sign in sign up
<<
Home , Adenoacanthoma, Adenocarcinoma

J Clin Pathol: first published as 10.1136/jcp.33.11.1064 on 1 November 1980. Downloaded from

J Clin Pathol 1980;33:1064-1067

Adenoacanthoma of the endometrium: a separate entity or a histological curiosity?

H BARROWCLOUGH AND KW JAARSMA From the Laboratory for Pathological Anatomy, University of Amsterdam, Wilhelmina Gasthuis, Amsterdam, The Netherlands

SUMMARY In a series of 42 patients with endometrial and of 53 with adeno- , age at the time of diagnosis, age at the onset of the menopause, gravidity, pathological staging, and survival were compared to see if there was any significant difference, apart from mor- phology, between the two tumours. No significant differences could be established, and it was concluded that adenoacanthoma should be regarded as a histological variant of and not as a separate entity.

Adenoacanthoma is a malignant epithelial tumour them with a similar group of patients with adeno- composed of both glandular and squamous ele- carcinoma. The intention was to establish if there ments. This was first described by Herxheimer in was an essential difference, apart from morphology, 19071 in gall bladder, stomach, caecum, , between the two tumours. and parotid as well as in the endometrium. copyright. Since about 1940 it is principally the adenoacan- Material and methods thoma of the endometrium that has held the atten- tion of pathologists. There is no agreement in the All patients who had been operated upon for a literature as to the frequency of this lesion; some malignant tumour of the corpus uteri in the Wilhel- reports regard the lesion as relatively rare,2-6 while mina Gasthuis, Amsterdam, between 1960 and 1969 others record an incidence of about 30 % of all endo- inclusive were reviewed. All patients had undergone metrial .7-12 a vaginal after closure of the cervix

A similar discrepancy exists in the prognosis. In over an alcohol-soaked swab. The patients were not http://jcp.bmj.com/ 1957, Novak and Nalley13 reported a case of endo- irradiated either pre- or post-operatively. metrial adenoacanthoma and stated that the occur- The epithelial malignancies were reclassified into rence of squamous nests within an adenocarcinoma adenocarcinoma, clear cell carcinoma, and adeno- indicated a better prognosis than for an ordinary . As has already been pointed out, there adenocarcinoma. Since then, a worse prognosis has are varying definitions in the literature as to what been suggested6 714-17 while others could establish precisely consitutes an adenoacanthoma. Since the

no apparent difference between adenoacanthoma and only characteristic feature of this lesion is squamous on September 30, 2021 by guest. Protected adenocarcinoma.2 1218-21 The consensus now is contiguous with adenocarcinoma it was that the two lesions have a similar prognosis. decided that all tumours containing unmistakable Furthermore, there is no uniform definition of squamous epithelium intimately admixed with adenoacanthoma in the literature. Babid18 refers to a glandular carcinoma would be classified as adeno- tumour in which there are 'acanthomatous elements acanthoma, irrespective of the quantity of that intimately admixed with glandular carcinoma'. squamous epithelium. Pokoly'0 defines it as an adenocarcinoma in which From the total patient material 95 cases were one nest of squamous epithelium can be found per 10 selected for further study on the basis ofthe adequacy high-power fields. Woodington,2' on the other hand, of the clinical and pathological documentation. specifies that the ratio of squamous to adenoid Documentation was considered to be adequate in epithelium must be 1:30. those patients in whom staging of the tumour could It was the object of this study to examine a series be accurately assessed and who had been followed of patients with adenoacanthoma and to compare up for at least five years. The following classification was used: stage 0, Received for publication 25 March 1980 superficial tumour growth without myometrial 1064 J Clin Pathol: first published as 10.1136/jcp.33.11.1064 on 1 November 1980. Downloaded from

