Cancer survival in Africa, Asia, and Central America: a population-based study.

BACKGROUND: Population-based cancer survival data, a key indicator for monitoring progress against cancer, are not widely available from countries in Africa, Asia, and Central America. The aim of this study is to describe and discuss cancer survival in these regions. METHODS: Survival analysis was done for 341 658 patients diagnosed with various cancers from 1990 to 2001 and followed up to 2003, from 25 population-based cancer registries in 12 countries in sub-Saharan Africa (The Gambia, Uganda), Central America (Costa Rica), and Asia (China, India, Pakistan, Philippines, Saudi Arabia, Singapore, South Korea, Thailand, Turkey). 5-year age-standardised relative survival (ASRS) and observed survival by clinical extent of disease were determined. FINDINGS: For cancers in which prognosis depends on stage at diagnosis, survival was highest in China, South Korea, Singapore, and Turkey and lowest in Uganda and The Gambia. 5-year ASRS ranged from 76-82% for breast cancer, 63-79% for cervical cancer, 71-78% for bladder cancer, and 44-60% for large-bowel cancers in China, Singapore, South Korea, and Turkey. Survival did not exceed 22% for any cancer site in The Gambia; in Uganda, survival did not exceed 13% for any cancer site except breast (46%). Variations in survival correlated with early detection initiatives and level of development of health services. INTERPRETATION: The wide variation in cancer survival between regions emphasises the need for urgent investments in improving awareness, population-based cancer registration, early detection programmes, health-services infrastructure, and human resources. FUNDING: Association for International Cancer Research (AICR; St Andrews, UK), Association pour la Recherche sur le Cancer (ARC, Villejuif, France), and the Bill & Melinda Gates Foundation (Seattle, USA).

Awareness of cancer risk factors among patients and attendants presenting to a tertiary care hospital in Karachi, Pakistan.

OBJECTIVE: To determine awareness of cancer risk factors in the patients and attendants of Out-patient Clinics at a University Hospital in Karachi, Pakistan. METHODS: A cross-sectional survey was conducted on 315 respondents reporting to a tertiary care hospital in Karachi, Pakistan, to assess their level of awareness regarding risk factors of cancer. RESULTS: The respondents belonged to an urban population with the mean time spent in Karachi of 29.1 years (SD +/- 13.94). There were 213 (67%) males and 102 (33%) females. All respondents had heard of the word 'cancer', while only 57.5% were aware of cancer risk factors. However, only 42.8% could identify age, 33% diet, 35% drugs and 31% obesity as risk factors for cancer. Even those who were aware of the risk factors were not able to appreciate personal risk of cancer. CONCLUSION: Despite awareness regarding some of the risk factors, the surveyed population was not aware of intrinsic risk factors for cancers like increasing age and obesity. It is important to create awareness through educational programs on cancer prevention, dissemination of knowledge pertaining to the preventable and avoidable cancer risk factors, the benefits of early diagnosis, and availability of screening tests.

Correlation of intra-operative frozen section consultation with the final diagnosis at a referral center in Karachi, Pakistan.

BACKGROUND: The correlation of intra-operative frozen section diagnosis with final diagnosis on permanent sections is an integral part of quality assurance in surgical pathology laboratories. However, there is scant data on this topic from Pakistan. Similarly, no local study has looked at frozen section turnaround times. AIMS AND OBJECTIVES: To analyze indications, discrepancies and deferrals for all frozen sections performed or received at our institution over a 1-year period and to determine the turnaround time for frozen section diagnoses in our cases. DESIGN: A retrospective study, was undertaken, of all frozen sections reported at our institution between 1 st January 2006 and 31 st December 2006. The records of these cases were reviewed. The number and types of discrepancies, including sampling and interpretation errors were determined. The deferred cases and causes for deferral were also determined. The turnaround times of all cases were recorded. Agreement rates were calculated as percent agreement, sensitivity/specificity and positive and negative predictive values. RESULTS: A total of 356 specimens were received. Out of these, 14 cases (3.93%) were deferred to permanent sections. Of the remaining 342 cases, the discordant diagnostic frequency was 2.92% while the concordant diagnostic frequency was 97.08%. The most common pathological processes encountered were presence/typing of neoplasm (51.12%) and assessment of surgical margins (27.53%). The average turn-around time for frozen section diagnosis was 23 minutes; 60% of the cases were reported in 20 minutes or less. CONCLUSIONS: The accuracy of frozen section diagnosis at AKUH pathology department can be interpreted as comparable with most international quality control statistics for frozen sections. The overall error and deferral rates are within the range of previously published errors in pathology. Deferrals and errors in some sub-specialties were higher than in others. The results suggest specific measures should be taken to reduce the number of discrepancies. The overall goal is to reduce errors, reduce the number of deferrals and improve frozen section diagnosis turnaround times.

Pediatric brain tumours at a tertiary care hospital in Karachi.

