Journal of Oral and Maxillofacial Pathology

: 2004  |  Volume : 8  |  Issue : 2  |  Page : 73--81

An assessment of oral epithelial dysplasia using criteria of `Smith & Pindborg Grading System' & `Ljubljana Grading System' in oral precancerous lesions

Mahendra C Mahajan, VK Hazarey 
 Department of Oral and Maxillo Facial Pathology, Government Dental College & Hospital, Nagpur, India

Correspondence Address:
Mahendra C Mahajan
Rameshwar Nagar, Ambad-Link Road, Nashik - 10, Maharashtra


Histopathologically assessed severity of Oral Epithelial Dysplasia (OED) are considered as «SQ»gold standards«SQ» for the prediction of malignant transformation of precancerous lesions, since the role of immunohistochemical and molecular biological parameters is still uncertain. Grades such as mild, moderate and severe dysplasia obtained with conventional grading system do not provide adequate information regarding therapeutic management of OED. Also assessing risk and deciding management of «SQ»moderate dysplasia«SQ» an intermediate category remains problematic. H & E stained slides of precancerous lesions (30 cases) were comparatively assessed using «SQ»Smith & Pindborg«SQ» grading system and «SQ»Ljubljana«SQ» grading system for «DQ»Epithelial Hyperplastic Laryngeal Lesions«SQ» (EHLL). Our study showed that, Ljubljana grading system provides more objectivity in evaluation of OED. It attempts to utilize a two-grade system, i.e. A «SQ»Low«SQ» and «SQ»High«SQ» grade thus having merit since it eliminates intermediate category. Also, it is a therapeutically informative system of grading, which provides immense information to the clinician and surgeons regarding accurate diagnosis and management of the same.

How to cite this article:
Mahajan MC, Hazarey V K. An assessment of oral epithelial dysplasia using criteria of `Smith & Pindborg Grading System' & `Ljubljana Grading System' in oral precancerous lesions.J Oral Maxillofac Pathol 2004;8:73-81

How to cite this URL:
Mahajan MC, Hazarey V K. An assessment of oral epithelial dysplasia using criteria of `Smith & Pindborg Grading System' & `Ljubljana Grading System' in oral precancerous lesions. J Oral Maxillofac Pathol [serial online] 2004 [cited 2021 Jun 16 ];8:73-81
Available from:

Full Text


Histological assessment of premalignant lesions is important at three sites in human body i.e., Cervix, Oral cavity and Vocal cord of larynx. At all the above three sites the mucosa is lined by stratified squamous epithelium. There is a generally held view that assessment of epithelial dysplasia in premalignant lesions is important because it is felt that the chances of malignant transformation increases with increasing severity of dysplasia.

 I. Precancer of Uterine Cervix

The most earliest effort to assess dysplasia was made at this site. Invasive squamous cell carcinoma of cervix are considered to be preceeded by a series of random cellular changes which are termed as dysplasia and `carcinoma in situ' (CIS), but now more popularly known as CIN (Cervical Intraepithelial Neoplasias). Reichart in 1973, subdivided CIN into 3 grades - CIN I, CIN II and CIN III equivalent to mild, moderate and severe dysplasias respectively, and followed by CIS. Later on in 1988 a Bethesda classification was proposed by National Cancer Institute Workshop which suggested the term SIL (Squamous Intraepithelial Lesions) with subdivisions into Low and High grade. Low grade SIL corresponds to CIN I, whereas High grade to GIN II & III. Rcichart in 1990 also suggested the term `Borderline CIN' or 'CIN with HPV- related changes' inorder to distinguish CIN I from HPV - related flat condylomas and to avoid overdiagnosis of CIN [1] .

 II. Laryngeal Precancer

Preneoplastic lesions of the larynx have been reported in the literature since Durant 1880 described `white scars' adjacent to histologically proven cancer on a vocal cord. Fisher 1956 classified CIS of larynx in 3 patterns as basal type, squamous type and bowenoid types. Kleinsasser 1963 graded laryngeal precancerous lesions as Class I - Hyperplasia without atypia, Class II - Hyperplasia with atypia and Class III - Carcinoma in situ. Kambic 1978 from `Ljubljana University' proposed a classification of precancerosis into 4 groups, hyperplasia simplex, hyperplasia abnormalis, hyperplasia atypica and Carcinoma (a) Carcinoma-in-situ - Preinvasive cancer and (b) Carcinoma invasivum - invasive cancer. The modification of this classification is the `Ljubljana classification' used in the present study [4] .

