intensity-modulated radiotherapy in the treatment of nasopharyngeal carcinoma: an update of the ucsf...

PII S0360-3016(02)02724-4

CLINICAL INVESTIGATION Head and Neck

INTENSITY-MODULATED RADIOTHERAPY IN THE TREATMENT OFNASOPHARYNGEAL CARCINOMA: AN UPDATE OF THE

UCSF EXPERIENCE

NANCY LEE, M.D., PING XIA, PH.D., JEANNE M. QUIVEY, M.D., KHALIL SULTANEM, M.D.,IAN POON, M.D., CLAYTON AKAZAWA, C.M.D., PAM AKAZAWA, C.M.D., VIVIAN WEINBERG, PH.D.,

AND KAREN K. FU, M.D.

Department of Radiation Oncology, University of California–San Francisco, San Francisco, CA

Purpose: To update our experience with intensity-modulated radiotherapy (IMRT) in the treatment of naso-pharyngeal carcinoma (NPC).Methods and Materials: Between April 1995 and October 2000, 67 patients underwent IMRT for NPC at theUniversity of California–San Francisco (UCSF). There were 20 females and 47 males, with a mean age of 49(range 17–82). The disease was Stage I in 8 (12%), Stage II in 12 (18%), Stage III in 22 (33%), and Stage IV in25 (37%). IMRT was delivered using three different techniques: 1) manually cut partial transmission blocks, 2)computer-controlled auto-sequencing segmental multileaf collimator (SMLC), and 3) sequential tomotherapyusing a dynamic multivane intensity modulating collimator (MIMiC). Fifty patients received concomitantcisplatinum and adjuvant cisplatinum and 5-FU chemotherapy according to the Intergroup 0099 trial. Twenty-six patients had fractionated high-dose-rate intracavitary brachytherapy boost and 1 patient had gamma kniferadiosurgery boost after external beam radiotherapy.The prescribed dose was 65–70 Gy to the gross tumor volume (GTV) and positive neck nodes, 60 Gy to the clinicaltarget volume (CTV), 50–60 Gy to the clinically negative neck, and 5–7 Gy in 2 fractions for the intracavitarybrachytherapy boost. Acute and late normal tissue effects were graded according to the Radiation Therapy OncologyGroup (RTOG) radiation morbidity scoring criteria. The local progression–free, local-regional progression–free,distant metastasis–free rates, and the overall survival were calculated using the Kaplan–Meier method.Results: With a median follow-up of 31 months (range 7 to 72 months), there has been one local recurrence atthe primary site. One patient failed in the neck. Seventeen patients developed distant metastases; 5 of thesepatients have died. The 4-year estimates of local progression–free, local-regional progression–free, and distantmetastases-free rates were 97%, 98%, and 66% respectively. The 4-year estimate of overall survival was 88%.The worst acute toxicity documented was as follows: Grade 1 or 2 in 51 patients, Grade 3 in 15 patients, andGrade 4 in 1 patient. The worst late toxicity was Grade 1 in 20 patients, Grade 2 in 15 patients, Grade 3 in 7patients, and Grade 4 in 1 patient. At 3 months after IMRT, 64% of the patients had Grade 2, 28% had Grade1, and 8% had Grade 0 xerostomia. Xerostomia decreased with time. At 24 months, only one of the 41 evaluablepatients had Grade 2, 32% had Grade 1, and 66% had Grade 0 or no xerostomia. Analysis of the dose–volumehistograms (DVHs) showed that the average maximum, mean, and minimum dose delivered were 79.3 Gy, 74.5Gy, and 49.4 Gy to the GTV, and 78.9 Gy, 68.7 Gy, and 36.8 Gy to the CTV. An average of only 3% of the GTVand 3% of the CTV received less than 95% of the prescribed dose.Conclusion: Excellent local-regional control for NPC was achieved with IMRT. IMRT provided excellent tumortarget coverage and allowed the delivery of a high dose to the target with significant sparing of the salivary glandsand other nearby critical normal tissues. © 2002 Elsevier Science Inc.

Nasopharynx, Carcinoma, Intensity-modulated, Radiotherapy.

INTRODUCTION

Nasopharyngeal carcinoma (NPC) is common amongAsians, especially the Southern Chinese. However, it israrely seen among the Caucasian population, representingless than 1% of all cancers in the United States (1). Thestandard treatment for NPC is radiotherapy alone for early(T1N0) disease, and combined radiotherapy and chemother-

apy for more advanced lesions, including those with nodalinvolvement or T2–4 disease (2, 3). The local control ratefor American Joint Committee on Cancer (AJCC) 1992Stage T1 and T2 tumors ranged from 64% to 95%. How-ever, the control rate decreased to 44–68% in AJCC 1992Stage T3/T4 tumors. Five-year survival was between 36%and 58% (4–10).

Reprint requests to: Nancy Lee, M.D., Department of RadiationOncology, University of California–San Francisco, 505 ParnassusAvenue, L-08, San Francisco, CA 94143. Tel: (415) 353-8900;

Fax: (415) 353-8679; E-mail: [email protected] Aug 9, 2001, and in revised form Dec 5, 2001.

Accepted for publication Dec 11, 2001.

Int. J. Radiation Oncology Biol. Phys., Vol. 53, No. 1, pp. 12–22, 2002Copyright © 2002 Elsevier Science Inc.Printed in the USA. All rights reserved

0360-3016/02/$–see front matter

12

Tumor control for carcinoma of the nasopharynx ishighly correlated with the dose delivered to the tumor. In aseries of 107 patients with NPC, local control was signifi-cantly improved when �67 Gy was delivered to the tumortarget (11). In another series of 118 patients, the improve-ment of tumor control was attributed not only to the pre-scription of higher doses of radiation, but also to improve-ments in technical accuracy (12). Because the nasopharynxis surrounded by many critical normal tissues, accuracy indose delivery is essential in any dose escalation study.

