Genetic Polymorphism of Cypiai, Gstm1 and Gstt1 Among Various Tobacco Habit Groups with Susceptibility to Oral Pre-Cancerous Lesions and Squamous Cell Carcinoma

GENETIC POLYMORPHISM OF CYPIAI, GSTM1 AND GSTT1 AMONG VARIOUS TOBACCO HABIT GROUPS WITH SUSCEPTIBILITY TO ORAL PRE-CANCEROUS LESIONS AND SQUAMOUS CELL CARCINOMA

MUHAMMAD MOHIUDDIN ALAMGIR

MBBS, M.Phil (Path)

Department of Pathology, Faculty of Medicine/ Ziauddin University

A thesis submitted in partial fulfillment of the requirements for the

degree of PhD in Pathology

Karachi/ Pakistan

December, 2016

Dedicated to

Late Prof. Dr. Naeem Aon Jafarey

in the memory of his enthusiastic guidance and kind support

vii

Acknowledgement

First I offer my thanks and gratitude to Allah Almighty, for providing the opportunity and means to do this work. All respect for his Last Prophet (PBUH).

I am grateful to my supervisor Prof. Dr. Qamar Jamal for her able guidance and continued encouragement throughout my research. She always remained positive regarding accomplishment of this endeavor. Her efforts in critical reading of the thesis are commendable. My special thanks are to my co-supervisor Late Prof. Naeem Aon Jafarey. He was my main inspiration for initiation of this research project. His guidance and interest in this work remained throughout right from the beginning of my synopsis writing until his demise. He helped me a lot in retrieval of literature and critical evaluation of my results.

I am grateful to my co-supervisor Prof. Dr. Talat Mirza, for the valuable feedback and limitless encouragement. She helped me not only in sample collection, but guided me in data analysis, paper writing and finally thesis writing. Without her support and interest it may not be possible to complete this work.

My special thanks to Prof. Dr. Kamran Hameed, Prof. Dr. Masood Hameed Khan and Prof. Dr. Khalid Mahmood for their support in ethical approvals from respective Institutes.

My sincere thanks are for Dr. Israr Nasir for his guidance and technical support. I carried out the tedious task of molecular analysis of my samples under his guidance. Without his technical expertise and trouble-shooting my work could not have been completed. I am also thankful to his supporting lab staff specially Miss Nazneen, Miss Maheen, Mr.Haris Lucky, Mr.Tariq, Mr.Nazuk and Mr.Waqar. I express my sincere thanks to Mr. Manzoor Asi for his help in retrieval of histopathology slides of cancer cases as well as their photomicrography.

At this point, I would like to offer my gratitude to Dr. Mushtaq at DUHS and Prof. Dr. Saeeda Baig at ZU, who introduced me to the fascinating world of molecular biology and genetics.

viii

I am thankful to Professor Dr. Iqbal Khayani at CHK, Mr. Illtifat at Radiotherapy department of ZUH, North Nazimabad, and my students Salman and Shahjee who helped and supported me in collecting patient samples and data. In fact, there is along list of persons who supported me in collection of studied cases and I am thankful to all of them. I am also thankful to Mr. Ejaz (ZU) and Mr. Faisal Raza (Panjwani Centre, KU) for helping me in statistical analysis.

I am thankful to Prof. Dr. Zahida Memon and Prof. Dr. Nikhat Siddiqui for their guidance and help.

I am grateful to my colleagues, subordinates at BUM&DC, and friends for their invaluable cooperation, support and sacrifices.

My special thanks go to my beloved family, my wife Dr. Hina Zaman and my daughter Maryam, who have been pillars of support throughout this project. I am indebted to them for being there when I felt depressed and tired.

Finally, I am grateful to Higher Education Commission of Pakistan and Ziauddin University, Karachi, for providing me financial support to complete this project.

List of Abbreviations

AARs Age-adjusted rates AHUH Aga Khan University Hospital APCR Aga Khan University Pathology-based Cancer Registry ASRs Age-standardized incidence rates CDK Cyclin dependent kinase CDKI Cyclin dependent kinase inhibitor CHK Civil Hospital Karachi CISH Chromogenic In situ Hybridization CYP1A1 Cytochrome P4501A1 CYPs Cytochrome P450s DNA Deoxy-ribonucleic acid DPX Distrenedibutylphthalatexylene EBV Epstein Bar Virus EDTA Ethylenediaminotetra acetic acid EGFR Epidermal growth factor receptor FISH FluorecentIn situ Hybridization

GAP GTPase-activating proteins

GDP Guanine Diphosphate

GEFs Guanine-nucleotide exchange factors

GSTM1 Glutathione S-tranferase M1

GSTs Glutathione S-transferases

GSTT1 Glutathione S-tranferase T1

GTP Guanine Triphosphate

HNSCC Head and neck squamous cell carcinoma HPV Human Papilloma Virus IARC International Agency on Cancer ICD International Classification of Disease JPMC Jinnah Post-Graduate Medical Centre KCR Karachi Cancer Registry KIRAN Karachi Institute of Radiotherapy and Nuclear Medicine

KPK Khyber Pakhtunkhwa LOH Loss of heterozygosity NAACCR North American Association of Central Cancer Registries

NATs N-acetyltransferases NCCP National Cancer Control Programme NCRP National Cancer Registry Programme NNK Nicotine-derived nitrosamine ketone NNN Nitrosonornicotine OPLs Oral precancerous lesions OR Odds ratio

OSCC Oral squamous cell carcinoma OSF Oral submucous fibrosis PAHs Polycyclic aromatic hydrocarbons PBCRs Population-based Cancer Registries PCLs Pre-Cancerous Lesions

PCR Polymerase Chain Reaction

PCR-RFLP PCR- Restriction Fragment Length Polymorphism

MDSCC Moderately Differentiated Squamous Cell Carcinoma

PMRC Pakistan Medical Research Council

PSCC Papillary squamous cell carcinoma qRT – PCR Quantitative Real Time PCR

ROS Reactive oxygen species

SAARC South Asian Association for Regional Cooperation

SCE Sister chromatid exchange

SKMCH & RC Shaukat Khanum Memorial Cancer Hospital and Research Centre

SLT Smokeless tobacco

SSCP Single Stranded Conformational Polymorphism

ST Sulfotransferases

TNM Tumor Node Metastases

UGTs UDP-glucuronosyltransferases

UP Uttar Pradesh

VC Verrucous carcinoma

XMEs Xenobiotic metabolizing enzymes

ZU Ziauddin University

xii

List of Tables

TABLE NO. DESCRIPTION PAGE

1.1 Top 3 malignancies by gender and age in an 18-year time period (1994-2012) at SKMCH & RC ………………………...... 02 Epidemiological facts about cancer of oral cavity according to 1.2 World regions ...……………………………………………………... 04 1.3 ASRs per 100,000 in women, Karachi South ……………………….. 09

1.4 ASRs per 100,000 in men, Karachi South …………………………... 10

1.5 Basic constituents of a gutka ………………………………………... 29

1.6 Overview of CYP1A1 Polymorphism Nomenclature ……………….. 34

3.1 Characteristics of the study subjects ………………………………… 60

3.2 Ethnic distribution of PCLs, OSCC & Control cases ……………….. 61

3.3 OSCC Cases According to Age, Sex & Intra-oral Sub-site …... 62

3.4 Tobacco indices in OSCC cases and controls ………………………. 69

3.5 CYP1A1MspI, GSTM1 and GSTT1 genotype variants in OSCC cases and controls ……………………………………………...... 70

3.6 Genotype distribution among different tobacco exposure groups for oral cancers and controls ……………………………………………. 72

3.7 Effect of genotype combinations on OSCC cases and control …….... 75

3.8 Risk analysis according to intra-oral sub-site (Cheek)………………. 80

3.9 Risk analysis according to intra-oral sub-site (Tongue)……………... 83

3.10 Mean and median tobacco indices in Controls and PCL cases ……... 86

3.11 CYP1A1, GSTM1 and GSTT1 genotype variants in PCLs and controls ………………………………………………………...... 88 3.12 Genotype distribution among different tobacco exposure groups in PCL cases and controls ……………………………………………… 90 3.13 Effect of genotype combinations on PCL cases and controls ………. 92

3.14 Distribution of genotypes according to ethnicity…...... …………….. 97

xiii

List of Figures

FIG.NO. DESCRIPTION PAGE

1.1 Structures located in the Head and Neck region……………………… 3

1.2 Comparison of AARs for mouth cancer in males across all PBCRs – India …………………………………………………………………... 6 1.3 Comparison of AARs for mouth cancer in females across all PBCRs – India ………………………………………………………………… 7

1.4 Comparison of AARs for Tongue cancer in males across all PBCRs – India …………………………………………………………………... 7

1.5 Natural history of oral carcinogenesis ………………………………... 22

1.6 Tobacco culture of Karachi …………………………………………... 30

1.7 Polymorphism of human CYP1A1 gene ...…...... 34

2.1 Steps in Polymerase Chain Reaction ………………………...... 53

2.2 Exponential amplification in PCR cycling ………………………...... 53

3.1 Carcinoma of buccal mucosa, Stage III, Case ID-105………….. 63

3.2 Carcinoma of left cheek, Stage II, Case ID-122………………... 64

3.3 Carcinoma of the lip, Stage IV, Case ID-48……………………. 64

3.4 Metastatic carcinoma, Stage IV, Case ID-03…………………… 65

3.5 Moderately-differentiated oral squamous cell carcinoma (H&E;× 200 magnification), Case ID-142……………………... 65 3.6 Moderately-differentiated SCC of tongue showing muscle invasion (H&E; × 200 magnification), Case ID-132…………… 66

3.7 Intact stratified squamous epithelium, tumor pushing from below (H&E;× 200 magnification), Case ID-110………………. 66 Cont…...

xiv

1. INTRODUCTION 1.1 Epidemiology In the past, a number of cancer registries have been established in Pakistan but there has been a problem as regards to their sustenance. In the first five decades after its creation the country lacked any substantive cancer data other than a few sporadic institution-based frequencies (Pakistan Medical Research Council, 1982). This data lacks complete demographic details and there was the problem of continuit. In subsequent decades, a number of proper cancer registries were established namely; The Karachi Cancer Registry (KCR) in 1995, The Aga Khan University Cancer Surveillance for Pakistan (ACSP) in 2000 at the Aga Khan University Pathology-based Cancer Registry (APCR), the hospital-based cancer registry at the Shaukat Khanum Memorial Cancer Hospital and Research Centre (SKMCH & RC) Lahore 1994 and the Punjab cancer registry in 2005 (Bhurgri, 2006; Badar, 2015; The Punjab cancer registry, 2016). The Karachi Cancer Registry assessed the magnitude of cancer burden in Karachi for the period between 1995 and 1997 and then between 1998 and 2002. The sample data was collected from Karachi South. Data from The Aga Khan University Cancer Surveillance for Pakistan was compared with that of Karachi Cancer Registry for the analysis of geographical variation. The same has been done for cities like Quetta, Hyderabad, Larkana and Peshawar (Bhurgri, 2006). In Pakistan there are over 1.4 million cancer patients. Annually there are 139,200 new cancer cases comprising of 10.73% among SAARC countries, 101,600 deaths comprising of 11.45% among SARRC countries and 72.99% death as percentage of incidence within the country (Noronha et al., 2012). When literature was reviewed it became apparent that incidences of cancers are different in different parts of Pakistan. SKMCH & RC Registry published its data of 18-year duration from 1994 to 2012 (Table 1.1). During the 18-year time period, a total of 58,761 tumors were registered comprising of 55,974 malignant tumors. Three most common malignancies for the said period among adults were those of breast (23.81%), lip and oral cavity (6.61%) and liver and intra-hepatic bile ducts (5.70%), respectively. Among adult males the commonest cancer was of liver/ intrahepatic bile ducts (8.66%) followed by cancer of lip and oral cavity (8.54%) and Non-Hodgkin‘s lymphoma (7.42%). The commonest cancer among adult females was

1 that of breast (45.46%) followed of ovary/ uterine adnexa (5.91%) and lip/ oral cavity (4.82%) [Badar and Mahmood, 2015]. Table 1.1 Top 3 malignancies by gender and age in an 18-year time period (1994-2012) at SKMCH & RC [Source: Modified and adapted from Badar and Mahmood, 2015]

Top 3 malignancies 1-Count (%) 2-Count (%) 3-Count (%) Total All age-groups, both Breast Cancer Leukemia Lip & Oral 55,974 genders combined 11,853 (21.18%) 3,368 (6.02%) cavity 3.336 (5.96%) Adults ( >18 years) Breast Cancer Lip & Oral Liver/intrahepa 49,765 11,848 (23.81%) cavity tic bile ducts 3,291 (6.61%) 2,836 (5.70%)

Adult males (>18 Liver/intrahepatic Lip & Oral Non- 23,971 years) bile ducts cavity Hodgkin‘s 2,076 (8.66%) 2,047 (8.54%) lymphoma 1,779 (7.42%) Adult females (>18 Breast Cancer Ovary & uterine Lip & Oral 25,794 years) 11,726 (45.46%) adnexa cavity 1,524 (5.91%) 1,244 (4.82%)

Children (<18 years) Acute Hodgkin‘s Non- 6,209 lymphoblastric lymphoma Hodgkin‘s leukemia 1,099 (17.70%) lymphoma 1,345 (21.66%) 694 (11.18%)

Another hospital-based registry at Karachi Institute of Radiotherapy and Nuclear Medicine (KIRAN) covering a large area of Sindh province, reported for a 9-year period (2000-2008) the top three malignancies out of a total of 18,351 cancer cases in males as head and neck cancer (32.6%), gastrointestinal tract cancer (6.9%) and lymphomas (6.1%) while in females breast cancer (38.2%) peaked followed by head and neck cancer (15.1%) and cervical cancer (5.5%) [Badar and Mahmood, 2015]. Retrospective four-year (1989-1992) data of a pathology-based cancer registry at Aga Khan University Hospital (AHUH) Karachi comprising of 2632 malignant cases revealed common cancers in males as lung cancer (15.2%), head and neck cancer (10.6%) and lymphoma (10.2%). In females, breast cancer (32.5%), ovarian cancer (7.9%) and cancer of gall bladder (7.7%) were found to be common malignancies (Malik et al., 1998). KCR reported cancer burden in Karachi South for the period between 1995 to 1997 as ASRs of 139.11/100,000 in males and 169.5/100,000 in females. In males, the mean age

2 for all cancer sites was 51.2 years while that for females was 50.0 years. Common malignancies among males were those of lung, oral cavity, urinay bladder and larynx having ASRs of 21.3, 14.2, 9.0 and 8.8, respectively. In females predominant malignancies were those of breast, oral cavity and ovary with corresponding ASRs of 53.1, 14.5 and 10.9, respectively. ASRs for cancers registered between the period 1998 and 2000 for males and females were 179.0/100000 and 204.1/100000, respectively. In males, the mean age for all cancers was 49.5 years while that for females was 53.7 years. For this later period, common cancers among men were lung, larynx and urinary bladder with ASRs of 25.5, 11.8 and 9.9, respectively. In females, cancer of breast again ranked top followed by esophagus and cervix, ASRs being 69.1, 8.6 and 8.6, respectively (Bhurgri et al., 2006). Head and neck cancer This includes cancers that occur in the oral cavity, pharynx, paranasal sinuses, nasal cavity, larynx, and salivary glands (Figure. 1.9). Forty percent of all head and neck cancers originate in the oral cavity. This cancer is located in any of the eight anatomic sub-sites including the tongue, floor of the mouth, palate, lips and oropharynx.

Figure 1.1 Structures located in the Head and Neck region [Source: Adapted and Modified from web site: http//www.aboutcancer.com] The squamous cell carcinoma of the upper aerodigestive tract is defined by the International Classification of Disease (ICD) as: intra-oral sites [ICD-10 C00-C06], oropharynx [ICD-10 C09-C10], and other ill-defined sites of the lip, oral cavity and pharynx

[ICD-10 C12-C14] (Slootweg et al., 2005), Larynx [ICD-10 C32] and esophagus [ICD- 10 C15] (Richiardi et al., 2012). Oral squamous cell carcinoma (OSCC) Malignant neoplasms of oral cavity, lip, tongue and oro-pharynx are often clustered jointly in epidemiological data. They account for 3-4% of all cancers and majority (around 90%) are squamous cell carcinomas (Choi et al., 2008; Chen et al., 2008). World wide prevalence of oral cancer Oral cancers when combined with oro-pharyngeal cancers are the 6thmost common cancers world-wide (Warnakulasuriya, 2009). According to IARC data for the year 2012, the oral cavity cancer has an annual incidence of 300,373 cases with an ASR of 4.0. Out of these 198,975 cases were from men having an ASR of 5.5 and 101,398 cases were from women with an ASR of 2.5 (IARC, Globocan, 2012). This cancer exhibits a wide variation in incidence across the globe (Table 1.2). Table 1.2 Epidemiological facts about cancer of oral cavity according to world regions [Source: Modified and adapted from GLOBOCAN 2012]

Area New cases Deaths Incidence Mortality (Thousands) (Thousands) ASRs ASRs M F M F M F M F World 199 101.4 97.9 47.4 5.5 2.5 2.7 1.2 Europe 42.6 18.8 17.6 6.0 7.5 2.5 3.0 0.7 South America 10.1 5.8 4.2 1.8 5.3 2.4 2.2 0.7 Caribbean 1.1 0.5 0.5 0.2 4.8 1.8 2.0 0.6 Asia 112 56.9 65.0 32.4 5.2 2.5 3.0 1.4 South East Asia 10.5 7.6 5.0 3.5 9.9 4.7 1.9 1.2 Australia/ New 1.7 0.9 0.3 0.2 8.3 3.7 1.4 0.6 Zealand Melanesia 0.6 0.5 0.3 0.3 22.9 16.0 14.4 10.2

Incidence rates in India and countries across South and South East Asia are amongst the highest in the world (Gupta et al., 2014). Countries with a very high incidence rates for oral cancer include India, Pakistan, Sri Lanka, Taiwan, France, Hungary, Slovakia, Slovania, Brazil, Uruguay, Caribean Islands, Papua New Guinea and Melanesia (Warnakulasuriya, 2009).

(I) Europe Oral and pharyngeal cancer ranks 7th among all cancers in countries of European Union (Boyle et al., 2005). Western Europe shows higher incidence than other regions. France and Hungary have the highest reported incidence rates while Greece and Cyprus the lowest. In northern France among males the incidence is 42.3 per 100,000. In recent decades some Eastern European countries have also shown a rise in oral cancer incidence (Banoczy et al., 2004). Oral cancer is not common in the United Kingdom while very low rates are reported from Greece, Finland and Sweden (Conway et al., 2006). Lately, in Hungary the incidence and mortality have almost doubled posing a serious threat from this cancer (Banoczy et al., 2004). (II) United States of America Approximately 49,200 new cases of Oral Squamous Cell Carcinoma are documented in 2012 from the United States of America. ASRs for males and females are 5.9 and 2.6 per 100,000, respectively. Black males are affected for the most part by cancers of oropharynx (Ferlay et al., 2014). More than 95% of oral cancer cases in the United States are seen in persons over 40 years of age, with a median age at diagnosis of 63 years (Silverman and Gorsky, 1990). (III) South America and Caribbean In this region 17,500 new oral cavity cancer cases were seen in 2012 (Ferlay et al., 2014). Argentina, Southern Brazil and Uruguay are the high incidence countries. Highest rates are reported from Brazil among males, which is the highest risk observed for OSCC in the world after France and India. Puerto Rico has the highest incidence of more than 15 per 100000 (Wunsch-Fiho et al., 2001). Cuba, where heavy cigar smoking is a common tobacco habit, intermediate incidence was observed. In males reported incidence of 7.2 per 100,000 was observed in 1986 and the same remained stable for over a decade (Garrote et al., 2001). (IV) Africa Oral cancer has not been found to be a serious problem in Africa, although the data is limited to few hospital frequencies. However, a high incidence is reported from Sudan, where the use of toombak, a product of oral snuff, in males is common (Idris et al., 1995).

In Harare, Zimbabwe, lip and oral cavity malignancies comprised 2.1% and 1.8% of all malignancies in women and men, respectively (Basset et al., 1995). A study conducted on 1723 cancer cases found mandibular gingival, floor of the mouth and tongue as most common sites of OSCC (Chidzonga et al., 2006). Lowest incidences were reported from Kenya and Congo (Onyango et al., 2004; Kayembe et al., 1995). (V) Asia Asia comprises of countries with the highest incidence rates of OSCC across the world. Indian subcontinent alone accounts for one third of global lip and oral cavity cancers. Here the disease largely affects the poor population (Johnson et al., 2011). Melanesia in the Pacific region has the highest global incidence in both men and women of 22.9 and 16.0, respectively (Table 1.2). India, Pakistan and Sri Lanka were the other top-ranking countries. Very high AARs of 43.8 to 114.9 per 100,000 are reported from India (Consolidated report, NCRP 2001-2004) [Figure 1.2 to 1.4].

Figure 1.2 Comparision of AARs for mouth cancer in males across all PBCRs – India [Source: Adapted and Modified from ―Consolidated report of population based cancer registries 2001-2004‖].

OSCC is one of the common cancers in Bangladesh mostly found in older men (Ahmed et al., 1990). Sri Lanka in the last decade was reported to have the highest incidence of this cancer in South Asia (Ferlay et al., 2004) accounting for around 30% of all cancers in males (Chiba et al., 1999). Oral cancer is not very common in Japan. Data from ten population-based cancer registries for the year 2001 revealed an incidence rate of 5.3 per

100,000. A total of 9612 cancers were recorded in the intra-oral sites (C00-C14) constituting 6984 in men and 2628 in women (Marugame et al., 2007).

Figure 1.3 Comparision of AARs for mouth cancer in females across all PBCRs – India [Source: Adapted and Modified from ―Consolidated report of population based cancer registries 2001-2004‖].

Figure 1.4 Comparision of AARs for Tongue cancer in males across all PBCRs – India [Source: Adapted and Modified from ―Consolidated report of population based cancer registries 2001-2004‖].

Oral cancer in Pakistan OSCC ranks as the eight most frequent malignancy globally but the incidence in developing countries like Pakistan, India and Sri Lanka are very high (Zygogianni et al., 2011). It accounts for more than 15% of all cancers in Pakistan in comparison to 3% internationally. In the 6th and 7th decades of the last century, oral cancer was reported to constitute 10% of all cancers in the country (Zaidi et al., 1966; Zaidi et al., 1974). In comparison to world-wide reported ASRs of 5.5 and 2.5 in males and females, ASRs for Pakistan are 11.0 and 8.6, respectively (Farley et al., 2014). In Pakistan ASRs for oral cancer are 11.0 and 5.2 in males and 8.6 and 4.1 in females, respectively (IARC Globocan, 2012). More recently, the data from Shaukat Khanum Memorial Cancer Hospital ranks oral cancer as the second most frequent cancer in adults constituting 6.69% of all malignancies (Badar and Mahmood, 2015)). The oral cancer incidence in Pakistan has risen considerably with time, as seen with two studies carried out in Karachi South, a sample population of Pakistan. The first study conducted from 1st January 1995 to 31st December 1997 show ASRs per 100,000 for oral cancer to be 14.2 in males and 14.5 in females. The second study done from 1st January 1999 to 31st December 2002 reveals ASRs of oral cancer in males as 22.5 and in females as 20.4 (Bhurgri et al., 2006). Oral cancer incidence in Karachi The very first record of oral cancer in Karachi dates back to mid sixties of the last century. The Department of Radiotherapy, JPMC, reported for the period 1960 to 1964 a total of 932 laryngo-pharyngeal and 828 buccal cavity cancers. Oropharyngeal cancers constituted 38% of all malignant cases (Zaidi et al., 1966). A survey conducted on prevalence of oral cancer in the sample population from three localities namely Korangi, Pir Illahi Bux Colony-Martin Road Quarters and Nazimabad areas of Karachi, has been found to be 0.05% (Jafarey et al., 1972). An appraisal was published from the same hospital (JPMC) in 1976 reporting 1192 oral and oropharyngeal carcinoma cases treated by the centre between 1967 and 1972. The same study documented a prevalence rate of 25/100,000 for carcinoma of oral cavity and 705/100,000 population for pre-malignant lesions (Jafarey et al., 1976). PMRC reported incidences of oral cancer in four cities inclusive of Karachi as 15.74 for males and 15.70 for females in 1974 while 8.6 for males and 10.33 for females in 1979 (Mahmood et al., 1986). For Karachi city, population-

8 based incidence or mortality figures are not available other than the ones for Karachi South documented by KCR. Hospital-based frequency data by Aga Khan University Hospital, Karachi puts head and neck cancers as the second most common cancer among males having a frequency of 10.6%. In females these cancers rank fifth with a frequency of 6.9% (Malik et al., 1998). Of the two subsequent studies from AKU, the first reported 68.1% oral cancers out of 72 reviewed head and neck cancers treated during 2007-2008 at this facility (Kazmi et al., 2012). The second study documented 72.91% oral cancers out of 288 reviewed head and neck cancers treated during 2006 to 2008 (Hameed et al., 2012). In another hospital, JPMC, out of all head and neck malignancies, 43% cancers treated at the radiology department were oral cancers (Kadri et al., 2015). Published data by KCR placed oral squamous cell carcinoma as the second most frequent malignancy in both males and females (Bhurgri et al., 2006). There was a gradual increase in incidence rates especially among men. The ASRs per 100,000 population during the period from 1998 to 2002, were 22.5 in men and 20.4 in women (Tables 1.3 and 1.4). Previously for the period from 1995 to 1997 ASRs per 100000 in males and females were 14.2 and 14.5, respectively (Bhurgri et al., 2006). By the year 2007, OSCC superseded lung cancer among men in Karachi (Bhurgri, 2007).

Table 1.3 ASRs per 100,000 in women, Karachi South. SITE 1995-1997 1998-2002 2003-2007 Breast 52.8 69.1 74.3 Oral cavity 14.9 20.4 25.3 Cervix 6.9 8.6 9.8 Esophagus 7.0 8.6 9.2 Ovary 11.1 7.8 8.5 Lymphoma 4.8 7.2 8.1 Gall bladder 5.4 5.8 6.1 Skin 5.6 5.5 5.8 Colo-rectum 5.3 5.5 5.9 Uterus 6.4 5.0 5.3

Table 1.4 ASRs per 100,000 in men, Karachi South.

SITE 1995-1997 1998-2002 2003-2007 Lung 21.0 25.2 26.3 Oral cavity 15.6 22.5 27.8 Larynx 8.8 10.7 11.9 Prostate 5.3 10.1 11.9 Lymphoma 7.9 9.6 10.1 U. Bladder 9.0 9.3 9.8 Pharynx 8.1 8.1 8.5 Esophagus 6.5 6.7 6.7 Colo-rectum 5.4 6.5 6.9 Stomach 3.9 6.0 6.5

1.2 Studies conducted on oral cancer in Karachi and Pakistan at large: A historical perspective In order to appreciate current trends observed in recent studies, an effort has been made to review the work done on oral cancer from the early years after inception of Pakistan, especially in the Karachi city. Studies on oral cancer were started in Karachi as early as in 1960 by the Department of Radiotherapy, in then Jinnah Hospital Karachi, a federally run cancer treatment facility covering the entire southern provinces of Pakistan. Later on a number of hospitals and research bodies participated in the process. Studies conducted by Jinnah Post Graduate Medical Centre, Karachi The first ever organized study on oral cancer in Karachi was conducted in the Department of Radiotherapy, Jinnah Post-Graduate Medical Centre Karachi for the period between 1960 and 1964. This study included oral cancer as part of the frequency of various neoplasms seen in the institute during the said period. According to this study oropharyngeal cancer formed about 38% of all malignancies seen at the centre and for the first time realized it to be a major cancer problem in Karachi population (Zaidi et al., 1966). The same centre published another valuable study in 1969 on the association of tobacco chewing and smoking habits with oral lesions. The study recognizes habits like chewing

10 of paan containing tobacco, supari, naswar, smoking cigarette and bidi as most prevalent among patients with oral lesions (Jafarey et al., 1969). The next study from this facility was published in 1972 and conducted on prevalence of oral lesions in three localities of Karachi. Localities included were Korangi, Pir Illahi Bux Colony-Martin Road Quarters and Nazimabad areas of Karachi City. The prevalence of oral cancer in the sample population has been found to be 0.05 percent. All persons included were 21 years and above, forming 48% of the study population. Premalignant lesions, mostly leukoplakia, were found in 1.4% of the cases. Persons showing high level of abnormalities were observed in those born in parts of India from where high prevalence was reported, i.e U.P., Gujrat and Bombay (Jafarey et al., 1972). On a subsequent assessment of oral cavity and oropharyngeal cancers in Karachi which was published in 1976 as a continuation of the study started in 1960. The prevalence rate found was 25/100,000 population for carcinoma of oral cavity and 705/100,000 for pre- malignant lesions like leukoplakia, erythroplakia and oral submucous fibrosis (OSF). Age group 50 to 54 showed the highest number of cancer cases in both genders. The commonest site of involvement was tongue in males and buccal mucosa of cheek in females. This series included 683 male and 509 female patients. Bidi smoking increased the risk by 36 times when used alone as compared to controls. This study also gave the all important clinical stage of these cancers. Largest number of oral cancers were presented in clinical stage III (369 cases) followed by stage IV (223 cases). Finally statistical analysis suggested that the prevalence of oral lesions among people of different origin is attributable to the chewing and smoking habits of the people rather than their place of birth (Jafarey et al., 1976). A study aimed to evaluate histological grading and clinical staging was performed during 1967 to 1973 and published in 1977. It concluded that clinical staging is a better predictor of outcome than histological grade. The survival rates of cases of a given clinical stage were found to be independent of histological grade (Mandviwala et al., 1977). Another substantial study, published in 1997, examined the demographic, etiological and clinic-pathological features of 130 oral cancer cases at JPMC, Karachi. This study attributed most of the cancer cases to tobacco chewing, smoking and betel nut chewing

11 habits. HPV type 16 and 18 were found in 17.6% of cases. p53 gene mutation was observed in 73.8% of cases (Mirza et al., 1998). Reports published by the Pakistan Medical Research Council This body conducted a multi-centre study on the frequency of various types of cancers during 1973-74 and 1977-80 for four cities of Pakistan, i.e., Karachi, Hyderabad, Lahore and Peshawar. Incidence rates of oral cancer per 100,000 population documented were 15.74 for males and 15.70 for females in 1974 while 8.6 for males and 10.33 for females in 1979. According to this report the proportion of oral cancer cases had gone down from 1974 to 1979 while that of carcinoma pharynx had gone up significantly in both sexes. As far as age distribution is concerned, the incidence of oral cancer started rising from 45 years onwards and peaked around 60 to 64 years. Oral cancer came out to be a disease of elderly individuals (Mahmood, 1986). Studies based on patient data from Civil Hospital Karachi Apart from its importance as a contributory centre for the KCR data bank, CHK has been the major centre for the diagnosis and treatment of OSCC in the city. Extensive oral cancer resections as well as tiresome reconstruction surgeries are a routine in this hospital. Moreover, oral cancer coverage area of CHK has expanded beyond the borders of Karachi and it has assumed the status of the main referral centre for OSCC from across the Sindh province. A multicentre study, conducted over 1256 OSCC patients, has been published in 2014, with CHK being the major contributor. The focus of this study was on clinico-pathological parameters of OSCC. Results showed that 63.4% of patients were between 41 to 60 years of age, 73.4% were males. Buccal mucosa of cheek and tongue came out to be the commonest sites of occurrence. In over 70% of cases use of tobacco in paan and with betel nut were the most common etiological factors. Lastly, of all the cancers studied 96.6% were squamous cell carcinomas (Memon et al., 2014). Another valuable contribution from CHK is through a study conducted between 2011 and 2012. The mean age was 48.3+7.3 years with M/F ratio of 1.7:1 and most patients (89.9%) admitted the consumption of carcinogenic agents. Gutka chewing came out to be the most prevalent tobacco habit (66.8%) in studied patients. The most common reason given by the study subjects for delayed presentation was lack of knowledge about the danger posed by their presenting complaints (Zahid et al., 2014).

A recently published study evaluated interleukin 6 & 8 (IL-6, IL-8) and HPV 16 & 18 in saliva of 35 OSCC patients from CHK. IL-6 and IL-8 were detected in 37.1 and 94.3% of cases respectively. HPV 16 was detected in 20% of cases while HPV 18 was found in 17.1% of cancer patients. More than half of all enrolled cases presented with clinical stage T3N0M0. The same study documented premalignant lesions as OSF (77.1%), Leukoplakia (28.6%), Erythroplakia (17.1%) and Lichen plannus (5.7%). HPV-16 was detected in 8.6% of PCLs. Many studied subjects had more than one PCL (Khyani et al., 2014). Cancer data reported from Shaukat Khanum Memorial Cancer Hospital and Research Centre Cancer registry in this hospital has been accumulating data of different cancers reporting for treatment since 1994. This is a specialized cancer treatment facility covering most of the northern and central Pakistan. According to its published data for an 18-year period (1994-2012), lip and oral cavity cancer is the third most common cancer in adults of both sexes (5.96%) and the second most common cancer in adult men (8.54%) while ranks third in women (4.82%). Most of the males were addicted to smoking while females to betel leaf chewing. Main bulk of cancers was of histological grade 1 or 2 and the most seen TNM stages among men and women were stage IV and III, respectively (Badar & Mahmood, 2015). Published data of Karachi Cancer Registry (KCR) on oral cancer According to KCR published data for the period 1995 to 1997, ASR/ 100,000 population for oral cancer was 14.2 in males only to be surpassed by lung cancer at 21.3. For females during the same period ASR was 14.5, second to breast at 53.1. It is the Karachi South data for the same period (1998-2002) that showed an astonishing ASRs of 22.5 in males and 20.4 in females (Bhurgri et al., 2006). Role of HPV in oral carcinogenesis; Studies conducted at Ziauddin University, Karachi Ziauddin University (ZU) has been engaged in oral cancer research for the last six years with special focus on HPV infection in these cases. A study conducted on gutka chewers presenting with oral lesions revealed 17.9% of 262 study subjects HPV positive. Lesions encountered included: oral ulcers (25%), rough mucosa (62%), OSF (24%), leukoplakia

(20%) and erythroplakia (10.6%) [Baig et al., 2012]. A second study on OSF showed 19% HPV positivity out of 187 study subjects presenting with trismus (Rubab et al., 2013) Another study revealed 11 to 20 years as the most common age group suffering from OSF (Wahab et al., 2014). Studies contributed by other centres In this category the first study which can be considered was conducted in two major treatment facilities of oral cancers from Lahore in central Punjab and Multan in southern Punjab. Data of 127 studied cases revealed a mean age of 51.46+12.28 years and M/F ratio of 1.5:1. The most common presenting complaint was a non-healing ulcer (66.1%). Grade I and II tumors were seen in 44.9% of cases while 10.2% cases were grade III tumors. Sites of involvement included tonge (46%) followed by buccal mucosa of cheek (39%) [Kashif et al., 2015]. Researchers at Aga Khan University Hospital, Karachi, conducted a retrospective review of 130 subjects treated at the facility for upper aero-digestive tract cancers between 2007 and 2008. Out of the total number of cases 68.1% were oral cancers, 22.2% laryngeal cancers and 8.3% were hypopharyngeal cancers (Kazmi et al., 2012). 1.3 General trends & risk factors In Asia, sizeable proportions of OSCC cases occur in users of smokeless tobacco (SLT) products (Boffetta et al., 2008; Jeng et al., 2001). There are also evidences for aggregation of oral cancers in the first degree relatives (Goldstein et al., 1994; Foulkes et al., 1996). Polymorphisms of genes for XMEs have also been implicated (Sato et al., 1999; Sreelekha et al., 2001; González et al., 1998). Human papilloma virus has been documented as an etiologic factor in the West especially among non smokers (Braakhuis et al., 2004; Mao et al., 2004). Finally, there are certain proven precancerous conditions like leukoplakia, erythroplakia, oral submucous fibrosis, linchen planus, nicotine stomatitis, tobacco pouch keratosis etc (Neville et al., 2002). Gender associations For decades oral cancer has been regarded as a disease of men. However, in the recent past the ratio has lessened down to two men for each woman. This has been attributed to increased tobacco use by women especially smoking during the last two decades. In Netherlands, a retrospective review of over 300 women diagnosed with OSCC indicated

14 that rates closely approximated those of men (De Boer et al., 1997). In India Ahmedabad, Banglore and the Bombay Cancer Registries showed an annual incidence rate of 12.2, 4.32 and 7.8 per 100,000 for males and 3.94, 6.1 and 4.13 per 100,000 for females (Bhurgri et al., 1998). In Pakistan oral cancers are amongst the most frequent cancers in both males and females. Frequency data from different oral cancer treatment facilities in the country variably reported percentages of oral cancer in males and females (M/F) as: 63.3/36.7 (Ara et al., 2014), 57.3/ 42.7 (Jafarey et al., 1976), 54.6/45.3 (Mahmood et al., 1986), 74/26 (Akram et al., 2013), 60.8/39.2 (Shah et al., 2010), 71/29 (Ashfaq et al., 2014), 59.8/ 40.2 (Kashif et al., 2015) and 63/37 (Kadri et al., 2015). Karachi Cancer Registry data, covering the period from 1995 to 2002, reveals a progressive increase in incidence rates in both genders, more readily apparent in males (Bhurgri et al., 2006). All the above mentioned studies documented variable results although a general male predominance is noted. However, global female tobacco consumption, an important etiological agent for OSCC, is increasing (Amos et al., 2011). 61% of pregnant women belonging to low-income strata of society were found to be smoking tobacco while 51% were also using smokeless tobacco (Das et al., 2012). Age of the patient Historically, OSCC had been regarded as a disease of middle and older individuals. However, there is an increase in oral cancers diagnosed in younger patients (Khyani et al., 2014, Warnakulasuriya et al., 2007). The study from Netherlands found women presenting with oropharyngeal cancers being younger, had a higher incidence of smoking and a positive history of heavy drinking (De Boer et al., 1997). Another cause in those affected at a younger age, as documented in peer reviewed research, is the infection of oral mucosa with HPV type 16 (van Monsjou et al., 2012). The study from Punjab province of Pakistan published in 2015 showed an age range in males of 27 to 80 years and in females 25 to 95 years, although most fall in the range of 40 to 60 years (Kashif et al., 2015). One more study from the same province reported a mean age of 52.56 years in males and 53.43 years in females (Badar et al., 2015). A study from Karachi city during late 90s revealed most of the patients between the age bracket of 41 to 70 years (92/130 cases), only 03 cases were < 30 years of age (Mirza et al., 1998). More recently the study published from AKUH reported an average age of 52+ 13 years (Kazmi et al., 2012). A

15 study done at DUHS on 190 oral cancer patients gave the mean age of males as 50.4 which is higher than that of females at 44.7 while most of these patients belong to low socioeconomic status (Zahid et al., 2014). Mirza et al., reported more p53 positive cases in relatively younger patients (mean age 50.31) as compared to p53 negative cases (mean age 64.91) [Mirza et al., 1998]. Ethnic considerations Karachi is a cosmopolitan city accommodating a number of ethnicities residing in its different localities. These include Urdu-speaking community, that has migrated in large numbers from mostly Nothern India, natives like Sindhi-speaking and Balochi-speaking, Punjabi-speaking who migrated from Central and Northern Pakistan, Memoni-speaking from South-western areas of India, Pashto-speaking, Saraiki-speaking, Pashto-speaking Afghans, Bangali-speaking, Christian communitities from South of India, people from Far-east countries etc. These communities have different genetic lineage and diverse cultural influences which have a special bearing on OSCC. Coming from diverse backgrounds these ethnicities exhibit a variety of chewing and smoking habits. In Karachi studies by Mirza and Bhurgri showed a very high percentage of Urdu-speaking community affected by oral cancer than any other ethnic group (Bhurgri et al., 2003; Mirza et al., 1998). The study done in early 70s in three densely populated localities of Karachi, namely, Korangi, Nazimabad and Pir Colony, revealed a prevalence of 0.05 percent in the sample population, which is comparable to Ahmedabad in India. Korangi was a known high risk population in Karachi city predominantly inhabited by immigrants from Western Uttar Pradesh (UP) province of India and belonging to low-income group. Nazimabad used to be a middle and upper class community from all parts of subcontinent. While Pir Colony was a lower middle class community mainly composed of immigrants from UP but with a mixture of people from other parts of India and Pakistan (Jafarey et al., 1972). A similar study a few years later in mid 70s attributed the difference in the prevalence of oral lesions among people of different origin to the chewing and smoking habits of the people rather than their place of birth. (Jafarey et al., 1976) A recent study (2011- 2012) showed that Urdu-speaking population in Karachi constitute majority of OSCC cases at 52.1%, Sindhi-speaking 27.9%, Pashto-speaking 10.5%, Balochi-speaking 7.9% and Punjabi-speaking 1.6% (Zahid et al., 2014).

