Article Text

Download PDFPDF

Breast cancer diagnosis in a resource poor environment through a collaborative multidisciplinary approach: the Kenyan experience
  1. Shahin Sayed1,
  2. Zahir Moloo1,
  3. Peter Bird2,
  4. Ronald Wasike1,
  5. Wambui Njoroge3,
  6. Joseph Karanu4,
  7. Ancent Nzioka2,
  8. Omar Sherman5,
  9. Satya Prasad6,
  10. Costa Mariwa6,
  11. James Obondi Otieno7,
  12. David Chumba8,
  13. David Koech8,
  14. Daniel Mbinga5,
  15. Musa Mohammed9,
  16. Richard Njoroge9,
  17. Rajendra Chauhan1,
  18. Sudhir Vinayak1,
  19. Catherine Kyobutungi10,
  20. Mansoor Saleh11
  1. 1Department of Pathology, Surgery and Radiology, Aga Khan University Hospital Nairobi, Nairobi, Kenya
  2. 2Department of Surgery and Pathology, African Inland Church Kijabe Hospital, KIjabe, Kenya
  3. 3Department of Surgery, Provincial General Hospital Nyeri, Nyeri, Kenya
  4. 4Department of Surgery, St. Mary’s Mission Hospital Nairobi, Nairobi, Nairobi, Kenya
  5. 5Department of Pathology and Surgery, Aga Khan Hospital Mombasa, Mombasa, Kenya
  6. 6Department of Pathology and Surgery, Aga Khan Hospital Kisumu, Kisumu, Kenya
  7. 7Department of Surgery, New Nyanza General Hospital, Kisumu, Kenya
  8. 8Department of Pathology and Surgery, Moi Teaching and Referral Hospital, Eldoret, Kenya
  9. 9Department of Surgery and Pathology, Provincial General Hospital Garissa, Garissa, Kenya
  10. 10African Population and Health Research Centre, Nairobi, Kenya
  11. 11Department of Medicine, University of Alabama at Birmingham Comprehensive Cancer Center, Birmingham, Alabama, USA
  1. Correspondence to Dr Shahin Sayed, Department of Pathology, Aga Khan University Hospital Nairobi, P.O. Box 30270, 3rd Parklands Avenue, Nairobi 00100, Kenya; shaheen.sayed{at}


INTRODUCTION The majority of women with breast cancer in Kenya present with node-positive (stage II) or locally advanced Q7 disease (stage IIIB). Diagnosis is made on fine needle aspirate cytology and treatment is with surgery if resectable. Diagnostic core biopsy is available only at subspecialty hospitals. Processing and reporting of biopsy tissue are not standardised. Hormone receptor and HER2 analyses are rarely done preoperatively.

METHODS As part of a larger study investigating the prevalence of triple negative breast cancer in Kenya, a multidisciplinary workshop of collaborators from 10 healthcare facilities was held. Process gaps were identified, preanalytic variables impacting on ER/PR/HER2 discussed and training in core biopsy provided. Local remedial strategies were deliberated.

CONCLUSION We describe our experience and outcome from the workshop, which can be modelled for other resource poor settings.

  • Breast Cancer
  • Fixation
  • Immunohistochemistry
  • Breast Pathology
  • Hormone

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


According to the Nairobi Cancer Registry, breast cancer is the second most common female malignancy in Kenya.1

Although ER, PR and Human Epidermal Growth Factor Receptor 2 (HER2) are critical prognostic and predictive markers and standard of care in breast cancer diagnosis and treatment,2–5 these markers are not routinely performed in Kenya due to cost considerations, availability and/or lack of appreciation of their significance.

Interestingly, while Western data show oestrogen receptor (ER)-positive breast cancers comprising about 75% of all breast cancers,6 published data from Eastern Africa report estimates of 25%–30% ER-positive cases.7–10

Importantly, however, these studies were based on small cohorts, mostly retrospective in design, and crucial preanalytic variables such as tissue sampling, fixation and pathological assessment were not standardised.

As part of a study (funded through the GSK Ethnic Research Initiative Innovations Grant), the aim of which is to investigate the prevalence of triple negative breast cancer in the three major ethnic groups in Kenya, we organised a collaborators’ workshop to discuss standardisation of diagnostic variables. We identified gaps in current practice and shared experiences at a multidisciplinary platform, with onsite training of surgeons, pathologists, radiologists, nurses and technologists. Government officials at the Ministry of Health were also engaged in the deliberations.


The 2-day workshop was conducted on 10 and 11 February 2012 at the Aga Khan University Hospital, Nairobi (AKUHN), a 254-bed private not-for-profit University hospital that provides tertiary healthcare services with 31 satellite centres in various parts of Kenya.

A dedicated one-stop breast clinic runs daily at AKUHN. Core biopsy is the standard mode of diagnosis for all clinically and radiological suspicious breast lumps. Preoperative analysis of ER/PR/HER2 is routinely performed on an automated immunohistochemistry platform.


The main objective of the 2-day workshop was to establish a standard practice in the diagnostic process of breast cancer. Collaborating institutions included tertiary government and faith based health facilities representing each of the 10 provinces in Kenya.

Participants included a surgeon, nurse, pathologist and technologist from each of these hospitals, an epidemiologist from the African Population Health and Research Centre, heads of the departments of non-communicable diseases and research development from the Ministry of Health and a collaborating surgeon and pathologist from the Amadu Belo University, Zaria, Nigeria.

The faculty included members from the departments of radiology, surgery and pathology from AKUHN and lead co-investigator from the University of Alabama at Birmingham Comprehensive Cancer Center.

The meeting format concentrated on three components: first, overview presentations were given by faculty on the burden of breast cancer in Kenya, current diagnostic modalities and management. The next component highlighted the advantages of core biopsies as a means of diagnosis, as well as appropriate preanalytical tissue handling, testing and recommended reporting guidelines for ER/PR/HER2. The final component addressed human subject protection issues, including the consent process, patient confidentiality in clinical research and the importance of accurate data collection. Participants shared experiences in lively interactive sessions and enumerated challenges faced in their work environment.

The breakout sessions on Day 2 of the workshop was dedicated to hands-on training. The pathology technologists were trained on preparation of buffered formalin, tissue handling procedures and principles of immunohistochemistry. The pathologists were taken through the CAP grossing and reporting guidelines for breast cancer specimens and the importance of intra and inter laboratory quality control and quality assurance. The surgeons were trained in palpation and image guided core biopsy technique. Bard guns with gauge 11 and 14 needles were used to demonstrate core biopsy techniques using chicken breasts stuffed with tumour surrogates (beet roots and olives). The nurses were guided through a patient informed consenting procedure and the completion of clinical case forms. All participants were given a live demonstration of the online National Cancer Institute (NCI)-USA Good Clinical Practice (GCP)/Good Laboratory Practice (GLP) training course and encouraged to enrol.


Table 1 provides a tabulation of breast cancer services currently available at AKUHN and each of the 10 participating institutions.

Table 1

Baseline services at each participating institution

Participants shared experiences on local practice and acknowledged significant gaps and variability in breast cancer diagnostic modalities that ranged from simple clinical exam to fine needle aspirate (FNA) and/or incisional tissue biopsy with or without ultrasound support. Only three of the 11 centres were proficient in and performed core biopsies. Preanalytical handling of tissue specimens was highly variable and suboptimal (see figure 1).

Figure 1

Images highlighting the impact of suboptimal fixation of breast cancer tissues on the quality of ER/PR and HER2 immunostains. (A, B) Poorly fixed breast cancer tissue with suboptimal heterogeneous oestrogen receptor nuclear staining (H&E and IHC ×20). (C, D) Same case well-fixed section of lymph node with metastatic tumour showing homogenous 100% score for oestrogen receptor (H&E and IHC ×20).

The CAP synoptic reporting format for breast cancer was not followed as the pathologists often cited overwork. ER/PR/HER2 assessment was not routinely performed due to high costs to the patient and lack of skilled technical staff and institutional funding. Treatment in most cases remains empiric and not based on accepted prognostic and predictive factors. Correcting this diagnostic and therapeutic deficit was felt to be the most important goal of the workshop participants.

Table 2 tabulates the identified shortfalls, possible solutions and anticipated hurdles in optimising breast cancer diagnosis at various centres.

Table 2

Overview of the methodological gaps, solutions and anticipated hurdles in optimising breast cancer diagnosis in provincial centres

None of the participating centres, except AKUHN, performed onsite ER/PR/HER2 on a routine basis. Loss to follow-up following initial biopsy was a recurring theme at all centres and uniform documentation of demographic, staging and pathological diagnosis was lacking.

Participants suggested remedial strategies drawing upon their experience. Anecdotal data suggest that to address the issue of loss to follow-up, it was important to involve the patient's immediate family in the care and management process from the outset and also document contact details of next of kin and the local area chief. Data suggest that these measures were successful in one Nigerian setting.11

Onsite continuous medical education by local experts in breast cancer care was suggested to encourage appropriate triaging by clinical officers who run the primary health facilities where the first patient contact usually occurs.

Teleconferencing and use of mobile technology to enhance connectivity would further contribute to standardisation of care for the breast cancer patient.

Short-term goals would include sharing of resources and developing guidelines for a locally appropriate and sustainable referral system.

Medium-term goals would be to plan monthly tumour boards with collaborating institutions.

Table 3 demonstrates the preworkshop and 6-month postworkshop measurable improvement metrics in participating centres as an outcome of the workshop.

Table 3

Improvement in objectively measurable metrics as an outcome of the training workshop


Much has been made of the difference in prevalence of various ER/PR/HER2 based breast cancer subtypes, including triple negative breast cancer, in Africa as compared to the West;12 ,13 yet, little attention has been paid to differences in basic tissue procurement and fixation techniques that are known to significantly alter morphological and ER/PR/HER2 immunohistochemistry results.

The ‘Breast Health Global Initiative’ resource-stratified guidelines (2005) recommend testing for ER and PR and HER2 in breast cancers at all levels of resources, stating these as key drivers that determine standard of care and outcome of breast cancer patients. Consequently, pathology reports should follow established guidelines and include appropriate diagnostic and prognostic/predictive information based on the submitted specimen type.14 ,15

However, reliable data on the prevalence of ER/PR/HER2 in breast cancer are lacking.7–10 Studies from Sub-Saharan Africa have been based on small cohorts, mostly retrospective in design, and crucial preanalytic variables such as tissue sampling, fixation and pathological assessment were not standardised.

In addition to the lack of resources and expertise to perform immunohistochemistry at the provincial hospital level, one of the major gaps highlighted in our workshop was the lack of appreciation of the critical importance of preanalytic tissue processing, specifically the use of 10% buffered formalin and optimal tissue fixation time and technique.

The American Society of Clinical Oncology (ASCO)/CAP guideline document notes that as a result of variation in preanalytic processes, thresholds for positivity and interpretation criteria are often responsible for variability in ER/PR results which can be as high as 20%.16

As Akarolo-Anthony et al17 suggest, the age adjusted prevalence of hormone receptor subtypes from Africa may not be different from that in other populations and the reported differences may be a consequence of poor tissue handling and laboratory processing practices.

The workshop participants therefore stressed on the urgent need to address the standardisation of the preanalytic diagnostic processes in order to ensure accurate prospective data that will better inform policy on breast cancer management.

Parenthetically, the added cost of establishing and maintaining high quality standards of breast pathology services, including reliable and reproducible ER/PR/HER2 testing, is often recouped when appropriate treatment is offered thus leading to improvement in overall survival and productive life-years, as well as avoiding needless morbidities and associated costs. Bird et al8 reported a cost analysis of inappropriate treatment with hormonal therapies when ER/PR testing is not done in resource poor settings and showed that in the long run it is cheaper to perform ER/PR/HER2 upfront rather than providing empiric tamoxifen over 5 years.

In most resource challenged settings, FNA cytology is a cost effective modality for the evaluation of a breast mass. However, FNA cytology requires skilled personnel for interpretation. In addition, it does not define invasive disease, so occasional in situ cancer will be overtreated.

On the other hand, a diagnostic core biopsy can also be performed on an outpatient basis, is less amenable to fixation artefacts associated with lumpectomy/mastectomy specimens and if properly performed provides ample tissue for ancillary testing, including immunohistochemistry. In the local setting where most patients present late and with advanced stage disease that could be managed first by neo adjuvant therapy, a pretreatment core biopsy thus may be the only opportunity to determine de novo ER/PR/HER2 status and dictate selection of neoadjuvant therapy.

Our multidisciplinary workshop, the first of its kind in Kenya, proved very successful since it brought together healthcare providers dedicated to breast cancer care from institutions with limited resources and technical capabilities. As is evident in our 6-month progress report, the education, awareness and provision of handson experience was practice-changing.

A published report from Ghana demonstrates the success of such initiatives. A week long training programme in the performance of core needle biopsies of breast masses for medical staff involved with clinical care was held as part of an ongoing collaboration between the University of Michigan (UM) and Komfo Anokye Teaching Hospital (KATH) in Kumasi. Following the training, histological concordance on core biopsies between the KATH and UM pathologists was 100% for benign and malignant lesions of the breast. The authors concluded that the design and implementation of appropriate diagnostic biopsy programmes are important for the delivery of quality, efficient breast cancer care in developing nations. This study emphasised the multidisciplinary approach in the success of the training programme.18

The Ministry of Health in Kenya recently drafted a National Cancer Program the aims of which are to raise public awareness, screening and develop management guidelines for cancer in consultation with a multidisciplinary team of medical specialists, and thus their participation in this workshop was seen as integral to the future sustainability of practice improvement.

Despite resource limitations in developing countries, our report shows that it is possible to improve the practice of breast cancer diagnosis and treatment through educational workshops fostering a collaborative relationship with stakeholders and seeking local solutions through sharing of standard of care methodologies, centralising specialised testing and establishing objectively measurable improvement metrics. The outcome of our training workshop is a good example of what can be achieved with modest funding and support.

Our experience underscores that for practice change to be effective, relevant and sustainable, appropriate government institutions need to be involved from the outset and public private partnerships encouraged.

Take-home messages

  • Breast cancer is the second most common malignancy in Kenya and yet treatment at most centres remains empiric and without consideration of key prognostic factors such as oestrogen receptor (ER), PR and HER2 receptor.

  • The accuracy of ER, PR and HER2 testing by immunohistochemistry dependent upon critical yet simple tissue handling methodologies, such as use of 10% neutral buffered formalin fixative and importance of fixation time, are often not adhered to at poorly equipped centres in the developing world.

  • Progress in the diagnosis and treatment of breast cancer in underserved populations can be achieved by educating healthcare providers from various disciplines through both didactics as well as hands-on workshops.


The authors would like to acknowledge GSK for funding the project through their GSK-ERI research grant. The authors would also like to acknowlegde the Director of Medical Services, the Departments of Non Communicable Diseases and Research and Development in the Ministry of Health, Kenya, for their adminstrative support for the study. In addition the authors would like to acknowledge the support of all the nurses and technologists at the participating centres, Dr Joshua Chege, senior resident in Anatomic Pathology and Innocent Abayo , research assistant at AKUHN. We would also like to appreciate the support of Dr El-Nasir Lalani, the Dean of Research and Post Graduate studies at Aga Khan University .



  • Contributors The research concept and study was designed by SS and MS with input from ZM. The manuscript was written by SS and MS, with review and input by PB and ZM. All the other authors are collaborators who enroled patients in the study and contributed to the data.

  • Funding 2011 GlaxoSmithKline–ERI award.

  • Competing interests None.

  • Ethics approval Aga Khan University Hospital Institutional Review Board—Nairobi.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Collaborators Dr Raymond Oigara, Dr Richard Nyagah (Departments of Surgery, St Mary's Mission Hospital, Nairobi) Dr Andrew Gachii, Dr Julius Githaiga (Departments of Pathology and Surgery, Kenyatta National Hospital).