second neoplasm in patients with head and neck cancer

SECOND NEOPLASM IN PATIENTS WITH HEAD ANDNECK CANCER

Xavier Leo n, MD,1 Miquel Quer, MD, 1 Santiago Diez, MD, 1 Cesar Oru s, MD,1

Antonio Lo´pez-Pousa, MD, 2 Joaquim Burgue ´s, MD1

1 Department of Otolaryngology, Hospital de la Santa Creu i Sant Pau, Avda. San Antoni Ma Claret, 167, 08025Barcelona, Spain2 Department of Oncology, Hospital de la Santa Creu i Sant Pau, Universitat Autonoma de Barcelona,Barcelona, Spain

Accepted 18 May 1998

Abstract: Background. The improvement in locoregional con-trol of head and neck carcinomas over the last decades does notappear to modify the final survival of these patients, mainly due tothe appearance of distant metastases and second neoplasms.We ran a study to evaluate the incidence of second neoplasmsand their characteristics in patients with head and neck carci-noma treated in our hospital.

Methods. A retrospective study was undertaken to analyzethe appearance of second neoplasm, its characteristics, and re-percussion on the survival in 1845 patients with carcinoma of theoral cavity, oropharynx, hypopharynx, and larynx diagnosed andtreated in our department between 1984 and 1995.

Results. A group of 302 patients (16%) developed a secondneoplasm. The incidence of second neoplasms was 4% per yearand was constant during the follow-up period. Second neoplasmswere mainly carcinomas located in head and neck (40%), lungs(31%), or esophagus (9%). Twenty percent of second neoplasmswere tumors located outside the aerodigestive tract. The appear-ance of a second neoplasm significantly worsened the survival.

Conclusions. Four percent per year of patients with carci-noma of the oral cavity, oropharynx, hypopharynx, and larynxdevelop a second neoplasm, most frequently a carcinoma of theaerodigestive tract. This leads to a significant worsening of sur-

vival. © 1999 John Wiley & Sons, Inc. Head Neck 21: 204–210,1999.

Keywords: second neoplasm; head and neck cancer; index tu-mor

Asignificant improvement in locoregional controlof head and neck squamous cell carcinomas hasbeen seen over the last decades, thanks to theintroduction of new surgical techniques, im-proved radiotherapy techniques, and the use ofchemotherapy. However, this improvement doesnot appear to influence the final survival in thisgroup of patients. The main factors accounting forthis divergency in patients who have achieved lo-coregional control of head and neck tumor are dis-tant metastases and second neoplasms.

The criteria to consider a tumor as a secondneoplasm were defined by Warren and Gates1 in1932. They stated that each tumor must be con-firmed histologically, the possibility that one ofthe tumors is a metastasis of the other must beruled out, and there must not be a submucousconnection between the two tumors. The indextumor is the first diagnosed tumor, and the sec-

Correspondence to: X. Leon

CCC 1043-3074/99/030204-07© 1999 John Wiley & Sons, Inc.

204 Second Neoplasms HEAD & NECK May 1999

ond neoplasm is any malignant tumor discoveredthereafter. From a chronological point of view, thesecond neoplasm is synchronous if it is diagnosedwithin 6 months after the diagnosis of the indextumor and metachronous if it is diagnosed after aperiod of 6 months. Synchronous tumors are si-multaneous if they are discovered at the sametime as the index tumor or during the workup ofthe index tumor.

The incidence of a second neoplasm in patientswho survive a head and neck carcinoma has beenreported by different authors to be between 3%and 7% per year.2,3 In patients with squamouscell carcinoma of the head and neck, there is astrong tendency for the second neoplasm to be asquamous carcinoma of the aerodigestive tract:mucous membranes of head and neck, esophagus,and lungs. In 1953, Slaughter et al4 proposed theconcept of the ‘‘condemned mucosa syndrome’’:the carcinogen effects from tobacco and alcoholthat promote the appearance of the squamous cellcarcinoma index tumor in a location of the aero-digestive tract also act on all the mucosas of theaerodigestive tract, thus promoting the appear-ance of new carcinomas.