Intra-oral sub-site of the lesion Early, small, symptomless tumors more frequently arise at sites such as floor of mouth, ventro-lateral surface of tongue, soft palate and subsequently spread to involve other sites (Mashberg et al., 1995). Contrary to the west, where tongue followed by floor of mouth are the commonest sites, in Pakistan cheek is the most prevalent site possibly due to keeping tobacco with or without additives in contact with the mucosa at this site for long hours. A study from Punjab province of Pakistan reported tongue to be the predominant site (46% of cases), followed by buccal mucosa (39%), lip (6%), floor of the mouth (5.5%) and palate (2.3%) [Kashif et al., 2015]. On analyzing the data for oral cavity cancers in Karachi south for two periods; from 1995 to 1998 and another period between 1998 and 2002, site of occurrence within the oral cavity shows a shifting trend. In males lip cancer has decreased (ASRs fell from 0.9 to 0.7 per 100,000) while it remained staged at 0.4 in females. For tongue cancers, ASRs showed an increase among both sexes. Most significant rise was for tumors arising from buccal mucosa i.e. from 9.1 to 15.3 in males and 9.3 to 12.3 in females (Bhurgri et al., 2006). Pre-cancerous lesions (PCLs) of oral cavity The oral cancer seems to originate normal epithelium through pre-malignant lesion to full blown metastatic phenotype (Santonlis et al., 2007). Early detection of lesions with malignant potential and prompt treatment offers best hope for these patients. Pre- malignant and early-malignant lesions of oral mucosa are asymptomatic. These early lesions may present as white or red patches and may regress after smoking is stopped. ―List of high-risk precursor lesions include leukoplakia, erythroplakia, oral submucous fibrosis(OSF), lichen planus, palatal lesions in reverse smokers, actinic keratosis, discoid lupus erythematosus, dyskeratosis and epidermolysis bullosa‖ (Wake, 1993;Warnakulasuriya et al., 2007). Out of these first four appear to be precursor lesions for a significant proportion of OSCC cases (Mehrotra et al., 2006). (I) Leukoplakia Is defined as ―a white patch or plaque that can not be characterized clinically or pathologically as any other disease‖ (Kramer et al., 1978). Most frequently found in middle-aged and older men. Prevalence increases with advancing age (Tanaka et al.,

2011). Common sites include buccal mucosa, alveolar mucosa, lower lip, floor of mouth and lateral tongue. White color results from a thickened hyperkeratosis of the surface epithelium and/or a thickened spinous layer (acanthosis), which masks the underlying microvasculature (redness) of the connective tissue. The conversion rates could be from 15.6% to as high as 39.2% (Shafer et al., 1961; Waldron et al., 1975). Occurrence of dysplasia in leukoplakia may represent a progressive genetic or molecular alteration and the rate of malignant transformation correlates with the degree of dysplastic change (Liu et al., 2000; Mehrotra et al., 2006). On appearance leukoplakia may be homogenous, nodular or verrucous. For verrucous leukoplakia common sites of involvement include mandibular alveolar ridge, gingivae, buccal mucosa, palate and tongue. The rate of malignant transformation reported is variable as 63% to 100%. It is more commonly found in older females (Lopes, 2010). In citizens of Karachi leukoplakia has been reported as 20% in tobacco chewing general population and 28.6% of all pre-cancerous lesions seen at a tertiary care hospital (Baig et al., 2012; Khyani et al., 2014). (II) Erythroplakia Erythroplakia is defined as a ―red patch that cannot be defined clinically or pathologically as any other condition‖ (Warnakulasuriya et al., 2007). Usually seen in older man as an asymptomatic, well demarcated, velvety, red macule or plaque that is only sometimes associated with sourness and burning sensation. The redness of the lesion is a result of atrophic epithelium that fails to produce keratin, allowing the underlying microvasculature to show through. Leuko-erythroplakia is a mixed red and white lesion whose majority has been reported to harbor dysplastic change with increased rate of malignant transformation (Hsieh et al., 2001). In a series of 65 erythroplakia cases 51% showed invasive carcinoma while 40% revealed severe dysplasia (Shafer et al., 1975). A study conducted in five different localities of Karachi comprising of 262 subjects chewing tobacco with additives revealed erythroplakia in 10.6% of individuals (Baig et al., 2012). (III) Oral submucous fibrosis (OSF) First described by Schwartz in 1952 and is characterized by inflammation and progressive fibrosis of submucosal tissue (Angadi et al., 2011). It is regarded as ―a

18 precancerous condition of insidious onset comprising of deposition of fibrous tissue in the sub-mucosal layer sites like retromolar area, buccal mucosa, soft palate, uvula, anterior faucial pillars, tongue, labial mucosa, and lips‖ (Liu et al., 2004). As the disease progresses there appears hyalinization of lamina propria causing mucosal rigidity of varying intensity. This results in gradual reduction in mouth opening which is a reflection of intensity of involvement (Haider et al., 2000; Pandya et al., 2009). Reduced mouth opening is graded as follows: Grade I: Normal mouth opening. Inter-incisal space > 35mm. Grade II: Restricted mouth opening. Inter-incisal space is 23-35mm. Grade III: Mouth opening is reduced to an inter-incisal space of 15 to 25mm. Grade IVA: Marked reduction in mouth opening, inter-incisal space <15mm. Diffuse fibrosis. Grade IVB: Extensive condition with pre-malignant changes observed all over the mucosa. (Khanna and Andrade 1995) A Study from rural Sindh constituting 280 OSF cases reported involvement of younger males with 99% consuming Gutka, Paan and betel nut. The same study reported 2% conversion of OSF into SCC (Memon et al., 2015). The study conducted on OSF in Clifton area of Karachi revealed more male patients among cases than controls. 55% of patients were Urdu-speaking, trailed by Balochi-speaking (17.5%), Sindhi-speaking (12.5%), Punjabi-speaking (10%) and Pashto-speaking (5%). Chewing habits included gutka (62.5%), areca nut (25%) and betel quid (12.5%). 85% of cases showed severe mouth opening defect while 10% have moderate defect (Akhlaq et al., 2014). Another study from Kemari, Clifton and North Nazimabad areas of Karachi, comprising of 70 OSF cases, showed 57% having severe trismus with grade III mouth opening, 73% showed involvement of palate while 64% have cheek mucosa and lips fibrosis (Wahab et al., 2014). In another study conducted at CHK OSF was reported to involve mostly cheek, palate and tonsillar pillars (Khyani et al., 2014). Rubab and co-workers found in 187 OSF cases, HPV positive oral mucosa in 19.25% subjects (88.9% were males), gutka was found to be the most prevalent chewing habit, followed by naswar, chalia and paan (Rubab et al., 2013).

Alcohol People consuming alcohol in moderate to high amounts have two to three fold more chances of developing oral malignancies than non-drinkers (Khalili, 2008). In Pakistan because of religious and social influences, alcohol is not readily available in the open market. Still we expect hidden exposures in certain subgroups of population. It has been observed that most heavy drinkers were also heavy smokers. When complemented by tobacco use, alcohol has been found to have a synergistic multiplicative effect (Petti, 2009). The risk of oral cancer increases 15 times for a smoker who also consumes alcohol (Alvi et al., 1996). Alcohol is a solvent that facilitates entry of carcinogens into target tissues. Post- metabolism, its metabolite acetaldehyde has been identified as a tumor promoter (Harty et al., 1997). Tobacco Out of several risk factors that have been linked to oral cancer, none is more closely associated than tobacco. Tobacco in all its forms is hazardous. In regions comprising of South and South-East Asia smokeless tobacco is used along with several additives which can alter cancer risk (IARC, 1985). Oral cancer incidence is six times greater among smokers while almost 90% of men and 60% of women with OSCC use tobacco in any form (Baker, 1993). Tobacco smoke contains nearly 60 carcinogenic compounds and their concentration varies in different forms of tobacco used (Hecht, 2003, IARC, 1985; Bhide et al., 1986). For instance, tobacco smoke contains pyrolysis products whereas chewable forms are rich in nitrosamines (Hecht, 2003). Role of chewable tobacco as a risk factor for oral cancer in Karachi The people of Karachi were mostly immigrants from various parts of the subcontinent who settled down here on the formation of Pakistan in 1947. Coming from diverse background they had a variety of chewing and smoking habits. The commonest chewing habit was that of Paan (betel quid), tobacco and supari. For smoking, cigarettes, bidi (tobacco rolled in temburini or tandoleaves) and Hookah (water pipe) were the popular modes (Jafarey et al., 1976). Tobacco habits are equally popular among women of Karachi. According to the National Health Survey which was conducted in 1994, 12.5%

20 of Pakistani women consumed tobacco and 10% were using tobacco in any chewable form (Sufia et al., 2003). Paan users have seven-fold higher risk while users of other smokeless tobacco products have five-time higher risk. These risks remained persistent even after adjustment for other risks factors (Boffetta et al., 2008). The higher risk for paan as compared to other SLTs may be attributable to its areca nut content which has been shown to harbor carcinogenic properties on its own and hence adds to carcinogenicity caused by SLT (Sharma, 2003). Another culprit found is the slaked lime which is used in paan and facilitates reactive oxygen species (ROS) generation in saliva (Khan et al., 2014). 1.4 Molecular mechanisms involved in oral carcinogenesis OSCC is a multifactorial disease with evolution of tumor as an outcome of cumulative molecular events. These events are influenced by both individual‘s genetic predisposition as well as exposure to chemical carcinogen with a predominant role of tobacco exposure. Different carcinogens may involve varied set of genetic aberrations occurring in different sequences and trigger a number of signaling pathways which affect cancer development and progression (Matta et al., 2009; Ha et al., 2009). Multiple factors act in conjunction in the pathogenesis of oral and oropharyngeal cancers (Pfeifer et al., 2009). Although oral cancer may arise de novo, mostly it develops through premalignant lesions revealing hyperplasia and various stages of dysplasia, influenced by carcinogenic agents. The dysplasia then progresses on to invasive carcinoma and subsequently to metastatic disease (Chen et al., 2008). OSCC, like many other cancers, progresses through two biologic stages, first stage is the loss of cell cycle control showing increased proliferation, decreased apoptosis and formation of a cell mass. The second stage is of increased tumor cell motility leading to invasion and metastases (Figure 1.5) [Tanaka & Ishigamori, 2011]. p53 gene mutations p53 guards the stability of the genome as it regulates the cell growth cycle, repair damaged DNA and control cell proliferation through apoptosis of damaged cells. Mutation or inactivation of p53 leads to deregulation of the p53-dependent DNA repair pathway and rendered it incapable of repairing the damaged DNA or inducing apoptosis in mutated cells (Raybaud-Diogene et al., 1996). Thus mutant p53 protein loses the

21 function of tumor suppressor activity and cells replicate with the damaged DNA transforming into malignancy (Stricker et al., 2007). The reported percentages of p53 mutations in head and neck cancers vary between 30 to 70% (Amelia et al., 2009). Indian studies have demonstrated a lower percentage of mutations (17 – 23%) but much higher protein expression rates with most of the p53 being functionally inactive (Vora et al., 2010; Ralhan et al., 2001; Jaysurya et al., 2014).

Figure 1.5 Natural history of oral carcinogenesis [Tanaka T and Ishigamori R, 2011].

Elevated levels of p53 protein are common in SCC of oral cavity (Field et al., 1991; Gusterson et al., 1991; Maestro et al., 1991). Point mutations of p53 in OPLs are seen in 10 to 17% of cases (Cruz et al., 2002; Abraho et al., 2011). From this we can infer that p53 alterations in oral cancer may represent an early event only to be followed later by tobacco-induced carcinogenesis and conversion to malignant phenotype. Mirza and co-researchers published the first report on p53 protein over–expression in Karachi city population. They found p53 protein aberrant expression in 96/130 oral

22 cancer cases (73.8%) and 14/25 cases of oral dysplasias (56%). None of the normal mucosal biopsies revealed the mutated phosphoprotein. They found p53 over-expression to be predominantly related to heavy smoking and tobacco chewing in oral cancer patients than with HPV infection (Mirza et al., 1998). Saleem and colleagues performed mutational screening on 250 OSCC cases and about 90% of their cases showed the mutation of the gene. These mutations may lead to the aberrant expression of p53 protein and may be associated with loss of allele as well as loss of heterozygosisty (LOH) in p53 gene (Saleem et al., 2011). p16 inactivation P16 is also known as p16 INK4A which is a 16kD protein encoded by CDKN2A. It is a negative regulator of cell cycle progression which functions as a CDKI inhibiting CDK4 which otherwise forms a complex CDK4-6/ Cyclin D1 and helps in progression of cell through G1-S phase. There are two states of p16 i.e. loss of tumor suppression and elevated expression due to Rb loss of function (Witkiewicz et al., 2011). High frequency of p16 inactivation, by epigenetic or genetic causes, is found early in development of HNSCCs by loss of heterozygosity, hypermethylation, deletions or mutations of p16INK4A locus with dominance of genetic mechanisms (Pannone et al., 2012). Loss of chromosomal regions CDKN2A in gene 9p21, coding for p16 and p14, are seen in 70- 80% of oral dysplastic lesions (Choi et al., 2008). Pande et al reported p16 loss in 83% of oral precancers (Choi et al., 2008; Pande et al., 1998). CDKN2A inactivation is seen in 75% of HNSCC (Rothenberg et al., 2012; Pérez-Sayáns et al., 2009). HPV-positive tumors are associated with defects in Rb and overexpression of p16. This high level of correlation between p16 and HPV status on OSCC is the basis for considering p16 as an easily detectable surrogate marker for HPV associated tumors (Pande et al., 1998). Epidermal growth factor receptor It is one of the members of Erb B family of transmembrane tyrosine kinase receptors which also include HER2, HER3 and HER4 (Olayioye et al., 2000). Erb B pathway activation is linked to oral carcinogenesis. Erb B receptor pathways may be activated by (i) receptor over-expression, (ii) mutant receptor with ligand independent activation, (iii) autocrine activation by overproduction of ligand and (iv) trans-activation through other receptor systems (Kalyankrishna et al., 2006).

Epidermal growth factor receptor (EGFR) is overexpressed in many tumors of epithelial origin more importantly the carcinomas of breast, colorectum and head and neck squamous cell carcinoma (HNSCC). Risk factor exposure effects on recruitment of various pathways may be appreciated from the fact that in chronic tobacco exposure this pathway is activated downstream from EGFR whereas for cases with chronic areca quid chewing associated tumors, constitutive Ras/Raf/MAPK pathway activation is observed (Matta et al., 2009). 50-80% of OSCCs overexpress EGFR as an early event in malignant transformation with positive expressivity seen in a high percentage of OPL as well, especially in severe than in mild dysplasia (Bernardes et al., 2010; Mendes et al., 2009). EGFR was detected in 40- 80% of OSCCs by Campo-Trapero (Campo-Trapero et al., 2008). H-ras mutations H-ras is amember of Ras family of proteins which are encoded by Ras genes. These are the most commonly affected oncogenes in humans with mutations detectable in 30% of all human cancers (White et al., 2011). Ras proteins are activated via growth factor activation of receptor tyrosine kinase (RTK) which recruits the Ras to the plasma membrane or by constitutive activation due to constant GTP binding. Ras genes function as ON and OFF switch which is in ‗on‘ state conducts signals when attached to GTP but when bound to GDP it switches ‗off‘. Conversion of GTP to GDP by GTPase actively results in termination of downstream signals (Vairaktaris E et al., 2008). This process is controlled by GAP (GTPase-activating proteins) which inactivates Ras genes and GEF‘s (Guanine-nucleotide exchange factors) which facilitates its activation (Murgan et al., 2012). Studies have revealed that particular tumors show mutations in specific Ras isoforms (Fernández-Medarde et al., 2011). K-ras mutation is more commonly associated with pancreatic ductal adenocarcinoma, lung and colon cancers. H-ras is frequently mutated in bladder tumors and OSCC in the Indian Subcontinent in tobacco and HPV-associated tumors (Rothenberg et al., 2012). N-ras mutations have been detected in melanomas and hematopoietic neoplasms (Fernández-Medarde et al., 2011). Saranath in 1991 showed H- ras mutations in association with tobacco chewing related OSCC in India (Sarnath et al., 1991). In Western countries H-ras rate is around 3-5%, whereas in the Indian

24 subcontinent, especially in association with tobacco chewing, the percentage of affected individuals is quite high at approximately 35% (Zushi et al., 2011; Campo-Trapero et al., 2008). C-myc over expression C-myc belongs to myc family of proteins encoded by oncogenes located at chromosome 8q24. It is the most commonly involved member of its family in the process of carcinogenesis. The myc family of proteins comprises of C-myc, L-myc and N-myc. Myc is normally involved in cell proliferation, differentiation, metabolism and apoptosis (Papakosta et al., 2006). It binds to specific DNA sites and regulates transcription of other genes involved in conducting messages for cells. Zushi and co-workers showed its role in increased tumorigenesis (Zushi et al., 2011). It is found to be upregulated in a number of malignancies like Burkitt‘s lymphoma, carcinoma of colon, breast, lung and stomach. In tumor cells it serves to amplify already active genes (Lin et al., 2012). C-myc protein over expression is observed in 10-40% of OSCC (Papakosta et al., 2006). Vora in 2007 has also agreed to overexpression in 26-40% of oral cancers and 75% if tongue cancers are considered alone, with a higher positivity for tobacco-related cancers (Vora et al., 2010). A study in South Indian population revealed C-myc overexpression in 80% OSCCs which is much higher than seen in East Indian population in an older study (Pai et al., 2009; Baral et al., 1998). Cyclin D1 Cyclin D1also called PRAD 1, is anuclear protein encoded by the gene CCND1 located on chromosome 11q13 (Swati et al., 2012). Cyclins are activated when GFs activate Ras/Raf/ERK pathways which activates downstream MAPK pathway. This in turn activates myc which alters the transcription genes in cell cycle including Cyclin D1. Cyclin D1 protein overexpression has been noted in head and neck, oral, laryngeal and nasopharyngeal carcinomas especially in tobacco-related OSCC (Das et al., 2011). It is more common in the HPV-negative tumors; however, a positive relationship with tobacco has been shown by Satya Das (Smeets et al., 2007; Das et al., 2011; Rothenberg et al., 2012). Perez has reviewed has reviewed expression of Cyclin D1 and found it to be overexpressed in 30-60% of HNSCC and 40% of oral dysplasias (Pérez-Sayáns et al., 2009).

Human papillomavirus (HPV) It is an epitheliotropic virus which is commonly transmitted via sexual contact. Once inside the epithelium, HPV replicates in the nuclei, leading to maturation of virons in the suprabasal epithelial cell layers (Del Mistro et al., 2001). More than hundred species of HPV have been identified until now. At least 15 of them fall in the high risk categories which induce malignant transformation (Marur et al., 2010). HPV 16, 18 and 31 are more commonly linked to this malignant potential (Feller et al., 2012). HPV 16 has been attributed in 85 to 95% while HPV 18 in 5-10% of HPV-associated oral cancers (Hunt, 2011). In vitro studies have shown the ability of high-risk HPV genotypes to immortalize oral keratinocytes (Park et al., 1991; Shin et al., 1994). E6 oncoprotein results in rapid degradation of p53 via ubiquitin-directed pathway (Scheffner et al., 1990). Inactivation of p53 abrogates its protective effect and leads to chromosomal instability. In this setting, tobacco could potentiate cancer formation alone or in conjunction with HPV by further contributing to chromosomal instability (Mirza et al., 1998). Binding of E7 to pRb results in release of E2F transcription factor leading to activation of gene transcription (Longworth et al., 2004). In addition E7 mediates degradation of pRb (Wang et al., 2001). HPV association with OSCC and oral pre-neoplastic lesions (OPL) shows a wide variation ranging from 0-100% in different studies done around the world. Luo et al. reported a prevalence of 30.4% in precancerous lesions (Luo et al., 2007). IARC study published in 2003 reported a prevalence of 3.9% in oral cancers and 18.3% in oropharyngeal cancers (Herrero et al., 2003). In Karachi city population, HPV was first reported in the late 90s when high-risk genotypes 16 and 18 were detected in 17.6% of oral cancer cases (Mirza et al., 1998). A later study conducted in 2012 on patients presenting with trismus and are habitual users of chewable tobacco showed HPV positivity in 19.25% of studied subjects (Rubab et al., 2013). More recently Khayani and co-workers reported HPV-16 and 18 in salivary samples of 20% and 17.1% untreated oral cancer cases, respectively (Khyani et al., 2014). pRb gene mutations It is commonly altered TSG in oncogenesis functional at G1-S transition. Cyclin D1/ CDK 4/6 complex phophorylates and inactivates Rb releasing ESF from its combination.

This E2F leads to transcription of factors which drive the cell cycle and are responsible for proliferation of cells. Protein pRb of Rb gene interacts with c-myc gene causing reduced transcription and so reduced proliferative activity. HPV protein E7 binds to Rb and serves to inactivate it. Loss or mutations in Rb gene are found in 66% OSCCs and 64% PCLs (Campo-Trapero et al., 2008). However, a low rate of pRb mutations are quoted in other studies (Choi et al., 2008). 1.5 Tobacco: modes of consumption, metabolism & gene polymorphisms Use of tobacco is as old as human civilization, although mode of its usage has been changing from one era to another. Until the 18th century, in areas now comprising Pakistan, the use of tobacco had been restricted to chewable forms. It was only during the 19th century that forms like cigar and in 20th century as cigarette became common (Rao & Chaturvedi, 2010). Modes of consumption In modern times, there are two basic forms via which tobacco is consumed: smokeless tobacco and smoked tobacco. Smokeless tobacco (SLT) is consumed through mouth or nose (snuffing) without burning it. This includes tobacco chewing, by sucking, by applying its paste on teeth, by stuffing tobacco between teeth and cheek etc. Smoking is done through products like cigarette, cigar, sheesha, chillum (old version of modern sheesha), huqqa and bidi etc (Shaik et al., 2016). (I) Smokeless tobacco (SLT) SLT mode of consumption varies from chewing pure tobacco to a mixture of tobacco with additives such as seen in products like paan (quid), naswar, pan-masala, gutka, Khaini, and Mishri, etc (Nisar et al., 2011; Bhawana, 2013). The habit of chewing these tobacco containing products dates back thousands of years in this part of the world. It has been an ancient custom deeply entrenched in the culture of these areas. Besides the variation in what is chewed, there are many variations in how it is chewed. This fact is of considerable importance in the context of OSCC because many people are in the habit of keeping the gutka or quid inside the cheek for hours. Many even carry these products throughout their night‘s sleep. The common denominators among all chewing habits appear to be tobacco, lime and areca nut (Khan Z et al., 2014)

Betel quid (Paan) Betel quid is locally named as ‗paan‘ and is used to be linked with the Muslim aristocracy. Paan was also a custom of Hindu maharajas. There used to be a person especially employed to prepare paan for the maharaja. Hindus embrace paan in high regard. For instance, Hindus use the pungent leaf in their religious ceremonies (Auluck et al., 2009). The basic ingredients of a Paan include: fresh betel leaf, catechu paste, lime and betel nut cut into small pieces. To this a variety of ingredients are added depending upon personal, ethnic and regional preferences. The most significant addition is that of tobacco either as dried cured leaves or a paste (Toprani & Patel., 2013). All the items are placed on the leaf and folded into a quid which is placed in the mouth and chewed for 10-15 minutes. A regular habitual chewer takes on average 10 quids per day. The quid induces copious salivation producing a deep red mixture which is swallowed or spitted out (Auluck et al., 2009) Gutka Gutka is a fairly recent addition to the constellation of tobacco products in the subcontinent. It is made by combining lime, tobacco, betel nut, zafran (fragrance), flavor and catechu (locally named as Katha). Basic ingredients of gutka are shown in Table 1.5. Recently gutka has assumed importance as being the most popular and commonly used tobacco containing SLT product among citizens of Karachi. It has in fact taken over the place of most of the traditionally used tobacco products because of its low cost and strong effect. Habitual eaters place the gutka in the mandibular or labial furrow, suck it slowly for few minutes and then either through it away or simply let it there till it slowly dissolves (Baig et al., 2012). Naswar Naswar is another form of smokeless tobacco (Nicotiana tabacum) typically produced and used in Central and Southeast Asia (Gupta and Ray, 2003). It contains tobacco, ash, cotton or sesame oil, water, and sometimes gum. Naswar is held in the mouth for 10 to 15 minutes. Alkaloids are also found in larger quantity in naswar which may be responsible for nicotine like effects (Ullah et al., 2011). Nicotine from tobacco is absorbed through oral mucosa and from the mucous membrane of GIT as small amount of snuff is also

28 ingested during the process. It causes addiction if used five times in a week (Zahid et al., 2014). Table 1.5 Basic constituents of a Gutka. Source [Taken and modified from Pindborg, 1989 & Jafarey, 1993].

Ingredient Description Catechu (Aracia The resinous extract of the matrix of the plant is used as a paste. It catechu) largely contains tannins and polyphenols. It gives red color to the gutka. No adverse effects on health are recorded. Lime (Calcium Prepared from sea shells or quarried limestones and used as a Hydroxide) water-based paste. Lime is a local irritant causes mucosal ulceration, epithelial hyperplasia and cellular atypia. Betel nut (Areca The fruit of Areca catechu – a member of the palm family – is catechu) boiled roasted or sun cured and used in finely cut pieces. It contains alkaloid like arecoline and tannins. Arecoline causes fibroblast proliferation and collagen synthesis. Tobacco Various forms of dried tobacco leaves are used in making gutka.

(II) Smoked tobacco In addition to chewing habits people are also addicted to a variety of smoking habits. Predominant tobacco smoking habit is that of cigarette while others include Bidi, Cigar, Pipe and the more recently introduced Sheesha. Smoking increases the oral cancer risk by six times and the risk conferred is dependent upon the amount and duration of tobacco smoked and the same is progressively lowered on cessation of smoking (Moore, 1971). However, the relationship between smoking and site of OSCC in the buccal cavity is less clear as compared to that for smokeless tobacco (Boffetta et al., 1992). ‗Bidi‘ is another smoking product which is a small hand-rolled cigarette that consists of tobacco wrapped in Tendu or Temburni leaf. Its use is widespread among people of low socio-economic status in our part of the world (Rahman et al., 2000). Bidi smokers have a 3.1 fold increased risk of oral cancer as compared to non-smokers (Rahman et al., 2003). Reverse smoking also has significant association with certain forms of OSCC. Some studies have suggested a lesser risk of OSCC with cigar and pipe as compared to cigarettes (Wynder et al., 1977). A study from Puerto Rico reported a four fold increased risk of potentially malignant disorders (Li et al., 2011). Another study from North East

India reported involvement of the anterior aspect of tongue with OSCC in smokers in as much as 93% of the studied cases (Taranikanti et al., 2013). Tobacco habits prevalent among different ethnicities of Karachi Karachi is a cosmopolitan city where a substantial proportion of its population has migrated from different areas of Pakistan. Their eating habits are consistent with their place of origin. Although Karachi is a part of Sindh province, majority of its Sindhi- speaking population has migrated from interior of Sindh (Rubab et al., 2013). Figure 1.6 potrays the typical tobacco culture of this city.

A B

C D

Figure 1.6 Tobacco culture of Karachi: Making of 100 plus infamous Zahoor-Raja-Jani mixed Patti Paans in a posh locality (A), Cigarettes of different brands (B), Typical Paan Khoka selling Paan, Gutka, Supari, cigarettes etc (C,D). The difference in the prevalence of oral lesions among people of different origin has been attributed in the past to the chewing and smoking habits of the people rather than place of birth (Jafarey et al., 1976). Another phenomenon observed is the adoption of tobacco eating culture of one ethnicity by another. For instance, Punjabi-speaking, Sindhi- speaking and Balochi-speaking communities have adopted the Paan culture brought by Urdu-speaking community from India (Rubab et al., 2013). Gutka, introduced in late 60s

30 and 70s, have become widely available and consumed by all ethnicities of the city (Khan et al., 2012). On evaluating 262 subjects from the city for Gutka-habit, Sindhi-speaking have now been found to be more addicted to Gutka (27%) than Urdu-speaking (20%). Interestingly, Pashto-speaking community, well known for naswar habit, has developed much liking for Gutka (9.2% of subjects) and other varieties of tobacco while living in Karachi. Similarly Punjabi-speaking (10.3%) and Balochi-speaking (18.3%) citizens have also developed craving for Gutka (Baig et al., 2012). Another study conducted on 190 diagnosed OSCC cases revealed 66.8% cases addicted to Gutka followed by Paan-masala (53.1%). Other habits included cigarette smoking (45.2%), areca nut (34.7%) and Paan (32.1%) [Zahid et al., 2014]. Tobacco-related OSCC status of Karachi Until late 70s, the commonest tobacco habit in men with oral and oro-pharyngeal cancers in Karachi was paan chewing plus smoking. In female carcinoma cases the commonest habit was chewing of paan with tobacco. At that time the reported risk as compared to non-users for paan chewing was about four times in males and three times in females. Smoking increased the risk by about five times in males and 12 times in females. Paan with tobacco and simultaneous smoking had a relative risk of 23 times in men and 35 times in women (Jafarey et al., 1976). Earlier in late 60s, paan with tobacco, tobacco alone and supari were the commonest habits among males having oral cancers. Naswar was found to be the least tobacco product associated with oral cancers. In females paan with tobacco and supari dominate the habits among cancer patients. In both genders paan with tobacco out numbered any other habit. Smoking of cigarettes, Bidi and Hookah were the habits almost exclusively seen in male patients with oral cancer (Jafarey et al., 1969). Out of 170 oral cancer patients, 30% were found to be addicted to Gutka while only 6.5% to betel nut only. Other tobacco habits included raw tobacco, manpuri, paan and naswar. More importantly in some of these patients development of OSCC had occurred within a span of five to ten years from the start of tobacco habit (Khan et al., 2012). As for PCLs, a recently published study from Karachi on 337 of its citizen‘s revealed gutka as the most prevalent chewebale form of tobacco in OSF patients (45.5%) followed by Naswar (37.56%). In the same study severe trismus was mostly observed in Urdu-speaking

31 community, moderate in Sindhi-speaking and mild in Pashto-speaking (Rubab et al., 2013). Researchers have linked tobacco smoking with oral leukoplakia and oral epithelial dysplasia (Mayne et al., 2006) while smoking habit does not seem to play any role in the causation of OSF (Sinor et al., 1990). Tobacco metabolism There are more than 30 known carcinogens in various tobacco products (Zakiullah et al., 2015). Three major classes include tobacco specific nitrosamines (TSNAs), poly aromatic hydrocarbons (PAHs) and aromatic amines (Rickert et al., 2009). These compounds are procarcinogens and need metabolism by XMEs for conversion into carcinogens. XMEs like Cytochrome P450 (CYPs) are involved in their bioactivation to carcinogenic species while Glutathione-S-transferases (GSTs) cause detoxification of these carcinogens (Zakiullah et al., 2014). Xenobiotics in tobacco Xenobiotics are chemical substances that are foreign to the human biological system. They include naturally occurring compounds, drugs, environmental agents and carcinogens (Omiecinski et al., 2011). Xenobiotic intermediate metabolites exert their adverse effects via covalent interrractions with genetic material or proteins and their related metabolites (Taspinar et al., 2008). Humans have been continuously exposed to naturally occurring xenobiotics contained in food, tobacco, alcoholic beverages, coffee, tea and smoke from burning wood etc. Generation of free radicals is a significant mechanism behind tobacco-induced carcinogenesis (Aust et al., 1993; Kulkarni, 2001; Hechtet al., 1986; Hoffmann et al., 1997). Xenobiotic metabolizing enzyme systems (XMEs) Catabolism of the above mentioned carcinogenic moieties by xenobiotic metabolizing enzymes can be categorized into phase I and phase II. The coordinated expression and regulation of these XMEs determines the outcome of carcinogen exposure. Phase I biotransformation Phase I biotransformation is the function of many enzyme systems. Among this scheme the most prominent pathway is the mono-oxygenation function catalyzed by the cytochrome P450s (CYPs: P450s). A vast number of xenobiotic chemicals is detoxified and/or bioactivated by CYPs. This is achieved through functionalization reactions

32 comprising of aliphatic and aromatic hydroxylation, N- and O-dealkylation, N- and S- oxidation and deamination. Xenobiotics handled include nicotine and acetaminophen, benzene and polyaromatic hydrocarbons (Omiecinskiet al., 2011). Phase II biotransformation Species generated during Phase I form the substrate for phase II reactions. An important effect of Phase II reactions is to conjugate these reactive moieties with endogenous molecules, resulting in their elimination from the body (Devasena et al., 2007). Phase II reactions comprise of glucouronidation, sulfation, methylation, acetylation, glutathione conjugation and amino acid conjugation. Enzyme families associated with phase II reactions include Glutathione S-transferases (GSTs), UDP-glucuronosyltransferases (UGTs), N-acetyltransferases (NATs), sulfotransferases (STs) and various methyltransferases (Omiecinski et al., 2011). Polymorphisms of XME genes By virtue of their genotype for enzymes involved in carcinogen metabolism, certain individuals are more susceptible to cancer formation when exposed to these compounds (Abbas et al., 2014). This could probably be the reason behind the fact that only a small number of persons among those addicted to tobacco or alcohol develop cancer in their lifetime (Zakiullah et al., 2014). (I) Cytochrome P450s (CYPs) Carcinogens present in tobacco are transformed into DNA-reactive metabolites by cytochrome P450 (CYP)-related enzymes, several of which display genetic polymorphism. The enzyme P4501A1 or CYP1A1 encodes for the aryl hydrocarbon hydroxylase involved in the activation of PAHs and aromatic amines. It is expressed in the oral tissue (Bartsch et al., 2000). Genetic polymorphisms as SNPs affect its expression levels. These SNPs may modify the actions of enzymes and hence enhance carcinogen activation (Figure 1.7, Table 1.6). One base substitution of thymine by cytosine in a non-coding region of the gene at position 3801 creates an MspI restriction site (CYP1A1*2A), which does not exist in the wild type genotype (Kawajiri, 1999). Among all polymorphisms, CYP1A1MspI is the most common and is associated with increase in enzyme activity and hence generation of more carcinogenic moieties (Petersen et al., 1991; Landi et al., 1994). This polymorphism results in three genotypes:

wild-type (m1/m1), heterozygous variant (m1/m2) and homozygous (m2/m2) variant (Zhou et al., 2009).

Figure 1.7 Polymorphism of human CYP1A1 gene: Localization of currently known mutations of CYP1A1 gene. Mutation numbers are in chronological order of discovery. m1 and m3 gain a new MspI cleavage site, m2 leads to the loss of BsrDI site and m4 can be detected by Bsarl digestion [Source: Adapted from Cascorbi et al.,1996].

Table 1.6 Overview of CYP1A1 Polymorphism Nomenclature. a wt, wild type. b Mutation m2 is in strict linkage disequilibrium with mutation m1 in Caucasians. (Bartsch et al., 2000). Point Nomenclature Historical mutations Systematic Systematic Nomenclature proposed by nomenclature of nomenclature nomenclature proposed by Nebert et al., polymorphism for mutations for alleles from IARC, 1999 1999 from Cascorbi Cascorbi et al., et al., 1996 1996

Wild-type None wta *1 *1 *1 allele, m1

MspI Allele 3′ 6235 T → m1 *2A *2 *2A non-coding C region, m2 Ile → Val, exon 4889 A → m2 *2B (m1+ m2b) *3 (m2) *2B (m1+ 7, codon 462 G m2)*2C (m2) African- 5639 T → m3 *3 *4 *3 American C specific allele, intron 7 Thr → Asn, 4887 C → m4 *4 *5 *4 exon 7, codon A 461

(II) Glutathione S-transferases (GSTs) Glutathione S-transferases (GSTs) are phase II transformation enzymes involved in the detoxification of hazardous agents (Shukla et al., 2010; Dckant, 2009). Activated PAHs and N-nitroso compounds produced by phase I XMEs are substrates for the glutathione- S-transferase M1 and T1 (GSTM1 and GSTT1) phase II enzymes (Ihsan et al 2014). The main role of GSTs is to detoxify xenobiotics by catalyzing the nucleophilic attack by glutathione synthetase on electrophilic carbon, sulfur, or nitrogen atoms and transform these into non-polar substrates which avert their interaction with crucial cellular proteins and nucleic acids (Josephy, 2010). GST family includes GST α, GST µ, GST θ, and GST π. Gene clusters of GST µ (GSTM1, M2, M3, M4, and M5) and GST θ (GSTT1 and T2) are located on chromosomes 1 and 22, respectively (Nasr et al., 2015). Genetic polymorphisms of these GSTs have been studied extensively for their potential role in lung cancer susceptibility (Raimondi et al., 2006; Calsten et al., 2008). Glutathione S-transferase Mu (GSTM1) present in human lung tissue is characterized by two active alleles GSTM1*A, GSTM1*B and a non-functional null allele which results from the entire GSTM1 gene deletion mutation. Contrarily to GSTM1, Glutathions S- tranferase θ (GSTT1) is polymorphic and characterized by a functional (wild) allele and a non-functional (null) allele. This null allele results from total or partial deletion of the gene. Individuals who are carriers of such genotypes may, therefore, be at increased cancer risk (Lόpez-Cima et al., 2012; Dzian et al., 2012; Ada et al., 2012; Ramzy et al., 2011). Studies conducted on polymorphisms of XME genes Due to variable expression profile for XMEs, polymorphisms of their respective genes can alter the cancer risk posed by tobacco-related carcinogens. Although several studies have evaluated this relationship but the results were inconsistent. Interactions between genotype and environment exposures have long been postulated and studied, however, it has assumed great significance for Pakistan in general and Karachi in particular because of the widespread tobacco use among its citizens and rising prevalence of OSCC and PCLs. In addition, tobacco is consumed here in many forms and its adverse effects are compounded by the addition of several additives making it difficult to evaluate the isolated effect of tobacco on the oral mucosa.