The objectives of this study were to determine the epidemiology of brain tumors during infancy and childhood and to define and segregate childhood brain tumors vis-a-vis their morphological characteristics. The present study includes pediatric brain tumors, ICD-10 category C71 encountered during 10 years (January 1989 through December 1998) at a tertiary care hospital in Karachi. Eighty one cases were included, 58 (71.6%) in males and 23 (28.4%) in females with a male to female ratio of 2.5:1. The cases were divided into 3 age groups each covering five years of life (0-4, 5-9, 10-14 years), with the greatest number in the second age group i.e. 5-9 years followed by the third age group and the 0-4 year age group. The mean age for all cases, both genders was 8.8 years (95% CI 7.9; 9.6) with a marginal variation for cases occurring in the cerebrum and cerebellum. The malignancies occurred at a younger age in the males for each subcategory by site and morphology. The morphological distribution of cases was astrocytoma (28 cases, 34.6%), primitive neuroectodermal tumor or PNET (40 cases; 49.4%), ependymoma (8 cases, 10%), mixed glioma (4 cases; 5%) and a case of oligodendroglioma. The 81 malignancies included in this study were further categorized by site into two groups, supratentorial (27 cases; 33.3%) and infratentorial (54 cases; 66.7%). The morphological categorization of supratentorial tumors was astrocytoma (17 cases; 63%), ependymoma (5 cases; 18.5%), mixed glioma (2 cases; 7.4%). PNET with rhabdoid differentiation, oligodendroglioma and pinealoblastoma comprised 1 case (3.7%) each. The 17 supratentorial astrocytoma were sub-categorized as follows - pilocytic astrocytoma (5 cases; 29.4%), grade II astrocytoma (6 cases; 35.3%); grade III astrocytoma (2 cases; 11.8%), anaplastic astrocytoma (1 case; 5.9%) and glioblastoma multiforme (3 cases; 17.7%). The morphological categorization of infratentorial tumors was astrocytoma (11 cases; 20.4%), medulloblastoma (38 cases; 70.4%), ependymoma (3 cases; 5.6%) and mixed glioma - astroependymoma (2 cases, 3.7%). The morphological sub-categorization of infratentorial astrocytoma was pilocytic astrocytoma (7 cases, 63.6%), with gemistocytic astrocytoma, grade II, grade III and anaplastic astrocytoma comprising 1 (9.1%) case each. The morphological categorization of medulloblastoma was classical medulloblastoma (15 cases; 39.5%), desmoplastic medulloblastoma (8 cases; 21.1%), medulloblastoma with astrocytic differentiation (12 cases; 31.5%), medulloblastoma with neural differentiation (2 cases; 5.3%), and neuroblastic medulloblastoma (1 case; 2.6%). The pediatric brain tumors in Karachi reflect a developing country scenario, with a strong male predisposition and a late presentation with a peak in the 5-9 year age group. There is a predominance of medulloblastoma and a paucity of astrocytomas. The current study is a single institution study and needs cautious interpretation. Population-based studies are required to determine the cancer burden due to pediatric malignancies of the brain in this population and for the morphological categorization of brain tumors in Karachi.

Patho-epidemiology of breast cancer in Karachi '1995-1997'.

OBJECTIVE: Provide an overview of the demographics and pathology of breast cancer in the female population of Karachi South during a 3 year period, 1995-1997. METHODS: Epidemiological data for 709 incident breast cancer cases, ICD-10 category C50 registered at Karachi Cancer Registry during 1st January 1995 to 31st December 1997 were reviewed. RESULTS: Breast cancer accounted for approximately one-third of the cancers in females. The age standardized incidence rate (ASR) world per 100,000 was 53.8, the crude incidence rate was 30.9. In KS 60% of the newly diagnosed breast cancers were observed in women below 50 years. The age-specific curves showed a gradual increase in risk from the third up till the seventh decade, followed by an actual/apparent decrease in risk. The socio-economic distribution was 24.9% in category I the financially deprived class, 38.9% in category II the middle class and 35.9% in category III, the affluent class. Microscopic confirmation of malignancies was 99%. Invasive breast cancers predominated with 99.4%, with in-situ cancers contributing to 0.6% of the malignancies. The morphology of cancers was tilted towards duct cell carcinoma (DCC), pure DCC (92%), combinations of DCC /Paget's disease (0.6%) and lobular carcinoma (0.4%). Approximately 45% of duct cell carcinoma were seen in the premenopausal age group (<45 years). All bilateral breast cancers were duct cell carcinoma with a family history of first degree relative with breast cancer. The majority of the cases presented as moderately differentiated or grade 2 lesions (59.0%). Approximately 56% cancers had spread to the regional lymph nodes and 8.3% to a distant site at the time of diagnosis. A family history of first degree relative with breast cancer was present in 3% and second degree relatives in 7% of the cases. Odds ratio (OR) for 680 breast cancer cases with complete demographic information was calculated with 675 gender matched controls. A slightly higher risk was observed in non-Muslims and migrant ethnicities: two to three fold elevation in the Indian migrants (Gujrati speaking Mohajirs OR 3.86 (95% CI 2.51; 5.92) Urdu speaking Mohajirs OR 2.85 (95% CI 2.05; 3.96), Memon Mohajirs OR 2.21 (95% CI 1.48; 3.29) and Afghan migrants [OR 2.99 (95% CI 11.20; 7.44)]. The risk was also high in the females of Punjabi ethnicity settled in KS [OR 2.73 (95% CI 1.87; 3.99)]. The risk seems much less for the ethnicities belonging to North Western Pakistan i.e. Pathans [OR 1.684 (95% CI 0.89; 3.17)] and Baluchs [OR 0.90 (95% CI 0.58; 1.39)]. A marginally higher risk was observed in the higher socio-economic categories. The risk of developing breast cancer increased gradually for each age category from illiterate [OR 1.2 (95% CI 0.94; 1.55)] to college graduates [OR 13.12 (95% CI 7.31; 23.73)]. CONCLUSIONS: The incidence of breast cancer in Karachi South (KS) for the period 1995-1997 was the third highest in Asia. The hallmarks were a high reproductive age malignancy involving a higher socio-economic class, an invasive duct cell carcinoma diagnosed at an advanced stage, in younger more educated females and a low in-situ malignancy. More studies are required to obtain a deeper insight into this breast cancer epidemic in Karachi. Implementation of breast cancer screening with stress on public health education is today a major responsibility of the government.

