Journal of Cardiac Failure Vol. 12 No. 4 2006

Perspective

Redefining the Therapeutic Objective in DecompensatedHeart Failure: Hemoconcentration as a Surrogate

for Plasma Refill Rate

ANDREW BOYLE, MD,1 AND PAUL A. SOBOTKA, MD2

Minneapolis, Minnesota

ABSTRACT

Background: Acute decompensated heart failure is a growing epidemiologic problem about which littleconsensus exists on guidelines and recommendations for therapy.Methods and Results: Available databases suggest that a large percentage of patients are being inade-quately decongested while hospitalized, resulting in poor clinical outcomes. This is partly from a lackof an appropriate target to define therapeutic success. The demonstration of a prerenal state by bloodwork does not indicate adequate decongestion but rather means that the rate of fluid removal has exceededthe plasma refill rate. Hemoconcentration, as evidenced by a rising hematocrit is an appropriate surrogateto indicate that the plasma refill rate has been exceeded by the rate of fluid removal. This surrogate ofplasma refill rate can be easily and continuously measured by using an in-line hematocrit sensor duringultrafiltration therapy.Conclusion: We propose that the therapeutic objective in acute decompensated heart failure should beredefined and that the rate of volume extraction should be adjusted to approximate the plasma refillrate and that complete decongestion will have occurred only once hemoconcentration is observed at min-imal rates of volume extraction.Key Words: Diuresis, Prerenal azotemia, Hematocrit sensor, Ultrafiltration.

The determinant of success of a potential therapy fora patient admitted with acute decompensated congestiveheart failure (adCHF) should be the adequacy of removalof the excess volume and restoration of the ideal bodyweight, the so-called dry weight. The benefit is inherentlylinked to the success of the intervention of attaining thatgoal without adversely affecting the neurohumoral system.Unfortunately, defining a patient’s dry weight is enigmaticas is defining the safe rate of volume reduction whichwill not result in intravascular depletion and subsequentneurohumoral activation. We propose that the optimalgauge for the titration of decongestion therapy for adCHFpatients is hemoconcentration, as a surrogate for plasma

From the 1Division of Cardiology, University of Minnesota, Minnepolis;and 2CHF Solutions Inc., Minneapolis, Minnesota.

Manuscript received November 1, 2005; revised manuscript receivedDecember 9, 2005; revised manuscript accepted January 17, 2006.

Reprint requests: Andrew Boyle, MD, Division of Cardiology, Universityof Minnesota, MMC 508, 420 SE Delaware St, Minneapolis, MN 55455.

1071-9164/$ - see front matter

� 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.cardfail.2006.01.011

247

refill rate (PRR), which is impossible to measure directly,because it satisfies both objectives by simultaneously defin-ing both the dry weight and a safe rate of volume reduction.

It is concerning, given the overwhelming agreement onthe chronic management of heart failure, that little consen-sus exists on therapeutic guidelines and recommendationsfor acute adCHF despite the fact that heart failure remainsthe leading cause of hospitalization in the United States,now for the 13th consecutive year, and the single largest ex-pense for Medicare. The absence of evidence-based recom-mendations is conspicuous.

Clinical Prognosis Linked to the Adequacyof Decongestion

In the patient with adCHF, clinical prognosis is directlyrelated to the adequacy of decongestion as represented bya reduction in the pulmonary capillary wedge pressure.1

Yet, 49% of patients hospitalized for adCHF either loseless than 5 pounds or gain weight while hospitalized, sug-gesting they are being inadequately decongested.2 If the

248 Journal of Cardiac Failure Vol. 12 No. 4 May 2006

success of decongestion therapy is dependent on reachinga desired dry weight, the failure to substantially reduceweight in patients hospitalized for adCHF might logicallyexplain their reportedly high readmission rates.3 Superfi-cially, the observation that clinical prognosis is linked toadequate decongestion suggests that as long as the thera-peutic goal remains complete decongestion of the patient,the means by which this objective is achieved is irrelevant.It is possible that strategies for decongestion might belinked with adverse events independently of their abilityto attain a dry weight. In current experimental designs,both the effect of a given treatment and the ability toachieve a desired therapeutic target are simultaneously as-sessed without a mechanism to differentiate the two. Di-vorcing the treatment effect from the effect of reachinga therapeutic target is critical in identifying the importanceof attaining specific therapeutic goals as compared with theefficacy of the tools we choose in attaining them.

