planned neck dissection as an adjunct to the management of patients with advanced neck disease...

PLANNED NECK DISSECTION AS AN ADJUNCT TOTHE MANAGEMENT OF PATIENTS WITH ADVANCEDNECK DISEASE TREATED WITH DEFINITIVERADIOTHERAPY: FOR SOME OR FOR ALL?

Kailash Narayan, MD, 1 Christopher H. Crane, MD, 2 Stephen Kleid, MD, 3

Peter G. Hughes, MD, 1 Lester J. Peters, MD 1

1 Department of Radiation Oncology, Peter MacCallum Cancer Institute, Melbourne, Victoria, Australia2 Department of Therapeutic Radiation and Oncology, University of Virginia, Charlottesville, Virginia3 Department of Otolaryngology/Head and Neck Surgery, Royal Melbourne Hospital and Peter MacCallumCancer Institute, Melbourne, Victoria, Australia

Accepted 25 January 1999

Abstract: Background. Management of patients with headand neck carcinoma and advanced nodal disease is controver-sial. The purpose of this analysis was to evaluate the efficacy andtoxicity of definitive radiotherapy followed by planned neck dis-section in patients with bulky neck disease.

Materials and Methods. The records of 52 patients who weretreated between 1989 and 1995 at the Peter MacCallum CancerInstitute with a planned neck dissection after radical radiotherapywere reviewed. All had advanced neck disease with one or morenodes $3 cm in maximum diameter, 94% being staged N2-3.The most common primary site was the oropharynx (56%). Sixtypercent of patients had either T2 or T3 primaries and all wereAJCC stage IV. Treatment consisted of high-dose radiotherapy tothe primary and involved neck sites using various fractionationprotocols followed by radical or modified radical neck dissectionafter confirmation of a complete response at the primary site. Themedian follow-up for living patients was 58 months (range 32–97).

Results. There were nine regional failures, of which threewere outside the dissected neck, yielding a 5-year actuarial over-all neck control rate of 83% and an in-field control rate of 88%.In-field control rates by neck stage were N1 3/3; N2 31/35; N311/13 and NX 1/1. There was only one in-field failure among 28

patients who had pathologically negative neck specimens com-pared with five in 24 patients with morphologic evidence of re-sidual disease. Of the 24 patients with pathologically positivenecks, 5 were long-term survivors and were probably cured bytheir surgery. Another 4 died of intercurrent disease without docu-mented recurrence of their head and neck cancer. Ten patientsrecurred at their primary sites (5-year actuarial control 79%) and8 developed distant metastases (5-year actuarial rate 20%). Atotal of 21 patients failed at one or more sites and none wassalvaged. Five-year actuarial disease-free survival was 57% andoverall survival 38%. Nine patients (17%) sustained significantcomplications following neck dissection.

Conclusions. In patients with advanced neck disease whoare treated primarily with radical radiotherapy, planned neck dis-section provides excellent regional control and appears to cure asubset of patients. However, routine neck dissection adds signifi-cant morbidity to treatment and should ideally be avoided in thosepatients in whom surgery is either unnecessary (no residual tu-mor) or futile (unsalvageable disease recurrence outside the dis-sected neck). Based on our analysis and other recently reportedseries, we now recommend observing patients who have a com-plete response to high-dose radiotherapy (± chemotherapy). Theability of PET imaging to detect residual viable tumor in the headand neck or at distant sites is under investigation. © 1999 JohnWiley & Sons, Inc. Head Neck 21: 606–613, 1999.

Keywords: neck nodes; radiotherapy; neck dissection; residualdisease; functional imaging

Correspondence to: L. J. Peters

CCC 1043-3074/99/070606-08© 1999 John Wiley & Sons, Inc.

606 Planned Neck Dissection Following Radiotherapy HEAD & NECK October 1999

Three approaches are commonly used in themanagement of the neck in patients presentingwith advanced nodal disease from squamous cellcarcinoma of the head and neck: surgical resec-tion of the primary and neck disease followed bypostoperative radiotherapy, radical radiotherapywith planned neck dissection regardless of nodalresponse, and radical radiotherapy with surgeryonly for those with resectable persistent or recur-rent disease.

