three-dimensional radiotherapy of head and neck and esophageal carcinomas: a monoisocentric...

Three-Dimensional Radiotherapy of Head andNeck and Esophageal Carcinomas: A

Monoisocentric Treatment Technique to AchieveImproved Dose Distributions

Munir Ahmad, Ph.D. and Ravinder Nath, Ph.D.Department of Therapeutic Radiology, Yale-New Haven Hospital and Yale University School

of Medicine, Hew Haven, CT 06504 and Radiation Therapy Center, William W. BackusHospital, Norwich, CT 06360

SUMMARY The specific aim of three-dimensional conformal radiotherapy is to deliveradequate therapeutic radiation dose to the target volume while concomitantly keeping thedose to surrounding and intervening normal tissues to a minimum. The objective of thisstudy is to examine dose distributions produced by various radiotherapy techniques usedin managing head and neck tumors when the upper part of the esophagus is also involved.Treatment planning was performed with a three-dimensional (3-D) treatment planningsystem. Computerized tomographic (CT) scans used by this system to generate isodosedistributions and dose-volume histograms were obtained directly from the CT scanner,which is connected via ethernet cabling to the 3-D planning system. These are usefulclinical tools for evaluating the dose distribution to the treatment volume, clinical targetvolume, gross tumor volume, and certain critical organs. Using 6 and 18 MV photonbeams, different configurations of standard treatment techniques for head and neck andesophageal carcinoma were studied and the resulting dose distributions were analyzed.Film validation dosimetry in solid-water phantom was performed to assess the magnitudeof dose inhomogeneity at the field junction. Real-time dose measurements on patientsusing diode dosimetry were made and compared with computed dose values. With regardto minimizing radiation dose to surrounding structures (i.e., lung, spinal cord, etc.), themonoisocentric technique gave the best isodose distributions in terms of dose uniformity.The mini-mantle anterior-posterior/posterior-anterior (AP/PA) technique producedgrossly non-uniform dose distribution with excessive hot spots. The dose measured on thepatient during the treatment agrees to within ± 5 % with the computed dose. The pro-tocols presented in this work for simulation, immobilization and treatment planning ofpatients with head and neck and esophageal tumors provide the optimum dose distribu-tions in the target volume with reduced irradiation of surrounding non-target tissues, andcan be routinely implemented in a radiation oncology department. The presence of areal-time dose-measuring system plays an important role in verifying the actual delivery ofradiation dose. Int. J. Cancer (Radiat. Oncol. Invest.) 96, 55–65 (2001). © 2001 Wiley-Liss, Inc.

Key words: 3-D treatment planning; head and neck and esophageal carcinomas; im-proved dose distributions

Correspondence to: Munir Ahmad, Ph.D., Radiation Therapy Center, The William W. Backus Hospital, 326 WashingtonStreet, Norwich, CT 06360. Phone: (203) 688-4261; Fax: (203) 688-8682; E-mail: [email protected]

Received 8 June 1999; Revised 16 November 1999; Accepted 17 November 2000. Published online 6 February 2001.

Int. J. Cancer (Radiat. Oncol. Invest):96, 55–65 (2001)© 2001 Wiley-Liss, Inc. Publication of the International Union Against Cancer

INTRODUCTION

The evolution of treatment planning systems utiliz-ing computed tomography (CT) and magnetic reso-nance imaging (MRI) modalities has tremendouslyenhanced our ability to develop three-dimensional(3-D) conformal irradiation techniques for thetreatment of cancer. The Photon Treatment Plan-ning Collaborative Group published a comprehen-sive document evaluating the role of 3-D treatmentplanning in the treatment protocols for eight differ-ent anatomical sites [1]. This study evaluated vari-ous aspects related to 3-D treatment planning in-cluding calculation algorithms, dose volumehistograms, uncertainty analysis, numerical scoringof treatment plans, and inhomogeneity corrections.

The clinical importance of 3-D planning ofhead and neck cancers arises from the need to de-liver high but uniform therapeutic doses to local-ized target volumes in close proximity to normalstructures, which can produce acute and long-termmorbidity. Multiple problems occur in treatmentplanning at this site due to rapidly changing patientcontours from the head to the neck to the thorax [2].Additionally, the effect of tissue inhomogeneity,such as the air in the upper aerodigestive tract ondose delivered to the localized volumes, must betaken into account [3]. For these reasons, sophisti-cated 3-D treatment planning is very important forhead and neck tumors in which definitive radiationtherapy plays the major curative role.