Adenoacanthoma of the endometrium: a separate entity or a histological curiosity invasion and confined to the body of the ; marised in the Table. There were no stage IV cases stage LA, invasion of the myometrium-less than among the selected patients of either group. The half the full thickness of the myometrium invaded; total survival of the pure adenocarcinoma group, stage IB, invasion of the myometrium-more than defined as tumour-free after five years or dead after half the full thickness of the myometrium invaded; not less than three years without tumour, was 83 %, stage II, invasion of the cervix; stage III, tumour while the total survival for the adenoacanthoma present outside the uterus but confined to the true group was 78.6 %. The survival figures for the pelvis; stage IV, distant metastases and/or invasion individual stages are given in the Table. of bladder and/or rectum. A limited number of clinical details was also con- Discussion sidered, namely, age of patient at the time of diagnosis, age at onset of the menopause, number of Adenoacanthoma of the body of the uterus is a pregnancies, and survival. The selected cases of tumour surrounded by confusion. There is no adenocarcinoma and adenoacanthoma were then adequate estimate of its frequency, there is still some compared on the basis of these data. disagreement with regard to prognosis, and, further- more, there is no uniformly accepted definition of Results what precisely constitutes the lesion. This point readily explains the great variation in the reported Altogether 149 patients were operated upon for an frequency of the tumour. epithelial of the corpus uteri during the The definition used in the present series is less 10-year period beginning in 1960. Of these patients, strict than others whereby a moderately large 96 (65.8 %) had an adenocarcinoma, 5 (3.3 %) had a percentage of tumours was classified as adenoacan- clear cell carcinoma, and 46 (39.9 %) had an thoma. Even so, the incidence in this study cor- adenoacanthoma. responds with that of other workers.1"'4 If those Of the 149 patients, 53 with adenocarcinoma, one tumours in this series that fulfil the definition of

of which showed the features of a clear cell Pokoly10 are selected, the incidence falls to 15%, copyright. carcinoma, and 42 with adenoacanthoma were which is lower than the 23-9 % found by Pokoly, but considered to be adequately documented and were other factors remain similar to those of the original selected for further study. The results are sum- adenoacanthoma group. The rather generous http://jcp.bmj.com/ on September 30, 2021 by guest. Protected

High-power detail ofa focus ofsquamous metaplasia within an adenoacanthoma (Haematoxylin and eosin x 140). J Clin Pathol: first published as 10.1136/jcp.33.11.1064 on 1 November 1980. Downloaded from

1066 Barrowclough and Jaarsma A comparison between adenocarcinoma and differentiation, and tumours arising from them are adenoacanthoma with regard to pathological staging and gradually being recognised. Since the cervix and the survival endometrium both arise from Mullerian epithelium, Adenocarcinoma Adenoacanthoma similar cells might be expected to be present in the endometrium, but, as yet, no one has been able to No. ofpatients 53 42 demonstrate their existence. No. premenopausal 12 5 Mean age at menopause (yr) 48 (SD 5-4) 49-3 (SD 4-1) A more acceptable theory is that the squamous Mean age at diagnosis (yr) 60-6 (SD 83) 63-2 (SD 9-0) epithelium arises by metaplasia of the tumorous Mean no. ofpregnancies 2-2 (SD 2-2) 2-6 (SD 2-3) Percentage nulliparous patients 22-6 16-7 glandular epithelium.12-14 23 This theory would Stage 0 explain the localisation of the squamous nests deep No. ofpatients 7 6 down in the tumour. Furthermore, a diligent search Percentage oftotal 13-2 14-3 No. premenopausal 3 1 will sometimes reveal transitional zones between the Mean age at menopause (yr) 40-0 (SD 6-1) 51-4(SD 2-9) two epithelial types, such as was described by Mean age at diagnosis (yr) 52-3 (SD 3-1) 59-5 (SD 7-5) Tweedale et al.'2 Mean no. ofpregnancies 2-1 (SD 1-3) 3-2 (SD 1-6) % survival 100 100 The consequences of this theory are of vital Stage IA importance in understanding the nature of adeno- No. ofpatients 27 23 Percentage oftotal 50-9 54-8 acanthoma, for it follows that the squamous No. premenopausal 7 3 epithelium has arisen directly from malignant Mean age at menopause (yr) 50-0 (SD 3-5) 49-8 (SD 4-0) epithelial cells and therefore must itself be malignant, Mean age at diagnosis (yr) 61-5 (SD 7-1) 62-2 (SD 9-5) Mean no. ofpregnancies 2-4 (SD 2-7) 2-3 (SD 2-5) however benign it may appear histologically. It % survival 92-6 86-9 further follows that the mixed or adenosquamous Stage IB carcinoma is simply an adenoacanthoma in which No. ofpatients 9 9 Percentage oftotal 17-0 21-4 the squamous component reveals its inherent No. premenopausal 0 0 malignancy. It is not, therefore, an entity apart but a Mean age at menopause (yr) 47-5 (SD 5-3) 47-1 (SD 4-3) well differentiated variant of Mean age at diagnosis (yr) 60-4 (SD 5-5) 67-8 (SD 7-1) less adenoacanthoma.

