Lymphatic Mapping for Staging of Head and Neck Cancer
Jai Balkissoon, Barry M. Rasgon, and Laurie Schweitzer
c(ppml
StShtrcudmiait
eicabbiSs
fiCatpdnrSmridb
ymphatic mapping with sentinel lymph node (SLN)iopsy can accurately stage the nodal basins in patientsith melanoma of the trunk and extremities and hasecome a routine, well-accepted diagnostic method forelanoma at these anatomic locations. Melanoma of
he head and neck (16% of all cases of melanoma) isomplex and difficult to manage because of the richbundant interlacing lymphatic drainage patterns, asell as watershed areas, which can lead to unusual andnexpected drainage patterns. Radioguided surgery inombination with blue dye facilitates localization of theLN in the head and neck; however, this type of ra-ioguided surgery is an evolving technique of someifficulty and thus requires careful coordination amonghe surgeon, nuclear medicine physician, and patholo-ist. Applications of this technique to other sites in theead and neck are currently being investigated for con-itions including squamous cell carcinoma (SCC) of theral cavity, thyroid cancer, and Merkel cell cancer.ore studies of patients with head and neck cancer are
RIMARY CUTANEOUS melanoma of thehead and neck is uncommon and occurs on all
reas of the scalp, face, and neck. The condition iseen more commonly in males and is more aggres-ive and has a worse prognosis and higher recur-ence rate than melanoma of similar thickness inhe trunk or extremities.1
The head and neck lymphatics are locatedithin six nodal basins in addition to the parotid
egion (Fig 1).2 Because these areas are highlyascularized and contain an abundant network ofverlapping lymphatic channels, collateral drain-ge, and watershed areas, drainage patterns arenpredictable and variable. Primary melanoma ofhe head and neck frequently develops superfi-
From the Department of Surgery, Alta Bates Medical Center,erkeley, CA; Department of Otolaryngology, Head and Neckurgery, Kaiser Permanente Medical Center, Oakland, CA; andhe Department of Otolaryngology, Head and Neck Surgery, Eastay Cancer Center, Summit Medical Center, Oakland, CA.Address reprint requests to Jai Balkisson, MD, 2999 Regent St,
ially or adjacent to the sentinel lymph nodeSLN) basins, which can make preoperative lym-hoscintigraphy and introperative localizationroblematic. Because the head and neck are theost visible parts of the body, any surgery in this
ocation must consider cosmesis.Blue dye has been used by itself to localize the
LN. The drawback of using this method is thathe afferent blue lymphatic channel leading to theLN cannot be visualized unless a skin incisionas been made. When this procedure is done inhe head and neck, subplatysmal flaps must beaised and the blue channel observed along itsourse to the lymph nodes. In patients who havenusual ipsilateral, bilateral, or contralateralrainage, SLNs can be missed if this blue dyeethod is used. The method also complicates
dentification of SLNs in unusual locations, suchs the supraclavicular or axillary areas. Melanomasn the lower parts of the neck and upper parts ofhe chest can drain to these areas.
Identifying SLNs in the parotid area can bespecially difficult. SLNs drain to the parotid arean 30% to 56% of affected patients and mostommonly accompany lesions on the ear, anteriorspect of the scalp, and temporal area.3,4 Whenlue dye alone is used, drainage to the parotidasin is difficult to detect: nodes may be concealedn the substance of the parotid gland, and theLNs—which are often very small—can resembleurrounding tissue.
Varying success rates have been reported fornding blue-stained SLNs in the head and neck.hao et al5 reported that 59% of SLNs in the head
nd neck stained blue compared with 68.6% in therunk and 74% in extremities. Other studies re-orted blue staining in 29% (J.B., unpublishedata, 2002) and 28%6 of SLNs in the head andeck. These rates are much lower than the rateeported by Bostick et al,7 who found blue-stainedLNs in 92% of patients with head and neckelanoma. This discrepancy in reported success
ates may reflect variation in elapsed time fromnjection to exploration or may reflect use of aifferent staining procedure (ie, not reinjectinglue dye).The blue dye migrates rapidly to the SLN and
ashes out easily because of extensive lymphatics
Seminars in Oncology, Vol 31, No 3 (June), 2004: pp 382-393
atibbwTm
3smdrsbwctb
cia1estt
gtpboltp
wa
LYMPHATIC MAPPING IN HEAD AND NECK CANCER 383
nd vascularity in the head and neck; Therefore,iming of injection of the blue dye is critical. Ifnjection is done to early, by the time the flaps areeing developed and subcutaneous dissection iseing done to find the blue channels, the blue dyeill have passed through the first echelon nodes.o prevent this rapid transit from being a problem,any surgeons recommend injecting the blue dye
Fig 1. Diagram shows lymph node basins of neck. Reprintedith permission.2 Copyright � 2002, American Medical Associ-tion. All rights reserved.
Table 1. Reported Sensitiv
StudyBlue DyeSensitivity
Leong et al, 19999 —Wagner et al, 200013 —Carlson et al, 200012 —Bostick et al, 19977 92% (basins)Wells et al, 199811 50% (patients)Alex et al, 199815 67% (patients)O’Brien et al, 19958 —Peralta et al, 199814 —Davison et al, 200110 82% (patients)
to 5 minutes before incision. The blue dyehould be reinjected intradermally around the pri-ary site every 15 to 20 minutes. Nontechnical
rawbacks of using blue dye in the head and neckegion include a cosmetic concern: possible bluetaining of skin surfaces that is easily visible. Mostlue-stained tissue eventually fades or is removedhen the melanomatous tissue is completely ex-ised. Also of concern is the small but real risk ofhe patient having an anaphylactic reaction to thelue dye.Imaging techniques using blue dye, radioactive
olloid, and gamma probe detection have togetherncreased the SLN detection rate to between 98%nd 100%.2 Most of these imaging studies (Table)7-15 have described melanoma of the trunk andxtremities and not of the head and neck. Mosteries describing large numbers of melanoma pa-ients do not include many with primary tumors ofhe head and neck.