Patho-epidemiology of Cancer Cervix in Karachi South.

INTRODUCTION: The present study was conducted with the objective of examining descriptive epidemiological and pathological characteristics of cancer cervix in Karachi South, an all urban district population of Karachi, Pakistan. METHODOLOGY: A total of 74 cases of cancer cervix, ICD-10 (International Classification of Diseases 10th Revision) category C53 were registered at the Karachi Cancer Registry, for Karachi South, during a 3 year period, 1st January, 1995 to 31st December 1997. RESULTS: The age standardized incidence rate (ASR) world and crude incidence rate (CIR) per 100,000 were 6.81 (5.2, 8.43) and 3.22 (2.49 to 3.96). Cancer cervix accounted for approximately 3.6% of all cancers in females and was the sixth malignancy in hierarchy. The mean age of the cancer cases was 53.27 years [standard deviation (SD) 11.6; 95% confidence interval (CI) 50.58, 55.96; range (R) 32-85 years)]. The distribution by religion was Muslims (90.5%), Christians (8.1%) and Hindus (1.4%). There were no cases reported in Parsees. The frequency distribution by ethnicity was Urdu speaking Mohajirs (20.3%), Punjabis (17.6%), Gujrati speaking Mohajirs (4.1%), memon Mohajirs (8.1%), Sindhis (10.8%), Baluchs (8.1%), Pathans (5.4%) and Afghan migrants (2.7%). The ethnicity was not known in approximately a fourth (23.0%) of the cases. The socio-economic distribution was 27.0% financially deprived class, 24.4% lower middle class and 48.7% upper middle and affluent classes. The majority of the women were married (86.5%); a smaller number were unmarried (2.7%) or widows (10.8%). The age-specific curves showed a gradual increase in risk from the fourth up till the seventh decade, followed by an actual apparent decrease in risk after 64 years of age. The peak incidence was observed in the 60-64 year age group. The morphological categorization was squamous cell carcinoma (86.5%), adenocarcinoma (10.9%) and adenosquamous carcinoma (2.6%). The majority of cases presented with moderately differentiated or grade 2 lesions (45.9%). There were no in-situ cases. Approximately half the cancers (58.1%) had spread regionally and 8.1% to a distant site at the time of diagnosis. Odds ratios (OR) were calculated for socioeconomic residential categories, religion, ethnicity, age groups and education. The OR for socioeconomic residential categories ranged between 0.69 and 2.9 with a marginally higher risk in the lower [OR 2.09 (95% CI .97; 4.49)] and lower middle class [OR 2.08 (95%CI 0.95; 4.58)]. Hindus [OR 1.2 (95% CI 0.18; 2.2)] had a slightly higher risk then the Muslims [OR 0.14 (95% CI 0.17; 1.2)]. A higher risk was also observed for Christians [OR 7.76 (95% CI 1.74; 34.5)]. CONCLUSION: The incidence of cervical cancer in Karachi South (1995-97) reflects a low risk population with a late presentation and a high stage disease at presentation. It is suggested that cervical screening if implemented should focus on once a life time methodology involving 36-45 year old women. This should be combined with HPV vaccination for the young and public health education for all. A regular cervical screening program would require mobilization of considerable financial, structural and human resources along with training for personnel. This may burden the already stretched health resources of a developing country.

Phylloides tumors in adolescent girls and young women in Pakistan.

The objective of this study was to determine the frequency of phylloides tumor (PT) in adolescent girls and young women (less than and equal to 25 years of age) and to define the clinico-pathological features of PT in this unusual clinical setting. This descriptive study was carried out at the Aga Khan University Hospital (AKUH) pathology department. All consecutive cases of PT diagnosed during the last sixteen years in the section of histopathology from 1st January 1990 to 31st June 2005 were included. Selection of cases was restricted to patients up to 25 years of age. A total of 42 cases of PT in up to 25 years of age were diagnosed. This comprised 11% of the total PT cases (total n=363). The number of benign (BPT), borderline (BLPT) and malignant (MPT) was identical i.e. 14 (33.3%) each. Clinically all cases presented with a solid, mobile, palpable mass. The mean age was 19.1 years (95% CI, 16.7-21.6), 21.9 years (95% CI, 20.7-21.9) and 19.7 years (95% CI, 17.2-22.3) in BPT, BLPT and MPT respectively. In majority of cases the surgical procedure performed was lumpectomy (50% of BPT, 78% of BLPT and 64% of MPT). High grade PT (BLPT and MPT) is an uncommon mammary tumor in adolescent girls and young women but seems to be occurring with increased frequency in the study population. This observation may indicate the biological behavior of PT in a high risk population, though chances of referral bias are also present. In view of the rarity of the disease, larger population studies are suggested to confirm our findings.

Metaplastic carcinoma of the breast, an intriguing rarity.