The concept of combined modality therapy forpatients with head and neck cancer was intro-duced in 1957 by MacComb and Fletcher.1 Theidea that large regional adenopathy in head andneck cancer was best treated with a combinationof radiation and surgery was developed by inves-tigators at the University of Texas M. D. Ander-son Cancer Center (MDACC). They presenteddata demonstrating that combined modalitytherapy improved control rates over either radia-tion or surgery alone.2 Other investigators havealso found improved control with combinedtherapy over surgery alone3,4 or irradiationalone.5,6

On the basis of the improved regional controlreported from the University of Florida5 usingplanned neck dissection after radiotherapy, asimilar treatment policy was instituted at the Pe-ter MacCallum Cancer Institute (PMCI) in 1989.Patients with primary tumors suitable for radio-therapy but with advanced cervical metastaseswere radically irradiated. Those who achieved acomplete response at the primary site then wenton to have either modified radical or radical neckdissection regardless of nodal response. It wasreasoned that such treatment would provide op-timal control of advanced nodal disease, and wasless morbid than primary surgery and postopera-tive radiotherapy.

This report reviews this experience and dem-onstrates that radical radiotherapy to the pri-mary and neck followed by neck dissection pro-vides excellent regional control with acceptablemorbidity. The analysis does not directly addressthe question recently raised by investigators atMDACC7 as to whether a neck dissection can besafely omitted in patients who obtain a completeregression of nodal disease, but this issue is con-sidered in the Discussion.

MATERIALS AND METHODS

Patient Population. All patients undergoingplanned neck dissection according to the policyoutlined above between 1989 and 1995 were in-

cluded. Patients were identified using the patientregistry at PMCI. Advanced neck disease was de-fined as the presence of one or more nodes $3 cmin maximum diameter. A total of 52 patients wereidentified, their treatment records were exam-ined, and the data were coded on specially de-signed forms. Neck dissection was carried out inall patients who had a complete clinical responseat the primary site based on postradiotherapy ex-amination under anesthesia. The diagnosticwork-up included history, physical exam, and CTscan in all patients. MRI was used, but not con-sistently in all patients. The nodal size was mea-sured directly with calipers and the primary andneck staging was determined on AJCC fourth edi-tion criteria using both physical and radiologicalfindings.

The pretreatment patient characteristics arelisted in Table 1. The median age was 62 (range40–83). The most common primary site was theoropharynx (56%). All patients were stage IV and

Table 1. Patient characteristics.

Characteristic Number (N = 52) %

AgeMedian 62Range 40–83

30–40 1 240–50 11 2150–60 13 2560–70 20 38>70 7 13

SexM:F 5.5:1M 44 85F 8 15

Primary siteOropharynx 29 56Hypopharynx 7 13Larynx 11 21Oral cavity 1 2Unknown 4 8

T stageTx 2 4T0 4 8T1 8 15T2 18 35T3 13 25T4 7 13

N stageNx 1 2N1* 3 6N2a 17 33N2b 16 31N2c 2 4N3 13 25

*All were 3 cm.

Planned Neck Dissection Following Radiotherapy HEAD & NECK October 1999 607

all had advanced nodal disease as defined above.Table 2 provides a breakdown of the number ofnodes and size of largest node for each patient.The median largest nodal diameter was 45 mm(range 30 mm–110 mm). The histology was squa-mous cell carcinoma in all patients.

Treatment Techniques. Treatment consisted ofdefinitive radiotherapy to the primary and in-volved neck sites in all patients using a variety oftechniques and doses (Table 3). For the purposesof analysis of dose response, the various fraction-ation schedules were converted into a 7-week 2Gy/fraction equivalent dose using an a/b ratio of10 and a dose equivalent of regeneration of 0.6Gy/day. Treatment fields generally included theprimary disease, bilateral upper neck, and lowerneck on the involved side or sides. Patients un-derwent neck dissection at variable intervals fol-lowing irradiation due to differences in their re-covery from acute radiation toxicity following thedifferent fractionation schedules. Thirty patientshad surgery within 2 months of radiotherapy, 16between 2 and 3 months after radiotherapy, and 6more than 3 months after radiotherapy. The samesurgeon operated on the majority of the patientsand the procedures were all either radical ormodified radical neck dissections with preserva-tion of the spinal accessory nerve.