Defining the Plasma Refill Rate

In a patient with adCHF, the primary objective is to re-move the excessive interstitial fluid in addition to the exces-sive intravascular fluid. By reducing the intravascularvolume through diuresis, capillary hydrostatic pressure de-clines to the point where interstitial pressure plus serum on-cotic pressure exceeds the luminal hydrostatic pressure andfluid is passively reabsorbed from the interstitium into theintravascular space. The rate at which this occurs is termedthe PRR and is proportional to the transcapillary pressuregradient and the permeability of the capillary membrane.Before the initiation of decongestion therapy, the PRR is0. When the equilibrium is perturbed, such as by initiationof diuresis or ultrafiltration, the resulting modification ofhydrostatic and oncotic pressures will result in a PRRgreater than 0. It is intuitive to expect that the PRR afterthis perturbation might not be identical in all patients giventhe considerable variability in body size, capillary mem-brane permeability, effective lymphatic flow, regional bloodflow, and serum protein concentration. The PRR would alsobe expected to decline throughout decongestion therapy asinterstitial edema is reabsorbed and removed and the trans-capillary pressure gradient continually declines.

Diuretics and the Plasma Refill Rate

Intravenous diuretics are the most commonly used thera-peutic agents to decongest patients with adCHF either asboluses or as a continuous infusion.2 The diuretic dose istitrated to achieve a ‘‘suitable’’ rate of diuresis until symp-tom resolution is achieved. Symptom resolution has notbeen shown to be particularly helpful in determining the ad-equacy of decongestion nor has it been shown to correlatewith improved clinical outcomes. Diuresis is often termi-nated and dry weight is considered achieved once a contrac-tion alkalosis or a rising serum blood urea nitrogen (BUN)and creatinine is observed because it is assumed that

a further reduction of intravascular volume will further im-pair renal perfusion resulting in activation of the renin-an-giotensin-aldosterone (RAAS) and sympathetic nervous(SNS) systems as well as triggering the cardio-renal syn-drome. These findings suggest a prerenal state has beenachieved with inadequate intravascular volume for renalperfusion. Patients are then declared euvolemic despitethe fact that their dyspnea, edema, or ascites may persist.The fundamental question to be addressed is whether ornot a rising serum BUN and creatinine and a contraction al-kalosis are truly representative of a euvolemic state.

In a patient with adCHF undergoing volume reductionvia diuretics, volume is removed from the intravascularspace as it is filtered by the nephron and subsequently notreabsorbed because of the effect of the diuretic. Concur-rently, fluid is then passively reabsorbed from the interstiti-um into the intravascular space at a rate determined by thePRR. In clinical practice, the diuresis rate is never titratedto match the PRR and, therefore, may randomly either ex-ceed the PRR or remain below the potential maximal PRRthat could be achieved. If the intravascular volume declinesduring diuresis at a rate greater than the PRR, renal perfu-sion must also decline that will be detected as a contractionalkalosis or a rising serum BUN and creatinine. The risk ofdeveloping acute renal failure by diuresing at a rate that ex-ceeds the PRR is not insignificant. Several studies havedemonstrated a worse clinical prognosis in adCHF patientswho develop renal failure while being diuresed.4,5 How-ever, the patient may not necessarily have reached his orher dry weight at that time. Rather, we propose, the correctresponse is to decrease the rate of diuresis so it does not ex-ceed the PRR and to continue diuresis. Because the PRR iscontinuously changing as the transcapillary pressure gradi-ent changes, the rate of diuresis also needs to continuouslychange to prevent a prerenal state and maximize volume re-moval to reach the patient’s true dry weight. It would seemrational then to remove excessive volume at a rate equiva-lent to the maximal instantaneous PRR so as to shorten theduration of diuresis and therefore, shorten the length ofstay, but not to exceed the PRR in order to avoid developingprerenal azotemia.

Hemoconcentration as a Surrogate for PRR

Therefore, a mechanism to continuously measure thePRR is necessary to optimize the management of adCHFwithout activation of the RAAS and SNS. Unfortunately,the PRR is impossible to measure directly and thereforea surrogate must be used to estimate these compartmentalvolume shifts. Dynamic changes in PRR can be indirectlymonitored with the concentration of agents that remain con-fined to the intravascular space such as red blood cells.Thus the hematocrit is diluted when vascular volume is en-gorged and further hemoconcentration occurs only underconditions when diuresis has exceeded the potential of in-terstitial reservoirs to restore intravascular volume becausethe therapy is progressing too rapidly. We hypothesize that

Redefining the Therapeutic Objective in Decompensated HF � Boyle and Sobotka 249

adjusting fluid removal rates to maintain a targeted level ofhematocrit marries the decongestion therapy rate to PRR.Because PRR varies throughout decongestion therapy, therate of volume extraction must also vary correspondinglyto maintain a targeted hematocrit. Hemoconcentrationmarks that diuresis has caused intravascular volume reduc-tion with the risk of end-organ hypoperfusion or neurohu-moral activation.

