68 Gy||69.45||68.11||67.6||72.41||70.31|
|Mandible||#*Dmax 70 Gy||74.61||75.17||74.25||74.84||73.57|
|Lens(R)||#*Dmax 7 Gy||14.29||14.82||12.05||13.2||23.4|
|Eye(R)||**Dmean 35 Gy||15.24||16.75||10.63||15.25||26|
|Eye(L)||**Dmean 35 Gy||14.77||17.08||10.41||14.61||23.02|
|Parotid (R)||**Dmean 25 Gy||39.77||37.89||37.91||41.65||45.3|
|Optic chiasm||#*Dmax 55 Gy||3998||4233||2805||4597||5427|
|Larynx||#*Dmax 44 Gy||4588||4716||4305||4526||5338|
|#*Dmax 30 Gy||47.26||4846||4684||4649||4761|
|Oral cavity||$$D1cc(Gy) <70||71Gy||71.5Gy||71Gy||70Gy||71.5Gy|
|Optic nerve(r)||#*Dmax55 Gy||39.08||39.98||32.32||4321||48..84|
|Optic nerve(l)||#*Dmax55 Gy||39.15||43.61||30.11||4223||4705|
|Middle ear(L)||**Dmean45 Gy||44.86||44.5||4332||45.93||47.65|
|Inner ear(R)||**Dmean45 Gy||47.95||48.11||4690||50.61||42..95|
|Inner ear(L)||**Dmean45 Gy||47.32||48.28||4577||52.16||35.25|
|Brainstem PRV#$||#*Dmax60 Gy||62.79||62.63||62.65||62.68||64.27|
|Spinalcord PRV#$||#*Dmax50 Gy||52.87||51.91||51.13||55.2||54.18|
#Total GTV (GTV T) volumes : Group 1 <30cc, Group 2 30-60cc, Group 3 60-90cc, Group 4 >90cc. #* Dmax-dose maximum, *V69-volume receiving 69Gy, $V75-volume receiving 75Gy, **Dmean- dose mean, ##V30-volume receiving 30Gy, $$D1cc(Gy)-dose to 1cc volume, ###V45-volume receiving 45Gy, #$ PRV- planning organ at risk volume
Tab. 3. Mean doses achieved for OARs in patients with nasopharyngeal carcinomas against GTV
The OARs which showed a significant positive correlation with the three GTVs in the study population as a whole and in the 4 GTV groups are shown in table 4. For the study population as a whole, significant positive correlation was noted with GTV primary for brainstem, brainstem PRV (planning organ at risk volume), spinal cord, spinal cord PRV, lens, parotids v30, pituitary and optic nerves. A positive correlation was also noted for eyes against GTV total. In group 1, a significant positive correlation was noted for lens, eyes and spinal cord PRV against GTV primary and parotid v30 against GTV total. In group 2, significant positive correlation was noted for eyes, optic nerve and spinal cord PRV against GTV primary and for mandible, oral cavity D1(Gy) and optic nerve against GTV total. In group 3, significant positive correlation was noted for optic nerve and brainstem PRV against GTV primary; There was also a significant positive correlation between eyes and oral cavity D1cc (Gy)[dose to 1cc volume] with GTV total in this group. In group 4, no positive correlation was noted for any OARs against GTV primary, GTV nodes and GTV total.
|GTV group||GTV primary||GTV total|
|Organ||p value||Organ||p value|
|Group1 (GTV total<30cc)||Lens||0.02|
|Spinal Cord prv||0.02|
|Group2 (GTV total30-60cc)||Eyes||0.036||Mandible||0.009|
|Optic nerve||0.04||Oral cavity D1cc(Gy)||0.027|
|Spinal cord prv||0.02||Optic nerve||0.04|
|Group3(GTV total 60-90cc)||Optic nerve||0.04||Eyes||0.047|
|Brain stem prv||0.03||Oral cavity D1cc(Gy)||0.036|
|Group 4 (GTV Total>90cc)||None||None|
|Study Population As A Whole||Brain stem||0.02||Eye||0.03|
|Brain stem prv||0.003|
Tab. 4. OARs with significant positive correlation with GTV groups in patients with nasopharyngeal carcinoma
Though IMRT is the widely used advanced radiation technique for the management of nasopharyngeal carcinoma, its effectiveness in sparing OARs around nasopharynx such as temporal lobe, parotid and cochlea is largely unclear. The data on the relationship between GTV and excess rates of dose to the OARs are also sparse. We did this study to analyse the dose received by organs at risk in patients receiving radiotherapy for nasopharyngeal cancer, the achieved degree of adherence to recommended dose constraints, and also the dose distribution achieved in target volumes.
In the majority of our patients, the desired target coverage was achieved for PTV HD, ID and LD. The target coverage of at least v95 of 95% was achieved for 93% of patients for PTV HD, 100% of patients for PTV ID and 98% of patients for PTV LD. Only 4 patients had a v110 of more than 1%. All except one patient had a dose less than 115% of the prescribed dose of 69.3Gy.
The dose distribution of OARs with respect to various GTVs in NPC patients treated with IMRT was prospectively analysed by Ji-Jin Yao et al in China . The study showed that, with a larger GTV the radiation dose to OARs increased significantly, and GTV was a useful predictor of radiation dose to OARs around the nasopharynx. The OARs like spinal cord,optic nerve, mandible, TM joint, eye, oral cavity and pharynx Constrictors were able to tolerate radiation dose easily.
