Abstract

Background: The association of myasthenia gravis(MG) with thymoma is well recognized. Our clinicalimpression has been that MG associated withthymoma may be more common in patients of Poly-nesian descent than in other races.

Aim: To determine the influence of ethnicity on theassociation of MG with thymoma in our population.

Method: Review of all cases of thymectomy per-formed at Greenlane Hospital in Auckland for the20-year period from June 1978 to June 1998.

Results: There were 103 thymectomies performed inthe study period. Fifty-five thymomas were identified,15 in subjects of Maori or Pacific Island ethnicity and40 in subjects of other races, predominantlyCaucasian.Ten of 15 Maori or Pacific Island subjectswith thymoma had MG (67%), compared with 15 of 40 subjects of other races (37.5%, P = 0.05). Themean age of Maori or Pacific Island subjects with

thymoma and MG was 42.5 years, compared with56.3 years in subjects from other races (P = 0.06). Allfive Maori and Pacific Island subjects with invasivethymoma had MG, whereas only four of 15 subjects(27%) from other races with invasive tumours hadMG (P < 0.01). The overall incidence of thymomaand the proportion of thymomas that were invasivedid not differ between the ethnic groups.

Conclusions: Myasthenia gravis with thymoma occursmore frequently among Maori or Pacific Islandpeople than in other racial groups in our population.This is due to an increase in the proportion of caseswith thymoma who have MG in this group, while theoverall frequency of cases of thymoma is similarbetween groups. MG with thymoma in the Maori orPacific Island populations also presents at a youngerage and is more often associated with tumourinvasion. (Intern Med J 2001; 31: 206–210)

Key words: Maori, myasthenia gravis, Polynesian,thymoma.

ORIGINAL ARTICLE

Myasthenia gravis with thymoma is more common in theMaori and Pacific Island populations in New Zealand

J. N. FINK,1 W. E. WALLIS1 and D. A. HAYDOCK2

1Department of Neurology, Auckland Hospital and 2Department of Cardiothoracic Surgery, Greenlane Hospital,Auckland, New Zealand

INTRODUCTION

The association of myasthenia gravis (MG) withthymic abnormalities is well recognized.1–4 The mostcommon abnormality is thymic hyperplasia, butthymoma may occur in up to 22% of cases of MG.5

Most series report that thymoma is present in 9–14%of cases of MG.6–11 MG with thymoma tends topresent at an older age, does not show the female sexpreponderance seen in non-thymomatous MG and

carries a poorer prognosis, with more severe myas-thenic symptoms, a poorer response to thymectomyand a higher mortality rate than MG withoutthymoma.5,6,9–16 A higher than expected occurrencerate of thymoma with MG among the Chinese popu-lation in Hong Kong has been reported.17 It has beenour clinical impression in Auckland that thymomaassociated with MG may be more common in patientsof Polynesian descent than in other races. This obser-vation has not been subject to systematic investigationpreviously.

METHODS

Greenlane Hospital (GLH) is the sole centre forcardiothoracic surgery for the upper North Island of

Correspondence to: John N. Fink, Stroke Division, Dana 710, Beth IsraelDeaconess Medical Center, 330 Brookline Ave, Boston, MA 02215,USA. Email: jfink@caregroup.harvard.edu

Received 20 July 2000; accepted 10 January 2001.

