1 Nov Episode 79 Increasing diagnostic capacity by pooling and increasing sensitivity by ddPCR

Sun, 11/01/2020 - 18:55

In this episode, a guideline paper from Israel on pooling strategy is presented and we examine also the evolution towards use of ddPCR to obtain greater sensitivity.  Clearly, ddPCR on respiratory samples can reveal SARS-CoV-2 positivity in formerly negative COVID patients. In the most recent paper, it becomes clear that virus can also be detected in blood and that clearance from blood (estimated by ddPCR) is slower than in respiratory sample (by classical RT-PCR).


  1. First a paper from Israel on the practical aspects of sample pooling of NP and OP swabs 
  • Size of pool depends on prevalence: pools of 8 at low prevalence (≈1 %), pools of 5 at high prevalence (≈6 %).
  • False positivity (no individual sample positive): only 4-5 %
  • Sparing 76 % of reagents
  • Loss in sensitivity: would be acceptable (calculated based on Ct values).      


  1. Comparing various specimens for positivity in RT-PCR: Wang in JAMA: RT-PCR (targeting ORF 1a) in hospitalized patients: positivity highest in lower respiratory tract samples (BAL 93 % and sputum 72%), lower in nasal swab (63 %),  pharyngeal swab (32 %) and feces (29 %).  Almost absent in blood and urine.  Limitations:
  • Diagnostic criteria not fully clear (clinical  + RX, confirmed by SARS-CoV-2 detection), but no data on severity
  • Lower respiratory samples mainly in severely ill subjects
  • Few blood and urine samples tested.


  1. Paper by Yu in CID: Only patients who were positive in RT-PCR and/or ddPCR (commercial targeting either ORF1a and N) with respective thresholds at 38 Ct and 10 copies. 
  • Four negative and 41 single-gene positive samples tested by RT-PCR were positive according to ddPCR with viral loads ranging from 11.1 to 123.2 copies/test
  • Average VL  in sputum (17 429 ± 6920 copies/test) >  throat swabs (2552 ± 1965 copies/test, P < .001) and nasal swabs (651 ± 501 copies/test, P < .001).
  • Sputum VL in the early and progressive stages were significantly higher than that in the recovery stage (46 800 ± 17 272 vs 1252 ± 1027, P < .001)
  • Blood and urine negative (but few samples tested)


  1. A PLoS paper from Italy focuses on 55 COVID-suspected patients with RT-PCR negative NP swabs: 19 of those were positive with ddPCR (in house targeting RdRP), later 100 % confirmed with serology (only done in 14?), whereas of the 36 ddPCR negative subjects, only 5 % showed antibodies.  Table 1 suggests that the ddPCR-positives had more dyspnea and lung involvement.


  1. The  fifth selected paper from China confirms the higher sensitivity of ddPCR on throat (OP) swabs, but serology is not done here.  An interesting (and worrying) aspect is that they show a positive ddPCR in 6/14 “convalescent” patients, who were discharged, based on 2 consecutively negative RT-PCR test + clinical and radiological “cure”.   As the authors state: “Although the risk of viral transmission is unknown, the virus is replicating, leading to the increase of viral load.  Therefore, the current clinical practice that the convalescent continues to be quarantined for at least two weeks is reasonable and necessary”


  1. The paper in J Mol Diagn investigates the dynamics of plasma viral load in 52 patients with pneumonia (either moderate, severe or critical).  Importantly, plasma VL was measured with a sensitive “one step RT-ddPCR (LOD 80 copies) and throat OP VL with regular RT-PCR.
  • Remarkably, 48/52 showed viremia by one step RT-ddPCR.  The concordance between regular RT-PCR on OP swabs was only 60 %.  there were 33 paired tests with plasma viral positivity but negative oropharyngeal swab.
  • Clearance of plasma VL occurred in a proportion of moderate and severe patients, but not in critical ones.
  • At discharge 50 % of (formerly) moderate/severe patients had still plasma VL, but most of them eventually cleared plasma VL and there was no case of transmission to close contacts.