Search tips
Search criteria 


Logo of brjopthalBritish Journal of OphthalmologyVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Br J Ophthalmol. 2007 September; 91(9): 1239–1243.
Published online 2007 April 12. doi:  10.1136/bjo.2007.118661
PMCID: PMC1954920

Expression of apoptosis‐related markers in malignant epithelial tumours of the lacrimal gland and their relation to clinical outcome



To investigate the correlation between the expression of apoptosis‐related markers and prognosis in malignant epithelial tumours of the lacrimal gland.

Materials and Methods

Series of cases.


Twenty one cases with malignant epithelial tumours of the lacrimal gland. Histological diagnosis was re‐examined and blocks selected were evaluated for the following parameters: incidence of apoptosis with TUNEL assay, expression of p53 and Bcl‐2 using monoclonal antibody. Predictors factors for survival, local recurrence and cumulative probability of death were statistically evaluated.


Re‐eximination of the 21 specimens was as follow: 11 adenoid cystic carcinomas, 4 mucoepidermoid carcinomas, 3 squamous cell carcinomas and 3 adenocarcinomas. Eleven of the 21 patients (53%) died during the follow‐up period (4–192 months; mean 71). Bcl‐2 staining >6% was significantly correlated with the death of patients. A statistically significant positive relationship for TUNEL and p53, and an inverse correlation for Bcl‐2 staining, was demonstrated with overall survival.


The correlation with survival of apoptotic index, p53 and Bcl‐2 expression suggest the more tumour cells go in apoptosis, upregulating p53 and down‐regulating Bcl‐2, the better the survival of patients. This study establishes a role of apoptosis‐regulatory proteins in the pathogenesis of malignant epithelial lacrimal gland tumours, and supports the hypothesis that evaluation of the expression of apoptosis‐related markers in these tumours may provide a prognostic tool.

Primary malignant neoplasms of the lacrimal gland are very rare, accounting for only 2% of biopsy specimens from expanding orbital masses.1,2 Histologically half of these carcinomas are adenoid cystics with the remaining divided between malignant mixed tumours and other carcinomas. These neoplasms have high morbidity and mortality, and the correlation of histological type to survival is still under discussion.3,4,5,6,7,8,9,10,11 The effect of different surgical procedures, such as local resection or wide cranio‐orbital resection, as well as the use of the adjuvant radiotherapy on tumour recurrence remains uncertain.7

Since it has been proposed that defects in the apoptotic pathway may represent a critical element in the progression of neoplastic disease and may also determine treatment efficacy at the cellular level,12,13,14 we have studied the role of apoptosis in the progression of malignant tumours of the lacrimal glands in order to improve the knowledge of the prognostic factors of these tumours. Apoptosis is a form of cell death, which is regulated at the gene level and plays a central role in cell number control during embryonic development and organogenesis.15,16 Through TUNEL technique, apoptosis can be detected at single‐cell level in formalin‐fixed, paraffin‐embedded tissue sections, thereby allowing the study of retrospective cases.17

In the present study, we analysed by TUNEL assay the incidence of apoptosis in a group of 21 cases of malignant epithelial tumour of the lacrimal glands and examined the expression of two apoptosis regulators—p53 and Bcl‐2. We correlated these results with the pathological features of the neoplasms and the clinical outcome.

As these tumours are rare, very little information about the molecular changes leading to their development and progression has so far been reported.

Materials and methods


Tissues from 21 malignant epithelial tumours of the lacrimal glands were used for this study. All patients were diagnosed and treated at the “Federico II” University of Naples. Survival data were collected from hospital charts and from periodic interviews with patients and their relatives. The pertinent clinical and histopathological data of all the patients are listed in table 11.

Table thumbnail
Table 1 Clinical and histopathological characteristics of patients


The formalin‐fixed, paraffin‐embedded samples were sectioned at 5 μm and stained with H&E. The histological diagnosis was re‐examined. In addition, the most representative blocks were selected to be cut into new 5 μm‐thick sections for immunohistochemical studies.


Immunohistochemical staining was carried out using the standard avidin‐biotin‐peroxidase complex technique and the LV Dako LSAB kit (DAKO A/S, Carpinteria, CA, USA) after antigen retrieval by pressure cooking. After quenching in 3% hydrogen peroxide and blocking, the sections were incubated with primary antibodies overnight at 4°C, using the following dilutions: p53 monoclonal antibody (clone n° DO‐7, DAKO) at 1:150 and Bcl‐2 monoclonal antibody (clone n°124, DAXO) at 1:250. Biotynilated secondary antibody and streptavidin conjugated to horseradish peroxidase were subsequently applied with colour development with 3,3′‐diaminobenzidine and haematoxylin counterstaining ((figsfigs 1 and 22).). Negative control slides in the absence of primary antibody were included for each staining.

figure bj118661.f1
Figure 1 Expression of Bcl‐2 in in a case adenoid cystic carcinoma.
figure bj118661.f2
Figure 2 Expression of p53 in mucoepidermoid carcinoma.

