Cancer survival in Malawi: a retrospective cohort study

Introduction Cancer is a leading cause of morbidity and mortality worldwide with the burden in sub-Saharan Africa projected to double by year 2030 from 715,000 new cases and 542,000 deaths in 2008. However, cancer survival data to inform interventions for early detection, diagnosis and treatment are lacking. Methods Cancer survival analysis was conducted on 842 cancer patients registered and followed-up from 2006 to 2013 at NdiMoyo Palliative Care Centre in Salima District, central Malawi. Cancer survival was measured from the time of diagnosis. Results In both sexes, the common types of cancer were; Kaposi's sarcoma (KS) (48.0%), cervical cancer (21.1%), cancer of oesophagus (14.8%), liver cancer (3.1%) and breast cancer (2.5%). In Males; KS, cancer of the oesophagus, cancer of the liver, bone cancer and non-Hodgkin's lymphoma were the commonest accounting for 67.4%, 19.4%, 3.9%, 1.0% and 1.0% respectively. In females; cancer of the cervix, KS, cancer of the oesophagus, cancer of the breast and cancer of the liver were the top five cancers accounting for 41.6%, 29.2%, 10.3%, 4.9% and 2.3% respectively. Of the 830 cancer patients with complete 5-year follow-up data, the overall median survival time was 9 months. Absolute survival rates at 1, 2, 3, 4 and 5 years or more were 31.8%, 18.0%, 12.5%, 7.8% and 6.0% respectively. The survival rates for top five cancers at 1, 2, 3, and 4 years or more were; KS (n= 397): 47.1%, 30.2%, 21.4% and 13.1%; cancer of the cervix (n = 174): 31.0%, 10.3%, 5.2% and 2.9%; cancer of the oesophagus (n = 124): 4.0%, 2.4%, 1.6% and 1.6%; liver cancer (n = 26): 19.2%, 3.8%, 3.8% and 3.8% and breast cancer (n = 21): 9.5%, 0%, 0%, 0% respectively. The risk of death was high in females than males, in those aged 50 years or more than in those aged less than 50 (p < 0.05). Conclusion This study demonstrated that cancer survival from the time of diagnosis in Malawi was poor with median survival time of about 9 months and only 6% of patients survived for 5 years or more. Improvement of early detection, diagnostic capability, access to treatment and palliative care services could improve cancer survival.


Introduction
Cancer is a leading cause of morbidity and mortality worldwide. In 2008, globally, there were 12.7 million new cancer cases and 7.6 million cancer deaths (around 13% of all deaths) with 56% of the new cases and 63% of the cancer deaths occurring in developing countries. It is projected that by 2030, the number of new cancer cases and deaths will increase by 69% and 72% to 21.4 million and 13.2 million respectively [1][2][3].
In sub-Saharan Africa, it has been projected that the burden of cancer will double by year 2030 from 715,000 new cases and 542,000 deaths in 2008 [3]. Cancer survival tends to be poor in this region because of a combination of a late stage at diagnosis and limited access to timely and standard treatment [4][5][6]. For example, in Uganda and Zimbabwe, 5-year relative survival for colorectal cancer and cervical cancer were as low as 8.3% and 17.7%, 17.4% and 30.5% compared with 63.9%, 58.1% respectively, for black American patients [7,8]. At the time of diagnosis, over 80% of cancer patients are in advanced and incurable stage, making the need for palliative care more important in this region [9][10][11]. World Health Organization defines palliative care as an approach that improves the quality of life of patients and their families facing the problem associated with life-threatening illness, through the prevention and relief of suffering by means of early identification, assessment and treatment of physical, psychosocial and spiritual problems. Palliative care, if initiated soon after diagnosis, has been found to improve the treatment outcomes and survival of cancer patients [12,13].
In Malawi, cancer is a major public health problem with estimated age-standardised incidence rate (ASR) per 100,000 population per year of 55.5 in males and 68. 8 Table 3, Table   4). Cox proportional hazard revealed that the hazard of death was 1.56 times higher in the age group of 50 years or more compared to those aged less than 50 years (p=0.0001, Log Rank=24.8, DF=1). The hazard of death was found to be 1.2 times higher in females with esophageal cancer compared to males (p=0.62, n=125).

Characteristics of cancer cases enrolled in the study
Females with skin cancer were also found to have a higher hazard of death and the results were statistically significant (P=0.001). The hazard of death was 36.5% higher in males with liver cancer compared to females, but the results were not statistically significant (p=0.33) because of small sample size (n=26). Survival in males and females for Kaposi's sarcoma was similar (hazard ratio=1.0). Stratifying for age, there was an increased hazard of death in those aged 50 years or more for cervical, esophageal, bone and bladder cancers while for Kaposi's sarcoma, liver, breast, and skin cancers, the risk of death was higher in those aged less than 50 years.

Discussion
Data on cancer survival from eastern and southern Africa are scarce because of lack of functional population-based registries that collect data regularly [14]. This paper has provided insight on common types of cancers and cancer survival from the time of diagnosis in Malawi using the available seven-year follow up data of over 800 patients from a palliative care centre. The pattern on common types of cancer by gender and age was similar and comparable to national cancer of the liver (3.9% vs 1.5%) were most common in males [14]. This could suggest that estimates on common types of cancer by gender and age from palliative care centre could be used a proxy where population-based cancer registry is un-functional or not regularly conducted and/ or in triangulation provided that sample size is large enough.
This study also highlighted that cancer patients in Malawi were  [7][8][9]. Inadequate/lack of cancer screening programmes, late presentation of patients, late diagnosis, inadequate/lack of diagnostic facilities, unavailability of treatment, limited access of patients to diagnosis and treatment sites and HIV have been documented as some of factors that are contributing to poor cancer survival in this region [7][8][9][10]. All these factors are present in Malawi and therefore addressing them may lead to early detection, diagnosis and treatment which in turn will lead to improved cancer survival [14,16]. Some of the efforts underway in

Limitations of the study
This study was facility-based and only from one palliative care centre hence the survival estimates presented could not be generalised. Population-based cancer survival estimates from cancer registry could have been the best but data collection by the Malawi Cancer Registry is irregular, mainly confined to one district (Blantyre) and does not have follow-up data [14]. The other limitation was that sample sizes for some cancers were small making survival estimates for those specific cancers less reliable.
Lack of information on the grade or stage of cancer was another limitation although this was not specific to this study but a general problem in Malawi where majority (>80%) of cancer cases are not laboratory confirmed and clinical staging is also not usually done [14]. Nonetheless, this study has provided an insight on cancer survival estimates based on the available seven year follow-up data of over 800 cancer patients which could be used inform interventions and strategies to improve early detection, diagnosis, treatment and palliative care services in Malawi.

Conclusion
In Malawi, cancer survival from the time of diagnosis was very poor with median survival time of about 9 months and only 6% of patients survived for 5 years or more. Improvement of public awareness and early detection, diagnosis, treatment options including chemotherapy and radiotherapy, and palliative care services could improve cancer survival.

Competing interests
The authors declare no competing interests.

Authors' contributions
GM, BT, CT and KPM conceived the idea; LKF and KH were involved in data collection, analysis, report writing and drafting of manuscript. All authors approved the submission of this manuscript for publication. All authors have read and agreed to the final version of this manuscript and have equally contributed to its content and to the management of the case.

Acknowledgments
This study was funded by the University of Malawi, College of Medicine. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of this manuscript.