Internal Medicine Journal 2001; 31: 206–210

207Myasthenia gravis in New Zealand

Internal Medicine Journal 2001; 31: 206–210

New Zealand and serves a population of over 1.5million people. The records of all patients who had a thymectomy for MG at GLH between June 1978and June 1998 were examined. Cases of MG were included only if the clinical diagnosis of MGwith muscle weakness and fatiguability had beenconfirmed by neurophysiological criteria, thepresence of anti-acetylcholine receptor (AChR) anti-bodies or unequivocal response to the edrophoniumtest. In addition, the records of all patients who hadthymectomy for thymoma and did not have MGwere examined as a control group. Patients who had had thymectomy for anterior mediastinal massand had non-thymomatous histologies, such asHodgkin’s disease, carcinoid tumours etc., wereexcluded from the analysis. Demographic data,including gender, age and ethnicity, were obtainedfor each case. Thymoma staging was according tothe scale of Masaoka and colleagues.18 Stage Itumours were classified as benign and stage II–IVtumours as invasive. The ethnic distribution of thegeneral population was obtained from the results ofthe 1996 New Zealand census. The presentation,preoperative treatment and severity of MG wererecorded for all patients with MG, with and withoutthymoma, along with the severity and treatment ofMG at long-term follow up. However, these data arenot presented here and will be the subject of anotherpaper.

RESULTS

The total number of thymectomies performed in the20-year study period was 103 (Table 1). Sixty-threethymectomies were performed for MG; thymoma waspresent in 25. Of the 40 subjects undergoing thymec-tomy for an anterior mediastinal mass without MG,30 had thymoma and 10 had other diagnoses, forexample Hodgkin’s disease, carcinoid tumour,multiple endocrine neoplasia syndrome, teratoma,benign thymic cyst and thyroid nodule. These caseswere excluded from further analysis. The totalnumber of cases with thymoma was 55.

Ethnicity and thymoma

Myasthenia gravis was present in association withthymoma more frequently among Maori or PacificIsland subjects than among those from other racialbackgrounds (Table 2). Of the 55 subjects withthymoma, 15 were of Maori or Pacific Islandethnicity and 40 were of other races, predominantlyCaucasian. Ten of 15 Maori or Pacific Island subjects

with thymoma also had MG (67%), a significantlygreater proportion than in subjects from other raceswith thymoma, 15 of 40 of whom had MG (37.5%,χ2 = 3.74, P = 0.05). Overall, 25 of 55 subjects withthymoma had MG (45%). Confining the analysis tothose who resided in the greater Auckland regionallowed comparison with census data.There were 42subjects with thymoma residing in the greaterAuckland region, with 12 (29%) of Maori or PacificIsland ethnicity. This proportion was not signifi-cantly different from the general population, whereMaori or Pacific Island people comprised 24% of theAuckland region’s population at the time of the 1996census (χ2 = 0.64, P = 0.42). Of the 12 Maori orPacific Island subjects in this region with thymoma,nine (75%) had MG, whereas 14 (47%) of the 30subjects with thymoma from other races had MG (χ2 = 2.78, P = 0.09). The overall incidence ofthymoma was therefore no greater than expectedamong Maori or Pacific Island people, but there wasa greater association of MG with thymoma in thispopulation. A multivariate model incorporatingethnicity, sex, age ≤ 50 or > 50 and thymoma stagewas analysed by logistic regression. The associationof Maori or Pacific Island ethnicity and MG wasconfirmed with a similar level of statistical signifi-cance in this model to the univariate analysis (P = 0.065). None of the other variables approachedstatistical significance.

Table 1 Thymectomies at Greenlane Hospital betweenJune 1978 and June 1998

Thymoma No Totalthymoma

Thymectomy for MG 25 38 63Thymectomy without MG 30 10 40All thymectomies 55 48 103

MG, myasthenia gravis.

Table 2 Ethnicity of subjects with thymoma

M/PI Other

All thymomaMG 10 15No MG 5 25 P = 0.05

Auckland regionCensus 1996 24% 76%All thymoma 12 (29%) 30 (71%) P = 0.42MG 9 14No MG 3 16 P = 0.09

M/PI, Maori and Pacific Island subjects; MG, myasthenia gravis.

Fink et al.208

Myasthenia gravis with thymoma:ethnic differences

There were 25 subjects with MG and thymoma, 10 ofMaori or Pacific Island ethnicity.

