# Tumor and Adjacent Normal Tissue

> The cells contained in tumor and adjacent normal tissue, the advantages of paired comparisons, and the interpretive limits of bulk RNA-seq.

Studies comparing tumor with adjacent normal tissue collect a tumor sample and a nearby tissue sample classified as non-tumor from each patient. These are not tissues from unrelated people. They form **a matched pair from the same patient**.

This comparison helps identify changes in tumor tissue while controlling for much of the genetic background and living environment that differs between patients. Interpretation goes wrong, however, if tumor tissue is treated as a sample containing only cancer cells or adjacent normal tissue as tissue from a healthy person.

## Tumor tissue does not contain only cancer cells

Tissue obtained by surgery or biopsy contains more than cancer cells.

| Component mixed into the tissue | Possible effect on the RNA-seq result |
| --- | --- |
| Malignant epithelial cells | Expression related to tumor proliferation, metabolism, and invasion |
| Immune cells | Expression related to inflammation and immune responses |
| Fibroblasts and stroma | Expression related to the extracellular matrix and tissue remodeling |
| Blood vessels and endothelial cells | Expression related to angiogenesis and oxygen supply |
| Remaining normal cells | Expression related to the organ's normal functions |

The cells, molecules, and blood vessels surrounding cancer cells are called the **tumor microenvironment**. The tumor and its microenvironment influence one another, so expression changes measured from tissue contain signals from both.

## Bulk RNA-seq is an average across the tissue

In bulk RNA-seq, RNA is extracted from every cell in a tissue piece at once. A count for one gene combines RNA produced by cancer cells, immune cells, fibroblasts, and other cells.

For example, an immune-related gene that appears higher in Tumor has at least two possible explanations.

1. The same number of immune cells expressed the gene more strongly.
2. The proportion of immune cells in the tissue increased.

Bulk RNA-seq alone cannot fully separate these explanations. The phrase "increased in tumor tissue" is therefore closer to the data than "increased in cancer cells."

Distinguishing the cell of origin requires additional information such as tumor purity from pathology slides, cell-type deconvolution, single-cell RNA-seq, or spatial transcriptomics.

## Why use adjacent normal tissue as the control?

Adjacent normal tissue, also called adjacent uninvolved tissue, is tissue near the tumor that researchers classify as non-tumor. Because it comes from the same patient, many factors that vary between people, including age, sex, genetic background, and lifestyle, can be substantially controlled.

| Advantage | Limitation |
| --- | --- |
| Can be compared directly as a pair with Tumor from the same patient | Not guaranteed to match the same organ in a healthy person |
| Controls for the person's genetic background and systemic condition | May be affected by inflammation and the tumor microenvironment |
| Lets a statistical model account for patient-specific baselines | Molecular changes around the tumor may already be present |

Adjacent normal tissue is therefore **tissue near a tumor that was classified as normal**, not normal tissue obtained from a cancer-free person. Healthy tissue, adjacent normal tissue, and tumor tissue are three distinct comparison groups.

## The question answered by a paired comparison

If both tissues were collected from the same patient, the analysis should preserve that pairing. A paired design such as [`design = ~ patient + condition`](/en/reference/design-formula/) first accounts for each patient's expression baseline, then finds changes that recur from Normal to Tumor within the same patient.

This design can support the following statements:

- How expression differs between tumor tissue and adjacent normal tissue in the patients included in the study
- Whether that difference recurs in the same direction across patients

A paired design alone cannot establish the following:

- Whether a change is a cause or consequence of cancer
- Whether a change occurred inside the cancer cells or reflects a change in cell composition
- How different the tissue is from the same organ in healthy people
- Whether cancer subtypes or age-at-onset groups differ from one another

Comparisons between groups require data from each group. Determining cellular origin requires single-cell or spatial data and validation experiments.

### Sources

- [US National Cancer Institute: cancer and the tumor microenvironment](https://www.cancer.gov/about-cancer/understanding/what-is-cancer)
- [US National Cancer Institute Dictionary of Cancer Terms: tumor microenvironment](https://www.cancer.gov/publications/dictionaries/cancer-terms/def/tumor-microenvironment)