Whitley was the most cited expert on HSV in his era. Kimberlin was one of
his proteges and took over that mantle from him. More recently Caviness did
her Ph.D. on neonatal HSV. Her work I believe surpasses Kimberlin's because
she actually accounts for the risks involved in maximizing therapy and also
computes a cost-benefit value.

Age of onset

All 10 of the cases in Caviness's study were under 21 days of age, 9/10
under 14 days. In Long's editorial accompanying that study, Long admits to
being much more liberal at handing out empiric acyclovir therapy, but
acknowledges in her experience as well 21 days is the limit unless the
presentation is CNS disease alone and presenting as seizures/obtundation in
the fourth week of age. She notes an Ohio colleague would not empirically
treat past 14 days and that Kimberlin (in the editorial just before Long's)
would empirically start acyclovir only for clear signs of HSV, signs of
sepsis, or CSF mononuclear pleocytosis outside of enteroviral season.

A medscape article circa 2004  also cites 6-21 days as typical,
with rarer CNS cases presenting in the fourth week of life.

Given those multiple "expert" viewpoints, it is hard to understand why I
have seen acyclovir empirically started on well appearing infants whose only
sign was fever and who were 30 days old. But such has been the slippery
slope of "just to be on the safe side." The acyclovir shortage last winter
finally curtailed some of that over treatment.

So, JaPe, I do not base this claim on one expert. Since HSV can be acquired
postnatally, indeed into adulthood, no doubt there are anecdotal reports of
later onset CNS disease. One would also have to wonder if those anecdotal
cases might also involve some mild immunodeficiencies. I would not argue
with treating proven HSV CNS disease in a 5 week old, but you would need
something beside CSF WBCs. Any articles you can cite to the contrary
supporting something beyond 21 days as a typical age cutoff?

Effectiveness of therapy

JaPe says "We know that early treatment with Acyclovir makes all the
difference" and

"the risk benefit ratio would overwhelmingly favor overtreatment ."

Both statements are incorrect. Acyclovir helps but is not highly effective
against HSV. That is why Kimberlin pushed the limits throughout the 1990's
and early part of this century. He increased dose and increased duration of
therapy. He urged earlier empiric therapy, hoping it would improve outcome,
without full consideration of the cost or number-needed-to-harm when
expanding use of empiric therapy on those without the disease. The dose was
pushed beyond the FDA approved 30 mg/kg/d up to 45 and then 60 mg/kg/d (HD,
high dose). The length of therapy increased from 10 days to 21 days. He even
began trying the addition of oral suppressive therapy until to 6 months of
age. To my knowledge, the final result of that investigational therapy has
never been published.

Best numbers for disseminated disease are 15% survival with no acyclovir,
40% with standard therapy and 70% with high dose therapy. Good, but nowhere
close to the benefits of antibacterial therapy of bacterial meningitis. For
CNS neonatal HSV, even with high dose therapy only 30% are developmentally
normal. So while I would use acyclovir, I would not portray it as highly
effective. When doing risk/benefit analyses, the effectiveness of a therapy
needs to be factored in. Now a key question, does earlier therapy help?
Delays of 2-5 weeks are clearly not good. Is there a study that has compared
starting therapy on admission versus waiting one or two days for the PCR
result? Nope. If the child has signs besides fever, (i.e. incr LFTs, decr
plt, lesions, sepsis, seizures) I wouldn't argue with therapy of a presumed
clinical diagnosis pending confirmation. That is very different from empiric
therapy on a much larger cohort.


Using the high dose regimen for 3 weeks, 20% of infants will have
neutropenia below 1000. Renal toxicity is frequent but can be reduced by
scrupulous attention to providing hydration. Both toxicities are less if the
acyclovir can be stopped in less that 48 hours by a rule out HSV PCR test
coming back that fast. It should be noted, however that false negative HSV
PCR tests on the CSF are common, as much as 30%, so therapy should be
continued if clinical suspicion is very high. Ampicillin is very safe, with
only the occasional severe allergic reaction. Acyclovir is different. That
changes the risk-benefit calculation.


