Trends in the full blood count blood test and colorectal cancer detection: a longitudinal, case-control study of UK primary care patient data

Pradeep S. Virdee, Julietta Patnick, Peter Watkinson, Jacqueline Birks, Tim HoltNIHR 2022 · DOI

TL;DR

This study reveals that subtle, long-term trends in common blood test (Full Blood Count - FBC) parameters can precede a colorectal cancer diagnosis by up to four years. By analyzing a large UK primary care dataset, the researchers identified distinct FBC patterns in patients who later developed colorectal cancer compared to those who did not, suggesting these trends could be a valuable tool for earlier detection.

What Problem Does This Paper Address?

Colorectal cancer (CRC) remains a significant health challenge, with prognosis heavily dependent on the stage at diagnosis. Unfortunately, many patients are diagnosed at later stages when treatment is less effective and survival rates are lower. Current diagnostic pathways often rely on overt symptoms, which typically appear when the disease is already advanced. While the Full Blood Count (FBC) is a routine blood test, its potential for early, pre-symptomatic detection of CRC has been underexplored. This paper addresses the gap in understanding how subtle, longitudinal changes in FBC parameters might serve as an early warning signal for CRC, potentially enabling earlier intervention and improving patient outcomes.

Methods

The researchers employed a retrospective, case-control, longitudinal analysis of UK primary care patient data. They used LOWESS smoothing to visualize trends in various FBC parameters over a 10-year period before diagnosis, comparing patients diagnosed with CRC (cases) to those without (controls). Mixed-effects models were then used to statistically model these trends, accounting for age and sex. To quantify the association between FBC trends and cancer presence, they utilized joint modeling of longitudinal and time-to-event data. Further analyses explored differences in FBC trends between different CRC tumor stages and the association between microcytic anemia and diagnosis.

Datasets

The study utilized data from the Clinical Practice Research Datalink (CPRD) GOLD, a large UK primary care database, and the UK National Cancer Registration and Analysis Service (NCRAS) for diagnosis data.

Key Insights and Ideas

  • Subtle, Long-Term Trends Matter: The most striking insight is that significant differences in FBC trends between CRC patients and controls emerge years before diagnosis, not just in the immediate pre-diagnosis period. This challenges the notion that only drastic deviations from normal ranges are clinically relevant.
  • Beyond Single Thresholds: The study highlights that FBC parameters might remain within "normal" reference ranges but still exhibit trends indicative of underlying pathology. This suggests that relying solely on single-value thresholds for referral might miss early signs.
  • A Potential for Proactive Screening: The identified trends offer a promising avenue for developing predictive models that could flag individuals at higher risk for CRC, even in the absence of alarming symptoms, potentially integrating with existing screening programs.

Key Results

  • Within four years of diagnosis, many FBC parameters showed statistically significant differences between CRC cases and controls.
  • Trends in red blood cell count, hemoglobin, white blood cell count, and platelets differed significantly between cases and controls in the years leading up to diagnosis.
  • For some parameters, like hemoglobin, trends suggested that abnormal thresholds were reached approximately 6-12 months before diagnosis, but for older cases, levels were abnormal for up to three years prior.
  • Microcytic anemia prevalence was consistently higher in cases than controls in the years preceding diagnosis, with odds of diagnosis increasing as the time to diagnosis approached.

Limitations and Future Work

The authors acknowledge several limitations:

  • Confounding Factors: FBC tests are ordered for various reasons, and these reasons are not always recorded, meaning controls may not be entirely healthy. Other relevant confounders like ethnicity, family history, and comorbidities were not available.
  • Survivor Effect: In older age groups, the "survivor effect" might influence FBC levels, as individuals surviving to older ages may have had healthier FBCs earlier on.
  • Missing Data: Red blood cell distribution width was excluded due to extensive missing data.
  • Tumor Staging Data: Missing tumor stage data for a third of cases limited analyses in this area.

Future work suggested by the authors includes developing a dynamic statistical prediction model that leverages these FBC trends to derive an individual's risk of CRC diagnosis. This model could potentially be integrated into the UK screening program to identify individuals for further investigation, possibly via fecal immunochemical tests (FIT).

Summary updated

Abstract

Trends in FBC parameters are different between patients with and without colorectal cancer for up to four years prior to diagnosis. Such trends could help earlier identification.
BibTeX
@inproceedings{virdee2022trends,
  author = {Pradeep S. Virdee and Julietta Patnick and Peter Watkinson and Jacqueline Birks and Tim Holt},
  title = {{Trends in the full blood count blood test and colorectal cancer detection: a longitudinal, case-control study of UK primary care patient data}},
  booktitle = {NIHR},
  year = {2022},
  doi = {10.3310/nihropenres.13266.2},
}
Figure