Metaplastic carcinoma breast is categorized as a rare heterogenous neoplasm generally characterized by a mixture of adenocarcinoma with dominant areas of spindle cell, squamous and/or other mesenchymal differentiation. To determine the epidemio-demographic and histopathologic characteristics of this rare entity a retrospective study was conducted to review all cases at the Aga Khan University Hospital (AKUH) Karachi, received during 1st January 2000 to 31st August 2005. Twenty-four patients were identified with a mean age at diagnosis of 46.4 (+/-SD 3.8) years, and an age range of 28-68 years. The mean tumor size was 7.9 cm, range 2.0-17.0 cms (+/-SD 4.77). The specimens were mostly obtained by modified radical mastectomy(54.2%) followed by biopsy (29.2%), lumpectomy (8.3%), and total mastectomy (4.2%). Skin ulceration was found in 37.5% cases. Component sub-categorization showed 13 (54.2%) cases of infiltrating ductal carcinoma with squamous metaplasia, followed by 2 (8.3 %) cases with heterologous elements, 4 (16.7%) cases with spindle cell component, 2 cases of matrix producing carcinoma and one case of squamous cell carcinoma. The malignancy was high grade, modified Bloom Richardson's grade III (54.2 %) and grade II (12.5%). Such grading was not applicable to 4 cases of spindle cell component and 1 case with extensive chondroid areas. Twelve patients had information available on the nodal status. Five (41.6%) were node-negative, four (33.33%) had 1-3 lymph node involvement positive, and three (12.5%) had more than 3 lymph nodes positive. The median 3 year Event Free Survival (EFS) was 10% and overall survival was 30%. MCB is an aggressive disease with a poor prognosis. This aspect appears bleaker in our population either due to the biological characteristic of the malignancy in a high risk group or the lack of availability and accessibility of health coverage, resulting in a delayed presentation. MCB is an uncommon breast malignancy and due to the lack of sufficiently large studies there is limited knowledge as to the pathogenesis, progress, best treatment protocols and prognosis. Collaborative studies are therefore recommended to allow for better understanding of this intriguing neoplasm.

Epidemiological review of head and neck cancers in Karachi.

Head and neck cancers, categories lip, oral cavity, pharynx and larynx are placed amongst the top ten malignancies globally. The cancers have a similar epidemiology, risk factors, morphology, and control measures. The geographical variations in incidence are indicative of the global differences in the prevalence of risk factors. The present study was conducted with the objective of reviewing descriptive epidemiological characteristics, incidence and time trends of head and neck cancers in Karachi (1995-2002). Head and neck cancers accounted for approximately one-fifth (21%) of the cancers in males and about one-tenth (11%) in females in the study period. The age standardized incidence rate (ASR) was 37.1/100,000 in males and 21.7/100,000 in females. In males, oral cavity and larynx were the commonly affected sites, followed by pharynx. In females, oral cavity was the preponderant site. The mean age of the patients was 53.0 years (95% CI 48.0; 58.0). A rising incidence was observed in both genders, more apparent in males. About 30% of oral cancer cases, 28.6% of the nasopharyngeal, 6.3% of the oropharyngeal, and 2.6% of laryngeal cancers occurred in patients 40 years and younger. The age specific incidence rates (ASIR) for oral cancer in males showed a gradual rise from 10 to 64+ years of age, for pharynx from 20 to 64+ and for larynx at 25+. The ASIR for oral cancer in females showed a gradual rise from 14 to 64+ years of age, for pharynx from 20 to 64+, a decade after the oral cancer rise and cancer larynx showed a rise at 25+, a decade and a half after the oral cancer rise. The peak incidence was at 64-69 years for all three cancer sites, in both genders. Pakistan falls into a high risk head and neck cancer geographical zone Presentation is late and treatment is not optimum. Recommendations, therefore for NCCP Pakistan, for short term benefits are selected community-based screening for the high risk population, early diagnosis, better treatment, rehabilitation and palliative care. These measures will improve survival and also contribute to a better quality of life. Primary prevention remains the only strategy for absolute cancer control.

Patho-epidemiology of lung cancer in Karachi (1995-2002).

The objective of the study was to provide an overview of the demographics of lung cancer, the number one cancer killer of men in Karachi South (1995-2002). Lung cancer cases recorded at Karachi Cancer Registry during 1st January 1995 to 31st December 2004 were analyzed. To allow for maximum data completion, cases recorded from 1st January 1995 to 31st December 2002 were included for final analysis. Trends were studied by analyzing the age standardized incidence rates (ASR)s in 2 time periods, 1995-1997 and 1998-2002. Odds ratio for sex, age-groups, ethnicity, religion, and residence by socio-economic categories were calculated by considering all malignancies (except tobacco-associated malignancies) for each group, registered at KCR for the same period as controls. Cancer of the lung ranked the most frequent malignancy in men in Karachi in the entire 1995-2002 period, though it did not feature amongst the first 10 malignancies in the females. In the 1995-1997 period, the ASR per 100,000 population for cancer of the lung was 21.4 and 2.9 in males (M) and females (F) respectively. The mean age of the patients was 60.4 years (95% CI, 59.1-61.7) M and 53.7 years (95% CI 48.9-58.5) F. In the 1998-2002 period the incidence rate increased to 25.5 per 100,000 (M) and 4.2 per 100,000 (F). Thus between 1995 and 2002, the incidence of lung cancer registered a 19% increase in men and almost 100% in women. The component of adenocarcinoma in females remained stable during 8 years, but increased 55% in males. Histologic confirmation was 80%; majority of cancer cases presented as grade 3 and grade 4 lesions (62.3%), and were discovered at advanced stages (stage III 35.7%; stage IV 55.8%). The odds ratio (OR) in men was 4.5 (95% CI 3.7; 5.4). The risk of developing lung cancer increased with age, the highest risk being observed in the 65+ age group. A marginally higher risk was observed in the higher socio-economic categories for men and in the lower socio-economic categories for women. A higher risk was also observed for men who were residing along the coastal belt, and for ethnicities belonging to Southern Pakistan (Sindhi and Mohajir) residing in Karachi South. In conclusion, Pakistan at present falls into a low risk lung cancer region in females and a moderate risk region for males and the highest registered increase between 1995 and 2002 was observed in the older age groups (65+). It is however a cause of concern that the overall lung cancer incidence rates continue to rise. The age specific rates though stable in the younger age groups (35-49 years), are at present equivalent to contemporary rates in high- risk countries. These rates correspond with the trends of smoking prevalence in the younger age groups in the last 2 decades. Published studies have given alerts to increase in the smoking habits of the present day youngsters and with an expanding population the country can expect a substantial increase in lung cancer. This threat can only be averted by implementation of stringent anti-tobacco rules and health education; prohibition of smoking in educational institutions at all levels and a ban on the sale of cigarettes to minors.