Intensity-modulated radiation therapy (IMRT) has re-cently gained popularity in the treatment of head-and-neckcancer. With this technique, the intensity of the radiationbeams can be modulated so that a high dose can be deliv-ered to the tumor while significantly reducing the dose tothe surrounding normal tissues (13–16). Xia et al. (17)compared IMRT treatment plans with conventional treat-ment plans for a case of locally advanced NPC. Theyconcluded that IMRT provided improved tumor target cov-erage with significantly better sparing of sensitive normaltissue structures in the treatment of locally advanced NPC.Two recent papers also substantiated this finding. The au-thors stated that because of a lack of major benefit withconventional 3-dimensional treatment planning used onlyduring the boost phase of treatment for NPC, they arecurrently using IMRT to deliver the entire course of radia-tion (18, 19).

At our medical center, we have been using IMRT for thetreatment of NPC since 1995. Local-regional control ratewas 100% in our initial clinical experience of 35 patients(20). These patients were treated primarily with forwardlyplanned IMRT. Over the past 4 years, we have been usinginverse planning for all our patients because this techniqueallows better sparing of the surrounding normal criticalstructures (17). The majority of the radiation beams weredelivered using either computer-controlled auto-sequencingstatic multileaf collimator (MLC) or the Peacock systemusing a dynamic multivane intensity-modulating multileafcollimator, called the MIMiC. In this report, we update ourresults, and describe the evolution of our techniques in thetreatment planning and delivery of IMRT for NPC.

METHODS AND MATERIALS

Patient and staging evaluationBetween April 1995 and November 2000, 67 patients

underwent IMRT for NPC in the Department of RadiationOncology, University of California–San Francisco. For thisanalysis, the records of these 67 patients, including the first35 patients that were reported previously, were reviewedand updated (20).

Pretreatment evaluation included a complete history andphysical examination, direct flexible fiberoptic endoscopicexamination, complete blood counts, liver function tests,chest X-ray, magnetic resonance imaging (MRI) scans ofthe nasopharynx and neck, and dental evaluation. Bone scanand computed tomography (CT) scans of the abdomen or

chest, or both, were obtained when clinically indicated. Thedisease was staged according to the 1997 AJCC stagingclassifications (21).

Radiation treatmentAll patients received external beam radiation therapy.

Twenty-six patients also had high-dose-rate intracavitarybrachytherapy boost at the discretion of the treating physi-cian. Over the past 6 years, several different radiotherapytechniques and IMRT methods have evolved. In our initialexperience, the primary tumor was treated with IMRT. Theupper neck above the vocal cords was irradiated with op-posed-lateral fields. The lower neck and the supraclavicularfossae were treated with a single anterior field using con-ventional radiotherapy. The IMRT field was matched withthe opposed-lateral neck field with a split-beam technique.The opposed-lateral neck field was also matched to thelower neck and supraclavicular field with a split-beam tech-nique. This technique was described previously (20).

A second technique treated the primary and the upperneck above the vocal cords with IMRT and the lower neckand the supraclavicular fossae with an anterior field. Thesetwo fields were matched with a split-beam technique. Thedose uncertainties at the match lines led us to our thirdtechnique in which we used an extended-field IMRT (EF-IMRT) that treated the primary tumor along with all theregional lymph nodes, including the supraclavicular nodes.A detailed description of each of the above techniques willbe published in a separate paper (22).

The prescribed dose was 65–70 Gy to the gross tumorvolume (GTV) and positive neck nodes, 60 Gy to theclinical target volume (CTV) which included the GTV plusa margin of potential microscopic spread, and 50–60 Gy tothe clinically negative neck for the external beam radiother-apy. Our goal was to prescribe 1.8 Gy/fraction/day, 5 days/week to the CTV. The GTV received a higher dose perfraction and was typically 2.12–2.25 Gy/fraction/day. Doseto the neck nodes was delivered at 1.8–2.0 Gy/fraction/day.

Typically, the intracavitary brachytherapy boost occurred1–2 weeks after the external beam radiotherapy. In general,the intracavitary brachytherapy boost was delivered usingthe Rotterdam nasopharyngeal applicator (23) and a remote-controlled high-dose-rate (HDR) afterloader (microSelec-tron, Nucleotron Inc.) with an 192Ir source. The prescribedminimum tumor dose for the intracavitary brachytherapyboost was 5–7 Gy delivered in 2 fractions, 5–6 h apart.

Fifty (75%) patients, the majority of whom had Stage IIIor IV disease also received cisplatin during, and cisplatinand 5-fluorouracil (5-FU) after radiotherapy according tothe Head and Neck Intergroup protocol 0099 (2).

Simulation, immobilization, and treatment planning CTprocedures

The patient’s head position was hyperextended at theinitial simulation so that there was adequate separationbetween the primary and retropharyngeal lymph nodes andthe upper neck field (20). The tip of the uvula and the base

13IMRT of nasopharyngeal carcinoma: UCSF update ● N. LEE et al.

of the occiput was on a plane parallel to the beam axis. Theisocenter on the initial simulation film was placed at theanticipated treatment isocenter. The head, neck, and in somecases the shoulders were immobilized using a thermoplasticmask with the neck supported on a Timo (S-type, MED-TEC) (17, 23). A pair of orthogonal radiographs was takenfor isocenter localization. Treatment planning CT scans inserial 3–5-mm slices from the head down through the clav-icles were obtained.