Follow-Up. Patients were scheduled to be seenand evaluated at 2- to 3-month intervals for thefirst 2 years, 4-month intervals for the third year,and 6-month intervals thereafter. The median fol-low-up for living patients was 58 months (range32–97 months). The median follow-up for all pa-tients was 26 months (range 3–97 months).

Endpoints of Study. All 52 patients were evalu-ated for actuarial rates of neck control, primarycontrol, and distant metastasis at the time of first

relapse, as well as disease-free and overall sur-vival. The Life Table method was used to calcu-late actuarial rates. Time intervals were calcu-lated from the date of initiation of radiotherapy tothe time of relapse or last evaluation. The close-out date for analysis was September 1997.

RESULTS

Surgical Pathology. Pathology reports were re-viewed in all 52 patients and morphologically vi-able tumor was found to be present in 24 (46%)with gross tumor identifiable in 17 of the 24. Theremaining 28 dissected necks (54%) had negativepathology.

Extracapsular extension (ECE) was present in8 of the 24 positive specimens. Seventeen of 27specimens with original nodal size >40 mm in di-ameter were positive compared with 7 of the 24with original nodal size #40 mm (p 4 0.003).No pretreatment characteristic other than ini-tial maximum nodal size was correlated with apathologic complete response. Because clinical re-sponse data were not systematically recorded be-fore neck dissection, it was not possible retrospec-tively to correlate clinical and pathologic completeresponses.

There was no clear correlation between the

Table 2. Extent of nodal disease.

No. ofnodes

Size of largest node (cm)

3.0 3.1–4 4.1–5 5.1–6 6.1–7 7.1–8 8.1–9 9.1–10 >10

1* 3 9 6 2 7 2 12 4 5 1 13 1 1 2 1 14 1 15 16 1

*One patient had a single “large” node removed by excision biopsy before referral.

Table 3. Radiotherapy techniques.

Regimen DoseNo. of

fractionsNo. ofweeks

Number(N = 52)

Standard fractionation 70 Gy 35 7 11(QD) 66 Gy 33 6.5 2

62 Gy 31 6.2 160 Gy 30 6 7

Accelerated fractionation 66 Gy 33 3.2 1(BID) 60 Gy 30 3 4

54 Gy 30 3 16Concomitant boost (CB) 66 Gy 35 5 9Split course (SC) 50 Gy 25 7 1


biologically effective dose (calculated as describedabove) and the presence of pathological residualtumor. There was a nonsignificant trend toward ahigher rate of pathologic clearance at higherdoses: with a 7-week equivalent dose of $70 Gythere were 15/25 negative neck specimens (60%)compared with 13/27 (48%) with lower doses.

Neck Control. Five-year actuarial neck controlwas 83% (Figure 1). There were a total of nineneck relapses with or without concurrent primaryfailure or distant relapse. Six of these were withinthe irradiated/dissected neck and three were inthe untreated contralateral neck. Four were iso-lated neck relapses, three occurred with synchro-nous primary failure, and two with synchronousdistant failure. Of the four isolated neck relapses,one occurred in an unirradiated area. The char-acteristics of the patients who suffered in-fieldneck relapse are shown in Table 4. None of thesepatients had a time to failure greater than 1 year.Five of the six patients who recurred in the dis-sected neck had residual disease in the neck dis-section specimen, and three had evidence of ex-tracapsular spread. Patients with residual diseasein the neck specimen had significantly worse in-field neck control than those whose neck speci-mens were pathologically negative (p 4 0.033).

There was no significant difference in in-fieldneck control between patients with N1, N2, or N3neck disease (p 4 0.822, Figure 2). Thus, 3 of 3patients with 3 cm N1 nodes, 31 of 35 (5-yearactuarial 87%) patients with N2 neck disease, and11 of 13 (5 year actuarial 86%) patients with N3neck disease had in-field neck control for an over-all 5-year actuarial in-field neck control rate of88%.

Primary Control. There were 10 primary tumorfailures (5-year actuarial control 79%). Five pa-tients suffered isolated primary failure, four had

synchronous primary and neck failure (includingtwo in the untreated neck), and one sustained pri-mary and distant failure.

Distant Metastasis. There were eight distant re-lapses (5-year actuarial 20%). Five patients suf-fered isolated distant relapse and, as previouslymentioned, two had simultaneous neck failure,and one had synchronous primary failure.