PREOPERATIVE LYMPHOSCINTIGRAPHY
Use of preoperative lymphoscintigraphy hasreatly improved our ability not only to identify allhe nodal basins at risk but also to identify multi-le SLNs and their location within each nodalasin (Table 2). Preoperative lymphoscintigraphyf the head and neck shows between one and threeymph node basins and one to five SLNs per pa-ient. Balkissoon (unpublished data, 2002) has re-orted visualizing 1.31 nodal basins and 1.75 SLNs
es for Lymphatic Mapping
Gamma Probe �
LymphoscintigraphySensitivity
CombinedSensitivity
90% (patients) —— 99% (patients)
98% (patients) —— 96% (basins)— 95% (patients)
91% (patients) —98% (patients) —
— —94% (patients) 100% (patients)
94% 98%
t Williams & Wilkins.
ity Rat
ippincot
pabifttStvpO
SRt2peSa
tpStatagdtpbn�a
mtph
384 BALKISSOON, RASGON, AND SCHWEITZER
er patient with melanoma of the head and neckrea. Rasgon4 has reported visualizing 1.4 nodalasins and three SLNs per patient. The multiplic-ty and widespread distribution of SLNs may resultrom the complexity of the lymphatic system inhe head and neck area.16 In addition, the elapsedime between injection of radioactive colloid andLN biopsy is prolonged enough to allow theracer to diffuse into secondary-echelon nodes pre-iously identified falsely as SLNs.6 In a study of 97atients, 85% of whom had multiple SLNs,’Brien et al8 reported identifying one or more
Table 2. Nodal Basins in the Head and Neck RegionWhere SLNs May Be Located
CervicalAnterior cervicalPosterior cervicalSubmentalSupraclavicularParotidCervical, other (occipital, mastoid, ectopic)
Fig 2. Lymphoscintigrams compare use of high-dose (500elanoma of right pinna and left cheek. Shine artifact incor
reatment with lower activity levels of 99mTc sulfur colloid aermission from Rasgon BM. Use of low-dose technetium TC 9
ead and neck: preliminary study. Laryngoscope 2001;111:11366-1137
LNs in 95 of 97 preoperative lymphoscintigrams.asgon4 found SLNs in 26 (96%) of the 27 pa-
ients, whereas Balkissoon et al (unpublished data,002) identified SLNs on preoperative lym-hoscintigrams of 14 (78%) of 18 patients. Preop-rative lymphoscintigraphy can help identify theseLNs and can allow the surgeon to plan surgeryccordingly.
Background radiation from the primary injec-ion site can constitute a shine artifact that incor-orates the SLN and can thus make localization ofLNs difficult—or impossible—during preopera-ive lymphoscintigraphy. For example, this shinertifact occurs in the treatment of melanoma ofhe pinna and cheek because these locations aredjacent to or directly superficial to the parotidland nodes. Every attempt should be made toecrease the shine artifact from the primary injec-ion site and to increase the distance between therimary melanoma site and its potential nodalasin. Radiation doses commonly used for mela-oma of the extremities and trunk (500 to 2,000Ci) may be too high for melanoma of the headnd neck if the primary site is near or overlaps the
low-dose (15 �Ci) treatment with 99mTC sulfur colloid fors parotid, postauricular, and suboccipital nodal basins; thus,etter visualization of adjacent nodal basins. Reprinted with
fur colloid to locate sentinel lymph nodes in melanoma of the4
�Ci) vporatellows b9m sul
2.
nhahultn
fiaoaptaestlpd
stb
masa
cbe
i
LYMPHATIC MAPPING IN HEAD AND NECK CANCER 385
odal basin in the head or neck (Fig 2). Rasgon4
as suggested that for patients with primary mel-noma near or overlapping the nodal basin in theead and neck, SLN biopsy can be done accuratelysing technetium 99m sulfur colloid at low activityevels (10 to 60 �Ci). The closer the melanoma tohe nodal basin, the smaller the volume and doseeeded (Fig 2).Preoperative lymphoscintigraphy allows identi-
cation of unusual drainage patterns in the headnd neck (Fig 3), which can occur in 34% to 84%f cases7,8,17 (Table 3). All nodal basins showingn SLN must be identified so that no potentiallyositive SLNs remain undetected. For this reason,he entire head and neck, upper part of the chest,nd both axillary areas must be included in preop-rative lymphoscintigraphy. Lymphoscintigraphyhows bilateral drainage in 10% of cases7 and failso detect existing SLNs in 10% of cases.8 Electiveymph node dissection should be considered foratients with melanoma of the head and neck whoo not have an SLN visualized.10
The injection site must be covered with a leadhield to decrease the shine artifact and thus makehe SLNs more visible. The patient’s neck should
Fig 3. Preoperative lymphoscintigrams of patient with bilamportance of imaging in different planes and the unpredictab
e flexed or extended so that it is positioned to
aximize the distance between the injection sitend the nodal basin. Use of a cobalt radiationource also helps outline landmarks in the headnd neck.
INTRAOPERATIVE LYMPHATIC MAPPINGAND SLN BIOPSY
The surgeon and the nuclear medicine physi-ian must both review lymphoscintigrams beforeeginning the planned surgery. This procedurenables the surgeon to discuss with the patient any
andibular drainage from a right cheek melanoma illustratingage patterns in the head and neck.
Table 3. Primary Sites and Unusual or UnexpectedLocation of SLNs in the Head and Neck Region
dditional findings that could affect informed con-ent (eg, if an SLN is located over the parotid orubmandibular area and thus increases the risk ofacial nerve injury). Another example is drainageo parotid lymph nodes, an event that can occur in0% to 50% of cases: Being aware of this findingould lead the surgeon to discuss with the patient
he comparative operative risks of performing su-erficial parotidectomy versus sentinel lymphade-ectomy and the risk of injury to branches of the
acial nerve. Preoperatively, the surgeon may alsolan to use a facial nerve monitor during therocedure.4 In addition, an SLN shown by preop-rative lymphoscintigraphy in the tail of the pa-otid may actually be located in level 2b. Theurgeon must be aware of this finding before be-inning dissection in and around the parotid glandnd must take extra steps intraoperatively to iden-ify a level 2b SLN. Because structures in the headnd neck are close together, the surgeon must usell techniques that can help decrease the shinertifact emanating from the injection site. Thisill maximize identification of the SLN on preop-rative lymphoscintigraphy and intraoperatively.emoving the primary melanoma site initiallyith appropriate margins or shielding the injec-
ion site with malleable lead can accomplish thisoal.
After location of the SLN has been identifiednd designated by a mark on the overlying skin,he surgeon must decide what initial skin incisions needed to remove the SLN; this decision will beetermined by the surgical approach most likely toinimize morbidity. In addition, thought must be
iven to the possibility of the SLN being positivend thus mandating future lymph node dissection.he initial incision should be properly planned so
hat it will not interfere with either a standardarotid incision or neck dissection. Biopsy of SLNsn the head and neck should not violate standardncologic principles.SLNs in the parotid area can be problematic
nd should be approached through a standardreauricular parotid incision. This method givesxcellent visualization of small SLNs and mini-izes the risk of injury to branches of the facialerve. Some surgeons recommend superficial pa-otidectomy in patients who have a parotid SLNdentified by lymphatic mapping.11,16 Advocates ofhis approach believe that superficial parotidec-
omy reduces the risk of facial nerve injury. More- n
ver, if these nodes are found positive, superficialarotidectomy will already have been performed,hus avoiding the need for reoperation in thearotid area, with the attendant risk of facial nervenjury. Others18 believe that sentinel lymphade-ectomy in the parotid basin is a feasible and safepproach. Ollila et al18 described sentinel lymph-denectomy without parotidectomy in 37 patientsith SLNs found in the parotid gland; one (2.6%)f these 37 patients had surgical morbidity of theacial nerve (the patient had transient facial pare-is that resolved completely).18 Sentinel lymphad-nectomy without parotidectomy was done also inseries of 25 patients with drainage to the parotid
land.4 A facial nerve monitor was used for allatients, none of whom had any facial nerve mor-idity.Most identified parotid SLNs are superficial to
he facial nerve. If sentinel lymphadenectomy isone in the parotid area, removing an SLNhrough a transverse incision directly over thearotid gland is not recommended; this approachs not cosmetically acceptable and could lead toartial necrosis of the flap if parotidectomy be-omes necessary. Instead, the incision should belaced in the preauricular crease so that it can bencorporated into a standard parotid incision ifarotidectomy becomes necessary. The incision inhe neck should also be placed in a location thatill allow incorporation into a standard neck dis-
ection incision.When multiple SLNs are found in a nodal basin,
preferable approach may be to perform selectiveeck dissection or superficial parotidectomy to en-ompass all nodes in the basin (Fig 4).11 A recom-ended approach for removing all SLNs in aodal basin is to resect the entire sentinel nodalasin (sentinel basin lymphadenectomy) and iso-ate the SLN ex vivo; this approach reduces oper-tive morbidity and facilitates identification andsolation of SLNs.16 Because the SLNs are smallnd encased by fibrolymphatic tissue (makingdentification difficult), this approach is impor-ant, particularly in the suboccipital and posterioruricular areas. After all SLNs have been identi-ed, the decision to perform elective lymph nodeissection should be based not on frozen sectionsut on permanent histology sections and immuno-istochemical stains.The incidence of positive SLNs in the head and
eck ranges from 11% to 15%. Although mela-
naaipsmhma
firiarsrhfm
wlcbp
tuenBdtsniasdr
scSfh5dsct1naormr
cdi
tpostb
LYMPHATIC MAPPING IN HEAD AND NECK CANCER 387
oma in the head and neck is inherently moreggressive than in other anatomic areas, Chao etl5 reported a lower incidence of nodal metastasisn patients with head and neck melanoma than inatients with melanoma located at other anatomicites. According to these investigators, this findingay indicate that SLN biopsy for melanoma of theead and neck may be less reliable and technicallyore challenging than SLN biopsy done at other
natomic sites.Reliability of this technique depends on the
alse-negative rate. Several studies of SLN biopsyn the head and neck6,7,9,17 reported false-negativeates ranging from 0% to 10%. At least two stud-es6,7 as well as our own study reported false-neg-tive rates of 0% to 1%. The true false-negativeate may remain unknown until longer follow-uptudies are done and more cases of same-basinecurrence are seen. Carlson et al12 reported theighest false-negative rate, 21%. The incidence of
alse-negative results is higher in patients with
Fig 4. Lymphoscintigrams of patients with multiple SLNs inhe head. (A) At left, lateral right-brow melanoma with multi-le SLNs in parotid basin; at right, lead shielding removes mostf shine artifact from primary injection site. (B) Melanoma ofcalp with multiple SLNs. Injection site not close to nodal basin;herefore, regular doses (250 to 500 �Ci) of radiocolloid maye used.
elanoma of the head and neck than in patients c
ith melanoma of the trunk and extremities.8 Theowest false-negative results are seen in studies of aombined technique: lymphoscintigraphy pluslue dye and intraoperative use of the gammarobe.4,6,7
DISCORDANT RATES FOR LYMPHATICMAPPING
Lymphatic mapping in the head and neck isechnically challenging and can show many un-sual and unexpected drainage patterns. The high-st rates of discordance are seen in the head andeck area9 and can range from 34% to 84%.8,18
alkissoon (unpublished data, 2002) found discor-ance in 44% of patients. O’Brien et al8 reportedhat 22% of patients had SLNs in other anatomicites than the parotid area or in the five standardeck levels. That most of these SLNs were located
n the posterior auricular and occipital regions—reas where SLNs are not usually reached duringtandard neck dissection—explains most of theiscordance in reported results. This discordanceate was calculated as 34%.8
In 21 (84%) of 25 patients, lymphoscintigraphyhowed lymphatic drainage to sites not predictedlinically.19 Chao et al5 reported that only 25% ofLNs harvested from the head and neck were
ound near the parotid area. Rasgon4 reported aigher incidence of parotid SLNs, ie, as many as0%. In contrast to these high discordance rates,ata from the Sydney [Australia] Melanoma Unithowed a high rate of disease control using clini-ally predicted drainage patterns, and false-nega-ive results were rare.20 In that study,20 in which06 patients had modified elective or selectiveeck dissections done by the same surgeon within6-year period, only three patients had recurrenceutside the dissected fields. Despite these excellentesults, the variably discordant results of treatingelanoma in the head and neck are yet to be
econciled.
SENTINEL LYMPH NODE BIOPSY FORSQUAMOUS CELL CARCINOMA OF THE
ORAL CAVITY
Squamous cell carcinoma (SCC) is the mostommon cancer of the oral cavity, where the con-ition most often affects the tongue. The mostmportant prognostic factor in SCC of the oral
avity is the status of regional lymph nodes. In
ci
aen1l(tccd
ooeorBafbatTInldnmsargisma
cnmin
owet
3mo
mSsottiwtub
oUiliianc
cdasdoacssat
tabaafiSfibSs
388 BALKISSOON, RASGON, AND SCHWEITZER
ases where these nodes test positive, the cure rates decreased by half.
Management of the neck that is clinically neg-tive for SCC of the oral cavity continues tovolve. In the past, the main treatment was radicaleck dissection, as first described by Crile in906.21 This operation included removal of allymphatic basins in the neck (levels I through V)Fig 1), as well as removal of the sternocleidomas-oid muscle, internal jugular vein, and spinal ac-essory nerve (Fig 1). This operation resulted inlinically significant morbidity (including shoulderysfunction) and cosmetic deformity.Twenty percent to 34% of patients with stage I
r II SCC of the oral cavity are found to haveccult metastasis in the neck when undergoinglective neck dissection.22-25 For cases in whichnly the primary node level is resected, the recur-ence rate in the neck may be as high as 42%.26
ecause of this high recurrence rate, current ther-py attempts to reduce morbidity by using variousorms of selective neck dissection. This approach isased on the concept of orderly lymphatic drain-ge by known routes, a strategy that depends onhe anatomic site in the oral cavity or pharynx.27
he currently accepted practice for treating stagesand II SCC of the oral cavity is supraomohyoideck dissection (ie, selected neck dissection at
evels I through III) with sparing of the sternoclei-omastoid muscle, internal jugular vein, and spi-al accessory nerve. This practice is controversial;any critics view selective neck dissection as a
taging procedure only.23,25,28 If pathologic nodesre found, more comprehensive neck dissection oradiation therapy is required.23,25,28 Other investi-ators believe that more extensive neck dissections necessary because of the possibility of metastasiskipping to level IV. The estimated rate of suchetastasis in patients with an N0 neck varies from
bout 2% to 10%.24,29,30
Clinical examination and imaging techniques—omputed tomography (CT) scan, magnetic reso-ance imaging (MRI), and positron emission to-ography (PET) scan—have not accurately
dentified micrometastases in a clinically negativeeck.31,32
Currently, for patients with stage I and II SCCf the oral cavity and pharynx, the most accurateay to stage the clinically negative neck is withlective neck dissection and microscopic examina-
ion of the neck nodes. In approximately 20% to t
4% of patients with clinically negative necks,icrometastases are found22-25 thus, 66% to 80%
f patients may be receiving unnecessary surgery.Because of the success rate of SLN biopsy inelanoma of the head and neck, interest in usingLN mapping in patients with stages I and IIquamous cell carcinoma of the oral cavity andropharynx is growing. For these patients, the po-ential advantage of SLN biopsy would be to iden-ify patients with possible micrometastases; suchdentification would distinguish patients whoould benefit from elective neck dissection from
hose who would not. This method of minimizingse of surgery would not only reduce overall mor-idity but would reduce costs of treatment.Several reports on SLN biopsy for SCC of the
ral cavity and pharynx have been published.33-41
nfortunately, these studies included only a lim-ted number of patients, few of whom receivedong-term follow-up. Nonetheless, most of thesenvestigators stated their belief that SLN mappings a feasible technique for patients with stages Ind II SCC of the oral cavity and that this tech-ique seems to adequately predict presence of oc-ult metastases in the neck.
Currently, for patients with SCC of the oralavity, no uniformity exists for determining theose or volume of radiolabeled colloid injectedround the primary site or how the primary sitehould be injected. Some surgeons inject the ra-iolabeled colloid mucosally around the primary;thers inject the radiolabeled colloid submucosallyround the primary; still others inject both submu-osally and intramuscularly around the primaryite. Timing of injection also is inconsistent: someurgeons inject the radiolabeled colloid withinbout 2 hours before surgery, whereas others injecthe colloid approximately 24 hours before surgery.
Unlike melanoma—for which most investiga-ors agree that the combined technique is moreccurate then either technique used alone—use oflue dye in patients with SCC of the oral cavitynd pharynx gives mixed results. Pitman et al38
nd Rasgon et al (unpublished data, 1998) did notnd blue dye in any of their patients injected.toeckli et al39 found blue dye in only two of therst seven patients injected and abandoned use oflue dye in the remaining patients. However, bothhoaib et al35,36 and Mozzillo et al41 found blue-tained lymph nodes in many patients and stated
hat the combined technique is beneficial. Pitman
eeaTtstca
ooilfilctpiai
ou(m(dson
iimcjtgppwcs
istmtcsstnctc
tttit
b
LYMPHATIC MAPPING IN HEAD AND NECK CANCER 389
t al38 and Rasgon both used isosulfan blue; Sto-cki et al39 used methylene blue. Shoaib et al35,36
nd Mozzillo et al41 both used patent blue V dye.he discrepancy in success rates for localizing
hese blue lymph nodes may be due to any or all ofeveral factors: the technique of injection aroundhe primary tumor; the timing of injection (be-ause blue dye washes out very rapidly in the headnd neck); or the type of blue dye injected.
One of the most important technical difficultiesf lymphatic mapping in patients with SCC of theral cavity and pharynx results from close proxim-ty of the primary tumor site to the first-echelonymphatic basins. The shine artifact emanatingrom radiocolloid injection at the primary site mayncorporate levels Ia, Ib, 2a, or 2b, depending onocation of the primary cancer site in the oralavity or pharynx. This shine artifact can makehe SLN difficult to identify on preoperative lym-hoscintigrams and can make intraoperative local-zation problematic (ie, because of the largemount of radiation emanating from the primarynjection site) and thus can conceal the SLN.
Technical maneuvers to improve identificationf SLNs on preoperative lymphoscintigrams are tose a low dose (30 to 60 �Ci) and small volume0.2 mL) of radiocolloid when injecting the pri-ary tumor site4 (Rasgon, unpublished data, 1998)
Fig 5). Other techniques include injecting a largeose of radiocolloid (0.4 to 3 mCi the day beforeurgery. Because the half-life of 99mTc is 6 hours,nly a low dose of the radiocolloid remains the
33,35,36,40
Fig 5. Lymphoscintigrams compare injection of low dose (leetter visualization of SLNs in neck.
ext day at the injection site ; the decrease t
n background radiation facilitates intraoperativedentification of SLNs. Rinsing the patient’s
outh after injection of the primary tumor sitean decrease spillage of radiocolloid from the in-ection site; such spillage can interfere with iden-ification of SLNs on preoperative lymphoscinti-rams. Another helpful practice for facilitatingreoperative identification of SLNs on lym-hoscintigrams is to shield the primary tumor siteith malleable lead (whenever possible) to de-rease the shine artifact from the primary injectionite.
Intraoperative identification of SLNs can bemproved by initially excising the primary cancerite or by using a malleable lead shield (if feasible)o decrease both the shine artifact from the pri-ary injection site as well as background radia-
ion. However, this initial excision of the primaryancer site does not completely eliminate thehine artifact near the first-echelon nodes. Thehine artifact emanating from the primary injec-ion site may be decreased also by angling theeoprobe away from the primary site and using aollimator on the tip of the neoprobe. Despite allhese modifications, however, identifying the SLNan be difficult.
General consensus holds that SLN mapping inhe oral cavity can be technically challenging andhat further investigation should be pursued beforehis technique is recommended. All patients stud-ed so far have received SLN biopsy in conjunc-ion with selective neck dissection; in most of
igh dose (right) of radiocolloid. Injecting a low dose may allow
ft) and h
hese cases, the SLN could be found and reflected
tswfbfmttw
tbbwccScarn(dabwfor
bststrm
cthantworbil
as“IirnrassdseciStiwt
dmn
ps
390 BALKISSOON, RASGON, AND SCHWEITZER
he remainder of the nodal basin. However, thesetudies included small numbers of subjects, most ofhom had limited long-term follow-up. Currently,
or patients with SCC of the oral cavity, SLNiopsy should not be used to determine the needor elective neck dissection. Large, prospective,ulti-institutional studies are needed to determine
he accuracy of this technique and to resolve theechnical difficulties seen with this procedurehen used in patients with SCC of the oral cavity.Because SLN biopsy for patients with SCC of
he oral cavity and pharynx is still evolving andecause further studies are needed before this cane recommended as the standard of care, how cane use this technology now to help improve oururrent approach to diagnosing and treating thelinically negative neck of a patient with SCC?elective neck dissection at levels I through III—urrently the standard of care for treating stages Ind II SCC of the oral cavity—can fail for threeeasons. First, some important nodal basins mightot be included in the selective neck dissectionie, skip metastasis to level IV or V or contralateralrainage; Fig 6). Skip lesions have been reportedt rates ranging from 2% to 10%.24,29,30 Second,ecause of insufficient thoroughness in operatingithin the field of dissection, the surgeon might
ail to incorporate all the high-risk nodes. Thismission has been reported based on recurrence
Fig 6. Lymphoscintigram shows level IV SLN in patient withrimary cancer of the tongue. This SLN would be missed withtandard level I-III neck dissection.
ates of approximately 3% to 4% in the dissected i
asins of patients who received selective neck dis-ection at levels I through III and have necks thatested negative for pathology.25,42,43 Finally, intandard selective neck dissection at levels Ihrough III, the pathologist scrutinizes the high-isk nodes less closely than SLNs and thereforeight fail to detect micrometastases.44
An abundance of evidence indicates that mi-rometastatic disease carries prognostic connota-ions.45 Presence of micrometastatic disease alsoas therapeutic implications. In a study by Byers etl,28 patients with a clinically negative neck andodes positive for tumor who received postopera-ive radiation therapy had a 0% recurrence rate,hereas similar patients who did not receive post-perative radiation therapy had a 36% recurrenceate and all recurrences were in the dissected nodalasins. Thus, identification and microscopic exam-nation are essential for the lymph nodes mostikely to contain micrometastatic disease.
Information on high-risk nodal basins can bescertained by incorporating preoperative lympho-cintigraphy with standard therapy. In addition, ifhot” nodes are identified outside levels I throughII, selective neck dissection can be extended tonclude the nodal basins not normally incorpo-ated in selective neck dissection. Moreover, aftereck dissection is completed, the neck may beescanned using the gamma probe to ensure thatll “hot,” high-risk nodes have been removed. Theurgeon may also dissect out the SLNs from theelective neck dissection once the selective neckissection is complete. The SLNs are then labeledeparately and sent to pathology where they arexamined (by using serial sectioning and immuno-hemical staining) so that the SLNs can be exam-ned for micrometastases. This excision of theLNs would enable more accurate staging andreatment of the neck and would help surgeons todentify patients with positive histology resultsho would benefit from postoperative radiation
herapy23,28 or from additional surgery.
LYMPHATIC MAPPING AND SLN BIOPSYFOR THYROID CANCER
The utility of SLN biopsy for treating well-ifferentiated thyroid cancer has yet to be deter-ined. At present, thyroidectomy without lymphode dissection is the standard treatment for pap-
llary thyroid cancer in a patient whose neck is
cp
ist(w
fttmchttntm
biCtOfsasmpppcclho
mThcRbor
waapmptwscpmepnttirlwdt
mpwhlbecltideibtlbdui
al
LYMPHATIC MAPPING IN HEAD AND NECK CANCER 391
linically negative for pathology. However, 82% ofatients may have occult metastasis.46
Using blue dye only, Dixon et al47 found SLNsn 65% of patients with thyroid neoplasms. Theeries included two SLNs that yielded false-nega-ive results. Other investigators reported a higher88%) success rate for identifying SLNs in patientsith thyroid disease.48
Lymphatic mapping is also being investigatedor use in treating medullary thyroid cancer. Whenhe technique is used for that condition, 7% ofhyroid metastases appear in the lateral compart-ent only and bypass the central lymph node
ompartment.46 SLN mapping can thus identifyigh-risk nodes in the lateral compartments whenhese nodes otherwise would be missed by dissec-ion in the central compartment only. The useful-ess of this procedure for detecting occult metas-ases may help reduce the high recurrence rate foredullary thyroid cancer.The technique of lymphatic mapping and SLN
iopsy described by Dixon et al47 involves inject-ng blue dye directly into the tumor or nodule.are is taken not to stain any surrounding struc-
ures with blue dye. All SLNs are then removed.ne key technical point noted by Dixon et al47 to
acilitate visualization of SLNs is to minimize dis-ection before injecting the blue dye. Dixon et al47
lso noted that the parathyroid glands can alsotain blue and thus cause a parathyroid gland to beislabeled as an SLN. Because few data on lym-
hatic mapping and SLN biopsy for thyroid neo-lasms are available and because these methodsresent technical issues that must be resolved, weannot draw any conclusions at this time. Futureonsiderations should include preoperative use ofymphoscintigraphy in conjunction with theandheld gamma probe to facilitate identificationf SLNs in the lateral compartments.
LYMPHATIC MAPPING AND SLN BIOPSYFOR MERKEL CELL CANCER
Merkel cell cancer is an uncommon, highlyalignant neuroendocrine tumor of the skin.reatment for this neoplasm is controversial andas included surgery with wide excision alone orombined with elective lymph node dissection.adiation therapy and chemotherapy also haveeen used. Despite our best efforts at treatment,verall survival has not improved, and recurrence
ates remain clinically significant. Most surgeons t
ould agree that Merkel cell cancer in the headnd neck must be treated aggressively because thispproach may help improve the prognosis.49 As inatients with melanoma, regional nodal status re-ains the most important predictor of survival in
atients with Merkel cell cancer.50 A more selec-ive approach is possible for identifying patientsho might benefit from elective lymph node dis-
ection; this possibility for patients with Merkelell cancer is intriguing because many of theseatients are older and have an increased risk fororbidity. As in patients with melanoma, preop-
rative lymphoscintigraphy can be performed inatients with Merkel cell cancer to identify theodal basins at risk. The combined technique canhen be done to identify all SLNs. Several inves-igators have shown that this procedure for treat-ng Merkel cell cancer is technically feasible andesults in minimal morbidity.51 However, until aarge, randomized prospective trial is completed,e cannot recommend this procedure as the stan-ard of care for patients with Merkel cell cancer ofhe head and neck.
CONCLUSIONS
SLN biopsy remains an evolving treatment forelanoma of the head and neck. Clinicians should
resent all treatment options to patients diagnosedith intermediate-thickness melanoma in theead and neck. To reduce operative morbidity,
ymphatic mapping and SLN biopsy using a com-ined technique can be offered as an alternative tolective lymph node dissection in patients withlinically negative nodes. However, close fol-ow-up is necessary. Patients must be informedhat this technique is not yet the standard of caren the community. The success of this techniqueepends on particularly complex technical consid-rations as well as on the experience of the treat-ng surgeon. Patients who choose not to have SLNiopsy and who prefer elective lymph node dissec-ion should be encouraged to have preoperativeymphoscintigraphy. This procedure shows nodalasins at high risk and can identify ambiguousrainage patterns. These patterns may either besed in lymph node dissection or be closely mon-tored postoperatively.