Ethnicity and age

Thymoma with MG was present at a younger age inthe Maori and Pacific Island group than in other races(Table 3). The mean age of the 10 Maori or PacificIsland subjects with thymoma and MG was 42.5 years(range 16–69), compared with 56.3 years (22–75) inthe 15 subjects from other races. The difference justfailed to reach statistical significance (P = 0.06).Seven of the 10 (70%) Maori or Pacific Islandsubjects with thymoma and MG were aged 50 yearsor less, whereas only five of 15 (33%) of those of otherraces were ≤ 50 years of age (χ2 = 3.23, P = 0.07).There was no difference between the racial groups inthe age distribution of subjects without MG. Theincreased association of MG with thymoma in Maorior Pacific Island people was particularly strong in theyounger age group. Among the 55 subjects withthymoma in the study, 25 were ≤ 50 years of age.Seven of nine Maori or Pacific Island subjects in thisage group had MG, compared with five of 16 subjectsof other races (χ2 = 5.0, P = 0.03).

Ethnicity and gender

No significant gender differences were present (Table 3). Three of 10 (30%) Maori or Pacific Islandsubjects and eight of 15 (53%) subjects of other

races with MG and thymoma were female (χ2 = 1.32,P = 0.25). Overall, 25 of 55 subjects with thymomawere female.

Ethnicity and thymoma stage

There was no ethnic difference in the prevalence ofMG among those with benign thymomas (Table 3):MG was present in five of 10 Maori or Pacific Islandsubjects (50%) and 11 of 25 subjects of other races(44%, χ2 = 0.10, P = 0.75). In those with invasivetumours, there was a significant ethnic difference inthe prevalence of MG. All five Maori or Pacific Islandsubjects with invasive thymoma had MG, whereasonly four of 15 (27%) subjects from other races withinvasive tumours had MG (χ2 = 8.15, P < 0.01). Theproportion of thymomas that were invasive did notdiffer between the ethnic groups. Invasive tumourswere present in five of 15 (33%) Maori or PacificIsland subjects and in 16 of 40 (40%) subjects ofother races with thymoma (P = 0.76).

There were no statistically significant differencesbetween the racial groups in mode and severity ofpresentation of myasthenia or long-term outcome.There were no statistically significant differencesbetween Maori subjects and Pacific Island subjects forany of the recorded parameters.

DISCUSSION

The results of the present study confirm that MGoccurs with thymoma more frequently in the Maori orPacific Island populations than in other racial groupsin Auckland. In addition, MG with thymoma presentsmore frequently in the younger (less than 50 years)age group in Maori or Pacific Island people. Epidemi-ological data support using 50 years of age as theminimum age for late-onset MG;19 however, thefinding of an increased proportion of younger caseswith MG and thymoma in the Maori or Pacific Islandgroup retains the same degree of statistical signifi-cance when 30 or 40 years of age is used to define thegroups. No conclusion about the overall proportion ofthose in each racial group presenting with MG thatare found to have thymoma can be made from thepresent study, because there are no data available ofthe overall prevalence of MG in the community. It isnot known whether there is a similar increase in theincidence of MG in Maori and Pacific Island peoplewithout thymoma.

Benign or invasive thymomas themselves do notoccur more commonly in the Maori or Pacific Island

Internal Medicine Journal 2001; 31: 206–210

Table 3 Ethnic differences in MG with thymoma

M/PI Other

AgeMean (range) 42.5 (16–69) 56.3 (22–75) P = 0.06Age ≤ 50 7 (70%) 5 (33%)Age > 50 3 (30%) 10 (67%) P = 0.07

GenderMale 7 (70%) 7 (47%)Female 3 (30%) 8 (53%) P = 0.25

Histology and MGBenign

MG 5 (50%) 11 (44%)No MG 5 (50%) 14 (56%) P = 0.75

InvasiveMG 5 (100%) 4 (27%)No MG 0 11 (73%) P < 0.01

M/PI, Maori and Pacific Island subjects; MG, myasthenia gravis.