Delineation of target volumesAll target volumes were outlined slice by slice on the

treatment-planning CT images. GTV was defined as thegross extent of the tumor shown by imaging studies as wellas physical examination, and this included the primarytumor as well as all involved regional lymph nodes. CTVwas defined as the GTV plus a margin for potential micro-scopic spread, including the regional lymph node drainingareas. During the era when the patients in this series weretreated, the planning system, Peacock or Corvus, did notallow specification of asymmetric margin to the GTV or theCTV. Therefore, our GTV and CTV had a built-in PTV soas to account for patient setup errors. The CTV included theentire nasopharynx, retropharyngeal lymph nodal regions,clivus, skull base, pterygoid fossae, parapharyngeal space,inferior sphenoid sinus, and posterior third of the nasalcavity and maxillary sinuses. The surrounding critical nor-mal structures, namely the brainstem, spinal cord, opticnerves, chiasm, parotid glands, temporomandibular (T-M)joints, and middle and inner ears were also outlined. In allcases, we required a recent MRI scan of the nasopharynx todelineate the extent of the tumor. Whenever possible, fusionof the diagnostic MRI images and the treatment planningCT images was done to more accurately delineate the GTVand the surrounding critical normal structures. When sub-optimal fusion occurred, the tumor was outlined on both CTand MRI scans with careful comparison of the two studies.

Treatment planning and deliveryTwo different treatment planning software systems were

used. In our earlier experience, we used a modified CT-based planning system developed at the University of Mich-igan (U-M Plan) (24). Over time, we found that there was asignificant further advantage using inverse treatment-plan-ning systems developed by the NOMOS Corporation, i.e.,the Peacock, Version 1, or Corvus, Version 3.0 and Version4.0 planning systems (24–27). Treatment was deliveredusing either manually cut partial transmission blocks or acomputer-controlled auto-sequence multileaf collimator(MLC) system (Siemens Medical Systems, Concord, CA),or the MIMiC when using the Peacock plan, and dynamicMLC system (Varian Oncology Systems, Palo Alto, CA).The treatment planning and delivery details were describedpreviously (20). Examples of a 10-field forward plan, aninverse Peacock plan, and an inverse Corvus plan are shownin Figs. 1–3.

Dose–volume analysis of treatment plansDose–volume histograms (DVHs) of the GTV and CTV

and the critical normal structures, such as the brainstem,spinal cord, chiasm, optic nerves, middle and inner ears,T-M joints, and the parotid glands were retrospectivelyretrieved from our planning CT scans and analyzed accord-ingly. A total of 65 patients were included in this part of theanalysis. Data were not available for 2 patients. Thirty-threepatients were treated with a forward plan (FP) and 29 wereplanned using an inverse plan (IP). Three patients receivedpart of their treatment with a FP and part of their treatmentwith an IP.

For GTV and CTV, we evaluated the volume receivingless than 95% and 90% of the prescribed dose as quantita-tive endpoints to reflect the tumor target coverage. Thevolume receiving greater than 105% of the prescribed dosewas used to evaluate dose homogeneity. The maximumdose, minimum dose, and the mean dose to the targetvolumes were also calculated.

For the critical organs with functional subunits organizedin series such as the brainstem, spinal cord, optic chiasm,and optic nerves, doses to 5% and 10% of the volumes wereexamined. For critical organs with functional subunits or-ganized in parallel such as the parotids, T-M joints, and themiddle and inner ears, the doses delivered to 50% and 80%of the volume were examined.

Follow-upAll the patients were evaluated at least once a week

during radiotherapy. The patients were then evaluated every1–2 months for the first 6 months, followed by every 3months for the next 6–12 months, every 4–6 months from18 months through 3 years, and annually thereafter. At eachfollow-up visit, a physical examination, including a directflexible fiberoptic endoscopy examination, and palpation ofthe neck was performed. A baseline post-treatment MRIscan of the nasopharynx and neck was obtained within 2–6months after completion of radiotherapy and then yearly orwhen clinically indicated. A positron emission tomography(PET) scan was also obtained in some patients during fol-low-up, generally at least 4 months after therapy. Acute andlate normal tissue effects were graded according to theRadiation Therapy Oncology Group (RTOG) radiation mor-bidity scoring criteria (28).

Statistical methodsThis is a study of a single cohort of consecutively treated

patients with NPC at UCSF. Descriptive statistics (mean,median, and proportions) were calculated to characterize thepatient, disease, and treatment features as well as toxicityafter IMRT. The probability of failure due to local disease,locoregional disease, distant metastasis, and death wereestimated using the Kaplan–Meier product-limit method(29). Durations were calculated from the date of diagnosis.Univariate comparisons of the distribution of outcome wereanalyzed using the log–rank statistic. There are too fewfailure events to perform any multivariate analysis except

14 I. J. Radiation Oncology ● Biology ● Physics Volume 53, Number 1, 2002

for the occurrence of distant metastasis. Cox’s proportionalhazard model was used to identify independent predictors ofdistant failure (30).

RESULTS

Patient characteristicsThere were 20 females and 47 males, with a mean age of

49 (range 17–82). Table 1 shows the T- and N-stage dis-tribution of the patient population according to the 1997AJCC staging classification. The disease was Stage I in 8(12%), Stage II in 12 (18%), Stage III in 22 (33%), andStage IV in 25 (37%). There were 55 Chinese, 3 African-American, 2 Hispanic, 6 white, and 1 Saudi Arabian. Thirty-four patients had nonkeratinizing carcinoma (WHO II), and33 patients had undifferentiated carcinoma (WHO III).There was no patient with WHO I keratinizing carcinoma.

Treatment outcomeThe median follow-up was 31 months (range 7 to 72

months). This analysis of treatment outcome was based onthe follow-up data available as of October 1, 2001. Therehas been one local recurrence at the primary site, at 25months after completion of radiotherapy. This patient pre-sented with T4N1 disease and was not a candidate forchemotherapy due to underlying medical conditions. Onepatient developed a neck recurrence at 13 months after thecompletion of radiotherapy; he remained disease-free aftersalvage neck dissection. Seventeen patients developed dis-tant metastases; 5 of these patients have died. One otherpatient died of uncontrolled epistaxis approximately 6months after completion of radiotherapy, although 1 monthbefore death this patient was clinically free of disease.Unfortunately, the patient’s family refused autopsy. The4-year estimates of local progression–free, local-regionalprogression–free, and distant metastasis–free rates were

Fig. 1. Isodose curves of a 10-field forward IMRT plan for a patient with T1N1 carcinoma of the nasopharynx displayedon the axial (a), coronal (b), and sagittal (c) planes through the centroid of the primary tumor and the DVH for therelevant structures (d).