(I) CYP1A1 polymorphisms During the past two decades several studies have evaluated the contribution of CYP1A1 polymorphism in the occurrence of OSCC. However, the conclusions remained inconsistent. A recently published met-analysis on this association has been reviewed (Xie et al., 2016). It comprises of 10 studies from 1999 to 2012 (Tanimoto et al., 1999; Hashibe et al., 2003; Varela-Lema et al., 2008; Drummond et al., 2004; Deng et al., 2004; Evans et al., 2004; Geisler et al., 2005; Xie et al., 2004; Buch et al., 2002; Drummond et al., 2005). The overall analysis suggested that the CYP1A1 MspI gene variants (hetero and homozygous) impart an increased risk than those with wild type genotype especially among Asians. Camparision was drawn between Asians, mixed-race and Caucasian population while the difference in the risk profile was attributed to ethnicities sharing different gene-gene and gene-environment backgrounds (Xie et al., 2016). The case-control study by Zakiullah and co-workers documented a weak association of CYP1A1 variants with OSCC [OR: 1.121 (0.717-1.752)]. The same study showed an increasing association in the simultaneous presence of GSTM1 and/or GSTT1 null genotypes. In this study all included subjects were naswar users while 30.5% were also smoking cigarettes in addition to naswar (Zakiullah et al., 2015). Previously contradictory reports exist about the association of this genetic aberration and oral cancer risk. Meta-analysis of 4635 cases and 5770 controls by Hashibe and colleagues revealed a summary odds ratio (OR) for head and neck cancer of 1.35 (95% CI, 0.95–1.28) for carrying the CYP1A1 Val462 allele (Hashibe et al., 2003). There is an association between CYP1A1 Val462 allele and MspI variant as the former has been reported to be in linkage disequilibrium with the CYP1A1 MspI variant allele in Japanese and Finnish populations (Hayashi et al., 1991, Hironen et al., 1992). The studies included in the meta-analysis except one have also shown similar association results for either variant of CYP1A1 gene. Case-control studies from 1990 to 2000 on the effects CYPs alone or in grouping with detoxifying enzymes showed odds ratios ranging from 2 to 10. These studies suggested that some CYP1A1/GSTM1 0/0 genotype combinations predispose to lung, esophagus, and oral cavityof smokers to higher risk for cancer or DNA damage (Bartsch et al., 2000). In 100 Japanese patients with oral squamous cell carcinoma the presence of the

36 homozygous variant CYP1A1, m2/m2 was founf to be associated with increased risk of oral SCC, in particular, at low cigarette dose levels (Tanimoto et al., 1999). A relatively more recent meta-analysis published in 2008 and comprising of studies carried out on oro-pharyngeal cancers for CYP1A1 and GSTM1 polymorphisms up to 2007 has evaluated data of 30 studies (3177 cases) and another 21 datasets for pooled analysis (3130 cases). The same showed a considerable relationship between oral and pharyngeal cancer and the CYP1A1 Mspl homozygous m2/m2 variant, ORs 1.9 and 2.0 at 95% CI, respectively (Varela-Lema et al., 2008). When we reviewed seven studies which documented the role of CYP1A1 polymorphism in oral cancer (Tanimoto et al., 1999; Sato et al., 1999; Sikdar et al., 2003, Gronau et al., 2003; Katoh et al., 1999; Matthias et al., 1998, Devasena et al., 2007), only three reported a risk association (Tanimoto et al., 1999; Sato et al., 1999; Devasena et al., 2007). This inconsistency warranted further studies on the subject. (II) GSTM1 null genotype The KPK study by Zakiullah et al., on Pashto-speaking ethnicity of Pakistan showed a significant GSTM1 genotype difference between OSCC cases and controls. 79.5% of their 200 cases showed GSTM1 null genotype as compared to 57% of controls (Zakiullah et al., 2015). Another study from North-east India found no significant connection between GSTM1 gene copy numbers with lung cancer risk (Ihsan et al., 2014). Meta-analysis by Hashibi and colleagues showed an overall OR of 1.23 (95% CI, 1.06– 1.62) for all 31 studies reviewed (Hashibe et al., 2003). Nasopharyngeal carcinoma (NPC) patients in a high-risk area in Guangxi showed the deletion frequency of NPC patients as 61.5% (56/91) for GSTM1 with statistically significant difference between patients and controls (Deng et al., 2004). In another study comprising of 283 individuals with head and neck squamous cell carcinoma (HNSCC) and 208 population-based controls, no association with GSTM1 null genotype was observed (Evans et al., 2004). A prospective cohort of 190 patients showed no significant association with GSTM1 null genotype (Geisler et al., 2005). In a population-based case-control study in Puerto Rico, the null variant of GSTM1 was found to be linked with a slightly significant decrease in oral cancer risk (Xie et al., 2004). However, in the Indian case–control study a 1.29 times greater risk was observed (Devasena et al., 2007).

(III) GSTT1 null genotype The study by Zakiullah and colleagues showed the frequency of GSTT1 null genotype more in cases at 47.5% than in controls at 23.2% (Zakiullah et al., 2015). Similarly, the North-East Indian study reported the GSTT1 wild type (+/+) two copy number and hemizygous one copy number more frequent in lung cancer cases than in controls. Interestingly patients with null genotype conferred 68% reduced risk compared to patients with two copy number of GSTT1. Decreasing copy number of GSTT1 gene showed a positive dose relationship with lung cancer (Ihsan et al., 2014). In the meta-analysis by Hashibi M et al., 21 studies have been included for review out of a total of 27 studies that have investigated the risk to OSCC conferred by GSTT1 null allele. An overall OR of 1.17 (95% CI, 0.98–1.4) has been found (Hashibe et al., 2003). Of the remaining seven studies published afterwards, three found a positive association (Deng et al., 2004; Drummond et al., 2005; Barroso Duarte et al., 2006), two found no association (Xie et al., 2004; Sikdar et al., 2004) and only one study endorsed the protective effect (Evans et al., 2004). Among three studies which showed positive association the first was done on 87 patients and revealed a significantly higher frequency of GSTT1 null genotype in those having cancer of the floor of the mouth (Drummond et al., 2005). The second one, comprising of 91 patients with nasopharyngeal carcinoma, showed an increased deletion frequency among cases 59.3% than controls 40.7% (Deng et al., 2004). The third study was conducted on 72 tobacco smoking patients with oral leukoplakia showed that the frequency of the GSTT1 null genotype in the group with oral leukoplakia at 48.6% was statistically different from the controls at 27.8% (Barroso-Duarte et al., 2006). Later, the study from India by Devasena and co-researchers showed a protective effect of this genotype to oral precancerous lesions and oral cancer (Devasena et al., 2007). (IV) Combined gene effects Limited number of studies has analyzed combined gene effects of all our investigated genes i.e. CYP1A1MspI, GSTM1 and GSTT1 (Zakiullah et al., 2015; Sikdar et al., 2004; Sato et al., 1999). In the Indian case-control study, simultaneous presence of CYP1A1 homozygous variant with GSTM1 null genotype upgraded oral cancer risk by 1.6-fold (Devasena et al., 2007). Later the meta- and pooled analysis by Varela-Lema et al.,

38 observed a combined effect for GSTM1 homozygous deletion and CYP1A1 heterozyugous m1/m2 genotypes on cancer risk (Varela-Lema et al., 2008). Finally, the study by Zakiullah et al., reported a weak and non-significant association of CYP1A1 polymorphism with OSCC. However, when considered in combination with GSTM1 and GSTT1 gene alleles, a 16-fold increased risk of oral cancer as compared to controls was observed in the presence of other confounding risk factors (Zakiullah et al., 2015). 1.6 Molecular Diagnostic Techniques Rapid advances over the last few decades in molecular mechanisms for diagnosis of early cancers have revolutionalized the cancer detecton scenarios. Molecular diagnostics is a sensitive method which detects alterations at a molecular (nucleic acid) level, much before clinical presentation or microscopic diagnosis of the disease process. There are various approaches to understanding of molecular basis of oral cancer. Initially these were confined to research purposes only but lately a number of techniques have been incorporated into routine laboratory diagnostic procedures. The common techniques which detect the molecular genetic changes are: 1. Gene sequencing 2. In situ hybridization 3. Immunochemistry 4. Polymerase chain reaction 5. DNA microarray 6. Proteomics Molecular diagnostic techniques have given a deep insight into the molecular mechanisms of oral carcinogenesis. The mutation analysis of oral cancers and preneoplastic lesions can be carried out by DNA and RNA based assays and protein expression profiling and have proved helpful in detection, prognostication and development of appropriate therapy directed against specific molecular targets thus identified (Wong, 2006). (I) Gene sequencing Gene sequencing is the identification of exact arrangement of nucleotides in a DNA molecule. DNA sequencing may be used to find out the sequence of individual genes, larger genetic regions (i.e. clusters of genes or operons), full chromosomes or entire

39 genomes. Application of nucleic acid detection procedures to histological preparations made it possible to visualize molecules in their natural environment (Fournir et al., 1999). A Southern blot is a method routinely used in molecular biology for detection of a specific DNA sequence in DNA samples. The Northern blot is a simple and inexpensive test to study gene expression by detection of RNA (or isolated mRNA) in a sample which indicates cell control over structure and function (Josefen et al., 2011). (II) In situ Hybridization (ISH) It is a type of hybridization technique which uses labeled DNA or RNA strand known as Probes to localize the specific nucleic acid tarets within fixed tissues giving information about gene expression of genetic loci. Two important methods of In situ Hybridization in use are Fluorecent in situ Hybridization (FISH) and Chromogenic in situ Hybridization (CISH). They allow semi-quantitative assessment of gain, losses and amplification directly on tissue sections. These serve to identify amplifications in oncogenes which act as prognostic and predictive tumor markers. FISH uses fluorescence-labeled oligonucleotide probes. It can be performed on dividing or resting cells in fresh frozen tissues as well as paraffin embedded tissues and promptly identifies chromosomal deletions, translocations and gene amplifications (Fournir et al., 1999). (III) Immunohistochemistry It is the detection of protein expression in formalin-fixed papaffin-embedded tissues and is now being used as a common procedure in routine diagnostic surgical pathology. It is commonly used in detection of abnormal cancer cells and can give the exact location of the specific protein in the tissue examined. It is helpful in diagnosis and classification of tumors as it gives information regarding cell lineage and tissue type. Also, it is now a helpful tool in demonstration of prognostic markers for tumors. The procedure is based on the principle of reaction of antibody to specific antigen which is then visualized by immune-peroxidase staining technique, which employs conjugation of antibody to an enzyme which can catalyze color producing reactions. The stains are then evaluated qualitatively or semi-quantiatively by the pathologist (Taylor et al., 2006). (IV) Polymerase chain reaction (PCR) It is a rapid and highly sensitive method which is finding a place in practical laboratory diagnostic procedures. The procedure employs three main steps: denaturing, annealing

40 and polymerization which ultimately results in multiple copies of targeted chimeric gene. This then matches sequence information available from databases. It is very helpful in identification of infectious agents, cancers and inherited genetic disorders (Netto et al., 2003). Reverse Transcriptase PCR (RT-PCR) was developed to amplify RNA targets. In this method, RNA targets are first changed into complimentary DNA (cDNA) by an enzyme called the Reverse Transcriptase and then amplified by PCR (Netto et al., 2003). Nested PCR was developed mainly to increase sensitivity (detect smaller quantities of target), uses two sets of amplification primers with results of first round of reactions subjected to a second round of amplification by primers specific for a sequence within first set of primers (Netto et al., 2003). Multiplex PCR is a type of PCR in which multiple genes are targeted simultaneously using a single mixture of multiple primers specific for different DNA sequences. It has the advantage of using less time and reagents as a single run can give information of more than one target. It utilizes different probes labeled with dyes using specific emission spectra. The annealing temperatures are optimized to work for some reaction. Commercial multiplexing mixtures are available (Netto et al., 2003). Quantitative Real time PCR (Q-PCR) provides thereal time quantitation of PCR product following each cycle. The quantity of fluorescence is relative to the number of copies of the amplification product. It is a rapid procedure as most of times the result can be obtained in two hours (Netto et al., 2003). (V) DNA Microarray A microarray is a fragmentation of DNA fragments marked in a planned pattern on grids onto a solid surface (a ‗chip‘). Each individual spot contains a high concentration of exclusive DNA fragments that map to exact genes or genomic sequences, and correspond to the target for the hybridization of test cDNA derived by reverse transcription of mRNA extracted from tissue samples or cell lines. It allows simultaneous analysis of thousands of genes in one sample, in one assay. These are then utilized to build up data bases for pupose of future comparisons for histogenesis, prognosis and information regarding therapeutic outcomes (Bertucci et al., 2003).

(VI) Proteomics It is the comprehensive learning of a specific ‗Proteome‘ which is an entire complement of proteins together with its modifications under stressful conditions. Clinical proteomics is a sub category where proteomics is applied to clinical specimens as blood. The application of proteomics to cancer can identify biomarkers responsible for diagnosis, prognosis and therapeutic outcomes. As proteins are the main functional units, understanding the exact protein functioning is essential to understand the biological processes. Various technologies such as mass spectrophotometry, protein microarray and nanotechnology may be used to detect and quantify the proteins acting as biomarkers (Wong, 2006). 1.7 Histopathological variants of OSCC (I) Conventional type OSCC Malignant tumors originating from squamous epithelium of the mouth range widely in their degree of differentiation. SCC are categorized as well, moderately and poorly differentiated depending on histological similarity to the normal squamous epithelium. Well differentiated tumors resemble the normal mucosa closely with mild atypia and abundant keratinization and at times formation of ‗Keratin Pearls‘. Moderately differentiated squamous cell carcinoma shows slightly greater atypia than well differentiated SCC with nuclear pleomorphism, normal and abnormal mitosis and less keratinization. Poorly differentiated SCC reveals still greater degree of atypia, mitoses, minimal or absence of keratinization, necrosis and tends to occur at the base of the tongue. Overall the tumors are characterized by cytological atypia, archietectural disarray and invasion into the surrounding tissue and distant metastasis. Invasion is usually accompanied by variable inflammatory infiltration and desmoplastic reaction (Wright et al., 1985). (II) Verrucous carcinoma It is a variant of well-differentiated SCC characterized by a broad based, exophytic, warty, slowly-growing tumor with pushing borders. Club-shaped projections are lined by well differentiated squamous epithelium devoid of malignant features. Surface shows abundant keratosis. Most patients are elderly males engaged in tobacco chewing or snuff

42 dipping (Sandstörm et al., 1982). It may grow through soft tissues of cheek, invade mandible or maxilla, and penetrate perineural spaces (Demian et al, 1973). (III) Basaloid carcinoma It is a variant containing both basaloid and squamous components (Barnes et al., 2005). It is a high grade tumor and has a poor prognosis. Microscopically the tumor is composed of small, basaloid cells with hyperchromatic nuclei and scant cytoplasm occasionally associated with peripheral palisading appearance. A squamous component superficially or focally is common. Necrosis is often present. Immunoreactivity for high molecular weight keratin is a consistent feature (Morice et al., 1988). (IV) Sarcomatoid carcinoma Sarcomatoid or spindle cell carcinoma is composed of sarcoma like formation blending with obvious squamous cell carcinoma. It presents as a polypoid, usually ulcerated mass. More commonly the major bulk of tumor comprises of mesenchymal sarcomatoid tissue with only a minor component of atypical squamoid cells. The tumor cells are plump, fusiform, or epitheloid with mild to moderate pleomorphism arranged in a storiform or herring bone pattern or as interlacing bundles. Immunohistochemistry is often needed to ascertain the epithelial origin. A quarter of cases show regional metastasis but distant dissemination is rare (Thompson et al., 2006). (V) Papillary carcinoma This tumor appears as a soft, friable, polypoid growth with a dominant papillary pattern and cytological atypia. Papillae have a thin fibrovascular core covered by immature basaloid cells. Foci of necrosis and hemorrhage are common. This type of carcinoma needs to be differentiated from squamous papilloma, verrucous carcinoma, and exophytic SCC. Recurrence is common. Some squamous carcinomas present as exophytic growths with papillary fronds (Crissman et al., 1988). (VI) Adenoid (pseudoglandular) carcinoma Adenoid variant shows a pseudoglandular or alveolar appearance due to acantholysis. True glandular formation and mucin production is absent. The tumor is composed of SCC with deeper tumor areas revealing foci of acantholysis in tumor nests, creating glandular differentiation. Differential diagnosis is from adenosquamous carcinoma and mucoepidermoid carcinoma. This is an uncommon variant with a relatively favorable

43 prognosis. Frequently seen on lips and probably induced by actinic radiation (Takagi et al., 1977). (VII) Adenosquamous carcinoma This tumor usually presents as small indurated submucosal nodule or sometimes as an exophytic mass. Surface may be ulcerated. The tumor is characterized by areas of squamous differentiation mixed with areas having true glandular differentiation. The two components lie close but distinct from each other with adenocarcinoma in the deeper parts of the tumor and can be tubular, alveolar or glandular. Intraluminar or intracellular mucin production is present (Martinez-Madrigal et al., 1991). (VIII) Lymphoepithelioma-like carcinoma This tumor which usually occurs in nasopharynx and tonsil, only rarely found in the oral cavity. It is a high-grade tumor which is poorly differentiated and it has reactive lymphocytic infiltrate. Most sinunasal carcinomas are associated with EBV infection (Evans et al., 1991). 1.8 Staging and grading Clinical Staging of OSCC TheTumor Node Metastases (TNM) staging is a clinical staging system that represents the best estimate of the extent of disease as determined from clinical and radiographic examination. The purpose of staging is to give a pre-therapeutic prognosis, determine treatment planning, and compare end result with other centres. In general early stage disease (localized tumor) has a favorable outcome than advanced stage disease (metastasized tumor). The ―American Joint Committee for Cancer Staging and End Result Reporting (AJCC)‖ was formed in 1959 for the reason of developing a system of clinical staging of cancers. In 1954, Union for International Control of Cancer (UICC) was also working on the clinical staging of cancer. The AJCC and UICC joined together with the similar objective of developing an internationally accepted classification system and in 1988 agreed upon the TNM classification system. This system defines a cancer based on variables like the primary tumor (T), spread to regional lymph nodes (N), and distant metastases (M). There are four grades of T, three grades of N, and two grades of M, resulting in 24 TNM categories. 24 TNM categories were condenced into four TNM stages for easing analysis (Edge et al., 2010).

Staging parameters Primary tumor T1 Tumor 2cm or less in greatest dimension T2 Tumor > 2cm but not > 4cm in greatest dimension T3 Tumor > 4cm in greatest dimension T4 Invasion of adjacent structures Regional Lymph Nodes N1 Single ipsilateral lymph node, 3cm or less N2 Single ipsilateral lymph node, > 3cm but not > 6cm, multiuple ipsilateral lymph nodes, none > 6cm, bilateral or contralateral lymph nodes, none > 6cm N3 More than 6 cm. Distant Metastases M0 No distant metastasis M1 Distant metastasis Stage Groupings Stage I T1 N0 M0 Stage II T2 N0 M0 Stage III T1,T2 N1 M0 T3 N0,N1 M0 Stage IVA T1,T2,T3 N2 M0 T4 N0,N1,N2 M0 Stage IVB Any T N3 M0 T4 Any N M0 Stage IVC Any T Any N M1

Histological grading of OSCC Tumor grade considers the structure, differentiation, nuclear pleomorphism, mitoses and degree of keratinization of tissue as observed on microscopic examination. OSCC can be graded by different grading systems. The most frequently adopted system is the WHO system based on four increasing grades that in turn depend on worsening cytological characteristics and architectural organization.

. G1: Well differentiated (low grade) . G2: Moderately differentiated (intermediate grade) . G3: Poorly differentiated (high grade) . G4: Undifferentiated (high grade) [Adapted from Akhter et al., 2011]. A study from Punjab province conducted on 127 OSCC cases reported grade distribution as: 44.9% belonging to well differentiated and moderately differentiated category, while only 10.2% were poorly differentiated (Kashif et al., 2015). A large treating facility from Karachi reported 59.53% of its OSCC cases as moderately differentiated, 32.55% as well differentiated and only 7.9% as poorly differentiated tumors (Hameed et al., 2012) Study from another large tertiary-care private hospital in the city reported 76.4% of the OSCC as moderately differentiated; 13.9% as well differentiated and 2.8% as poorly differentiated SCC (Kazmi et al., 2012).

1.9 Purpose of study

1. To detect polymorphisms of CYP1A1, GSTM1 and GSTT1 gene loci among various tobacco habit groups.

2. To establish susceptibility of these individuals to clinically suspicious oral precancerous lesions and squamous cell carcinoma.

2. METHODS 2.1 Study design It is a case-control study. 2.2 Sampling technique Non-probability (purposive sampling) was performed, researcher choosing the most appropriate cases in terms of case information and research material from the target group of OSCC cases, PCLs, and habit- matched controls. The sample size at 95% confidence level was calculated to find out the genetic polymorphism of CYP1A1MspI, GSTM1 and GSTT1 among various tobacco habit groups having precancerous lesions and oral cancer as well as controls on the basis of previous study documenting 47%, 45% and 44% CYP1A1 genes found in precancerous lesions, oral cancers and controls, respectively (Devasena et al., 2007) with 10% precision and the design effect is 2, using computer software package EPI-Info version 6.0 for calculating the sample size (Daniel, 1987). 2.3 Settings Study cases were collected from Oncology Department of Ziauddin University Hospital, North Nazimabad, Karachi, Otolaryngology ward of Civil Hospital, Karachi, Dr. Ishrat ul Ebad Khan Institute of Oral Health Sciences (DIKIOHS) at Dow University of Health Sciences, Karachi and by setting up camps (atleast 10) at different localities of Karachi (for precancers and controls). The routine histopathology and laboratory procedures for reporting were carried out at respective histopathology labs at Ziauddin University Hospital North Nazimabad, Karachi and Dow Diagnostic Reference and Research Laboratory (DDRRL). Genetic analysis was performed by Dr. Israr Nasir (Incharge Molecular Lab, Ziauddin University Hospital, North Nazimabad, Karachi) and by the principal investigator under Dr. Israr Nasir supervision. Multi-Disciplinary Lab (MDL) at Ziauddin University, Clifton Campus, Karachi, was also utilized for the same. Special assistance was provided by the molecular lab of Liaquat National Hospital and Medical College, Karachi.

2.4 Ethical approvals The study protocol has been approved by the Ziauddin University Ethical Review Committee (ZU ERC) for human research and Institutional Review Board (IRB), Dow University of Health Sciences, Karachi. 2.5 Duration of study The cases of oral cancer, pre-cancerous lesions (PCLs) and controls which fulfilled the selection criteria were mostly collected during a span of five years from January 2010 to December 2014. During the first half of 2015 case collection continued with major emphasis on laboratory procedures related to PCR and gel electrophoresis. Second half of 2015 and first half of 2016 were the periods of thesis writing and submission. 2.6 Inclusion and exclusion criteria Inclusion criteria for cases - Presence of histopathologically diagnosed SCC of oral cavity - Clinically diagnosed oral precancerous lesions - Age above 10 years. Exclusion criteria for cases - Cancer of any site other than oral cavity. - Any other serious disease like diabetes mellitus, hypertension, ischemic heart disease, systemic autoimmune diseases etc. - Those who did not give consent for research. Inclusion criteria for controls - Absence of prior history of any cancer including oral cancer. - Absence of oral pre-cancerous lesions. - Tobacco habit which approx. matches in duration and frequency to those of diseased cases plus habit-free individuals. - Age above 10 years. Exclusion criteria for controls - Cancer of any site within the body. - Any other serious disease like diabetes mellitus, hypertension, ischemic heart disease, systemic autoimmune diseases etc. - Those who did not give consent for research.

2.7 Population of interest This study was carried out on one hundred and fifty (150) cases of clinically diagnosed, histopathologically proven Oral Squamous Cell Carcinoma (OSCC), hundred (100) cases of clinically diagnosed Oral Precancerous Lesions (PCLs) and hundred and eight (108) habit-matched controls for tobacco usage. After obtaining an informed consent from all participants personal details were recorded in a proforma from patient files as well as by direct patient interviews. Information regarding age, gender, ethnicity, tobacco-habit, site of primary lesion within the oral cavity, its size, nodal status, histologic grade and clinical stage were recorded (in cases of OSCC). In cases of studied three PCLs, i.e., leucoplakia, erythroplakia and oral submucous fibrosis, type, location, extent or size within the oral cavity were recorded. In order to analyze the gene-environment interaction, all study subjects were enrolled in the above mentioned three categories. Category one belongs to already diagnosed oral cancer cases from two tertiary care hospitals of Karachi, one being a public sector Civil Hospital Karachi and the other is a private sector Ziauddin University Hospital Karachi. All cancers were re-confirmed by review of histopathology to be squamous cell carcinomas. Category two comprises of patients having oral precancerous lesions like leucoplakia, erythroplakia and oral submucous fibrosis. These cases were enrolled from Dr. Ishrat ul Ibad Institute of Oral Health, Dow University of Health Sciences Karachi, ENT OPD of Civil Hospital Karachi, Edhi free dispensary at Kharader, Karachi as well as from setting up of camps (atleast 10) in various city areas like Railway colony, Landhi, Orangi Town, Liaquatabad etc. Category three were controls having no lesions in the oral cavity but with a history of tobacco use in any form for a period matching patients and inducted in a manner similar to category two. For the purpose of analyzing the interaction between the nature of tobacco exposure and genetic susceptibility factors, the study subjects were further divided into different tobacco-habit groups (Devasena A et al., 2007):  Exclusive chewers– individuals who consumed tobacco only in the smokeless form either with or without additives such as betel nut and lime etc.  Exclusive smokers – individuals who smoked tobacco in forms like cigarettes, bidis etc.

 Mixed tobacco habitués – individuals who consumed tobacco in both the smokeless form and were also smokers.  Habit free group– Individuals who reported a lack of former or current consumption of tobacco in any form. Lifetime exposure was calculated as chewing and smoking index among all tobacco users for both cases and controls as follows: Chewing index= Freq. of chewing events per day × Duration in years Smoking index= Number of cigarettes/10 × Duration in years (Devastana et al., 2007) Selection criteria for clinically suspicious oral pre-cancerous lesions The following criteria, based on visual observation, were utilized for the selection of clinically suspicious oral precancerous lesions: 1. Nodular type or having a single nodule within the lesion 2. Erosion or ulceration 3. A lesion that is hard in its periphery 4. Lesions of anterior floor of mouth and tongue Mehul Shinde, 2015. 2.8 Laboratory Procedures Histopathology reporting Tissues obtained at the time of biopsy or surgery were fixed in 10% neutral buffered formalin and carried to the diagnostic labs for histopathological diagnosis. The representative sections were subjected to routine embedding, processing and staining with Hematoxylin and Eosin (H&E) followed by special stains, e.g. periodic acid sciff, trichrome etc, if required. All cases of squamous cell carcnoma (SCC) were graded into well differentiated, moderately well differentiated and poorly differentiated on the basis of WHO classification. The cases of pre-neoplastic oral lesions were all clinically diagnosed. In only a few cases did we see initial report of a precancerous lesion only to be followed later by the report declaring the diagnosis of SCC. These cases were then included in the cancer category. Routine Laboratory Procedure in Histopathology Fixation and Processing of Tissue 1. Tissues were fixed in 10% neutral buffered formalin at least for 24-48 hours.

2. Tissues were passed through a series in ascending strength of ethyl alcohol starting from 70% through 80% and 95% for a period of one hour in each, followed by two changes in absolute alcohol, one hour each. 3. Tissues were passed through two changes of xylene, each for an hour, to render the tissues transparent. 4. Tissues were infiltrated with molten paraffin at 58°C. Two changes were given for one hour each. 5. Properly processed tissues were correctly orientated in cassettes; labeled and liquid paraffin was poured and allowed to cool in a refrigerator. 6. The blocks were trimmed free of excess paraffin leaving some free margin around the embedded tissues. 7. 2-3μm thin sections were cut and floated in warm water to remove wrinkles and picked up on a glass slide and labeled with a diamond pencil. 8. Slides were placed in warm oven for ten minutes to make sections strongly adherent to the glass slides. Hematoxylin and Eosin (H & E) Staining 1. Sections were de-waxed in two changes of xylene for ten minutes each. 2. Rehydrated in two changes of absolute alcohol, for 5 minutes. 3. Then immersed in 95% alcohol for two minutes and 70% alcohol for two minutes. 4. Washed in distilled water. 5. Sections were stained with Harris‘s Hematoxylin (BDH, USA) (appendix II) for 1-2 minutes. 6. Washed well in running tap water until water was clear. 7. Immersed in Eosin for 1-2 minutes. 8. Washed well in running tap water until water was clear. 9. Dehydrated in ascending grades of alcohol solutions (50%, 70%, 80%, 95% x2 and 100%). 10. Cleared in two changes of xylene. 11. Mounted in DPX (distrene dibutyl phthalatexylene)

Polymerase chain reaction (PCR) Basic Principle of PCR PCR is a method of amplifying target sequences from a DNA specimen and provides a higher degree of sensitivity and specificity than traditional hybridization methodologies. Target DNA strands are denatured and then annealed with two oligonucleotide primer pairs of about 20 nucleotides in length (Figure 2.1).

Figure 2.1 Steps in Polymerase Chain Reaction [Source: Adapted and modified from web site: http://alserv.rug.ac.be/~avierstr/principles/pcrani.html]

Figure 2.2 Exponential amplification in PCR cycling [Source: Adapted and modified from web site: http://allserv.rug.ac.be/~avierstr/principles/pcrani.html]

The primers anneal to sequences located about 100-2000 base pairs apart and act as primers for DNA synthesis by Taq DNA polymerase. Multiple cycles of template denturation and primer extension by the Taq polymerase result in millions of copies of a target DNA sequence (Figure 2.2). Polymerase chain reaction requires relatively small mounts of DNA (25-500ng); consequently, the technique can be applied to DNA extracted from cell suspensions and from formalin-fixed paraffin-embedded tumor tissues. PCR-RFLP (Restriction fragment length polymorphism) PCR-RFLP (Restriction fragment length polymorphism) was the technique utilized for detecting MspI polymorphism in the amplified CYP1A1 gene fragment. The technique involved treatment of PCR reaction mixture with the restriction enzyme MspI (Hpall) Cat. # ERO541, Thermo-Scientific, USA. The recommended protocol for digestion of PCR products directly after amplification involved addition of PCR reaction mixture 10μL, nuclease-free water 18μL, 10X buffer Tango 2μL and MspI 1μL. The mixture was mixed gently and spinned for a few seconds. The same was incubated at 37⁰C for 3 hours. The MspI restriction digestion product of CYP1A1 was loaded on agarose gel stained with ethidium bromide. The gel was carefully taken out after electrophoresis and placed in ultraviolet light for analysis of the amplified products and gel documentation. DNA extraction DNA was extracted from blood samples. Buffy Coat preservation: For this purpose whole blood (5ml) was collected in EDTA (Ethylenediaminetetra acetic acid) tubes, mixed well and centrifuged at 10,000 r.p.m for 10 minutes. With the help of Pasteur pipette buffy coat layer between plasma and red blood cells was removed and transferred to 1.5 ml eppendrof tube. Washing of buffy coat was accomplished with DNA/ RNA free water for several times. Finally, the buffy coat (WBC pellet) was preserved in 1X PBX buffer for further analysis.

DNA Extraction and Purification DNA Extraction from preserved Buffy coat was achieved using PureLink® Genomic DNA Kits for DNA purification manufactured by Invitrogen Life Technologies, Carlsbad USA. Cat. # K182001 as per manufacturer‘s instructions.. Steps for DNA Purification Blood Lysate The following protocol was used to prepare lysate from stored Buffy coat from blood samples: 1. Heat block was set at 55°C. 2. Stored Buffy coat was taken; spinned and 200 μL PBS buffer was added. 3. 200 μL Lysis Buffer plus 20 μL Proteinase K (supplied with the kit) was added to the sample, mixed/ vortex. 4. Incubated at 55°C for 10 minutes to promote protein digestion. 5. 200 μL 96–100% ethanol was added to the lysate. Vortexed well and spinned. 6. Proceeded immediately to Binding DNA. Binding DNA 1. A PureLink® Spin Column was removed in a Collection Tube from the package. 2. The lysate (~640 μL) prepared with PureLink® Genomic Lysis/Binding Buffer and ethanol was added to the PureLink® Spin Column. 3. The column was centrifuged at 10,000 × g for 1 minute at room temperature. 4. The collection tube was discarded and the spin column was placed into a clean PureLink® collection tube supplied with the kit. 5. Proceeded to Washing DNA. Washing DNA 1. 500 μL Wash Buffer 1 prepared with ethanol was added to the column. 2. The column was centrifuged at room temperature at 10,000 × g for 1 minute. 3. The collection tube was discarded and the spin column was placed into a clean PureLink® collection tube supplied with the kit. 4. 500 μL Wash Buffer 2 prepared with ethanol was added to the column. 5. The column was centrifuged at maximum speed for 3 minutes at room temperature. 6. Collection tube was discarded.

7. Proceeded to Eluting DNA. Eluting DNA 1. The spin column was placed in a sterile 1.5-mL microcentrifuge tube. 2. 80 μL of PureLink® Genomic Elution Buffer was added to the column. 3. Incubated at room temperature for 1 minute. The column was centrifuged at maximum speed for 1 minute at room temperature. The tube contained purified genomic DNA. 4. To recover more DNA, a second elution step was performed using the same elution buffer volume as first elution in another sterile, 1.5-mL microcentrifuge tube. 5. The column was centrifuged at maximum speed for 1.5 minutes at room temperature. The tube contained purified DNA. The column was removed and discarded. Storing DNA The purified DNA was stored at –20°C for PCR amplification. To avoid repeated freezing and thawing of DNA, for immediate use the purified DNA was stored at 4°C. Genetic polymorphisms tested 1- CYP1A1 MspI polymorphism: - wild type (m1/m1) - heterozygous variant (m1/m2) - homozygous variant (m2/m2) 2- GSTM1 null polymorphism 3- GSTT1 null polymorphism CYP1A1 MspI polymorphism (Gene location 15q24.1) A 340bp fragment of exon-7 of CYPIAI containing the polymorphic region was amplified by using PCR as described previously (Sivaraman Let al., 1994). Primers sequence: Forward (5‘-CAGTGAAGAGGTGTAGCCGCT-3‘) Reverse (5‘-TCCGTACTCTGTTCTGAGGATT-3‘) [Adapted from: Rozati et al., 2008]. The 10µl polymerase chain reaction (PCR) reaction contained 4.0 µl DNA, 1.5 µl each primer, 10X buffer 1µl, Taq DNA polymerase 0.25 µl, MgCl2 0.2 µl and H2O 1.55µl.

PCR conditions Initial denaturation at 94ºC for 3 minutes, followed by 35 cycles of denaturation at 94ºC for 30 seconds, annealing at 54ºC for 60 seconds and elongation at 72 ºC for 45 seconds and final elongation at 72 ºC for 5 minutes (Rotorgene Thermal Cycler). The PCR product was digested with MspI (Hpall, Thermo Scientific) restriction enzyme for 12hrs at 37ºC (Rozati et al., 2008). After digestion, the CYP1A1 amplified products were loaded on agarose gel stained with ethidium bromide. The gel was carefully taken out after electrophoresis and placed in UV light for analysis and gel documentation. The presence or absence of bands at specified regions corresponded to the presence or absence of CYP1A1 MspI polymorphism. Identification of PCR products after restriction enzyme treatment: Presence of MspI restriction site resulted in splicing of the original 340 bp CYP1A1 fragment into two 200 bp and 140 bp fragments. On the gel each of the three polymorphisms were identified as: - Wild type (m1/m1) → only one 340 bp band. - Heterozygous variant (m1/m2) → three bands of 340 bp, 200 bp and 140 pb, respectively. - Homozygous variant (m2/m2) → only two bands of 200 bp and 140 bp. GSTM1 null polymorphism (Gene location 1p13.3) The GSTM1 deletion polymorphism was identified by amplification of a 219 bp fragment with specific primers as described previously (Rozati et al., 2008). Primer sequence: Forward (5‘-GAACTCCCTGAAAAGCTAAAGC-3‘) Reverse (5‘-GTTGGGCTCAAATATACGGTGG-3‘) [Adapted from: Rozati et al., 2008]. The 10µl polymerase chain reaction (PCR) reaction contained 4.0 µl DNA, 1.5 µl each primer, 10X buffer 1µl, Taq DNA polymerase 0.25 µl, MgCl2 0.1 µl and H2O 1.65µl. PCR conditions Initial denaturation at 94ºC for 2 minutes, followed by 35 cycles of denaturation at 94ºC for 30 seconds, annealing at 62ºC for 45 seconds and elongation at 72 ºC for 45 seconds and final elongation at 72 ºC for 5 minutes.

After PCR amplification, the products of GSTM1 were loaded on agarose gel stained with ethidium bromide. The gel was carefully taken out after electrophoresis and placed in UV light analysis of the amplified products and gel documentation. The presence or absence of a band at specified region corresponded to the presence or absence of GSTM1gene. GSTT1 null polymorphism (Gene location 22q11.23) The GSTT1 deletion polymorphism was identified by amplification of 480 pb fragment with specific primers as described previously (Pemble et al., 1994). Primer sequence Forward (5‘-TTCCTTACTGGTCCTCACATCTC-3‘) Reverse (5‘-TCACCGGATCATGGCCAGCA -3‘) [Adapted from: Ada et al., 2007]. The 10µl polymerase chain reaction (PCR) reaction contained 3.75µl DNA, 1.5 µl each primer, 10X buffer 1µl, Taq DNA polymerase 0.25 µl and H2O 2.0µl. PCR conditions Initial denaturation at 94ºC for 3 minutes, followed by 35 cycles of denaturation at 94ºC for 30 seconds, annealing at 54ºC for 60 seconds and elongation at 72 ºC for 45 seconds and final elongation at 72 ºC for 5 minutes. After PCR amplification, the product of GSTT1 was loaded on agarose gel stained with ethidium bromide. The gel was carefully taken out after electrophoresis and placed in UV light for analysis of the amplified products and gel documentation. The presence or absence of a band at specified region corresponded to the presence or absence of GSTT1. Gel electrophoresis Steps for making 2% agarose gel 1. 10X TE buffer was taken 2. 500 ml 1X TE buffer was made via formula: Desired conc. X desired vol. = 1 X 500 = 50 Stock conc. 10 3. 50ml of 10X TE was taken –deionized water was added making it 500ml in a cylinder. That made 1X TE buffer 4. 1gm of agarose powder was weighed on electric balance keeping in view that the bubble remained in the centre.

5. This 1gm agarose was added to a conical flask. 6. 1X TE buffer was added to make it 50ml. 7. This was placed in the microwave oven for some minutes until the gel melted and bubbles came out. 8. 4µl of Ethidium bromide solution (Cat# SLE1144 Sciencelab.com Inc, USA) was added having a concentration of 10mg/ml into the melted gel and mixed. 9. This gel was poured into the electrophoresis bath after placing the comb making sure that the combs are dipped in it. 10. Any bubble was also removed by a tip while the gel was still liquid. 11. Cooled it. 12. The comb was removed. 13. Now the gel was ready to use. 14. 100bp ladder (Cat# SM0323 M/s Fermentas) was loaded along with sample on a separate well to judge the product size. 2.9 Statistical analysis We analyzed demographic data by descriptive statistics and compare the results among cases and controls. Different genotypes (CYP1A1, GSTM1 and GSTT1) were distributed among different habit groups, ethnicity and site of lesion for control, cancer and precancer cases. The results were expressed as percentage of total number of cases from each category. To evaluate the association between tobacco exposure and oral cancer outcome, odds ratios were calculated among cases and controls while the precision of odds ratios was adjusted by 95% confidence interval (CI). Mean age of patients was adjusted to exclude the aging factor. Number of control cases was also adjusted to match the age and tobacco lifetime exposure of cancer cases. Finally 98 (90 with tobacco habit and 8 no habit) controls were used for comparison. The risk due to studied genes was determined by binary logistic regression model with CYP1A1 m1/m1, GSTM1 not null and GSTT1 not null considered as the reference category. Multi-variant analyses were performed to understand gene–gene and gene–environment interactions. SPSS (Statistical program for social sciences) software version 20 was used for the statistical analysis of data. Chai square test was applied for the determination of significant difference between groups. The significance level was considered as P<0.05.