Survival. A total of 31 patients were disease-freeat last encounter or death, for a 5-year actuarialdisease-free survival (DFS) of 57%; 20 were aliveat last follow-up for a 5-year actuarial overall sur-vival (OS) of 38% (Figure 3). No patients werealive with disease. Twenty-one patients recurredat one or more sites and all died of or with uncon-trolled disease. Another patient died of treat-ment-related complications with residual diseasein the neck dissection specimen. Of the remaining10 deceased patients, 4 died of intercurrent ill-ness, 4 died of uncertain causes without evidenceof disease at last encounter, and 2 died after de-veloping a metachronous lung cancer. The 5-yearactuarial incidence of intercurrent death or deathdue to second primary was 27%. Of the 24 pa-tients whose planned neck dissection revealed re-sidual tumor, 9 remained cancer-free (5 were

Table 4. In-field neck failure: patient characteristics.

Time tofailure (mo) Primary site

Pretreatmentnodal staging

N stage

Maxsize(mm)

RT dose &fractionation

Neck specimenpathology*

Neck alone 11.3 T3 Pyriform sinus N2a 40 60 BID +, +ECE5.0 TX Soft palate N3 110 60 QD +, +ECE

10.7 T1 Pyriform sinus N2a 50 54 BID +, −ECENeck & primary 10 T2 Supraglottis N3 80 70 QD −

7.5 T2 Tonsil N2a 35 60 QD +, −ECENeck & distant 8.7 T2 Supraglottis N2b 35 66 BID +, +ECE

*ECE = extracapsular extension.

FIGURE 1. Actuarial overall and in-field neck control.


alive at last contact and 4 died of intercurrentdisease). One patient died of postoperative com-plications. The remaining 14 died of their cancer,including all but 1 patient with ECE in the neck.Of the 28 patients whose neck specimens werenegative, 21 remained disease free (15 were aliveat last contact and 6 died free of disease). Of the 7who had disease recurrence, 3 failed with distantdisease alone, 2 with local disease alone, 1 withconcurrent local and neck disease, and 1 with con-current distant and neck disease. Patients withpositive disease in the neck specimen had statis-tically worse DFS (p 4 0.006) and OS (p 4 0.004).

Toxicity. Nine patients (17.3%) sustained 10 sig-nificant complications related to the planned neckdissection (Table 5). A significant complicationwas considered to be one requiring either reop-eration or long-term medical treatment. Threepatients had sterile wound breakdown and onehad a skin flap necrosis. These were repaired us-ing pectoralis major myocutaneous flaps. Two pa-tients had chyle leaks requiring repair, one ofwhom died of anesthetic complications. One pa-tient had a postoperative wound infection and

persistent sinus formation, one had a venousbleed requiring reoperation, and two patients de-veloped severe painful neck stiffness followingtheir neck dissection requiring long-term analge-sics. There were no fistulae, carotid blowouts,temporary or permanent tracheostomies, or per-manent feeding tubes needed.

DISCUSSION

Advanced neck disease in patients with head andneck cancer can be approached in a variety ofways. Surgical resection of both primary and necknodal disease followed by postoperative radio-therapy is the most commonly used treatmentstrategy.8–10 Although this approach is appropri-ate for many patients, it results in unnecessaryfunctional morbidity in patients with radiocur-able primary disease who are first seen with ad-vanced neck disease. An alternative and equallyefficacious approach in these patients is radicalradiotherapy to all gross disease followed byplanned neck surgery in patients who achieve acomplete response at the primary site. This strat-egy must be distinguished conceptually from theuse of low- to medium-dose preoperative radio-therapy (in which the intent of treatment is lim-ited to sterilization of subclinical tumor exten-sions), which does not obviate the need to resectthe primary tumor. We report here on 52 patientswith neck disease $3 cm who underwent defini-tive radiotherapy followed by planned neck dis-section. In-field neck control was achieved in 88%and overall survival was 38% in this group ofstage IV patients. Isolated neck failure was anuncommon event (8%). This compares favorablywith other reports in the literature using eithersingle modality or combined modality treatmentFIGURE 3. Actuarial disease-free and overall survival.

FIGURE 2. Actuarial in-field neck control by neck stage.

Table 5. Postsurgical complications.