Cancer profile of Larkana, Pakistan (2000-2002).

This present study was conducted with the objective of providing the cancer profile of Larkana, the fourth largest city of Sindh, Province. The study included two sets of patients. The first set included the incident cancer cases, residents of Larkana, who reached Karachi for diagnosis or treatment. The second set consisted of the incident cancer cases registered at the Aga Khan University (AKU) Pathology collection points at Larkana during 1st January 2000 to 31st December 2002. The age-standardized rates (ASR) for cancer (all sites) were 134.2/100,000 in males and 110.3/100,000 in females. The most common malignancies in males were lymphoma, oral cavity, prostate, liver, and urinary bladder. Cancers in females were breast, oral cavity, lymphoma, skin and thyroid. Tobacco-associated cancers accounted for approximately 35.0% of the tumors in males and 18.6% in females. The gender ratio was 1.5, the mean age of cancer all sites was 45.5 years (95% CI 34.6; 56.4) in males and 42.9 years (95% CI 33.6; 52.2) in females. This is the first attempt to determine the cancer incidence pattern of Larkana and should serve as a guideline for estimation of the cancer burden of Pakistan and the cancer control program of the country. The data must be interpreted with care, as they are largely pathology-based with approximately 15-30% population-wise under-registration. However, the chances of selective collection bias were minimized as the AKU pathology specimens were collected from 2 centers within the city, Each collection centre provided diagnostic service to several health centres giving wide population coverage, thus ensuring adequate sampling from the entire city.

Small B cell Non-Hodgkins Lymphoma in Pakistan.

OBJECTIVE: To study the pattern of small B cell lymphomas in Pakistan. METHODS: This descriptive study was carried out at the Aga Khan University Hospital pathology department including 1721 cases of Non-Hodgkins Lymphoma (NHL) diagnosed during a period of five years (1998-2002) and classified according to REAL/WHO classification. The antibodies used included Leukocyte Common Antigen (LCA), Pan B (CD20, CD79a), Pan T (UCHL-1), Bcl 2, Mib 1(Ki 67) and Cyclin D1 (Dako, Denmark). RESULTS: Out of the 1721 NHL cases, only 140 (8.1%) could be categorized as small B-cell NHL. The study group comprised small lymphocytic lymphoma/chronic lymphocytic leukemia (58 cases; 41.4%) followed by follicular lymphoma (46 cases; 32.9%), mantle cell lymphoma (15 cases; 10.7%), extra nodal marginal zone B cell lymphoma of MALT type (15 cases; 10.7%), lymphoplasmacytic lymphoma (5 cases; 3.6%) and splenic marginal zone B-cell lymphoma (1 case; 0.7%). No case of nodal marginal zone lymphoma was diagnosed. The age ranged from 18 to 98 years with a mean and median of 54.64 and 58.50 years respectively. Small B-cell NHL was more common in males; with male to female ratio of 2.1. Majority of the small B-cell NHL were nodal at presentation with a nodal to extranodal ratio of 3.4. CONCLUSION: It is concluded that the frequency of these small B-cell NHL is very low in our population in contrast to the western literature. Further studies based on epidemiologic and etiological factors are required to look into this marked difference of occurrence of these indolent lymphomas.

Cancer of the oral cavity - trends in Karachi South (1995-2002).

The objective was to study the time trends in site-specific oral cancer incidence and to determine the age-and socio-economic profile over time in Karachi South. Oral cancer ranks second in this population, in both genders. The incidence is the highest reported worldwide. Incident oral cancer cases received at the Karachi Cancer Registry during 1(st) January 1995 to 30(th) June 2004 were reviewed. To ensure maximally complete data, cases registered between 1(st) January 1995 and 31(st) December 2002 were considered for the present study. Cases of lymphoma, leukemia and melanoma were not included. Trends were studied by grouping cases into two periods, 1995-1997 and 1998-2002. A total of 2253 cases of oral cancer were registered in Karachi South for the 8 year study period accounting for 8.8% of all cancer cases. Overall, the most common site was the mucosa cheek (55.9%), followed by the tongue (28.4%), palate (6.8%), gum (4.4%), lip (3.1%) and floor of the mouth (1.4%). About 30% of cases occurred in patients 40 years and younger and 23% occurred in patients 65 years and older. Sub-categories of oral cancer showed variation in trends, but an earlier onset of disease in period two was evident for all categories. The incidence of lip cancer in men decreased, the rates remained level in females. An increased incidence was observed for tongue, but a more dramatic increase in the cheek was evident in both sexes, despite no improvement during the past decade in detection of early, localized lesions. A strong socio-economic factor with a poorer, low literacy profile of oral cancer was apparent in the entire study period. The evidence that the largest increase in incidence has occurred in this population may unfavorably affect the mortality rates. Oral cancer trends are an interplay of prevalent risk factors, the level of prevalence, preventive education and intervention. Cost effective and efficient cancer control focused around the target populations would be beneficial for Pakistan. Educational campaigns should include information on oral hygiene, awareness of risk factors and symptoms and the importance of seeking early professional help when any of these are recognized. Audio-visual media involvement is imperative in view of the literacy status of the target population. Capacity building is required by the Government to increase the availability and accessibility of professionals. Population screening would reduce the incidence of oral cancer, but requires careful planning, and extensive financial resources. Mobilization of general practitioners, health visitors, volunteer organizations and medical students for early detection of oral cancer is the essential need of today.

