Malignant Hyperthermia Deaths Related to Inadequate Temperature Monitoring, 2007-2012: A Report from The North American Malignant Hyperthermia Registry of the Malignant Hyperthermia Association of the United States

Abstract

Background:
AMRA (adverse metabolic or muscular reaction to anesthesia) reports submitted to The North American Malignant Hyperthermia Registry of the Malignant Hyperthermia Association of the United States from 1987 to 2006 revealed a 2.7% cardiac arrest and a 1.4% death rate for 291 malignant hyperthermia (MH) events. We analyzed 6 years of recent data to update MH cardiac arrest and death rates, summarized characteristics associated with cardiac arrest and death, and documented differences between early and recent cohorts of patients in the MH Registry. We also tested whether the available data supported the hypothesis that risk of dying from an episode of MH is increased in patients with inadequate temperature monitoring.

Methods:
We included U.S. or Canadian reports of adverse events after administration of at least 1 anesthetic drug, received between January 1, 2007, and December 31, 2012, with an MH clinical grading scale rank of "very likely MH" or "almost certain MH." We excluded reports that, after review, were judged to be due to pathologic conditions other than MH. We analyzed patient demographics, family and patient anesthetic history, anesthetic management including temperature monitoring, initial dantrolene dose, use of cardiopulmonary resuscitation, MH complications, survival, and reported molecular genetic DNA analysis of RYR1 and CACNA1S. A one-sided Cochran-Armitage test for proportions evaluated associations between mode of monitoring and mortality. We used Miettinen and Nurminen's method for assessing the relative risk of dying according to monitoring method. We used the P value of the slope to evaluate the relationship between duration of anesthetic exposure before dantrolene administration and peak temperature. We calculated the relative risk of death in this cohort compared with our previous cohort by using the Miettinen and Nurminen method adjusted for 4 comparisons.

Results:
Of 189 AMRA reports, 84 met our inclusion criteria. These included 7 (8.3%) cardiac arrests, no successful resuscitations, and 8 (9.5%) deaths. Of the 8 patients who died, 7 underwent elective surgeries considered low to intermediate risk. The average age of patients who died was 31.4 ± 16.9 years. Five were healthy preoperatively. Three of the 8 patients had unrevealed MH family history. Four of 8 anesthetics were performed in freestanding facilities. In those who died, 3 MH-causative RYR1 mutations and 3 RYR1 variants likely to have been pathogenic were found in the 6 patients in whom RYR1 was examined. Compared to core temperature monitoring, the relative risk of dying with no temperature monitoring was 13.8 (lower limit 2.1). Compared to core temperature monitoring, the relative risk of dying with skin temperature monitoring was 9.7 (1.5). Temperature monitoring mode best distinguished patients who lived from those who died. End-tidal CO2 was the worst physiologic measure to distinguish patients who lived from those who died. Longer anesthetic exposures before dantrolene were associated with higher peak temperatures (P = 0.00056). Compared with the early cohort, the recent cohort had a higher percentage of MH deaths (4/291 vs 8/84; relative risk = 6.9; 95% confidence interval, 1.7-28; P = 0.0043 after adjustment for 4 comparisons).

Conclusions:
Despite a thorough understanding of the management of MH and the availability of a specific antidote, the risk of dying from an MH episode remains unacceptably high. To increase the chance of successful MH treatment, the American Society of Anesthesiologists and Malignant Hyperthermia Association of the U.S. monitoring standards should be altered to require core temperature monitoring for all general anesthetics lasting 30 minutes or longer.

Full text

December 2014 • Volume 119 • Number 6 www.anesthesia-analgesia.org 1359
Copyright © 2014 International Anesthesia Research Society
DOI: 10.1213/ANE.0000000000000421
BACKGROUND: AMRA (adverse metabolic or muscular reaction to anesthesia) reports submitted
to The North American Malignant Hyperthermia Registry of the Malignant Hyper thermia Association
of the United States from 1987 to 2006 revealed a 2.7% cardiac arrest and a 1.4% death rate for
291 malignant hyperthermia (MH) events. We analyzed 6 years of recent data to update MH cardiac
arrest and death rates, summarized characteristics associated with cardiac arrest and death, and
documented differences between early and recent cohorts of patients in the MH Registry. We also
tested whether the available data supported the hypothesis that risk of dying from an episode of
MH is increased in patients with inadequate temperature monitoring.
METHODS: We included U.S. or Canadian reports of adverse events after administration of at least 1
anesthetic drug, received between January 1, 2007, and December 31, 2012, with an MH clinical grad-
ing scale rank of “very likely MH” or “almost certain MH.” We excluded reports that, after review, were
judged to be due to pathologic conditions other than MH. We analyzed patient demographics, family and
patient anesthetic history, anesthetic management including temperature monitoring, initial dantrolene
dose, use of cardiopulmonary resuscitation, MH complications, survival, and reported molecular genetic
DNA analysis of RYR1 and CACNA1S. A one-sided Cochran-Armitage test for proportions evaluated asso-
ciations between mode of monitoring and mortality. We used Miettinen and Nurminen’s method for
assessing the relative risk of dying according to monitoring method. We used the P value of the slope
to evaluate the relationship between duration of anesthetic exposure before dantrolene administration
and peak temperature. We calculated the relative risk of death in this cohort compared with our previous
cohort by using the Miettinen and Nurminen method adjusted for 4 comparisons.
