Buccal swabs: cheek cells when a tube isn't practical
A cotton or flocked swab rubbed against the inner cheek lifts buccal epithelial cells for DNA. Yield is lower than saliva and far below blood — but the swab earns its place with infants, frail patients, and dried-card archiving. An educational explainer from the PROMPT Registry desk.
From cheek to sequencer
The buccal route is short, but each step protects a small DNA mass — so technique matters more here than with a high-yield saliva tube.
- 01
1 · Prepare the mouth
No food, drink, gum, or smoking for ~30 minutes before sampling. For infants, no pacifier or finger for an hour and no breastfeeding for ~3 hours — milk and oral debris dilute the cell harvest.
- 02
2 · Rub firmly, both cheeks
Press the swab against the inner cheek and rotate with moderate pressure for 30–60 seconds, then repeat on the second cheek with the same swab. Firm contact dislodges nucleated buccal epithelial cells — light dabbing leaves most cells behind.
- 03
3 · Air-dry completely
Stand the swab tip-up in a clean, dry holder for 30–60 minutes until visibly dry. Drying suppresses mould and bacterial overgrowth that degrade DNA in transit; sealing a damp swab is the most common avoidable failure.
- 04
4 · Stabilise or card
Sheath the dry swab in its transport tube, or press the head onto an
FTAcard. Card matrices immobilise nucleic acids, strip PCR inhibitors during washing, and shield DNA from UV. - 05
5 · Extract & amplify
The lab elutes or punches the sample, extracts genomic DNA, then PCR-enriches a hereditary-cancer panel (
BRCA1,BRCA2,PALB2,ATM,CHEK2,TP53…) for sequencing and ACMG variant classification.
Swab vs saliva vs blood
Yields and stability are order-of-magnitude figures from device specifications and peer-reviewed comparisons — individual results vary widely.
| Method | Sample | Typical DNA yield | Stability | Notes |
|---|---|---|---|---|
| Buccal swab | 1–2 swab heads | ~5–15 µg (mean ~11 µg) | Dry swab ~days; on FTA card >10 yr | Lowest yield; PCR success ~62–89% of subjects; best when a tube is impractical |
Saliva (OG-500) | 2 mL | ~110 µg (range 15–>300 µg) | Room temp, up to ~30 months | Lysis buffer stabilises at collection; richest non-invasive yield |
| Whole blood (EDTA) | 3–10 mL | Tens–hundreds of µg | 2–8 °C, days; frozen for years | Highest-quality DNA; needs phlebotomy and cold handling |
Table 1. Typical genomic-DNA yield and handling by collection method (germline workflows).
Relative DNA yield by method
Mean genomic-DNA mass recovered, normalised to the buccal-swab baseline. Saliva and blood deliver far more starting material than a cheek swab.
~11 µg mean
~110 µg typical
method-dependent, highest quality
Children, the frail, and the field
The swab's advantage is not yield — it is tolerability and logistics. A neonate or young child who cannot spit fills a saliva tube poorly, but holds still for a few seconds of gentle cheek rubbing. The same is true for elderly or unwell patients with dry mouth, reduced muscle control, or swallowing difficulty.
Swabs also travel light: dried onto an FTA card they ship by ordinary post, archive at room temperature for years, and need no buffer or refrigeration — useful for remote collection and biobank backup. The trade-off is the small DNA mass, so good technique and a clean dry-down are non-negotiable.
Glossary
Terms that recur across collection and downstream analysis.
- Buccal epithelial cell
- A cheek-lining cell carrying a full nuclear genome — the DNA source a swab collects.
- FTA card
- A chemically treated paper matrix that lyses cells on contact, immobilises and preserves DNA at room temperature, and removes PCR inhibitors during washing.
- Germline DNA
- Inherited DNA present in every nucleated cell; the substrate for hereditary-cancer testing, equally accessible from cheek, saliva, or blood.
- VUS
- Variant of uncertain significance — an ACMG tier where evidence is insufficient to call a finding pathogenic or benign; not actionable for clinical decisions.
Common questions
Is a swab result less reliable than blood?
For SNP and panel genotyping, properly collected buccal DNA gives concordant results — the genome is identical regardless of source. The difference is quantity, not identity: a low-yield or contaminated swab is more likely to fail amplification and need a recollect, whereas blood and saliva have ample margin.
Why does a swab yield so much less DNA than saliva?
A saliva tube captures cells suspended throughout 2 mL of fluid in a stabilising buffer; a swab lifts only the cells it physically contacts on the cheek surface. Reported means are roughly 11 µg for a swab versus ~110 µg for an OG-500 saliva sample.
How do I get the most DNA from a swab?
Avoid food and drink for ~30 minutes beforehand, press firmly and rotate for the full 30–60 seconds on each cheek, use both cheeks with the same swab, and air-dry completely before sealing. Firm pressure and a clean dry-down are the two highest-leverage steps.
Can a swab be used for a baby or an elderly relative?
Yes — this is the swab's strongest use case. It needs no spitting and tolerates dry mouth or limited cooperation. For infants, wait an hour after any pacifier or feeding so milk and debris don't dilute the sample. Any collection for a real diagnostic decision should be arranged through a clinician or genetic counsellor.
- [1]NCI. National Cancer Institute. BRCA Gene Changes: Cancer Risk and Genetic Testing Fact Sheet — >60% lifetime breast-cancer risk in carriers vs ~13% baseline.↗
- [2]Cancer Epidemiol Biomarkers Prev 2006. Rylander-Rudqvist T et al. Quality and quantity of saliva DNA obtained from the self-administered Oragene method vs buccal swabs; FTA-card vs swab archive stability.↗
- [3]DNA Genotek. Oragene OG-500 product specification — ~110 µg mean DNA from 2 mL saliva (range 15–>300 µg); room-temperature stability up to ~30 months.↗
- [4]BMC Research Notes 2018. Evaluation of buccal swabs for pharmacogenetics — DNA yield, quality, and PCR performance relative to saliva and blood.↗
Match the method to the patient
Lower yield is acceptable when a swab is the only practical option. Read the wider collection set and how stored samples hold up before a sequencing run.