TUNEL assays

TUNEL assay was carried out essentially as described previously.18 After removal of paraffin with xylene and rehydration in ethanol solutions of decreasing concentrations, sections were digested for 10 min with proteinase K (20 mg/ml), washed in distilled water, and exposed briefly to 3% hydrogen peroxide to inactivate endogenous peroxidase. The TUNEL reaction was performed using the peroxidase‐based Apoptag kit (Oncor, Gaithersburg, MD, USA). TUNEL positive cells were detected with diaminobenzidine and H2O2 according to the supplier's instructions. Finally, stained sections were lightly counterstained with haematoxylin. Cells were defined as apoptotic if they were TUNEL positive ((figsfigs 3 and 44)

figure bj118661.f3
Figure 3 High expression of TUNEL assay in a case adenoid cystic carcinoma.
figure bj118661.f4
Figure 4 Low expression of TUNEL assay in a case adenoid cystic carcinoma.

Scoring and quantification of the immunoreactivity

Two observers (AB and GB) estimated the staining pattern of TUNEL, p53, and Bcl‐2 separately, and scored each specimen for the percentage of positive cells ([less-than-or-eq, slant]6% and >6% of cells TUNEL positive or expressing Bcl‐2 and [less-than-or-eq, slant]10% and >10% of cells expressing p53 or Bcl‐2). The level of concordance, expressed as the percentage of agreement between the observers, was 90% (19 out of 21 specimens). In the remaining specimens, the score was obtained after collegial revision and agreement. Analysis of the data using such necessarily arbitrary cut‐off was highly statistically significant and, therefore, functionally operative.

Statistical analyses

Spearman's rank correlation or Fisher's exact test were used to assess relation between ordinal data (correlation matrix between immunostaining indicators). The immunostaining variables were stratified as follow: TUNEL: < or >6%; Bcl‐2: < or >6%; p53: < or >10%. Predictors for survival or local recurrence were evaluated either by an univariate model or multivariate analysis. The multivariate survival analysis was evaluated according to the Cox proportional hazards model. The cumulative probability of death was calculated for each group according to Kaplan–Meier analysis and compared using log rank test ((figsfigs 5, 6 and 77).). A p value of <0.05 was regarded as statistically significant in two‐tailed tests. SPSS software (version 9.00, SPSS, Chicago, IL, USA) was used for statistical analysis.

figure bj118661.f5
Figure 5 Cumulative probability of death for p53.
figure bj118661.f6
Figure 6 Cumulative probability of death for Bcl‐2.
figure bj118661.f7
Figure 7 Cumulative probability of death for TUNEL assay.



There were 11 females and 10 males; median age was 51 years (range 14–81 years). The mean follow‐up was 71 months—the shortest follow‐up was 4 months (for patient 1, who had a dramatic spread of the disease and developed liver and bone metastasis), up to a maximum of 192 months. Among the group of patients alive the minimum follow‐up was 24 months to a maximum of 192 months.

All patients underwent a lateral orbitotomy with complete removal of the mass and lateral orbital wall, except for one case (patient 1) who had replacement of the bone. Thirteen cases received radiotherapy of 60 Gy.

Information about survival time was available for all patients; 11 of the 21 patients (53%) died during the follow‐up period.

One patient (patient 8), alive during follow‐up, had a local recurrence of the disease 21 months after the first treatment and underwent surgery for a second time. Out the 11 patients who did not survive, 10 cases died from distant metastasis (lymphatic, liver, bone, lung, brain) and one case from a local aggressive relapse of the disease. The death of these patients occurred from 4–72 months after the first treatment with a mean follow‐up of 39.9 months.


We evaluated 21 specimens: 11 adenoid cystic carcinomas, 4 mucoepidermoid carcinomas, 3 squamous cell carcinomas and 3 adenocarcinomas (table 11).).

Immunohistochemistry and TUNEL assays

TUNEL and p53 immunostainings were always nuclear, with a low to absent background, while Bcl‐2 positivity was always cytoplasmatic (fig 11).). The number of positive cells varied in different specimens. Using Spearman's rank correlation matrix test, a significant correlation was found between all the molecular markers (table 22).

Table thumbnail
Table 2 Spearman's rank correlation matrix (and statistical significance) for molecular markers

In particular TUNEL index was positively correlated with p53 staining and inversely correlated with Bcl‐2 staining; moreover p53 and Bcl‐2 levels were inversely correlated. In multivariate analysis, according to the Cox proportional hazards model, neither histological type (TUNEL, p53 and Bcl‐2 staining) or therapy type (surgery vs surgery plus radiotherapy) were predictors of local recurrence (table 33).