209Myasthenia gravis in New Zealand

populations, but are more likely to be associated with MG, particularly in the case of invasive thymoma.Theproportion of non-Polynesian subjects with thymomawho developed MG in our series (37.5%) was similarto that reported in a review of 1655 published cases of thymoma (30%),20 but the rate of MG in Maori orPacific Island subjects with thymoma was significantlygreater (67%). One possibility is that this is due tofailure of Maori or Pacific Island people to presentwith thymoma in the absence of symptoms of MG.We consider this to be an unlikely explanation for the observed difference, because the proportion of Maori or Pacific Island people among subjects withthymoma was at least equal to that in the generalpopulation according to census data, suggesting thatthymoma is no more likely to be diagnosed in oneethnic group than in the other. In addition, it seemslikely that benign thymomas would be more likely toremain undetected than invasive thymomas and anapparent increase in the rate of MG associated withbenign thymomas would result from underdiagnosisof these tumours in the absence of MG. This was notseen in our study, where the increase in MG was seenin those with invasive tumours among the Maori orPacific Island group.

Ethnic differences in the presentation of MG withthymoma have not been well described previously.Teoh et al. have suggested that there is an increasedincidence of thymoma among Hong Kong Chinesepatients with MG compared with that described inother populations.17 They have identified theEpstein–Barr virus (EBV) genome in two thymomasand have postulated an association with the increasedrate of EBV-related nasopharyngeal and salivarygland carcinoma in their population. Their data werebased on a surgical series of Hong Kong Chinesepatients without a direct control group, however, andthe relatively high proportion of thymomas found intheir population could have reflected a low rate ofthymectomy in non-thymomatous MG. A laterepidemiological study of adult Hong Kong Chinesepatients with MG8 has not confirmed an increasedincidence of thymoma among all patients with MG(14%) compared with historical series elsewhere,6–11

although a high proportion of invasive thymomas wasfound. An increased rate of EBV-related nasopharyn-geal or salivary gland carcinomas is not recognizedwithin the Maori or Pacific Island populations in NewZealand and none of our patients with thymoma weretested for the presence of the EBV genome.

In comparison with thymoma-associated MG, ethnicdifferences in MG without thymoma have been studied

in more detail. A study of Chinese people in Taipei andCaucasian people in London with MG has discoveredthat Chinese people have a greater proportion of veryearly onset (less than 20 years) and ocular MG andlack the late onset peak seen in the Caucasian popula-tion.21 An increased incidence and prevalence of MGhas been found among African-American people inVirginia compared with Caucasian people.22 Ethnicdifferences have also been demonstrated in juvenileMG.23 Differences in human leucocyte antigen (HLA)and immunoglobulin allotype associations with MGare also seen between ethnic groups.24,25

The pathogenesis of MG with thymoma differs fromthat of non-thymomatous MG, but the exact role ofthe thymoma in the pathogenesis of MG in thesepatients has not been determined. Antibodies toAChR are found in both thymomatous and non-thymomatous MG, but while the AChR is found inmyoid cells in tumour-free thymuses, it is notexpressed in thymoma.26 Antibodies to titin, a striatedmuscle antigen, are present in a high proportion ofcases of MG with thymoma and may be found insome cases of late-onset MG, but are rare in MG withthymic hyperplasia.27,28 It is likely that specificity ofthe autoimmune response is initiated by non-receptorproteins in thymomas that possess AChR or titinepitopes and that these positively select helper T cellsmaturing in the tumours or actively sensitize otherT cells.23,26–30 These epitopes are expressed morefrequently in thymomas with cortical differentiationthan in medullary thymomas.28 Cortical and predom-inantly cortical thymomas have a greater associationwith MG than medullary or mixed thymomas.4 Thishistological classification was not used in our seriesbefore 1993; thus, classification according to stageonly has been used in the present report. However,cortical and predominantly cortical thymomas alsohave a greater potential for local invasiveness and agreater metastatic potential than other thymoma typesand are likely to be overrepresented in the groups withinvasive tumours in our series. The increased associa-tion of MG with these tumours in the Maori andPacific Island populations may be a result of specificHLA immunotypes in this population that increasethe likelihood of producing an autoimmune responseagainst the AChR-related epitopes critical for thedevelopment of MG, which are expressed by cortical-type thymomas. Further study would be necessary toconfirm this hypothesis; however, different associa-tions of HLA and other T-lymphocyte antigens havebeen demonstrated between thymomatous and non-thymatous MG previously.31,32 A role for EBV