RESULTS: Of 189 AMRA reports, 84 met our inclusion criteria. These included 7 (8.3%) cardiac arrests,
no successful resuscitations, and 8 (9.5%) deaths. Of the 8 patients who died, 7 underwent elective
surgeries considered low to intermediate risk. The average age of patients who died was 31.4 ± 16.9
years. Five were healthy preoperatively. Three of the 8 patients had unrevealed MH family history.
Four of 8 anesthetics were performed in freestanding facilities. In those who died, 3 MH–causative
RYR1 mutations and 3 RYR1 variants likely to have been pathogenic were found in the 6 patients in
whom RYR1 was examined. Compared to core temperature monitoring, the relative risk of dying with
no temperature monitoring was 13.8 (lower limit 2.1). Compared to core temperature monitoring, the
relative risk of dying with skin temperature monitoring was 9.7 (1.5). Temperature monitoring mode
best distinguished patients who lived from those who died. End-tidal CO2 was the worst physiologic
measure to distinguish patients who lived from those who died. Longer anesthetic exposures before
dantrolene were associated with higher peak temperatures (P = 0.00056). Compared with the early
cohort, the recent cohort had a higher percentage of MH deaths (4/291 vs 8/84; relative risk = 6.9;
95% condence interval, 1.7–28; P=0.0043 after adjustment for 4 comparisons).
CONCLUSIONS: Despite a thorough understanding of the management of MH and the availability of a
specic antidote, the risk of dying from an MH episode remains unacceptably high. To increase the chance
of successful MH treatment, the American Society of Anesthesiologists and Malignant Hyperthermia
Association of the U.S. monitoring standards should be altered to require core temperature monitoring
for all general anesthetics lasting 30 minutes or longer. (Anesth Analg 2014;119:1359–66)
From *The North American Malignant Hyperthermia Registry of the
Malignant Hyperthermia Association of the United States, Pittsburgh,
Pennsylvania; †Department of Anesthesiology, University of Pittsburgh
Medical Center, Pittsburgh, Pennsylvania; ‡Olympia Anesthesia Associates,
Olympia, Washington; §Department of Anesthesiology and Pain Medicine,
University of California at Davis, Davis, California; and ∥Department of Public
Health Sciences, Penn State College of Medicine, Hershey, Pennsylvania.
Accepted for publication July 9, 2014.
Funding: Funding from the Department of Anesthesiology, University of
Pittsburgh Medical Center, through the Safar Fund and other research funds.
Conict of Interest: See Disclosures at the end of the article.
This report was previously presented, in part, at the Anesthesiology
2013 (Annual Meeting of the American Society of Anesthesiologists), San
Malignant Hyperthermia Deaths Related to Inadequate
Temperature Monitoring, 2007–2012: A Report from The
North American Malignant Hyperthermia Registry of the
Malignant Hyperthermia Association of the United States
Marilyn Green Larach, MD, FAAP,* Barbara W. Brandom, MD,*† Gregory C. Allen, MD, FRCPC,‡
Gerald A. Gronert, MD,§ and Erik B. Lehman, MS∥
Francisco, California, October 13, 2013, and the Malignant Hyperthermia
Scientific Conference 2013 (Clinical Significance of Ryanodine Receptor-1
Gene Variants: New Insights into Pathophysiology, Diagnosis, and
Treatment of RYR1-Linked Diseases), Toronto, Canada, November 1,
2013.
Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions of
this article on the journal’s website (www.anesthesia-analgesia.org).
Section Editor: James DiNardo
Society for Pediatric Anesthesia
Reprints will not be available from the authors.
Address correspondence to Marilyn Green Larach, MD, FAAP, The
North American Malignant Hyperthermia Registry of the Malignant
Hyperthermia Association of the United States, The North American MH
Registry, Department of Anesthesiology, University of Pittsburgh Medical
Center (UPMC) Mercy Hospital, 8th Floor, Bldg. B, Ermire 1400 Locust St.,
Pittsburgh, PA 15219. Address e-mail to mlarach@gmail.com.