True vocal cord (pair of inferior vocal cord) [Figure 1] is the only site in larynx which is lined by a non-keratinized stratified squamous epithelium [Figure 2], the rest is a typical respiratory epithelium i.e. a pseudostratified ciliated columnar epithelium (Stiblar-Martincic D 1997) (1 b). The normal histology of vocal cord epithelium bears marked resemblance to non-keratinised oral epithelium.Laryngeal keratosis' or 'laryngeal leukoplakia' is the precancerous lesion which most frequently occurs on 'vocal cord'. Bouquot & Gnepp 1991 [4] found that the clinicopathologic and epidemiologic characteristics of `Laryngeal keratosis' and `Oral leukoplakia' were similar. Also the histological features of these two entities were that of `keratosis with or without atypia' and therefore proposed that there is a striking similarity between the two lesions. Therefore it was conceived to use `Ljubljana grading system', an laryngeal dysplasia grading system for evaluation of OED.

Ljubljana grading system, proposed by Working Group on E.H.L.L of European Society of Pathology (Heliquist et al 1999) [8]

E.H.L.L. is the term used to describe epithelial hyperplastic and dysplastic lesions of laryngeal mucosa. Ljubljana grading system recognizes 4 categories, Simple hyperplasia' [Figure 3], `Abnormal hyperplasia' [Figure 4], `Atypical or Risky hyperplasia' [Figure 5] and `Carcinoma in situ' [Figure 6]. The criterias laid down to categorize are as given in [Table 1]. This classification was accepted as rational and suitable for everyday clincal and diagnostic work at the international conference on `E.H.LL' in `Ljubljana' in 1996.

The first two groups i.e. Simple and Abnormal hyperplasias are considered 'benign' with no risk of progression to carcinoma. Whereas `Atypical hyperplasia' (Low grade), and `Carcinoma in situ' (High grade) as precancerous with `some' risk and `high' risk respectively of development of invasive carcinoma. `Carcinoma in situ' is also considered as actually malignant but without invasion. Thus `Ljubljana' classification of E.H.L.L. utilizes a two grade system as shown in [Table 2], thus having merit since it eliminates intermediate category.

 III. Oral Precancer

Clinical management options of oral premalignant lesions depends upon clinical and bistopathological assessment of malignant potential of a lesion. Various grading systems for OED have been proposed till date. The base for these is the progression of the histologic features of these lesions to cancer. Considering various criteria thought to be typical for the transformation of `dysplastic lesion' to `carcinoma' they are most frequently graded into 3 different groups, i.e. Mild, Moderate and Severe dysplasias. Some authors also include `Carcinoma in situ' at an extreme end of OED considering it as 'pre-invasive malignancy' (Bukhardt & Maerker 1981 [6] , Lumerman et al 1995 [12] , Bouquot 1996 [5] ). However, these grades do not offer clear therapeutic guidelines to the clinicians and surgeons for the appropriate management of a particular premalignant lesion. Assessing risk and deciding management of `moderate dysplasia', an intermediate category remains problematic.

One of the earliest attempt to classify OED objectively was proposed by Smith & Pindborg grading system (1969) [15] . It involves the evaluation of 13 dysplastic characteristics or features. The scores for each characteristic depends on whether it is scored in `None', `Slight' or `Marked' category after comparing with photographic standards. These scores are eventually added together, to achieve the "Epithelial Atypia Index" (E.A.T.). The maximum possible index value is 75 and depending upon the score, a "grade" is assigned as suggested by Katz et al 1985 [11] as shown in [Table 3]. Depending upon E.A.I. if it is 0 - 10; 11 - 25;26 - 45,46 - 75 the grade is No dysplasia, Mild dysplasia, Moderate dysplasia & Severe dysplasia respectively. This grading system is one of the first systematic approach for assessing O.E.D. and is valuable for standardization (Katz et al 1985) [11] .