97%, 98%, and 66%, respectively (Figs. 4–6). The 4-yearestimate of overall survival was 88% (Fig. 7). There were13 patients surviving at least 48 months from diagnosiswithout evidence of disease, with the longest survival being72 months.

Prognostic factorsAt this time, there are too few events to carry out subset

analysis of endpoints for local failure, locoregional failure,or death (1, 1, 6 failures, respectively). With 17 of 67patients experiencing distant failure, only stage of dis-ease was a significant predictor of outcome favoringthose with early-stage (I and II) disease (p � 0.01). Noneof the 20 patients with Stage I or II disease developeddistant metastasis.

Acute and late toxicityAcute side effects of radiation therapy � chemotherapy

were well tolerated. Fifty-one patients had Grade 1 or 2; 15had Grade 3; and 1 had Grade 4 acute toxicity. The most

common acute toxicities were mucositis (47 Grade 2; 15Grade 3; 1 Grade 4) and pharyngitis (48 Grade 2; 15 Grade3; 1 Grade 4).

The late normal tissue effects were scored according tothe RTOG criteria. Table 2 shows the frequency of the worstlate toxicity by type and grade. The worst late toxicity wasGrade 1 in 20; Grade 2 in 15; Grade 3 in 7 patients; andGrade 4 in 1 patient. The most common late effect wasxerostomia. Other late toxicities observed included hearingimpairment (all of whom had chemotherapy) nasopharyn-geal stenosis, TMJ pain on chewing, trismus, and chondro-necrosis of the torus.

Xerostomia appeared to decrease with time after IMRT.Figure 8 shows the frequency of Grade 0, 1, and 2 xerosto-mia over a period of 2 years after completion of radiother-apy. At 3 months after IMRT, 64% of the patients hadGrade 2, 28% had Grade 1, and 8% had Grade 0 or noxerostomia. At 24 months, only 1 of the 41 evaluablepatients had Grade 2, 32% had Grade 1, and 66% had Grade0 or no xerostomia.

Fig. 2. Isodose curves of an inverse IMRT plan delivered using MIMiC for a patient with T4N1 carcinoma of thenasopharynx displayed on the axial (a), coronal (b), and sagittal (c) planes through the centroid of the primary tumorand the DVH for the relevant structures (d). The GTV is shown in red and the CTV is shown in orange.


Dose–volume analysisTable 3 shows the DVH statistics for the target volumes.

Although the prescription dose was 70 Gy to the GTV and60 Gy to the CTV, the mean dose to the GTV was 74.5 Gyand to the CTV was 68.7 Gy. This was because the pre-scribed dose was the minimum dose that encompassed the

Table 1. Distribution of patients by the 1997 American JointCommittee on Cancer Staging Classification

Stage N0 N1 N2 N3 Total

T1 9 7 10 1 27T2 2 2 6 1 11T3 1 4 5 5 15T4 3 7 3 1 14Total 15 20 24 8 67

Fig. 3. Isodose curves of an inverse IMRT plan using 9 coplanar gantry angles delivered with conventional MLC fora patient with T1N0 carcinoma of the nasopharynx displayed on the axial (a), coronal (b), and sagittal (c) planes throughthe centroid of the primary tumor and the DVH for the relevant structures (d). The GTV is shown in red and the CTVis shown in magenta.

Fig. 4. Kaplan–Meier estimate of local progression–free probabil-ity.


tumor target volume. On average, the target volumes hadexcellent coverage and only 3% of the GTV and 3% of theCTV received �95% of the prescribed dose. The majorityof the GTV and CTV actually received more than 105% ofthe prescribed dose. It should be noted that the maximumand minimum doses are point doses, although the maximumdose described by the International Commission on Radia-tion Units and Measurements Report 50 (ICRU 50) is theregion that is encompassed by a 2-cm area.

Table 4 shows the DVH statistics for the critical normalstructures organized in series; Table 5 shows the DVHstatistics for the critical normal structures organized inparallel. There was significant sparing of all critical struc-tures without compromising tumor target coverage. Dose tothese critical organs would be significantly higher if con-ventional radiotherapy were used. The results of the quan-titative analyses of the DVHs for the tumor target andcritical structures were consistent with the clinical results ofexcellent local control and decreased xerostomia. In addi-tion, there was no increase in late tissue toxicity. Of note,excellent target coverage with a high dose could beachieved with either a forward plan or an inverse planmethod. We found that the inverse plan method delivered

lower doses to the surrounding normal tissues than theforward plan method (31).

DISCUSSION

Radiotherapy � chemotherapy is the primary treatmentmodality for NPC, preferred over surgery due to its ana-tomic location. The reported local control rate for T1/T2tumors ranges between 64% and 95% whereas for T3/T4tumors, the control rate drops to 44–68% according to the1992 AJCC staging classification (4–10). Studies haveshown that local control rate increased with dose (11, 12).However, due to the anatomic location of the nasopharynxin proximity to many critical normal tissues, it is ratherchallenging for the radiation oncologist to deliver an ade-quate dose without causing potentially serious complica-tions. For early lesions, studies have shown that the additionof interstitial or intracavitary brachytherapy can further im-prove the local control without increasing morbidity (32–34). However, brachytherapy has its limitations, as it cannoteffectively treat lesions that are more advanced. In thisseries, 26 patients (39%) received brachytherapy boost aspart of their treatment. These patients all presented with T1or T2 lesions. The role of brachytherapy boost in combina-tion with IMRT was evaluated and discussed in a separatestudy (32). Our results suggest that IMRT alone or com-bined with brachytherapy boost is equally effective in con-trolling early-stage disease.