1360 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Fatal Malignant Hyperthermia: Inadequate Temperature Monitoring
Malignant hyperthermia (MH) is an autosomal dom-
inant myopathy (MIM No. 145600)a manifested
by sustained skeletal muscle hypermetabolism
related to altered calcium homeostasis. In human suscep-
tible individuals, MH is usually triggered by exposure to
volatile anesthetic drugs and/or the depolarizing muscle
relaxant, succinylcholine. MH has been associated with
abnormalities in the RYR1 (ryanodine receptor type 1 gene),
encoding the skeletal muscle isoform of the calcium release
channel of the sarcoplasmic reticulum,1 and CACNA1S,
encoding the α1 subunit of the L-type calcium channel
isoform of the sarcolemma (dihydropyridine receptor).2 In
the United States, clinical testing for RYR1 MH-associated
mutations began in 2005. Thirty-one MH-causative muta-
tions have been described in RYR1.b
MH mortality has been estimated as 0.0082 per 100,000
U.S. surgical inpatients, constituting 1% of all anesthesia-
related deaths for the years 1999 to 2005.3 We have previously
reported our ndings on MH events voluntarily submitted
to The North American Malignant Hyperthermia Registry
(Registry) of the Malignant Hyperthermia Association of
the United States (MHAUS) from 1987 through 2006. This
study found a 2.7% cardiac arrest rate and a 1.4% mortality
rate for 291 MH events. In this early cohort, cardiac arrest
and death were associated with muscular build, the compli-
cation of disseminated intravascular coagulation, and a lon-
ger time period between anesthetic induction and maximal
end-tidal carbon dioxide (CO2).4 Also, temperature abnor-
malities were the rst to third MH sign in 63.5% of patients
with a median temperature maximum of 39.1°C. As would
be expected, there was an association between peak temper-
ature and risk of disseminated intravascular coagulation.5
The MHAUS Hotline continues to receive reports of MH
events complicated by cardiac arrest and death. We therefore
evaluated reports received by the Registry from 2007 through
2012 (recent cohort) to update MH cardiac arrest and death
rates, summarized the characteristics associated with car-
diac arrest and death, and documented differences between
early and recent cohorts of patients in the MH Registry. We
also tested whether the available data support the hypoth-
esis that the risk of dying from an episode of MH is increased
in patients with inadequate temperature monitoring.
METHODS
The University of Pittsburgh IRB deemed this study exempt.c
The study cohort was identied as shown in Figure1. We
examined AMRA (adverse metabolic or muscular reaction
to anesthesia) reports received by the Registry from January
1, 2007, through December 31, 2012. The AMRA report is
a standardized form including relevant medical and anes-
thetic history that clinicians complete and voluntarily
submit to the Registry after a suspected MH incident.d As
shown in Figure1, we included AMRA reports document-
ing episodes that occurred in the United States or Canada,
involved at least 1 anesthetic drug or neuromuscular
blocker given before the event, and ranked as a “very likely”
or “almost certain” MH event on the MH clinical grading
scale (CGS).e,6
We excluded 3 AMRA reports that described events
before January 1, 2005. We also excluded 5 AMRA reports
when all authors agreed that the underlying pathologic
condition was not MH (e.g., aspiration, asthma, systemic
absorption of insufated CO2 during laparoscopic surgery,
myopathies other than central core disease, seizure, sepsis,
surgical complication). For in-depth analysis, reviewers
were not blinded to outcome and studied the entire case
report complete with molecular genetic analysis and free
text added by reporting health care professionals.f
The dataset included demographic data (age, gender,
muscular body build), family MH history, previous anes-
thetic history of unusual metabolic responses, adverse anes-
thetic response (anesthetic drugs, ventilation mode, monitor
type including specic temperature monitor site), location
in which the suspected MH episode was rst recognized
Figure 1. Flow sheet to identify the cohort of patients included in
this analysis.
eDue to unreliability of reported myoglobin units in the AMRA forms, myo-
globinuria and myoglobinemia were excluded from the muscle breakdown
category of the clinical grading scale calculation. Also, points for rapid
reversal of metabolic and/or respiratory acidosis with IV dantrolene were
awarded if these conditions “existed,” dantrolene was given, and a decrease
in end-tidal P2 or arterial P2 was reported.
fFor cardiac arrest/death cases, reviewers were MGL, BWB, GCA, GAG. For
non-cardiac arrest/death cases, reviewers were MGL, BWB, GCA.
ahttp://www.omim.org/entry/145600. Accessed September 21, 2013.
bhttp://www.emhg.org/genetics/mutations-in-ryr1. Accessed July 4, 2014.
cUniversity of Pittsburgh IRB# PRO12070525.
dhttp://www.mhreg.org/Forms/1106AMRA.pdf. Accessed September 21,
2013.

December 2014 • Volume 119 • Number 6 www.anesthesia-analgesia.org 1361
(operating room, postanesthesia care area, intensive care
area), and clinical signs of MH (masseter spasm, generalized
muscular rigidity, hypercarbia, hypertension). The dataset
also included the time intervals from anesthetic induction
to documentation of the rst sign of an adverse anesthetic
reaction, the maximal end-tidal partial pressure of CO2
(P2), maximal temperature, discontinuation of volatile
anesthetic drugs, and rst dantrolene dose. The dataset also
included the maximal temperature, maximal serum potas-
sium, maximal end-tidal P2, maximal arterial P2, lowest
pH, maximal base decit,g initial dantrolene dose (mg/kg),
need for cardiopulmonary resuscitation, and survival to
hospital discharge. Also included were results of molecular
genetic DNA analysis of RYR1 and CACNA1S when these
were available.h
Statistical Analysis
Mean and standard deviations are reported unless other-
wise noted. Wilcoxon rank sum tests were used to compare
the medians of time intervals.
The relationship between the likelihood of dying from
an MH episode and the mode of temperature monitoring
(no temperature monitoring, skin temperature monitoring
only, and core temperature monitoring [1 or more of the
following: pulmonary artery, nasopharyngeal, esophageal,
and/or tympanic]) was evaluated. The primary hypothesis
of an association between mode of monitoring versus mor-
tality after an MH episode was assessed by the Cochran-
Armitage test for proportions, with the hypothesis that the
proportional mortality would be no probe > skin probe >
core probe.7,8 The relative risk of dying with no probe versus
core probe and skin probe versus core probe was assessed
using the method of Miettinen and Nurminen.9 The relative
risk was adjusted for 2 comparisons by setting α to 0.05/2
and then adjusting for a single-sided test (decreasing risk
with increased monitoring). Because a single-sided test
was assumed for both the Cochran-Armitage test and the
calculation of the condence interval, only the lower bound
of the condence interval is reported.
