figure 1.jpg
figure 3.jpg
figure 2.jpg
figure 4.jpg

Clinical Outcomes of Colorectal Cancer in Kenya


Authors: Saidi H1,2 MBChB, MMed, FACS, Abdihakin M2 MBChB, Njihia B1 Bsc, Jumba G1 Phd, Kiarie G 1 MMed, Githaiga J1 MMed, Abinya


NO1,2 MMed. Affiliations: 1-University of Nairobi and Kenyatta National Hospital 2-Aga Khan University Hospital, Nairobi Correspondence: Prof. Hassan Saidi, Dept of Human Anatomy, University of Nairobi. P.O. Box 30197, GPO 00100, Nairobi, Kenya. E-mail




The incidence of colorectal cancer in Africa is increasing. True data on clinical outcomes of the disease is hampered by follow up challenges.



Follow up data of 233 patients treated for colorectal cancer between 2005 and 2010 at various Nairobi hospitals were evaluated. The primary outcome was mortality while secondary outcomes included recurrence rates, time to recurrence and the patient, disease and treat-ment factors associated with mortality and recurrence. Kaplan Meir charts were charted for survival trends.



Half of the lesions were located in the rectum. There was no relation-ship between the sub-site location and recurrence and mortality.



The mean follow-up period was 15.9 months. Overall recurrence and mortality rates were 37.5% and 29.4% respectively. Most recurrences occurred within one year of surgery. Recurrence was not influenced by age, gender, sub-site, chemotherapy receipt or presence of co-morbidity. Factors significantly associated with mortality included the male gender ( p 0.04), presence of co-morbidity (p 0.029), recurrence (p 0.001), curative intent (p 0.01), disease stage (p 0.036) and receipt of chemotherapy ( p< 0.01).



Follow up of colorectal cancer patients is still challenging. The mortal-ity and recurrence rates are high for the short follow up periods. Fur-ther studies are needed to explore the determinants of both survival and recurrences, especially with longer follow ups.

table 2.jpg
table 1.jpg


The incidence of colorectal cancer (CRC) in the devel-oped countries is stable or decreasing (1) while that in Africa is increasing (2,3). African CRC presents in late stages and in relatively younger patients. Outcome stud-ies from America suggest worse outcomes for African Americans in comparison to the whites (1). Worse treat-ment outcomes are tied to treatment access, screening practice, and presence and nature of co-morbidities (1). In an earlier Kenyan study, mortality in CRC was worse for men and emergency surgery (2). The follow-up chal-lenges are major and the true clinical outcomes remain largely unknown.


We sought to further define CRC outcomes and relate these to patient, tumor and treatment variables


Patients and Methods


This was a chart review of incident cases treated between 2005 and 2009 and followed up by the authors (HS, GJ, GK, JG, NOA) at the Kenyatta National and the Aga Khan University hospitals in Nairobi. A smaller group accrued from private offices of the authors, having had their treatments at other hospitals in Nairobi. Only patients with sufficient information on CRC pathology, treatment and follow up were included. Patient profile, tumor sub-site, pathology details, recurrence and mor-tality data were collected.


The primary outcome was CRC mortality. Secondary outcomes included rate of recurrences, time to mortality and recurrence and the associated patient and disease factors.


Means (SD) or frequencies (%) for distributions were calculated. Outcome groups were compared for continu-ous and categorical variables using Students t-test and X2 as appropriate with significance level set at p < 0.05. Sur-vival trends were charted using the Kaplan Myer method.



Two hundred and thirty three patients were studied. Fifty eight others were excluded for lack of key data elements. The proportion of black African patients was 91%. Males comprised 58.8% of the group while the propor-tion with co-morbid diseases was 38.3%. The peak age (25.9%) affected was 41-50 years (all-group mean age 53 years). The proportion of patients 40 years of age or younger was 17.6% (Fig 1).


Half (50.5%) of the lesions were located in the rectum (rectosigmoid lesions comprised 63.3% of all cases) (Fig. 2). There was no relationship between tumor sub-site location and age, gender, recurrence and mortality.



The surgical procedures undertaken and chemotherapy regimens administered are shown in table 1. The mean follow-up period was 15.9 months (range 1-20 months). Overall recurrence and mortality rates were 37.5% and 29.4% respectively. Most recurrences occurred within one year of the surgery (fig. 3). Recurrence was not in-fluenced by age, gender, sub-site, chemo receipt, and co-morbidity status.



Most lesions (70%) were well or moderately differentiat-ed. The disease stage at treatment was I (7%), II (22.7%), III (40.1%) and IV (29.1%).


For a variable proportion of reports, key data elements were missing. Staging data was missing in 22.9% of re-ports, differentiation unstated in 52.9%, lympho-vas-cular status not shown in 41.3% of reports and nodal characterization absent in 37.2%. In 34.3% of reports, the nodal harvest was inadequate (less than 12 nodes harvested).


Statistical Analysis


On univariate analysis, the factors that were significant-ly associated with mortality included the male gender, presence of co-morbidity, recurrence, receipt of chemo-therapy, disease stage and curative intent (table 3). The following factors did not influence mortality: post-oper-ative morbidity (p 0.08), age (p 0.55), sub-site (p 0.86), race (p 0.47), tumor differentiation (p 0.33), and type of chemotherapy (p 0.22). The Kaplan Meir curves depict-ing survival trends are shown in figure 4a-d.



The results of the current study have provided additional outcome data on colorectal cancer in Kenya. The data on young age and advanced stage at presentation corrobo-rate earlier data from the region (2, 4, 6). The youngest patient in this group was 18 years with peak age of 41-50 years. In Lagos and Sagamu in Nigeria (6) the youngest was 10 years with a peak of 60-69 (mean 50.7 and 23% younger than 40 years).


The pattern of young age of presentation contrasts Western data (1).


Click to view figure 1

Click to view figure 2

Folfox- 5FU-leucovorin/oxaliplatin combination, Xelox – oral capecitabine/ox-aliplatin combination, Folfiri- 5FU-leucovorin/irinotecan combination, Xeloda– oral capecitabine, avastin - bevazixumab


Click to view table 1

Click to view table 2

Click to view figure 3


A trend towards younger age at presenta-tion amongst Americans has however been recently reported. Meyer et al, analyzing the surveillance, epi-demiology and end results (SEER) database reported a 3.8% increase in the rate of rectal cancer in the young between 1984 and 2005 but could not demonstrate a similar trend for colon cancer (6). The authors had no explanation for the observation. Future research direc-tion will entail defining this younger population better through genetic and family analyses (4,6).


The advanced stage at presentation also contrasts the presentation amongst Americans, especially the non-black population. Whereas 29% of cases in the current study have presented with metastatic disease, in the re-port by Polite et al. the respective proportions for Ameri-cans was 18% for whites and 24% for African Americans


(1). These differences are reflected in the survival pat-terns. The three-year overall survival for our local CRC, derived from the survival curves was 40%, in contrast to better rates at even longer follow up, 65% for whites and 55% for African Americans at five years . Differing tumor biologies may explain the discrepancies as may factors related to access to diagnosis and treatment (1).


In one of the earlier Kenyan studies, factors that signifi-cantly influenced mortality included – gender and emer-gency surgery. The effect of gender has persisted in the current analysis. Additional factors in the present study included the presence of co-morbidity, recurrence, cura-tive intent, disease stage and use of adjuvant chemother-apy. Our results are comparable to others for the effect of stage and receipt of chemotherapy (7). Further local studies are needed to show the real association with age, race and type of chemotherapy.


The study has additionally shown a number of areas that need improvement in our CRC care pathways. The follow-up period, although 10 months longer than the earlier study (2), attests to the need to develop pro-spective databases that can more reliably monitor CRC outcomes long term. Pathology reports were discordant with international guidelines in a significant number of reports. A quality monitoring tool, if effected, will en-sure concordance with chemotherapy for node positive cases, surgical margins and nodal harvest, median time intervals of receipt of treatment and pathology report-ing. The Kenya oncological research database (8) is one such effort designed to improve the care of CRC in Ke-nya. Using data from this source, care givers can inter-rogate the follow-up data more thoroughly and analyse survival more comprehensively than the overviews we have documented in this report.

Click to view figure 4




The study results were presented at the East Africa Oncology Experience meeting, Serena Hotel, Mombasa, Kenya April 2010




  1. Polite BN, Dignam JJ: A colorectal cancer model of health disparities:Understanding mortality differences in minority populations. J Clin Oncol 24:2179-2187, 2006

  2. Saidi H, Nyaim EO, Githaiga JW, et al. Colorectal cancer sur-gery trends in Kenya, 1993-2005. World J of Surg, 2008; 32: 217-223

  3. Iliyasu Y, Ladipo JK, Akang EE, et al. A twenty-year review of malignant colorectal neoplasms at University Col-lege Hospital, Ibadan, Nigeria. Dis Colon Rectum. 1996 May;39(5):536-40.

  4. Saidi H, Nyaim EO, Karuri D, et al.. Young patients with colorectal cancer at a tertiary hospital in Kenya, 1993-2005. Ann Afr. Surg, 2007; 1: 10-15

  5. Abdulkareem FB, Abudu EK, Awolola NA, et al. Colorec-tal carcinoma in Lagos and Sagamu,Southwest Nigeria histopathological review. World J Gastroenterol. 2008; 14;(42):6531-5.

  6. Meyer JE, Narang T, Schnoll-Sussman FH, et al. Increasing incidence of rectal cancer in patients aged younger than 40 years: an analysis of the surveillance, epidemiology, and end results database. Cancer. 2010 Sept 15 (published online Au-gust 23

  7. Andre T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment for colon cancer. N Engl J Med 2004;350:2343-2351

  8. The Kenya Oncological Research Database, accessed 02 No-vember 2010, <>