Table 2. Frequency of worst late toxicity by type and grade

Type Grade1 Grade 2 Grade 3 Grade 4

Xerostomia 20 39 — —Hearing loss — — — 5Neuropathy 1 3 — —Trismus 2 — — —Chronic dysphagia — — — 1Subcutaneous fibrosis 5 — — —Soft tissue necrosis — — — 1Chondronecrosis of

the torus — — 1 —

Fig. 5. Kaplan–Meier estimate of local-regional progression–freeprobability.

Fig. 6. Kaplan–Meier estimate of distant-metastasis–free probabil-ity.

Fig. 7. Kaplan–Meier estimate of overall survival.


For advanced lesions with tumor extension into the skullbase, other means such as stereotactic radiosurgical boost orchemotherapy have been used to improve local control (2,35). Tate et al. (35) published their experience with stereo-tactic radiosurgical boost with a mean dose of 12 Gy afterconventional radiotherapy with a dose of 66 Gy in 23patients with primary NPC. With a median follow-up of 21months, a local control rate of 100% was achieved. It shouldbe noted that this study accrued only 23 patients over aperiod of 6 years. Twelve patients had T4 lesions; in theother patients, the T-stage was not specified. Only patientswith residual volumes �30 cm3 after external beam radia-tion therapy were candidates for the stereotactic boost.Therefore, this study population represents a highly selec-tive group of patients. Unfortunately, because stereotacticradiosurgery was used only as a boost, all the patients hadpermanent xerostomia.

Wolden et al. (18) recently published the Memorial

Sloan-Kettering Cancer Center experience in the treatmentof NPC. Between 1988 and 1998, 68 patients receivedconventional external beam radiotherapy followed by a 3Dconformal radiotherapy boost. The median follow-up ofthese patients was 42 months (range 12 to 118 months).Fifty-two percent of the patients also received chemother-apy. Their 5-year actuarial local control rate was 77% witha progression-free survival of only 56%. Patients withStages I–III disease had a progression-free survival of 65%vs. 40% for those with Stage IV disease. When compared totheir historical series of 33 patients treated with 2D tech-niques, the local control rate of 77% was not significantlyimproved over the 75% local control rate with the noncon-formal 2D technique. They concluded that perhaps the rea-son for a lack of improvement with the 3D plan was that itwas only used for a portion of the treatment. Currently, theyare using IMRT to treat all patients with NPC.

Intensity-modulated radiation therapy allows the deliveryof a very high dose to the tumor target while sparing thesurrounding critical structures. In addition, IMRT allows thedelivery of a higher dose per fraction which results in aTable 3. Dose–volume histograms (DVHs) statistics for target

volumes

GTV average(range)

CTV average(range)

Volume (cc) 104 (10–669.2) 301 (82–1248)Maximum dose (Gy) 79.3 (65.8–93.8) 78.9 (66.4–94.3)Mean dose (Gy) 74.5 (63.3–94.6) 68.7 (61.0–82.8)Minimum dose (Gy) 49.4 (16.2–71.5) 36.8 (12.5–61.1)% volume receiving �95%

of the prescribed dose 2.7 (0–93) 3.4 (0–17)% volume receiving �90%

of the prescribed dose 1.6 (0–66) 2.1 (0–12.3)% volume receiving �105%

of the prescribed dose 75.5 (0–99) 87 (33–100)

Table 4. Dose–volume statistics derived from DVHs for serialnormal critical structures

Organ

Dose to5% volume (Gy)Average (range)


Brain stem 46.3 (26.6–67) 43.5 (17.6–65)Spinal cord 36.5 (9.8–46.3) 30 (1.9–45.7)Chiasm 28.7 (3.6–55.7) 26.9 (3.5–54)Optic nervesRight 25.7 (8–67.7) 23.3 (2.9–65.8)Left 21.5 (7.5–61.8) 20.6 (7.5–62)

Fig. 8. Frequency of Grade 0–2 xerostomia at various times after radiotherapy (RT).


biologically more effective dose to the tumor target. There-fore, IMRT offers a significant advantage over conventionalconformal therapy. The nasopharynx is a site where IMRTcan play a major role in improving the therapeutic ratio. Thelack of organ motion and its central location makes thenasopharynx ideally suited for IMRT. A high dose can begiven to the nasopharynx while sparing critical normalstructures such as the brainstem, the spinal cord, the opticchiasm, and salivary glands. The advantages of IMRT havepopularized its use in the treatment of NPC.

Sultanem et al. (20) reported our initial experience of 35patients treated with IMRT between April 1995 and March1998. With a median follow-up of 21 months, the local-regional progression–free rate was 100%. In the presentseries, we included an additional 32 patients treated withIMRT between April 1998 and October 2000 and updatedthe follow-up for the original series. To date, we haveobserved only one local failure. This patient had T4N1disease and was not a candidate for concurrent chemother-apy due to underlying medical conditions. The excellentlocal control reported in our initial series is substantiatedwith additional patients and longer follow-up in this updatedseries. Only 1 patient failed in the upper neck. This patientreceived IMRT for his primary tumor. His neck was treatedwith conventional radiotherapy using opposed-lateral fieldsfor his upper neck nodes, and an anterior field for his lowneck as well as the supraclavicular fossae. The patientunderwent a neck dissection at 13 months after treatmentand has been disease-free since then.

Although excellent local control has been achieved withIMRT, unfortunately, patients still failed distantly. It isdisappointing that the chemotherapy regimen used in theIntergroup trial is not sufficient in preventing distant me-tastasis. A pilot study at UCSF (36) suggests a significantcorrelation between microvessel density and risk of distantmetastasis and survival in patients irradiated for NPC. Thissuggests that anti-angiogenic agents may have a role in thetreatment of this disease. Future trials using more effectivechemotherapy or anti-angiogenic agents may further im-prove the outcome of NPC patients.

Accurate delineation of the target volumes is probablyone of the most critical steps in IMRT treatment planning.