Cancer profile of Hyderabad, Pakistan 1998-2002.

Hyderabad is the third largest city of Pakistan, the second largest city of Sindh Province and one of the oldest cities of the sub-continent. This administrative headquarter is located just east of the River Indus and is an important commercial and industrial center. Once a provincial capital, it is at a distance of approximately 200-km from Karachi. This present study was conducted with the objective of providing the cancer profile of Hyderabad, which has an urban population of 2,840,653 (52.2% M, 47.8% F) annual growth rate 1.13. The city is inhabited by all ethnicities of the country, however the predominant ethnicity is Sindhi, followed by Mohajirs (post-partition immigrants from India), and a lesser extent other ethnicities of Pakistan viz. Baluchs, Punjabis and Pathans. The study includes two sets of patients. First the incident cancer cases, residents of Hyderabad, who reached Karachi for diagnosis or treatment. Second the incident cancer cases registered at the Aga Khan University Pathology-based Cancer Registry (APCR) Pathology collection points at Hyderabad and subsequently registered at APCR, during 1st January 1998 to 31st December 2002. The pathology department of the AKU has 3 centers in Hyderabad, which provide diagnostic pathology especially oncopathology services to the city. The age-standardized rates (ASR) for cancer (all sites) 1998 to 2002 in Hyderabad were 91.6/100,000 in males and 96.0/100,000 in females. The most common malignancies (ASR per 100,000) in males were oral cavity (11.8), lymphoma (10.6), lung (8.0), urinary bladder (6.8), prostate (4.8), liver (4.4), pharynx (4.2), colo-rectum (3.6), larynx (3.2), and skin (3.2). The cancers in females (ASR per 100,000) were breast (22.4), oral cavity (11.5), gall bladder (4.8), esophagus (4.2), cervix (3.6), ovary (3.4), colo-rectum (3.4), lymphoma (3.4), uterus (3.4), and thyroid (2.4). Tobacco-associated cancers were responsible for approximately 40.0% of the tumors in males and 20.0% in females. Histological confirmation remained 96.3%, with 44.5% presenting in grade II or I, 55.5% presenting as stage III and IV. Information on grade and stage of malignancy was available in 70% and 50% of the cases respectively. Males comprised 53.1%, and females 46.9% of the cases. The mean age of cancer all sites, both genders was 45.2 years (95% CI 44.4; 45.9), males 45.4 years (95% CI 44.3; 46.5); females 44.9 years (95% CI 43.9; 45.9). Conclusions drawn from this database must be interpreted with care, as it may be identified as data from selected medical institutions. Chances of selective collection bias are minimized as the data of the AKU pathology is collected from multiple centers in the city of Hyderabad, dispersed at distances, which allows adequate sampling from the entire city. There is a slight preponderance of lymphomas in males which we feel is a true higher risk, yet it may indicate an over representation of easily accessible sites in pathology based-data. Nonetheless, this is the first attempt to determine the cancer incidence pattern of Hyderabad, and should serve as a guideline for estimation of the cancer burden and risk assessment statistics of Pakistan and the cancer control program of the country.

Increasing incidence of non-Hodgkin's lymphoma in Karachi, 1995-2002.

This first population-based study of non- Hodgkin lymphoma (NHL) from any region in Pakistan, provides an overview of the incidence pattern and time trends in Karachi and generates hypotheses for future experimental research. Epidemiological data for 429 incident (1(st) Jan 1995 to 31(st) Dec 2002), microscopically verified nodal and extra-nodal NHL cases, registered at the Karachi Cancer Registry (KCR) for Karachi South, were reviewed. The age standardized incidence rate (ASIR) was 5.3/100,000 in males (M) and 4.1/100,000 in females (F), in 1995. A gradual increase in the annual incidence was observed during the study period, with NHL incidence rate increasing in 2002 to 8.4/100,000 in men and 6.5/100,000 in women, almost double the 1995 rates. NHL affected all age groups in both genders and for each group the ASIR was higher among men than women, with an overall gender ratio of 1.9. The mean ages of the patients were 41.5 years (95% CI 39.1; 43.8) in males and 44.0 years (95% CI 40.8; 47.1) in females. The adult to childhood ratios were 8.6 (M) and 10.7 (F). B-cell NHL comprised 81.0% of NHL in males and 87.3% in females. One fourth of the NHL cases were extra-nodal, the largest group was of gastrointestinal origin (54.1% M, 38.5% F). The gastric component was 21% M and 25.6% F. Odds Ratios for sex, age-groups, ethnicity, religion, and subdivision by socio-economic categories were calculated by considering all malignancies, except lymphoproliferative disorders as controls. The odds ratio (OR) in men was 2.2 (95% CI 0.6; 3.0). Children and adolescents were at the highest risk of developing NHL, especially the 5-9 year olds, in both genders. A marginally higher risk was observed for the lower socioeconomic categories and for ethnicities belonging to Northern and North Western Pakistan (Punjabi, Pushtu and Baluch) residing in Karachi South. The incidence rates of NHL registered in Karachi South are likely to be a reflection of non-AIDS-associated NHL. Estimated HIV/AIDS incidence was too low during the study period in this population to have an impact on NHL incidence. The preponderance of low and intermediate grade lymphomas, paucity of central nervous system NHL and a higher childhood NHL component support this hypothesis. As yet unpublished reports, however, are raising the alarm on rising HIV positivity. NHL correlation with HIV/AIDS status and studies identifying risk factors of non- HIV/AIDS associated NHL (childhood viral infections, Hepatitis C virus, and Helicobacter pylori) are potential areas for future experimental and epidemiological research.