Internal Medicine Journal 2001; 31: 206–210

Fink et al.210

Internal Medicine Journal 2001; 31: 206–210

infection in the initiation of either specific tumourtypes or the autoimmune response in the Maori andPacific Island populations also remains possible.17

The investigation of thymoma tissue (with or withoutassociated MG) in Maori and Pacific Island peoplefor the presence of the EBV genome may also providegreater insights into the pathogenesis of this disorder.

To our knowledge, no study of MG and the Maori orPacific Island populations has been performed previ-ously. We conclude that MG with thymoma occursmore frequently in the Maori or Pacific Island popula-tions than in other racial groups in Auckland. This isdue to an increase in the proportion of cases withthymoma and MG in this group, while the overallnumber of cases of thymoma is similar. In addition, MGwith thymoma in the Maori or Pacific Island popula-tions presents more frequently at a younger age than inother races and is associated with invasive tumours.

REFERENCES1 Schleup M, Wilcox N, Ritter MA, Newsom-Davis J, Larche

M, Brown AN. Myasthenia gravis thymus: Clinical, histologicaland culture correlations. J Autoimmun 1988; 1: 445–67.

2 Whitaker JN. Myasthenia gravis and autoimmunity. AdvIntern Med 1980; 26: 489–510.

3 Weiner HL, Hauser SL. Neuroimmunology. I:Immunoregulation in neurological disease. Ann Neurol 1982;11: 437–49.

4 Hohlfield R, Wekerle H. The thymus in myasthenia gravis.Neurol Clin North Am 1994; 12: 331–42.

5 Beghi E, Antozzi C, Batocchi AP et al. Prognosis ofmyasthenia gravis: a multicenter follow-up study of 844patients. J Neurol Sci 1991; 106: 213–20.

6 Kula RW. Neuromuscular disorders associated with systemicneoplastic diseases. In: Vinken PJ, Bruyn GW (eds).Handbook of Clinical Neurology, Vol. 41. Amsterdam:Elsevier Science Publishers, 1979; 317–403.

7 Oosterhuis HJ. The natural course of myasthenia gravis: along term follow up study. J Neurol Neurosurg Psychiatr1989; 52: 1121–7.

8 Yu YL, Hawkins BR, Ip MS, Wong V, Woo E. Myastheniagravis in Hong Kong Chinese. 1. Epidemiology and adultdisease. Acta Neurol Scand 1992; 86: 113–19.

9 Donaldson DH, Ansher M, Horan S, Rutherford RB, RingelSP. The relationship of age to outcome in myasthenia gravis.Neurology 1990; 40: 786–90.

10 Christensen PB, Jensen TS, Tsiropooulos I et al. Mortalityand survival in myasthenia gravis: a Danish population basedstudy. J Neurol Neurosurg Psychiatr 1998; 64: 78–83.

11 Papatestas AE, Genkins G, Kornfield P et al. Effects ofthymectomy in myasthenia gravis. Ann Surg 1987; 206: 79–88.

12 Beekman R, Kuks JB, Oosterhuis HJ. Myasthenia gravis:diagnosis and follow-up of 100 consecutive patients. J Neurol1997; 244: 112–18.

13 Masaoka A,Yamakawa Y, Niwa H et al. Extended thymectomyfor myasthenia gravis patients: a 20-year review. Ann ThoracSurg 1996; 62: 853–9.

14 Monden Y, Nakahara K, Kagotani K et al. Myasthenia graviswith thymoma: analysis of postoperative prognosis for 65

patients with thymomatous myasthenia gravis. Ann ThoracSurg 1984; 38: 46–52.