We evaluated the secondary hypothesis that peak tem-
perature was correlated with the duration of anesthetic
exposure before dantrolene administration (time to dan-
trolene). The P value of the slope of the relationship between
time to dantrolene and peak temperature was evaluated
using the Fisher exact test adjusted for 3 comparisons.
We examined the ability of 6 physiologic measurements
(temperature, potassium, pH, base decit, arterial CO2, and
end-tidal CO2) to distinguish between patients who died
from MH and patients who survived an MH event. For each
physiologic measurement, we calculated the P value and
condence interval of the difference in the measurement
between patients who lived and died. The P value was used
to rank order the measurements from best (the measurement
differed in the patients who died) to worst (the measurement
was not different in patients who lived and died). The con-
dence interval was used to graphically display the data by
(1) taking the mean of both groups (patients who died and
patients who lived), (2) subtracting the mean of the 2 sub-
group means from all data points, and (3) dividing the result
by the condence interval, yielding an estimate of effect size.
We tested the relative risk of death in this cohort versus
our previous cohort, using the Fisher exact test, adjusted for
4 comparisons. Exact χ2 tests were used to compare other
characteristics of the earlier and recent cohorts.
RESULTS
Figure1 shows the process of identifying the study cohort.
One hundred eighty-nine reports were received by the
Registry between January 1, 2007 through December 31,
2012. Eighty-four (44.4%) met entry criteria and constitute
the recent cohort. Of these 84 patients, 8 died of MH. (The
data for the 84 patients may be found in Supplements 1–3,
Supplemental Digital Content 1, http://links.lww.com/
AA/B2; Supplemental Digital Content2, http://links.lww.
com/AA/B3; Supplemental Digital Content 3, http://links.
lww.com/AA/B4).
Temperature Monitoring
Figure 2 shows the relationship among MH deaths, peak
temperature, and temperature monitoring. The x-axis is the
peak temperature reported to the Registry. The y-axis is the
cumulative number of deaths. The type of monitoring probe
36 38 40 42 44
0
2
4
6
8
Maximum Temperature (°C)
Cumulative Deaths
Other Probe
Core Probe
Skin Probe
No Probe
Died
Figure 2. The relationship among peak tempera-
ture, malignant hyperthermia death, and method
of temperature monitoring. The risk of death
increases with increasing temperature.
gIf unreported, base decit was calculated using calculator found at this
website and accessed during the months of August through October 2013:
http://www-users.med.cornell.edu/~spon/picu/calc/basecalc.htm.
hReports of RYR1 screening were available from a few survivors. RYR1 clini-
cal screening results were available in 5 of those who died. These clinical
tests were limited to sequence analysis of the hotspots of RYR1 where all the
proven MH-causative mutations are located. The family of another subject
who died consented to a research protocol with screening of all 106 exons in
RYR1 and all 44 exons in CACNA1S.1

1362 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Fatal Malignant Hyperthermia: Inadequate Temperature Monitoring
is indicated by color. Patients who died are indicated by an
open circle with a “+” in the center.
All deaths occurred in patients whose peak temperature
was 38.9°C or higher (mean temperature of 41.6 ± 1.7°C).
Most patients with high temperatures had either no tem-
perature probe or only a skin temperature probe. Only 3 of
9 patients whose temperature exceeded 41°C survived. Of
the 9 patients whose temperature exceeded 41°C, only 2 had
core temperature monitoring. Both survived.
Table1 shows the mortality from an MH event as a func-
tion of temperature monitoring: 30%, 21%, and 2% for no
monitoring, skin temperature monitoring only, and core
temperature monitoring, respectively (P = 0.0012). The rela-
tive risk (lower bound) for no probe versus core probe was
13.8 (2.1). The relative risk (lower bound) for skin probe ver-
sus core probe was 9.7 (1.5).
Figure3 shows the difference in peak temperature, potas-
sium, pH, base decit, arterial CO2, and end-tidal CO2 in
patients who lived and died. The y-axis is the effect size,
calculated as described in Methods. The P value is used to rank
order the measurements but not for statistical inference. In this
dataset, temperature difference best distinguished patients
who lived from those who died. Had we used the P values for
statistical inference, the P value of 0.0022 for peak temperature
would have been signicant, even after adjusting for multiple
comparisons. End-tidal CO2 was the worst physiologic mea-
sure to distinguish patients who lived from those who died.
Figure4 shows the relationship between the time from
anesthetic induction to dantrolene administration (x-axis)
and peak temperature (y-axis). Longer anesthetic exposures
before dantrolene are associated with higher peak tempera-
tures (P = 0.00014 without correction, P = 0.00042 corrected
for 3 comparisons).
