Polymerase chain reaction (PCR) analysis to study loss of heterozygosity (L
OH) and microsatellite instability (MSI) in tumors is widely used. Microdis
section techniques are applied to obtain tumor-specific tissue cells. By mi
crodissection, however, the amount of template DNA extracted may vary consi
derably and interfere with optimal PCR amplification. To circumvent LOH and
MSI misinterpretations due to low DNA input, we have assessed the critical
level of DNA input for reliable PCR analysis. PCR analysis was performed b
y using 18 polymorphic markers (mono-, di-, tri-, and tetranucleotide) on D
NA derived from both paraffin-embedded, formalin-fixed, and fresh frozen tu
mor specimens at template input levels ranging from 0.05 to 25.0 ng. We sho
w a highly significant relation between DNA input and the occurrence of LOH
and MSI artifacts. Furthermore, for DNA extracted from paraffin-embedded m
aterial, the percentage of LOH artifacts is significantly higher compared w
ith DNA extracted from frozen tissue. For reliable PCR analyses using a mon
o-, di-, tri-, or tetranucleotide marker, a minimum of 10.0 ng DNA is requi
red when DNA is isolated from formalin-fixed, paraffin-embedded tissue and
5.0 ng when isolated from fresh frozen tissue. HUM PATHOL 31: 1414-1419. Co
pyright (C) 2000 by W.B, Saunders Company.