In the present study, the terms `Potentially malignant lesion'. 'Precancerous lesion', 'Precancerous conditions' and `Oral Epithelial Dysplasia' (OED) are defined as follows:

Potentially Malignant Lesions

The two major oral precancerous lesions, namely leukoplakia and erythroplakia are termed as `potentially malignant lesions' by Johnson NW 1991 [9] . This is because a proportion of them will eventually become overtly malignant; this subgroup might then reasonably be termed as 'premalignant lesion', whereas the remainder (probably the majority - certainly so in cases of leukoplakia) will not become malignant within the lifespan of patient.

Precancerous Lesions of oral cavity are defined as a morphologically altered tissue in which cancer is more likely to occur than its apparently normal counterpart. The examples are leukoplakia, erythroplakia and palatal changes associated with reverse chutta smoking [2] .

Precancerous condition is defined as "generalised state of oral mucosa associated with significantly increased risk for cancer". Examples are oral submucous fibrosis, sidcropenic dysphagia, syphilis, and possibly oral lichen planes [2] .

Oral Epithelial Dysplasia (OED) as defined by Lumermand et al (1995) [12] is a diagnostic term used to describe the histopathological changes seen in chronic progressive and premalignant disorder of oral mucosa. Clinically it may present as leukoplakia, erythroplakia and leukoerythroplakia.

The present study was undertaken in the department of `Oral Pathology and Microbiology'. Government Dental College & Hospital, Nagpur with following aims & objectives:

1) To comparatively evaluate histopathological grading of OED in precancerous lesions at the light microscopic level using Smith & Pindborg grading system for OED and Ljubljana grading system for E.H.L,L.

2) Comparison of grade of dysplasia obtained by Smith & Pindborg grading system with clinical parameters such as site of lesion and clinical appearance of the lesion.

3) To see whether the Ljubljana grading system offers any help in providing information regarding therapeutic management of OED.

 Materials And Methods

Study group

Precancerous Lesions: 30 cases

(Leukoplakia [Figure 7]/

Erythroplakia [Figure 8] / Leucocrythroplakia)

These were the randomly selected cases of precancerous lesions reported between September 1996 to March 2001, biopsied in the Department of Oral & Maxillofacial Pathology, Govt. Dental College & Hospital. Nagpur.


1. Histopathological slides containing 6 micron thick H & E stained sections of biopsies of patients in the `study group' were retrieved from the registry of department. The oral epithelial dysplastic features were evaluated at light microscopic level using both grading system, with the help of a `Binocular compound light microscope' (MICRON UNITEK)

2. Histopathological sections (20 biopsies) showing dysplastic laryngeal epithelium were retrieved from Government Medical College & Hospital, Nagpur. These slides were graded according to Ljubljana classification & used for the purpose of standardisation.

3. An autopsy specimen containing "larynx" was obtained from an unclaimed body in the mortuary of Govt. Medical College & Hospital, Nagpur. Larynx was surgically dissected. "True vocal cord" of the right side was excised, fixed in 10% formalin & processed for routine H&E staining. This was done in order to study histology of normal vocal cord epithelium, so that laryngeal dysplastic lesions and its grading system (i.e. Ljubljana classification) are better understood & applied to oral dysplastic lesions.

Statistical tests can only be applied to observations involving numerical values. Ljubljana grading system do not involve determination of any numerical value, so statistical tests could only be applied to observations related to Smith & Pindborg grading system, as discussed subsequently. Chi-square test [13] is used to find out the significance of difference between two or more than two proportions. It was used at various steps in the present study. Value of p Mild dysplasia

Within this category most of the cases (approx. 84.6%) were graded as abnormal hyperplasia and remaining were gradal as simple hyperplasia. Basically all the cases graded as mild dysplasia with Smith & Pindborg grading were categorized by Ljubljana grading system as benign and Showed benign augmentation of either basal and parabasal cell layers or prickle cells and so close follow ­up is not mandatory lilt this group as suggested by working groupof Ljubljaua classification.

Moderate dysplasia

The therapeutic approach to Moderate dysplasia is always a query since it forms an intermediate group. Within this category most of the cases i.e. 87.5%, were graded as atypical hyperplasia by Ljubljana grading system. This finding warrants that this intermediate category of moderate dysplasia Constitute, mostly of atypical hyperplasia therefore were risky / potentially malignant, requiring a dose follow-up which is mandatory to recognize any progression to severe atypia. Thereby such cases can be separated front a relatively smaller group of abnormal hyperplasia which are benign or nor risky for which follow-up is nut mandatory.

