Comparison of the incidence of oesophageal cancer in two 6-year periods from selected hospitals in and around Gauteng Province, South Africa

GENERAL SURGERY Comparison of the incidence of oesophageal cancer in two 6-year periods from selected hospitals in and around Gauteng Province, South Africa A Gould;1 H Morgan;1 N Motha;1 M Makda;1 A Domingo;1 S Tiedt;1 J Wing;1 M Munanga;1 J Tembo;1 M Hale,2 MB ChB, FCPath (SA); D Bizos,3 MB BCh, FCS(SA), MMed (Surg) Graduate Entry Medical Programme (GEMP) Year 3 / Bachelor of Medicine and Bachelor of Surgery (MB ChB) Year 5 Department of Anatomical Pathology, National Health Laboratory Services 3 Department of Surgery, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa 1 2 Corresponding author: D Bizos () Introduction. Global trends suggest that the incidence of squamous cell carcinoma (SCC) has decreased but that the incidence of adenocarcinoma (AC) has increased. In South Africa, outdated data exist, thereby prompting this investigation. Objectives. To determine the incidence of oesophageal cancer in two 6-year periods in and around Gauteng Province. Further distinction was made based on gender, race and histological types. Methods. A retrospective audit was conducted including histologically documented oesophageal cancer cases collected over a 12-year period from 2001 to 2012, which was divided into two 6-year periods. Incidence was calculated based on the 2013 Gauteng estimated population size. Statistical analysis was performed using the χ2 test. Results. On the whole, there was a significant decrease in the incidence of SCC (p=0.0001). Significant decreases were seen in the African male and female groups (p=0.001 and p=0.0006, respectively). No significant difference was seen in the non-African male and female groups. A non-significant decrease was seen in the AC type with regards to gender and race. Conclusion. Reasons for the decline shown here are unknown as patient risk factors were not available. Furthermore, major healthcare centres were not included. These are points for future investigation. The incidence of oesophageal cancer has decreased since 2001, owing to the decrease in SCC in African males and females. Although the decreases in the AC type were not significantly different, they do not parallel global trends. S Afr J Surg 2015;53(2):55-58. DOI:10.7196/sajsnew.7857 In 2008 an estimated 482 000 new cases of oesophageal cancer were diagnosed worldwide, resulting in 407 000 deaths globally and making it the 8th most frequently diagnosed cancer and the 6th most likely cause of cancer-related deaths.[1] Globally, squamous cell carcinoma (SCC) is the more prevalent type, accounting for 90% of cases in most European and Asian countries.[1,2] Observed trends over the last few decades show that there is a decreasing incidence of SCC in both high- and low-risk areas and an increase in adenocarcinoma (AC), especially in Western countries,[3-5] which is estimated to increase by 140% by 2025.[2] In South Africa for the period 1991 - 1995, the incidence of oesophageal cancer was greatest in the Transkei region with an incidence in SCC of approximately 76/100 000 and 36/100 000 in males and females, respectively. [2] While it is generally accepted that SCC is the predominant type, recent observations show that the incidence of AC is on the rise in certain population subgroups,[6] a point not yet explored in the South African setting. The latest data from the 2008 GLOBOCAN database shows that southern Africa has the highest incidence of oesophageal cancer in the world.[7] However, no data exist as to the incidence of AC and there are no current data in South Africa regarding trends in incidence. Objectives The objective of this study was to determine the incidence of oesophageal cancer in two 6-year periods as diagnosed at the National Health Laboratory Services (NHLS) at Charlotte Maxeke Johannesburg Academic Hospital (CMJAH). Additional aims were to determine the incidence of oesophageal cancer with regards to gender, race group and histological type. Methods Ethical clearance was obtained from the University of the Witwatersrand Human Ethics committee. A retrospective audit was conducted on data from the NHLS at CMJAH and included all cases of confirmed oesophageal carcinoma based on histological diagnosis from specimens processed by the NHLS and whose results were subsequently captured onto the database. A time frame of 12 years from 2001 to 2012 was analysed. Variables considered for each histological subtype included gender, race (African or non-African) and referral site, with the top 10 sites reported separately and the rest grouped together. Patient anonymity was maintained. Incidence was calculated using the 2013 estimated population size in Gauteng Province, obtained from StatsSA. A comparison was drawn between two 6-year periods: 2001 - 2006 (period 1) and VOL. 53 NO. 2 JULY 2015 SAJS 55 56 SAJS VOL. 53 NO. 2 JULY 2015 Incidence (patients / 100 000 people / year) When comparing the combined groups (Fig. 1) presenting with the AC type, there was no significant difference between the incidence (0.24/100 000 v. 0.14/100 000, p=0.964) for the two time periods. In the SCC group, there was a significant difference between the two periods (2.27/100 000 v. 1.14/100 000, p=0.0001). Without considering gender and race within both groups, the incidence was significantly higher in period 1. Comparison of the incidence of the SCC type (Fig. 2) between the two periods with regards to gender and race showed a significant difference in the African male (7.17/100 000 v. 4.06/100 000, p=0.001) and African female groups (4.49/100 000 v. 2.75/100 000, p=0.0006). There was no significant difference in the incidence of SCC in the non-African male (0.44/100 000 v. 0.46/100 000, p=0.116) and non-African female (0.43/100 000 v. 0.15/100 000, p=0.264). All groups demonstrated a decrease from period 1 to period 2. The same comparisons regarding AC type incidence (Fig. 3) between the two periods also showed a decrease in all groups from period 1 to period 2, but these decreases are not significantly different. The incidence for periods 1 and 2 in the African males was 0.43/100 000 v. 0.28/100 000 (p=0.803) and in the African female groups 0.23/100 000 v. 0.19/100 000, (p=0.254), which both show a decrease in the incidence of AC. In the non-African males, the incidence for periods 1 and 2 was 0.57/100 000 v. 0.37/100 000 (p=0.055) and non-African females the incidence was 0.10/100 000 v. 0.07/100 000 (p=0.35). When comparing the sites of origin of patient histological samples, the majority demonstrated a decrease in the number of samples sent to the NHLS labs at CMJAH during both periods (Fig. 4). There were, however, increases in the number of samples sent from Tembisa, Boksburg and Polokwane regarding the two periods. It is unclear why no specimens were received from Sebokeng during period 2. p=0.0001* p=0.964 AC SCC Fig. 1. Comparison of incidence between periods 1 and 2 within the AC and SCC groups (* denotes statis (...truncated)