Complication

RTdose(Gy) Technique

XRT/Surginterval

(mo)Nodalpath

Seroma and wounddehiscence 70 QD 2.0 N−

Wound dehiscence 70 QD 2.2 N+Wound dehiscence 60 QD 1.5 N+Skin flap necrosis 54 BID 2.0 N+Chyle leak; painful

stiff neck 54 BID 2.5 N−Chyle leak; anesthetic

death 66 CB 1 N+Wound infection with

persistent sinus 66 CB 1.3 N−Postoperative bleed 54 BID 1.3 N−Painful stiff neck 54 BID 1.5 N−


and is consistent with reports of a similar ap-proach at the University of Florida of high-doseradiotherapy followed by neck dissection.11,12

The purpose of the planned neck dissection inthese patients was to maximize the probability ofachieving regional control, as it is well estab-lished that control of neck nodes by radiotherapyalone decreases with nodal size. Parsons et al.11,12

reported improved neck control rates in patientswith N2 and N3 (but not N1) disease treated withtwice daily radiotherapy followed by neck dissec-tion versus radiotherapy alone. In a large series ofpatients with advanced neck disease treated withradical radiotherapy alone at the Institut Curie,Dubray et al.13 reported overall 3-year actuarialneck failure rates of 33% for N2 (n 4 103) and45% for N3 (n 4 699) using UICC (1987) stagingcriteria. However, in the same series the inci-dence of isolated neck failure was only 7%.14 Onecan conclude from these reports and our data thatplanned neck dissection contributes to neck con-trol in unselected N2 and N3 patients treated byradiotherapy, but its exact role remains unde-fined.

The question of whether patients could be se-lected for neck dissection following radical radio-therapy on the basis of their response to treat-ment was addressed by Peters et al.7 Theyreported on 75 patients with node-positive oro-pharyngeal cancer treated definitively with theUT MDACC concomitant boost technique. Theneck nodes had a median size of 3 cm with a rangeof 1–9 cm. Of these patients, 62 had a completeclinical and radiological response to radiotherapyand did not undergo planned neck dissection.Neck failure occurred in 8 patients, 5 of whomhad an associated primary recurrence. The re-maining 13 patients had residual palpable nodesin the neck and underwent planned neck dissec-tion. Only 6 of these had pathologic evidence ofresidual cancer, and there were no neck recur-rences in this group. In the M. D. Anderson series,the probability of obtaining a complete responsewas inversely related to nodal size as expected.However, there was no correlation between initialnode size and the risk of neck failure provided acomplete response (CR) was achieved. Lavertu etal.15 reported on 53 patients with N2-3 diseasewho underwent radical treatment with radio-therapy with or without chemotherapy. Of 30 pa-tients who obtained a complete response in theneck, 18 had a planned neck dissection (pathologi-cally positive in 4) and 12 were observed for un-stated reasons. Of these, 3 recurred in the neck.

More recently, Johnson et al.16 reported on a se-ries of 81 patients with cervical lymphadenopathy(62 N2-3) at the time of referral. Of these, 58 ob-tained a complete response to treatment with aconcomitant boost accelerated hyperfractionatedradiotherapy schedule and were subsequently ob-served. As in the M. D. Anderson series, isolatedneck failure was uncommon, occurring in onlythree patients. Boyd et al.17 described 25 patientsof whom 24 had N2-3 disease and all of whomunderwent planned neck dissection after radicalradiotherapy. Nineteen of the 28 dissected necks(68%) contained no pathologic evidence of re-sidual disease. This is similar to our own findingof pathologically negative specimens in 54% of pa-tients. Although there is obviously a finite falsenegative rate associated with a pathologic end-point because of sampling error, it is clear whenall the data are taken together that a significantproportion of radically irradiated patients whopresent with advanced neck disease are region-ally controlled by radiotherapy alone. A clinicalCR is indicative of nodal control but is not, in andof itself, sufficient to exclude the presence of re-sidual viable disease. Conversely, the presence ofa palpable residuum is associated with pathologicevidence of tumor in less than half the cases.

The most cogent argument for planned neckdissection is the importance of securing neck con-trol with the initial treatment, because salvage ofrecurrent disease in the neck is rarely achieved.For example, in a series of 139 patients at theUniversity of Florida with positive nodes treatedwith radiotherapy alone, 35 developed neck recur-rence; salvage was attempted in 9 patients butwas successful in only two.18 Likewise, in the M.D. Anderson series referred to above only one ofseven attempted salvages of patients who re-lapsed in the neck after radiotherapy was suc-cessful.7 Significantly, in both of these series, thenumber of attempted salvages was small relativeto the number of neck failures because of uncon-trolled disease beyond the neck.