Severe dysplasia

Within this category most of these cases i.e 89% were graded as carcinoma in situ (C IS) by Ljubljana system, and are considered precanceous with high risk of malignant transformation and should be treated aggressively, either with surgery or radiotherapy. Thereby these cases are separated from atypical hyperpasia which requires a close follow-up which is manidatory to recognize any progression to severe atypia .

Of the 13 dysplastic characteristics used in Smith & Pindborg grading system, those frequently observed in descending order were (a) drop-shaped role ridges, (b) increased N/C ratio. (c) irregular epithelial stratification, (d) hyperchromatic nuclei and (e) loss of intercellular adherence. These findings were similar to those observed by Kalz et al 1985 [11] and Wright A and Shear M 1985 [18] . These features can be considered as early features of premalignancy. Whereas feature, such as bizarre mitoses and mitotic activity in upper cell layers occurred infrequently.

The distribution of 13 dysplastic features between the mild- moderate and severe dysplasia were stastically significant (Chi-Square test, p [ 11] . In descending order they were as follows (a ) Level of mitotic activity -93% (b ) Anisocytosis and anisonuclcosis - 87%, (c) Pleomorphrc cell In, nuclei -87% (d) Increased mitosis (in upper ½) - 77% (c) Bizarre mitoses - 67% and (f) Hyperchromaticnuclei-60%.

Clinical Parameters

Clinical details of study group regarding age and sex of patient, site and clinical appearance of lesion are given in [Table 5]. All the cases had history of tobacco usage in some or other form, the duration and frequency varied in these cases. Three cases recurred as white lesions after treatment of primary cancer (Case no. 17, 24 & 30). Three cases (Case no. 3. 19 & 29) had associated oral submueous fibrosis.

There were 8 cases of erythroplakia. 518 cases of erythroplakia exhibited severe dysplasia whereas 3 cases exhibited mild dysplasia, when graded with Smith & Pindborg grading. 6:7 cases of CIS (Ljubljana) were either mixed red & white or red lesions. This finding suggests that clinical appearance of the lesion. i.e. The appearance of red component is associated with higher degree ofdysplasia.

The site distribution of cases were as follows: fluccal mucosa 17 cases, Gingivae and alveolar sulcus 9 cases, Tongue 2 cases, Palate I case, Retonnolar ridge 1 case. The pattern of site distribution reflects the tobacco habits, i.e. Quid placement followed by Bidi smoking are the habits prevalent in this part of the world. The site distribution didn't have any relation to grades of dysplasia (p>0.05. Chi. Sq. test not significant).

There was no significant difference between the distribution of mild, moderate and severe dysplasia in either different age groups (p>0.05, Chi. Sq, test not significant) or in males and females (p>0.05. Chi. Sq. test not significant).


Histopathologically assessed severity of Oral Epithelial Dysplasia (OED) is considered as gold standards for the predication of malignant transformation of oral precancerous lesions. In the present study 30 cases of oral precancerous lesions are comparatively assessed using Smith & Pindborg grading system (1969) and Ljubljana grading system for E.H.L.L. (1999). E.H.L.L means Epithelial Hyperplastic Laryngeal Lesions and is the term used to describe epithelial hyperplastic and dysplastic lesions of laryngeal mucosa. Ljubljana grading system divides the lesions into Non risky group (Simple &. Abnormal hyperplasia) and a Risky group (atypical hyperplasia [low grade] and CIS [high grade]).

An overlap in grades of the two grading system was observed. Mild dysplasia was largest group (approx 43%) which inturn constituted of non risky group in Ljubljana grading system i.e. 85% of abnormal hyperplasia and remaining simple hyperplasia moderated 27% of the cases, of which majority i.e. 87.5%. were atypical hyperplasia (low grade risky). 30% of totaI eases were graded as severe dyplasia with Smith & Pindborg grading system, ofwhich 89% were graded as CIS(high grade)with Ljubljana grading system.