3. RESULTS 3.1 Characteristics of study subjects The present study comprises of 150 cases of oral squamous cell carcinoma and 100 cases of oral pre-cancerous lesions. 108 age- and sex-matched controls were taken with and without tobacco-habit for comparison with both cancer and precancerous cases. The ages of OSCC cases ranged from 20-78 years with the mean age being 47.1 ± 12.22 (Table 3.1). Out of 150 OSCC cases 98 were males and 52 were females, the male to female ratio was 1.88:1. The ages of patients with precancerous lesions ranged from 16 to 78 years, the mean age was 34.17 ± 13.78 and male to female ratio was 4:1, 80% males and 20% females. In controls, age ranged from 15 to 87 years with mean age being 41.6 ± 14.58. Among controls, the male to female ratio was approximately 3:1, 74.08% males and 25.92% females, respectively. Table 3.1 Characteristics of the study subjects Variable Controls PCL Oral cancer Age Mean + SD 41.6 ± 14.58 34.17 ± 13.78 47.1 ± 12.22

Range 15-87 16-78 20-78

Gender, n (%) Male 80 (74.08) 80 (80) 98 (65.33)

Female 28 (25.92) 20 (20) 52 (34.66)

Tobacco status, Exclusive 70 (64.81) 85 (85) 87 (58) n (%) chewing Exclusive 8 (7.40) 3 (3) 8 (5.33) smoking Mixed habit 22 (20.3) 8 (8) 30 (20)

No Habit 8 (7.40) 4 (4) 25 (16.66)

Total 108 100 150

Detailed questions regarding the established risk factor of tobacco consumption in forms like cigarette smoking, tobacco-chewing (with or without additives) and mixed habit of both tobacco chewing and smoking revealed it to be intimately associated with oral cancers as 125 out of 150 patients (83.44%) were chronically exposed to atleast one of these habits. Only 25 patients (16.66%) did not give a positive history of addiction to these carcinogenic insults (Table 3.1). Similarly in cases of PCLs, 96 out of 100 patients

60 gave a positive history of tobacco use in one form or the other while only 4% patients denied any use of tobacco during their lives. In order to analyze the interaction between the nature of tobacco exposure and genetic susceptibility factors, the study population was divided into different habit groups. Exclusive chewers (70 controls, 85 PCLs and 87 cancers) comprised of individuals who consumed tobacco only in the smokeless form either with or without additives like betel nut, lime etc. Exclusive smokers (8 controls, 3 PCLs and 8 cancers cases) were individuals who consumed tobacco in the form of cigarettes or bidis. Mixed habit group (22 controls, 8 PCLs and 30 cancer patients) comprised of cases that consumed tobacco in smokeless form and were also smokers. Individuals who reported a lack of former or current consumption of tobacco formed the Habit free group (08 controls, 4 PCLs and 25 cancer cases). Out of all 250 PCLs plus OSCC cases, 172 (68.8%) subjects gave the history of exclusive tobacco chewing which turned out to be the most dominant tobacco habit and was much more prevalent than exclusive smoking, only 11 cases. Finally, a significant number of both PCLs and OSCC cases (n=38, 15.2%) were having a combination of chewing and smoking habits which emphasized the additive effect of these known risk factors in the causation of both cancers and pre-cancerous lesions. Distribution of all 350 subjects registered for the study along ethnic lines showed that the majority in each of the three categories, i.e. PCLs, cancers and controls, belonged to the Urdu-speaking community only to be followed by Memoni-speaking, Sindhi-speaking and Balochi-speaking communities (Table 3.2). Table 3.2 Ethnic distribution of PCLs, OSCC & Control cases

Ethnicity Control PCLs OSCC n (%) n (%) n (%) Urdu spk. 56 (51.85) 59 (59) 75 (50) Balochi-spk. 12 (11.11) 5 (5) 17 (11.3) Memoni-spk. 16 (14.81) 9 (9) 17 (11.3) Pashto-spk. 6 (5.5) 12 (12) 7 (4.6) Punjabi-spk. 7 (6.48) 0 8 (5.3) Sindhi-spk. 7 (6.48) 10 (10) 20 (13.3) Others 4 (3.70) 5 (5) 6 (4) Total 108 (100%) 100 (100%) 150 (100%)

Urdu-speaking persons comprised half (n=75) of oral cancer cases. Others ethnic groups include Sindhi-speaking (n=20), Memoni/Balochi-speaking (n=17 each), Punjabi/ Pashto- speaking (n=8/7) and others (n=6%). Within the same category of OSCCs, Urdu- speaking males dominate females and the same is true for Balochi-speaking and Punjabi- speaking while females dominate in numbers in communities like Sindhi-speaking, Memoni-speaking and Pashto-speaking. In PCLs, the majority (59%) again belonged to the Urdu-speaking community followed by Pashto-speaking (12%) and Sindhi-speaking (10%). 3.2 Distribution OSCC cases according to age, gender and site of lesion Table 3.3 gives the division of 150 OSCC cases according to the gender, age and intra- oral sub-site of occurrence. Out of 150 cases 46 (31%) were in the age group between 41 to 50 years, 36 (24%) cases were in 51 to 60 years age group while 35 (23%) cases were in 31 to 40 age bracket. The majority of male patients were seen in the fourth decade (29 cases) while most female patients fall within the fifth decade (21 cases).

Table 3.3 OSCC Cases According to Age, Sex & Intraoral Subsite

Location within the oral cavity

1 2 3 4 5 6 7 Age M F M F M F M F M F M F M F Total 11-20 1 1 2

21-30 1 6 3 1 2 1 14

31-40 1 20 1 3 1 5 4 35

41-50 1 1 18 11 2 4 1 4 4 46

51-60 5 10 9 1 2 1 1 1 4 2 36

61-70 3 3 1 3 1 2 1 14

71-80 1 1 1 3

Total 7 2 59 25 1 1 10 9 3 1 1 0 17 14 150

Key: 1-Alveolar, 2- Cheek, 3- Floor of mouth, 4- Lips, 5- Palate, 6-Retromolar area, 7-Tongue

Cheek turned out to be the most common site of involvement in both males and females. 84 cases were from this site alone out of which 59 were males and 25 were females. The next common sites were tongue (n=31, 20.66%) and lips (n=19, 12.66%) affecting males more often, i.e., 17 males and 14 females for tongue while 10 males and 9 females for lip

62 cancers, respectively. Alveolar mucosa was the primary site in 9 (6%) cases, out of which 7 were males and 2 were females. Palate was the site in 4 (2.66%) patients, out of which 3 were males and only one was female. The least common affected sites included floor of the mouth (n=2, 1.3%) and retro-molar area (n=1). Gender distribution for intra-oral sub- sites revealed that males had more number of cases in each of the three common sites of occurrence, i.e. buccal mucosa of cheek, tongue and lips. 3.3 Distribution of OSCC cases according to histological grade and clinical stage In this study, we examined oral cancers with varying presentations as regards to the site of the primary lesion, extent of disease or the clinical stage on hospitalization as well as three grades of carcinomas namely well-, moderately- and poorly-differentiated tumors. Few examples of different categories of lesions that were recorded, both photographs of lesions as well as those of histological preparations, are shown in figures 3.1 to 3.11.

Figure 3.1 Carcinoma of buccal mucosa, Stage III, Case ID-105.

Figure 3.2 Carcinoma of left cheek, Stage II, Case ID-122.

Figure 3.3 Carcinoma of the lip, Stage IV, Case ID-48.

Figure 3.4 Metastatic carcinoma, Stage IV, Case ID-03.

Figure 3.5 Moderately-differentiated oral squamous cell carcinoma (H&E;× 200 magnification), Case ID-142.

Figure 3.6 Moderately-differentiated SCC of tongue showing muscle invasion (H&E; × 200 magnification), Case ID-132.

Figure 3.7 Intact stratified squamous epithelium, tumor pushing from below (H&E;× 200 magnification), Case ID-110.

Figure 3.8 Moderately differentiated SCC with early invasion from rete ridge (H&E;× 200 magnification), Case ID-107.

Figure 3.9 Lymph node metastasis by poorly differentiated SCC (H&E;× 200 magnification), Case ID-141.

The histological grading of all 150 OSCC cases in the series showed maximum number of tumors in moderately differentiated grade II as compared to grade-I and III (Figure 3.10).

Grade I 36% Grade II Grade III

59%

Figure 3.10 Division of OSCC cases based on histological grade

13%

29% Stage I Stage II Stage III 35% Stage IV

23%

Figure 3.11 Division of OSCC cases based on clinical stage

There were 54 (36%) cases in grade-I, 89 (59%) cases in grade-II and only 7 (5%) cases in grade-III. The TNM clinical staging of the OSCC series revealed that the presentation was significantly distributed in late stages (Figure 3.11). Majority of cases belonged to stage-II and IV, followed by stage-III and I. There were 53 (35%) and 43 (29%) cases in stage-II and IV, respectively. The rest consisted of 35 (23%) cases in stage-III and 19 (13%) cases in stage-I.

3.4 Tobacco indices of OSCC cases and controls Lifetime tobacco exposure was calculated as chewing and smoking index among all tobacco users for 140 OSCC cases and 90 controls. Prior to this an age-adjustment was done so as to achieve a certain level of matching by cutting down the original 150 OSCC and 108 control cases to 140 and 90, respectively (08 controls were tobacco-habit free individuals while 10 with a positive tobacco history were excluded). The mean and median exposures were determined (Table 3.4).

Table 3.4 Tobacco indices in OSCC cases & controls Tobacco Indices Controls Oral cancers All chewers Mean ± SE 279.01±25.65 264.27±26.96 Median 200 200 Exclusive chewers Mean ± SE 277.46±29.0 230.17±28.01 Median 210 150 Smoking index All smokers Mean ± SE 27.15±4.62 28.22±5.55 Median 21 20 Exclusive smokers Mean ± SE 31.57±7.16 34.21±10 Median 30 40

For all chewers, comprising of individuals with exclusive tobacco chewing habit plus those having chewing as a component part of mixed habit, mean and median values were comparable. However, among exclusive chewers, cancer cases reflected lower mean and median values as compared to controls. Again among all smokers, these values were comparable, while for exclusive smokers the median value was slightly higher in cancer cases (Figure 3.12).

350 300 250 200 150 100 50 0 All chewers Exclusive chewer All smokers Exclusive smoker

Figure 3.12 Mean tobacco indices with standard error in OSCC cases (red) and controls (blue) for all tested four categories of tobacco users.

3.5 Genotype variant frequencies in OSCC cases and controls Table 3.5 gives the dispersal of CYP1A1 MspI, GSTM1 and GSTT1 gene variants in oral cancer cases and controls. The frequency distribution of CYP1A1 MspI heterozygous and homozygous variants was found to be 62.85% and 18.57% among OSCC cases and 62.24% and 11.22% in controls, respectively.

Table 3.5 CYP1A1MspI, GSTM1 and GSTT1 genotype variants in OSCC cases and controls

Genotype Control, Cancer, CI (95%) OR P-value n (%) n (%) CYP1A1 26 (26.53) 26 (18.57) m1/m1 m1/m2 61 (62.24) 88 (62.85) (0.71-2.52) 1.44 0.26 m2/m2 11 (11.22) 26 (18.57) (1.0-6.20) 2.36 0.05 GSTM1 67 (68.36) 97 (69.28) Not null Null 31 (31.63) 43 (30.71) (0.51-2.22) 0.95 0.91 GSTT1 Not 96 (97.95) 123 (87.85) null Null 2 (2.04) 17 (12.14) (1.49-29.4) 6.63 0.01

The distribution of GSTM1 null variant was found to be 30.71% and 31.63% in cases and controls, respectively. The GSTT1 null variant was seen in 12.14% of cancer cases and

2.04% of controls. The numbers of GSTT1 null and CYP1A1 m2/m2 (homozygous) gene variants were appreciably higher in cancer patients as compared to controls with p-values ≤ 0.05. The frequency of GSTM1 null and CYP1A1 m1/m2 (heterozygous) variants in controls approximates that of cancer cases. The dispersal of these genetic variants points towards a possible influence on the occurrence of OSCC (Figure 3.13). Overall, the CYP1A1 homozygous variant and GSTT1 null genotypes increased oral cancer risk by revealing ORs of 2.36 and 6.63, respectively. For both the mentioned genotypes, p values were found significant. On the contrary, GSTM1 showed a trend towards protection with an OR of 0.95, however, the results remained statistically insignificant.

P< 0.05

P= 0.05

Figure 3.13 ORs with 95% CI for CYP1A1 MspI, GSTM1 and GSTT1 genotype variants among OSCC patients and controls

3.6 Risk analysis according to type of tobacco exposure The risk contribution by CYP1A1, GSTM1 and GSTT1 to OSCC was calculated individually by utilizing the binary logistic regression model while considering CYP1A1 m1/m1 (wild type), GSTM1 not null and GSTT1 not null as the reference genotypes.

Table 3.6 and Figure 3.14 summarize results of genotype-tobacco interactions by giving the exact number of cases for each of the tested polymorphisms in subjects, both cases and controls, which are engaged in different types of tobacco habits.

Table 3.6 Genotype distribution among different tobacco exposure groups for oral cancer and controls Genotype Tobacco Genotype/ Control Oral CI (95%) OR P- value exposure Ref. cancer CYP1A1 Chewers m2/m2 4 18 m2/m2 Ref. m1/m1 16 10 1.8-27.5 7.2 0.003 < median m2/m2 3 5 Ref. m1/m1 10 7 0.42-13.38 2.3 0.32 > median m2/m2 1 13 Ref. m1/m1 6 3 2.2-304.5 26 0.009 Smokers m2/m2 4 1 Ref. m1/m1 1 3 0.0035-1.94 0.083 0.12 Mixed m2/m2 1 4 habits Ref. m1/m1 6 7 0.29-39.6 3.42 0.32

No habit m2/m2 2 3 Ref. m1/m1 3 6 0.078-7.20 0.75 0.81 Total m2/m2 11 26 Ref. m1/m1 26 26 0.97-5.7 2.36 0.057 GSTM1 null Chewers Null 16 28 Not-null 47 55 0.72-3.09 1.49 0.28 < median Null 6 11 Not-null 24 25 0.56-5.51 1.76 0.33 > median Null 10 17 Not-null 23 30 0.50-3.37 1.30 0.59 Smokers Null 4 4 Not-null 4 4 0.14-7.0 1 1 Mixed Null 6 5 habits Not-null 13 22 0.12-1.93 0.49 0.31 No habit Null 5 6 Not-null 3 16 0.04-1.24 0.22 0.08 Total Null 31 43 Not-null 67 97 0.54-1.67 0.95 0.88 GSTT1 null Chewers Null 2 8 Not-null 61 75 0.66-15.88 3.25 0.14 < median Null 1 2 Not-null 29 33 0.15-20.39 1.75 0.66 > median Null 1 6 Not-null 32 42 0.52-39.87 4.57 0.16 Cont….

Genotype Tobacco Genotype/ Control Oral CI (95%) OR P- value exposure Ref. cancer GSTT1 Smokers Null 0 1 null Not-null 8 7 N/R Mixed Null 0 7 habits Not-null 19 20 N/R No habit Null 0 1 Not-null 8 21 N/R Total Null 2 17 Not-null 96 123 1.49-29.4 6.63 0.01

P< 0.05

P<0.05

P< 0.05

Figure 3.14 ORs with 95% CI for genotypes among different tobacco- exposure groups.

The presence of CYP1A1MspI m1/m2 genotype and oral cancer showed no significant association. However, the homozygous (m2/m2) variant contributed an increased risk of oral cancer with an OR of 2.36 (95% CI, 1.0-6.20,p=0.05), as shown in the table 3.5. When observed in the exclusive chewer‘s category the risk and hence the OR increased to 7.2 (95% CI, 1.8-27.5, p=0.003). Upon dividing the tobacco chewer‘s category into above and below median exposure, the risk conferred was further increased, OR=26 (95% CI, 2.2-304.5, p=0.009) for the above median exposure group (Table 3.6). For GSTM1 null genotype, no significant association was observed with OSCC in any of the tobacco exposure category. However, the GSTT1 null polymorphism revealed an overall OR of 6.63 (95% CI, 1.49-29.4, p=0.01) independent of any tobacco exposure type and an enhanced risk for oral cancer (Table 3.5). 3.7 Gene Interactions and Tobacco exposure Interactions between different combinations of three tested polymorphisms and various tobacco exposures were analyzed based on the assumption that certain genotype combinations can insert a synergistic action on the risk to OSCC. Enhanced carcinogen potentiation by phase I enzymes with simultaneous reduction of detoxification by phase II enzymes can translate into increased risk than caused by polymorphism of an isolated gene. The same hypothesis was evaluated in the current study as regards to the type and quantity of tobacco used. Table 3.7 and figure 3.15 show results of gene combinations with different tobacco exposure groups. Results show that simultaneous occurrence of GSTM1 null and CYP1A1 homozygous (m2/m2) variants raised the OR to 12.8 (95% CI, 1.20-135.5 p=0.03) in tobacco chewers. Studied subjects carrying GSTM1 null and GSTT1 not null gene variants revealed an OR of 1.49 (95% CI, 0.71-3.13) among exclusive chewers and a trend towards protection among mixed habit group i.e. OR 0.24 (95% CI, 0.04-1.38). However, these observations lacked statistical power. The combination GSTM1 not null and GSTT1 null variants, in our overall studied subjects, conferred an OR of 4.58 to OSCC (95% CI, 0.99-21.2, p=0.05) and these values were statistically significant. However, upon stratifying these cases into different habit subgroups the risks conferred did not show statistical significance.

Table 3.7 Effect of genotype combinations on OSCC cases and controls

Gene variant Tobacco Genotype/ Control Cancer CI (95%) OR P combinations exposure Ref. value

CYP1A1 m2/m2 Chewers m2/m2 or 19 38 or GSTM1 null GSTM1 null Ref. m1/m1 or 55 60 0.94-3.55 1.83 0.07 GSTM1 not null median m2/m2 or 11 18 GSTM1 null Ref. m1/m1 or 30 32 0.62-3.77 1.53 0.35 GSTM1 not null Smokers m2/m2 or 5 4 GSTM1 null Ref. m1/m1 or 4 7 0.07-2.76 0.45 0.40 GSTM1 not null Mixed m2/m2 or 7 8 habits GSTM1 null Ref. m1/m1or 17 24 0.24-2.65 0.80 0.74 GSTM1 not null No habit m2/m2 or 6 9 GSTM1 null Ref. m1/m1 or 5 19 0.09-1.64 0.39 0.20 GSTM1 not null Total m2/m2 or 37 59 GSTM1 null Ref. m1/m1 or 81 110 0.71-1.93 1.17 0.54 GSTM1 not null CYP1A1 m2/m2 Chewers m2/m2 and 1 8 & GSTM1 null GSTM1 null Ref. m1/m1& 8 5 1.20-135.5 12.8 0.03 GSTM1 not null

Cont….

Gene variant Tobacco Genotype/ Control Cancer CI (95%) OR P combinations exposure Ref. value CYP1A1 m2/m2 >median m2/m2 and 0 5 & GSTM1 null GSTM1 null Ref. m1/m1 & 2 1 NR GSTM1 not null Smokers m2/m2 & GSTM1 3 1 null Ref. m1/m1& 1 0 NR GSTM1 not null Mixed m2/m2 & GSTM1 0 1 NR habits null Ref. m1/m1& 2 5 GSTM1 not null No habit m2/m2 & GSTM1 1 0 NR null Ref. m1/m1& 1 3 GSTM1 not null Total m2/m2 & GSTM1 5 10 0.48-6.97 1.84 0.37 null Ref. m1/m1 12 13 GSTM1 not null GSTM1 null & Chewers GSTM1 null & 0 2 GSTT1 null GSTT1 null Ref. GSTM1 not null 45 49 NR & GSTT1 not null < median NR >median Smokers GSTM1 null & 0 0 GSTT1 null Ref. GSTM1 not null 4 3 NR & GSTT1 not null Mixed GSTM1 null and 0 3 habits GSTT1 null Ref. GSTM1 not null 13 18 NR and GSTT1 not null No habit GSTM1 null and 0 0 GSTT1 null Ref. GSTM1 not null 3 15 NR and GSTT1 not null

Cont….

Gene variant Tobacco Genotype/ Control Cancer CI (95%) OR P combinations exposure Ref. value Total GSTM1 null & 0 5 GSTT1 null Ref. GSTM1 not null 65 85 NR & GSTT1 not null GSTM1 null & Chewers GSTM1 null & 16 26 GSTT1 not-null GSTT1 not null Ref. GSTM1 not null 45 49 0.71-3.13 1.49 0.29 & GSTT1 not null < median GSTM1 null & 6 15 GSTT1 not null Ref. GSTM1 not null 25 23 0.90-8.18 2.7 0.07 & GSTT1 not null > median GSTM1 null & 10 11 GSTT1 not null Ref. GSTM1 not null 20 26 0.3-2.48 0.84 0.76 & GSTT1 not null Smokers GSTM1 null and 4 4 GSTT1 not null Ref. GSTM1 not null 4 3 0.17- 1.33 0.79 & GSTT1 not null 10.25 Mixed habits GSTM1 null & 6 2 GSTT1 not null Ref. GSTM1 not null 13 18 0.04-1.38 0.24 0.11 & GSTT1 not null No habit GSTM1 null & 0 6 GSTT1 not null Ref. GSTM1 not null 0 15 & GSTT1 not null Total GSTM1 null & 26 38 GSTT1 not null Ref. GSTM1 not null 64 85 0.60-1.99 1.10 0.76 & GSTT1 not null GSTM1 not-null GSTM1 not null & GSTT1 null Chewers & GSTT1null 2 6 Ref. GSTM1 not null 45 49 0.52-14.3 2.75 & GSTT1not null

Cont…

Gene variant Tobacco Genotype/ Control Cancer CI (95%) OR P combinations exposure Ref. value GSTM1 not-null < median GSTM1 not null 1 3 & GSTT1 null & GSTT1null Ref. GSTM1 not null 25 23 0.31-33.6 3.2 & GSTT1not null >median GSTM1 not null 1 3 & GSTT1null Ref. GSTM1 not null 20 26 0.22-23.8 2.3 & GSTT1not null Smokers GSTM1 not null 0 1 & GSTT1null Ref. GSTM1 not null 4 3 NR & GSTT1not null Mixed habits GSTM1 not null 0 4 & GSTT1null Ref. GSTM1 not null 13 18 NR & GSTT1not null No habit GSTM1 not null 0 1 & GSTT1null Ref. GSTM1 not null 3 15 NR & GSTT1not null Total GSTM1 not null 2 12 & GSTT1null Ref. GSTM1 not null 65 85 0.99-21.2 4.58 0.05 & GSTT1not null

P= 0.05

P< 0.05

Figure 3.15 ORs with 95% CI for various genotype combinations and tobacco exposures

3.8 Differential susceptibility among intra-oral sub-sites The likelihood of variable predisposition among different intra-oral sub-sites has been done for the two commonest sites found in our oral cancer study subjects, i.e., buccal mucosa of cheek and the tongue. A dissimilar predisposition profile emerged for the two sites. An overall greater risk to cancers of cheek was observed than those of tongue. The susceptibilities that were seen for the overall study population to oral cancers when single gene polymorphisms were present either independently or in the presence of specific tobacco exposures were simulated at the buccal site alone. For tongue cancers GSTT1 null genotype revealed a somewhat distinct OR of 10.2 (95%CI, 1.57-67.0, p=0.01). To some extent similar observations were made for gene interactions as well. Table 3.8, 3.9 and figures 3.16, 3.17 summarize such results.

Table 3.8 Risk analysis according to intra-oral sub-site (Cheek).

Genotype Tobacco Genotype/ Control Check CI (95%) OR P exposure Ref. value

CYP1A1 Chewers m2/m2 4 12 m2/m2 Ref. m1/m1 16 4 2.48-57.95 12 0.002 median m2/m2 1 10 Ref. m1/m1 6 1 3.13-1146 60 0.006 Smokers m2/m2 4 0 Ref. m1/m1 1 1 NR Mixed habits m2/m2 1 3 Ref. m1/m1 6 6 0.23-37.65 3 0.40 No habit m2/m2 2 1 Ref. m1/m1 3 1 0.05-40.61 1.5 0.82 Total m2/m2 11 16 Ref. m1/m1 26 12 1.12-8.80 3.15 0.028 GSTM1 null Chewers Null 16 15 Ref. Not null 47 32 0.59-3.17 1.37 0.46 median Null 10 11 Ref. Not null 23 18 0.48-4.03 1.4 0.53 Smokers Null 4 01 Ref. Not null 4 2 0.03-7.99 0.5 0.63 Mixed habits Null 6 5 Ref. Not null 13 15 0.17-2.92 0.72 0.66 Cont…

Genotype Tobacco Genotype/Ref. Control Check CI (95%) OR P exposure value GSTM1 null No habit Null 5 3 Ref. Not null 3 5 0.04-2.72 0.36 0.32 Total Null 31 24 Ref. Not null 67 54 0.56-2.12 1.09 0.91 GSTT1 null Chewers null 2 6 Ref. Not null 61 41 0.85-23.1 4.46 0.07 median null 1 5 Ref. Not null 32 23 0.76-63.5 6.95 0.85 Smokers null 0 0 Ref. Not null 8 3 NR Mixed habits null 0 4 Ref. Not null 19 16 NR No habit null 0 0 Ref. Not null 8 8 NR Total null 2 10 Ref. Not null 96 68 1.49-33.2 7.05 0.01 CYP1A1 m2/m2 Chewers m2/m2 & & GSTM1 null GSTM1 null 1 4 Ref. m1/m1 & 8 1 1.56-655 32 0.02 GSTM1 not null median m2/m2 & 0 4 GSTM1 null Ref. m1/m1 & 2 0 NR GSTM1 not null Smokers m2/m2 & 3 0 GSTM1 null Ref. m1/m1 & 1 0 NR GSTM1 not null Mixed habits m2/m2 & 0 01 GSTM1 null Ref. m1/m1 & 2 4 NR GSTM1 not null No habit m2/m2 & 1 0 GSTM1 null Ref. m1/m1 & 1 0 NR GSTM1 not null Cont…

Genotype Tobacco Genotype/Ref. Control Check CI (95%) OR P exposure value CYP1A1 m2/m2 Total m2/m2 & 5 5 & GSTM1 null GSTM1 null Ref. m1/m1 & 12 5 0.47- 2.4 0.29 GSTM1 not null 12.12

P< 0.05

P<0.05

P< 0.05

Figure 3.16 Risk analyses for cancers of cheek mucosa.

Table 3.9 Risk analyses according to intra-oral sub-site (Tongue).

Genotype Tobacco Genotype/ Control Tongue CI (95%) OR P value exposure Ref. CYP1A1 m2/m2 Chewers m2/m2 4 2 Ref. m1/m1 16 1 0.57-111.90 8 0.1 median m2/m2 1 2 Ref. m1/m1 6 1 0.48-294 12 0.12 Smokers m2/m2 4 1 Ref. m1/m1 1 0 NR NR Mixed habits m2/m2 1 0 Ref. m1/m1 6 1 NR NR No habit m2/m2 2 0 Ref. m1/m1 3 1 NR NR Total m2/m2 11 3 Ref. m1/m1 26 3 0.41-13.58 2.36 0.34 GSTM1 null Chewers Null 16 3 Ref. Not null 47 8 0.26-4.66 1.1 0.90 median Null 10 2 Ref. Not null 23 8 0.10-3.20 0.57 0.53 Smokers Null 4 1 Ref. Not null 4 0 NR NR Mixed habits Null 6 0 Ref. Not null 13 3 NR NR No habit Null 5 0 Ref. Not null 3 2 NR NR Total Null 31 4 Ref. Not null 67 13 0.20-2.20 0.66 0.51 GSTT1 null Chewers null 2 1 Ref. Not null 61 10 0.25-36.84 3 0.38 median null 1 1 Ref. Not null 32 9 0.20-62.60 3.55 0.39 Smokers null 0 0 Ref. Not null 8 1 NR NR Mixed habits null 0 2 Ref. Not null 19 1 NR NR

Cont….

Genotype Tobacco Genotype/Ref. Control Tongue CI (95%) OR P exposure value GSTT1 null No habit null 0 0 Ref. Not null 8 2 NR NR Total null 2 3 Ref. Not null 96 14 1.57-67.0 10.2 0.01 CYP1A1 m2/m2 Chewers m2/m2 & 1 1 & GSTM1 null GSTM1 null Ref. m1/m1& 8 1 0.25-255 8 0.24 GSTM1 not null median m2/m2 & 0 1 GSTM1 null Ref. m1/m1 & 2 1 NR NR GSTM1 not null Smokers m2/m2 & 3 0 GSTM1 null Ref. m1/m1 & 1 0 NR NR GSTM1 not null Mixed m2/m2 & 0 1 habits GSTM1 null Ref. m1/m1& 2 1 NR NR GSTM1 not null No habit m2/m2 & 1 0 GSTM1 null Ref. m1/m1 & 1 1 NR NR GSTM1 not null Total m2/m2 & 5 2 GSTM1 null Ref. m1/m1 & 12 3 0.20-12.69 1.6 0.66 GSTM1 not null

P< 0.05

Figure 3.17 Risk analyses for cancers of tongue

CYP1A1 m2/m2 homozygous variant conferred OR of 3.15 (95% CI,1.12-8.80,p=0.028) specifically to cancer of cheek mucosa. For exclusive tobacco chewers, the OR was further increased to 12 (95% CI, 2.48-57.95, p= 0.002). Upon stratifying the same into above and below median lifetime exposures, the OR reached up to 60 (95% CI, 3.13- 1146, p=0.006) in the above median exposure category as compared to controls. All of the above mentioned results showed statistical significance. GSTT1 null polymorphism revealed OR of 7.05 (95%CI, 1.49-33.2, p=0.01) for cheek cancers. For GSTM1 null polymorphism no association was observed for two common sub-sites. When present in combination, CYP1A1 MspI and GSTM1 null polymorphisms showed significant value only for OSCC of cheek mucosa with an OR of 32 (95% CI,1.56-655, p=0.02) among exclusive chewers category for cancer at this site. The same gene combination increased the risk of tongue cancer among exclusive tobacco chewers by showing an OR of 8 (95% CI, 0.25-255), but the association lacked statistical power. 3.9 Tobacco indices in PCL cases and controls Table 3.10 shows mean and median tobacco indices in PCL cases and controls. Figure 3.18 summarizes mean tobacco indices in such subjects.

Table 3.10 Mean & median tobacco indices in Controls and PCL cases

Tobacco indices Controls PCLs All chewers Mean ± SE 279.01±25.65 204.84±21.65 Median 200 144 Exclusive chewers Mean ± SE 277.46±29.0 206.65±22.50 Median 210 144 Smoking index All smokers Mean ± SE 27.15±4.62 17.16±6.42 Median 21 10 Exclusive smokers Mean ± SE 31.57±7.16 27.26±6.79 Median 30 25

300

250 200 150 100 Tobacco Indices Tobacco 50 0 All chewers Exclusive All smokers Exclusive chewer smoker

Figure 3.18 Mean tobacco indices with standard error in PCL cases (red) and controls (blue) for all tested categories of tobacco users

In PCLs, overall tobacco indices for lifetime exposure were lower than in controls. In exclusive chewer‘s category, the median tobacco index was 144 in cases and 200 in controls. In exclusive smokers, mean and median indices were comparable in the two groups. 3.10 Genotype variant frequencies in PCLs and controls Table 3.11 and figure 3.19 show the percentage distribution of CYP1A1 MspI, GSTM1 and GSTT1 gene variants in PCL cases and controls. The proportions of CYP1A1 MspI hetro- and homozygous variants were found to be 67%, 16% in PCLs and 62%, 11% in controls, respectively. The GSTM1 null polymorphism was seen in 38% of PCLs and 33% of controls. GSTT1 null polymorphism was seen in 9% PCL cases and 1.85% controls. The proportions of CYP1A1 hetro and homozygous genotypes were found to be proportionately different in PCL cases and controls. Similarly, differences were noted in percentages of GSTM1 null and GSTT1 null genotypes among both PCLs and controls. However, all of the above differences were found to be statistically non-significant except in case of GSTT1 null polymorphism which shows a statistically significant OR of 5.24 (95%CI- 1.0-27.14, p=0.04). The risk contribution by CYP1A1, GSTM1 and GSTT1 to PCL was calculated individually by utilizing the binary logistic regression model while considering CYP1A1 m1/m1, GSTM1 not null and GSTT1 not null as the reference genotypes.

Table 3.11 CYP1A1, GSTM1 and GSTT1 genotype variants in PCLs and controls

Genotype Control, n (%) PCL, n (%) CI (95%) OR P value CYP1A1 m1/m1 29 (26.85) 17 (17) m1/m2 67 (62.03) 67 (67) 0.85-3.39 1.7 0.12 m2/m2 12 (11.11) 16 (16) 0.87-5.93 2.27 0.09

Total 108 100 GSTM1 Not null 72 (66.66) 62 (62) Null 36 (33.33) 38 (38) 0.57-2.60 1.22 0.60 Total 108 100 GSTT1 Not null 106 (98.14) 91 (91) Null 2 (1.85) 9 (9) 1.0-27.14 5.24 0.04 Total 108 100

Table 3.12 and figure 3.20 summarize results of genotype-tobacco interactions by giving the exact number of cases for each of the tested polymorphisms in subjects, both PCL cases and controls, which are engaged in different types of tobacco habits. In case of CYP1A1MspI m2/m2 homozygous variant no significant association was observed with the occurrence of PCL. However, the GSTM1 null variant revealed an increased risk by an OR of 2.94 (95% CI, 1.0-8.68, p<0.05) for PCLs (Table 3.12). This association was observed in the exclusive chewer‘s category having less than median lifetime exposure to tobacco.GSTT1 null genotype conferred on the whole an OR of 5.24 (95% CI, 1.10- 24.87, p=0.03). Upon stratifying this into various tobacco habits no significant association was noted.

P< 0.05

Figure 3.19 ORs with 95% CI for CYP1A1, GSTM1 and GSTT1 polymorphisms among PCL cases and controls

Table 3.12 Genotype distribution among different tobacco exposure groups in PCL cases and controls

Genotype Tobacco Genotype/ Control PCL CI (95%) OR P exposure Ref. value CYP1A1 m2/m2 Chewers m2/m2 5 13 m1/m1 18 15 0.90-10.76 3.12 0.07 < median m2/m2 3 8 m1/m1 11 6 0.93-25.66 4.88 0.06 > median m2/m2 1 5 m1/m1 7 9 0.36-41.31 3.88 0.26 Smokers m2/m2 4 0 m1/m1 1 0 Mixed habits m2/m2 1 2 m1/m1 7 1 0.57-338.59 14 0.10 No habit m2/m2 2 1 m1/m1 3 1 0.05-40.61 1.5 0.82 Total m2/m2 12 16 m1/m1 29 17 0.87-5.9 2.27 0.09 GSTM1 null Chewers Null 19 33 Not-null 51 52 0.85-3.37 1.70 0.12 < median Null 6 17 Not-null 26 25 1.0-8.68 2.94 0.049 > median Null 10 16 Not-null 25 27 0.56-3.86 1.48 0.42 Smokers Null 4 1 Not-null 4 2 0.03-7.99 0.5 0.63 Mixed habits Null 8 3 Not-null 14 5 0.19-5.60 1.05 0.95 No habit Null 5 1 Not-null 3 3 0.01-2.91 0.2 0.24 Total Null 36 38 Not-null 72 62 0.69-2.16 1.22 0.49 GSTT1 null Chewers Null 2 7 Not-null 68 78 0.61-15.18 3.0 0.17 < median Null 1 3 Not-null 33 39 0.25-25.57 2.53 0.43 > median Null 1 4 Not-null 35 39 0.38-33.65 3.58 0.26 Smokers Null 0 0 Not-null 8 3 N/R

Cont…

Genotype Tobacco Genotype/ Control PCL CI (95%) OR P exposure Ref. value GSTT1 null Mixed habits Null 0 1 Not-null 22 7 N/R No habit Null 0 1 Not-null 8 3 N/R Total Null 2 9 Not-null 106 91 1.10-24.87 5.24 0.03

P< 0.05

Figure 3.20 ORs with 95% CI for genotypes among different tobacco exposure groups in both PCL cases and controls

Interactions between different combinations of three tested polymorphisms and various tobacco exposures were analyzed based on the assumption that certain genotype combinations can exert a synergistic action on the risk to OSCC. Enhanced carcinogen potentiation by phase I enzymes with simultaneous reduction of detoxification by phase II enzymes can translate into increased risk than caused by polymorphism of an isolated gene. The same hypothesis was evaluated in case of PCLs as well as regards to the type and quantity of tobacco used. Table 3.13 and figure 3.21 show results of gene combinations with different tobacco exposure groups. Overall ORs of tested gene combinations were for CYP1A1m2/m2 or GSTM1 null 1.36 (95% CI-0.80-2.29, p=0.24), CYP1A1m2/m2 and GSTM1 null 2.0 (95% CI-0.51-8.33, p=0.3), GSTM1 null and GSTT1 not null 1.26 (95% CI-0.60-2.31, p=0.45), GSTM1not null and GSTT1 null 2.41 (95% CI- 0.88-13.64, p=0.32), respectively. However, none of our results was statistically significant. Table 3.13 Effect of genotype combinations on PCL cases and controls

Genotype Tobacco Genotype Control PCL CI (95%) OR P combinations exposure Condition/ Ref. value

CYP1A1 m2/m2 Chewers m2/m2 or 23 39 or GSTM1 null GSTM1 null Ref. m1/m1 or 60 58 0.94-3.55 1.75 0.07 GSTM1 not null median m2/m2 or 12 20 GSTM1 null Ref. m1/m1 or 29 29 0.69-4.0 1.66 0.25 GSTM1 not null Smokers m2/m2 or 5 1 GSTM1 null Ref. m1/m1 or 4 2 0.02-6.1 0.4 0.52 GSTM1 not null Mixed habits m2/m2 or 9 4 GSTM1 null Ref. m1/m1or 19 5 0.36-7.84 1.68 0.51 GSTM1 not null

Cont…

Genotype Tobacco Genotype Control PCL CI (95%) OR P combinations exposure Condition/ Ref. value CYP1A1 m2/m2 No habit m2/m2 or 6 2 or GSTM1 null GSTM1 null Ref. m1/m1 or 5 4 0.05-3.30 0.41 0.41 GSTM1 not null Total m2/m2 or 43 46 GSTM1 null Ref. m1/m1 or 88 69 0.80-2.29 1.36 0.24 GSTM1 not null CYP1A1 m2/m2 Chewers m2/m2 and 1 7 and GSTM1 null GSTM1 null Ref. m1/m1 and 9 9 0.70-69.0 7 0.09 GSTM1 not null median m2/m2 and 0 3 GSTM1 null Ref. m1/m1 and 3 6 N/R GSTM1 not null Smokers m2/m2 and 3 0 GSTM1 null Ref. m1/m1 and 1 0 N/R GSTM1 not null Mixed habits m2/m2 and 0 1 GSTM1 null Ref. m1/m1 and 2 1 NR GSTM1 not null No habit m2/m2 and 1 0 GSTM1 null Ref. m1/m1 and 1 0 NR GSTM1 not null Total m2/m2 and 5 8 GSTM1 null Ref. m1/m1 and 13 10 0.51-8.33 2.0 0.3 GSTM1 not null GSTM1 null & Chewers GSTM1 null & 0 4 GSTT1 null GSTT1 null Ref. GSTM1 not null & 49 49 NR GSTT1 not null < median NR >median

Cont…

Genotype Tobacco Genotype Control PCL CI (95%) OR P combinations exposure Condition/ Ref. value GSTM1 null & Smokers GSTM1 null & 0 0 GSTT1 null GSTT1 null Ref. GSTM1 not null & 4 2 NR GSTT1 not null Mixed habits GSTM1 null & 0 1 GSTT1 null Ref. GSTM1 not null & 4 5 NR GSTT1 not null No habit GSTM1 null& 0 0 GSTT1 null Ref. GSTM1 not null & 3 2 NR GSTT1 not null Total GSTM1 null & 0 5 GSTT1 null Ref. GSTM1 not null & 60 58 NR GSTT1 not null GSTM1 null & Chewers GSTM1 null & 19 29 GSTT1 not-null GSTT1 not null Ref. GSTM1 not null & 51 49 0.78-3.19 1.58 0.19 GSTT1 not null < median GSTM1 null & 8 15 GSTT1 not null Ref. GSTM1 not null & 26 25 0.53-70 1.95 0.2 GSTT1 not null > median GSTM1 null & 11 14 GSTT1 not null Ref. GSTM1 not null & 25 24 0.50- 1.32 0.57 GSTT1 not null 3.49 Smokers GSTM1 null & 4 1 GSTT1 not null Ref. GSTM1 not null & 4 2 0.03-7.99 0.5 0.63 GSTT1 not null Mixed habits GSTM1 null and 8 2 GSTT1 not null Ref. GSTM1 not null & 14 5 0.1-4.47 0.7 0.71 GSTT1 not null No habit GSTM1 null and 0 1 GSTT1 not null Ref. GSTM1 not null & 0 2 NR GSTT1 not null Total GSTM1 null & 31 33 GSTT1 not null Ref. GSTM1 not null & 69 58 0.60-2.31 1.26 0.45 GSTT1 not null

Cont….