Table thumbnail
Table 3 Analysis of histological, molecular and therapeutic predictors for local recurrence

The same statistical analysis was used to investigate the efficacy of these parameters as predictors of survival.

As shown in table 44,, in multivariate analysis only Bcl‐2 staining >6% was significantly correlated with patient outcome. Finally we looked at the relation between TUNEL, p53 and Bcl‐2 staining and overall survival of the patients affected by epithelial malignant tumours of the lacrimal glands. Indeed, a statistically significant positive relation was found for TUNEL and p53, while an inverse correlation was found for Bcl‐2 (table 55).).

Table thumbnail
Table 4 Analysis of histological, molecular and therapeutic predictors of survival
Table thumbnail
Table 5 Survival and immunohistochemical parameters

All the histological types were not statistically correlated with overall survival of the patients (data not shown). To support the validity of these data, survival curves were constructed using product‐limit survival analysis. The cumulative probability of death was calculated for each group according to the Kaplan–Meier product‐limit method and compared using log‐rank test (p<0.001).


Malignant epithelial tumours of the lacrimal gland have high morbidity and mortality. In the series reported here 53% of patients died during the follow‐up period which ranged from 4 months to 192 months. Factors that influence the prognosis of these tumours are still uncertain.

Age of onset seems to be relevant. Tellado reported that young patients with adenoid cystic carcinomas have a better prognosis than adult patients.10 Wide surgical removal versus local resection, results insignificant for the prognosis. Wright has reported that the disease‐free survival after treatment of lacrimal gland carcinoma appears unaltered by cranio‐orbital resection.7 The effectiveness of adjuvant radiotherapy is also a doubtful point.8 Although in this study the histological types were not statistically significant predictors of local recurrence and mortality, Zimmerman3 and Forrest4 have demonstrated the significance of histological features for the prognosis. Recently Khalil11 demonstrated that it is essential to differentiate a basal cell adenocarcinoma of the lacrimal gland from the solid basaloid type of adenoid cystic carcinoma, because the former has a better prognosis.

In the present study, we evaluated the apoptotic index and the immunohistochemical expression of p53 and Bcl‐2 in a group of 21 surgically resected epithelial malignant tumours of the lacrimal gland to clarify critical events in the pathogenesis of these tumours, and to evaluate the impact of these events on patient outcomes. We choose the immunohistochemical method for our investigation. This method avoids the contamination by non‐neoplastic cells that constantly affects both western blot analyses and nucleic acid‐based approaches. In addition, immunohistochemistry on formalin fixed, paraffin embedded sections is a relatively easy method, which is suitable for routine evaluations.

We found that the expression of p53 was directly correlated with the apoptotic index, while expression of Bcl‐2 was inversely correlated. This expression pattern has been already described in other tumours.19,20,21 Although several studies have confirmed genetic point mutations as a frequent cause of p53 accumulation,22 environmental conditions may also influence protein stability or gene transcription and therefore alter the cellular content of p53 protein.23,24 Presumably, in our case, the correlation between apoptotic index, p53 and Bcl‐2 indicates that tumour cells have lost or down‐regulated their survival gene Bcl‐2, and up‐regulated the expression of p53, presumably to attempt damage repair. If this is unsuccessful, apoptosis would occur. The fact that in multivariate analysis only Bcl‐2 staining >6% was significantly correlated with patient outcome suggests that Bcl‐2 expression is an important molecular event correlated with the progression of these tumours. The correlation we found with survival of apoptotic index, p53 and Bcl‐2 expression strengthens this functional hypothesis. The more tumour cells go into apoptosis, up‐regulating p53 and down‐regulating Bcl‐2, the better the survival of the patients.

Malignant lacrimal gland tumours are histologically related to those lesions that arise in the intra‐oral minor salivary glands.9 Similar abnormalities have been found in these tumours, suggesting possible common molecular mechanisms involved in their neoplastic proliferation.25 Indeed, studies on apoptosis and apoptosis‐related factors in salivary gland tumours have shown similar correlation as that found by us in lacrimal gland tumours.26,27

This study establishes a role of apoptosis‐regulatory proteins in the pathogenesis of malignant epithelial lacrimal gland tumours and suggest that the expression of apoptosis and correlated proteins may provide a tool for evaluating the prognosis of these tumours. Further studies are needed to better define the involvement of these proteins and prove the clinical utility of targeting this pathway.


Competing interests: None declared.