With regard to GTV, it is relatively more straightforward. Inour patients, the GTV included all the visible tumor (pri-mary as well as regional grossly enlarged lymph nodes)seen on the MRI, CT, or palpated on physical examination.MRI is more sensitive than CT scan as it is more sensitivein demonstrating, for example, the extent of tumor involve-ment of the base of skull. In our patients, we obtained MRIscans for every patient unless there was medical contrain-dication. Fusion of the diagnostic MRI scan and the treat-ment planning CT scan was performed whenever feasible toaccurately delineate the target volumes. A major difficultythat a radiation oncologist faces in treatment planning is thedefinition of CTV. What should the physician outline on thetreatment planning CT scans to ensure that all the areas ofpotential microscopic spread and the lymphatics are ade-quately covered? Recently, several papers examining theprecise definition of the lymphatics of the head-and-neckregion have been published (37–42). These nodal atlasguidelines may be helpful to radiation oncologists in delin-eating the CTVs during treatment planning. When doubtarises concerning the appropriate CTV, the lymph nodaltargets included in the traditional opposed-lateral fieldsshould be referenced.

Accurate and precise delivery of radiation beams is ab-solutely essential in IMRT. Therefore, adequate patient im-mobilization is critical in IMRT. In addition, our ability totailor the radiation beams in IMRT such that they con-formed to the tumor target shape three-dimensionally willrequire further research focusing on the improvement ofmechanical precision in linear accelerators. A study done atUCSF showed that both the GTV and the CTV were notvery sensitive to the random patient motion and setup errors,up to 3 mm (43). Under these circumstances, the averagechange in dose to sensitive structures was also not signifi-cant. However, consistent setup errors could introduce sig-nificant change in dose to sensitive structures such as thebrainstem and the chiasm (44). Several studies have exam-ined better immobilization techniques by using an individ-ualized bite block device which provides accurate and re-producible patient positioning for conformal therapy in thehead-and-neck region (45). At our institution, several dif-ferent techniques of immobilization have been used includ-ing thermoplastic head masks and head, neck, and shouldermasks. Skin toxicity was observed depending on whichimmobilization technique was used, especially when head,neck, and shoulder masks were used (46). A comparison ofthe different immobilization techniques will be reported in aseparate paper (22).

All of our patients tolerated IMRT well. Acutely, themost common complaints were mucositis and pharyngitiswith RTOG Grade 3 toxicity (requiring a temporary feedingtube) observed in 22%. This is not unexpected in view of thehigh doses to the mucosal regions with 50% of the CTVreceiving �116% of the prescribed dose of 60 Gy. (Table3). Excessive skin toxicity was also observed in some of ourpatients, especially those who had EF-IMRT which simul-taneously treated the primary and all the lymphatic regions

Table 5. Dose–volume statistics derived from DVHs for parallelnormal critical structures

Organ



ParotidsRight 34.8 (7.9–72.3) 24.6 (5.5–67.7)Left 33.9 (13.4–60) 24.4 (6.2–52.8)

T-M jointsRight 49.0 (19.4–79) 44.3 (16.5–78.2)Left 49.5 (24.9–78.3) 43.1 (21.1–74)

EarsRight 49.5 (12.8–74.4) 42.4 (8.8–72)Left 48.7 (20.2–64.3) 42.9 (12.5–60.6)


including the supraclavicular fossae. Multiple factors mayhave contributed to the increased skin reaction in patientstreated with EF-IMRT. However, by taking into consider-ation the skin as a sensitive structure during inverse plan-ning, it is possible to reduce the skin dose to within toler-ance without compromising tumor target coverage (46).

One of the major complaints from patients treated withconventional external beam radiation therapy to the naso-pharynx is xerostomia due to a high dose to the majorsalivary glands bilaterally. Salivary flow is markedly re-duced after 10–15 Gy of radiation delivered to most of thegland (47–49). The recovery of the salivary function ispossible over time even with doses up to 40–50 Gy. How-ever, higher doses to most of the gland will result in irre-versible and permanent xerostomia. As a result, patients’quality of life is compromised as they experience changes inspeech and taste. The oral dryness also predisposes thepatients to fissures, ulcers, dental caries, infection, and inworst cases, osteoradionecrosis (50–53). The degree ofxerostomia is largely dependent on the radiation dose andthe volume of the salivary gland that is in the radiation field.IMRT is capable of reducing the dose to the salivary glandswhile simultaneously delivering a high dose to the tumortarget. In our series, the mean dose to 50% of the parotidswas 34 Gy. Figure 8 shows the frequency of Grade 0, 1, and2 xerostomia over a follow-up period of 2 years after thecompletion of radiotherapy. Sixty-four percent of the pa-tients complained of Grade 2 xerostomia 3 months afterradiotherapy, but by 24 months, only 1 patient out of the 41evaluable patients complained of Grade 2 xerostomia. Itappears that patients treated with IMRT recover their sali-vary flow faster and more completely than those treatedwith conventional radiotherapy.

In general, the treatment was very well tolerated without

significant severe late complications. In addition, nearly100% of our patients completed 3 cycles of adjuvant che-motherapy consisting of 5-fluorouracil and cisplatinum inaddition to the 3 cycles of cisplatinum during radiotherapy,whereas in the Intergroup trial only 65% of the patients inthe combined radiotherapy and chemotherapy group re-ceived 3 cycles of cisplatinum during radiotherapy. Further-more, only 55% of the patients completed 3 courses of theadjuvant chemotherapy and 45% of the patients received 2cycles or less adjuvant chemotherapy due to toxicity (2).Compared to the Intergroup trial, compliance to adjuvantchemotherapy was improved in patients treated with IMRT.

Our results with IMRT in the treatment of NPC showedexcellent local control with decreased normal tissue effects,in particular, xerostomia, and improved compliance to ad-juvant chemotherapy. Based on the UCSF experience, aPhase I/II study to evaluate the efficacy of IMRT in thetreatment of NPC in a multi-institutional setting is beingdeveloped by the Radiation Therapy Oncology Group(RTOG). We also plan to investigate whether the addition ofanti-angiogenic agents to combined IMRT and chemother-apy will decrease distant metastasis and improve survival.

