Jimmy Klemis, MD
CT Surgery Conference





Potentially lethal disease with varying etiologic agents and different clinical situations (NVE vs PVE, etc)
No “cookbook” approach to proper therapy, esp when considering surgery
In select patients, combined medical and surgical Rx offers substantial benefit compared with medical Rx alone
However, surgery carries risk and decision on whether or not to operate must be carefully thought out with good communication between surgical and medical teams





In pre-Abx era, largely fatal disease
1885 – Sir William Osler in Gulstonian lectures referred to IE as the “malignant endocarditis”, 30 years later he expressed pessimism about ever finding a
“cure” for IE
1940’s – PCN revived hope for a cure of IE, however morbidity and mortality only partially altered resistant organisms and shifting etiology (IVDA)
Chamoun. Am J Med Sci. Oct 2000; 320 (4)


1961 – Kay et al first to report surgical cure of pt with medically resistant IE (fungal TV)

1965 Wallace, et al – first report of successful valve replacement in active endocarditis
 early success in many studies of selected patients led to “paradigm shift” in management of complicated endocarditis


Hemodynamic compromise/ Heart failure

Persistent sepsis

Peripheral embolization

Extravalvular extension of infxn


Mills, et al. UCSF 1974 1




79/144 pt developed CHF within 6mos of admit
60% moderate-severe
MR – 50% developed CHF, 1/2 severe
AR – 80% CHF, 2/3 severe

6 month survival with severe CHF/AR
 medical 7 % med/surgical 64%
1 Mills J, et al. Chest 66:151-157, 1974





Lewis, et al. Johannesburg, South Africa, 1975-80 1 early valve replacement in 95 hemodynamically unstable pt – 64% emergent 88% 48hrs
Mortality


 urgent surgery 15% (13/84) elective 18% (2/11)
5 year survival 60%
Periprosthetic leaks in 13% (10/80) of survivors
1 Lewis BS, et al. J Thorac Cardiovasc Surg 84:579-84, 1982


Johannesburg, SA 1982-1988 1

203pt with active IE and early valve replacement

Urgent surgery (<48hrs) in 53%

Mortality

Urgent 7%

Overall 4%
 long term 6% pt followed 38± 22mos
1 Middlemost S, et al. JACC 18:663-667, 1991

Medical
Mortality
Med/Surgical
No CHF 15% 11%
CHF
Moon, et al. Prog Cardiovasc Dis. 1997
60% 29%





 nonsterile Bld Cx 3-5d after dx lack of improvement sxs after 1wk appropriate Abx usually due to

Bacterial resistance

 valvular/perivalvular infections non cardiac septic foci (splenic, renal, cerebral abcess, mycotic aneurysm

GNR, staph or fungal infxn surgery may eliminate septic focus, but not necessarily improve pt hemodynamic condition unless significant valvular regurg
+Bld Cx at surgery predict adverse outcome


Postive Cx @ time of surgery predicts poorer outcome

D`Agostino, et al Ann Thor Surg 1985

108pt with NVE

87pt Bld Cx (-) >90% 1 year complication free survival (no perivalvular leak, IE recurrence)

19 pt Bld Cx (+) <70%

 although ↑ complication if Bld Cx +, still important to intervene esp in face of further destruction of valvular/annular tissue

Boyd, et al. NYU 1977 1
 operative mortality risk in uncontrolled infxn better when operated earlier (within 10d of admit) (17%) than when abx continued for 4-6wks (90%)
1 Boyd et al. J Thorac Cardiovasc Surg 73:23-30, 1977

Mortality Risk
Recurrent IE after successful medical Rx
PVE after valve replacement in active IE
Alsip et al, Am J Med 78:138-148, 1985
10%
10%
10%
Approaches 50%

 may also be from extracardiac source/emboli
 splenic, renal, cerebral abcesses

? proper Rx – surgery?, incidence of recurrent endocarditis in these situations?

Image: Roberts, Cornell Univ Web Site:Vascular infections


S. aureus
 highly destructive
 meta-analysis showed higher mortality with medical
(39/76 56% ) compared with med/surgical Rx
(24/77 31% ) p<.03
 not absolute indication but more aggressive surgical approach should be considered, esp if other factors

Gram (-)/serratia/pseudomonas


Fungal

 most common : Aspergillus, Candida, Torulopsis glabrata risk: prev cardiac surgery, Abx use and hyperalimentation, long therm IV cath, IVDA


 clinical: neg Bld Cx/fever, changing murmur, chorioretinitis, and large peripheral emboli overall survival with medical Rx 25% c/w med/surgical rx 58% compelling if not absolute indication for surgery
Rubenstein and Lang. Fungal Endocarditis. Eur Heart J 1995



 embolic events common 30-40% of IE


 brain>limbs, coronary, spleen, kidney directly responsible for ~25% of fatalities 1 recurrence rate 54% within 30d incidence falls after initiation of Abx therapy ~ 2wks risk


 size > 10mm (47% vs 19%) 2 staph, candida, GNR mobile, pedunculated, mitral>aortic
1 Acar, et al. Eur Heart J, 16 (supplement B), 94-98. 1995
2 Mugge et al. JACC 14:631-638. 1989