The overlap in grads observed in the present study are compared with similar studies undertaken by Michacls 1997 [14] and Kambie & Gale 1995 [10] . who compared Ljubljana classification with WHO (dysplasia grade method) for assessment of laryngeal dvsplastic lesions. No such similar study exists for comparison of OED

1. In the present study. majority of cases in Mild dysplasia were graded as abnormal hyperplasia i.e. 84.6%. and all the cases in mild dysplasia were non-risky or benign. Similarly in the study by Michaels, abnormal hyperplasia constituted 67% of grade I (mild dysplasia) and 92 % were reported to be non-risky.

2. In the present study, in Moderate dysplasia most of the cases i.e. around 87.5 % we graded as typical hyperplasia i.e. potentially malignant or risky group. In contradiction to our results in a study by Michael, Grade II (moderate dysplasia) was split between abnormal hyperplasia (53%) which constituted a relatively larger group than atypical hyperplasia(32%).

3. In the present study, 89%, of Severe dysplasia cases were graded in Ljubljana grading system as carcinoma in situ. These findings were similar to those reported by Kambic and Gale 1995 [10] . In contradiction must of the grade III (WH0) wets assigned to atypical hyperplasia group (61%) by Michaels.

In a study by Michaels, few cases of moderate dysplasia (grade III were reclassified as carcinoma in situ (CIS) (1/19 caases i.e. 5.3%) and also some cases of grade III were reclassified as non-risky group ie. Benign (5/14 cases i.e. 36%). Such a high degree of overlap was not observed in our study.

Helliwell in 1999 [7] had critical analysed Ljubljana classification and expressed his concern regarding following points:

I. Reproducibility

The use of this classification requires re-education of the histopathologist. No studies have been found till date to determine the intra-observer and inter-observer variation for Ljubljana classification. Present study is the first study which comparatively assess OED with Smith & Pindborg and Ljubljana classification,

2. Terminology

The term hyperplasia is probably appropriate for simple and abnormal though proof of regression is difficult to acquire. There may be reservations about using the term hyperplasia to describe atypical lesions which have greater propensity to progress and sonic of w hich may be in the evolutionary line of neoplasia. We suggest the term Risky dysplasia as an alteniative to Atypeial." Risky hyperplasia.

3. Another important disadvantage we noted in the present study that when Ljubljana classfication is used for grading OED it cannot categorize certain oral lesions which exhibit epithelium that is atrophic and without significant atypia, since they neither exhibit hyperplasia of prickle or basal and parabasal cell layers nor is there any significant atypia to categorize as atypical hyperplasia. This is particularly true while grading atrophic epithelia of oral submucous fibrosis and oral lichen planus.

Despite these above mentioned relative draw backs of the Ljubljana classification, there are certain important advantages of this grading system as given below Helliwell 1999 [7] due to which its use for routine diagnostic work is proposed. It provides helpful guidelines for the clinicians to successfully treat and rnanage dysplastic lesions, since it focusses on important clinical decision points which is the primary concern for the clinician, which include, identification of:.

1. Patients with purely hyperplastic lesions i.e. benign (simple and abnormal hyperplasia) that do not require close follow-up.

2. Patient with mild degree of atypia low grade (i.e atypical/ risky hyperplasia) that require close follow-tip to recognize any progression to severe atypia

3. Patients with severe atypia high grade (carcinoma in situ) are required to be treated aggressively either with surgical excision or radiation therapy.

Thus the result of the present study shows that the Ljubljana classification is advantageous over the Smith & Pindborg classification in two major aspects:

1) An intermmediate category which is therapeutically ill defined i.e. moderate dysplasia which is an overlap between abnormal hvpcrplasia (non-risky group) and typical hyperplasia (risky group). Thereby, it identifies those putenitally malignant lesions in the intermediate category (i.e. atypical hvpcrplasia) which requires close follow-up to recognize; my progression towards malignancy. Thus this two grade system gives clear guidelines to the clinicians and pathologist.

2) It provides distinction in the severe dysplasia category as atypical hyperplasia, in which progression may occur, thus requiring close follow-up and carcinoma in situ which is considered actually malignant without invasion i.e. high risk. Thus it identities carcinoma in situ which needs to be treated relatively aggressively with surgery or radiotherapy.