Mean no. ofpregnancies 1-9 (SD 2-4) 2-8 (SD 2-2) It must now be considered how the metaplasia copyright. % survival 88 77-8 takes place. During the course of the present study Stages II and III No. ofpatients 10 4 we noticed that in about 80% of adenoacanthomas Percentage oftotal 18-9 9-5 inflammatory infiltrate was present and that in No. premenopausal 2 approximately 35 % this was quite dense. Only 60% Mean age at menopause (yr) 47-9 (SD 6-3) 49-0 (SD 5-0) Mean age at diagnosis (yr) 64-1 (SD 12-1) 64-0(SD 11-0) of the pure were associated with Mean no. of pregnancies 1-9 (SD 1-1) 2-5 (SD 2-6) inflammation, the infiltrate being dense in approxi- % survival 40 0 mately 20%. It is possible, therefore, that the

metaplasia takes place, at least in part, as a result of http://jcp.bmj.com/ inflammatory irritation. selection in this study does not therefore seem to Squamous metaplasia within an adenocarcinoma influence the validity of the comparisons. appears to have a modal distribution whereby some A further problem arose in 1956 when Glucksmann tumours undergo spontaneous metaplasia while and Cherry22 described a mixed carcinoma in the others do not, even in the presence of severe inflam- uterine cervix. This tumour consisted of malignant mation. Between these two extremes there is a wide squamous epithelium within an adenocarcinoma. A range of tumours whose metaplastic response to similar entity in the endometrium was later described inflammation is extremely variable. on September 30, 2021 by guest. Protected both as a mixed carcinoma8 9 23 24 and, latterly, as The question then arises whether this metaplastic an .1"19 25 This tumour change indeed identifies a separate entity, or merely is now usually held to be a lesion distinct from represents a histological variant of adenocarcinoma adenoacanthoma. which is of no further clinical significance. In order to understand the adenoacanthoma and The results presented in this study reveal no its doubly malignant counterpart, adenosquamous significant differences between the two lesions. With carcinoma, the origin of the squamous epithelium the exception of stage II and stage III tumours, the must be considered. Several theories have been prognosis was similar, the small differences being no considered in the past, of which two have received more than could be expected to occur by chance. the most support: the first is that the squamous The difference in stages IL and III can be explained epithelium arises from indifferent cells in the by the very small numbers involved. Furthermore, endometrium,2 5 619 26 which correspond with the the distribution of staging was similar in both reserve cells in the cervix. These cervical reserve tumours. cells are capable of both glandular and squamous The mean ages at onset of the menopause and at J Clin Pathol: first published as 10.1136/jcp.33.11.1064 on 1 November 1980. Downloaded from