Additional Patient and Event Characteristics
Nineteen (23%) patients developed signs of MH after the
procedure was completed. Twelve were still in the operat-
ing room, and 7 (8.3%) were in either the postanesthesia
Table 1. Mortality Associated with Type of Temperature Monitoring: Relative Risk for None and Skin
Temperature Versus Core Temperature Monitoring
Probe Dead Alive Mortality Relative risk versus core Lower limit confidence intervalb
None 3 7 30%a13.8 2.1
Skin only 4 15 21%a9.7 1.5
Core 1 45 2%a1
Other 0 9 0 Excluded
The Table reports the mortality associated with each type of temperature monitoring present before the rst sign of a malignant hyperthermia (MH) event and the
relative risk for none and skin temperature monitoring versus core temperature monitoring.
aNone > skin > core, P = 0.0012 by the Cochran-Armitage test.
bCondence intervals constructed using alpha = 0.025, then *2 for single-sided test. Only the lower limit of the condence interval is given, corresponding to a
one-sided Cochran-Armitage test.
-2
-1
0
1
2
Normalized Value
Figure 3. The difference in physiologic parameters between patients who died (mean = gray line) and patients who lived (mean = black line).
Red = no probe, blue = skin probe, green = core probe, yellow = other probe. Patients who lived appear as solid circles. Patients who died
appear as an open circle with a “+” in the center. Peak Temp = peak temperature. pH is arterial pH. Base decit is arterial base decit. Peak
K = peak serum potassium; Etco2 = peak end-tidal partial pressure of carbon dioxide. The number in parentheses is the P value, used to help
identify the physiologic measure that best identies patients at risk of death.

December 2014 • Volume 119 • Number 6 www.anesthesia-analgesia.org 1363
care or intensive care units. No patient developed MH
after discharge from the postanesthesia care unit. The time
between anesthetic induction and the rst adverse sign was
longer in those who presented after a surgical procedure
with a median of 145 (103 rst quartile, 199 third quartile,
range 23–480) minutes vs 90 (27 rst quartile, 135 third
quartile, range 0–600) minutes; P = 0.02. For the 82 patients
in whom volatile anesthetics were reported, all had received
either isourane, sevourane, or desurane, and 6 received
succinylcholine.
After their MH events, 8 of 84 (9.5%) patients were dis-
covered to have a family anesthetic history consistent with
possible MH.i Because the abnormal family anesthetic his-
tory was not revealed preoperatively, 5 of these patients
(including 3 who died) received both a volatile anesthetic
drug and succinylcholine, and 3 received a volatile anes-
thetic drug without succinylcholine.
Patient and Event Characteristics for
Those Who Died
Of 84 MH patients, there were 7 cardiac arrests during the
initial MH event and 8 (9.5%) deaths before discharge from
the hospital. One 21-year-old patient was declared brain
dead on postoperative day 8. All patients who suffered a car-
diac arrest died despite resuscitation efforts. Four patients
were Caucasian, 2 were Hispanic, 1 was Native American,
and 1 was of undocumented race. Seven of the 8 who died
were anesthetized for elective low- or intermediate-risk
surgery, and 5 of 8 were healthy preoperatively.j Those who
died were 31.4 ± 16.9 (range 18–67) years old.
Diagnoses of thyroid storm and airway obstruction
were pursued before the administration of dantrolene in 2
cases. Four of 8 (50%) fatal anesthetics were administered
in free-standing facilities. All fatal anesthetics lasted longer
than 30 minutes. Time between anesthetic induction to rst
adverse sign of a fatal anesthetic reaction was 131.4 ± 60.5
(range 60–199) minutes. Seven of 8 fatal MH events were
diagnosed in the operating room, with 2 of these detected
after surgery had been completed. One event was recog-
nized in the postanesthesia care unit.
Six of 8 patients had postmortem genetic studies, with
the ndings of 3 MH–causative RYR1 mutationsk and 3
RYR1 variants of uncertain signicance. Two of these vari-
ants have been found in other well-studied MH-susceptible
subjects1 (Table2).
Differences Between Early and Recent MH
Cohort Survival
In our earlier MH cohort, there were 4 deaths in 291 events,
for a death rate of 1.4%.4 The death rate of 9.5% in this cohort
is signicantly higher than our previous cohort (relative risk
= 6.9; 95% condence interval, 1.7–28; P = 0.0043, adjusted
for 4 comparisons). There was low power to detect other dif-
ferences between cohorts (34% for cardiac arrest mortality,
0 100 200 300400 50
06
00
36
38
40
42
44
Induction to Dantrolene (min)
Peak Temperature (°C)
Other Probe
Core Probe
Skin Probe
No Probe
Died
P = 0.00042
Figure 4. The relationship between peak temperature and the time from anesthetic induction to rst dantrolene dose (induction to dantrolene).
The P value is adjusted for 3 comparisons. Three of 8 patients suffering malignant hyperthermia death are not shown because the time at which
they received dantrolene was not reported.
iExamples include patients with a family history of a possible MH death; a
mother who experienced heat stroke and was intolerant to anesthetic gases
and a brother who may have had King-Denborough syndrome; 2 paternal
great uncles who died on the operating table; a sister who had an intraop-
erative hyperthermic episode; and a paternal grandfather who had 2 cardiac
arrests during separate general anesthetics.
jUnderlying conditions in 3 of those who died included insulin-dependent
diabetes mellitus in a 22-year-old male undergoing a urologic procedure;
cardiac ablation performed for recurrent vasovagal bradycardia in a 28-year-
old male with a recent crush injury and abdominal free air for an emergent
exploratory laparotomy; and diabetes mellitus type 2, coronary artery dis-
ease, hypercholesterolemia, and hypertension in a 67-year-old male undergo-
ing repeat back surgery for single level fusion.
khttp://www.emhg.org/genetics/mutations-in-ryr1/. Accessed October 6,
2013.

1364 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Fatal Malignant Hyperthermia: Inadequate Temperature Monitoring
34% for operating facility location, 65% for muscular build,
39% for temperature probe site).
DISCUSSION
MH associated with general anesthesia continues to be
lethal. Nearly 10% of MH episodes in this most recent
cohort resulted in death, higher than the risk of death from
MH in our earlier cohort reported 6 years ago. How can MH
continue to be so lethal when there is a specic and effec-
tive antidote and when nearly all anesthesia providers are
taught how to recognize and treat MH?
These data suggest the answer is the failure to adopt
uniform temperature monitoring standards. In this study,
30% of the subjects died from an MH episode if their tem-
peratures were not monitored, 21% died if only their skin
temperature was monitored, but only 2% (1 subject) died
when core temperature was monitored. The difference was
highly statistically signicant. No temperature monitoring
at least doubles the risk of death compared to core tempera-
ture monitoring. Skin temperature monitoring in lieu of
core temperature monitoring increases the risk of mortality
by at least 50%.
Perhaps the consistent lack of monitoring is driven by
a belief that change in temperature is a late sign of MH.
We have previously shown that temperature abnormali-
ties are an early MH sign in the majority of events.5 Now,
as shown in Figure3, we demonstrate that deranged tem-
perature better identies patients who will die from an
MH episode than potassium, pH, arterial CO2, base de-
cit, or end-tidal CO2.
Relative to the early cohort, the recent MH event mortal-
ity rate has increased from 1.4% to 9.5%, although no other
signicant differences were detected between the 2 cohorts.4
Our power to detect other differences was limited by the
small numbers of patients who had suffered cardiac arrest
or death in each cohort.
The mean temperature of those who died in the recent
cohort was 41.6°C. In a 1978 study, Bynum et al.10 suggested
that temperatures of 41.6°C to 42°C for 45 minutes to an hour
constitute a critical thermal maximum in humans. Once the
critical thermal maximum is reached, the high temperatures
cause irreversible tissue damage due to apoptosis, making
it difcult for patients to recover.11 Our study is limited by
the lack of data regarding the length of time each patient
spent at his or her maximal temperature. Going forward,
the Registry will query maximal temperature duration.
American Society of Anesthesiologists monitoring stan-
dards state that “every patient receiving anesthesia shall have
temperature monitored when clinically signicant changes in
body temperature are intended, anticipated or suspected.”l
This is in contrast to the MHAUS recommendation that “all
patients undergoing general anesthetics that exceed 30 min-
utes in duration should have their temperature monitored
using an electronic temperature probe.”m The guidance on
monitoring after 30 minutes reects the difculty of inter-
preting core temperature changes during this initial anes-
thetic period.12 In the 2007–2012 cohort, the rst sign of fatal
MH events occurred >30 minutes after anesthetic induction.
The MH deaths associated with inadequate tem-
perature monitoring documented in this study should
prompt re-examination of current American Society of
Anesthesiologists and MHAUS temperature monitoring
standards. Our data demonstrate that without continu-
ous core temperature monitoring, clinicians are unable to
“anticipate or suspect” what is going to happen in time
to intervene in fatal MH events. Continuous and reli-
able monitoring of core temperature could have allowed
lStandards for basic anesthetic monitoring. Approved by the ASA House of
Delegates on October 21, 1986, and last amended on October 20, 2010, with
an effective date of July 1, 2011. www.asahq.org/For-Members/Standards-
Guidelines-and Statements.aspx. Accessed October 3, 2013.
mwww.mhaus.org. Accessed October 3, 2013.
Table 2. Selected Characteristics of Each Malignant Hyperthermia Event in Subjects with Molecular
Genetic Findings
RYR1
Known
MH causal
mutation CACNA1S CGS score
Max K+
(mEq/L)
Max T (°C)
(site)
Max Etco2
(mm Hg) Lowest pH
Max arterial
Pco2
(mm Hg)
Base deficit
(mEq/L)
Thr2206MetaYes Not done 68 na 41.7 (E) 200 6.46 200 22.9
Gly2434ArgaYes Not done 71 8.0 42.2 (U) na 6.50 120 27.0
Arg3283Glna,b,cVal875Met b,c 75 7.5 41.4 (E) 119 6.98 119 5.4
Gly341ArgaYes Not done 61 8.4 41.6 (E) 245 7.08 92 5.9
Lys487Proa,b Not done 73 6.9 45.0 (E) 80 7.10 63 9.3
Arg2508Hisa,b Not done 76 6.8 38.9 (S) 85 6.91 67 17.7
Arg614Cys Yes Not done 73 4.6 40.0 (E) 85 6.99 82 14.0
Gly2434Arg Yes
Gly341Arg Yes Not done 78 7.5 39.6 (Ax) na 7.10 48 13.4
Gly2375ArgdNot done 38 4.8 38.6 (NP) 76 na na na
The Table lists selected characteristics of each malignant hyperthermia event in subjects with molecular genetic ndings. The rst 6 rows contain data from
subjects experiencing lethal malignant hyperthermia events. In the 3 cases with survivors, 1 infant suffered a watershed central nervous system injur y. The
survivor with the compound RYR1 mutations was reported by Newmark et al.17
Ax = axillary temperature probe site at the time maximal temperature was observed; CGS = clinical grading scale score with a score of ≥50 = “almost cer tain”
malignant hyperthermia event6; E = esophageal temperature probe site at the time maximal temperature was observed; Max K+ = maximal serum potassium;
Max T = maximal temperature; Max Etco2 = maximal end-tidal partial pressure of carbon dioxide; Max arterial Pco2 = maximal ar terial partial pressure of carbon
dioxide; na = not available; NP = nasopharyngeal temperature probe site at the time maximal temperature was observed; S = skin temperature probe site at the
time maximal temperature was observed; U = unreported temperature probe site at the time maximal temperature was observed in an emergency department.
aLethal malignant hyperthermia event.
bVariant previously reported in Brandom et al.1
cNovel genetic ndings after screening 106 exons of RYR1 and 44 exons of CACNA1S.
dNovel genetic nding.

December 2014 • Volume 119 • Number 6 www.anesthesia-analgesia.org 1365
rescue before patients’ temperatures became critical.
Revision of temperature monitoring standards should
alert clinicians regarding life-threatening nonspecic
MH-presenting signs such as inappropriate tachycardia
and hypercarbia.5
In 23% of all cases in this report, the rst MH sign was
noted after the surgical procedure was complete, with 8.3%
identied in either the postanesthesia or intensive care
units. This is higher than the 1.9% observed by Litman et
al.13 in an earlier Registry cohort. The reasons for this differ-
ence are unclear. Our ndings emphasize the value of con-
tinued vigilance for MH, including accurate temperature
monitoring even after surgery has been completed.
Our study is limited by incomplete patient data, under-
reporting, and/or biased reports inherent in a fragmented
medical system and a Registry dependent on voluntary
reporting of an infrequent event. The small number of cases
reduced power. These limitations may impact our ndings.
Nevertheless, we report 8 lethal events.
When death highly suspicious for MH occurs, the anes-
thesia care team should take responsibility for ensuring that
blood or muscle is collected for postmortem RYR1 analysis.n
When an MH-causative mutation is found, familial muta-
tion analysis may serve as the rst step in identifying
surviving relatives who now require non-MH-triggering
anesthetics.14 There should be a low threshold for initiating
the MH treatment protocol in susceptible individuals even
in the absence of anesthetic drugs.15
We report 3 individuals in whom a possible famil-
ial MH history was discovered only after death. This
underscores the importance of identifying those who are
susceptible to this potentially lethal disorder and dissem-
inating this information appropriately to family members
and health care providers. Patients undergoing general
anesthetics should be encouraged to ask all family mem-
bers about adverse anesthetic events, including perioper-
ative deaths, before they meet their anesthetic providers
for evaluation.
Muscle contracture testing remains a valuable method
for documenting pathogenicity when a genetic variant of
uncertain signicance is found. Muscle contracture testing
remains the only validated method for demonstrating that
an individual does not have an increased risk of MH.16
Finally, to prevent MH deaths, clinicians should monitor
core temperatures intraoperatively whenever their patients
undergo general anesthesia for at least 30 minutes. E
DISCLOSURES
Name: Marilyn Green Larach, MD, FAAP.
Contribution: This author helped design the study, conduct the
study, analyze the data, and write the manuscript.
Attestation: Marilyn Green Larach has seen the original study
data, has reviewed the analysis of the data, has approved the
nal manuscript, and is the author responsible for archiving the
study les.
Conicts of Interest: The author has no conicts of interest to
declare.
Name: Barbara W. Brandom, MD.
Contribution: This author helped design the study, conduct the
study, analyze the data, and write the manuscript.
Attestation: Barbara W. Brandom has seen the original study
data, reviewed the analysis of the data, and approved the nal
manuscript.
Conicts of Interest: Barbara W. Brandom received research
support from the Malignant Hyperthermia Association of the
United States, which is a not-for-prot group. She is the cur-
rent director of The North American MH Registry, which is a
subsidiary of MHAUS.
Name: Gregory C. Allen, MD, FRCPC.
Contribution: This author helped design the study, conduct the
study, analyze the data, and write the manuscript.
Attestation: Gregory C. Allen has seen the original study
data, reviewed the analysis of the data, and approved the nal
manuscript.
Conicts of Interest: The author has no conicts of interest to
declare.
Name: Gerald A. Gronert, MD.
Contribution: This author helped design the study, conduct the
study, analyze the data, and write the manuscript.
Attestation: Gerald A. Gronert has seen the original study
data, reviewed the analysis of the data, and approved the nal
manuscript.
Conicts of Interest: The author has no conicts of interest to
declare.
Name: Erik B. Lehman, MS.
Contribution: This author helped design the study, conduct the
study, analyze the data, and write the manuscript.
Attestation: Erik B. Lehman has seen the original study data,
reviewed the analysis of the data, and approved the nal
manuscript.
Conicts of Interest: The author has no conicts of interest to
declare.
This manuscript was handled by: Steven L. Shafer, MD.