CONCLUSION

Excellent local-regional control for NPC was achievedwith intensity-modulated radiotherapy. IMRT provided ex-cellent tumor target coverage and allowed the delivery of ahigh dose to the target with significant sparing of the sali-vary glands and other nearby critical normal tissues. Futuretrials evaluating more effective chemotherapy regimens �anti-angiogenic agents in decreasing the rate of distantmetastasis in NPC are warranted.

REFERENCES

1. Yu MC. Diet and nasopharyngeal carcinoma. Prog Clin BiolRes 1990;346:93.

2. Al-Sarraf M, LeBlanc M, Shanker Giri PG, et al. Chemora-diotherapy versus radiotherapy in patients with advanced na-sopharyngeal cancer: Phase III randomized intergroup study0099. J Clin Oncol 1998;16:1310–1317.

3. Cooper JS, Lee H, Torrey M, et al. Improved outcome secondaryto concurrent chemoradiotherapy for advanced carcinoma of thenasopharynx: Preliminary corroboration of the intergroup expe-rience. Int J Radiat Oncol Biol Phys 2000;47:861–866.

4. Chu A, Flynn MB, Achino E, et al. Irradiation of nasopha-ryngeal carcinoma: Correlations with treatment factors andstage. Int J Radiat Oncol Biol Phys 1984;10:2241–2249.

5. Hoppe RT, Goffinet DR, Bagshaw MA. Carcinoma of thenasopharynx: Eighteen years’ experience with megavoltageradiation therapy. Cancer 1976;37:2605.

6. Mesic JB, Fletcher GH, Goepfert H. Megavoltage irradiationof epithelial tumors of the nasopharynx. Int J Radiat OncolBiol Phys 1981;7:447–453.

7. Sanguineti G, Geara F, Garden A, et al. Carcinoma of thenasopharynx treated by radiotherapy alone: Determinants oflocal and regional control. Int J Radiat Oncol Biol Phys1997;37:983–996.

8. Bailet JW, Mark RJ, Abemayor E, et al. Nasopharyngealcarcinoma: Treatment results with primary radiation therapy.Laryngoscope 1992;102:965–972.

9. Vikram B, Strong EW, Manolatos S, et al. Improved survivalin carcinoma of the nasopharynx. Head Neck Surg 1984;7:123–128.

10. Wang CC. Carcinoma of the nasopharynx. In: Wang CC,editor. Radiation therapy for head and neck neoplasms: Indi-cations, techniques, and results. 2nd ed. Chicago: Year BookMedical Publishers; 1990. p. 261–283.

11. Marks JE, Bedwinek JM, Lee F, et al. Dose-response analysisfor nasopharyngeal carcinoma. Cancer 1982;50:1042–1050.

12. Vikram B, Mishra UB, Strong EW, et al. Patterns of failure incarcinoma of the nasopharynx: I. Failure at the primary site.Int J Radiat Oncol Biol Phys 1985;11:1455–1459.

13. Nutting C, Dearnaley DP, Webb S. Intensity-modulated radi-ation therapy: A clinical review. Br J Radiol 2000;73:459–469.

14. Butler EB, The BS, Grant WH, et al. Smart (simultaneousmodulated accelerated radiation therapy) boost: A new accel-erated fractionation schedule for the treatment of head andneck cancer with intensity modulated radiotherapy. Int J Ra-diat Oncol Biol Phys 1999;45:21–32.


15. Eisbruch A, Ship JA, Martel MK, et al. Parotid gland sparingin patients undergoing bilateral head and neck irradiation:Techniques and early results. Int J Radiat Oncol Biol Phys1996;36:469–480.

16. Eisbruch A, Ten Haken RK, Kim HM, et al. Dose, volume,and function relationships in parotid salivary glands followingconformal and intensity-modulated irradiation of head andneck cancer. Int J Radiat Oncol Biol Phys 1999;45:577–587.

17. Xia P, Fu KK, Wong GW, et al. Comparison of treatmentplans involving intensity modulated radiotherapy for nasopha-ryngeal carcinoma. Int J Radiat Oncol Biol Phys 2000;48;329–337.

18. Wolden SL, Zelefsky MJ, Hunt MA, et al. Failure of a 3Dconformal boost to improve radiotherapy for nasopharyngealcarcinoma. Int J Radiat Oncol Biol Phys 2001;49:1229–1234.

19. Hunt MA, Zelefsky MJ, Wolden S, et al. Treatment planningand delivery of intensity-modulated radiation therapy for pri-mary nasopharynx cancer. Int J Radiat Oncol Biol Phys 2001;49:623–632.

20. Sultanem K, Shu H-K, Xia P, et al. Three-dimensional inten-sity-modulated radiotherapy in the treatment of nasopharyn-geal carcinoma: The University of California–San Franciscoexperience. Int J Radiat Oncol Biol Phys 2000;48:711–722.

21. Fleming ID, S CJ, Henson DE, et al. AJCC manual for stagingof cancer. 5th ed. Philadelphia: J.B. Lippincott; 1997. p.31–39.

22. Xia P, Phillips TL, Lee T, et al. Comparison of differenttreatment techniques involving intensity-modulated radiationtherapy in the treatment of head and neck carcinoma. Manu-script in preparation.

23. Levendag PC, Peters R, Meeuwis CA, et al. A new applicatordesign for endocavitary brachytherapy of cancer in the naso-pharynx. Radiother Oncol 1997;45:95–98.

24. Thornto AF Jr, Ten Haken RK, Gerhardsson A, et al. Three-dimensional motion analysis of an improved head immobili-zation system for simulation, CT, MRI, and PET imaging.Radiother Oncol 1991;20:224–228.

25. Lichter AS, Sandler HM, Robertson JM, et al. Clinical expe-rience with three dimensional treatment planning. Semin Ra-diat Oncol 1992;2:257–266.