In our study, recommended doses were achieved for OARs around the nasopharynx (brainstem, spinal cord, optic chiasm, optic nerves, eyes and mandible) in more than 80% of the patients. One reason for the above result could be the high priority given for these OARs, especially the neuronal structures, during radiation treatment planning. Positive significant correlations were noted for OAR doses with volume of GTV primary for those OARs close to the primary site. Hence, as the volume of GTV primary increases, the doses to OARs around the primary site increases. GTV total also correlated positively with the OARs around the primary site.
No significant positive correlations were observed for all OARs with GTV nodes except for brainstem PRV in group 3 patients.
Another observation in our study was that both dose volume constraints for parotids including Dmean and v30 <50% were not achieved in the majority of our patients. One reason for not achieving the dose volume constraint could be the close proximity of the GTVs to parotids, especially so when the primary is large and when there are enlarged jugulodigastric lymph nodes. Assigning a higher priority for the parotids during radiation treatment planning might be effective in reducing parotid doses.
The factors influencing the parotid function in NPC treated with parotid sparing radiotherapy were studied by Wen-Shan Liu et al. They observed that the mean dose to the parotid gland was the most important factor that influenced parotid function. The parotid function could recover one year after parotid sparing radiotherapy .
The use of parotid sparing IMRT for preserving the parotid function for NPCs was also studied by Ching-Yeh Hsiung et al. . They used salivary scintigraphy to quantitatively analyse preserved parotid function after IMRT and compared with historical data after conventional radiotherapy. A significant dose-function relationship was noted for the parotid gland. Significant preservation of parotid function was achieved with IMRT for NPC patients.
Another organ which needs more attention is the temporal lobe. A potentially lethal complication of the central nervous system that can occur in patients treated with radiotherapy for NPC is the temporal lobe radiation necrosis .The incidence of brain radiation necrosis increases as doses exceed 60 Gy in conventional fractionation. In our study, only 35% of patients achieved a dose of less than or equal to 100% of the recommended maximum dose of 68Gy. However, the constraint of v69<1cc was achieved in 85% of the patients. Hence, a higher priority should be considered for temporal lobes during radiation treatment planning for NPC.
The dose volume parameters of oral cavity, pharyngeal constrictors, pituitary and esophagus were more than 110% of the recommended dose in the majority of patients. But no significant positive correlation could be observed for these OARs against GTV primary, GTV nodes or GTV total. One of the limitations of our study is the small number of patients which is probably the reason for not observing a positive correlation for these instances. Since carcinoma of nasopharynx is rare compared to endemic areas, multicentric studies may be required in order to observe a statistically significant correlation.
In our study, adequate target dose coverage was achieved in the majority of patients for PTV HD, PTV ID as well as PTV LD. Adherence to the recommended dose volume constraints were noted for neuronal structures close to the primary site as well as for eyes and mandible in the majority of patients. Significant positive correlation was also noted for OARs close to the primary site with GTV primary and GTV total. The achieved doses for parotids and temporal lobes in particular were higher. Hence, a higher priority needs to be given for parotids and temporal lobes during radiation treatment planning.
- Source: Globocan 2020 - Global Cancer Observatory. 2020
- Kam MK, Teo PM, Chau RM, Cheung KY, Choi PH, et al. Treatment Of Nasopharyngeal Carcinoma With Intensity- Modulated Radiotherapy: The Hong Kong Experience. Int J Radiat Oncol Biol Phys. 2004;60:1440-1450.
- Lin S, Pan J, Han L, Zhang X, Liao X, et al. Nasopharyngeal Carcinoma Treated With Reduced-volume Intensity-modulated Radiation Therapy: Report On The 3-year Outcome Of A Prospective Series. Int J Radiat Oncol Biol Phys. 2009;75:1071-1078.
- Mao YP, Yin WJ, Guo R, Zhang GS, Fang JL, et al. Dosimetric Benefit To Organs At Risk Following Margin Reductions In Nasopharyngeal Carcinoma Treated With Intensity-modulated Radiation Therapy. Chin J Cancer. 2015; 34:189-197.
- La wrence YR, Li XA, Naqa I, Hahn CA, Marks LB, et al. Radiation dose-volume effects in the brain. Int J Radiat Oncol Biol Phys. 2010;76:S20-S27.
- Kam MK, Leung SF, Zee B, Chau RM, Suen JJ, et al. Prospective randomized study of intensity-modulated radiotherapy on salivary gland function in early-stage nasopharyngeal carcinoma patients. J Clin Oncol. 2007;25:4873-4879.
- Ji-Jin Y, Fo-Ping C, Guan-Qun Z, Zhang WJ, Xu L, et al, A prospective study on radiation doses to organs at risk (OARs) during intensity-modulated radiotherapy for nasopharyngeal carcinoma patients. Oncotarget. 2016;7:21742-21752.
- Liu W, Lee S, Lee J, Su M, Chen G, et al Factors Influencing the Parotid Function in Nasopharyngeal Carcinoma Treated with Parotid-Sparing Radiotherapy. Japanese J Clin Oncol. 2006;36:626-631.
- Hsiung C, Ting H, Huang H, Lee C, Huang E, et al. Parotid-sparing intensity-modulated radiotherapy (IMRT) for nasopharyngeal carcinoma: Preserved parotid function after IMRT on quantitative salivary scintigraphy, and comparison with historical data after conventional radiotherapy. Int J Radiat Oncol Biol Phys. 2006;66:454-461.
- Dropcho EJ. Neurotoxicity of radiation therapy. Neurol Clin. 2010; 28:217-234.