15 Durelli L, Giuliano M, Casadio C, Ferri R, Rendine S,Bergamini L. Actuarial analysis of the occurrence ofremissions following thymectomy for myasthenia gravis in 400patients. J Neurol Neurosurg Psychiatr 1991; 54: 406–41.

16 Fujii N, Itoyama Y, Machi M, Goto I. Analysis of prognosticfactors in thymectomized patients with myasthenia gravis:correlation between thymic lymphoid cell subsets andpostoperative clinical course. J Neurol Sci 1991; 105: 143–9.

17 Teoh R, McGuire L, Wong K, Chin D. Increased incidence ofthymoma in Chinese myasthenia gravis: possible relationshipwith Epstein–Barr virus. Acta Neurol Scand 1989; 80: 221–5.

18 Masaoka A, Monden Y, Nakahara K, Tanioka T. Follow-upstudy of thymomas with special reference to their clinicalstages. Cancer 1981; 48: 2485–92.

19 Aarli JA. Late-onset myasthenia gravis: a changing scene. ArchNeurol 1999; 56: 25–7.

20 Namba T, Brunner NG, Grob D. Myasthenia gravis inpatients with thymoma, with particular reference to onsetafter thymectomy. Medicine 1978; 57: 411–33.

21 Chiu HC, Vincent A, Newsom-Davis J, Hsieh KH, Hung TP.Myasthenia gravis: population differences in diseaseexpression and acetylcholine receptor antibody titers betweenChinese and Caucasians. Neurology 1987; 37: 1854–7.

22 Phillips LH, Torner JC, Anderson MS, Cox GM. Theepidemiology of myasthenia gravis in central and westernVirginia. Neurology 1992; 42: 1888–93.

23 Andrews PI, Massey JM, Howard JF, Sanders DB. Race, sexand puberty influence onset, severity, and outcome in juvenilemyasthenia gravis. Neurology 1994; 44: 1208–14.

24 Hawkins BR, Chan-Lui WY, Choi EK, Ho AY. Strongassociation of HLA BW6 with juvenile onset myastheniagravis in Hong Kong Chinese. J Neurol Neurosurg Psychiatr1984; 47: 555–7.

25 Chiu HC, Lange GG, Wilcox N et al. Immunoglobulinallotypes in Caucasian and Chinese myasthenia gravis:differences from Japanese patients. J Neurol NeurosurgPsychiatr 1988; 51: 214–17.

26 Kirchner T, Tzartos S, Hoppe F et al. Pathogenesis ofmyasthenia gravis. Acetylcholine receptor-related antigenicdeterminants in tumor-free thymuses and thymic epithelialtumors. Am J Pathol 1988; 130: 268–80.

27 Aarli JA. Late onset myasthenia gravis. Eur J Neurol 1997;4: 203–9.

28 Marx A, Schultlz A, Wilisch A et al. Paraneoplasticautoimmunity in thymus tumors. Dev Immunol 1998; 6:129–40.

29 Marx A, Wilisch A, Schultz A et al. Expression ofneurofilaments and of a Titin epitope in thymic epithelialtumors. Implications for the pathogenesis of myastheniagravis. Am J Pathol 1996; 148: 1839–50.

30 Nagvekar N, Moody AM, Moss P et al. A pathogenetic rolefor the thymoma in myasthenia gravis. Autosensitization ofIL-4-producing T cell clones recognizing extracellularacetylcholine receptor epitopes presented by minority class IIisotypes. J Clin Invest 1998; 101: 2268–77.

31 Spurkland A, Gilhus NE, Ronningen KS, Aarli JA, Vartdal F.Myasthenia gravis patients with thymus hyperplasia andmyasthenia gravis patients with thymoma display differentHLA associations. Tissue Antigens 1991; 37: 90–3.

32 Huang D, Liu L, Noren K et al. Genetic association of Ctla-4to myasthenia gravis with thymoma. J Neuroimmunol 1998;88: 192–8.