There is no question that high rates of regionalcontrol can be achieved with routine planned neckdissection, but at what cost? In our series therewas a 17.3% incidence of significant postoperativemorbidity (including one treatment- relateddeath), and all patients who have a radical neckdissection sustain some functional deficit. Thishas to be balanced against the fact that plannedneck dissection can only affect the ultimate out-come of patients who have residual disease in theneck but are controlled at the primary site and


have no distant metastases. In our series, al-though regional control at 5 years was 84%, dis-ease-free survival was only 57%, reflecting a5-year primary failure rate of 21% and a distantmetastasis rate of 20%. The survival of patientswho had residual disease in their neck dissectionspecimen was significantly worse that that of pa-tients with negative pathology; nonetheless, 5 ofthe 24 patients with residual disease were appar-ently cured by the surgery and another 4 died ofintercurrent disease without evidence of recur-rent cancer.

What can be done to improve the utility oftreatment this group of patients? Clearly, a betterway of identifying patients with residual viabledisease who would benefit from a neck dissectionis needed. One potential way of improving the dis-crimination of clinical assessment is by functionalimaging with positron emission tomography(PET). Quantitative PET scanning has beenshown to correlate with tumor response in smallpilot series.19,20 However, other studies havefailed to show an association.21 Lack of specificityof PET imaging may be attributable to the inflam-matory response associated with recent radio-therapy and it would be predicted that the dis-crimination of functional imaging would improvewith time after treatment. We are currently in-vestigating this modality with 18FDG scans 12weeks after completion of radiotherapy.

Another possible approach is to improve theefficacy of treatment such that the probability ofachieving regional control is so high in completeresponders as to make the false negative rate neg-ligible. The most promising developments in thisregard are with rationally designed concomitantchemoradiation protocols. Recent data from M. D.Anderson show a negligibly small rate of neck re-lapse (1 of 27) in patients with advanced neckdisease who obtained a complete response aftertreatment with radical radiotherapy plus infu-sional 5FU and cisplatin during the last 2 weeksof treatment.22

Finally, for circumstances in which there isperceived need to proceed with planned neck dis-section, consideration should be given to reducingthe morbidity of the operation by making it selec-tive rather than comprehensively radical. In theseries of Boyd et al.,17 the conclusion (based onanatomic distribution of residual disease in theneck) was that a selective dissection of levels II–IV would suffice in the great majority of patients.Alternatively, and perhaps more rationally, one

could plan modified dissections in which only theoriginally involved nodal stations were removed.

In conclusion, our data, like other series, dem-onstrate that planned neck dissection after radi-cal radiotherapy achieves a very high level of re-gional control but is ultimately beneficial to onlya subset of patients. In the majority the operationis either unnecessary because there is no residualdisease in the neck, or futile because of unsal-vageable primary recurrence or distant metasta-ses. Potential avenues for improvement are betterways of identifying residual disease after radio-therapy, more effective initial treatment proto-cols, and more selective dissections in patientsundergoing postradiotherapy neck surgery. Forthe moment, in the absence of compelling datafrom randomized trials, the subject of plannedneck dissection is bound to remain controversial.

REFERENCES1. MacComb WS, Fletcher GH. Planned combination of sur-

gery and radiation in treatment of advanced primaryhead and neck cancer. Am J Roentgenol 1957;77:397–415.

2. Barkley HT, Fletcher GH, Jessee RH, Lindberg RD. Man-agement of cervical lymph node metastases in squamouscell carcinoma of the tonsillar fossa, base of tongue, su-praglottic larynx, and hypopharynx. Am J Surg 1972;124:462–467.

3. Huang DT, Johnson CR, Schmidt-Ullrich R, Grimes M.Postoperative radiotherapy in head and neck carcinomawith extracapsular lymph node extension and/or positiveresection margins: a comparative study. Int J Rad OncolBiol Physics 1992;23:737–742.

4. Vikram B, Strong EW, Shah JP, Spiro R. Failure in theneck following multimodality treatment for advancedhead and neck cancer. Head Neck Surg 1984;6:724–729.