26. Carol MP. Integrated 3-D conformal multivane intensity-mod-ulation delivery system for radiotherapy. In: Hounsell AR,Wilkinson JM, Williams PC, editors. XIth International Con-ference on the Use of Computers in Radiation Therapy.Manchester, UK: Christie Hospital, NHS Trust; 1994. p. 172–173.

27. Carol MP. Peacock. A system for planning and rotationaldelivery of intensity modulated fields. Int J Imag Sys Tech1995;6:56–61.

28. Cox JD, Stetz J, Pajak TF. Toxicity criteria of the RadiationTherapy Oncology Group (RTOG) and the European Organi-zation for Research and Treatment of Cancer (EORTC) [Ed-itorial]. Int J Radiat Oncol Biol Phys 1995;31:1341–1346.

29. Kaplan EL, Meier P. Nonparametric estimation from incom-plete observations. J Am Stat Assoc 1958;53:457–481.

30. Cox DR. Regression models and life-tables. J R Stat Soc1972;34:187–220.

31. Poon I, Lee N, Akazawa P, et al. A DVH analysis of naso-pharynx cancer patients treated with inverse planned intensitymodulated radiation therapy. Manuscript in preparation.

32. Lee N, Hoffman R, Phillips TL, et al. Managing nasopharyn-geal carcinoma with intracavitary brachytherapy: One institu-tion’s 45 year experience. J Brachyther. In press.

33. Syed AM, Puthawala AA, Damore SJ, et al. Brachytherapy forprimary and recurrent nasopharyngeal carcinoma: 20 years’experience at Long Beach Memorial. Int J Radiat Oncol BiolPhys 2000;47:1311–1321.

34. Chang JT, See LC, Tang SG, et al. The role of brachytherapy

in early-stage nasopharyngeal carcinoma. Int J Radiat OncolBiol Phys 1996;36:1019–1024.

35. Tate DJ, Adler JR, Chang SD, et al. Stereotactic radiosurgicalboost following radiotherapy in primary nasopharyngeal car-cinoma: Impact on local control. Int J Radiat Oncol Biol Phys1999;45:915–921.

36. Roychowdhury DF, Tseng A, Fu KK, Weinberg V, WeidnerN. New prognostic factors in nasopharyngeal carcinoma. Can-cer 1996;77:1419–1426.

37. Gregoire V, Coche E, Cosnard G, et al. Selection and delin-eation of lymph node target volumes in head and neck con-formal radiotherapy: Proposal for standardizing terminologyprocedure based on the surgical experience. Radiother Oncol2000;56:135–150.

38. Som P, Curtin HD, Mancusco AA. Imaging-based nodal clas-sification for evaluation of neck metastatic adenopathy. AJR2000;174:837–844.

39. Nowak PJCM, Wijers OB, Lagerwaard FJ, et al. A three-dimensional CT-based target definition for elective irradiationof the neck. Int J Radiat Oncol Biol Phys 1999;45:33–39.

40. Nowak P, Van Dierene, Vander Est H, et al. Treatment portalsfor elective radiotherapy of the neck: An inventory in theNetherlands. Radiother Oncol 1997;43:81–86.

41. Som PM, Curtin HD, Mancuso AA. An imaging-based clas-sification for the cervical nodes designed as an adjunct torecent clinically based nodal classification. Arch OtolaryngolHead Neck Surg 1999;125:388–396.

42. Martinez-Monge R, Fernandes PS, Gupta N, et al. Cross-sectional nodal atlas: A tool for the definition of clinical targetvolumes in three-dimensional radiation therapy planning. Ra-diology 1999;211:815–828.

43. Xia P, Chuang C, Nguyen-Tan F, et al. Computer simulatedpatient motion and setup uncertainties in intensity modulatedradiotherapy. In: Schlegel W, Bortfeld T, editors. The use ofcomputer in radiation therapy, XIIIth International Confer-ence. Heidelberg, Germany: Springer; 2000. p. 317.

44. Gilbeau L, Octave-Prignot M, Loncol T, et al. Comparison ofsetup accuracy of three different thermoplastic masks for thetreatment of brain and head and neck tumors. Radiother Oncol2001;58:155–162.

45. Willner J, Hadinger U, Neumann M, et al. Three dimensionalvariability in patient positional using bite block immobiliza-tion in 3D-conformal radiation treatment for ENT tumors.Radiother Oncol 1997;43:315–321.

46. Lee N, Chuang C, Quivey JM, et al. Methods of reducingacute skin toxicity due to extended-field intensity-modulatedradiation therapy (EF-IMRT) for the treatment of head andneck carcinoma. In press.

47. Eisbruch A, Kim HM, Terrell JE, et al. Xerostomia and itspredictors following parotid-sparing irradiation of head-and-neck cancer. Int J Radiat Oncol Biol Phys 2001;50:695–724.

48. Leslie MD, Dische S. The early changes in salivary glandfunction during and after radiotherapy given for head and neckcancer. Radiother Oncol 1994;30:26–32.

49. Mira JG, Wescott WB, Starcke EN, Shannon IL. Some factorsinfluencing salivary function when treating with radiotherapy.Int J Radiat Oncol Biol Phys 1981;7:535–541.

50. Harrison LB, Zelefsky MJ, Pfitzer DG, et al. Detailed qualityof life assessment in patients treated with primary radiother-apy for cancer of the base of tongue. Head and Neck 1997;19:169–175.

51. Balogh JM, Sutherland SE. Osteoradionecrosis of the mandi-ble: A review. J Otolaryngol 1989;18:245–250.

52. Cooper JS, Fu K, Marks J, Silverman S. Late effects ofradiation therapy in the head and neck region. Int J RadiatOncol Biol Phys 1995;31:1141–1164.

53. Shannon IL, Trodahl JN, Wescott WB. Effects of radiotherapyon whole saliva flow. J Dent Res 1977;56:693.


intensity-modulated radiotherapy in the treatment of nasopharyngeal carcinoma: an update of the ucsf...

Documents