Genotype Tobacco Genotype Control PCL CI (95%) OR P combinations exposure Condition/ Ref. value GSTM1 not-null Chewers GSTM1 not null & 2 3 & GSTT1 null GSTT1null Ref. GSTM1 not null & 49 49 0.24-9.37 1.5 0.67 GSTT1not null < median GSTM1 not null & 1 1 GSTT1null Ref. GSTM1 not null & 25 25 0.05- 1 1 GSTT1not null 16.88 >median GSTM1 not null & 1 2 GSTT1null Ref. GSTM1 not null & 24 24 0.16- 2 0.59 GSTT1not null 23.54 Smokers GSTM1 not null & 0 0 GSTT1null Ref. GSTM1 not null & 4 2 NR GSTT1not null Mixed habits GSTM1 not null & 0 0 GSTT1null Ref. GSTM1 not null & 4 5 NR GSTT1not null No habit GSTM1 not null & 0 1 GSTT1null Ref. GSTM1 not null & 3 2 NR GSTT1not null Total GSTM1 not null & 2 4 GSTT1null Ref. GSTM1 not null & 70 58 0.88- 2.41 0.32 GSTT1not null 13.64

Figure 3.21 ORs with 95% CI for various genotype combinations and tobacco exposures

3.11 Ethnic distribution of genotypes in overall cases and controls In our study, we also evaluated the distribution of all the above genotype variants among common ethnicities of our studied subjects (Table 3.14, figure 3.24). Some very interesting observations we document here in this regard. As almost half of all cases in the present study belonged to one ethnic group, i.e., Urdu-speaking, we record here for the first time their genetic profile for tobacco as the risk factor. In this ethnic group, CYP1A1 MspI heterozygous variant came out to be the most prevalent genotype variant among all the three categories of cases, namely OSCC, PCLs and controls. This is the variant we found in 66.6% of OSCC cases. The percentage of homozygous variant we found equal among PCLs and OSCC cases, 13.5% and 13.3%, respectively.

Table 3.14 Distribution of genotypes according to ethnicity

Ethnicity Genotype Control PCL OSCC n / total (%) n / total (%) n / total (%) Balochi-spk. CYP1A1 m1/m1 2/12 (16.66) 0 0 CYP1A1 m1/m2 8/12 (66.66) 4/5 (80) 12/17 (70.5) CYP1A1 m2/m2 2 /12 (16.66) 1/5 (20) 5/17 (29.4) GSTM1 null 6/12 (50) 1/5 (20) 6 /17 (35.2) GSTT1 null 0 0 2/17 (11.7)

Memoni-spk. CYP1A1 m1/m1 3/16 (18.7) 0 4/17 (23.5) CYP1A1 m1/m2 10/16 (62.5) 6/9 (66.6) 10/17 (58.8) CYP1A1 m2/m2 3/16 (18.7) 3/9 (33.3) 3/17 (17.6) GSTM1 null 5/16 (31.25) 1/9 (11.1) 6/17 (35.2) GSTT1 null 0 0 2/17 (11.7)

Pashto-spk. CYP1A1 m1/m1 2/6 (33.33) 1/12 (8.3) 1/7 (14.2) CYP1A1 m1/m2 3/6 (50) 8/12 (66.6) 2/7 (28.5) CYP1A1 m2/m2 1/6 (16.66) 3/12 (25) 4/7 (57.1) GSTM1 null 4/6 (60) 7/12 (58.3) 3/7 (42.8) GSTT1 null 1/6 (20) 2/12 (16.6) 0

Cont…

Ethnicity Genotype Control PCL OSCC n / total (%) n / total (%) n / total (%) Punjabi-spk. CYP1A1 m1/m1 2/7 (28.57) 0 2/8 (25) CYP1A1 m1/m2 5/7 (71.42) 0 4/8 (50) CYP1A1 m2/m2 0 0 2/8 (25) GSTM1 null 2/7 (28.57) 0 2/8 (25) GSTT1 null 0 0 1/8 (12.5) Sindhi-spk. CYP1A1 m1/m1 3/7 (42.85) 2/10 (20) 6/20 (30) CYP1A1 m1/m2 2/7 (28.57) 7/10 (70) 11/20 (55) CYP1A1 m2/m2 2/7 (28.57) 1/10 (10) 3/20 (15) GSTM1 null 5/7 (71.42) 8/10 (80) 9/20 (45) GSTT1 null 0 1/10 (10) 0 Urdu-speaking CYP1A1 m1/m1 16/56 (28.57) 14/52 (26.9) 15/75 (20) CYP1A1 m1/m2 36/56 (64.28) 37/59 (62.7) 50/75 (66.6) CYP1A1 m2/m2 4/56 (7.14) 8/59 (13.5) 10/75 (13.3) GSTM1 null 12/56 (21.4) 19/59 (32.2) 21/75 (28) GSTT1 null 1/56 (1.78) 6/59 (10.1) 12/75 (16) Others CYP1A1 m1/m1 0 0 2/6 (33.3) CYP1A1 m1/m2 04/4 (100) 5/5 (100) 03/6 (50) CYP1A1 m2/m2 0 0 01/6 (16.6) GSTM1 null 2/4 (50) 2/5 (40) 01/6 (16.6) GSTT1 null 0 0 0

The next ethnic group that showed high proportions of heterozygous variant of CYP1A1 among its PCL and OSCC cases was Sindhi-speaking, 70% and 55%, respectively. This was closely balanced by Memoni-speaking community, 66.6% and 58.8%, respectively. Both the above mentioned ethnicities were represented in much smaller proportions in our sample population as compared to Urdu-speaking community. The highest percentage of homozygous variant was found in oral cancer cases belonging to Pashto- speaking, (57.1%). The highest proportion of homozygous variant among PCL cases as well as control population was seen in Memoni-speaking, 33.3% and 18.7%, respectively. In Urdu-speaking ethnicity, GSTM1 null genotype was mostly found in PCL cases (32.2%), while GSTT1 null was most common in OSCC patients (16%). Other than Urdu- speaking, GSTM1 null variant was most common in PCL cases belonging to Sindhi- speaking (80%) while GSTT1 null among PCLs in Pashto-speaking (16.6%).

Figure 3.22 Distribution of genotypes according to ethnicity in all studied subjects

4. DISCUSSION Most of the studies from Pakistan have reported relative frequencies of different cancers. Population-based incidence or mortality figures are mostly not available. Although data from hospital-based and pathology-based registries are available, such studies are subject to several areas of bias. For instance, referral bias may reflect the availability of certain services at a particular institution. Malignancies seen at a radiotherapy centre may reflect more frequent referral of tumors where radiation is the main mode of therapy as in case of head and neck tumors. Another bias in figures is due to the socio-economic status of patients. In certain cancer treatment facilities there is under-representation of patients belonging to low socio-economic status or vice-versa. This may cause excess of cancers that are more prevalent in middle to higher socio-economic class. For oral cancer in Pakistan, the latest data that can be referred to is from Shaukat Khanum Memorial Cancer Hospital & Research Centre, Lahore and according to its eighteen-year records oral cancer has been the second most common cancer in adults of both sexes accounting for 6.69% of all malignancies (Badar et al., 2015). Published data by the population-based Karachi Cancer Registry has also placed oral cancer in the second position amongst all cancers in both genders, while the ASRs reported for the city are the highest worldwide (Bhurgri et al., 2006). The incidence data published by The Aga Khan University Pathology-based Cancer Registry from Aga Khan University Hospital for oral cancer is not very dissimilar, again placing head and neck cancer as the second commonest cancer in males with a frequency of 10.6%. In females these cancers are fifth inline with a frequency of 6.9% (Malik et al., 1998). This slight deviation from the frequency reported by hospital–based registries, in our opinion, could be because of referral biases. Overall, the hospital-based registry at Shaukat Khanum and pathology based registry at Aga Khan Hospital although situated way appart in the country still they are validating the data of each other with only slight differences. The same is the relevance with population-based KCR data. From these facts we can draw the inference that the frequency of head and neck cancers inclusive of oropharyngeal cancer as reported by these registries is in fact a reflection of the status of these cancers in the overall population of Pakistan.

100

There are differences between cancer prevalence in Pakistan and the Western world. Although smoking-related cancers in the developing countries appear to be achieving the same epidemic proportions as are evident in the developed world, still there are certain cancers that are certainly more prevalent in our part of the world (Badar & Mahmood, 2015). Two such cancers are head and neck cancers and lymphoma. Cancers origination from structures in the head and neck cancer region have been reported as the commonest cancers in Southern Pakistan, consistent with other reports from South Asia (Malik et al., 1998). In the USA, cancers arising from oral cavity and pharynx rank 9th among males while very low incidence is reported in females (American cancer society, 2016). Higher rates of oropharyngeal cancers are observed among black males, while ASRs in overall population has been 15.6 per 100,000 in males and 6.1 per 100,000 in females. Three factors contribute to this cancer in the western world namely; tobacco smoking, alcohol and HPV infection of oral mucosa, specially type 16 and 18. However, there are certain very peculiar differences that exist in our part of the world, particularly in tobacco consuming habits. For instance, in Pakistan it is the much prevalent habit of tobacco chewing in forms like gutka, paan as well as a variety of chewable tobacco products that are strongly linked to OSCC cases, constituting the major bulk of head and neck cancers in this country. More recently, HPV positivity has been reported among citizens of Karachi in the range of 17.9-19%. Studied subjects included those having a tobacco chewing habit with or without oral premalignant lesions (Baig et al., 2012; Rubab et al., 2013). Our findings add to the existing body of evidence which suggest use of tobacco with or without additives, in either smoked or smokeless form as a highly considerable risk factor for oral cancer (Jusawalla et al., 1971; Boyle et al., 1990; Vingneswaran et al., 1995). Oral cancer in technologically advanced countries like United States, Japan, United Kingdom and France, is mostly found to be associated with tobacco smoking with or without concomitant alcohol consumption. In Pakistan generally and in Karachi specifically, a variety of tobacco habits are common and they differ from region to region. A substantial proportion of individuals use SLT in forms like naswar, khaini, mishri, paan, gutka, paan-masala or smoke tobacco in forms like cigarettes, bidi, hooka etc. The use of SLT is rampant in our society and the most preferable site of tobacco

101 placement is the buccal groove situated between cheek mucosa and the alveolar ridge. This may partly explain why in Pakistan the buccal mucosa of cheek is the primary site for cancer development in contrast to the tongue and floor of the mouth in Western countries (Critchley et al., 2003). The magnitude of the problem not only remains unresolved but appears to be growing. The effects of various risk factors appear to be not only complementary but rather multiplicative. Historically, the high incidence of oral cancers in Indo-Pak subcontinent has long been associated with the habit of Paan (Quid) chewing, especially when tobacco is incorporated in it, and smoking being synergistic in its effects. Karachi is a metropolis city with a highly diversified population that has migrated not only from regions within the country but from many parts of South-east Asia. They exhibit diverse ethnic, economic and cultural characteristics. The relationship between ethnic origin, cultural practices and oral cancer further supported a tobacco- related etiology as the population most commonly affected was the one massively involved in tobacco chewing with several harmful additives as compared to rest of the citizens. Furthermore, use of SLT in various chewable forms is widely prevalent in our population alongwith a substantial number engaged in a mixed, chewing plus smoking, habit. Predominance of males in the results correlates well with large number of male tobacco users. Males are financially better-off in our society and have more access to tobacco products as compared to home-bound females. The involvement of intra-oral subsites, i.e., buccal mucosa of cheek, tongue, floor of the mouth and palate also corresponds to the same habit. Similarly, the buccal mucosa appears to be affected very often in females possibly connected with the role of tobacco chewing and betel quid habit. According to Western figures oral cancer is regarded as a disease mainly affecting males in their sixth to eight decades of life. In our study it presented at least two decades earlier in fourth to sixth decades with a definite male predominance. In males, we observed earlier start and more frequent use of tobacco-containing products and the trend of simultaneous use of SLT as well as smoking is more often seen in our male population. However, the number of females is higher in our study, is in sharp contrast to the western population (Keller, 1970), further pointing towards the significance of tobacco chewing and betel quid as popular habits among females in this part of the world.

102

Well established oral precancerous lesions like leukoplakia, erythroplakia and conditions like OSF were also seen in a certain group of people. Oral precancers are supposed to remain dormant for several years and if left untreated, about three to six percent become malignant over a period of ten years (World Health Organization, 1992). Similarly, in our series we have been able to observe these lesions in association with oral cancers, although in a small number of our cases, which further supports the evolution of a progressive disease in atleast some cancers. Although a major risk factor in the western world, alcohol was found to be a very uncommon event in our study group. The stigma attached with alcohol drinking declares it to be a social evil and hence under-reporting could be a possibility. Misinformation regarding the habit is a great possibility and its effect may be stronger than that observed in the society. There was an overwhelming majority of patients belonging to low socioeconomic class. This may reflect the social class distribution of the population of Karachi or may support the postulate that the nutritional deficiency states existing in the poor class might act as fertile soil for the concomitant carcinogens. In addition to this, tobacco containing products used by this class are quite substandard and the level of carcinogens may hence be quite high as compared to some similar but expensive products. If we consider ethnicity-wise distribution of our total studied cases, we observed that half of all collected OSCC cases and more than half of PCLs as well as tobacco-consuming control cases belonged to Urdu-speaking community. Other than this, the communities we encountered in our series included Sindhi-speaking, Balochi-speaking, Memoni- speaking, Pashto-speaking, Punjabi-speaking and others. Second highest number of cancer cases was seen in Sindhi-speaking ethnicity followed by equal numbers in Balochi-speaking and Memoni-speaking. While the frequency of PCLs after Urdu- speaking community was highest among Balochi-speaking followed by Sindhi-speaking and Memoni-speaking. All the above communities residing in Karachi have adopted the tobacco culture prevalent in the city, mostly chewable forms of tobacco. Sindhi-speaking community, however, we found more often addicted to betel nut chewing. Overall, the high frequency of oral lesions in ethnicities like Urdu-speaking, Balochi-speaking and

103

Memoni-speaking can readily be linked to the wide-spread consumption of various chewable products available in cheaper prizes within the city and are affordable to all. In contrast to the respiratory system, the role of occupational exposure has not been unequivocally demonstrated for cancer of the oral cavity (Elwood et al., 1984). The five year survival rate has not improved for this cancer beyond 45-50% which is quite unfortunate for such an accessible cancer (Bagan et al., 2008). TNM stage, histological grade and to some extent the intra-oral subsite of the lesion form the mainstay of predicting prognosis in these patients (Rubin, 1993; Sobin, 2003). WHO grading system proposes three grades of differentiation, i.e. well differentiated, moderately differentiated and poorly differentiated (Pindborg et al., 1997). The grades are based on subjective assessment of the degree of keratinisation, cellular and nuclear pleomorphism, and mitotic activity (Woolgar, 2006). Well-differentiated cancers have better prognosis (Scully et al., 2009). Oral cancer treatment facilities from Karachi have variably reported grade-I tumors as 59.53%, 60%, 25%, grade-II tumors as 32.55%, 36%, 55% and grade-III tumors as 7.9%, 4%,20%, respectively (Hameed et al., 2012, Akram et al., 2013, Zafar et al., 2015). In the current study we found grade II tumors as most rampant (59%) followed by grade I (36%). Only 5% of OSCC cases in our series were the most anaplastic grade III tumors. Early stage cancers exhibit a better outcome (Schroeff et al., 2009). Metastatic involvement of neck lymph nodes has been recognized as the most reliable single prognostic factor that defines the outcome of the oral cancer patient better than any other prognostic factor considered. Advanced-stage OSCC cases, i.e. stageIII/IV tumors, have been reported in proportions like 70%, 56.1% and 55.7% in different studies from Karachi city (Jafarey et al., 1976, Mirza et al., 1998, Zafar et al., 2015). In our study clinical stage III and IV make out 52% of all cases while in contrast to other studies stage II came out to be the largest group with 35% of all cases. We are of the opinion that stage II is over-expressed in our series and it is a great possibility that many of these stage II cases were in fact stage III or IV tumors. This may partly be explained by the lack of proper facilities required for correct clinical staging, especially reliable scanning reports, available to the treating oncologists or surgeons. In addition we also found deficiencies in proper documentation of the extent of disease at the time of surgery.

104

In the final analysis of clinico-pathological observations it was reaffirmed that chewing of tobacco with various additives like betel nut and smoking remains the high priority risk factor in the population of Karachi. Therefore, it was worthwhile to integrate the clinical data with molecular events and evaluate the role of tobacco metabolism alongwith the level of tobacco exposure of the oral mucosa in causing oral lesions. The importance of clinical stage and histologic grade in describing outcome was further emphasized in our study. CYP1A1 and GSTs have been the most implicated enzyme systems that are involved in tobacco metabolism and generation of related potential carcinogens. Although, a sizable number of studies have analyzed mutational polymorphisms in CYP1A1, GSTM1 and GSTT1 genes and their association with oral cancer but the results documented were quite variable (Zakiullahet al., 2015). Studies have also evaluated the risk of OSCC bestowed by these genes either singly as well as in various combinations (Ref. Xie et al., 2016; Zhang et al., 2011)). In the current study, we have tried to evaluate the relationship between tobacco-metabolizing enzyme gene polymorphisms and oral cancer risk and reinforce our observations by similar analysis in various precancerous lesions. All of these associations were done under the setting of various modes of tobacco exposure as an environmental trigger. In our opinion, this is probably the first study of its kind that has been conducted on the indigenous population of Karachi which exhibits a very high incidence of this cancer. Our results indicate that isolated polymorphisms in two of our tested genes, CYP1A1MspI and GSTM1, do not separately enhanced risk to OSCC. However, GSTT1 null polymorphism surfaced in our series as an independent risk factor, conferring a significant increased risk to oral cancer and this risk was not associated with tobacco exposure in any of its forms. Among all examined genotype variants, CYP1A1MspI homozygous genotype posed a significant risk in the presence of tobacco, i.e., in exclusive chewers. These results indicate that the magnitude of risk caused by tobacco is dependent upon the type of gene polymorphism inherited by an individual as well as on the nature and extent of tobacco exposure. The observed association with the type of tobacco used assumes significance in our population as large strata of tobacco consuming subjects are addicted to tobacco chewing habit.

105

The role of enzymes coded by CYP1A1 gene in the metabolism of tobacco has been extensively studied and proven (Bartsch et al., 2000). ―Tissue-specific expression of this isoform has been demonstrated in buccal cells of the oral cavity‖ (Vondracek et al., 2001). However, the association between genetic aberrations of CYP1A1 and oral cancer risk has been inconsistently reported. Meta-analysis of 4635 head and neck cancer cases and 5770 controls by Hashibe and co-workers, revealed an OR of 1.35 (95% CI, 0.95– 1.28) for carrying the CYP1A1 Val462 allele (Hashibe et al., 2003). There is an association between CYP1A1 Val462 allele and MspI variant as the former has been found to be in linkage disequilibrium with the CYP1A1 MspI variant allele in Japanese and Finnish populations (Hayashi et al., 1991, Hironen et al., 1992). Studies included in the meta-analysis except one have also shown similar association results for either variant of CYP1A1 gene. We have also found a similar OR of 1.44 (95% CI, 0.71-2.52) for the heterozygous variant of CYP1A1 but the result lacked statistical strength. However, for the homozygous variant an OR of 2.36 (95% CI, 1.0-6.20, p=0.05) was found and the result was statistically significant. A further increase in risk by homozygous variant was observed among exclusive tobacco chewers. When these individuals were stratified into above and below median exposure groups, the risk increased to several folds in the above median lifetime exposure category and the results were statistically significant. Hence a possible interaction with tobacco can be attributed. This observation becomes even more significant when we examine it in light of the fact that majority of OSCC cases (58%) as well as PCL cases (85%) in our series belonged to exclusive tobacco chewer category. In our series, the percentages of CYP1A1MspI wild-type, heterozygous and homozygous variants in control group was found to be 26.53%, 62.24% and 11.22% respectively, which is quite low for the wild type genotype and high for the polymorphisms when compared to the variants reported by Zakiullah from naswar-consuming KPK population of Pakistan, i.e., 63.3% for wild type and 36.4% for polymorphisms (Zakiullah et al., 2015). This may partly be explained by the different genetic inheritance of our study population, comprising mostly of Urdu-speaking emigrants from northern and central parts of India, in comparision to the KPK Pashto-speaking population. This fact also puts some light on the high prevalence of oral cancer in the Urdu-speaking ethnicity of Karachi mostly engaged in SLT consumption.

106

The reported frequencies of CYP1A1 homozygous variant allele ranged between 0-30.0% for OSCC cases and 0-10.5% for controls (Xie et al., 2016). In Japanese OSCC patients, the frequency of m2/m2 variant was reported to be 15% in cancers and 8% in controls (Tanimoto et al., 1999). Devasena and colleagues reported for the mixed-habit group ORs of 3.2 for OSCC cases and 4.29 for PCLs among Indians harboring the homozygous variant (Devasena et al., 2007). Another meta-analysis by Varela-Lema et al., reported for m2/m2 variant in oral and pharyngeal cancers ORs of 1.9 and 2.0, respectively (Varela-Lema et al., 2008). We reviewed seven articles which have evaluated the role of CYP1A1 polymorphism in the causation of oral cancer (Tanimoto et al., 1999; Sato et al., 1999; Sikdar et al., 2003; Gronau et al., 2003; Katoh et al., 1999; Matthias et al., 1998, Devasena et al., 2007). Out of the seven studies only three reported a risk association (Tanimoto et al., 1999; Sato et al., 1999; Devasena et al., 2007). This inconsistency warranted further studies on the subject. In the present study, we observed CYP1A1 homozygous variant in 18.57% of cancer cases and 11.2% of controls. This is in confirmation to the previously reported data (Xie et al., 2016). However, the heterozygous variant (m1/m2) of the same CYP1A1MspI gene polymorphism we found in as much as 62.85% of our OSCC cases and in 62.24% of our controls. Both OSCC cases and controls in the present study have higher proportions of m1/m2 variant than any of the ten studies included in the meta-analysis by Xie and colleagues (Xie et al., 2016). This may partly explain why our population has such a high prevalence of oral cancer especially in the presence of tobacco habit as the risk modulator, confirming the gene- environment equation. However, no statistically significant association was observed between oral cancer and CYP1A1 gene polymorphisms among exclusive smokers and mixed tobacco habitués (smoking plus chewing). There was a relatively small representation of tobacco smokers in our study population. This in turn could be a reflection of the overall trend of reduction in cigarette smoking due to higher cost of this product as compared to other chewable forms of tobacco. On comparing the frequency of CYP1A1 genotype variants between PCL and OSCC cases we found values for heterozygous variant more in PCLs (67%) while lower than cancers for the homozygous variant (16%). This suggests that PCLs may be more

107 common among carriers of heterozygous variant of this gene while the homozygous variant is more linked to OSCC. However, our observation can only be validated by studies done on much larger scales with large sample sizes than ours. GSTs are important for conjugating phase I-derived substrates. During the process of tumorigenesis, this reduced detoxification increases susceptibility to cancers (Parl, 2005). Studies conducted on cytogenetic end-points favor this hypothesis. Higher frequencies of sister chromatid exchange (SCE) and higher intensity of chromosomal aberrations (CA) have been seen in those smokers having GSTM1 null polymorphism (Norppa, 2004, Salama et al., 1999, Wienke et al., 1995). In cancers, there is a correlation between tumor clinical stage/ grade and GST levels (Chen et al., 1997; Patel et al., 2005). In cancers that are resistant to the effects of radiation and chemotherapy, GSTs have been found to play a role (Adler et al., 1999). They also regulate post-translational glutathionylation reactions (Townsend et al., 2005; Mcllwain et al., 2006). Finally, GSTM1 and GSTT1 null polymorphisms have been linked to prognosis in oral cancers (Geisler et al., 2005). The GSTM1 null polymorphism resulting from homozygous deletion causes functional loss of GSTM1 enzyme (Gronau et al., 2003). Studies have pointed towards its role in the genesis of several cancers, including lung cancer, breast cancer and urinary bladder cancer (Carlsten et al., 2008, Shi et al., 2008, London et al. 2000, Yu et al., 2009, Gracia- Closas et al., 2005). Individuals harboring this polymorphism and are also engaged in smoking have been shown to have higher SCE and CA levels as compared to those smokers that are positive for this gene (Norppa, 2004). Several of the above studies have implicated GSTM1 null polymorphism as an independent risk factor for cancer. However, GSTM1 analysis in our study population did not appear to be in agreement with such observations. An increased risk to head and neck cancers, including oral cancer, in the presence of GSTM1 null polymorphism has been documented in the meta-analysis with an OR of 1.23 (95% CI, 1.06-1.62) (Hashibe et al., 2003). Out of six studies that were conducted later on the same association, three reported an increased risk (Drummond et al., 2004; Deng et al., 2004, Devasena et al., 2007), two documented no association (Evans et al., 2004; Geisler et al., 2005) and one research found a protective effect (Xie et al., 2004). In the study from India by Devasena and colleagues GSTM1 null polymorphism surfaced as

108 a risk factor in individuals having moderate to high consumption of smokeless tobacco. This observation is of special significance as ―an increase in oral cancer risk with increase in lifetime exposure to tobacco has previously been demonstrated‖ (Devasena et al., 2007). Zakiullah and co-researchers, documented in the Pashto-speaking population of Pakistan an increased risk conferred by GSTM1 null polymorphism with an adjusted OR of 3.019, 95%CI, 1.861-4.898 (Zakiullah et al., 2015). In our series, the frequency of GSTM1 null genotype among control subjects was 31.63% which is almost half to what has been reported by Zakiullah and co-workers, i.e., 57% from the KPK province of Pakistan (Zakiullah et al., 2015). The values for GSTM1 null genotypes are within the range of 17-38% as reported from India (Nair et al., 1999; Buch et al., 2002; Sikdar et al., 2004; Naveen, 2004; Sreelekha et al., 2001; Srivastava et al., 2005). We further observed overall ORs of 1.22 (95%CI, 0.57-2.60) and 0.95 (95% CI, 0.51-2.22) in pre-cancers and cancers, respectively, for this polymorphism but the results lacked statistical power. Hence we can say that we do not found any statistically significant relationship between GSTM1 null polymorphism and OSCC risk not only in our overall cancer cases but also neither of tobacco habit groups. However, in our PCL cases among tobacco chewers having less than median lifetime exposure a statistically significant OR value of 2.94 (95%CI, 1.0-8.68, p=0.04) has been observed. From these observations we can state that in cases of below median lifetime exposure of chewable tobacco, it is the precancerous lesion that may evolve first in individuals carrying GSTM1 null genotype within their oral cavity and this may progress to develop invasive OSCC later. In the meta-analysis by Hashibi and co-researchers, in total 27 studies have documented the risk created by GSTT1 null allele to oral cancer. Out of these 21 have documented an overall OR of 1.17 (95% CI, 0.98–1.4) (Hashibi et al., 2003). Out of seven studies conducted later, three found a positive association (Drummond et al., 2005; Deng et al., 2004; Barroso-Duarte et al., 2006), two found no association (Sikdar et al., 2004; Xie et al., 2004) and two studies declared GSTT1 null genotype a protective polymorphism for head and neck cancers (Evans et al., 2004; Devasena et al., 2007). Of the three studies which showed positive association of GSTT1 null genotype with cancer risk, the one done by Drummond et al., documented this association in cancers of floor of mouth

109

(Drummond et al., 2005). The second by Deng et al., showed this association in nasopharyngeal carcinoma cases (Deng et al., 2004). The last of the three studies was conducted on oral leukoplakia cases in smokers by Barroso-Duarte and his colleagues and it demonstrated a significant association of leukoplakia cases with GSTT1 null polymorphism (Barroso- Duarte et al., 2006). In the present series, the percentage of GSTT1 null polymorphism among control subjects was 2.04% which is remarkably lower than what has been reported by Zakiullah and co- workers, i.e., 23.2% in the subjects studied in the KPK province of Pakistan (Zakiullah et al., 2015). This value is even lower than the ranges (8-27%) reported from India (Nair et al., 1999; Buch et al., 2002; Sikdar et al., 2004; Naveen 2004; Sreelekha et al., 2001; Srivastava et al., 2005). The KPK study showed the distribution of GSTT1 null polymorphism in higher proportions in its OSCC cases (47.5%) as compared to controls (23.2%). The same study further demonstrated, in the naswar chewing Pashto-speaking population of Pakistan, an increased risk conferred by GSTT1 null polymorphism as an adjusted OR of 3.011 (95%CI, 1.865-4.862) (Zakiullah et al., 2015). In contrast to this, the study from north- east India interestingly showed that smoking conferred 68% reduced risk to lung cancer in those having GSTT1 null genotype as compared to those having two copy number of this gene (Ihsan et al., 2014). More or less similar result has previously been documented in the Indian study by Devasena and colleagues who observed a 0.57 times reduced risk among users of smokeless tobacco, i.e., tobacco chewers (Devasena et al., 2007). The GSTT11 null polymorphism in the present study was much higher in our OSCC cases at 12.14% as compared to controls at 2.04%, while in pre-cancers we found this genotype to be present in 9% of cases. Moreover, this genotype variant appeared as an independent risk factor in our study subjects, conferring an OR of 6.63 (95% CI, 1.49-29.4, p=0.01) to oral squamous cell carcioma. The results were statistically significant. This OR is almost double of what has been reported by Zakiullah and colleagues among tobacco chewers from the KPK province (Zakiullah et al., 2015). In other tobacco exposure categories, either we could not found any significant association with GSTT1 null variant allele or there was lack of representation of null genotype among these tobacco habitués in our series. This could be again because of reduction in the use of smoked tobacco products

110 we observed in our study population which in turn could be due to the high cost of cigarettes when compared to cheaper chewable tobacco alternatives available in the city. There is paucity of studies that have evaluated the effects of aforementioned gene combinations. To date, relatively fewer researchers have analyzed the risk posed by simultaneous occurrence of two or more of the tested CYP1A1 MspI, GSTM1 null and GSTT1 null mutations (Sato et al., 1999; Sikdar et al., 2004; Devasena et al., 2007). There were studies conducted previously that analyzed the effects of combination of CYP1A1 and GST genes polymorphisms and compared them with the results when the CYP1A1 polymorphism existed alone. Bartsch and co-researchers analyzed such studies for the period between 1990 and 2000. They reported ORs ranging between 2 and 10 for common CYP1A1 polymorphisms. The same analysis suggested that some CYP1A1 polymorphisms when combined with GSTM1 null genotype increase cancer risk among smokers in atleast some cancers like those of lung, esophagus and oral cavity (Bartsch et al., 2000). Devasena and colleagues reported the presence of either CYP1A1 MspI or GSTM1 null variant enhanced chances of oral cancer by 1.6-fold while the same genes in combination increased the risk by 7.42 fold in the above median tobacco chewer category (Devasena et al., 2007). In the present study gene interactions also displayed a pattern of risk association which is quite different from single gene effects. Simultaneous presence of CYP1A1 homozygous variant and GSTM1 null polymorphism in our series increased oral cancer risk among overall chewer category, OR 12.8 (95%CI, 1.20-135.5, P=0.03). This increase in risk is almost twice the risk we found for CYP1A1 MspI homozygous variant alone in tobacco chewers, i.e., OR 7.2 (95% CI, 1.8-27.5, P=0.003). Therefore, OSCC risk in the presence of smokeless tobacco has significantly increased by this genotype combination. This finding of ours support the hypothesis that ―combined gene variants can lead to increased risk by acting synergistically‖ and the same has been validated by similar studies from other areas of the world (Tanimoto et al., 1999; Sato et al., 1999). In the current study, GSTT1 null polymorphism surfaced as an independent risk factor for oral cancer irrespective of any other examined genotype or tobacco exposure. When GSTT1 null polymorphism combined with GSTM1 not null variant, we observed an overall increased risk (OR 4.58, 95% CI, 0.99-21.2, p=0.05). This observation in fact reflects the contribution for increased risk by GSTT1 null genotype.

111

This finding of ours differs from reports that have documented the presence of GSTT1 null genotype as a protection against oral cancer in Indian population (Devasena et al., 2007). For precancerous lesions, it was postulated that ―certain genotype combinations can increase risk by acting synergistically‖. Devasena and his colleagues tested this hypothesis in different combinations of CYP1A1 MspI alongwith either null or not null variants of GSTM1 and GSTT1. They found a protective effect in one of the combinations, i.e., when GSTM1was present while GSTT1 was absent. In exclusive chewers PCL risk was reduced to 0.26 fold (Devasena et al., 2007). In our study we also tested the same hypothesis while analyzing different categories and lifetime exposures of tobacco. Contrary to the Indian study, we found an overall increased risk conferred by various combinations of the three tested genes. However, none of our results were statistically significant. In our series, we came across distinct susceptibility patterns of oral cancer for two common intra-oral sub-sites i.e., cheek mucosa and tongue. Previously genetic dissimilarities have been reported for risk to cancers at various sub-sites within the oral cavity (Pradhan, 1991; Tanimoto et al., 1999). These could be the result of ―altered molecular signature patterns in the process of carcinogenesis in these regions‖ (Huang et al., 2002). Dissimilarities also exist in the expression levels of phase I and phase II enzymes within the oral cavity which may contribute towards altered risk-profiles of these sub-sites (Pelkonen et al., 1997, Vondracek et al., 2001). The pattern of gene-tobacco interactions we observed previously in our overall oral cancer cases was simulated in the case of buccal mucosa cancers, i.e. CYP1A1 MspI m2/m2, GSTT1 null and CYP1A1 m2/m2 plus GSTM1 null. None was replicated for the tongue cancers, except the one for GSTT1 null genotype which increased the risk to tongue cancer (OR: 10.2, 95%CI-1.57-67.0, p=0.01) even more than cheek mucosa (OR: 7.05, 95%CI-1.49-33.2, P=0.01) and the results we found were statistically significant. The other significant difference we observed in case of CYP1A1 m2/m2 (homozygous) variant which showed an OR of 3.15 (95% CI, 1.12-8.80, p=0.028) specifically to cancer of cheek mucosa. A further increase in risk, upto several times, was observed among exclusive chewers and when these were stratified into above and below median lifetime

112 exposures the risk conferred showed an OR of as high as 60 (95% CI, 3.13-1146, p=0.006) in the above median exposure category. No association was observed in case of GSTM1 null genotype at both of the common sites. Finally, simultaneous presence of two polymorphisms, i.e. CYP1A1 MspI m2/m2 and GSTM1 null, conferred several fold increased risk among chewers to cancer of cheek mucosa, OR 32 (95% CI, 1.56-655, p=0.02). The same gene combination increased the risk of tongue cancer also, i.e. an OR of 8 (95% CI, 0.25-225, p=0.24) but the association lacked statistical power. Our results are more or less in agreement to what has been reported by Devasen and colleagues among Indians. It is also important to mention here that this increased genetic susceptibility of buccal mucosa may be compounded by the habit of placing gutka between alveolar ridge and cheek for long periods in our study subjects, explaining in part the much higher incidence of OSCC at this site in Karachi. In our final analysis of molecular events, we submit here our observations regarding ethnic distribution of the three genotypes we tested in the current study. Our‘s is the very first study that has evaluated genotype distribution of tobacco metabolizing-enzymes among various ethnicities of Karachi affected by tobacco-related oral lesions. Zakiullah and co-workers reported among tobacco-addicted Pashto-speaking ethnicity of KPK province the CYP1A1 rs4646903 polymorphism in 38% of cancer cases and 36.4% of controls. They found the wild type variant of this gene in 62% of OSCC cases and 63.6% of controls (Zakiullah et al., 2015). We found in our study, CYP1A1MspI m1/m2 and m2/m2 polymorphisms in 85.7% of OSCC cases and 66.6% of control subjects belonging to the Pashto-speaking community. Hence we observed a very high proportion of these mutations in Pashto-speaking population of Karachi, both cases and controls alike, although the total number of Pashto-speaking cases was only 25 inclusive of all three categories, i.e., PCLs, OSCC and controls. Among Urdu-speaking community, a significantly higher percentage of CYP1A1 MspI homozygous variant we observed in our pre-cancers and cancers as compared to controls. Heterozygous variant we found as equally distributed in all three categories of study subjects. This fact points towards a very significant finding that it is the homozygous variant of CYP1A1 MspI polymorphism that may be an important contributing factor towards high proportions of oral cancer in the tobacco-chewing Urdu-speaking community of Karachi. A somewhat similar pattern

113 has been observed in the Balochi-speaking ethnicity of the city. In our series, the heterozygous variant of CYP1A1 MspI gene was found to be more prevalent among PCL cases in ethnic entities like Balochi-speaking, Sindhi-speaking, Memoni-speaking and Pashto-speaking. Percentage-wise this heterozygous genotype has been observed to be associated with precancerous lesions in these ethnicities even more than that seen in Urdu-speaking community. The GSTM1 null polymorphism has been reported among Pashto-speaking ethnicity in as much as 79.5% of oral cancers (Zakiullah et al). The same null genotype has been represented in lung cancer cases from north-east Indian population in 37% of cases (Ihsan et al., 2014). Different studies around the world have shown a variable association of oral cancer with GSTM1 null genotype as decreased risk, no risk, moderate risk to a 3- fold increased risk (Deng et al., 2004; Evans et al., 2004; Geisler et al., 2005; Xie et al., 2004; Buch et al., 2002). In our study, we found a weak association of this genotype with only PCLs in the oral cavity. Overall no association has been observed with oral cancer. When considering all studied ethnicities, only Urdu-speaking community has a higher proportion of GSTM1 null variant in PCL and OSCC cases as compared to controls. From this we can infer that GSTM1 null polymorphism is not a significant contributing genetic aberration for oral cancer risk in tobacco users among any of the ethnic entities living in Karachi. As far as GSTT1 null genotype is concerned, an increased prevalence of this gene variant has been reported in oral cancer cases from tobacco-addicted Pashto-speaking ethnicity (Zakiullah et al, 2015). In the present study, we found this genotype in PCL cases among Pashto-speaking while none of our OSCC cases belonging to this ethnic group showed this polymorphism. Among Urdu-speaking community a significantly high frequency was seen in oral cancers as compared to PCLs and controls. Others ethnicities like Balochi-speaking and Memoni-speaking showed increased frequency of this polymorphism in OSCC cases while in Sindhi-speaking we found it to be present only in PCL cases. We found GSTT1 null polymorphism to be significantly associated with oral cancer in the Urdu-speaking community of Karachi especially among individuals having more than median lifetime exposure to chewable tobacco.