There have been several studies [4–11] focus-ing on the treatment for carcinomas of the head andneck, but little has been reported on a practicalapproach to treat head and neck tumors in whichthe superior part of the esophagus is also involved.Traditionally, standard 3-field techniques, consist-ing of bilateral neck fields and a low anterior su-praclavicular field with a midline block to elimi-nate overlap of multiple fields on the spinal cord,would underdose central structures at risk for re-currence. The radiation oncologist is faced with thetechnical challenge of treating the entire neck,treating a low-lying primary site, such as theesophagus, while avoiding overlap of multiplefields on the spinal cord. One should also avoidmatching lateral and anterior fields at the tumor sitesince a “cold spot” would underdose tissue at riskand a “hot spot” could damage sensitive structuresclose to the primary site. Sohn et al. [12] havedescribed an elegant 3-field isocentric beam ar-rangement using asymmetric collimators and asingle isocenter that gives a uniform dose to theneck and supraclavicular lymph nodes.

Previously, Ahmad et al. [13] studied isodose

distributions related to different half-field irradia-tion techniques for treating pituitary adenoma. Inthis report, we present a comparative treatmentplanning study of head and neck and esophagealcarcinomas using (1) a parallel-opposed mini-mantle anterior-posterior/posterior-anterior (AP/PA) technique, (2) a monoisocentric technique totreat the head and neck tumors bilaterally with half-field parallel-opposed beams while treating theesophagus using an AP/PA split field, or with acombination of an AP and two posterior obliquefields. The effect of beam energy on dose distribu-tions was also examined. The comparative plansare evaluated in terms of dose uniformity in theclinical target volume and dose maximum “hotspots” in the irradiated non-target tissue volume.We also studied dose inhomogeneity at the junctionas a result of over- or underlapping of matchingfields to evaluate the tolerance of digital display ofour asymmetric collimators.

MATERIALS AND METHODSThe routine use of a 3-D radiation treatment plan-ning system generally involves a series of tradi-tional and CT-assisted procedures shown in Table 1and is discussed in the following section.

Simulation and CTScanning ProceduresA conventional simulator with fluoro capability(Varian Ximatron XC) is used in our institution todetermine the isocenter and all treatment param-eters. For esophageal and head and neck cancers,patients are set up supine on the Vac-Lok immobi-lization cushion (MED-TEC, Inc., Orange City,Iowa) on the simulator flattop couch with theirheads immobilized using thermoplastic casts. Anorthogonal pair of films, anterior-posterior (AP)and lateral (LAT) are taken, and triangulationmarks, showing the entry points of anterior andlateral fields are tattooed on the patient’s skinthrough cutaway portions of the head and neckthermoplastic cast. The isocenter is defined on themid-plane at the C6-C7 junction. This arrangementallows the user to design independent superior andinferior half fields for treating head and neck, andesophageal carcinomas, independently with no di-vergence at the spinal cord.

Following the simulation, the patient under-goes CT scanning (GE High-Speed Advantage spi-ral scanner, GE Medical Systems, Cincinnati,Ohio) in the planned treatment position using thesame immobilization devices and reference local-ization marks as per simulation. Radiopaquemarker wires are placed on the patient’s skin simu-lating the planned isocenter position on the CT

56 Ahmad and Nath: 3-D RTP of Head and Neck and Esophageal Carcinomas

scans. To produce high-resolution 3-D reconstruc-tion, contiguous CT images with a 5-mm slicethickness ranging from C1 to T12 are obtained.These CT images are transferred via ethernet to the3-D treatment planning system to initiate planningand dose calculation process.