Caviness found that empiric treatment with acyclovir in febrile infants
without other signs/symptoms of HSV is cost prohibitive. Even in the face of
CSF pleocytosis, she estimated the cost at $44,000 per year of life (QALY)
saved, not per life saved. While many some investigators recommend a cutoff
of $100,000 per QALY, that is a luxurious standard that is unsustainable.
Britain's NIH has talked about $40,000, but I think even that is too high
except for clearly transient halfway technologies. So the cost-benefit ratio
is not highly in favor of empiric therapy.


I'm not sure what 3 cases you are talking about. Can you cite any research
supporting a correlation of CSF RBCs with HSV in the neonatal population? I'
ve never seen it. Then calculate the PPV and NPV of considering the RBC
count given all the traumatic taps. The specificity of a CSF RBC count is
simply too low and will only mislead.

Group B strep

There is synergy in using amp and gent - faster killing of GBS in vivo.
That, per my training, is the reason for preferring amp/gent to
cephalosporins for GBS. For coliforms, the cephalosporins have been the
superior choice. So as the risk of sepsis/meningitis shifts from GBS to
coliforms around 4 weeks of age, I recommend shifting from amp/gent to
amp/cefotax. Clinically, the differences are probably too small to measure
and I just go with the local flow. The increasing incidence of cephalosporin
resistant coliforms may be changing this wisdom, but I haven't reviewed that
in detail recently. Any articles?

Kevin Powell MD PhD FAAP

Saint Louis, MO

  -----Original Message-----
  From: JaPe [mailto:[log in to unmask]]
  Sent: Saturday, August 08, 2009 9:51 PM
  To: 'Kevin Powell'; [log in to unmask]
  Subject: RE: question


  Thanks for reminding us of the study from Baylor. I do not agree with some
of your comments:

  1.     You said...." RBCs should be considered part of a traumatic tap and
their presence/absence should not be allowed to mislead decision making.
Just one more thing you were taught in residency that has proven to be
erroneous. Drop it in the trash heap."  I am not ready to "drop it" based on
3 (Yes, three!) cases of CNS herpes out of a total cohort of 499 viral
pathogens from one center. Recall that this teaching is based on the fact
that some viruses can cause a hemorrhagic meningoencephalitis and HSV is one
of the treatable causes of meningoencephalitis in the neonatal age group. It
is not isolated RBCs' that we are referring to, but RBC's in conjunction
with CSF pleiocytosis.

  2.     You said."The risk of acyclovir increases further if your PCR isn't
back from the lab in 48 hrs. Ampicillin is very effective against Listeria.
Acyclovir's benefit to the patient (not the malpractice lawyer) is more
marginal. " I am not sure what you are suggesting here. The patient that
Todd described had meningitis. Early CNS-herpes infection has to be in the
differential diagnosis of this febrile neonate. We know that early treatment
with Acyclovir makes all the difference in a disease with devastating
neurologic sequelae. I would initiate & continue anti-viral therapy until
HSV culture or PCR is available, and the neonate remains "well."  I don't
care how long the PCR takes because the risk benefit ratio would
overwhelmingly favor overtreatment for bacterial and HSV meningitis in this
age group.

  3.     You said." There is literature to support almost all neonatal HSV
occurring before 3 weeks of age." While mostly true for disseminated & SEM
disease, CNS herpes can occur from 4-55 days. I had the privilege of
training with one of the foremost authorities on herpes infections, Dr. Rich
Whitley @ UAB, and I would urge you to read his chapter on HSV infection in
the neonatal period in Nelson's Textbook.  You are hastily embracing &
referencing a commentary by one so called "expert." I would urge caution
here before discarding the possibility of HSV meningitis in this patient.

  4.     "For Group B strep, a more likely pathogen at 5 weeks of age, amp
and gent remains the best combination." Yes, late onset GBS is the likely
bacterial etiology for meningitis @ 5 wks. However, if this is proven GBS,
gentamycin is not necessary. Ampicillin should suffice. The gent. is for
coverage of enteric gram negatives/coliforms. Although, I do not believe
that this is the best combination for expectant antimicrobial coverage. At
our center, we prefer cefotaxime because it does not require monitoring of
drug levels and is not associated with nephro- or oto-toxicity.

  Jay Pershad

  Le Bonheur

For more information, send mail to [log in to unmask] with the message: info PED-EM-L
The URL for the PED-EM-L Web Page is: