Prognostication of Breast Cancer Patients Using the Nottingham Prognostic Index in Kenyatta National Hospital

Breast cancer is a disease where cells in the breast grow out of control, and how well a patient recovers often depends on the specific characteristics of the tumor when it is first found. Doctors have long used tools to predict a patient's future, estimating the likelihood of survival based on factors like the size of the tumor, how quickly the cells are dividing, and whether the cancer has spread to nearby lymph nodes. One such tool, known as the Nottingham Prognostic Index, was developed in the West to sort patients into groups with good, moderate, or poor outlooks. It works by combining these physical measurements into a single score that suggests how long a person might live after diagnosis. However, these tools were created using data from Western populations, and it is not always clear if they work the same way for people in other parts of the world, where genetics, environment, and access to treatment can differ significantly.

In Kenya, breast cancer is the most common cancer affecting women, yet many patients arrive at the hospital with the disease already in an advanced stage. At Kenyatta National Hospital, a team of researchers decided to test whether the standard Western prediction tool could accurately forecast the future for their local patients. They looked back at the medical records of 258 women who had been diagnosed with breast cancer between 2010 and 2016. The team gathered details on the women's ages, the size and type of their tumors, whether the cancer had spread to lymph nodes, and what treatments they received. Using the standard formula, they calculated a prognostic score for each woman and placed her into one of three categories: those with a good outlook, a moderate outlook, or a poor outlook. They then compared what the tool predicted would happen against what actually happened over a five-year period.

The results revealed a striking difference between the prediction and reality. While the tool correctly identified that larger tumors, higher grades of cell abnormality, and spread to lymph nodes were linked to worse outcomes, the actual survival rates were much higher than the tool had suggested. In fact, the women in the study lived longer than the index predicted across every single group. The gap was most dramatic for the women the tool labeled as having a poor prognosis; while the index suggested a very low chance of five-year survival (31.6%), the actual survival rate for this group was nearly triple what was expected (94.5%). Even for those with moderate and good prognoses, the women survived longer than the standard model anticipated. Although the individual components of the tool correlated with prognosis, the final scores failed to distinguish survival outcomes between the groups; statistical analysis showed no significant difference in median survival rates between the good, moderate, and poor prognostic groups. This suggests that while the tool could rank patients based on tumor biology, it lost its ability to predict actual survival differences in this specific population.

This underestimation appears to be driven by the effectiveness of modern treatments available at the hospital. Nearly all the women in the study received adjuvant therapy, which includes treatments like chemotherapy or radiation given after surgery to kill any remaining cancer cells. The researchers suggest that the prediction tool, which was developed decades ago, does not fully account for the power of these contemporary treatments. The tool was built on data from a time when such therapies were less common or less effective, leading it to be overly pessimistic for patients today who receive full medical care. Additionally, the study highlighted that most patients in Kenya still present with advanced disease, with more than half arriving at stage three, a reflection of limited screening programs and barriers to early detection.

The study concludes that while the components of the prediction tool are useful for understanding the biology of the tumor, the final score should not be used to make routine clinical decisions for patients in Kenya without further testing. The tool suggests a grim future that does not match the reality of patients who receive modern care. The authors emphasize that before this index can be trusted to guide treatment plans in this region, it needs to be validated specifically for the Kenyan population. Until then, doctors should rely on the individual factors the tool measures rather than the final score, recognizing that the standard Western model may not tell the whole story for women in East Africa.