Ductal carcinoma in situ (DCIS) in Karachi.

OBJECTIVE: To study the frequency of ductal carcinoma in situ (DCIS) in a large pathology series. DCIS is a proliferation of non-invasive, malignant epithelial cells within the ductolobular system of the breast. It is a heterogeneous entity with several morphologic variants that differ in gross appearance, growth pattern, cytologic features, mammography, and malignant potential. METHODS: The data of The Aga Khan University' Pathology Department, diagnosed on the basis of histopathology, during a 6-year period (1st January 1998 to 31st December 2003) was reviewed, all cases of DCIS studied, and data was analyzed with the help of analytical software SPSS. RESULTS: Thirty-eight cases of DCIS were reported to the Aga Khan University Pathology Department, during a 6-year period (1998 to 2003), comprising approximately 1% of all breast cancers reported to the unit in the same period. The mean age of the patients at diagnosis was 48.95 years (CI 95% 44.6; 53.3). Approximately half the cases occurred in the 45-54 year age group (figure 1). Two cases (5.3%) were recurrences with previous lumpectomy scars. Comedo necrosis was observed in five (13.2%) cases, whereas 33 (86.8%) cases were non-comedo type. The clinical presentation was a palpable mass (92.1%), nipple discharge (5.3%) or clinically occult lesions diagnosed on mammography (2.6%). Approximately half the patients presented with a grade 2 disease. Atypical ductal hyperplasia was observed in a third of the cases, predominantly associated with a grade 1 and 2 disease. The estrogen and, progesterone receptor status was studied in 12 (31.6%) cases. Estrogen positivity was observed in 11 (91.7%) cases and progesterone positivity in 7 (58.3%) cases. Microcalcification was observed in four (10.6%) cases. CONCLUSION: The cases reported in this study are the indolent grade 1 or 2 cases with a non-comedo pattern, and a positive estrogen and progesterone receptor status. If untreated, only 40% of these innocuous forms of DCIS become invasive over a time span of approximately 25-30 years. In Pakistan we are missing the more aggressive forms of DCIS which have a shorter transition to invasive carcinoma.

Malignant smooth muscle tumours of soft tissue--a demographic and clinicopathological study at a tertiary care hospital.

OBJECTIVE: To determine the demographic characteristics and distribution of various reported prognostic factors of leiomyosarcoma (LMS) among subjects reporting to histopathology diagnostic centers of Aga Khan University (AKU) all over Pakistan between 2000-2004. METHODS: This study analyzed 205 consecutive confirmed cases of LMS received during a period of four years. The data regarding age, sex and size of tumor was obtained from the pathology medical records. Data on grade and positivity of immunohistochemical stains was assessed and all the variables analyzed using SPSS version 12.5. RESULTS: Of the 205 specimens of LMS, 31 were received as blocks for second opinion. A hundred specimens were multiple fragments, and 74 were intact masses. Of the latter, 40% had clear margins. The mean size of the tumor was 7.23 cms. (95% CI 6.1; 8.4). There were no cases of childhood LMS. Teenage LMS comprised 3% of all tumors. Grade 1 tumors were 16.6% of the total, 56.6% were grade 2, and 8.5% were grade 3. Vimentin positivity was observed in 97% of the tumors, desmin positivity in 56.7%, HHF35 positivity in 81.0%, s-100 positivity in 15.6% and cytokeratin positivity in 11% of the cases. The median age of LMS patients was 48.4 years, the mean age was 48.6 years (95% CI 45.9; 50.8) and male to female ratio was 1:1.2. The malignancy occurred a decade earlier in the females as compared to the males. The mean age of male cases was 52.3 years (95% CI 48.8; 55.8); and of females was 45.2 years (95% CI 41.8; 48.6). The mean age of teenage LMS was 17.2 years (95% CI 15.6; 18.7). The most common symptom was a painless swelling. The most common sites were lower limb (24.4%) and pelvis (24.4%) followed by abdomen (20.6%) and head and neck (12.7%). The least common sites were upper limb and chest (8.0% each). Seventy five percent of the tumors with involved margins were more than 5.0 cms. in size whereas 47% of tumors with free margins were above 5.0 cms. in size. Approximately half the tumors in the males were less than 5.0 cms, as compared with 31.0% in the females. LMS in the extremities was equally common in both genders, but trunk LMS was nearly twice as common in the females. CONCLUSION: In our population, leiomyosarcoma occurs at a relatively younger age, has a late presentation; is more common in females as compared to males and usually excised incompletely. No genetic study for LMS has been published in Pakistan; these studies are recommended to determine the biological pattern of LMS in our population.

Large-scale pathology-based cancer data--a reflection of population-based cancer data.

Pathology-based cancer data is a high quality reflection of the patterns of cancer in the population it represents, provided the demographic details of the patients are carefully recorded. Relative frequency data is neither a replacement for population-based data nor a suggested alternative; it simply enhances the quality of population data and in very large data sets reflects the cancer patterns observed in the representative populations. Aware of the standard shortfalls of pathology-based data, the department of pathology, The Aga Khan University Hospital (AKUH) standardized its data, representing 53.4% of the cancer data of Karachi Division (Pakistan) and also reflecting the cancer pattern of other provinces of Pakistan. This data was compared with 4 different population and institutional-based data sets. The findings substantiate the observation that despite the problems of interpretation of data from pathology laboratories, they are an invaluable source of information on cancer patterns in much of the world where incidence data are unavailable . If developing countries, unable to organize National Population- Based Registry should as an alternate develop National Pathology-based Registers a well targeted and monitored, a Cancer Control Program would be possible. A good quality, large-scale pathology data with demographic details of the patient recorded can also be extended to give coverage to the population.