The objectives of our study were to analyze theincidence of second neoplasms in patients withhead and neck carcinomas, to evaluate the signifi-cance of the different epidemiologic characteris-tics that could influence their appearance, to es-tablish the possible relationships between theindex tumor and the second neoplasm, and toquantify the impact that the appearance of secondneoplasms has on survival.

PATIENTS AND METHODS

The data used in this study were obtained retro-spectively from a database which prospectivelycollects the information about the clinical statusof the patient, including carcinogen use, tumorcharacteristics, treatment, and follow-up of allthe patients with a head and neck malignant tu-mor diagnosed in our hospital since 1984.

A total of 2109 patients with head and necksquamous cell carcinoma were diagnosed in ourdepartment in the period 1984–1995. We ex-cluded 57 patients who were treated in other hos-pitals, and 43 patients were lost to follow-up inthe first 2 years. The study was thus made up ofthe 2009 patients with head and neck carcinomadiagnosed and treated in our department andwith a minimum follow-up of 2 years. We consid-ered two groups: those with index tumors at loca-tions epidemiologically related to the abuse of to-

bacco and alcohol (oral cavity, orohypopharynx,and larynx) and those with locations not relatedto this abuse (rhinopharynx, ear, nose and para-nasal sinuses, and salivary glands).

The routine follow-up program in our hospitalconsists of the recent medical history and locore-gional examinations at 2-month intervals duringthe first year, 3-month intervals in the secondyear, 4-month intervals between the third andfifth years, and every 6 months thereafter. All thepatients underwent an annual chest radiography.The control visits were carried out in a rotatorymanner by the departments involved in the treat-ment of the patients: otorhinolaryngology, radio-therapy, and medical oncology. We considered asindex tumor the first head and neck carcinomadiagnosed. When more than one carcinoma of thehead and neck was diagnosed simultaneously, weconsidered as index the carcinoma that broughtabout the symptoms leading to the diagnosis ofhead and neck carcinoma.

Eighty-eight patients had a malignant tumorprior to the study. In 26, the previous tumor waslocated in the aerodigestive tract (16 in the lungs,2 in the esophagus, and 9 in the head and neck).

We follow the Warren and Gates1 criteria fordefining a second tumor. We evaluated the inci-dence of second neoplasms as well as their char-acteristics: location of the index tumor, locationand histology of the second neoplasm, interval be-tween the index tumor and the second neoplasm,and patient survival in relation to the appearanceof a second neoplasm. The possibility that someepidemiologic factors could increase the risk ofappearance of a second neoplasm was studied bya Cox proportional hazard regression model.

The statistical analysis was done with anSPSS program. The chi-square method was usedto compare frequencies and the actuarial Kaplan-Meier method was used to calculate the risk ofappearance of a second neoplasm and the surviv-als. The log-rank test was performed to comparethe survival functions. In the Cox-regressionmodel, the ‘‘forward conditional’’ technique wasused. We accepted p < .05 as the limit of statisti-cal significance.

RESULTS

Global of Patients. Three-hundred fifteen pa-tients (15.7%) had a second malignant tumor, 26a third, and 9 a fourth malignant tumor. Table 1shows the distribution of the second tumors inrelation to the location of the index tumor.Twenty-four of the 26 third tumors were carcino-

Second Neoplasms HEAD & NECK May 1999 205

mas of the aerodigestive tract (5 located in theesophagus and 19 in the head and neck). All ofthem appeared in patients in whom the secondtumor was a head and neck carcinoma.

Seven of the 9 fourth tumors were carcinomasof the aerodigestive tract (1 located in the lungs, 2in the esophagus, and 4 in the head and neck). Allof them appeared in patients in whom the thirdtumor was a head and neck carcinoma.

