PLoS ONE (Jan 2016)

Randomized Controlled Trials to Define Viral Load Thresholds for Cytomegalovirus Pre-Emptive Therapy.

  • Paul D Griffiths,
  • Emily Rothwell,
  • Mohammed Raza,
  • Stephanie Wilmore,
  • Tomas Doyle,
  • Mark Harber,
  • James O'Beirne,
  • Stephen Mackinnon,
  • Gareth Jones,
  • Douglas Thorburn,
  • Frank Mattes,
  • Gaia Nebbia,
  • Sowsan Atabani,
  • Colette Smith,
  • Anna Stanton,
  • Vincent C Emery

DOI
https://doi.org/10.1371/journal.pone.0163722
Journal volume & issue
Vol. 11, no. 9
p. e0163722

Abstract

Read online

To help decide when to start and when to stop pre-emptive therapy for cytomegalovirus infection, we conducted two open-label randomized controlled trials in renal, liver and bone marrow transplant recipients in a single centre where pre-emptive therapy is indicated if viraemia exceeds 3000 genomes/ml (2520 IU/ml) of whole blood.Patients with two consecutive viraemia episodes each below 3000 genomes/ml were randomized to continue monitoring or to immediate treatment (Part A). A separate group of patients with viral load greater than 3000 genomes/ml was randomized to stop pre-emptive therapy when two consecutive levels less than 200 genomes/ml (168 IU/ml) or less than 3000 genomes/ml were obtained (Part B). For both parts, the primary endpoint was the occurrence of a separate episode of viraemia requiring treatment because it was greater than 3000 genomes/ml.In Part A, the primary endpoint was not significantly different between the two arms; 18/32 (56%) in the monitor arm had viraemia greater than 3000 genomes/ml compared to 10/27 (37%) in the immediate treatment arm (p = 0.193). However, the time to developing an episode of viraemia greater than 3000 genomes/ml was significantly delayed among those randomized to immediate treatment (p = 0.022). In Part B, the primary endpoint was not significantly different between the two arms; 19/55 (35%) in the less than 200 genomes/ml arm subsequently had viraemia greater than 3000 genomes/ml compared to 23/51 (45%) among those randomized to stop treatment in the less than 3000 genomes/ml arm (p = 0.322). However, the duration of antiviral treatment was significantly shorter (p = 0.0012) in those randomized to stop treatment when viraemia was less than 3000 genomes/ml.The results illustrate that patients have continuing risks for CMV infection with limited time available for intervention. We see no need to alter current rules for stopping or starting pre-emptive therapy.