ACKNOWLEDGMENTS
We express our sympathy to the families whose relatives died
during MH events and our gratitude to anesthesiologists, inten-
sive care physicians, surgeons, nurse anesthetists, other health
care professionals, and MHAUS MH Hotline Consultants who
anonymously submitted AMRA report forms. This project would
not have been possible without the nancial support of MHAUS
for The North American Malignant Hyperthermia Registry
during the study period. We acknowledge the support of John
Williams, MD, Peter and Eva Safar Professor of Anesthesiology,
University of Pittsburgh Medical Center, who allowed the Safar
Fund to support The North American Malignant Hyperthermia
Registry of MHAUS so that this work could be continued. We
thank Kristee Adams, BA, The North American Malignant
Hyperthermia Registry of MHAUS administrative assistant,
and Michael Young, MS, The North American Malignant
Hyperthermia Registry of MHAUS technical database manager,
for their assistance in preparing and disseminating data. The
authors thank Steven L. Shafer, MD, Editor-in-Chief of Anesthesia
& Analgesia, for assistance with the statistical analysis.
REFERENCES
1. Brandom BW, Bina S, Wong CA, Wallace T, Visoiu M, Isackson
PJ, Vladutiu GD, Sambuughin N, Muldoon SM. Ryanodine
receptor type 1 gene variants in the malignant hyperthermia-
susceptible population of the United States. Anesth Analg
2013;116:1078–86
nwww.mhaus.org. Accessed September 25, 2013. Both addresses of clinical
laboratories where RYR1 screening is performed and those of MH Diagnostic
Muscle Biopsy Centers are listed at this website.

1366 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA
Fatal Malignant Hyperthermia: Inadequate Temperature Monitoring
2. Monnier N, Kozak-Ribbens G, Krivosic-Horber R, Nivoche Y, Qi
D, Kraev N, Loke J, Sharma P, Tegazzin V, Figarella-Branger D,
Roméro N, Mezin P, Bendahan D, Payen JF, Depret T, Maclennan
DH, Lunardi J. Correlations between genotype and pharmaco-
logical, histological, functional, and clinical phenotypes in malig-
nant hyperthermia susceptibility. Hum Mutat 2005;26:413–25
3. Li G, Warner M, Lang BH, Huang L, Sun LS. Epidemiology of
anesthesia-related mortality in the United States, 1999-2005.
Anesthesiology 2009;110:759–65
4. Larach MG, Brandom BW, Allen GC, Gronert GA, Lehman EB.
Cardiac arrests and deaths associated with malignant hyper-
thermia in North America from 1987 to 2006: a report from
The North American Malignant Hyperthermia Registry of
the Malignant Hyperthermia Association of the United States.
Anesthesiology 2008;108:603–11
5. Larach MG, Gronert GA, Allen GC, Brandom BW, Lehman EB.
Clinical presentation, treatment, and complications of malig-
nant hyperthermia in North America from 1987 to 2006. Anesth
Analg 2010;110:498–507
6. Larach MG, Localio AR, Allen GC, Denborough MA, Ellis FR,
Gronert GA, Kaplan RF, Muldoon SM, Nelson TE, Ørding H,
Rosenberg H, Waud BE, Wedel DJ. A clinical grading scale to
predict malignant hyperthermia susceptibility. Anesthesiology
1994;80:771–9
7. Cochran WG. Some methods for strengthening the common
chi-square tests. Biometrics 1954;10:417–51
8. Armitage P. Tests for linear trends in proportions and frequen-
cies. Biometrics 1955;11:375–86
9. Miettinen O, Nurminen M. Comparative analysis of two rates.
Stat Med 1985;4:213–26
10. Bynum GD, Pandolf KB, Schuette WH, Goldman RF, Lees DE,
Whang-Peng J, Atkinson ER, Bull JM. Induced hyperthermia in
sedated humans and the concept of critical thermal maximum.
Am J Physiol 1978;235:R228–36
11. Bouchama A, Knochel JP. Heat stroke. N Engl J Med
2002;346:1978–88
12. Sessler DI. Temperature monitoring and perioperative thermo-
regulation. Anesthesiology 2008;109:318–38
13. Litman RS, Flood CD, Kaplan RF, Kim YL, Tobin JR.
Postoperative malignant hyperthermia: an analysis of cases
from the North American Malignant Hyperthermia Registry.
Anesthesiology 2008;109:825–9
14. Girard T, Treves S, Voronkov E, Siegemund M, Urwyler A.
Molecular genetic testing for malignant hyperthermia suscepti-
bility. Anesthesiology 2004;100:1076–80
15. Brandom BW, Muldoon SM. Unexpected MH deaths with-
out exposure to inhalation anesthetics in pediatric patients.
Paediatr Anaesth 2013;23:851–4
16. Urwyler A, Deufel T, McCarthy T, West S; for the European
Malignant Hyperthermia Group. Guidelines for molecular
genetic detection of susceptibility to malignant hyperthermia.
Br J Anaesth 2001;86:283–7
17. Newmark JL, Voelkel M, Brandom BW, Wu J. Delayed onset of
malignant hyperthermia without creatine kinase elevation in a
geriatric, ryanodine receptor type 1 gene compound heterozy-
gous patient. Anesthesiology 2007;107:350–3