Seventy-nine percent of the second neoplasmswere carcinomas of the aerodigestive tract. Thefrequency of second neoplasm in index tumorswith locations typically related to the overuse oftobacco and alcohol, such as in the oral cavity,orohypopharynx, and larynx was 16%, 80% ofthese being carcinomas of the aerodigestive tract.For other locations in the head and neck, in whichthe appearance of carcinomas is not related to theoveruse of tobacco and alcohol, such as rhinophar-ynx, ear, nose and paranasal sinuses, or salivaryglands, the frequency of second neoplasm was 8%,significantly lower (chi-square 4 7.9, p 4 .005).In these locations not related to tobacco and alco-hol, almost 50% of the second neoplasms were notcarcinomas of the aerodigestive tract.

Index Tumors Located in the Oral Cavity, Orohypo-pharynx, and Larynx. Due to the differences in thefrequency of second neoplasms and their charac-teristics related to the location of the index tumor,to achieve a more homogeneous group, from thispoint we are going to study only the second neo-plasms that appeared in patients with index tu-mors in the oral cavity, orohypopharynx, and lar-ynx. Of the 1845 patients with index tumors inthese locations, 302 (16%) had a second neoplasm.Of these, 57 (19%) were synchronous simulta-neous tumors, 29 (10%) were synchronous nonsi-multaneous tumors, and 216 (71%) were meta-

chronous. Two hundred forty-two secondneoplasm (80%) were carcinomas of the aerodi-gestive tract. Figure 1 shows the actuarial sur-vival free of second neoplasm. From the diagnosisof the index tumor, patients with carcinomas lo-cated in the oral cavity, orohypopharynx, and lar-ynx had an actuarial risk for second neoplasms of4% per annum. Looking at the shape of the sur-vival curve, it is of interest that this risk wascontinuous throughout the follow-up period.

We studied different epidemiologic variables(age when the index tumor was diagnosed, sex,overuse of tobacco and alcohol, differentiation,stage and location of the index tumor) with theCox proportional hazard regression model toevaluate their possible influence on the appear-ance of second neoplasm (Table 2). Of the vari-ables studied, only previous history of tobaccooveruse and location of the index tumor appearedas significant prognostic factors. Patients whosmoked fewer than 20 cigarettes per day had arisk 1.6 higher than nonsmokers of second neo-

Table 1. Distribution of second neoplasms in relation to the location of the index tumor.

Lungs EsophagusHead

and neck Others Total

Oral cavity (n = 215) 4 2 24 7 37 (17%)Oropharynx (n = 369) 12 9 36 9 66 (18%)Hypopharynx (n = 187) 13 3 12 2 30 (16%)Larynx (n = 1074) 66 13 48 42 169 (16%)

Supraglottis (n = 439) 35 5 18 19 77 (18%)Glottis* (n = 635) 31 8 30 23 92 (14%)

Rhinopharynx (n = 106) 2 0 2 4 8 (8%)Others† (n = 58) 3 0 0 2 5 (7%)Total 100 (32%) 27 (9%) 122 (39%) 66 (20%) 315

*Grouping glottis, transglottis and subglottis†Including ear, nose, paranasal sinuses and salivary glands.

FIGURE 1. Actuarial survival free of second neoplasm for pa-tients with index tumors in the oral cavity, orohypopharynx, andlarynx.


plasm. Patients who smoked more than 20 ciga-rettes per day had a risk 1.7 higher than did non-smokers of second neoplasm.

Using index carcinomas in the oral cavity asreference category, the patients with index carci-nomas located in the oro-hypopharynx showed nosignificant differences in the risk of second neo-plasm, whereas the relative risk for the patientswith index tumors located in the larynx was 0.7with regard to the reference category (p 4 .002).

To analyze the location of the second neoplasmthat appeared in the aerodigestive tract in rela-tion to the location of the index tumor, wegrouped the second neoplasms of the aerodiges-tive tract in accordance with two axes: secondneoplasms located in the larynx and lungs wereconsidered inside the respiratory axis, and secondneoplasms in the hypopharynx and esophaguswere considered inside the digestive axis. Secondneoplasms in the oral cavity and oropharynx wereconsidered mixed. Table 3 shows the proportion ofsecond tumors in each of the axes in relation tothe location of the index tumor.

There were significant differences in the loca-tion of the second aerodigestive tract neoplasmsin relation to the location of the index tumor (chi-square 4 107, p < .0001). When the index tumorwas located in the oral cavity and oropharynx, themost frequent location of the second neoplasmwas again in the oral cavity and oropharynx.When the index tumor was located in the hypo-pharynx, the most frequent location of the secondneoplasm was in the digestive axis (hypopharynxand esophagus). Finally, when the index tumorwas located in the larynx, the most frequent loca-tion of the second neoplasm was in the respiratoryaxis (larynx and lungs).

When we analyzed the actuarial observed sur-vival of the patients in relation to the appearanceof second neoplasm, we observed that patients

with a second neoplasm had a 22% 10 years’ ob-served survival rate (95% confidence interval[CI], 14% to 31%), significantly lower than the55% 10 years’ survival rate for patients withoutsecond neoplasm (95% CI, 51% to 58%) (Figure 2).

Considering the actuarial observed survivalfrom the diagnosis of the second neoplasm for pa-tients with a second neoplasm, we observed thatthe better prognosis was for patients with a sec-ond neoplasm located in the head and neck, witha 49% 5 years’ actuarial survival, as comparedwith the 15% survival when the second neoplasmwas in the lungs or the 0% survival when it was inthe esophagus. When the second neoplasms ap-peared outside the aerodigestive tract, the 5years’ actuarial survival was 25% (Figure 3).

DISCUSSION

The appearance of a second neoplasm is one of thefactors that limits the final survival of patientswith head and neck carcinomas when locoregionalcontrol of the tumor is achieved.

A second malignant tumor was seen in 15.7%

Table 3. Location of the second aerodigestive tract neoplasmsin relation to the location of the index tumor.*

Index tumorRespiratory

axisDigestive

axis Mixed

Oral cavity (n = 30) 7 (23%) 3 (10%) 20 (67%)Oropharynx (n = 57) 16 (28%) 9 (16%) 32 (56%)Hypopharynx (n = 28) 12 (43%) 14 (50%) 2 (7%)Larynx (n = 127) 100 (79%) 18 (14%) 9 (7%)Total 135 (56%) 44 (18%) 63 (26%)

*Respiratory axis: larynx and lungs; digestive axis: hypopharynx andesophagus; mixed: oral cavity and oropharynx.

Table 2. Results of the multivariate analysis according to theCox proportional hazard regression model.

OD95% Confidence

interval p

Nonsmoker 1Smoker < 20 cig/day 1.6 1.1–2.3 .01Smoker $ 20 cig/day 1.7 1.2–2.3 .002Location index tumor

Oral cavity 1Oropharynx 1.2 0.9–1.6 .1Hypopharynx 1.2 0.9–1.6 .3Larynx 0.7 0.6–0.8 .0002

FIGURE 2. Actuarial observed survival for patients with indextumors in the oral cavity, orohypopharynx, and larynx in relationto the appearance of second neoplasm.