At the University of Minnesota, we have been experi-menting with an inline continuous hematocrit sensor inconjunction with ultrafiltration (UF) therapy as a surrogatemeasure for the PRR. Consequently, hemoconcentration, asevidenced by an increasing hematocrit, indicates that theUF rate has exceeded the PRR and the UF rate should bedecreased until it matches the PRR, as evidenced by a stabi-lized hematocrit. With the ability to continuously monitorthe hematocrit, it is imaginable that the UF rate could becontinuously adjusted so that it would never exceed thePRR and the patient would never, even temporarily, experi-ence a decline in renal perfusion. Ultrafiltration could con-tinue, albeit at a variable rate, without risking adverseneurohumoral or renal consequences until hemoconcentra-tion was observed at a minimal rate of volume extraction,indicating the patient’s dry weight had been reached.

Nephrologists exploit this concept by continuously mea-suring the hematocrit during hemodialysis and defininga patient’s ‘‘crash crit’’ or the hematocrit value at which hy-potension is precipitated. The ‘‘crash crit’’ is never ex-ceeded during hemodialysis because of the risk ofcausing systemic hypotension and inadequate end-organperfusion. We hypothesize that less aggressive hemocon-centration than the ‘‘crash crit’’ targeted during hemodialy-sis could be used to prevent the reduction of intravascularvolume in heart failure patients during decongestion ther-apy and might protect against activation of the neurohu-moral system or worsening renal function.

Universality of Concept

The hypothesis of titrating decongestion therapy to re-duce interstitial edema without embarrassing intravascularvolume is applicable to all forms of decongestion therapy.Using hemoconcentration as a surrogate for PRR is obvi-ously more amenable to therapies in which there is contin-uous access to blood such as ultrafiltration. However, if thehypothesis is proven, then it becomes incumbent on inves-tigators to find adequate similar surrogates for PRR whichare adaptable for other forms of decongestion therapy. Weare calling for a universal target of treatment for deconges-tion therapy that would be independent of the therapeuticdecongestion modality.

By offering a testable hypothesis that the PRR defines theupper limit of safe decongestion therapy, a host of clinicallyvaluable experiments are possible. These include: (1) Isthere an upper limit of volume removal which potentiallyactivates the neurohumoral system? (2) Is there a clinicalcost to prolonging decongestion therapy by removing fluid

too slowly? and (3) Are there differing advantages and dis-advantages of each therapeutic treatment strategy whencompared with a common volume goal. Decongestion ther-apy is plagued by questions such as how much volume canbe safely extracted and how quickly, what is the clinicalcost of delaying effective treatment, and what are the inher-ent adverse events associated with the device or pharmaco-logic strategy employed. The hospital economicenvironment makes the question of how rapidly can we re-duce volume even more prescient. If economic pressurescause clinicians to remove fluid too rapidly or excessively,we may simply exchange clinical congestion for advancingrenal disorders or untoward activation of the neurohumoralsystem.

Conclusion

In conclusion, the therapeutic objective for patients ad-mitted with adCHF needs to be redefined. Prompt therapy,at a rate not exceeding the PRR, should preserve renal per-fusion while optimizing the rate of removal of interstitialsalt and water. Ultimately, the total volume removed shouldbe governed by the limits of recovery of excess volumefrom both the intravascular and interstitial spaces, ratherthan titrated against activation of the RAAS and SNS oran adverse clinical event, such as a rising serum BUN orcreatinine. The hematocrit is a reasonable surrogate forthe PRR and a logical target for decongestion therapy be-cause hemoconcentration observed at minimal rates of vol-ume extraction should indicate that euvolemia has beenachieved. As a consequence of selecting hemoconcentrationas the target of decongestion therapy, a treatment modalitymust have the ability to continuously measure the hemato-crit to prescribe a rate of fluid removal which can approx-imate the PRR and safely achieve complete decongestionof the patient and thereby improve their clinical prognosisand reduce their readmission rate.

References

1. Fonarow GC, Stevenson LW, Steimle AE, Hamilton MA,

Moriguchi JD, Walden JA, et al. Persistently high left ventricular filling

pressures predict mortality despite angiotensin converting enzyme

inhibition in advanced heart failure. Circulation 1994;90:I-488.

2. Adhere. Acute Decompensated Heart Failure National Registry. Avail-

able from: www.adhereregistry.com. Accessed February 9, 2006.

3. Stewart S, Horowitz JD. Home-based intervention in congestive heart

failure: long-term implications on readmission and survival. Circulation

2002;105:2861–6.

4. Butler J, Forman DE, Abraham WT, Gottlieb SS, Loh E, Massie BM,

et al. Relationship between heart failure treatment and development

of worsening renal function among hospitalized patients. Am Heart J

2004;147:331–8.

5. Hillege HL, Girbes ARJ, De Kam PJ, Boomsma F, de Zeeuw D,

Charlesworth A, et al. Renal function, neurohormonal activation, and

survival in patients with chronic heart failure. Circulation 2000;102:

203–10.