Lymphatic mapping in SCC of the oral cavity islso evolving and presently appears feasible, butarge multi-institutional studies are needed before
he technology can be recommended for this con-
dancsihbLtcsht
st
p
ma7
tHlN
sv
ca2
n1
sc2
ln
pn
bom
nr
lR
mdl4
sn1
aL
rA
la
lf1
lo
lR
cl
sd
ct
at
ovo
sn9
lc
td
392 BALKISSOON, RASGON, AND SCHWEITZER
ition. Until then, the technology can be used asn adjunct to standard treatment for the clinicallyegative neck. Preoperative lymphoscintigraphyan allow for better preoperative planning so thatelective neck dissection can be modified accord-ngly. The gamma probe can thus assure that alligh-risk nodes have been removed from the nodalasin after selective neck dissection is complete.ast, the gamma probe can then be used to identifyhe SLN in the just-dissected tissue. These SLNsan be dissected out from the selected neck dis-ection, labeled separately, and sent for detailedistopathologic examination. This approach isherefore consistent with oncologic principles.
Finally, lymphatic mapping for thyroid cancer istill experimental and warrants further investiga-ion before any recommendations can be made.
ACKNOWLEDGMENT
The Medical Editing Department, Kaiser Foundation Hos-itals, Inc, provided editorial assistance.
REFERENCES
1. Urist MM, Balch CM, Soong SJ, et al: Head and neckelanoma in 534 clinical stage I patients. A prognostic factors
nalysis and results of surgical treatment. Ann Surg 200:769-75, 19842. Robbins KT, Clayman G, Levine PA, et al: Neck dissec-
ion classification update: Revisions proposed by the Americanead and Neck Society and the American Academy of Oto-
aryngology—Head and Neck Surgery. Arch Otolaryngol Headeck Surg 128:751-758, 20023. Kapteijn BA, Nieweg OE, Liem IH, et al: Localizing the
entinel node in cutaneous melanoma: gamma probe detectionersus blue dye. Ann Surg Oncol 4:156-160, 19974. Rasgon BM: Use of low-dose technetium Tc 99m sulfur
olloid to locate sentinel lymph nodes in melanoma of the headnd neck: Preliminary study. Laryngoscope 111:1366-1372,0015. Chao C, Wong SL, Edwards MJ, et al: Sentinel lymph
ode biopsy for head and neck melanomas. Ann Surg Oncol0:21-26, 20036. Eicher SA, Clayman GL, Myers JN, et al: A prospective
tudy of intraoperative lymphatic mapping for head and neckutaneous melanoma. Arch Otolaryngol Head Neck Surg 128:41-246, 20027. Bostick P, Essner R, Sarantou T, et al: Intraoperative
ymphatic mapping for early-stage melanoma of the head andeck. Am J Surg 174:536-539, 19978. O’Brien CJ, Uren RF, Thompson JF, et al: Prediction of
otential metastatic sites in cutaneous head and neck mela-oma using lymphoscintigraphy. Am J Surg 170:461-466, 19959. Leong SP, Achtem TA, Habib FA, et al: Discordancy
etween clinical predictions vs lymphoscintigraphic and intra-perative mapping of sentinel lymph node drainage of primaryelanoma. Arch Dermatol 135:1472-1476, 1999
10. Davison SP, Clifton MS, Kauffman L, et al: Sentinel
ode biopsy for the detection of head and neck melanoma: Aeview. Ann Plast Surg 47:206-211, 2001
11. Wells KE, Rapaport DP, Cruse CW, et al: Sentinelymph node biopsy in melanoma of the head and neck. Plasteconstr Surg 100:591-594, 199712. Carlson GW, Murray DR, Greenlee R, et al: Manage-ent of malignant melanoma of the head and neck using
ynamic lymphoscintigraphy and gamma probe-guided sentinelymph node biopsy. Arch Otolaryngol Head Neck Surg 126:33-437, 200013. Wagner JD, Park HM, Coleman JJ 3rd, et al: Cervical
entinel lymph node biopsy for melanomas of the head andeck and upper thorax. Arch Otolaryngol Head Neck Surg26:313-321, 200014. Peralta EA, Yarington CT, Glenn MG: Malignant mel-
noma of the head and neck: Effect of treatment on survival.aryngoscope 108:220-223, 199815. Alex JC, Krag DN, Harlow SP, et al: Localization of
egional lymph nodes in melanomas of the head and neck.rch Otolaryngol Head Neck Surg 124:135-140, 199816. Shah JP, Kraus DH, Dubner S, et al: Patterns of regional
ymph node metastases from cutaneous melanomas of the headnd neck. Am J Surg 162:320-323, 1991
17. Morton DL, Wen DR, Foshag LJ, et al: Intraoperativeymphatic mapping and selective cervical lymphadenectomyor early-stage melanomas of the head and neck. J Clin Oncol1:1751-1756, 199318. Ollila DW, Foshag LJ, Essner R, et al: Parotid region
ymphatic mapping and sentinel lymphadenectomy for cutane-us melanoma. Ann Surg Oncol 6:150-154, 199919. Wells KE, Cruse CW, Daniels S, et al: The use of
ymphoscintigraphy in melanoma of the head and neck. Plasteconstr Surg 93:757-761, 199420. O’Brien CJ, Petersen-Schaefer K, Ruark D, et al: Radi-
al, modified, and selective neck dissection for cutaneous ma-ignant melanoma. Head Neck 17:232-241, 1995
21. Crile G: Excision of cancer of the head and neck. Withpecial reference to the plan of dissection based on one hun-red and thirty-two operations. JAMA 47:1780-1786, 190622. Shah JP, Candela FC, Poddar AK: The patterns of
ervical lymph node metastases from squamous carcinoma ofhe oral cavity. Cancer 66:109-113, 1990
23. Yuen AP, Lam KY, Chan AC, et al: Clinicopathologicalnalysis of elective neck dissection for N0 neck of early oralongue carcinoma. Am J Surg 177:90-92, 1999
24. Brazilian Head and Neck Cancer Study Group: Resultsf a prospective trial on elective modified radical classicalersus supraomohyoid neck dissection in the management ofral squamous carcinoma. Am J Surg 176:422-427, 199825. Pitman KT, Johnson JT, Myers EN: Effectiveness of
elective neck dissection for management of the clinicallyegative neck. Arch Otolaryngol Head Neck Surg 123:917-22, 199726. Cunningham MJ, Johnson JT, Myers EN, et al: Cervical
ymph node metastasis after local excision of early squamousell carcinoma of the oral cavity. Am J Surg 152:361-366, 1986
27. Lindberg R: Distribution of cervical lymph node metas-ases from squamous cell carcinoma of the upper respiratory andigestive tracts. Cancer 29:1446-1449, 1972
issections for squamous carcinoma of the upper aerodigestiveract: patterns of regional failure. Head Neck 21:499-505, 1999
29. Khafif A, Lopez-Garza JR, Medina JE: Is dissection ofevel IV necessary in patients with T1-T3 N0 tongue cancer?aryngoscope 111:1088-1090, 200130. Byers RM, Weber RS, Andrews T, et al: Frequency and
herapeutic implications of “skip metastases” in the neck fromquamous carcinoma of the oral tongue. Head Neck 19:14-19,99731. Bergman SA, Ord RA, Rothman M: Accuracy of clin-
cal examination versus computed tomography in detectingccult lymph node involvement in patients with oral epider-oid carcinoma. J Oral Maxillofac Surg 52:1236-1239, 199432. Braams JW, Pruim J, Freling NJ, et al: Detection of
ymph node metastases of squamous-cell cancer of the head andeck with FDG-PET and MRI. J Nucl Med 36:211-216, 199533. Taylor RJ, Wahl RL, Sharma PK, et al: Sentinel node
ocalization in oral cavity and oropharynx squamous cell can-er. Arch Otolaryngol Head Neck Surg 127:970-974, 2001
34. Civantos FJ, Gomez C, Duque C, et al: Sentinel nodeiopsy in oral cavity cancer: Correlation with PET scan andmmunohistochemistry. Head Neck 25:1-9, 2003
35. Shoaib T, Soutar DS, Prosser JE, et al: A suggestedethod for sentinel node biopsy in squamous cell carcinoma of
he head and neck. Head Neck 21:728-733, 199936. Shoaib T, Soutar DS, MacDonald DG, et al: The accu-
acy of head and neck carcinoma sentinel lymph node biopsy inhe clinically N0 neck. Cancer 91:2077-2083, 2001
37. Koch WM, Choti MA, Civelek AC, et al: Gammarobe-directed biopsy of the sentinel node in oral squamous cellarcinoma. Arch Otolaryngol Head Neck Surg 124:455-459,99838. Pitman KT, Johnson JT, Edington H, et al: Lymphaticapping with isosulfan blue dye in squamous cell carcinoma of
he head and neck. Arch Otolaryngol Head Neck Surg 124:90-793, 199839. Stoeckli SJ, Steinert H, Pfaltz M, et al: Sentinel lymph
ode evaluation in squamous cell carcinoma of the head andeck. Otolaryngol Head Neck Surg 125:221-226, 2001
40. Chiesa F, Mauri S, Grana C, et al: Is there a role forentinel node biopsy in early N0 tongue tumors? Surgery 128:6-21, 200041. Mozzillo N, Chiesa F, Botti G, et al: Sentinel node
iopsy in head and neck cancer. Ann Surg Oncol 8:103S-105S,001 (suppl)42. Carvalho AL, Kowalski LP, Borges JA, et al: Ipsilateral
eck cancer recurrences after elective supraomohyoid neckissection. Arch Otolaryngol Head Neck Surg 126:410-412,00043. Hosal AS, Carrau RL, Johnson JT, et al: Selective neck
issection in the management of the clinically node-negativeeck. Laryngoscope 110:2037-2040, 200044. Enepekides DJ, Sultanem K, Nguyen C, et al: Occult
ervical metastases: Immunoperoxidase analysis of the patho-ogically negative neck. Otolaryngol Head Neck Surg 120:713-17, 199945. Ferlito A, Devaney KO, Rinaldo A, et al: Micrometas-
ases: Have they an impact on prognosis? Ann Otol Rhinolaryngol 108:1185-1189, 199946. Noguchi S, Noguchi A, Murakami N: Papillary carci-
oma of the thyroid. II. Value of prophylactic lymph nodexcision. Cancer 26:1061-1064, 1970
47. Dixon E, McKinnon JG, Pasieka JL: Feasibility of sen-inel lymph node biopsy and lymphatic mapping in nodularhyroid neoplasms. World J Surg 24:1396-1401, 2000
48. Kelemen PR, Van Herle AJ, Giuliano AE: Sentinelymphadenectomy in thyroid malignant neoplasms. Arch Surg33:288-292, 198949. Shack RB, Barton RM, DeLozier J, et al: Is aggressive
urgical management justified in the treatment of Merkel cellarcinoma? Plast Reconstr Surg 94:970-975, 1994
50. Victor NS, Morton B, Smith JW: Merkel cell cancer: Isrophylactic lymph node dissection indicated? Am Surg 62:79-882, 199651. Alex JC, Krag DN: The gamma-probe-guided resection
f radiolabeled primary lymph nodes. Surg Oncol Clin N Am:33-41, 1996