114

5. CONCLUSIONS

 In our study subjects CYP1A1 MspI and GSTM1 null polymorphisms do not independently confer risk to oral cancer. The homozygous variant genotype of CYP1A1MspI enhanced oral cancer risk only when there is a history of tobacco use, i.e., in exclusive tobacco chewers, confirming the gene-environment equation.

 GSTT1 null polymorphism is an independent risk factor to OSCC in the absence of an association with tobacco consumption.

 Different combinations of CYP1A1, GSTM1 and GSTT1 null gene variants revealed a pattern of vulnerability quite dissimilar to that of an individual gene. The concomitant occurrence of CYP1A1 MspI homozygous variant and GSTM1 null polymorphism amplified oral cancer risk by several fold.

 There was increased predisposition to oral cancer related to tobacco metabolizing gene polymorphisms at the cheek sub-site as compared to the tongue.

 Our results indicate that ethnicity and the nature of tobacco exposure along with the distribution of XME genes can be major determinants of oral cancer risk.

 Finally, the importance of clinical stage and histologic grade in describing outcome of oral cancer was further emphasized in our study.

6. RECOMMONDATIONS  Ban on the sale, manufacture, storage, advertising, sponsorships and promotion of all tobacco based chewable products.  Large-scale educational interventions. Intervention can be the greatest tool if the target population is identified along with its tool.  Marketing candies with similar flavor or taste can be an alternative attraction for habitual teenagers.  Awareness campaigns through media and educational institutes regarding information on oral hygiene, risk factors, symptoms and the importance of seeking early professional help.

115

 Population screening would reduce the mortality associated with oral cancer but requires carefull planning, new research projects and monitory support.  Participation of family physicians and medical students for early detection of oral cancer is the need of time. 7. LIMITATIONS OF THE STUDY  Gene sequencing could have been employed for finding out exact arrangement of nucleotides in the target DNA and hence identified special set of genes responsible for the increased risk observed in our study subjects. However, it was not attempted because of financial limitations as well as time constraint.  Histological examination of the tissue remains the gold standard for diagnosis and identification of early malignant changes. As it is an invasive technique with surgical implications, it is not routinely recommended in oral pre-cancerous lesions until there is a very high level of suspicion on visual examination and was not included in our study protocol.

116

8. BIBLIOGRAPHY  Abbas, M., Srivastava, K., Imran, M., & Banerjee, M. (2014). Association of CYP1A1 gene variants rs4646903 (T>C) and rs1048943 (A>G) with cervical cancer in North Indian population. Eur J Obstet Gynecol Reprod Biol, 176, 68-74.  Abraho, A.C., Bonelli, B.V., Nunes, F.D., Dias, E.P., & Cabral, M.G. (2011). Immunohistochemical expression of p53, p16 and TERT in oral squamous cell carcinoma and potentially malignant disorders. Brazillian Oral Res, 25(1), 34-41.  Ada, A.O., Kunak, S.C., Hancer, F., Soydas, E., Alpar, S., Gulhan, M., & Iscan, M.l. (2012). Association between GSTM1, GSTT1 and GSTP1 polymorphisms and lung cancer risk in a Turkish population. Mol Biol Rep, 39(5), 5985-5993.  Ada, A.O., Yilmazer, M., Suzen, S., Demiroglu, C., Demirbag, A.E., Efe, S., et al. (2007). Cytochrome P450 (SYP) and glutathione S-transferases (SGT) polymorphisms (CYP1A1, CYP1B1, GSTM1, GSTP1, and GSTT1) and urinary levels of 1- hydroxypyrene in Turkish coke oven workers. Genetic and Molecular Biology, 30(3), 511-519.  Adler, V., Yin, Z., Fuchs, S.Y., Benezra, M., Rosario, L., Tew, K.D., et al. (1999). Regulation of JNK signaling by GSTp. Embo J, 18(5), 1321-1334.  Ahmed, F., & Islam, K.M. (1990). Site predilection of oral cancer and its correlation with chewing and smoking habit – a study of 130 cases. Bangladesh Med Res Counc Bull, 16(1), 17-25.  Akhlaq, H., Jamal, Q., & Ismail, M.O. (2014). Risk factors associated with decreased mouth opening among patients with oral submucous fibrosis (OSMF). Annals Abbasi Shaheed Hospital & Karachi Medical & Dental College, 19(1), 3-9.  Akhter, M., Hossain, S., Rahman, Q.B., & Molla, M.R. (2011). A study on histological grading of oral squamous cell carcinoma and its co-relationship with regional metastasis. J Oral Maxillofac Pathol, 5(2), 168-176.  Akram, S., Mirza, T., Mirza, M.A., & Qureshi, M. (2013). Emerging patterns in clinico- pathological spectrum of oral cancers. Pak J Med Sci, 29(3), 783-787.  Alvi, A., Myers, E.N., & Johnson, J.T. (1996). Cancers of the oral cavity. In: Myers EN, Suen JY, eds. Cancer of the head and neck. Philadelphia: WB Saunders. p. 321-360.

117

 Amelia, A.S., Maria, T.R., Carlos, R., Jose, I.A., Marian, M.P., Josef, A.C., et al. (2009). P53 mutation is rare in oral mucosa brushings from patients previously treated for a head and neck squamous cell carcinoma. Oral Oncol, 45, 661-664.  American Cancer Society. (2016). Cancer Facts & Figures. Atlanta: American Cancer Society.  Amos, A., Greaves, L., Nichter, M., & Bloch, M. (2011). Women and tobacco: a call for including gender in tobacco control research, policy and practice. Tob Control, 21, 236- 243.  Angadi, P.V., & Rekha, K. (2011). Oral submucous fibrosis: a clinicopathologic review in 205 cases in Indians. Oral and Maxillofacial Surgery, 15(1), 15-19.  Antunes, J.L.F., Biazevic, M.G.H., de Araújo, M.E., Tomita, N.E., Chinellato, N.E., & Narvai, P.C. (2001). Trends and spatial distribution of oral cancer mortality in São Paulo, Brazil, 1980-1998. Oral Oncol, 37, 345-350.  Ara, N., Atique, M., Ahmed, S., & Bukhari, S.G.A. (2014). Frequency of p53 gene mutation and protein expression in oral squamous cell carcinoma. J Coll Physic & Surg Pak, 24(10), 749-753.  Ariyawardana, A. & Johnson, N.W. (2013). Trends of lip, oral cavity and oropharyngeal cancers in Australia 1982-2008: overall good news but with rising rates in the oropharynx. BMC Cancer, 13, 333.  Ashfaq, K., Ashfaq, M., Ahmed, A., Khan, M., & Azhar, M. (2014). Submandibular Gland Involvement in Early Stage Oral Cavity Carcinomas: Can the Gland be left behind? J Coll Physic & Surg Pak, 24(8), 565-568.  Auluck, A., Hislop, G., Poh, C., Zhang, L., & Rosin, M.P. (2009). Areca nut and betel quid chewing among South Asian immigrants to Western countries and its implications for oral cancer screening. Rural Remote Health, 9(2), 1118-1127.  Aust S.D., Chignell C.F., Bray T.M., Kalyanaraman, B., & Mason, R.P. (1993). Free radicals in toxicology. Toxicol Appl Pharmacol, 120, 168–178.  Badar, F. & Mahmood, S. (2015). Hospital-Based Cancer Profile at the Shaukat Khanum Memorial Cancer Hospital and Research Centre, Lahore, Pakistan. J Coll Physic & Surg Pak, 25(4), 259-263.

118

 Bagan, J.V. & Scully, C. (2008). Recent advances in Oral Oncology 2007: epidemiology, aetiopathogenesis, diagnosis and prognostication. Oral Oncol, 44(2), 103-108.  Baig, S., Lucky, M.H., Qamar, A., Ahmed, F., Khan, S., Ahmed, W., et al. (2012). Human papilloma virus and oral lesions in Gutka eating subjects in Karachi. J Coll Physic & Surg Pak, 22(3), 135-138.  Baker, S.R. (1993). Malignant neoplasms of the oral cavity. St. Louis: Mosby. p1248- 1305.  Banoczy, J. & Squier, C. (2004). Smoking and disease. Eur J Dent Educ, 4, 8(Suppl 4), 7-10.  Baral, R., Patnaik, S., & Das, B.R. (1998). Co-overexpression of p53 and c-myc proteins linked with advanced stages of betel- and tobacco-related oral squamous cell carcinoma from Eastern India. Eur J Oral Sci, 106(5), 907-913.  Barnes, L., Eveson, J., Reichart, P., & Sidransky, D. (2005). Pathology and Genetics of Head and Neck Tumors. Lyon: IARC Press.  Barroso Duarte, E.C., Da Silva, M.S., Gomez, M.V., & Gomez, R.S. (2006). GSTT1 polymorphism and oral leukoplakia. Anticancer Res, 26, 427–430.  Bartsch, H., Nair, U., Risch, A., Rojas, M., Wikman, H., & Alexandrov, K. (2000). Genetic Polymorphism of CYP Genes, Alone or in Combination, as a Risk Modifier of Tobacco-related Cancers. Can Epid Bio & Prev, 9, 3-28.  Basset, M.T., Cokunonga, E., Mauchaza, B.G., Thomas, D.B., & Parkin, D.M. (1995). Cancer in African population of Harare, Zimbabwe, 1990-1992. Int J Cancer, 63, 29-36.  Bernardes, V.F., Gleber-Netto, F.O., Sousa, S.F., Silva, T.A., & Aguiar, M. (2010). Clinical significance of EGFR, Her-2 and EGF in oral squamous cell carcinoma: a case control study. J Exp Clin Cancer Res, 29(1), 40.  Bertucci, F., Viens, P., Tagett, R., Nguyen, C., Houlgatte, R., Birnbaum, D. (2003). DNA arrays in clinical oncology: promises and challenges. Lab Invest, 83: 305-316.  Bhawana, G. (2013). Burden of smoked and smokeless tobacco consumption in India— results from the global adult tobacco survey India (GATS-India)- 2009–2010. Asian Pac J Can Prev, 14(5), 3323–3329.

119

 Bhide, S.V., Nair, U.J., Nair, J., Spiegelhalder, B., & Preussmann, R. (1986). N- nitroamines in the saliva of tobacco chewers or masher users. Food Chem Toxicol, 24, 293-297.  Bhurgri, Y., Bhurgri, A., Hussainy, A.S., Usman, A., Faridi, N., Malik, J., et al. (2003). Cancer of the oral cavity and pharynx in Karachi- Identification of potential risk factors. Asian Pac J Cancer Prev, 4(2), 125-130.  Bhurgri, Y., Bhurgri, A., Nishter, S., Ahmed, A., Usman, A., Pervaiz, S., et al. (2006). Pakistan – Country Profile of Cancer and Cancer Control 1995-2004. J Pak Med Assoc, 56(3), 124-130.  Bhurgri, Y., Bhurgri, A., Usman, A., Pervez, S., Kyani, N., Bashir, I., et al. (2006). Epidemiological review of head and neck cancers in Karachi. Asian Pac J Cancer Prev, 7(2), 195-200.  Bhurgri, Y., Rahim, A., Bhutto, K., Bhurgri, A., Pinjani, P.K., Usman, A., et al. (1998). Incidence of carcinoma of oral cavity in Karachi- district south. J Pak Med Assoc, 48(11), 321-325.  Bhurgri, Y. (2007). Age standardized incidence of oral cancer rate per 100,000 in males of Karachi South. Unpublished data, Karachi cancer registry, Karachi.  Bhurgri, Y. (2005). Cancer of oral cavity – trends in Karachi South (1995-2002). Asian Pac J Cancer Prev, 6(1), 22-26.  Boffetta, P., Hecht, S., Gray, N., Gupta, P., & Straif, K. (2008). Smokeless tobacco and cancer. Lancet Oncology, 9(7), 667-675.  Boffetta, P., Mashberg, A., & Winkelmann, R. (1992). Carcinogenic effect of tobacco smoking and alcohol drinking on anatomic sites of the oral cavity and oropharynx. Int J Cancer, 52, 530-533.  Boyle, P. & Ferlay, J. (2005). Cancer incidence and mortality in Europe. Ann Oncol, 16(3), 481-488.  Boyle, P., Zheng, T., Macfarlane, G.J., McGinn, R., Maisonneuve, P., LaVecchia, C., et al. (1990). Recent advances in the aetiology and epidemiology of head and neck cancer. Curr Opin Oncol, 2, 539-545.  Braakhuis, B.J.M., Snijders, P.J.F., Keune, W.J.H., Meijer, C.J.L.M., Ruijter-Schippers, H.J., Leemans, C.R., et al. (2004). Genetic patterns in head and neck cancers that contain

120

or lack transcriptionally active human papilloma virus. J Nat Cancer Inst, 96(13), 998- 1006.  Bradley, K.T., Budnick, S.D., & Logani, S. (2006). Immunohistochemical detection of p16INK4a in dysplastic lesions of oral cavity. Modern Pathology, 19(10), 1310-1316.  Buch, S.C., Notani, P.N., & Bhisey, R.A. (2002). Polymorphism at GSTM1, GSTM3 and GSTT1 gene loci and susceptibility to oral cancer in an Indian population. Carcinogenesis, 23, 803–807.  Calsten, C., Sagoo, G.S., Frodsham, A.J., Burke, W., & Higgins, J.P. (2008). Glutathione S-transferase M1 (GSTM1) polymorphisms and lung cancer: a literature-based systemic HuGE review and meta-analysis. Am J Epidemiol, 167, 759-774.  Campisi, G. & Giovannelli, L. (2009). Controversies surrounding human papilloma virus infection, head and neck vs. oral cancer, implications for prophylaxis and treatment. Head Neck Oncol, 1, 8-10.  Campo-Trapero, J., Cano-Sánchez, J., Palacios-Sánchez, B., Sanchez-Gutierrez, J., Gonzalez-Moles, M., & Bascones-Martinez, A. (2008). Update on molecular pathology in oral cancer and precancer. Anticancer research, 28(2B), 1197-1205.  Carlsten, C., Sagoo, G.S., Frodsham, A.J., Burki, W., & Hijjins, J.P. (2008). Glutathione S-transferase M1 (GSTM1) polymorphisms and lung cancer: a literature-based systematic HuGE review and meta-analysis. Am J Epidemiol, 167(7), 759-774.  Cascorbi, I., Brockmoller, J., & Roots, I. (1996). A C4887A polymorphism in exon 7 of human CYP1A1: population frequency, mutation linkages and impact on lung cancer susceptibility. Cancer Res, 1996, 56(21), 4965–4969.  Chen, C., Méndez, E., Houck, J., Fan, W., Lohavanichbutr, P., Doody, D., et al. (2008). Gene expression profiling identifies genes predictive of oral squamous cell carcinoma. Cancer Epidemiology biomarkers & Prevention, 17(8), 2152-2162.  Chen, J.K., Katz, R.V., & Krutchkoff, D.J. (1990). Intraoral squamous cell carcinoma: epidemiologic patterns in Connecticut from 1935 to 1985. Cancer, 66, 1288-1296.  Chen, Y.J., Chang, J.T.C., Liao, C.T., Wang, H.M., Yen, T.C., Chiu, C.C., et al. (2008). Head and neck cancer in the betel chewing area: recent advances in molecular carcinogenesis. Cancer science, 99(8), 1507-1514.

121

 Chen, Y.K. & Lin, L.M. (1997). Evaluation of glutathione S-transferase activity in human buccal epithelial dysplasias and squamous cell carcinomas. Int J Oral Maxillofac Surg, 26, 205-209.  Cheng, T.J., Christiani, D.C., Wiencke, J.K., Wain, J.C., Xu, X., & Kelsey, K.T. (1995). Comparison of sister chromatid exchange frequency in peripheral lymphocytes in lung cancer cases and controls. Mutat Res, 348, 75-82.  Chiba, I., Muthumala, M., Yamazaki, Y., Uz Zaman, A., Lizuka, T., Amemiya, A., et al. (1998). Characteristics of mutations in the p53 gene of oral squamous-cell carcinomas associated with betel-quid chewing in Sri Lanka. Int J Cancer, 77(6), 839-842.  Chidzonga, M.M. & Mahoma, L. (2006). Squamous cell carcinoma of the oral cavity, maxillary antrum and lip in a Zimbabwean population: A descriptive epidemiological study. Oral Oncology, 42, 184-189.  Choi, S. & Myers, J. (2008). Molecular pathogenesis of oral squamous cell carcinoma: implications for therapy. J Dent Res, 87(1), 14-32.  Cia: The World Fact book 2005. Pakistan. Retrieved June 2005, from http://www.odei.gov. cia/publications/ facebook/geos/pk.html  Consolidated report of population based cancer registries 2001-2004. National Cancer Registry Programme: Accessed April 29, 2016, from, http://www.icmr.nic.in/ncrp/cancer_reg.htm  Conway, D., Macpherson, L.M.D., Warnakulasuriya, K.A.A.S., Ogden, G., & Stockton, D.L. (2006). Incidence of oral and oropharyngeal cancer in the United Kingdom (1990- 1999) – current status, recent trends and intercountry comparisons. Oral Oncol, 42, 586- 592.  Crissman, J.D., Kessis, T., Shah, K.V., et al. (1988). Squamous papillary neoplasia of the adult upper aerodigestive tract. Hum Pathol, 19, 1387.  Critchley, J.A. & Unal, B. (2003). Health effects associated with smokeless tobacco: a systemic review. Thorax, 58, 435-443.  Cruz, I., Napier, S., Van Der Waal, I., Snijders, P., Walboomers, J., Lamey, P., et al. (2002). Suprabasal p53 immunoexpression is strongly associated with high grade dysplasia and risk for malignant transformation in potentially malignant oral lesions from Northern Ireland. J Clin Patho, 55(2), 98-104.

122

 Daniel, W.W. (1987). Biostatistics: a foundation for analysis in Health Sciences, 5th edition, John Wiley & Sons, ISBN 0-471-52514-6, p.157.  Das, C.M., Ghori, A., Khursheed, F., Zaheen, Z., & Sharma, M. (2012). Frequency, knowledge and practice of tobacco smoking in pregnant women at a out patient department of a tertiary care hospital. J Liaq Univ of Med & Health Sci, 11(3), 176-179.  Das, S.N., Khare, P., Singh, M.K., & Sharma, S.C. (2011). Correlation of Cyclin D1 expression with arrgessive DNA pattern in patients with tobacco-related intraoral squamous cell carcinoma. Ind J Med Res, 133(4), 381.  Dckant, W. (2009). The role of biotransformation and bioactivation in toxicity. EXS, 99, 57-86.  De Boer, M.F., Sanderson, R.J., Damhuis, R.A., Meeuwis, C.A., & Knegt, P.P. (1997). The effects of alcohol and smoking upon the age, anatomic sites and stage in the development of cancer of the oral cavity and oropharynx in females in the South west Netherlands. Eur Arch Otorhinolaryngol Suppi, 544, 177-179.  Del Mistro, A. & Chieco Bianchi, L. (2001). HPV related neoplasias in HIV infected individuals. Eur J Cancer, 37, 1227-1235.  Demian, S.D.E., Bushkin, F.L., & Echevarria, R.A. (1973). Perineural invasion and anaplastic transformation of verrucous carcinoma. Cancer, 32, 395-401.  Deng, Z.L., Wei, Y.P., & Ma, Y. (2004). Frequent genetic deletion of detoxifying enzyme GSTM1 and GSTT1 genes in nasopharyngeal carcinoma patients in Guangxi Province, China. Zhonghua Zhong Liu Za Zhi, 26, 598–600.  Devasena, A., Chaubal, P.C., Kannan, S., Bhisey, R.A., & Mahimkar, M.B. (2007). Susceptibility to oral cancer by genetic polymorphisms at CYP1A1, GSTM1 and GSTT1 loci among Indians: Tobacco exposure as a risk modulator. Carcinogenesis, 28(7), 1455- 1462.  Drummond, S.N., De Marco, L., Noronha, J.C., & Gomes, R.S. (2004). GSTM1 polymorphism and oral squamous cell carcinoma. Oral Oncol, 40, 52–55.  Drummond, S.N., Gomez, R.S., Noronha, J.C.M., Pordeus, I.A., Barbosa, A.A., & De Marco, L. (2005). Association between GSTT-1 gene deletion and the susceptibility to oral squamous cell carcinoma in cigarette-smoking subjects. Oral Oncol, 41, 515–519.

123

 Dyson, N., Howley, P.M., Munger, K., & Harlow, E. (1989). The human papilloma virus 16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science, 243(4893), 934-937.  Dzian, A., Halasova, E., Matakova, T., Kavcova, E., Smolar, M., Dobrota, D., et al. (2012). Lung adenocarcinoma and squamous cell carcinoma in association with genetic polymorphisms of GSTs in Slovak population. Neoplasma, 59(2), 160-167.  Edge, S., Byrd, D., Compton, C., Fritz, A., Greene, F., & Trotti, A. (2010). AJCC Cancer Staging Manual (American Joint Committee on Cancer). New York: Springer.  Elwood, J.M, Pearson, J.C.G., & Jackson, D.H. (1984). Alcohol, smoking, social and occupational factors in the aetiology of cancer of the oral cavity, pharynx and larynx. Int J Cancer, 34(5), 603-612.  Evans, A.J., Henner, W.D., Eilers, K.M., Montalto, M.A., Wersinger, E.M., Anderson, P.E., et al. (2004). Polymorphisms of GSTT1 and related genes in head and neck cancer risk. Head Neck, 26, 63–70.  Evans, A.T., & Guthrie, W. (1991). Lymphoepithelioma-like carcinoma of the uvula and soft palate: a rare lesion in an unsual site. Histopathology, 19, 184-186.  Feller, L. & Lemmer, J. (2012). Oral leukoplakia as it relates to HPV infection: a review. Int J Dent, 2012, 540561. doi: 10.1155/2012/540561  Ferlay, J., Shin, H., & Bray, F. (2012). GLOBOCAN 2010: Cancer incidence and mortality worldwide (IARC Cancer Base 2008 estimates). Lyon, France: IARC Press.  Ferlay, J., Pisani, P., & Parkin, D.M. (2004). GLOBOCAN 2002: Cancer incidence, mortality and prevalence worldwide (IARC Cancer Base 2002 estimates). Lyon, France: IARC Press.  Ferlay, J., Soerjomataram, I., Diskshit, R., Eser, S., Mathers, C., Rebelo, M., et al. (2014). Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J cancer, 136, E359-E386.  Fernández-Medarde, A. & Santos, E. (2011). Ras in cancer and developmental diseases. Genes & cancer, 2(3), 344-358.  Field, J.K., Spandidos, D.A., Malliri, A., Gosney, J.R., Yiagnisis, M., & Stell, P.M. (1991). Elevated p53 correlates with a history of heavy smoking in squamous cell carcinoma of the head and neck. Br J Cancer, 64, 573-577.

124

 Field, J.K., Spandidos, D.A., & Stell, P.M. (1992). Overexpression of p53 gene head and nek cancer, linked with heavy smoking and drinking. Lancet, 339, 502-503.  Foulkes, W.D., Brunet, J.S., Sieh, W., Black, M.J., Shenouda, G., & Narod, S.A. (1996). Familial risks of squamous cell carcinoma of head and neck: retrospective case-control study. British Medical Journal, 313(7059), 716-721.  Fournir, J.G. & Escaig-Haye, F. (1999). In Situ molecular hybrizization techniques for ultrathin sections. Electron Microscopy Methods and Protocols. Springer. p167- p181.  Garcı´a-Closas, M., Malats, N., Silverman, D., Dosemeci, M., Kogevinas, M., & Hein, D.W., et al. (2005). NAT2 slow acetylation, GSTM1 null genotype, and risk of bladder cancer: results from the Spanish Bladder Cancer Study andmetaanalyses. Lancet, 366(9486), 649–659.  Garrote, L.F., Herrero, R., Ortiz Reyes, R.M., Vaccarella, S., Lence Anta, J., Ferbeye, L., et al. (2001). Risk factors for cancer of the oral cavity and oro-pharynx in Cuba. Br J Cancer, 85(1), 46-54.  Garte, S., Boffetta, P., Caporaso, N., & Vineis, P. (2001). Metabolic gene allele nomenclature. Cancer Epidemiol. Biomarkers Prev, 10, 1305–1306.  Geisler, S.A., Olshan, A.F., Cai, J., Weissler, M., Smith, J., & Bell, D. (2005). Glutathione S-transferase polymorphisms and survival from head and neck cancer. Head Neck, 27(3), 232-342.  Goldstein, A.M., Blot, W.J., Greenberg, R.S., Schoenberg, J.B., Austin, D.F., Peston- Martin, S., et al. (1994). Familial risk in oral and pharyngeal cancer. Eur J Cancer Part B, 30(5), 319-322.  González, M.V., Alvarez, V., Pello, M.F., Menéndez, M.J., Suárez, C., Coto, E. (1998). Genetic polymorphism of N-acetyltransferase-2, glutathione S-transferase-M1 and Cytochromes P450IIE1 and P450IID6 in the susceptibility to head and neck cancer. J Clinic Pathol, 51(4), 294-298.  Gronau, S., Koenig-Greger, D., Jerg, M., & Riechelmann, H. (2003). GSTM1 enzyme concentration and enzyme activity in correlation to the genotype of detoxification enzymes in squamous cell carcinoma of the oral cavity. Oral Dis, 9(2), 62-67.

125

 Gronau, S., Koenig-Greger, D., Jerg, M., & Riechelmann, H. (2003). Gene polymorphisms in detoxification enzymes as susceptibility factor for head and neck cancer. Otolaryngol Head Neck Surg, 128, 674-680.  Gulland, A. (2014). Global cancer prevalence is growing at ―alarming pace,‖ says WHO. BMJ, 348, g1338. http://dx.doi.org/10.1136/bmj.g1338  Gupta B., & Johnson, N.W. (2014). Emerging and established global life-style risk factors for cancer of the upper aero-digestive tract. Mini Review. Asian Pacific J Can Prev, 15, 5983-5991.  Gupta, P.C., & Ray, C.S. (2003). Smokeless tobacco and health in India and South Asia. Respirology, 8(4), 419-431.  Gusterson, B.A., Anbazhagan, R., Warren, W., Midgely, C., Lane, D.P., O'Hare, M., et al. (1991). Expression of p53 in pre-malignant and malignant squamous cell epithelium. Oncogene, 6, 1785-1789.  Ha, P.K., Chang, S.S., Glazer, C.A., Calofano, J.A., & Sidransky, D. (2009). Molecular techniques and genetic alterations in head and neck cancer. Oral oncology, 45(4), 335- 339.  Haider, S.M., Merchant, A.T., Fikree, F.F., & Rahbar, M.H. (2000). Clinical and functional staging of oral submucous fibrosis. Br J Oral Maxfac Surg, 38, 12-15.  Hameed, A., Sobani, Z.A., Nazir, M.R., & Ghaffar, S. (2012). Recurrence patterns of head and neck cancer at a tertiary care center in Karachi, Pakistan. Pak J Otolaryngol, 28, 72-74.  Harty, L.C., Caporaso, N.E., Hayes, R.B., Winn, D.M., Bravo-Otero, E., Blot, W.J., et al. (1997). Alcohol dehydrogenase 3 genotype and risk of oral cavity and pharyngeal cancers. J Natl Cancer Inst, 89, 1698-1705.  Hashibe, M., Brennan, P., Strange, R.C., Bhisey, R., Cascorbi, I., Lazarus, P., et al. (2003). Meta- and pooled analyses of GSTM1, GSTT1, GSTP1, and CYP1A1 genotypes and risk of head and neck cancer. Cancer Epidemiol Biomarkers Prev, 12, 1509–1517.  Hayashi, A., Watanabe, J., Nakachi, K., & Kawajiri, K. (1991). Genetic linkage of lung cancer-associated MspI polymorphisms with amino acid replacement in heme binding region of the human cytochrome P4501A1 gene. I Biochem (Tokyo), 110, 407-411.

126

 Hecht, S.S., Rivenson, A., Braley, J., DiBello, J., Adams, J.D., & Hoffmann, D. (1986). Induction of oral cavity tumors in F344 rats by tobacco-specific nitrosamines and snuff. Cancer Res, 46, 4162–4166.  Hecht, S.S. (2003). Tobacco carcinogens, their biomarkers and tobacco-induced cancer. Nat Rev Cancer, 3, 733-744.  Herrero, R., Castellsagué, X., Pawlita, M., Lissowska, J., Kee, F., Balaram, P., et al. (2003). Human papilloma virus and oral cancer: the International agency for research on cancer multicentre study. J Nat Canc Inst, 95(23), 1772-1783.  Hironen, A., Husgafvel-Pursiainen, K., Karjalainen, A., Anttila, S., & Vainio, H. (1992). Point-mutational MspI and Ile-Val polymorphisms closely linked in the CYP1A1 gene: lack of association with susceptibility to lung cancer in Finnish study population. Cancer Epidemiol Biomark Prev, 1, 485-489.  Hoffmann, D. & Djordjevic, M.V. (1997). Chemical composition and carcinogenicity of smokeless tobacco. Adv Dent Res, 11, 322–329.  Hsieh, L.L., Wang, P.F., Chen, I.H., Liao, C.T., Wang, H.M., Chen, M.C., et al. (2001). Characteristics of mutations in the p53 gene in oral squamous cell carcinoma associated with betel quid chewing and cigarette smoking in Taiwanese. Carcinogenesis, 22(9), 1497-1503.  Huang, Q., Yu, G.P., McCormick, S.A., Mo, J., Datta, B., Mahimkar, M., et al. (2002). Genetic differences detected by comparative genomic hybridization in head and neck squamous cell carcinomas from different tumor sites: construction of oncogenetic trees for tumor progression. Genes Chromosomes Cancer, 34, 224-233.  Hung, R.J., Boffetta, P., Brockmöller, J., Butkiewicz, D., Cascorbi, I., Clapper, M.L., et al. (2003). CYP1A1 and GSTM1 genetic polymorphisms and lung cancer risk in Caucasian non-smokers: a pooled analysis. Carcinogenesis, 24(5), 875-882.  Hunt, J.L. (2011). An update on molecular diagnosis of squamous and salivary gland tumors of head and neck. Arch Pathol Lab Med, 135(5), 602-609.  IARC. (1985). Tobacco habits other than smoking; betel quid and areca nut chewing; and some related nitrosamines. IARC Working Group. Lyon, 23-30 October 1984. IARC Monogr Eval Carcinog Risk Chem Hu, 37, 1-268.

127

 Idris, A.M., Ahmed, H.M., Mukhtar, B.I., Gander, A.F., & El-Beshir, E.I. (1995). Descriptive epidemiology of oral neoplasms in Sudan 1970-1985 and the role of toombak. Int J Cancer, 61, 155-158.  Ihsan, R., Chauhan, P.S., Mishra, A.K., Singh, L.C., Sharma, J.D., Zomawia, E., et al. (2014). Copy number polymorphism of glutathione-S-transferase genes (GSTM1 & GSTT1) in susceptibility to lung cancer in high-risk population of north-east India. Indian J Med Res, 139, pp720-729.  Imam, S.Z., Nawaz, H., Sepah, Y.J., Pabaney, A.H., Ilyas, M., Ghaffar, S. (2007). Use of smokeless tobacco among groups of Pakistani medical students—a cross sectional study. BMC Public Health, (7), 231.  Ingelman-Sundberg, M., Oscarson, M., Daly, A.K., Garte, S., Nebert, D.W. (2001). Human cytochrome P-450 (CYP) genes: a web page for the nomenclature of alleles. Cancer Epidemiol Biomarkers Prev, 10, 1307–1308.  International Agency for Research on Cancer. (2012). Globocan 2012: Fact Sheets by Population. Retrieved from http://globocan.iarc.fr/Pages/fact_sheets_population.aspx.  Jafarey, N.A., & Zaidi S.H.M. (1976). Carcinoma of oral cavity and oropharynx in Karachi. An appraisal. Tropical Doctor, 6, 63-67.  Jafarey, N.A., Sheikh, S., & Zaidi S.H.M. (1972). Prevalence of oral lesions in three localities of Karachi, 1972. J Pak Med Assoc, 22(5), 125-129.  Jafarey, N.A., Zaidi, S.H.M., & Jinnah, M. (1969). Preliminary report on the association of chewing and smoking habits with oral lesions. J Pak Med Assoc, 19(2), 1-64.  Jafarey, N.A. (1993). Betel quid, betel nut and other chewing habits. Community health, 15(3), 42-44.  Jaysurya, R., Francis, G., Kannan, S., Lekshminarayanan, K., Nalinakumari, K., Abraham, T., et al. (2014). p53, p16, and cyclin D1: molecular determinants of radiotherapy treatment response in oral carcinoma. Int J Can, 109(5), 710-716.  Jeng, J.H., Chang, M.C., & Hahn, L.J. (2001). Role of areca nut in betel quid-associated chemical carcinogenesis: current awareness and future perspectives. Oral Oncology, 37(6), 477-492.

128

 Johnson, N.W., Warnakulasuriya, S., Gupta, P.C., Dimba, E., Chindia, M., Otoh, E.C., et al. (2011). Global oral health inequalities in incidence and outcomes of oral cancer: causes and solutions. Adv Dent Res, 23(2), 237-246.  Jorquera, R., Castonguay, A., & Schuller, H.M. (1994). DNA single-strand breaks and toxicity induced by 4-(methyl-nitrosamino)-1-(3-pyridyl)-1-butanone or N- nitrosodimethylamine in hamster and rat liver. Carcinogenesis, 15, 389–394.  Josefen, K., & Nielsen, H. (2010). Northern blotting analysis. In H. Neilsen (Ed.), RNA; Methods in Molecular Biology, 703, 87-105. Springer. Retrieved from link.springer.com  Josephy, P.D. (2010). Genetic variations in human glutathione transferase enzymes: significance for pharmacology and toxicology. Hum Genomics Proteonomics, 876940. doi: 10.4061/2010/876940  Jusawalla, D.J., & Desphande, V. (1971). Evaluation of cancer risk in tobacco chewers and smokers: An epidemiologic assessment. Cancer, 28, 244-252.  Kadri, S., Samiuddin, Ahmed, N., & Mahmood, T. (2015). Malignant head and neck tumors in Radiology Department JPMC Karachi—a tertiary care experience. J Pak Med Assoc. 65(8), 862-864.  Kalyankrishna, S., & Grandis, J.R. (2006). Epidermal growth factor receptor biology in head and neck cancer. Journal of Clinical Oncology, 24(17), 2666-2672.  Kashif, M., Minhas, S., & Nagi, A.H. (2015). A Clinic-morphological Study of Oral Squamous Cell Carcinoma in Pakistan. Int J Curr Res, 7(6), 17401-17405.  Katoh, T., Kaneko, S., Kohshi, K., Munaka, M., Kitagawa, K., Kunugita, N., Ikemura, K., & Kawamoto, T. (1999). Genetic polymorphisms of tobacco- and alcohol-related metabolizing enzymes and oral cancer. Int J Cancer, 83, 606-609.  Kawajiri, K. (1999). CYP1A1. In Vineis, P., Malats, N., Lang, M., d'Errico, A., Caporaso, N., Cuzick, J. and Boffetta, P. (eds) Metabolic Polymorphisms and Susceptibility to Cancer. IARC Scientific Publications no. 148. IARC, Lyon, France, pp. 159–172.  Kayembe, M.K., & Kalengayi, M.M. (1999). Histological and epidemiological profile of oral cancer in Congo (Zaire). Odontostomatol Trop, 22(88), 29-32.

129

 Kazmi, F.N., Adil, A., Ghaffar, S., & Ahmed, F. (2012). Association between tumor volume and recurrence of squamous cell carcinoma of head and neck. J Pak Med Assoc, 62, 1129-1133.  Keller, A.Z. (1970). Cellular types, survival race nativity, occupation, habits and associated diseases in the pathogenesis of lip cancer. Am J Epidemiol, 91, 486-499.  Khalili, J. (2008). Oral cancer: risk factors, prevention and diagnosis. Exp Oncol, 30(4), 259-264.  Khan, M.A., Saleem, S., Shahid, S.M., Hameed, A., Qureshi, N.R., Abbasi, Z., et al. (2012). Prevalence of oral squamous cell carcinoma (OSCC) in relation to different chewing habits in Karachi, Pakistan. Pak J Biochem Mol Biol, 45(2), 59-63.  Khan Z, Tönnies T, &Müller S. (2014). Smokeless Tobacco and Oral Cancer in South Asia: A Systematic Review with Meta-Analysis. J Can Epidemiology. doi.org/10.1155/2014/394696  Khanna, J.N., & Andrade, N.N. (1995). Oral submucous fibrosis: a new concept in surgical management. Report of 100 cases. Int J Oral Maxillofacial Surg, 24(6), 433-439.  Khyani, I.A.M., Qureshi, M.A., Farooq, M.U., & Mirza, T. (2014). Molecular diagnosis of oral pre-malignant lesions and oral squamous cell carcinoma in saliva- A breakthrough in Pakistan. Int J Endor Health Sci Res, 2(2), 108-116.  Kolahdoozan, S., Sadjadi, A., Radmard, A.R., & Khademi, H. (2010). Five Common Cancers in Iran: Mini Review. Arch Iran Med, 13 (2), 143-146.  Kramer, I.R., Lucas, R.B., Pindborg, J.J., & Sobin, L.H. (1978). WHO Collaborating Centre for Oral Precancerous Lesions. Definition of leukoplkia and related lesions: An aid to studies on oral precancer. Oral Surg Oral Med Oral Pathol, 46(4), 518-539.  Kulkarni, A.P. (2001). Role of biotransformation in conceptal toxicity of drugs and other chemicals. Curr Pharm Des, 7, 833–857.  Landi, M.T., Bertazzi, P.A., Shields, P.G., Clark, G., Lucier, G.W., Garte, S.J., et al. (1994). Association between CYP1A1 genotype, mRNA expression and enzymatic activity in humans. Pharmacogenetics, 4(5), 242-246  Landi, S. (2000). Mammalian class theta GST and differential susceptibility to carcinogens: a review. Mutat Res, 463, 247-283.

130

 Li, L., Psoter, W.J., Buxo, C.J., Elias, A., Cuadrado, L., & Morse, D. (2011). Smoking and drinking in relation to oral potentially malignant disorders in Puerto Rico: a case- control study. BMC Cancer, 11, 324.  Lian, B., Tseng, Y.T., Su, C.C. & Tsai, K.Y. (2013). Progression of precancerous lesions to oral cancer: results based on the Taiwan National Health Insurance Database. Oral Oncol, 49, 427-430.  Lin, C.Y., Lovén, J., Rahl, P.B., Paranal, R.M., Burge, C.B., Brander, J.E., et al. (2012). Transcriptional amplification I tumor cells with elevated c-Myc. Cell, 151(1), 56-67.  Liu, C.J., Lee, Y.J., Chang, K.W., Shih, Y.N., Liu, H.F., & Dang, C.W. (2004). Polymorphism of the MICA gene and risk for oral submucous fibrosis. J Oral Pathol Med, 33(1), 1-6.  Liu, S.C., & Klein-Szanto, A. (2000). Markers of proliferation in normal and leukoplakic oral epithelia. Oral oncology, 36(2), 145-151.  Liu, Y., Sundqvist, K., Belinksy, S.A., Castonguay, A., Tjälve, H., & Grafström, R.C. (1993). Metabolism and macromolecular interaction of the tobacco-specific carcinogen 4- (methylnitrosamino)-1-(3-pyridyl)-1-butanone in cultured explants and epithelial cells of human buccal mucosa. Carcinogenesis, 14, 2383–2388.  London, S.J., Yuan, J.M., Chung, F.L., Gao, Y.T., Coetzee, G.A., Ross, R.K., et al. (2000). Isothiocyanates, glutathione S-transferase M1 and T1 polymorphisms, and lung- cancer risk: a prospective study of men in Shanghai, China. Lancet, 356(9231), 724–729.  Longworth, M.S., & Lamins, L.A. (2004). Pathogenesis of human papilloma viruses in differentiating epithelia. Microbiol Mol Biol Rev, 68, 362-372.  Lopes, M.A. (2010). Proliferative verrucous leukoplakia, squamous cell carcinoma and axillary metastasis. Med Oral Pathol Oral Cir Bucal, 15(5):e704-708.  Luo, C.W., Roan, C.H., & Liu, C.J. (2007). Human papilloma viruses in oral squamous cell carcinoma and pre-cancerous lesions detected by PCR-based gene-chip array. Int J Oral Maxillofac Surg, 36, 153-158.  Lόpez-Cima, M.F., Alvarez-Avellόn, S.M., Pascual, T., Fernández-Somoano, A., & Tardón, A. (2012). Genetic polymorphisms in CYP1A1, GSTM1, GSTP1 and GSTT1 metabolic genes and risk of lung cancer in Asturias. BMC Cancer, 12, 433.