Karachi Cancer Registry Data--implications for the National Cancer Control Program of Pakistan.

Cancer registries play a major role in providing the data to justify establishment, implementation and monitoring of cancer control programs, therefore stability in cancer registration is of pivotal importance. An erroneous assessment of the cancer burden can have long-term negative implications for the limited health resources of a country. Thus, registries starting simultaneously with cancer control programs clearly cannot be adequate for the purpose. The Karachi Cancer Registry (KCR) is the first population-based registry of Pakistan, with 9 years proven data stability (1995-2003) for Karachi South (KS), a location with a population distribution similar to that for the country in general as regards age, gender, and religion. It also has the distinction of being the only district in the country with representation of all ethnic and socio-economic groups of the country. The primary recommended strategy for the 'National Cancer Control Program' (NCCP), Pakistan based on the assessment of eight common cancers in Karachi and the WHO estimates would be identical. A curb on the epidemic levels of tobacco and areca nut use would reduce malignancies in males by 43.7% and in females by 17.8%. WHO estimates put these figures at 45% and 18.5% for males and females respectively. Primary prevention in the form of diet control, checks on preservatives, dyes, and pesticides; protection from occupational hazards, control of biological agents and solar UV protection would help control of another half of the malignancies. Resource restrictions put high technology methods beyond the scope of Pakistan today. Early detection of cancers of accessible sites, though not an urgent requirement, would be warranted for oral, cervical and breast cancer, after sufficient capacity building, initially in the high-risk groups. In females, this could help target 47.6% (approximately half) of the malignancies and in men 13% of the total. Establishment of equitable pain control and a palliative care network throughout the country is an urgent and essential measure as more than 70% of cancer patients report with very advanced stages of malignancy. The estimated annual incident cancer cases for Pakistan, year 2000 on the basis of KCR data were 138,343 for males and 135,054 for females; approximately twice the number cited by WHO for the same year. The argument that higher KCR estimates reflect an urban catchment population may be justified, the urban: rural ratio being 2:1 in Pakistan. Evidence-based strategies, however,support the counter argument, that the rising incidence of cancer in Pakistan is primarily attributable to risk factors equally prevalent in the rural and urban areas viz. increasing tobacco use, low socio-economic conditions, dietary deficiencies and prevalence of oncogenic viruses. Pakistan has a significant cancer burden and rising trends of risk factors - it is a country in dire need of a Cancer Control Program. KCR data along with WHO estimates can form the initial framework of a NCCP in Pakistan; the lack of a national cancer registration should not deter initiatives. Benefits of an immediate, prompt and targeted implementation established today will be realized after 20-30 years. Otherwise the country should be prepared to face epidemic proportions of the disease in the next decade or two. Prerequisite 'qualification criteria' or 'sincerity of intent test' for NCCP funding by international donors should be legislation against tobacco and areca nuts in Pakistan and stringent evaluative criteria.

Cancer of the oral cavity and pharynx in Karachi--identification of potential risk factors.

The objective of the study was to provide an overview of the demographics of cancer of the oral cavity and pharynx in Karachi South (1995-2001), and identify potential risk factors. Cases recorded for Karachi South, at Karachi Cancer Registry during 1(st) January 1995 to 31(st) December 2002 were analysed. For maximum completion of data cancer cases, recorded from 1(st) January 1995 to 31(st) December 2001 were included for final analysis. The age standardized incidence rates per 100000 population (ASIRs) for cancer of the oral cavity (excluding salivary gland) in Karachi South were 17.1 and 16.5 in males and females whereas the ASIRs for cancer of the pharynx (excluding nasopharynx) were 7.1 and 2.4 in males and females, respectively. The oral pharyngeal ratios were 2.4 and 6.9 for males and females and gender ratios (M F) were 1.04 for the oral cavity and 3.0 for the pharynx. The mean ages were 51 years (95% CI 49.6; 52.2) and 56.1 years (95% CI 54.4; 57.8) respectively. Cancer of the oral cavity ranked 2(nd) in Karachi in both genders. Cancer of the pharynx ranked 7(th) in males and 14(th) in females. Approximately 97% of the oral cavity and pharyngeal cancers were histologically confirmed. The majority of the oral (47.1%) and pharyngeal (51.9%) cancer cases presented as grade II lesions, and were discovered at advanced stages. Of the cancers reported during 1995-2001, 60.4% of the oral and 78.1% of the pharyngeal lesions had spread to a distant site at the time of diagnosis. Squamous cell carcinoma comprised 96.5% and 91.8% of the totals. The incidences of these cancers are comparable to the highest risk regions of the world. As distinct from other geographical areas oral cancer is as common in females as in males, which may reflect the pattern of exposure to known risk factors such as betal quid, arecanut and tobacco and the absence of alcohol as a risk factor in both genders. Apergillus contamination of arecanut could also be a risk factor but no confirmation studies or quantification is available. Despite the common risk factors, incidence of pharyngeal cancer is three times higher in men as compared with women. The keys to reducing the incidence and mortality due to oral and pharyngeal cancers are prevention and control, emphasizing cessation of tobacco use and cancer screening. However a targeted cancer and tobacco control program does not presently exist in Pakistan.