5. Mendenhall WM, Million RR, Cassisi NJ. Squamous cellcarcinoma of the head and neck treated with radiationtherapy: the role of neck dissection for clinically positiveneck nodes. Int J Rad Oncol Biol Physics 11986;2:733–740.

6. Fein DA, Lee WR, Amos WR, et al. Oropharyngeal carci-noma treated with radiotherapy: a 30-year experience. IntJ Rad Oncol Biol Physics 1996;34:289–296.

7. Peters LJ, Weber RS, Morrison WH, Byers RM, GardenAS, Goepfert H. Neck surgery in patients with primaryoropharyngeal cancer treated by radiotherapy. HeadNeck 1996;18:552–559.

8. Peters LJ, Goepfert H, Ang KK, et al. Evaluation of thedose for postoperative radiation therapy of head and neckcancer: first report of a prospective randomized trial. IntJ Rad Oncol Biol Physics 1993;26:3–11.

9. Peters LJ. The efficacy of postoperative radiotherapy foradvanced head and neck cancer: quality of the evidence.Int J Rad Oncol Biol Physics 1998;40:527–528.

10. Lundahl RE, Foote RL, Bonner JA, et al. Combined neckdissection and postoperative radiation therapy in themanagement of the high risk neck. A matched pair analy-sis. Int J Rad Oncol Biol Physics 1998;40:529–534.

11. Parsons JT, Mendenhall WM, Stringer SP, Cassisi NJ,Million RR. Twice-a-day radiotherapy for squamous cellcarcinoma of the head and neck: the University of Floridaexperience. Head Neck 11993;5:87–96.


12. Parsons JT, Mendenhall WM, Cassisi NJ, Stringer SP,Million RR. Neck dissection after twice-a-day radio-therapy: morbidity and recurrence rates. Head Neck1989;11:400–404.

13. Dubray BM, Bataini JP, Bernier J, et al. Is reseeding fromthe primary a plausible cause of node failure? Int J RadOncol Biol Physics 1993;25:9–15.

14. Bernier J, Bataini JP. Regional outcome in oropharyngealand laryngopharyngeal cancer treated with high dose perfraction radiotherapy: analysis of neck disease responsein 1646 cases. Radiother Oncol 1986;6:87–103.

15. Lavertu P, Adelstein DJ, Saxton JP, et al. Management ofthe neck in a randomised trial comparing concurrent che-motherapy and radiotherapy with radiotherapy alone inresectable stage III and IV squamous cell head and neckcancer. Head Neck 1997;19:559–566.

16. Johnson C, Silverman L, Clay L, Schmidt-Ullrich R. Ra-diotherapeutic management of bulky cervical lymphade-nopathy in squamous cell carcinoma of the head and neck:is post-radiotherapy neck dissection necessary? RadiatOncol Investig 1998;6:52–57.

17. Boyd T, Harari P, Tannehill S, et al. Planned postradio-therapy neck dissection in patients with advanced headand neck cancer. Head Neck 1998;20:132–137.

18. Mendenhall WM, Million RR, Bova FJ. Analysis of time-dose factors in clinically positive neck nodes treatedwith irradiation alone in squamous cell carcinoma of thehead and neck. Int J Rad Oncol Biol Physics 1984;10:639–643.

19. Greven KM, Williams DW III, Keyes JW Jr, et al. Positronemission tomography of patients with head and neck car-cinoma before and after high dose irradiation. Cancer1994;74:1355–1359.

20. Daemen BJ, Elsinga PH, Paans AM, Wieringa AR, Kon-ings AW, Vaalburg W. Radiation-induced inhibition of tu-mor growth as monitored by PET using L-[1-11C]tyrosineand fluorine-18-fluorodeoxyglucose. J Nuclear Med 1991;33:373–379.

21. Haberkorn U, Strauss LG, Dimitrakopoulou A, et al. PETstudies of fluorodeoxyglucose metabolism in patients withrecurrent colorectal tumors receiving radiotherapy. JNuclear Med 1991;32:1485–1490.

22. Garden A, Glisson B, Ang K, et al. A phase I/II trial ofradiation with chemotherapy “boost” for advanced squa-mous cell carcinomas of the head and neck: toxicities andresponses. J Clin Oncol 1999, in press.


planned neck dissection as an adjunct to the management of patients with advanced neck disease...

Documents