Delineation of Target Volume andCritical Structures

The appropriate tumor volumes for the head andneck and the esophagus treatment sites are identi-fied on each relevant CT slice. The adjacent normalstructures, including lungs and spinal cord, are out-lined by drawing contours around these structureson each CT slice where they appear. The targetvolume and normal organ images are reconstructedin 3-D and displayed in the beam’s eye view(BEV), using wire-frame graphics to outline sur-faces and color to differentiate between structures.This beam’s eye view concept [14] is an effectivetool in identifying optimum gantry, collimator, andcouch angles at which to irradiate the target, whileminimizing the inclusion of healthy critical organsin the high-dose region by interactively moving pa-tient and treatment field. Furthermore, it enablesthe planning physicist/dosimetrist to project a 2-Dimage of 3-D disease extension, especially in ob-lique fields, thus facilitating the design of custom-ized blocking and inadvertent blocking of tumorand unnecessary irradiation of normal tissues.

Beam Arrangementand Characteristics

The coplanar treatment techniques, namely, a mini-mantle AP/PA arrangement; a pair of half-beambilateral head and neck fields and an AP/PA esoph-ageal half-beam setup, and two parallel-opposedhalf-beam lateral neck fields and a 3-field, i.e., AP/RPO/LPO arrangement, have been investigated.These techniques use 6 MV photons for head andneck and 6 and 18 MV photons for esophagealtreatment from a dual energy linear accelerator

(2100C/D, Varian Associates, Inc., Palo Alto, Ca-lif.). The field sizes in each plan are chosen to givean adequate margin around the target volume. TheCT images are used for producing digitally recon-structed projections (DRPs), which are superim-posed onto simulator films. The latter are used astemplates for designing cerrobend blocks and forverification of actual treatment port films. Beamapertures in all plans are shaped by digitizingblocks drawn on the simulation films. All theseplans use beam apertures designed with amodified1-cm margin around the target volumes.

Treatment Planning System andDose Calculation

The treatment plans are designed on the 3-D RTPsystem (Precision Therapy, Inc., Columbia, Md.),which uses an irregular field algorithm, modified tocalculate dose distributions in an external-beamprogram environment. Scatter integration is accom-plished with a two-dimensional integration over thearea of the field so that the shadow and transmis-sion values of attenuators, which might be in thebeam, are taken into account. Inhomogeneity cor-rections are applied for the primary component ofthe dose only, by using the equivalent path lengththrough the patient to the point of calculation fromtable lookup of the primary component of the dose.Corrections are also made to the primary compo-nent for both change in beam quality off axis andchange in beam fluence off axis. The effect ofwedges is accounted for by reference to an in-air

Table 1. Flowchart for 3-D Treatment Planning and Delivery

Simulation Patient is simulated in the planned treatment position, simulation films are taken, and immobilization devices aredesigned. Patient positioning marks are tattooed.

CT Scanning Patient is CT scanned using the immobilization devices in the planned treatment position. The range andthickness of CT slices used are as specified in the CT protocol.

CT Images Transfer The CT images are downloaded into the 3-D planning system via ethernet.Delineating volumes The target volume and critical structures are defined on each CT slice for planning purposes.Treatment Planning An optimized isodose plan is generated and evaluated using dose-volume histograms; Beam’s eye view (BEV)

display is used to design cerrobend blocks.Treatment Verification Patient starts treatment according to the approved plan; port films are obtained to verify patient setup.

On-patient diode dosimetry is performed within the first three treatments to verify the treatment delivery.Dosimetry is repeated once a week for assurance of continuity of optimal setup and beam delivery conditions.

Table 2. Dose Variation as a Function ofField Gap

Field Gap % Dose at Isocenter

2 mm overlap 1211 mm overlap 109No Gap 1001 mm underlap 902 mm underlap 81

Ahmad and Nath: 3-D RTP of Head and Neck and Esophageal Carcinomas 57

fluence function that is derived from wedge pro-files measured in water. A beam-hardening correc-tion is accomplished for the effect of attenuatorsthat might be in the beam, such as wedges, blocks,and compensators, by reference to measured trans-mission-through-water data to adjust the primarycomponent.