131

 Maden, C., Beckmann, A.M., Thomas, D.B., McKnight, B., Sherman, K.J., Ashley, R.L., et al. (1992). Human papillomaviruses, herpes simplex viruses, and the risk of oral cancer in men. Am J Epidemiol, 135(10), 1093-1102.  Maestro, R., Dolcetti, R., Casprotto, D., Doglioni, C., Pelucchi, S., Barzan, L., et al. (1991). High frequency of p53 gene alterations associated with protein overexpression in human squamous cell carcinoma of the larynx. Oncogene, 7(6), 1159-1166.  Mahmood, Z., Khurshid, A. & Jafarey, N.A. (1986). Incidence of the cancer of oral cavity and pharynx in Pakistan. Pak J Med Res, 25(4), 173-181.  Malik, I.A., & Khan, W.A. (1998). Pattern of malignant tumors observed in a university hospital: A retrospective analysis. J Pak Med Assoc, 48, 120-122.  Mandviwala, S.S., & Jafarey, N.A. (1977). Histological grading and clinical staging in estimating prognosis of squamous cell carcinoma of the oral cavity. J Pak Med Assoc, 25(6), 343-345.  Mao, L.I., & Hong, W.K. (2004). Hoe does human papilloma virus contributes to head and neck cancer development. Journal of National Cancer Institute, 96(13), 978-979.  Martinez-Madrigal, F., Baden, E., Casiraghi, O., & Micheau, C. (1991). Oral and pharyngeal adenosquamous carcinoma. A report of four cases with immunohistochemical studies. Eur Arch Otorhinolaryngol, 248, 255-258.  Marugame, T., Matsuda, T., Kamo, K., Katanoda, K., Ajiki, W., & Sobue, T. (2007). Japan cancer surveillance research group. Cancer incidence and incidence rates in Japan in 2001 based on the data from 10 population-based cancer registries. Japanese J Clin Oncol, 37, 884-891.  Marur, S., DSouza, G., Westra, W.H., & Forastiere, A.A. (2010). HPV-associated head and neck cancer: a virus-related cancer epidemic. The lancet oncology, 11(8), 781-789.  Mashberg, A., & Samit, A. (1995). Early diagnosis of asymptomatic oral and oropharyngeal squamous cancers. CA Cancer J Clin, 45, 328-351.  Matta, A., & Ralhan, R. (2009). Overview of current and future biologically based targeted therapies in head and neck squamous cell carcinoma. Head & Neck Oncology, 1(1), 6.

132

 Matthias, C., Bockmuhl, U., Jahnke, V., Harries, L.W., Wolf, C.R., Jones, P.W., et al. (1998). The glutathione S-transferase GSPT1 polymorphism: effects on susceptibility to oral/pharyngeal and laryngeal carcinomas. Pharmacogenetics, 8, 1-6.  Mayne, S.T., Morse, D.E., & Winn, D.M. (2006). Cancer of the oral cavity and pharynx. In Schottenedfeld, D., & Fraumeni, J.F., Jr (Eds.), Cancer Epidemiology & Prevention (3rd ed., pp. 674-696). New York: Oxford University Press.  Mcllwain, C.C., Townsend, D.M., & Tew, K.D. (2006). Glutahtione S-transferase polymorphisms: cancer incidence and therapy. Oncogene, 25, 1639-1648.  Meher, R., Lee, A.J., Warnakulasuriya, A.A.S., Lewis, J.A, & Johnson, N.W. (1994). Role of areca nut in the causation of oral submucous fibrosis: a case-control study in Pakistan. J Oral Pathol Med, 23, 63-69.  Mehrotra, R., Gupta, A., Singh, M., & Ibrahim, R. (2006). Application of cytology and molecular events in diagnosing pre-malignant or malignant oral lesion. Mol Cancer, 5, 11-15.  Mehrotra, R., & Yadav, S. (2006). Oral squamous cell carcinoma: etiology, pathogenesis and prognostic value of genomic alterations. Indian J Cancer, 43(2), 60-66.  Mehul Shinde (2015). Oral precancerous lesions. Education, Published on Apr 30th, accessed on 14th July, 2017 from https://www.slideshare.net/mehulshinde2/oral- precancerous-lesions  Memon, I.M., Iqbal, S.M., Hussain, S.I., & Baig, M.N. (2014). Pattern of oral malignancies at tertiary care hospitals. Pak J Surg, 30(3), 268-271.  Memon, M.A., Shaikh, M.S., & Jaffery, M.H. (2015). Oral submucous fibrosis in rural Sindh. J Liaquat Uni Med Health Sci, 14(1), 44-47.  Mendes, R.A., Carvalho, J.F., & Waal, I.V.D. (2009). An overview on the expression of cyclooxygenase-2 in tumors of head and neck. Oral oncology, 45(10), e124-e128.  Merchant, A., Hussain, S.M., Hosain, M., Fikree, F., Pitiphat, W., Siddiqui, A.R., et al. (2000). Paan without tobacco: an independent risk factor for oral cancer. Int J Cancer, 86, 128-131.  Miller, C.S., & White, D.K. (1996). Human papillomavirus expression in oral mucosa, premalignant conditions, and squamous cell carcinoma: a retrospective review of the literature. Oral Surg Oral Med Oral Path Oral Radiol Endod, 82(1), 57-68.

133

 Mirza, T., Alam, S.M., Lauder, I., Pringle, J.H., & Zaidi, S.H. (1998). Molecular analysis of human papilloma virus and oncosuppressor genes in tobacco related oral cancer. Pak J Otolaryngol, 14, 27-33.  Moore, C. (1971). Cigarette smoking and cancer of the mouth, pharynx, and larynx: a continuing study. J Am Med Assoc, 218, 553-558.  Morice, W.G., & Ferreiro, J.A. (1988). Distinction of basaloid squamous cell carcinoma from adenoid cystic and small cell undifferentiated carcinoma by immunohistochemistry. Hum Pathol, 29, 609-612.  Morse, D.E., Pendrys, D.G., Neely, A.L., & Psoter, W.J. (1999). Trends in the incidence of lip, oral, and pharyngeal cncer: Connecticut, 1935-94. Oral Oncol, 35(1), 1-8.  Murgan, A.K., Munirajan, A.K., & Tsuchida, N. (2012). Ras oncogenes in oral cancer: The past 20 years. Oral oncology, 48(5), 383-392.  Murti, P.R., Bhonsle, R.B., Pindborg, J.J., Daftary, D.K., Gupta, P.C., & Mehta, F.S. (1985). Malignant transformation rate in oral submucous fibrosis over a 17 year period. Comm Dent Oral Epidemol, 13, 340-341.  Muttagi, S.S., Chaturvedi, P., Gaikwad, R., Singh, B., & Pawar, P. (2012). Head and neck squamous cell carcinoma in chronic areca nut chewing Indian women: Case series and review of literature. Ind J Med & Paed Oncol, 33(1), 32-35.  Nair, U.J., Nair, J., Mathew, B., & Bartsch, M. (1999). Glutathione S-transferase M1 and T1 null genotypes as risk factors for oral leukoplakia in ethnic Indian betel quid/tobacco chewers. Carcinogenesis, 20(5), 743-748.  Nasr, A.S., Sami, R.M., Ibrahim, N.Y., & Darwish, D.O. (2015). Glutathione S transferase (GSTP 1, GSTM 1, and GSTT 1) gene polymorphisms in Egyptian patients with acute myeloid leukemia. Ind J Can, 52(4), 490-495.  National Cancer Control Programme. 2012. Cancer Incidence Data: Sri Lanka Year 2006. Retrieved from https:/www.nccp.health.gov.lk/images/PDF_PUBLICATIONS/ CancerRegistry2006.pdf  Naveen, A.T., Adithan, C., Padmaja, N., Shashindran, C.H., Abraham, B.K., Saatyanarayanamoorthy, K., et al. (2004). Glutathione S-transferase M1 and T1 null genotype distribution in South Indians. Eur. J. Clin. Pharmacol, 60, 403-406.

134

 Netto, G.J., Saad, R.D., & Dysert, P.A. (2003). Diagnostic molecular pathology: current techniques and clinical applications. Part I proceedings, Baylor University Medical Centre, 16(4), 379.  Neville, B.W., & Day, T.A. (2002). Oral cancer and precancerous lesions. CA Cancer J Clin, 52(4), 195-215.  Newcomb, P.A., & Carbone, P.P. (1992). The health consequences of smoking: cancer. Med Clin North Am, 76, 305-331.  Nisar, M.I. & Iqbal, R. (2011). Smokeless tobacco use prevention and cessation (S- TUPAC): a need of the time. J Pak Med Assoc, 61(7), 711–712.  Noronha, V., Tsomo, U., Jamshed, A., Hai, M., Wattegama, S., Baral, R., et al. (2012). A fresh look at oncology facts on South Central Asia and SAARC countries. S Asian J of Can, 1(1), 1-4.  Norppa, H. (2004). Cytogenetic biomarker and genetic polymorphisms. Toxicol Lett, 149, 309-334.  Norppa, H. (2004). Cytogenetic biomarkers. IARC Sci Publ, 157, 179-205.  Ogden, G.R., Cowpe, J.G., Chisholm, D.M., & Lane, E.B. (1994). DNA and keratin analysis of oral exfoliative cytology in the detection of oral cancer. Eur J Cancer B Oral Oncol, 30B, 405-408.  Ogden, G.R. (2005). Alcohol and oral cancer. Alcohol, 35(3), 169-173.  Olayioye, M.A., Neve, R.M., Lane, H.A., & Hynes, N.E. (2000). NEW EMBO MEMBER‘S REVIEW: The ErbB signaling network: receptor hetrodimerization in development and cancer. The EMBO journal, 19(13), 3159.  Omiecinski, C.J., Vanden Heuvel, J.P., Perdew, G.H., & Peters, J.M. (2011). Xenobiotic metabolism, disposition, and regulation by receptors: From biochemical phenomenon to predictors of major toxicities. Toxicological sciences, 120(S1), S49-S75.  Onyango, J.F., Omondi, B.I., Njiru, A., & Awange, O.O. (2004). Oral cancer at Kenyatta National Hospital, Nairobi. East Afr Med J, 81(6), 318-321.  Pai, R., Pai, S., Lalitha, R., Kumaraswamy, S., Lalitha, N., Johnston, R., et al. (2009). Overexpression of c-Myc oncoprotein in oral squamous cell carcinoma in the South Indian population. Ecancermedicalscience, 9, 3:128.

135

 Pakistan Medical Research Council (1977). Collection of data of various types of tumors in Pakistan. PMRC Monograph, 1, 58.  Pakistan Medical Research Council (1982). Malignant tumor report of a multi-centre study. Karachi, Pakistan.  Pande, P., Mathur, M., Shukla, N., & Ralhan, R. (1998). pRb and p16 protein alterations in human oral tumorigenesis. Oral oncology, 34(5), 396-403.  Pandya, S., Chaudhary, A.K., Mantra, S., Mangal, S., & Mehrotra, R. (2009). Correlation of histopathological diagnosis with habits and clinical findings in oral submucous fibrosis. Head Neck Oncol, 1, 10.  Pannone, G., Rodolico, V., Santoro, A., Lo Muzio, L., Franco, R., Botti, G., et al. (2012). Evaluation of a combined triple method to detect causative HPV in oral and oropharyngealsquamous cell carcinoma: p16 immunohistochemistry, Consensus PCR HPV-DNA, and In Situ Hybridization. Infect Agent Cancer, 7, 4.  Papakosta, V., Vairaktaris, E., Vylliotis, A., Derka, S., Nkenke, E., Vassiliou, S., et al. The co-expression of c-myc and p53 increases and reaches a plateau early in oral oncogenesis. Anticancer res, 26(4B), 2957-2962.  Park, N.H., Min, B.M., Li, S.L., Huang, M.Z., Cherick, H.M., & Doniger J. (1991). Immortalization of normal human oral keratinocytes with type-16 human papilloma virus. Carcinogenesis, 12(9), 1627-1631.  Parkin, D.M., Bray, F., Farley, J., & Pisani, P. (2002). Global Cancer Statistics. CA Cancer J Clin, 55(2), 74-108.  Parkin, D.M., Bray, F., Farley, J., & Pisani, P. (2001). Estimating the world cancer burden: Globocan 2000. Int J Cancer, 94(2), 153-156.  Parl, F.F. (2005). Glutathione S-transferase genotypes and cancer risk. Cancer Lett, 221, 123-129.  Patel, B.P., Rawal, U.M., Shah, P.M., Prajapati, J.A., Rawal, R.M., Dave, T.K., at al. (2005). Study of tobacco habits and alterations in enzyme antioxidant system in oral cancer. Oncology, 68, 511-519.  Pelkonen, O., & Raunio, H. (1997). Metabolic activation of toxins: tissue-specific expression and metabolism in target organs. Environ Health Perspect, 105, Suppl 4, 767- 774.

136

 Pemble, S., Schroeder, K.R., Spencer, S.R., Meyer, D.J., Hallier, E., Bolt, H.M., et al. (1994). Human glutathione S transferase theta (GSTT1): cDNA cloning and the characterization of a genetic polymorphism. Biochem J, 300, 271-276.  Pérez-Sayáns, M., Samoza-Martín, J.M., Barros-Angueira, F., Reboiras-Lόpez, M.D., Gándara Rey J.M., & García, A. (2009). Genetic and molecular alterations associated with oral squamous cell cancer (Review). Oncology reports, 22(6), 1277-1282.  Petersen, D.D., McKinney, C.E., Ikeya, K., Smith, H.H.,, Bale, A.E., McBride, O.W., et al. (1991). Human CYP1A1 gene: cosegregation of the enzyme inducibility phenotype and an RFLP. Am J Hum Genet, 48(4), 720-725.  Petti, S. (2009). Lifestyle risk factors for oral cancer. Oral oncology, 45(4), 340-350.  Pfeifer, G.P., & Besaratinia, A. (2009). Mutational spectra of human cancer. Human genetics, 125(5-6), 493-506.  Pindborg, J.J., Murti, P.R., Bhosle, R.B., Gupta, P.C., Daftary, D.K., & Mehta, F.S. (1984). Oral submucous fibrosis as a pre-cancerous condition. Scand J Dent Res, 92, 224-229.  Pindborg, J.J., Reichart, P.A., Smith, C.J., & Waal, I. (1997). World Health Organization histological typing of cancer and precancer of the oral mucosa (2nded.). New York: Springer.  Pindborg, J.J. (1989). Oral submucous fibrosis: A Review. An Acad of Med Singapore, 18(5), 603-607.  Poeta, M.L., Manola, J., Goldwasser, M.A., Forestiere, A., Benoit, N., Califano, J.A., et al. (2007). Tp53 mutations and survival in squamous cell carcinoma of the head and neck. N Eng J Med, 357, 2552-2561.  Pradhan, S.A. (1991). Gingivo-buccal cancer – the Indian oral cancer. In R.S. Rao & P.B. Desai (Eds.), Oral Cancer. 63-76.  Presland, R.B., & Dale, B.A. (2000). Epithelial structural proteins of the skin and oral cavity: function in health and disease. Crit Rev Oral Biol Med, 11, 383-408.  Queyrat, L. (1911). ―Erythroplasie de gland‖. Bull Soc Bull Franc Derm Syph, 22, 378- 382.  Rahman, M., & Fukui, T. (2000). Bidi smoking and health. Public Health, 114, 123-127.

137

 Rahman, M., Sakamoto, J., & Fukui, T. (2003). Bidi Smoking And Oral Cancer: A Meta- analysis. Int J Cancer, 106, 600-604.  Raimondi, S., Paracchini, V., Autrup, H., Barros-Dios, J.M., Benhamou, S., Boffetta, P., et al. (2006). Meta- and pooled analysis of GSTT1 and lung cancer: a HuGE-GSEC review. Am J Epidemiol, 164, 1027-1042.  Ralhan, R., Agarwal, S., Nath, N., Mathur, M., Wasylyk, B., & Sivastava, A. (2001). Correlation between p53 gene mutations and circulating antibodies in betel and tobacco-consuming North Indian population. Oral Oncology, 37(3), 243-250.  Ramzy, M.M., Solliman, M.E.D.M., Abdel-Hafiz, H.A., & Salah, R. (2011). Genetic polymorphism of GSTM1 and GSTP1 in lung cancer in Egypt. Int J Collab Res Intern Med Public Health, 3, 41-51.  Rao, V. & Chaturvedi, P. (2010). Tobacco and health in India. Indian J Cancer, 47, 3-8.  Raybaud-Diogene, H., Tetu, B., Morency, R., Fortin, A., & Monteil, R.A. (1996). P53 overexpression in head and neck squamous cell carcinoma: review of literature. Oran Oncol Eur I Cancer, 32B, 143-149.  Richiardi, L., Corbin, M., & Marron, M., Ahrens, W., Pohlabein, H., Lagiou, P., et al. (2012). Occupation and risk of upper aerodigestive tract cancer: the ARCAGE study. Int J Cancer, 130(10), 2397-2406.  Rickert, W.S., Joza, P.J., Trivedi, A.H., Momin, R.A., Wagstaff, W.G., & Lauterbach, J.H. (2009). Chemical and toxicological characterization of commercial smokeless tobacco products vailable on the Canadian market. Regul Toxicol Pharmaco, 53(2), 121- 133.  Rothenberg, S.M., & Ellisen, L.W. (2012). The molecular pathgogenesis of head and neck squamous cell carcinoma. J Clin Invest, 122(6), 1951-1957.  Rozati, R., Reddy, S., Giragalla, S.B., Bakshi, H., Doddmaneni, S., Khaja, N., et al. (2008). The CYP1A1 and GSTM1 Genetic Polymorphisms and Susceptibility to Endometriosis in Women from South India. Int J Fert Ster, 2(3), 105-112.  Rubab, Z., Mughal, A.M., Baig, S., Lucky, M.H., & Khan, M.A. (2013). Relationship of human papilloma virus with trismus in chewable tobacco users. Pak J Med Dent, 2(1), 3- 11.

138

 Rubin, P.H. (Ed.). (1993). Clinical oncology: A multidisciplinary approach to physicians and students. Philadelphia: Saunders.  Salama, S.A., Abdel-Rahman, S.Z., Sierr-Torres, C.H., Hamada, F.A., & Au, W.W. (1999). Role of polymorphic GSTM1 and GSTT1 genotypes on NNK-induced genotoxicity. Pharmacogenetics, 9, 735-743.  Saleem, S., Hameed, A., Abbasi, Z.A., Qureshi, N.R., Khan, M.A., Ajmal, M., et al. (2011). Loss of heterozygosity in p53 gene found in oral squamous cell carcinoma patients of Pakistan. Pak J Biochem Mol Biol, 44(2), 51-55.  Sandstörm, B., Mörnstad, H., & Axéll, T. (1982). Oral carcinomas associated with snuff dipping. Some clinical and histological characteristics of 23 tumours in Swedish males. J Oral Pathol, 11, 245-251.  Santonlis, P.K., Kastramkis, N.G., & Touruas, A.D. (2007). Advance in biology of oral cancers. Oral Oncol, 43(6), 523-534.  Sarnath, D., Chang, S., Bhoite, L., Panchal, R., Kerr, I., Mehta, A., et al. (1991). High frequency of mutation in codons 12 and 61 of H-ras oncogene in chewing tobacco-related human oral carcinoma in India. British J Cancer, 63(4), 573.  Sato, M., Sato, T., Izumo, T., & Amagasa, T. (1999). Genetic polymorphism of drug- metabolizing enzymes and susceptibility to oral cancer. Carcinogenesis, 20, 1927-1931.  Scheffner, M., Werness, B.A., Huibragtac, J.M, Levine, A.J., & Howley, P.M. (1990). The E6 oncoprotein encoded by human papilloma virus Types 16 and 18 promotes the degradation of p53. Cell, 63, 1129-1136.  Schneider, A. (1994). Natural history of genital papillomavirus infections.(Review). Intervirology, 37, 201-214.  Schroeff, M.P., & Baatenburg de Jong, R.J. (2009). Staging and prognosis in head and neck cancer. Oral Oncol, 45(4-5), 356-360.  Scully, C., & Bagan, J.V. (2009). Recent advances in oral oncology 2008; squamous cell carcinoma imaging, treatment and prognostication and treatment outcomes. Oral Oncol, 45(6), 25-30.  Shafer, W.B., & Waldron, C.A. (1961). A clinical and histopathologic study of oral leukoplakia. Surg Gynecol Obstet, 112, 411-420.

139

 Shafer, W.G., & Waldron, C.A. (1975). Erythroplakia of the oral cavity. Cancer, 36, 1021-1028.  Shah, I., Sefvan, O., Luqman, U., Ibrahim, W., Mehmood, S., & Alamgir, W. (2010). Clinical stage of oral cancer patients at the time of initial diagnosis. J Ayub Med Coll Abbottabad, 22(3), 61-63.  Shaik, S.S., Doshi, D., Bandari, S.R., Madupu, P.R., & Kulkarni, S. (2016). Tobacco use cessation and prevention- A review. J Clin Diagn Res, 10(5), ZE13-ZE17.  Sharma, D.C. (2003). Betel quid and areca nut are carcinogenic without tobacco. The Lancet Oncology, 4(10), 587.

 Shi, X., Zhou, S., Wang, Z., Zhou, Z., & Wang, Z. (2008). CYP1A1 and GSTM1 polymorphisms and lung cancer risk in Chinese populations: a meta-analysis. Lung Cancer, 59(2), 155–163.  Shimada, T., Yamazaki, H., Oda, Y., Hiratsuka, A., Watabe, T., & Guengerich, F.P. (1996). Activation and inactivation of carcinogenic dihaloalkanes and other compounds by glutathione S-transferase 5-5 in Salmonella typhimurium tester strain NM5004. Chem Res Toxicol, 9, 333-340.  Shin, K.H., Min, B.M., Cherrick, H.M., & Park, N.H. (1994). Combined effect of human papilloma virus –18 and N-methyl-N`–nitrosoguanidine on the transformation of normal human oral keratinocytes. Mol Carcinog, 9, 76-86.  Shukla, R.K., Kant, S., Mittal, B., & Bhattacharya, S. (2010). Ploymorphism of cytochrome p450, glutathione-S-transferase and N-acetyltransferases: influence on lung cancer susceptibility. Niger J Med, 19(3), 257-263.  Sikdar, N., Mahmud, S.A., Paul, R.R., & Roy, B. (2003). Polymorphism in CYP1A1 and CYP2E1 genes and susceptibility to leukoplakia in Indian tobacco users. Cancer, 195, 33-42.  Sikdar, N., Paul, R.R., & Roy, B. (2004). Glutathione S-transferase M3 (A/A) genotype as a risk factor for oral cancer and leukoplakia among Indian tobacco smokers. Int J Cancer, 109, 95–101.  Silverman, S.J. & Gorsky, M. (1990). Epidemiologic and demographic update in oral cancer: California and national data 1973 to 1985. J Am Dent Assoc, 120, 49-9.

140

 Sinor, P.N., Gupta, P.C., Murti, P.R., Bhonsle, R.B., Daftary, D.K., Mehta, F.S., et al. (1990). A case-control study of Oral Submucous Fibrosis with special reference to the etiologic role of areca nut. J Oral Pathol Med, 19(2), 94-98.  Sivaraman, L., Leatham, M.P., & Yee, J. (1994). CYP1A1 genetic polymorphism and situ colorectal cancer. Cancer Res, 54, 3692-3695.  Slootweg, P.J., & Eveson, J.W. (2005). Introduction: tumors of the oral cavity and oropharynx. World Health Organization classification of tumours. In L. Barnes, J. W. Eveson, P. Reichart & D. Sidransky (Eds.), Pathology and genetics of head and neck tumours (pp. 166-167). Lyon: IARC.  Smeets, S.J., Hesselink, A.T., Speel, E.J.M., Haesevoets, A., Snijders, P.J., Pawlita, M., et al. (2007). A novel algorithm for reliable detection of the human papilloma virus in paraffin embedded head and neck cancer specimen. Int J Cancer, 121(11), 2465-2472.  Sobin, L.H. (2003). TNM: evolution and relation to other prognostic factors. Semin Surg Oncol, 21, 3-7.  Sreelekha, T.T., Ramdas, K., Pandey, M., Thomas, G., Nalinakumari, K.R., & Pillai, M.R. (2001). Genetic polymorphisms of CYP1A1, GSTM1 AND GSTT1 genes in Indian oral cancer. Oral Oncol, 37, 593-598.  Srivastava, D.S., Mandhani, A., Mittal, B., & Mittal, R.D. (2005). Genetic polymorphisms of glutathione S-transferase (GSTM1, GSTT1 and GSTP1) and susceptibility to prostate cancer in Northern India. BJU Int, 95, 170-173.  Stricker, T.P., & Kumar, V. (2007). Neoplasia. In: V. Kumar, A.K. Abbas, N. Fausto, R.N. Mitchell (Eds.), Robin’s basic pathology (8th ed., pp.173-223). Philadelphia: W.B. Saunders.  Sufia, S., Khan, A.A., & Ijaz, S. (2003). Patterns of tobacco use in Pakistan. Pak Oral Dent J, 23(1), 45-50.  Swati, S., Madhusudan, A., Nisheeth, S., Nidhi, D., & Shitalkumar, G.S. (2012). Cyclin D1 Expression and Its Correlation with Histopathological Differentiation in Oral Squamous Cell Carcinoma. The Scientific World Journal, Article ID 978327.  Takagi, M., Sakota, Y., Takayama, S., & Ishikawa, G. (1977). Adenoid squamous cell carcinoma of the oral mucosa. Report of two autopsy cases. Cancer, 40, 2250-2255.

141

 Tanaka, T., & Ishigamori, R. (2011). Review article: Understanding carcinogenesis for fighting oral cancer. J Oncol, 2011, 603740.  Tanimoto, K., Hayashi, S., Yoshiga, K., & Ichikawa, T. (1999). Polymorphisms of the CYP1A1 and GSTM1 gene involved in oral squamous cell carcinoma in association with a cigarette dose. Oral Oncol, 35, 191–196.  Taranikanti, M., & Das, B. (2013). Risk factor profile of oral cancer patients in North East India. Int J Biomed Res, 4(11), 615-622.  Taspinar, M., Aydos, S.E., Comez, O., Elban, A.H., Karabulut, H.G., & Sunguroglu, A. (2008). CYP1A1, GST gene polymorphisms and risk of chronic myeloid leukaemia. Swiss Med Wkly, 138, 12-17.  Taylor, C., & Levenson, R. (2006). Quantification of immunohistochemistry – issues concerning methods, utility and semiquantitatie assessment II. Histopathology, 49(4), 411-424.  The Punjab cancer registry. (2016). Retrieved April 25, 2016, from http://punjabcancerregistry.org.pk/  Thompson, L.D., & Goldblum, J.R. (2006). Head and neck pathology. Churchill Livingstone Elsevier.  Toprani, R., & Patel, D. (2013). Betel leaf: Revisiting the benefits of an ancient Indian herb. South Asian J Cancer, 2(3), 140-141.  Townsend, D.M., Findlay, V.L., & Tew, K.D. (2005). Glutathione S-transferase as regulators of kinase pathways and anticancer drug targets. Methods Enzymol, 401, 287- 307.  Ullah, N., Asif, A.H., Khan, M.A., Ahmad, W., Ali, N., Khan, T., Shah, A.A. (2011). Chemical analyses of naswar and cigarettes; a comparative study. Int J Basic and Clin Res, 1(1), 13-15.  Vairaktaris, E., Spyridonidou, S., Papakosta, V., Vylliotis, A., Lazaris, A., & Perrea, D., et al. (2008). The hamster model of sequential oral carcinogenesis. Oral oncology, 44(4), 315-324.  Van Monsjou, H., van Velthuysen, M., van den Brekel, M., Jordanova, E., Melief, C., & Balm, A. (2012). Human papillomavirus status in young patients with head and neck squamous cell carcinoma. Int J Can, 130(8), 1806-1812.

142

 Varela-Lema, L., Taioli, E., Ruano-Ravina, A., Barros-Dios, J., Anantharaman, D., Benhamou, S., Boccia, S., Bhisey, R.A., Cadoni, G., et al. (2008). Meta-analysis and pooled analysis of GSTM1 and CYP1A1 polymorphisms and oral and pharyngeal cancers: a HuGE-GSEC review. Genet Med, 10(6), 369-384.  Vingneswaran, N., Tilashalski, K., Rodu, B. (1995). Tobacco use and cancer. Oral Surg Oral Med Oral Pathol, 80, 178-182.  Vondracek, M., Xi, Z., Larsson, P., Baker, V., Mace, K., Pfeifer, A., Tjalve, H., Donato, M.T., Gomez-Lechon, M.J., Grafstrom, R.C. (2001). Cytochrome P450 expression and related metabolism in human buccal mucosa. Carcinogenesis, 22, 481-488.  Vora, H.H., Mehta, S.V., Shukla, S.N., & Shah, P.M. (2010). No mutation is detected in five hot spot codons of TP53 gene by restriction site muataion analysis in patients with carcinoma of the tongue. Int J Biol Mark, 25(1), 46-51.  Wahab, N., Ali, S.A., Khan, M., Khan, S., Mehdi, H., & Sawani, A. (2014). Frequency and clinical presentation of oral submucous fibrosis. Pak J Med Dent, 3(4), 48-53.  Wake, M. (1993). The urban/rural divide in head and neck cancer-the effect of atmospheric pollution. Clin Otolaryngol Allied Sci, 18, 298-302.  Waldron, C.A., & Shafer, W.G. (1975). Leukoplakia revisited: A clinicopathologic study of 3256 oral leukoplakias. Cancer, 36, 1386-1392.  Wang, J., Sampath, A., Raychaudhuri, P., & Baqchi, S. (2001). Both Rb and E7 are regulated by the ubiquitin-proteasome pathway in HPV-containing cervical tumor cells. Oncogene, 20(34), 4740-4749.  Warnakulasuriya, S., Sutherland, G., & Scully, C. (2005). Tobacco, oral cancer and treatment of dependence. Oral oncology, 41(3), 244-260.  Warnakulasuriya, S., & van der Waal, J.N. (2007). Nomenclature and Classification of potentially malignant disorders of the oral mucosa. J Oral Pathol Med, 36, 575-580.  Warnakulasuriya, S. (2009). Global epidemiology of oral and oropharyngeal cancer. Oral oncology, 45(4), 309-316.  Werness, B.A., Levine, A.J., & Howley, P.M. (1990). Association of human papilloma virus types-16 and 18 E6 protein with p53. Science, 248, 76-79.

143

 White, A.C., Tran, K., Khuu, J., Dang, C., Cui, Y., Binder, S.W., et al. (2011). Defining the origins of Ras/p53-mediated squamous cell carcinoma. Proceedings of the National Academy of Sciences, 108(18), 7425-7430.  Wiencke, J.K., Pemble, S., Ketterer, B., & Kelsey, K.T. (1995). Gene deletion of glutathione S-transferase theta: correlation with induced genetic daage and potential role in endogenous mutagenesis. Cancer Epidemiol Biomarkers Prev, 4, 253-259.  Wingo, P.A., Reis, L.A., Giovino, G.A., Miller, D.S., Rosenberg, H.M., Shopland, D.R., et al. (1999). Annual report to the nation on the status of cancer, 1973-1996, with a special section on lung cancer and tobacco smoking. J Natl Cancer Inst, 91(8), 675-690.  Witkiewicz, A.K., Knudsen, K.E., Dicker, A.P., & Knudsen, E.S. (2011). The meaning of p16ink4a expression in tumors: Functional significance, clinical associations and future developments. Cell Cycle, 10(15), 2497-2503.  Wong, D.T. (2006). Salivary Diagnostics powered by nanotechnologies, proteomics and genomics. J Am Dent Assoc, 137:313-321.  Woolgar, J.A. (2006). Histopathological prognosticators in oral and oropharyngeal squamous cell carcinoma. Oral Oncol, 42(3), 229-239.  World Health Organization. (1992). Recent advances in oral health. Geneva, Switzerland. Report of a WHO Expert Committee.  Wright, A., & Shear, M. (1985). Epithelial dysplasia immediately adjacent to oral squamous cell carcinoma. J Oral Pathol, 14, 559-564.  Wunsch-Fiho, V., & de Camargo, A. (2001). The burden of mouth cancer in Latin America and the Caribbean: epidemiologic issues. Seminars Oncol, 28, 158-168.  Wynder, E.L., Mushinski, M.H., & Spivah, J.C. (1977). Tobacco and alcohol consumption in relation to the development of multiple primary cancers. Cancer, 40, 1872-1878.  Xie, H., Hou, L., Shields, P.G., Winn, D.M., Gridley, G., Bravo-Otero, E., et al. (2004). Metabolic polymorphisms, smoking, and oral cancer in Puerto Rico. Oncol. Res, 14, 315– 320.  Xie, S., Luo, C., Shan, X., Zhao, S., He, J., & Cai, Z. (2016). CYP1A1 MspI polymorphism and the risk of oral squamous cell carcinoma: Evidence from a meta- analysis. Molecular and clinical oncology, 4, 660-666.

144

 Yu, K.D., Di, G.H., Fan, L., Wu, J., Hu, Z., Shen, Z., et al. (2009). A functional polymorphism in the promoter region of GSTM1 implies a complex role for GSTM1 in breast cancer. FASEB J, 23(7), 2274-2287.  Zafar, M., Hadi, N.I., Baig, S., & Zehra, N. (2015). Association between interleukin 6 gene polymorphism and human papilloma virus infection in oral squamous cell carcinoma patients. Br J Med Med Res, 10(6), 1-9.  Zahid, T., Hussain, S.I., Siddiqui, A.H., Junaid, S.T., Ahmed, Z., & Noorani, M.M. (2014). Health seeking behavior of oral cancer patients of low socioeconomic status: A cross sectional study in a tertiary care hospital of Karachi. J Dow Univ Health Sci, 8(2), 72-79.  Zaidi, S.H.M. & Jafarey, N.A. (1966). Cancer problem in Pakistan with special reference to oropharyngeal carcinoma. J Pak Med Assoc, 14(2), 44-52.  Zaidi, S.M.H., Jafarey, N.A., & Ali, S.A. (1974). Cancer trends in Karachi. J Pak Med Assoc, 24, 84-92.  Zakiullah, Ahmadullah, Khisroon, M., Saeed, M., Khan, A., Khunda, F., et al. (2015). Genetic susceptibility to oral cancer due to combined effects of GSTT1, GSTM1 and CYP1A1 gene variants in tobacco addicted patients of pashtun ethnicity of Khyber Pakhtunkhwa province of Pakistan. Asian Pac J Can Prev, 16(3), 1145-1150.  Zakiullah, Saeed, M., Ali, S., Khisroon, M., Muhammad, B., Khuda, F., et al. (2014). Genetic susceptibility to esophageal cancer due to CYP1A1 gene variant rs4646903 in tobacco addicted patients of Pashtun ethnicity: a case-control study in Khyber Pakhtunkhwa province of Pakistan. Asian Pac J Can Prev, 15(16), 6715-6720.  Zhang, Z.J., Hao, K., Shi, R., Zhao, G., Jiang, G.X., Song, Y., et al. (2011). Gluthione S- Transferase M1 (GSTM1) and Glutathione S-Transferase T1 (GSTT1) Null Polymorphisms, Smoking, and Their Interraction in Oral Cancer: A HuGE Review and Meta-Analysis. Am J Epidemiol, 173(8), 847-857.  Zhou, S.F., Liu, J.P., & Chowbay, B. (2009). Polymorphism of human cytochrome P450 enzymes and its clinical impact. Drug Metab Rev, 41(2), 89-295.  Zushi, Y., Narisawa-Saitom M., Noguchi, K., Yoshimatsu, Y., Yugawa, T., Egawa, N., et al. (2011). An in vitro multistep carcinogenesis model for both HPV-positive and negative human oral squamous ell carcinoma. Am J Cancer Res, 1(7), 869-881.

145

 Zygogianni, A., Kyrgias, G., Mystakidou, K., Antypas, C., Kouvaris, J., Papadimitriou, C., et al. (2011). Potential role of the alcohol and smoking in the squamous cell carcinoma of the head and neck: review of the current literature and new perspectives. Asian Pac J Can Prev, 12(2), 339-344.

146

9. APPENDICES

Appendix Image I

Ladder 100bp GSTM1 219bp

GSTM1 219bp gel image after electrophoresis. From left to right genotypes were interpreted as: not null, not null, repeat, not null, not null, null, not null, not null, not null, not null, repeat, not null, not null, not null & not null

147

Appendix Image II

Ladder GSTT1 480bp 100bp

GSTT1 480bp gel image after electrophoresis. From left to right genotypes were interpreted as: Not null, not null, not null, not null, not null, null, not null, not null, null & not null

148

Appendix Image III

Ladder CYP1A1 340bp 100bp 200bp 140bp

CYP1A1 MspI 340bp gell image after electrophoresis of PCR products treated by the restriction enzyme. From left to right polymorphisms were interpreted as: m1/m2, m1/m2, m1/m1. m1/m2, m1/m2, m1/m2, m1/m2, m1/m2,m1/m2, m1/m2, m1/m1, m1/m2, m1/m2 & m1/m1.