In view of the above observations it can be concluded that Ljubljana grading system provides immense information to the clinicians. since it focusses on important clinical decision points. It demarcates Risky lesions front Non risky lesions which do not require either fullow up or any treatment. It further categorizes risky group into two grade system Low grade which comprised majority of moderate dysplasia group in our study and requires close follow up to recognize any changes in the clinical behaviour or progression in severity of dysplasia. Whereas High grade (CIS) requires aggressive treatment, either surgical excision or radiotherapy. These guidelines therefore improves the accuracy of diagnosis and as a result, the efficacy of treatment of oral epithelial dysplastic lesions. Therefore its use in day to day practice in assessing risk of malignant transformation in OED would be an useful adjunct to pathologists and clinicians.

It is also suggested that numerical system should be incorporated into Ljubljana system. so that statistical test for significance can be applied to it. Further studies with inter-observer and inra-observer reliability. For Ljubljana grading system using Kappa statistics and multi variate analysis are required to substantiate the above observations anti to monitor due reproducibility of this new grading system.


1) Prof D.K. Datary- Tata Institute of Fundamental Research, Mumbai

2) Dr. (Mrs) S.M. Ganvir, Professor. Department of Oral Pathology, Government Dental College and Hospital, Nagpur, for her guidance and help during the study.


1Ackerman's. Surgical pathology, Vol. 1, 8 th Ed.. Mosby publications.
2Axell T. Holmstrup P. Kramer IRH et al. International seminar on oral leukoplakia and associated lesions related to tobacco habits. Lund univ.. Malmo, Sweden. June 27 - 30; 1983. Journal of Communtiy Dental Oral Epidemiology 1984: 12(3): 145-154.
3Bouquot JE Weiland LH and Kurland LT Leukoplakia and carcinoma in situ synchronously, associated with invasive oral oropharyngeal carcinoma in Rochester, Minn 1935 - 1984. Oral Surg. Oral Med. Oral Pathol. 1988:65: 199-207.
4Bouquot JE and Gnepp DR. Laryngeal precancer: A review of lilerature. commentary and comparison with oral laukoplakia. Head and Neck1991:13:488-497.
5Bouguot JE The pathology and progression of oral. premalina ncy: 1946.
6Burkhardt A and Maerker R. A colour atlas of oral cancer. Chicago: Year Book: 1981.
7Helliwel TR. Commentary - Risky epithelium in the lawns - a practical diagnosis? Histopathology 1999; 34: 262-265.
8Hellquist H. Cardesa A. Gale N, Kambic V Michaels L (Working Group): Criteria for grading in the Ljabljana classification of epithelial hyperplastic laryngeal lesions. A study by members of the Working Group on epithelial hyperplaslic laryngeal lesions of the European Society of Pathology. Histopathology 1990: 14: 226-233_
9Johnson NW. Risk markers for oral diseases: Vol 2: Oral cancer, Chapter 8: Structural indicators of high risk lesions, Editorial, 165.
10Kambic V and Gale N. Epithelial hyperplaslic lesion of the larynx. Amsterdam: Elscvicr. 1995.
11Katz HC, Shear M and Altini M. A critical evaluation of epithelial dysplasia in oral mucosal lesions using the Smith & Pindborg method of standardization. Journal of Oral Pathology 1985:14:476-482.
12Lumerman H Freedman P, Kerpel S. Flushing. Oral epithelial dysphasia and the development of invasive squamous cell carcinoma. Oral Sung. Oral Med. Oral Pathul. 1995: 79: 321 329.
13Mahajan BK. Methods in Biostastics. IV Ed. 1984. Smt. lndu Mahajan Publications.
14Michaels L The Kambic-Gale method of assessment of epithelial hyperplaslic laryngeal lesions of the larynx in comparison with dysplasia grade method. Acta. Otolayngol. 1997(Stockh):Suppl.527:17-20.
15Smith CJ and Pindborg JJ. Histological grading of oral epithelial atypia - by using photographic standards. Copenhagen: WHO reference centre for oral precancerous conditions (1969)
16Stiblar - Martincic D. Histology of laryngeal mucosa. Acta. Otolaryngol. (Stockh) 1997: Supp1527:138-141.
17Warnakulsuriya S.Editorial-Lack of molecular markers to predict malignant potential of oral precancer. J.Pathol. 2000: 190:407-109.
18Wright A and Sheer M Epithelial dysplasia immediately adjacent to oral squamous cell carcinoma. J of Oral Pathology & Medicine 1985; 14: 559-64. Malignant potential.