The prescription doses are independently de-livered to the two mid-sagittal off-axis points, oneselected in the head and neck tumor volume, andthe other in the center of the esophagus tumor vol-ume. Khan et al. [15] has described a system ofmonitor unit calculations for an off-axis prescrip-tion point in the open portion of the asymmetricfield and is given by:

MU =DP~d,r!

k.TMR~d,rd!.Sc~rc!.Sp~rd!.~SDcal/SPD!2.OARd~x!.TF,

(1)

whereDP(d,r) is the prescription dose specified at

point P at depthd in the open portion of the field;k is the calibration dose perMU at the referencepoint of calibration at a distanceSDcal from thesource and depthdcal; TMR(d,rd) is the tissue-maximum ratio at depthd for the equivalent squarefield of dimensionrd at pointP, Sc(rc) is the colli-mator scatter factor for the equivalent field of sizerc (ignoring asymmetry of the field);Sp(rd) is thephantom scatter factor at the calibration depth for asymmetric field of dimensionrd; SPDis the sourceto point P distance;OAR(x)is the off-axis ratio atpoint P at a distancex from central axis for a sym-metrical field; andTF is the transmission factor forany beam attenuator between the collimator endand the patient surface.

The efficacy of target coverage and dose to thesurrounding normal tissues are evaluated by iso-dose surface distributions on mid plane axial, sag-ittal and coronal CT images. The target dose isprescribed to the maximum isodose surface distri-bution that completely covers the target volume(ICRU “minimum dose”) [16]. The dose gradient

Fig. 1. Dose distribution of 6 MV photons for a mini-mantle anterior-posterior/posterior-anterior (AP/PA) field arrangement.Isodoses are superimposed on the 3-D reconstructed: a) mid-coronal and b) mid-sagittal planes. 120%, 115%, 110%, 100%,and 70% isodoses are drawn. The 100% isodose is displayed in color-wash format. The asterisk (*) shows the maximum dosepoint (“hot spot”) for each plane.


within the target volume for an acceptable treat-ment plan should be kept between 5% to 8% de-pending on the treatment site and field arrangementused.

Match-Plane Dose VerificationClinac 2100C/D accelerator with a set of asymmet-ric collimator was used for this study to define thehalf-beam fields longitudinally. To determine thedosimetric tolerance at the match-line, an XV2 film(Kodak, Rochester, NY) in ready pack was sand-wiched horizontally between two 5-cm thick and30 × 30 cm2 solid water slabs. The isocenter wasplaced at 5-cm depth on the film. The anterior fieldwas defined by closing the inferior half of theasymmetric jaws [Y14 0 cm, Y24 8 cm, and X4 12 cm]. Note that this field was set up by usingthe digital displays on the gantry and was verifiedby visually checking the jaw position on the film.One exposure of 50 monitor units was given fromthis field. Then the superior half of the field was

closed, and the inferior half was opened [Y14 8cm, Y24 0 cm, and X4 12 cm] to give a secondexposure using the same monitor units. The dose inthe match plane was measured by setting the asym-metric collimator (Y2) of AP half fields to 0 mm, ±1 mm, and ± 2 mmposition. These measurementswere performed using 10 MV photons. The filmswere scanned at 5-cm depth in the superior-inferiordirection. The measured percentage dose at the iso-center as a function of field-setup uncertainty isgiven in Table 2.

Plan and Treatment VerificationTo confirm a reproducible and clinically effica-cious daily dose delivery,in vivo dosimetry wasperformed for every head and neck patient using an8-channel microprocessor-controlled semi-conductor diode system (PC-Rainbow, Model 37-703, Nuclear Associates, Carle Place, NY). The useof semi-conductor diodes for patient dose verifica-tion has been discussed in the literature [17–18].

Fig. 1. Continued


For each treatment site (the head and neck andesophagus), one diode was placed on the skin sur-face at the appropriate CT level containing thepoint of dose prescription. The entrance and exitdoses were determined from the diode readings us-ing the equations:

Dentrance 5 Fentrance 3 Rentrance 3 )i Ci, (2)

Dexit 5 Fexit 3 Rexit 3 )i Ci, (3)

where Dentrance and Dexit are beam entrance andexit doses,Rentrance and Rexit are the diodereadings at the beam entrance and exit surfaces,Fentrance and Fexit are the dose calibration factorsfor the diodes, andPi Ci represents a combinedcorrection factor which accounts for various treat-ment parameters including SSD, wedges, patientseparation, beam aperture, and beam incidence

angle, etc. The product of known parameters in theequation gives this factor:

)i Ci 5 Cfield size 3 Cgantry angle 3 Cwedgex . . .(4)

From the measured entrance and exit doses fromeach beam, the dose at the prescription point iscomputed using the tissue-maximum ratio(TMR)tables. The treatment verification index,ƒ, is de-fined by the ratio:

f 5 Dmeasured/Dcalc (5)

Alternatively, the corrected diode reading, whichrepresents the entrance or exit dose, can be directlycompared with the calculated dmax dose from thetreatment plan. An acceptable value ofƒ should fallbetween 1 ± 0.05.

Fig. 2. A half-beam anterior-posterior/posterior-anterior (AP/PA) irradiation technique using 6 MV photons with 15° wedgeon anterior (ANT) for treating esophageal carcinoma. Dose distributions are displayed on the a) mid-coronal and b) mid-sagittal planes. 108%, 102%, 100%, 98%, 90%, and 80% isodoses are drawn. The 100% isodose is displayed in color-washformat. The asterisk (*) shows the maximum dose point (“hot spot”) for each plane.


RESULTS AND DISCUSSION

Isodose distributions of 6 MV photons in the mid-coronal and mid-sagittal planes for the mini-mantletreatment plan using equal beam weights are shownin Fig.1. The prescription dose to the isocenter isselected to encompass the target volume with the100% isodose line displayed in color-wash format.Our comparison of dose distributions for 6 MV and18 MV photons (not shown) demonstrates the factthat the 100% isodose volume of 6 MV photons isgreater than that of 18 MV photons because ofreduced skin sparing, as expected. In other words,the integral dose to the surrounding tissue would beless with the 18 MV photon beams. The 3-D dosedistributions of 6 MV photons result in large 130%“hot spots of 2 cm2” anteriorly and laterally in theneck as compared to 120% for 18 MV photons. Thefield apertures are designed to treat the primary siteto approximately 45 Gy while keeping the corddose under tolerance.

Doppke et al. [8] used a weighting of AP/PA4:1 with a full-length spinal cord block to achievethe desired dose limits. Their studies indicate that ifthe primary site is at risk, a lateral boost is requiredto bring the anterior midline dose to adequate lev-els. However, this can result in a very high dose tothe soft tissue of the neck. Therefore, a mini-mantletechnique involving irregular field apertures has aninherent problem in that it produces excessivelyinhomogeneous dose distributions and one is facedwith the monumental task of boosting a region ofpotential failure while delivering minimal dose tothe spinal cord.

In the monoisocentric technique, the treatmentfield is split in two half-beam sections: head andneck is treated bilaterally with 6 MV photons, andthe esophagus is treated with a half-beam AP/PA orAP/LPO/RPO field arrangement using 6 or 18 MVphotons. The mid-coronal and sagittal isodose dis-tributions resulting from 6 MV photons to treat

Fig. 2. Continued


esophageal carcinoma with AP/PA are shown inFig. 2. It is found that dmax doses (“hot spots of2 cm2”) for 6 and 18 MV (not shown) are 110 and105%, respectively, indicating a slight differentialsparing of normal tissue with higher energy pho-tons. Following approximately 45 Gy via AP/PAfields, additional dose is delivered using two orfour oblique fields to spare spinal cord.

The mid-coronal and sagittal isodose distribu-tions of 6 MV photons resulting from the AP/LRO/RPO field arrangement are shown in Fig. 3. Equalbeam weights and 30° wedges on the obliquebeams deliver uniform doses to the target volumes.Unlike the half-field AP/PA esophageal treatmentplans, the isodose distributions of 6 and 18 MVphotons are dosimetrically similar. The “hot spots”on the order of 112% are produced by the two-photon energies. Therefore, either of the two-

photon beams can be safely used for the treatmentof patients with esophageal tumors.

Transverse isodose distributions of 6 MV pho-tons resulting from bilateral head and neck half-field arrangement, superimposed on various CT im-ages, are displayed in Fig. 4. Use of 15° wedgesand a 1-cm bolus produces uniform doses to thetarget volumes. The additional dose to neck nodesis delivered by appropriately redesigning the fieldapertures.

The results of match-line film dosimetryacross the adjacent radiation fields are listed inTable 2 showing that the dose inhomogeneityranged from 81% to 121% as a result of under- oroverlapping of fields. When the fields were abutted(no gap), a uniform dose of 100% was measured atthe match-line. For ± 1 mm over- or underlappingof fields, about 10% dose variation was observed.

Fig. 3 A half-beam, 3-field, anterior-posterior/left posterior oblique/right posterior oblique (AP/LPO/RPO) irradiation tech-nique using 6 MV photons for treating esophageal carcinoma. Dose distributions are displayed on the a) mid-coronal and b)mid-sagittal planes. Oblique fields use 30° wedges. 110%, 105%, 100%, 90%, 80%, and 50% isodoses are drawn. The 100%isodose is displayed in color-wash format. The asterisk (*) shows the maximum dose point (“hot spot”) for each plane.


When the over- or underlapping was increased to± 2 mm, it produced a 21% overdose or 19% un-derdose, respectively. Therefore, to achieve an ac-ceptable uniform dose within ± 10% at the junction,systematic errors related to mechanical accuracy ofthe jaw alignment with the mechanical isocenter fornondiverging fields must be minimized. This de-mands a tighter tolerance of less than ±1 mm on thedigital display of asymmetric collimators.

The magnitude of “hot” and “cold” spot at thefield match-line can be substantially affected by therandom errors resulting from daily setup variationsand patient motion during treatment. Especially, re-duction in the setup error along the superior-inferior direction is very important to achieve ac-ceptable dose uniformity across the match-line.This translation shift is defined as the differencebetween the positions of the isocenter as seen in thelongitudinal direction on the simulation radiographand the corresponding treatment verification portfilm. As mentioned earlier, torso of all head andneck patients are fixed into the Vac-Lock immobi-lization cushion with their heads immobilized using

thermoplastic masks in order to minimize patientmotion [19]. Several authors [20,21] have investi-gated the effect of setup uncertainties (referred toasrandom errors) on the radiation therapy of headand neck patients. Hunt et al. [21] have observednet positional error of less than 1 mm in the supe-rior-inferior direction. Since random errors and sys-tematic errors are uncorrelated, they are combinedin quadrature to quantify the composite error. It isbelieved that overall setup and localization errorscan be reduced to 1.5-mm levels by using effectivepatient fixation devices and by maintaining stifftolerances on the geometric accuracy of asymmet-ric jaw collimators.

The in-vivo dose measurements performed onall head and neck patients using a dedicated diodedosimetry system have shown that the measureddose agrees with the planned prescription dose towithin ± 5%. Such routine measurements have be-come an integral part of our quality assurance pro-gram to establish accuracy in patient setup and dosedelivery.

Fig. 3. Continued


CONCLUSIONS

The protocols developed in this work for simula-tion and treatment planning of head and neck andesophageal tumors are based on monoisocentrictechnique, which provides the superior dose distri-butions in the target volume with markedly reducedmorbidity to the surrounding normal tissues, andcan be easily implemented in a radiation oncologydepartment. The asymmetric collimators allow easyand reproducible patient setup with decreased treat-ment time per patient. Film validation studies haveshown that, in order to maintain a dose deviationwithin ± 10% from the normalization dose at thejunction, the digital display tolerance of asymmet-ric collimators should be kept below ± 1 mm. The3-D treatment planning system used in these stud-ies has proven to be a powerful tool in terms of itscomputational speed and accuracy. Its 3-D displaycapability of isodose distributions and dose volumehistogram features allows efficient plan evaluation.The use of a real-time dose-measuring system playsan important role in verifying the actual delivery ofradiation dose.

REFERENCES1. Three-dimensional photon treatment planning. Report

of the Collaborative Working Group on the evaluationof treatment planning for external photon beam radio-therapy. Int J Radiat Oncol Biol Phys. 1991;21:1-265.

2. Paradelo JC, Ucmakli A, Schiller B, Sternick E, MowerH. Mid-sagittal plane dosimetry in patients with exten-sive head and neck malignancy. Int J Radiat Oncol BiolPhys 1981;7:115–120.

3. Young MEJ, Kornelson RO. Dose corrections for low-density tissue inhomogeneities and air channels for 10-MV x-rays. Med Phys 1983;10:450–455.

4. Gillin M, Kline RW. Field separation between lateraland anterior fields on a 6 MV linear accelerator. Int JRadiat Oncol Biol Phys 1980:6:233–237.