149

Appendix Table I - Cancer

ID Sex Age Ethnicity Tobacco Duration Freq./ Site Size Grade Stage CYP1A1 GSTM1 GSTT1 habit in years day Non- 1 F 62 Punjabi Chewer 35 5 Lips 2 II II m2/m2 null null Mixed Non- Non- 2 M 45 Urdu habit 25 17 Tongue 1.5 II II m2/m2 null null Non- 3 M 38 Urdu Chewer 7 10 Tongue 2 I IV m1/m2 null null Non- Non- 4 F 78 Sindhi No habit -- -- Cheek 3.4 II IV m1/m1 null null Non- Non- 5 M 60 Urdu Chewer 15 6 Alveolar 4 I IV m1/m2 null null Non- Non- 6 M 35 Baloch Chewer 25 10 Cheek 4 I III m1/m2 null null Non- 7 F 55 Baloch Chewer 30 8 Cheek 5 I III m1/m2 null null Mixed Non- Non- 8 M 72 Memon habit 60 6 Cheek 4 II IV m2/m2 null null Non- 9 F 75 Urdu No habit -- -- Tongue 2 III III m1/m1 null null Non- Non- 10 M 58 Baloch Chewer 26 30 Alveolar 6 II IV m1/m2 null null Non- 11 M 35 Urdu Chewer 10 10 Cheek 5 II IV m1/m2 null null Non- Non- 12 M 40 Memon Chewer 38 4 Lips 3 I III m1/m2 null null Non- Non- 13 M 40 Urdu Chewer 10 12 Cheek 6.6 II IV m1/m2 null null Non- Non- 14 M 35 Urdu Chewer 20 30 Cheek 7 II IV m1/m2 null null Mixed Non- Non- 15 M 50 Urdu habit 30 17 Cheek 5 I III m1/m2 null null -- -- Non- Non- 16 F 46 Memon No habit Tongue 3 II II m1/m2 null null -- -- Non- Non- 17 F 49 Urdu No habit Tongue 4.5 II IV m1/m1 null null -- -- Non- Non- 18 M 62 Pathan No habit Tongue 4 III IV m2/m2 null null -- -- Non- 19 F 50 Urdu No habit Cheek 3.5 II IV m1/m1 null null Mixed Non- Non- 20 M 56 Urdu habit 20 12 Cheek 6.3 I IV m1/m2 null null Non- Non- 21 M 30 Urdu Chewer 4 6 Tongue 2 II II m1/m2 null null Mixed Non- 22 M 40 Balooch habit 20 20 Tongue 2 II II m1/m2 null null Marwari Non- Non- 23 M 35 (others) Chewer 1 2 Cheek 3.7 I III m1/m2 null null Non- 24 M 26 Urdu Chewer 6 15 Cheek 4 II III m1/m2 null null Non- Non- 25 M 55 Memon Chewer 1 2 Tonsil 4 I IV m1/m2 null null Non- Non- 26 M 30 Urdu Chewer 8 6 Cheek 4.7 II III m1/m2 null null Non- Non- 27 M 35 Baloch Chewer 5 7 Cheek 2.5 II II m1/m2 null null Non- 28 M 38 Urdu Smoker 8 20 Tongue 7.8 II III m1/m1 null null Non- 29 M 62 Urdu Chewer 20 10 Cheek 3 III II m2/m2 null null Non- Non- 30 M 22 Urdu Chewer 4 4 Cheek 7.5 II III m1/m2 null null

Table I cont…

150

Appendix Table I Cancer

ID Sex Age Ethnicity Tobacco Duration Freq./ Site Size Grade Stage CYP1A1 GSTM1 GSTT1 habit in years day Mixed Non- 31 M 48 Urdu habit 25 15 Cheek 2 II II m1/m2 null null Machi Mixed Non- Non- 32 M 60 (others) habit 40 35 Cheek 4 II IV m2/m2 null null Non- Non- 33 M 40 Punjabi Chewer 15 12 Lips 0.4 I II m1/m2 null null 34 F 50 Memon Chewer 30 5 Lips 6 I IV m1/m2 null null Non- 35 F 35 Baloch No habit -- -- Tongue 1.5 I I m1/m2 null null Non- 36 F 70 Urdu Chewer 40 10 Cheek 2.5 II II m1/m2 null null Non- 37 F 62 Urdu Chewer 35 12 Cheek 2.5 II II m2/m2 null null Marvari Non- Non- 38 F 35 (Others) Chewer 20 12 Lips 7 I III m1/m2 null null Non- 39 F 35 Memon Chewer 25 30 Cheek 7 I III m1/m1 null null Non- Non- 40 F 60 Baloch Chewer 30 12 Lips 5 II III m1/m2 null null Non- Non- 41 F 70 Urdu Chewer 30 5 Lips 7 I III m1/m2 null null Burmi Non- Non- 42 F 50 (Others) No habit -- -- Cheek 4 II III m1/m2 null null Non- 43 F 50 Sindhi No habit -- -- Cheek 1.5 I I m1/m2 null null Non- 44 F 53 Memon Chewer 25 8 Cheek 5 II III m1/m2 null null Non- Non- 45 F 33 Pashto No habit -- --. Tongue 1.4 II IV m2/m2 null null Non- 46 F 34 Urdu Chewer 8 6 Tongue 3.5 I IV m1/m2 null null Non- Non- 47 F 50 Sindhi No habit -- -- Cheek 1 II I m2/m2 null null Non- Non- 48 F 50 Sindhi No habit -- -- Lips 5.3 I IV m1/m2 null null Mixed Non- Non- 49 M 42 Urdu habit 30 50 Tongue 3 I II m1/m2 null null Mixed Non- Non- 50 M 50 Urdu habit 25 22 Cheek 6 II IV m2/m2 null null Mixed Non- Non- 51 M 27 Punjabi habit 10 24 Cheek 3 II II m1/m1 null null Mixed Non- Non- 52 M 43 Urdu habit 12 28 Cheek 2 II II m1/m2 null null Non- 53 M 51 Urdu Smoker 25 20 Cheek 3 II II m1/m1 null null Mixed Non- 54 M 46 Urdu habit 25 8 Lips 3 II II m1/m2 null null Non- 55 M 53 Punjabi Smoker 25 20 Cheek 4 III III m1/m2 null null Non- 56 M 44 Memon Chewer 30 10 Tongue 5 II IV m1/m2 null null Mixed Non- Non- 57 M 45 Urdu habit 20 30 Lips 3 I II m1/m2 null null Non- Non- 58 M 25 Urdu Chewer 15 25 Lips 4 I III m1/m2 null null Mixed 59 M 55 Urdu habit 25 20 Cheek 2.5 I II m1/m2 null null Non- Non- 60 M 30 Sindhi Smoker 5 7 Cheek 3 II II m1/m2 null null Table I cont…

151

Appendix Table I Cancer

ID Sex Age Ethnicity Tobacco Duration Freq./ Site Size Grade Stage CYP1A1 GSTM1 GSTT1 habit in years day Mixed Non- Non- 61 M 25 Baloch habit 15 16 Cheek 2.5 II III m1/m2 null null Non- 62 M 25 Sindhi Smoker 10 10 Lips 5.5 I IV m2/m2 null null Non- 63 M 35 Baloch Chewer 15 10 Tongue 2 I II m1/m2 null null Mixed Non- Non- 64 M 43 Urdu habit 30 50 Cheek 5.5 I III m1/m2 null null Mixed Non- Non- 65 M 50 Urdu habit 40 10 Cheek 8 I IV m1/m2 null null Mixed Non- Non- 66 M 40 Urdu habit 15 25 Cheek 2.5 I II m1/m2 null null Non- Non- 67 M 45 Sindhi Chewer 40 2 Cheek 1.3 I I m1/m2 null null Mixed 68 M 36 Urdu habit 15 13 Cheek 8.5 I IV m2/m2 null null Non- Non- 69 M 57 Sindhi Chewer 15 10 Alveolar 2.5 I II m1/m2 null null Non- Non- 70 M 54 Urdu No habit -- -- Tongue 3.5 I II m1/m2 null null Non- Non- 71 M 42 Urdu Chewer 27 27 Cheek 3 I II m1/m2 null null Non- Non- 72 M 60 Urdu Chewer 40 30 Cheek 3 II IV m1/m2 null null Mixed Non- Non- 73 M 57 Urdu habit 20 31 Cheek 3.8 II II m1/m1 null null Non- Non- 74 F 50 Memon No habit -- -- Tongue 2 I II m1/m2 null null Non- Non- 75 F 55 Urdu Chewer 30 4 Cheek 5 II IV m1/m2 null null Non- Non- 76 F 60 Memon Chewer 40 2 Cheek 3.5 I III m1/m1 null null Non- 77 F 60 Sindhi Chewer 20 5 Cheek 2.2 II III m1/m1 null null Non- Non- 78 F 50 Baloch Chewer 5 4 Cheek 2 II IV m1/m2 null null Non- 79 F 50 Sindhi Chewer 30 2 Lips 3 I III m1/m2 null null Non- Non- 80 F 45 Sindhi No habit -- -- Cheek 2 I I m1/m2 null null Non- 81 M 43 Sindhi Chewer 5 18 Tongue 5 I IV m1/m1 null null Mixed Non- 82 M 52 Urdu habit 30 30 Lips 3 II II m1/m1 null null Mixed Non- 83 M 50 Memon habit 30 30 Cheek 4 I III m1/m1 null null Mixed Non- Non- 84 M 40 Memon habit 25 45 Cheek 3 II II m1/m1 null null Mixed Non- Non- 85 M 40 Urdu habit 12 18 Cheek 3 I II m1/m1 null null Mixed 86 M 45 Urdu habit 9 14 Cheek 4 I III m1/m1 null null Non- Non- 87 M 29 Punjabi Chewer 3 3 Tongue 3 II IV m1/m1 null null 88 M 50 Urdu Chewer 25 24 Cheek 4 I III m1/m2 null null Non- 89 M 42 Memon Chewer 15 16 Cheek 3 II II m1/m2 null null Non- 90 M 51 Urdu Chewer 20 20 Tongue 4 II III m1/m1 null null Table I cont…

152

Appendix Table I Cancer

ID Sex Age Ethni- Tobacco Durat- Freq./ Site Size Grade Stage CYP1A1 GSTM1 GSTT1 city habit ion yrs day Non- Non- 91 M 52 Urdu Chewer 25 8 Tongue 3 I IV m2/m2 null null Non- 92 M 38 Sindhi No habit -- -- Cheek 5 II IV m1/m2 null null Non- 93 M 52 Pathan Chewer 30 70 Cheek 3 II II m2/m2 null null Mixed Non- Non- 94 M 55 Urdu habit 40 15 Alveolar 1.5 II I m1/m2 null null Non- 95 M 45 Urdu Chewer 6 10 Cheek 0.5 I I m1/m1 null null Saraiki Non- Non- 96 M 27 (Other) Chewer 12 20 Lips 2.8 II IV m1/m1 null null Non- Non- 97 M 40 Punjabi Smoker 20 20 Cheek 3 I III m1/m2 null null Non- Non- 98 M 42 Urdu Chewer 15 10 Cheek 3 II II m1/m2 null null Non- Non- 99 M 32 Punjabi Chewer 10 5 Cheek 4 I III m1/m2 null null Non- Non- 100 M 38 Urdu Chewer 20 10 Cheek 5 II III m1/m2 null null Non- 101 M 50 Urdu Chewer 5 5 Cheek 1 II I m1/m2 null null Non- Non- 102 M 32 Urdu Chewer 5 15 Tongue 1.5 II I m1/m1 null null Non- 103 M 55 Memon Chewer 8 12 Alveolar 2 II II m1/m2 null null Non- Non- 104 M 50 Baloch Chewer 8 7 Alveolar 2.5 II II m1/m2 null null Non- 105 M 40 Urdu Chewer 30 12 Cheek 4 II III m1/m2 null null Non- Non- 106 M 20 Sindhi Chewer 10 14 Cheek 2 II II m1/m2 null null Non- Non- 107 M 40 Urdu Chewer 30 12 Lips 1 II I m1/m2 null null Non- Non- 108 M 38 Baloch Chewer 18 15 Cheek 3.5 II II m2/m2 null null Non- Non- 109 M 38 Urdu Chewer 20 20 Cheek 3 II II m1/m2 null null Mixed Floor of Non- 110 M 65 Urdu habit 15 20 mouth 3 II II m1/m2 null null Non- Non- 111 F 60 Memon Chewer 35 2 Cheek 4 II IV m2/m2 null null Non- Non- 112 F 50 Pathan No habit -- -- Lips 1.5 II I m1/m1 null null Non- 113 F 45 Sindhi Chewer 30 12 Cheek 7 II IV m1/m2 null null Non- Non- 114 F 60 Sindhi No habit -- -- Cheek 2.5 I II m1/m2 null null Non- Non- 115 F 50 Urdu Chewer 30 5 Alveolar 2 I II m1/m2 null null Non- Non- 116 F 30 Urdu Chewer 1 6 Alveolar 1 II I m1/m1 null null Non- Non- 117 F 50 Memon No habit -- -- Cheek 2 II I m1/m2 null null Non- Non- 118 F 70 Urdu Chewer 25 5 Cheek 5 II IV m1/m1 null null Non- Non- 119 F 60 Memon Chewer 30 12 Cheek 2.5 I II m2/m2 null null Non- Non- 120 F 45 Urdu C hewer 20 10 Cheek 3.5 III II m2/m2 null null Table I cont…

153

Appendix Table I Cancer

ID Sex Age Ethni- Tobacco Durat- Freq./ Site Size Grade Stage CYP1A1 GSTM1 GSTT1 city habit ion yrs day Non- Non- 121 F 50 Urdu Chewer 20 12 Cheek 1.5 II I m1/m2 null null Non- Non- 122 F 45 Urdu Chewer 30 9 Cheek 3 II II m1/m2 null null Non- Non- 123 F 25 Pathan Chewer 10 3 Tongue 1.4 II I m1/m2 null null Non- 124 F 45 Pathan Chewer 5 10 Tongue 2 II IV m2/m2 null null Non- Non- 125 F 50 Urdu Chewer 8 5 Palate 6 II III m1/m2 null null Non- 126 F 20 Sindhi No habit -- -- Tongue 3 II IV m1/m1 null null Non- 127 F 60 Sindhi No habit -- -- Tongue 4 II IV m1/m1 null null Non- 128 F 40 Urdu No habit -- -- Tongue 2 II II m1/m2 null null Non- 129 F 65 Punjabi Chewer 40 24 Lips 1.5 II I m2/m2 null null Floor of Non- 130 F 60 Baloch Chewer 40 25 mouth 2.5 II II m1/m2 null null Non- Non- 131 F 65 Urdu No habit -- -- Tongue 2 II II m1/m1 null null Non- Non- 132 F 55 Urdu No habit -- -- Tongue 1.6 II IV m1/m2 null null Non- Non- 133 F 55 Baloch Chewer 30 6 Cheek 1.5 II I m2/m2 null null Non- Non- 134 M 47 Urdu Chewer 15 6 Cheek 3.5 I II m1/m2 null null Non- Non- 135 M 70 Sindhi No habit -- -- Tongue 2.2 II III m1/m1 null null Non- 136 M 52 Urdu Chewer 12 3 Tongue 3.5 II IV m2/m2 null null Retromo Non- Non- 137 M 54 Urdu Smoker 20 20 lar area 2.2 I II m1/m2 null null Mixed Non- 138 M 70 Urdu habit 25 52 Cheek 1.5 II I m1/m2 null null Non- 139 M 45 Baloch Chewer 25 50 Cheek 10 I IV m2/m2 null null Mixed Non- Non- 140 M 40 Urdu habit 25 15 Cheek 7 II IV m1/m2 null null Non- 141 M 60 Urdu Chewer 20 12 Cheek 7 III IV m2/m2 null null Potohar Non- 142 M 65 (Other) Smoker 40 20 Palate 2 II I m1/m1 null null Non- 143 M 50 Urdu Chewer 15 20 Cheek 2.5 II IV m1/m2 null null Non- 144 M 65 Baloch Chewer 20 12 Cheek 5 II III m2/m2 null null Non- Non- 145 M 40 Urdu Chewer 20 20 Cheek 2.2 II II m1/m2 null null Non- 146 M 35 Baloch Chewer 3 20 Alveolar 2.5 III II m2/m2 null null Non- 147 M 32 Urdu Chewer 12 30 Cheek 3 I II m1/m2 null null Non- Non- 148 M 60 Urdu Chewer 25 10 Lips 2 I II m2/m2 null null Non- Non- 149 M 28 Sindhi Chewer 2 12 Cheek 4 II III m2/m2 null null Mixed Non- 150 M 57 Pathan habit 35 32 Cheek 2.2 I II m1/m2 null null

154

Appendix Table II Precancerous lesions

Type of Dura- Freq. Site of ID Lesion Sex Age Ethnicity exposure tion /day lesion Size CYP1A1 GSTM1 GSTT1 Non- 1 SMF Male 26 Gujrati Chewer 6 12 -- -- m1/m2 Non-null null Non- 2 Leucoplakia Male 36 Urdu Chewer 20 12 Cheek 2 m1/m2 Non-null null Non- 3 SMF Male 30 Gujrati Chewer 5 5 -- -- m1/m2 Non-null null Non- 4 SMF Male 68 Gujrati Mixed habit 35 40 -- -- m1/m2 Non-null null Non- 5 SMF Female 60 Gujrati Chewer 30 4 -- -- m1/m2 Non-null null

6 Leucoplakia Male 30 Urdu Chewer 17 20 Cheek 3 m2/m2 Null Null

7 SMF Male 40 Urdu Chewer 25 15 -- -- m1/m1 Non-null null Non- 8 Erythoplakia Male 19 Sindhi Chewer 10 7 Cheek 1.5 m1/m2 null null Non- 9 SMF Male 36 Urdu Chewer 22 40 -- -- m1/m2 Non-null null Non- 10 SMF Male 18 Urdu Chewer 6 6 -- -- m1/m1 null null Non- 11 Leucoplakia Male 27 Urdu Chewer 16 9 Cheek 1 m1/m2 Non-null null Non- 12 SMF Male 19 Urdu Chewer 6 30 -- -- m1/m2 null null Non- 13 Leucoplakia Male 31 Urdu Chewer 10 30 Cheek 1 m1/m2 Non-null null Non- 14 SMF Male 19 Urdu Chewer 5 14 -- -- m1/m2 Non-null null Non- 15 SMF Male 50 Urdu Chewer 40 30 -- -- m1/m2 Non-null null

16 SMF Male 35 Urdu Chewer 15 20 -- -- m2/m2 Non-null null Non- 17 Leucoplakia Male 32 Urdu Chewer 15 3 Cheek 1.5 m1/m2 Non-null null Non- 18 SMF Male 40 Urdu Chewer 25 10 -- -- m1/m2 Non-null null Non- 19 Leucoplakia Female 35 Urdu Chewer 12 7 Cheek 2 m1/m1 Non-null null Non- 20 Leucoplakia Male 38 Urdu Chewer 5 6 Cheek 2 m1/m2 Non-null null Non- 21 Leucoplakia Female 60 Urdu Chewer 40 6 Cheek 1.2 m1/m2 Non-null null Non- 22 Leucoplakia Male 30 Urdu Chewer 15 20 Cheek 2.5 m1/m2 Non-null null Non- 23 SMF Male 21 Urdu Chewer 8 10 -- -- m1/m2 Non-null null Non- 24 SMF Male 42 Urdu Chewer 25 20 -- -- m1/m2 Non-null null Non- 25 Leucoplakia Male 37 Urdu Chewer 20 20 Cheek 2 m1/m2 Non-null null Non- 26 Leucoplakia Male 46 Urdu Chewer 25 12 Cheek 1 m1/m2 Non-null null Non- 27 SMF Female 55 Urdu Chewer 25 5 -- -- m1/m2 Non-null null Non- 28 SMF Male 25 Urdu Chewer 9 15 -- -- m1/m2 Non-null null Table II cont…

155

Appendix Table II Precancerous lesions

Type of Dura- Freq. Site of ID Lesion Sex Age Ethnicity exposure tion /day lesion Size CYP1A1 GSTM1 GSTT1 Non- 29 SMF Male 24 Urdu Chewer 4 10 -- -- m2/m2 Non-null null Non- 30 SMF Male 30 Urdu Chewer 25 12 -- -- m2/m2 Non-null null Non- 31 SMF Male 30 Urdu Chewer 15 16 -- -- m1/m2 null null Non- 32 Leucoplakia Male 24 Hazara Chewer 14 30 Cheek 0.5 m1/m2 Non-null null Non- 33 SMF Male 35 Hazara Chewer 15 35 -- -- m1/m2 null null Non- 34 SMF Male 20 Hazara Chewer 3 35 -- -- m1/m2 null null Non- 35 SMF Male 17 Pathan Chewer 5 12 -- -- m1/m2 Non-null null

36 SMF Male 78 Urdu Chewer 20 5 -- -- m1/m1 Non-null null Non- 37 SMF Male 28 Pathan Chewer 8 24 -- -- m1/m2 Non-null null Non- 38 SMF Male 39 Pathan Chewer 11 30 -- -- m1/m2 null null

39 SMF Male 24 Pathan Chewer 10 30 -- -- m1/m2 null null Non- 40 SMF Male 20 Pathan Mixed habit 5 30 -- -- m1/m2 Non-null null Non- 41 SMF Male 20 Pathan Chewer 3 30 -- -- m1/m2 Non-null null Non- 42 SMF Male 24 Hazara Mixed habit 6 20 -- -- m1/m2 Non-null null

43 SMF Male 24 Pathan Mixed habit 12 30 -- -- m1/m2 null null Non- 44 SMF Male 30 Urdu Chewer 15 18 -- -- m1/m1 Non-null null Non- 45 Leucoplakia Male 50 Baloch Smoker 12 14 Cheek 1.5 m1/m2 Non-null null Non- 46 SMF Male 40 Urdu Chewer 26 35 -- -- m1/m2 null null Non- 47 Leucoplakia Male 30 Sindhi Chewer 5 5 Cheek 2 m1/m2 Non-null null Non- 48 SMF Male 24 Urdu Chewer 15 12 -- -- m1/m2 Non-null null Non- 49 SMF Female 40 Sindhi Chewer 10 5 -- -- m1/m2 null null Non- 50 SMF Male 22 Sindhi Mixed habit 5 5 -- -- m1/m2 null null Non- 51 SMF Male 45 Sindhi Chewer 2 5 -- -- m1/m1 null null Non- 52 Erythoplakia Male 22 Pathan Chewer 7 5 Tongue 1.5 m2/m2 null null Non- 53 SMF Male 49 Urdu Chewer 25 5 -- -- m1/m2 null null Non- 54 SMF Female 24 Memon No habit 0 0 -- -- m1/m2 Non-null null Non- 55 SMF Male 19 Urdu Chewer 15 10 -- -- m1/m1 null null Table II cont…

156

Appendix Table II Precancerous lesions

Type of Dura- Freq. Site of ID Lesion Sex Age Ethnicity exposure tion /day lesion Size CYP1A1 GSTM1 GSTT1 Non- 56 SMF Male 41 Urdu Chewer 15 12 -- -- m1/m2 Non-null null Non- 57 SMF Female 40 Urdu Chewer 25 14 -- -- m1/m1 null null Non- 58 SMF Male 35 Urdu Chewer 15 10 -- -- m1/m2 Non-null null Non- 59 SMF Male 50 Balochi Chewer 20 5 -- -- m1/m2 Non-null null Hazara Non- 60 Leucoplakia Male 20 (Others) Chewer 10 12 Cheek 1.5 m1/m2 Non-null null Non- 61 SMF Male 39 Urdu Mixed habit 15 20 -- -- m1/m1 Non-null null Non- 62 Leucoplakia Female 24 Baloch Chewer 15 20 Cheek 4 m1/m2 Non-null null Non- 63 SMF Male 45 Pathan Chewer 10 15 -- -- m1/m1 null null Non- 64 SMF Male 17 Pathan Mixed habit 1 12 -- -- m2/m2 null null Non- 65 Leucoplakia Female 55 Urdu No habit 0 0 Cheek 1 m1/m1 null null Non- 66 SMF Female 65 Urdu Chewer 35 4 -- -- m1/m1 Non-null null Non- 67 SMF Female 20 Memon No habit 0 0 -- -- m1/m2 Non-null null Non- 68 Leucoplakia Male 43 Urdu Chewer 30 15 Cheek 3 m1/m2 Non-null null Non- 69 Erythoplakia Male 55 Sindhi Smoker 25 10 Cheek 2 m1/m2 null null Non- 70 SMF Female 45 Sindhi Chewer 25 6 -- -- m2/m2 null null Non- 71 SMF Female 16 Urdu Chewer 5 10 -- -- m1/m2 null null Non- 72 Erythoplakia Female 40 Baloch Chewer 25 5 Cheek 4 m2/m2 Non-null null Non- 73 SMF Male 29 Urdu Chewer 30 10 -- -- m1/m2 null null Non- 74 SMF Male 29 Urdu Chewer 9 25 -- -- m2/m2 null null Non- 75 SMF Male 65 Baloch Chewer 30 10 -- -- m1/m2 null null Non- 76 SMF Female 30 Sindhi Chewer 3 5 -- -- m1/m1 null null Non- 77 Leucoplakia Male 70 Sindhi Smoker 20 20 Tongue 1.5 m1/m2 Non-null null Non- 78 SMF Female 25 Urdu Chewer 6 5 -- -- m2/m2 null null

79 SMF Male 20 Urdu Chewer 2 29 -- -- m1/m2 null null Non- 80 SMF Male 30 Urdu Chewer 12 12 -- -- m1/m2 null null Non- 81 Leucoplakia Male 30 Urdu Chewer 10 20 Cheek 4 m1/m1 Non-null null Non- 82 Leucoplakia Male 30 Urdu Chewer 10 24 Cheek 4 m1/m1 Non-null null Table II cont…

157

Appendix Table II Precancerous lesions

Type of Dura- Freq. Site of ID Lesion Sex Age Ethnicity exposure tion /day lesion Size CYP1A1 GSTM1 GSTT1 Hard Non- 83 SMF Male 30 Urdu Chewer 10 16 palate 1.5 m1/m2 Non-null null

84 Leucoplakia Female 60 Urdu No habit 0 0 -- -- m2/m2 Non-null null Non- 85 SMF Male 22 Urdu Chewer 12 50 -- -- m1/m2 null null Non- 86 SMF Male 25 Urdu Chewer 10 10 -- -- m1/m2 Non-null null Non- 87 SMF Male 17 Urdu Chewer 4 6 -- -- m1/m2 null null Non- 88 SMF Male 21 Urdu Chewer 6 5 Cheek 2 m1/m2 null null

89 SMF Female 30 Sindhi Chewer 9 8 -- -- m1/m2 null null Non- 90 SMF Female 35 Urdu Chewer 10 20 -- -- m1/m1 Non-null null Non- 91 Erythoplakia Male 26 Urdu Chewer 6 15 Cheek 1.5 m1/m2 null null Non- 92 Leucoplakia Male 45 Pathan Chewer 4 25 Alveolus 1.5 m2/m2 null null Non- 93 Leucoplakia Male 56 Urdu Chewer 20 6 Cheek 1.5 m1/m2 Non-null null Mixed Non- 94 SMF Male 20 Gujrati habit 10 20 -- -- m2/m2 Non-null null Non- 95 Leucoplakia Male 50 Urdu Chewer 32 20 Cheek 1 m1/m2 Non-null null Non- 96 SMF Male 30 Urdu Chewer 15 10 -- -- m1/m1 Non-null null Non- 97 Leucoplakia Male 24 Pathan Chewer 5 12 Cheek 0.5 m1/m2 Non-null null Non- 98 Leucoplakia Male 28 Memon Chewer 12 6 Cheek 2 m2/m2 Non-null null Non- 99 SMF Male 27 Urdu Chewer 5 5 -- -- m2/m2 Non-null null Non- 100 Leucoplakia Female 26 Gujrati Chewer 6 8 Cheek 2.5 m2/m2 null null

158

Appendix Table III Controls

Duration Freq./ ID Sex Age Ethnicity Tobacco habit In years day CYP1A1 GSTM1 GSTT1

1 Male 30 Gujrati Chewer 4 12 m1/m2 Non-null Non-null

2 Male 54 Kathiawari Mixed habit 34 8 m1/m1 Null Non-null

3 Female 60 Gujrati Chewer 40 10 m1/m2 Non-null Non-null

4 Female 60 Gujrati Chewer 40 4 m1/m2 Non-null Non-null

5 Male 40 Gujrati Chewer 4 15 m1/m2 Non-null Non-null

6 Male 24 Urdu Chewer 4 30 m1/m2 Non-null Non-null

7 Male 20 Urdu Mixed habit 5 12 m1/m2 Non-null Non-null

8 Male 22 Urdu Mixed habit 7 20 m1/m1 Null Non-null

9 Male 32 Urdu Chewer 20 50 m1/m2 Non-null Non-null

10 Male 22 Gujrati Chewer 10 5 m1/m2 Null Non-null

11 Male 28 Urdu Chewer 6 15 m1/m2 Non-null Non-null

12 Male 27 Urdu Mixed habit 13 20 m1/m1 Null Non-null

13 Male 28 Urdu Chewer 5 8 m1/m2 Non-null null

14 Male 56 Urdu Chewer 40 25 m1/m2 Non-null Non-null

15 Male 62 Urdu Chewer 40 20 m1/m2 Non-null Non-null

16 Male 59 Urdu Mixed habit 30 40 m1/m1 Null Non-null

17 Female 32 Urdu Chewer 17 40 m1/m1 Non-null Non-null

18 Male 37 Urdu Chewer 17 24 m2/m2 Non-null Non-null

19 Male 30 Urdu Chewer 5 5 m1/m2 Non-null Non-null

20 Male 20 Pathan Chewer 6 5 m1/m1 Null Non-null

21 Female 38 Baloch Chewer 22 12 m1/m1 Null Non-null

22 Male 60 Urdu Mixed habit 34 6 m1/m2 Non-null Non-null

23 Male 56 Urdu Smoker 35 4 m1/m1 Non-null Non-null Table III cont…..

159

Appendix Table III Controls

Duration Freq./ ID Sex Age Ethnicity Tobacco habit In years day CYP1A1 GSTM1 GSTT1

24 Male 15 Saraiki Mixed habit 3 20 m1/m2 Null Non-null

25 Male 60 Urdu Mixed habit 30 41 m1/m2 Non-null Non-null

26 Male 60 Urdu Smoker 15 12 m2/m2 Null Non-null

27 Male 30 Sindhi Chewer 15 12 m1/m1 Null Non-null

28 Male 32 Memon Chewer 25 20 m1/m2 Null Non-null

29 Male 43 Urdu Chewer 8 20 m1/m1 Null Non-null

30 Male 34 Urdu Chewer 14 40 m1/m2 Null Non-null

31 Male 62 Urdu Chewer 45 15 m1/m2 Non-null Non-null

32 Male 30 Pathan Chewer 20 25 m1/m2 Null Non-null

33 Male 40 Urdu Chewer 25 25 m1/m1 Null Non-null

34 Male 32 Urdu Chewer 10 40 m1/m2 Non-null Non-null

35 Male 48 Urdu Chewer 20 15 m1/m2 Null Non-null

36 Male 56 Urdu Chewer 12 9 m1/m1 Non-null Non-null

37 Female 45 Urdu Chewer 20 12 m1/m2 Non-null Non-null

38 Female 40 Urdu Chewer 20 10 m1/m2 Non-null Non-null

39 Female 55 Baloch Chewer 20 12 m1/m2 Null Non-null

40 Female 40 Gugrati Chewer 22 5 m1/m1 Null Non-null

41 Male 64 Urdu Smoker 30 20 m1/m2 Non-null Non-null

42 Female 30 Baloch Smoker 6 5 m2/m2 Null Non-null

43 Male 21 Baloch Chewer 11 20 m1/m2 Null Non-null

44 Male 28 Punjabi Chewer 10 12 m1/m2 Non-null Non-null

45 Male 37 Gujrati Mixed habit 25 8 m1/m1 Non-null Non-null

46 Male 23 Baloch Chewer 13 6 m1/m2 Non-null Non-null

Table III cont…..

160

Appendix Table III Controls

Duration Freq./ ID Sex Age Ethnicity Tobacco habit In years day CYP1A1 GSTM1 GSTT1

47 Male 32 Urdu Chewer 13 24 m1/m2 Non-null Non-null

48 Male 35 Punjabi Chewer 9 24 m1/m2 Non-null Non-null

49 Male 27 Punjabi Chewer 7 28 m1/m2 Non-null Non-null

50 Male 30 Memon Mixed habit 18 63 m1/m2 Non-null Non-null

51 Female 45 Baloch Chewer 20 18 m1/m1 Null Non-null

52 Female 35 Memon Chewer 5 4 m2/m2 Null Non-null

53 Female 45 Baloch Chewer 34 10 m1/m2 Non-null Non-null

54 Male 37 Memon Mixed habit 20 34 m1/m2 Null Non-null

55 Male 40 Gujrati Chewer 20 7 m1/m2 Non-null Non-null

56 Male 60 Sindhi Smoker 24 24 m1/m2 Null Non-null

57 Female 35 Memon Chewer 20 10 m1/m1 Non-null Non-null

58 Female 50 Baloch Smoker 30 10 m2/m2 Null Non-null

59 Female 40 memon Chewer 10 4 m2/m2 Non-null Non-null

60 Male 35 Baloch Chewer 20 12 m1/m2 Non-null Non-null

61 Male 30 Urdu Chewer 9 30 m1/m2 Non-null Non-null

62 Male 17 Urdu Chewer 2 10 m1/m2 Non-null Non-null

63 Male 20 Urdu Chewer 6 10 m1/m1 Non-null Non-null

64 Female 20 Sindhi Chewer 2 15 m2/m2 Non-null Non-null

65 Female 70 Sindhi Mixed habit 50 24 m1/m1 Null Non-null

66 Male 56 Urdu Smoker 30 10 m2/m2 Non-null Non-null

67 Male 85 Urdu Mixed habit 40 10 m2/m2 Non-null Non-null

68 Female 55 Punjabi Chewer 20 5 m1/m2 Non-null Non-null

69 Male 38 Urdu Mixed habit 20 7 m1/m2 Non-null Non-null

Table III cont…..

161

Appendix Table III Controls

Duration Freq./ ID Sex Age Ethnicity Tobacco habit In years day CYP1A1 GSTM1 GSTT1

70 Male 36 Urdu Mixed habit 20 35 m1/m1 Non-null Non-null

71 Male 32 Urdu Chewer 15 15 m1/m2 Non-null Non-null

72 Male 60 Urdu Chewer 30 7 m1/m2 Non-null Non-null

73 Female 45 Urdu Chewer 10 3 m1/m1 Non-null Non-null

74 Male 42 pathan Chewer 12 20 m1/m1 Null Non-null

75 Male 39 Urdu Mixed habit 15 15 m1/m2 Non-null Non-null

76 Male 33 Pathan Chewer 12 35 m1/m2 Non-null Null

77 Male 46 pathan Chewer 15 5 m1/m2 Non-null Non-null

78 Male 43 Urdu Mixed habit 20 12 m1/m2 Non-null Non-null

79 Male 53 Urdu Chewer 30 8 m1/m1 Non-null Non-null

80 Male 40 Sindhi Chewer 25 5 m1/m1 Null Non-null

81 Male 32 Baloch Chewer 12 30 m1/m2 Non-null Non-null

82 Female 40 memon Chewer 15 12 m2/m2 Non-null Non-null

83 Male 46 Urdu Chewer 20 10 m1/m2 Non-null Non-null

84 Female 36 Urdu Chewer 16 8 m1/m1 Non-null Non-null

85 Male 29 Urdu Mixed habit 15 30 m1/m2 Non-null Non-null

86 Male 34 Hindko Mixed habit 15 15 m1/m2 Non-null Non-null

87 Male 55 Urdu Chewer 45 12 m1/m2 Non-null Non-null

88 Female 42 Sindhi Chewer 15 10 m1/m2 Null Non-null

89 Female 52 Baloch Chewer 25 6 m1/m2 Non-null Non-null

90 Female 70 Urdu Chewer 30 5 m1/m1 Non-null Non-null

91 Male 87 Urdu Chewer 50 20 m1/m2 Non-null Non-null

92 Male 52 Urdu Smoker 20 20 m1/m2 Non-null Non-null

Table III cont…..

162

Appendix Table III Controls

Duration Freq./ ID Sex Age Ethnicity Tobacco habit In years day CYP1A1 GSTM1 GSTT1

93 Male 40 Hindko Chewer 8 10 m1/m2 Non-null Non-null

94 Male 60 Urdu Mixed Habit 30 20 m1/m2 Null Non-null

95 Male 69 Urdu Chewer 20 4 m1/m1 Non-null Non-null

96 Male 44 Baloch Chewer 30 10 m1/m2 Non-null Non-null

97 Male 45 Punjabi Mixed Habit 25 46 m1/m2 Non-null Non-null

98 Male 43 Urdu Mixed Habit 20 24 m1/m2 Non-null Non-null

99 Male 50 Urdu Chewer 40 20 m1/m2 Null Non-null

100 Female 60 Memon Chewer 25 10 m1/m2 Non-null Non-null

101 Male 40 Sindhi No habit m2/m2 Non-null Non-null

102 Female 40 Punjabi No habit m1/m1 Null Non-null

103 Male 37 Urdu No habit m1/m2 Null Non-null

104 Female 55 Punjabi No habit m1/m1 Null Non-null

105 Female 29 Urdu No habit m1/m2 Null Non-null

106 Male 24 Urdu No habit m1/m1 Non-null Non-null

107 Male 22 Pathan No habit m2/m2 Null Non-null

108 Male 27 Urdu No habit m1/m2 Non-null Non-null

163

PROFORMA

Case No. ------OPD/Bed No.------

Name:------Age:------Sex: ------

Occupation:------Ethnicity: ------

Socio-economic status: ------Date of Admission: ------

Address: ------Tel No:------

Tobacco Habit:

Substance Type Duration (yrs) Freq/day <5 5-10 10-15 15-20 20-25 25-30 >30 1-5 5-10 >10 Exclusive tobacco chewer Exclusive smoker Mixed habit (chewing + smoking) No habit

DIET:

Fruits and vegetables:------Hot and spicy foods:------

DENTITION/ ORAL SEPSIS:

Number/ Missing teeth:------Broken/ Pointed/ Jagged teeth:------

Dentures:------Sepsis:------

PRENEOPLASTIC LESIONS/ CONDITIONS:

Leukoplakia: Erythroplakia: Oral submucous fibrosis: Lichen planus: PAST HISTORY:

164

FAMILY HISTORY:

DRUG HISTORY:

PRESENTING COMPLAINTS:

Swelling/ Growth: Pain: Dysarthria/ Dysphonia/ Bleeding:Dysphagia: Ulcers/ Vesicles:Trismus: Salivation/ Burning sensation:

LOCAL EXAMINATION: Site/ Size/ Shape of the lesion: Infiltration/ Induration of margins: Overlying mucosa/ Ulceration/ Fibrosis: Orodental disease/ Associated lesions:

GENERAL EXAMINATION: Pulse: B.P: Anemia: Jaundice:

Wt. loss: Lymph node status:

SYSTEMIC EXAMINATION:

Respiratory system: Cardiovascular system: Gastrointestinal system: Nervous system: INVESTIGATIONS:

RADIOLOGY/ IMAGING:

LAB.INVESTIGATIONS:

BIOPSY:

Dx: Grade: Stage:

MOLECULAR ANALYSIS:

CYP1A1: GSTM1: GSTT1:

165

Consent form

INFORMED CONSENT

Genetic polymorphism at CYPIAI, GSTM1 and GSTT1 among various tobacco habit groups and susceptibility to oral precancerous lesions and squamous cell carcinoma

You are a regular tobacco user which can lead to many complications including oral cancers. You are taken as per your confidence. Your consent will be the major contribution for this research. We will collect 5 ml of your blood and tissue from your biopsy/resected tumor which will be sufficient to help us carry on the research.

Please help us fill the given information which is essential to undertake the study. Your signature on this form will be taken as positive consent.

I______S/o, D/o ______Understand that I am included voluntarily as a subject for the sake of obtaining scientific information for the above mentioned study. I will be asked a few questions regarding my health and physical examination will be conducted. 5 ml venous blood and tissue from your biopsy/resected tumor will be taken for this research and these samples will not be used for any other purpose.

Compensation: I have been told that there will be no monitory benefits for participation in this study. Confidentiality: I understand that my participation in the study will be kept confidential and that the results from my blood examination will also be kept confidential and conveyed to me privately if I would ask for it. Cost of the tests: I understand that the test being done purely for research purposes are free and I will not have to pay for them. Contact: I have been told that I can contact Dr. Mohiuddin Alamgir at phone#(021)99240002 Ext:353, Bahria University Medical & Dental College, Bahria University, 13-National Stadium road, PNS Karsaz, Karachi during office timings for any question regarding my participation.

I hereby confirm that I have read and understood whatever has been stated above and based on the same I am voluntarily giving consent for participation in the study.

Subject Sig: ______Date: ______Investigator Sig: ______Date______Name: ______Name:______

166

Consent form in native language Urdu

167

10. SIMILARITY INDEX BY TURNITIN

168

169

11. PUBLICATION

170

171

172

173

174