Skin biopsy processing for rapid molecular diagnosis and histopathologic interpretation: application to Kaposi sarcoma in East Africa

Abstract

Background: Kaposi sarcoma (KS) is a cancer of viral origin (Kaposi sarcoma-associated herpesvirus; KSHV) for which the detection of KSHV DNA is an attractive target for a rapid, automatable diagnostic test. We previously demonstrated favorable diagnostic accuracy using loop-mediated isothermal amplification (LAMP) to quantitate KSHV DNA in lesional skin biopsies, though extracting DNA from the punch biopsies was the time-limiting step. Herein, we describe the development of a biopsy processing tool called Slicer to enable rapid nucleic acid testing in addition to traditional histopathological interpretation.

Methods: Slicer divides skin punch biopsies into two ½-cylinders and a thin, cross-sectional slice. The thin slice enables a previously demonstrated, equipment-free alkaline extraction termed ColdSHOT while the remaining ½-cylinders are available for histopathological diagnosis and additional molecular testing as needed. Slicer prototypes were used on skin punch biopsies collected from patients in Uganda who were referred for clinical suspicion of KS.

Results: For 27 patient samples, the combination of Slicer and ColdSHOT sample processing with LAMP testing resulted in qualitative KSHV DNA detection that was fully concordant with US-based histopathological diagnoses. Additional analysis demonstrated compatibility of Slicer and ColdSHOT with qPCR for KSHV DNA quantitation.

Conclusions: These results warrant further investigation using a larger set of skin biopsies and indicate that the Slicer and ColdSHOT could enable accurate KS diagnosis within a few hours of biopsy collection with minimal equipment.

Keywords: Kaposi sarcoma; Nucleic acid testing; Point-of-care testing; Skin sample processing.

Fig. 1

Slicer design and use. Slicer is composed of a blade holder which houses two razor blades in parallel and a biopsy holder which features a silicone-backed chamber to hold a skin punch biopsy (a). Slicer prototypes were 3D printed in polylactic acid and used to partition 5 mm diameter skin punch biopsies into two ½-cylinder portions and a thin, cross-sectional slice. The representative sample shown, oriented with the epidermis upwards, was collected from a thawed, once-frozen human skin sample (b). To use Slicer, the user loads a punch biopsy into the biopsy holder and presses it firmly into the blade holder. After retrieving the biopsy holder, the ½-cylinders are easily retrieved, and the spacer is used to lift the slice from between the blades for collection (c)

Fig. 2

Slicer development on thawed, once-frozen porcine and human skin punch biopsies. Prototypes with varying spacer thickness and razor blade were used on porcine (a) and human (b) skin punch biopsies (n = 3 and 2, respectively). After screening, two versions (shaded bars in b) were selected for further testing with eight additional human biopsies. The slices were evaluated quantitatively (C) and qualitatively (D). Slice thickness was estimated from Eq. 1. Data shown is mean and absolute range for a-b and mean and SD for c

Fig. 3

ColdSHOT optimization on thawed, once-frozen porcine tissue slices. DNA yields, quantitated by a LAMP assay targeting GAPDH, were compared across varying NaOH concentrations and final extraction volumes for a one-hour incubation (a). DNA yields were measured as a function of time using 100 mM NaOH and a final volume of 750 µl (b). Data shown is mean and absolute range

Fig. 4

LAMP data comparing the ColdSHOT and DNeasy yields for KSHV detection in patient samples. KSHV DNA content was quantitated via LAMP using DNA extracted by ColdSHOT (thin slice) and by DNeasy (¼-cylinder). Mean copy number and absolute range are shown with US-based histopathological diagnoses indicated below (a). DNA yields for GAPDH (all samples) and KSHV (KS-present samples only) were compared between the DNeasy (¼-cylinder) and ColdSHOT (thin slice). Mean copy number and standard deviation are shown (b). Bland-Altman analyses were used to compare DNeasy to ColdSHOT for the GAPDH (c) and KSHV (d) assays with 95% limits of agreement indicated

Fig. 5

PCR data for the ColdSHOT yields from thin tissue slices collected from patient samples. Copies of GAPDH and KSHV were quantitated per 5 µl reaction. Mean copy number and absolute range are shown with US-based histopathological diagnoses indicated below (a). Bland-Altman analyses were used to compare GAPDH (b) and KSHV (c) quantitation via PCR and LAMP for the ColdSHOT extraction with 95% limits of agreement indicated