1. Rootman J A. Disease of the orbit. A multidisciplinary approach. Philadelphia: Lippincott, 1988
2. Shields C L, Shields J A, Eagle R C. et al Clinicopathologic review of 142 cases of lacrimal gland lesions. Ophthalmology 1989. 96431–435.435 [PubMed]
3. Zimmerman L E, Sanders T E, Ackerman L V. Epithelial tumors of the lacrimal gland; prognostic and therapeutic significance of histological types. Int Ophthalmol Clin 1962. 2337–367.367
4. Forrest A W. Pathological criteria for effective management of epithelial lacrimal gland tumors. Am J Ophthalmol 1971. 71178–192.192 [PubMed]
5. Wright J E, Stewart W B, Krohel G B. Clinical presentation and management of lacrimal gland tumors. Br J Ophthalmol 1979. 63600–606.606 [PMC free article] [PubMed]
6. Wright J E. Factors affecting the survival of patients with lacrimal gland tumors. Can J Ophthalmol 1982. 173–9.9 [PubMed]
7. Wright J E, Rose G E, Garner A. Primary malignant neoplasms of the lacrimal gland. Br J Ophthalmol 1992. 76401–407.407 [PMC free article] [PubMed]
8. Font R L, Smith S L, Bryan R G. Malignant epithelial tumors of the lacrimal gland, a clinicopathological study of 21 cases. Arch Ophthalmol 1998. 116613–616.616 [PubMed]
9. Paulino A F, Huvos A G. Epithelial tumors of the lacrimal glands, a clinicopathological study. Ann Diagn Pathol 1999. 3199–204.204 [PubMed]
10. Tellado M V, McLean I W, Specht C S. et al Adenoid cystic carcinomas of the lacrimal gland in childhood and adolescence. Ophthalmology 1997. 1041622–1625.1625 [PubMed]
11. Khalil M, Arthurs B. Basal cell carcinoma of the lacrimal gland. Ophthalmology 2000. 107164–168.168 [PubMed]
12. Salomon R N, Diaz‐Cano S. Introduction to apoptosis. Diagn Mol Pathol 1995. 4235–238.238 [PubMed]
13. Canman C E, Kastan M B. Induction of apoptosis by tumor suppressor genes and oncogenes. Semin Cancer Biol 1995. 617–25.25 [PubMed]
14. Schmitt C A, Lowe S. Apoptosis and therapy. J Pathol 1999. 187136–149.149
15. Ellis R E, Yuan J, Horvitz H R. Mechanisms and functions of cell death. Ann Rev Cell Biol 1991. 73111–3120.3120
16. Katsikis P D, Leben W R. Apoptosis. Nature 1995. 374670
17. Gavrieli Y, Sherman Y, Ben‐Sasson S A. Identification of programmed cell death in situ specific labeling of nuclear DNA fragmentation. J Cell Biol 1992. 119493–501.501 [PMC free article] [PubMed]
18. Baldi A, Calia E, Ciampini A. et al Deafferentation‐induced apoptosis of neurons in thalamic somatosensory nuclei of the newborn rat: critical period and rescue from cell death by peripherally applied neurotrophins. Eur J Neurosi 2000. 122281–2290.2290
19. Hoehner J C, Gestblom C, Olsen L. et al Spatial association of apoptosis‐related gene expression and cellular death in clinical neuroblastoma. Br J Cancer 1997. 751185–1194.1194 [PMC free article] [PubMed]
20. Basolo F, Pollina L, Fontanini G. et al Apoptosis and proliferation in thyroid carcinoma, correlation with bcl‐2 and p53 protein expression. Br J Cancer 1997. 75537–541.541 [PMC free article] [PubMed]
21. Paradiso A, Simone G, De Lena M. et al Expression of apoptosis‐related markers and clinical outcome in patients with advanced colorectal cancer. Br J Cancer 2001. 84651–658.658 [PMC free article] [PubMed]
22. Greenblatt M S, Bennet W P, Hollstein M. et al Mutations in the p53 tumor‐suppressor gene, clues to cancer etiology and molecular pathogenesis. Cancer Res. 1994;544855–4878.4878
23. Lane D P. A death in the life of p53. Nature 1993. 362786–787.787 [PubMed]
24. Hall P A, Lane D P. p53 in tumor pathology, can we trust immunohistochemistry? J Pathol 1994. 1721–4.4 [PubMed]
25. Hrynchak M, White V, Berean K. et al Cytogenetic findings in seven lacrimal gland neoplasms. Cancer Genet Cytogenet 1994. 75133–138.138 [PubMed]
26. Soini Y, Tormanern U, Paakko P. Apoptosis is inversely related to bcl‐2 but not to bax expression in salivary gland tumors. Histopathology 1998. 3228–34.34 [PubMed]
27. Yin H F, Okada N, Takagi M. Apoptosis and apoptotic‐related factors in mucoepidermoid carcinoma of the oral minor salivary glands. Pathol Int 2000. 50603–609.609 [PubMed]

Articles from The British Journal of Ophthalmology are provided here courtesy of BMJ Publishing Group