of our patients with a head and neck carcinoma.Of these, 79% were carcinomas of the aerodiges-tive tract. When we considered two groups of headand neck carcinomas in relation to the locationsepidemiologically related with overuse of tobaccoand alcohol, we observed that the frequency ofsecond neoplasm for index tumors with locationsrelated to tobacco and alcohol (oral cavity, orohy-popharynx, and larynx) was more than doublethat for index tumors with locations not relatedwith these. Moreover, in the group with locationsrelated to tobacco and alcohol, 80% of the secondneoplasms were carcinomas of the aerodigestivetract, whereas the frequency of aerodigestivetract carcinomas in the group of index tumors notrelated to tobacco and alcohol was lower than50%. In 1953, in a study with patients with oralcavity carcinomas, Slaughter et al4 proposed thetheory of ‘‘condemned mucosa syndrome.’’ Accord-ing to this theory, the carcinomatous promotingaction of certain substances contained in the to-bacco and alcohol that promoted the first carci-noma would act over the rest of the mucous cov-erage of the aerodigestive tract, promoting theappearance of more carcinomas over these muco-sas. The epidemiologic results of our study sup-port Slaughter’s theory.

In the 1845 patients with index tumors in theoral cavity, oro-hypopharynx, and larynx, 4.7%(86/1845) had a second synchronous neoplasm,and 66% of these (57/86) were simultaneous tu-mors. In a meta-analysis carried out by Haugheyet al,5 the prevalence of a second synchronousneoplasm reported in the literature oscillated be-tween 0.8% and 18%. The differences between theresults could be due to the kind of study. In this

way, in prospective studies, the average preva-lence of synchronous second neoplasms was 9%,significantly higher than the global average. Webelieve it is of little value to compare the preva-lence of metachronous second neoplasms betweenthe different series, as the final frequency de-pends on the duration of the follow-up. It is morerelevant to compare the annual incidence of sec-ond neoplasms. Different authors have pointedout that patients with head and neck carcinomahave an annual risk between 3% and 7% of ap-pearance of a metachronous second neoplasm.2,3

In our study, we found an annual 4% risk of meta-chronous second neoplasm in the patients with anindex carcinoma in the oral cavity, orohypophar-ynx, and larynx. It is important to point out thatthis risk is constant throughout the follow-up pe-riod, as has already been established by studieswith long-term follow-up.6

In relation to the location of the second neo-plasm, the majority of authors agree that themost frequent location of the second neoplasm isin the head and neck, followed by lungs. The re-sults of Haughey’s meta-analysis5 showed that35% of the second neoplasms were head and neckcarcinomas, 25% lung carcinomas, 9% esophaguscarcinomas, and the remaining 31% were malig-nant tumors outside the aerodigestive tract.When the index tumor was located in the larynx,the most frequent location of the second neoplasmwas in the lungs, whereas when the index headand neck tumor had another location, the mostfrequent location of the second neoplasm was an-other head and neck carcinoma. Panosetti et al7

have pointed out that the location of the secondneoplasm is related to the location of the indextumor: the tumors have a tendency to follow arespiratory axis (larynx–lung) or a digestive axis(pharynx–esophagus). Silvestri et al8 studied therisk of second pulmonary cancer in patients withlaryngeal carcinoma in relation to the location ofthe index carcinoma inside the larynx. Theyfound that the risk of developing a second pulmo-nary tumor, especially of undifferentiated carci-nomas, was higher for supraglottic carcinomasthan for glottic carcinomas.

Our results agree with the literature: the mostfrequent second neoplasms were head and neckcarcinomas (39%), followed by lung carcinoma(32%) and esophagus (9%). Twenty per cent of tu-mors appeared outside the aerodigestive tract.When the index tumor was located in the larynx,the second neoplasms were located preferably inthe respiratory axis; when the index tumor was

FIGURE 3. Actuarial observed survival for patients with indextumors in the oral cavity, orohypopharynx, and larynx who devel-oped a second neoplasm in relation to the location of the secondneoplasm.


located in the hypopharynx, the second neoplasmwas located preferably in the digestive axis; andwhen the index tumor was located in the oral cav-ity or oropharynx, the second neoplasm was againlocated preferably in the oral cavity and orophar-ynx.