5. Sailer SL, Sherhouse GW, Chaney EL, Rosenman JG,Tepper JE. A comparison of postoperative techniquesfor carcinomas of the larynx and hypopharynx using3-d dose distributions. Int J Radiat Oncol Biol Phys1991;21:767–777.

6. Vikram B, Strong EW, Shah J, Spiro RH. Elective post-operative radiation therapy in stages III and IV epider-moid carcinoma of the head and neck. Am J Surg 1980;140:580–584.

7. Amdur RJ, Parsons JT, Mendenhall WM, Million RR,

Fig. 4. (a-e) Transverse dose distributions superimposed on multiple CT images for 6 MV bilateral field arrangement using30° wedges and a 1-cm bolus for treating oropharyngeal wall. 110%, 105%, 100%, 95%, and 90% isodoses are drawn. The100% isodose is displayed in color-wash format.


Stringer SP, Cassis NJ. Postoperative irradiation forsquamous cell carcinoma of the head and neck: ananalysis of treatment results and complications. Int JRadiat Oncol Biol Phys 1989;16:25–36.

8. Doppke K, Novack D, Wang CC. Physical consider-ations in the treatment of advanced carcinoma of thelarynx and pyriform sinuses using 10 MV × rays. Int JRadiat Oncol Biol Phys 1980;6:1251–1255.

9. Chiang TC, Culbert H, Wyman B, Cohen L, Ovadia J.The half field technique of radiation therapy for thecancers of head and neck. Int J Radiat Oncol Biol Phys1979;5:1899–1901.

10. Datta R, Mira JG, Pomeroy TC. Dosimetry study ofsplit beam technique using megavoltage beams and itsclinical implications. Int J Radiat Oncol Biol Phys1979;5:565–571.

11. Devineni VR, Heisch ME. Couch rotation technique fortreatment of head and neck cancer. Med Dosim 1992;17:57–60.

12. Sohn JW, Suh JH, Pohar S. A method for deliveringaccurate and uniform radiation dosages to the head andneck with asymmetric collimators and a single iso-center. Int J Radiat Oncol Biol Phys 1995;32:809–813.

13. Ahmad M, Fontenla DP, Lai PP, Del Row J, YaparpalviR, Deore SM, Curran J, Wojcicka J, Vikram B. Three-dimensional external-beam treatment planning andreal-time dose verification of pituitary adenoma: Clini-cal and physical considerations. Radiat Oncol Investig1995;3:141–148.

14. Goitein M, Abrams M, Rowell D, Pollari H, Wiles J.

Multi-dimensional treatment planning: II. Beam’s-eyeview, back projection through CT sections. Int J RadiatOncol Biol Phys 1983;9:789–797.

15. Khan FM. Dosimetry of wedged fields with asymmetriccollimation. Med Phys 1993;20:1447–1451.; Khan FM,Gerbi BJ, Deibel FC. Dosimetry of asymmetric x-raycollimators. Med Phys 1986;13:936–941.

16. Landberg T, Chavaudra J, Dobbs J, Hanks G, Johans-son K-A, Moller T, and Purdy J. Prescribing, recordingand reporting photon beam therapy. International com-mission on radiation units and measurements (ICRU)report 50, 1993.

17. Riker G, Grussel E. Patient dose measurements in pho-ton fields by means of silicon semiconductor detectors.Med Phys 1987;14:870–873.

18. Leunens G, Van Dam J, Dutreix A, Van der SchuerenE. Quality assurance in radiotherapy by in vivo dosim-etry. 1. Entrance dose measurements, a reliable proce-dure. Radiother Oncol 1990;17:141–151.

19. Marks M, Haus AG . The effect of immobilization onlocalization error in the radiotherapy of head and neckcancer. Clin Radio 1976;27:175–177.

20. Verhey LJ, Goitein M, McNulty P, Munzenrider JE,Suit HD, Precise positioning of patients for radiationtherapy. Int J Radiat Oncol Biol Phys 1982;8:289–294.

21. Hunt MA, Kutcher GJ, Burman C, Fass D, Harrison L,Leibel SL, Fuks Z. The effect of setup uncertainties onthe treatment of nasopharynx cancer. Int J Radiat OncolBiol Phys 1993;27:437–447.


three-dimensional radiotherapy of head and neck and esophageal carcinomas: a monoisocentric...

Documents