Moon, et al. Prog Cardiovasc Dis 1997

Vegetation on atrial surface of PMVL



Rohmann, et al 1

64% vegetations resolved/decreased




36% no change/increased valve replacement 2% vs 45% perivalvular abcess 2%vs 13% mortality 0% vs 10%
Vuille, et al 2

 persistent veg in 50% despite clinical healing, no independent association with late complications in the absence of valvular dysfxn, persistent vegetation on echo shouldn’t be criterion for valve replacement in absence of other indications
1 Rohmann, et al. J Am Soc Echo 4:465-474, 1991
2 Vuille, et al. Am Heart J 128: 1200-1209. 1994

 recurrent emboli are relative indication for surgery (class IIa) but should not be considered absolute indication


 surgical intervention with cardiopulm bypass can cause extension of infarct or hemorrhagic transformation of previously bland infarct
Eishi et al – cerebral emboli + surgery
24hrs 2wks 4wks
Extension or expansion of infarct
Mortality
50%
67%
<10%
<20%
2%
<10%
Eishi, et al. J Thorac Cardiovasc Surg 110:1745-1755, 1995

Fig. 1. Computed tomographic scans of a patient with right middle cerebral artery infarction resulting from infective endocarditis. This patient underwent a Bentall-type operation for graft infection on the same day, resulting in massive brain swelling, and died 3 days later.
Top row, Preoperative computed tomographic scans; bottom row, postoperative scans.
Eishi,et al. J Thorac Cardiovasc Surg 1995;110:1745-55



Ting, et al – smaller, bland cerebral infarcts 31pt 1
 operative mortality 19%
 survivors (81%)

5pt with cerebral hemorrhage  CVA
 others:


12% exacerbated CNS sxs
16% unchanged


20% partial resolution
52% complete resolution
Other studies have shown complete neurologic recovery in pt with coma or dense hemiparesis after valve replacement, but recommended delay if bleed 2
1 Ting, et al. Ann Thorac Surg 51:18-22, 1991
2 Zisbrod, et al. Circulation 76:V109-V112, 1987 (suppl V)

Ruptured mycotic aneurysm in MCA territory (causative agent: Aspergillus)



 single cerebral embolus not indication for surgery unless assoc with large mobile veg and that further
CNS injury might preclude meaningful chance at recovery/rehabilitation bland infarct – if stable hemodynamics, 2-3 wks Abx before considering surgery to minimize provoking further CNS injury hemorrhagic infarct – surgery postponed as long as possible – optimally if full course Abx can be given and recovery of neurologic dysfxn

 annular abscess


 operative mortality 19-43% (vs >75% medically treated) 1 extensive tissue necrosis/structural damage including interventricular septum, conduction system, and fibrous skeleton of heart
In NVE mitral (1-5%) < aortic (25-50%)

 clinically have more valvular regurgitation hi risk (staph/fungal, new heart block, PVE) should undergo TEE (90% detection vs 50% TTE)
1 Moon, et al. Prog Cardiovasc Dis 1997 Nov-Dec 40(3) p246







ECHO findings in Annular abscess anterior or posterior Ao root wall thickness≥
10mm perivalvular density in IVS ≥ 14mm sinus of valsalva defect/aneurysm rocking of prosthetic valve
Sens and Spec 85% if 1 of above seen

Cormier et al. Eur Heart J 1995 (16) suppl B 68-71

Otto. Textbook of Clinical Echocardiography 2 nd Ed. Chp 13
TTE (L) and TEE (R) showing evidence of AV vegetation and paravalvular abscess

Dec 2001 ECHO case of the month, www.acc.org


Conduction disturbances in 30% with abscess vs
<2% if no abscess

1 st degree > 7d, new 2 nd or 3 rd degree block requires eval for abcess - TEE

Moon, et al. Prog Cardiovasc Dis. 1997

Moon, et al. Prog Cardiovasc Dis 1997

Predictors of operative mortality
Moon, et al. Prog Cardiovasc Dis 1997



Combined medical/surgical rx of selected populations offers substantial morbidity and mortality benefit.
careful attention to hemodynamic status, infecting organism (staph aureus, fungi, GNR), valve(s) involved
(AV), clinical manifestations (emboli, abscess, conduction abnl, CHF), and findings on imaging
(TTE/TEE, etc) allow a tailored approach to proper Rx in each patient to minimize morbidity and mortality