When we analyzed the appearance of secondneoplasms in patients with an index tumor in thelarynx in relation with the sublocation of the in-dex tumor, our results were similar to those ob-tained by Silvestri et al.8 The tumors located inthe supraglottic area had a higher frequency ofsecond neoplasm than glottic carcinomas. This in-crease in the frequency of second neoplasms wasat the expense of a higher frequency of secondpulmonary neoplasms: the 5 years’ actuarial sur-vival free of a second pulmonary tumor was 88%for supraglottic carcinomas and 94% for glotticcarcinomas (log-rank test, p 4 .02).

The results of the multivariate analysisshowed that the factors that significantly influ-enced the appearance of a second neoplasm wereprevious history of tobacco overuse and laryngeallocation of the tumor.

Smokers had a risk of between 1.6 and 1.7higher than nonsmokers of developing a secondneoplasm. This means that between 36% and 39%of the incidence of second neoplasm could be at-tributed to the overuse of tobacco. Such estimatesrefer to the overuse of tobacco prior to the diag-nosis of the index tumor.

The majority of our patients gave up smokingafter diagnosis and treatment of the index tumor,but as we do not have information about the pro-portion of patients who continued to smoke, wecannot evaluate the influence of continuing smok-ing on the appearance of second neoplasms. Sil-verman et al9 and Day et al10 have shown a de-creased incidence of second neoplasm in patientswho stop smoking, but other authors11,12 have notfound such a decrease.

In the multivariate analysis, the laryngeal lo-cation of the index tumor had a lower risk of sec-ond neoplasm, whereas there were no significantdifferences between the other locations.

Attempts have been made to identify patientswith a higher susceptibility to develop a secondneoplasm. Hsu et al,13 measuring the number ofchromatid breaks per cell of cultivated lympho-cytes treated with bleomycin, developed a methodto evaluate the individual susceptibility to themutagenic action of certain carcinogens. In astudy carried out by Cloos,14 the susceptibility tothe action of bleomycin was significantly higher

in a group of patients with head and neck carci-nomas than in a control group (p < .0001). Thehighest risk of head and neck carcinoma was ob-served in patients with a high sensitivity to bleo-mycin and a history of overuse of alcohol and to-bacco. The relative risk to develop a head andneck carcinoma for subjects with high sensitivityto bleomycin and overuse of tobacco and alcoholwas 58 (95% CI, 18–188), whereas the relativerisk for patients with sensitivity to bleomycinwithout overuse of tobacco and alcohol was only2.3 (95% CI, 0.5–9.4). In patients with a history ofoveruse of tobacco and alcohol but with a normalsensitivity to bleomycin, the relative risk of headand neck carcinoma was 11 (95% CI, 4–30).

Schantz15 followed 84 patients with head andneck carcinoma and observed that the relativerisk to develop a second neoplasm for patientswith a high sensitivity to bleomycin was 4.4 (95%CI, 1.2–16) in relation to patients with a normalsensitivity. A logistic-regression analysis showeda significant relationship between an increasingrisk of multiple malignancies and an incremen-tally increasing mutagen-sensitive value (p 4.02).

Other authors have studied the role of onco-genes in the susceptibility to develop a secondneoplasm in patients with head and neck carci-noma. Shin et al16 did not find significant differ-ences in the frequency of second neoplasms in pa-tients with p53 oncoprotein overexpression,although when a second neoplasm appeared, theinterval between index tumor and second neo-plasm for patients with p53 overexpression wassignificant lower. Furthermore, Gallo et al17 didnot find significant differences in the proportionof overexpression of p53 oncoprotein in head andneck carcinoma patients who develop multiplemalignancies, but, confirming the findings ofSchantz et al,15 they observed a correlation be-tween the sensitivity to the mutagenic action ofthe bleomycin and the appearance of second neo-plasms.

The importance of these studies lies in the pos-sibility to quantify the risk of appearance of acarcinoma of the aerodigestive tract and, for thosepatients who have developed an aerodigestivecarcinoma, the possibility to identify subjectswith a greater risk of developing multiple malig-nancies. In theory, this group of patients wouldbenefit most from preventive treatment withmedications that could lower the frequency of sec-ond tumors, such as 13-cis-retinoic acid.18


REFERENCES1. Warren S, Gates O. Multiple malignant tumors: a survey

of literature and statistical study. Am J Cancer 1932;51:1358–1414.

2. Sturgis EM, Miller RH. Second primary malignancies inthe head and neck cancer patient. Ann Rhinol Otol Laryn-gol 1995;104:946–954.

3. Cooper JS, Pajak TF, Rubin P, et al. Second malignanciesin patients who have head and neck cancer: incidence,effect on survival and implications based on the RTOGexperience. Int J Radiat Oncol Biol Phys 1989;17:449–456.

4. Slaughter DP, Southwick HW, Smejkal W. Field cancer-ization in oral stratified squamous epithelium. Cancer1953;6:963–968.

5. Haughey BH, Gates GA, Arkfen CL, Harvey J. Meta-analysis of second tumors in head and neck cancer: thecase for an endoscopic screening protocol. Ann Otol RhinolLaryngol 1992;101:105–112.

6. Tepperman BS, Fitzpatrick PJ. Second respiratory andupper digestive tract cancers after oral cancer. Lancet1981;2:547–549.

7. Panosetti A, Luboinski B, Mamelle G, Richard JM. Mul-tiple synchronous and metachronous cancers of the upperaerodigestive tract: a nine-year study. Laryngoscope1989;99:1267–1273.

8. Silvestri F, Bussani R, Cosatti C, Bosatra A. High relativerisk of a second pulmonary cancer in patients affected bylaryngeal cancer: differences by specific site of occurrenceand lung cancer histotype. Laryngoscope 1994;104:222–225.

9. Silverman S, Gorsky M, Greenspan D. Tobacco usage inpatients with head and neck carcinomas: a follow-upstudy on habit changes and second primary oropharyn-geal cancers. J Am Dent Assoc 1983;106:33–35.

10. Day GL, Blot WJ, Shore RE, et al. Second cancer followingoral and pharyngeal cancer: role of the tobacco and alco-hol. J Natl Cancer Inst 1994;86:131–137.

11. Castigliano SG. Influence of continued smoking on theincidence of second primary cancers involving mouth,pharynx , and larynx. J Am Dent Assoc 1968;77:580–585.

12. Shottenfeld D, Gantt RC, Wynder EL. The role of alcoholand tobacco in multiple priamry cancers of the upper di-gestive system, larynx and lung: a prospective study. PrevMed 1974;3:277–293.

13. Hsu TC, Johnston DA, Cherry LM, et al. Sensitivity togenotoxic effects of bleomycin in humans: possible rela-tionship to environmental carcinogenesis. Int J Cancer1989;43:403–409.

14. Cloos J, Spitz MR, Schantz SP, et al. Genetic susceptibil-ity to head and neck squamous cell carcinoma. J NatlCancer Inst 1996;88:530–535.

15. Schantz SP, Spitz MR, Hsu TC. Mutagen sensitivity inpatients with head and neck cancers: a biologic marker forrisk of multiple primary malignancies. J Natl Cancer Inst1990;82:1773–1775.

16. Shin DM, Lee JS, Lippman SM, et al. p53 Expression:predicting recurrence and second primary tumors in headand neck squamous cell carcinoma. J Natl Cancer Inst.1996;88:519–529.

17. Gallo O, Bianchi S, Giovannucci-Uzzielli ML, et al. p53Oncoprotein overexpression correlates with mutagen-induced chromosome fragility in head and neck cancerpatient with multiple malignancies. Br J Cancer 1995;71:1008–1012.

18. Benner SE, Pajak TF, Lippman SM, Early C, Hong WK.Prevention of second primary tumors with isotretinoin inpatients with squamous cell carcinoma of the head andneck: long term follow-up. J Natl Cancer Inst 1994;86:140–141.


second neoplasm in patients with head and neck cancer

Documents