Adenoacanthoma of the endometrium: a separate entity or a histological curiosity? 1067 the time of diagnosis showed no statistically signifi- adenoacanthoma. Obstet Gynecol 1964;23:611-9. 13 Novak ER, Nalley WB. Uterine adenoacanthoma. Obstet cant differences between the two groups, nor did the Gynecol 1957 ;9:396-402. mean gravidity values. 14 Charles D. Endometrial adenoacanthoma. 1965; We conclude, therefore, that adenoacanthoma 18:737-50. does not differ significantly from adenocarcinoma in 15 Liggins GC, Way S. A comparison of the prognosis of clinical presentation or adenoacanthoma and adenocarcinoma of the corpus biological behaviour and uteri. J Obstet Gynaec Brit Emp 1960;67:294-6. that it should be regarded as a histological variant of 16 Morrison DL. Adenoacanthoma of the uterine body. J adenocarcinoma and not as a separate entity. Obstet Gynaec Brit Cwlth 1966;73 :605-10. 17 Renning EL, Javert CT. Analysis of a series of cases of References carcinoma of the endometrium treated by radium and operation. Am J Obstet Gynecol 1964;88:171-7. Herxheimer G. Uber heterologe Cancrolde. Beitr Pathol 18 Babid AO, Kurohara SS, Vongtama VY, Selim MA, 1907 ;41 :348-412. Webster JH. Biological behavior of adenoacanthoma of 2 Marcus SL. Adenoacanthoma of the endometrium. Am J endometrium. Am J Obstet Gynecol 1970;106:205-9. Obstet Gynecol 1961 ;81 :259-67. 19 Salazar OM, DePapp EW, Bonfiglio TA, Feldstein ML, 3 Healy WP, Cutler M. Radiation and surgical treatment of Rubin P, Rudolph JH. Adenosquamous carcinoma of carcinoma of the body of the uterus. Am J Obstet the endometrium. Cancer 1977;40:119-30. Gynecol 1930;19:457-89. 20 Williams GL. Adenoacanthoma of the corpus uteri. J 4 Piver MS. Distant of adenoacanthoma of the Obstet Gynaec Brit Cwlth 1965;72:674-6. endometrium. Am J Obstet Gynecol 1966;96:1011-6. 21 Woodington GF, Dockerty MB, Welch JS. Adenoacan- Postoloff AV, Rodenberg TA. Uterine adenoacanthoma thoma of the endometrium. Am J Surg 1963 ;106:915-21. with metastasis. Obstet Gynecol 1954;4:642-6. 22 Glucksmann A, Cherry CP. Incidence, , and 6 Schattenberg HJ, Ziskind J. Adenoacanthoma of the response to radiation of mixed (adenoacan- uterus. Am J Obstet Gynecol 1940;39:1 12-6. thomas) of the uterine cervix. Cancer 1956;9:971-9. 7 Chanen W. A clinical and pathological study of adenoacan- 23 Aikawa M, Ng ABP. Mixed (adenosquamous) carcinoma thoma of the uterine body. J Obstet Gynaec Brit Emp of the endometrium: electron microscopic observations. 1960;67:287-93. Cancer 1973;31:385-97. 8 Ng ABP. Mixed carcinoma of the endometrium. Am J 24 Skinner IC, McDonald JR. Mixed adenocarcinoma and Obstet Gynecol 1968;102:506-15. of the uterus. Am J Obstet Gynecol 1940;40:258-66. 9 Ng ABP, Reagan JW, Storaasli JP, Wentz WB. Mixed copyright. adenosquamous carcinoma of the endometrium. Amer J 21 Haqqani MT, Fox H. Adenosquamous carcinoma of the Clin Pathol 1973;59:765-81. endometrium. J Clin Pathol 1976;29:959-66. 10 Pokoly T. A comparison of the clinical behaviour of 26 Marshall D, Sedlis A, Tchertkoff V. Adenoacanthoma of uterine adenocarcinomas and adenoacanthomas. Am J the corpus uteri with distant metastases. Obstet Gynecol Obstet Gynecol 1970;108:1080-4. 1963 ;22:578-82. Sliverberg SG, Bolin MCr, De Giorgi LS. Adenoacanthoma and mixed adenosquamous carcinoma of the endo- Requests for reprints to: Dr H Barrowclough, Laboratory metrium. Cancer 1972;30:1307-14. for Pathological Anatomy, University of Amsterdam, 12 Tweedale DN, Early LS, Goodsitt ES. Endometrial Wilhelmina Gasthuis, Amsterdam, The Netherlands. http://jcp.bmj.com/ on September 30, 2021 by guest. Protected