Phosphodiesterase Inhibitors

Journals Reviewed: Journal of Cardiothoracic & Vascular Anesthesia August 2000; Anaesthesia July 2000
Abstracted by: Dr VM Oerder MB.BCh (Registrar, University of the Witwatersrand)

Summary of abstract

Years down the line, phosphodiesterase inhibitors (remember how they seemed to show so much promise?) still have to find their place in our armamentarium. We look at a few recent articles.. (You may wish to briefly browse our editorial comment ).

1. Milrinone during weaning from cardiopulmonary bypass

This is a comparative study, but at no point do the authors describe the haemodynamic effects they wish to compare. On initial inspection the sample size of 48 coronary artery bypass patients appears to be sufficient, but the group is further divided into 4 - to now draw any statistical conclusions from these small groups seems rather bold! The authors standardised the premedication and anaesthetic, but none of the other variables - in particular the inotropes used to come off bypass. A patient by the end of a difficult wean may find themselves on dopamine, dobutamine, nitrates, adrenaline/noradrenaline and milrinone. In such circumstances it's difficult to assess "what's doing what".

The conclusions of the study were that :-

  • Patients with lower cardiac indices (CI) had associated lower left ventricular ejection fractions, required more pre-operative medication and had worse pre operative NYHA functional class, with a higher risk of myocardial ischaemic damage. (One could assert that this is fairly obvious)!
  • Patients with lower cardiac indices required more inotropic support, which also seems fairly self-evident;
  • Milrinone dropped systemic vascular resistance, and was associated with a significant drop in systolic blood pressure, necessitating the use of noradrenaline, further complicating the 'polypharmacy' required;
  • Patients with high baseline CI had satisfactory CI and SVR on weaning but the group who had milrinone added dropped their SVRs and supranormalised their CI, needing to have noradrenaline added.
  • The authors felt that their study supported the use of prophylactic milrinone in low CI patients but not high CI patients.

Overall I found little in the study of clinical or statistical merit, and certainly nothing that would change my present anaesthetic practice to include prophylactic PDE inhibitors.

Article 1: Haemodynamic effects of milrinone during weaning from cardiopulmonary bypass: comparison of patients with a low and high pre-bypass cardiac index
Journal of Cardiothoracic and Vascular Anesthesia August 2000 Vol. 14 No. 4
Article type: Clinical Study
Authors: Yamada T., Takeda J, Katori N. et al

2. Amrinone in weaning from cardiopulmonary bypass

This double blind prospective study contrasts nicely with the preceding study. The authors set out to determine whether prophylactic amrinone decreased the incidence of failure to weaning from CPB. (They also describe the impact of prophylactic amrinone on failure to wean at different levels of pre-operative LV ejection fraction, although the power of the study appears to have been insufficient for categorical conclusions about such sub-groups).

This well-designed study shows the merits of careful study planning, including consulting statisticians to ensure a statistically significant sample size! Two hundred and thirty four patients undergoing CPB for valvular or CABG surgery were studied. The patients' anaesthetics were standardized as were their pre-operative cardiovascular assessments. Study participants were stratified into 4 groups according to LV ejection fraction - normal (>55%), mild (45-55%), moderate (35-45%) and severe (<35%). The inotropic support to be used to come off bypass was standardized and the criteria for failure to wean were well defined.

Conclusions drawn by the investigators were:

  • There was a statistical difference between difficult weaning in the amrinone and placebo group - 7% vs. 21%
  • Fewer patients in all 4 LV function groups with amrinone failed to wean but such differences were only statistically significant in the group with normal left-ventricular function;
  • There was no difference between the amrinone and placebo group post operatively with respect to haemodynamic stability/IABP use/other inotropes/cardiac pacing/ICU or hospital stay.
  • In general use of amrinone was associated with increased use of colloids and phenylephrine.
  • The authors comment that there appears to be no relationship between failure to wean and pre-operative LV function.
  • They failed to comment on any cost benefit, but this appears unlikely from their study.

I found the article easy to read and very informative. It provided me with some ammunition not to use PDE inhibitors in our cash strapped medical world!

Article 2: Prophylactic use of amrinone for weaning from cardiopulmonary bypass
Anaesthesia July 2000 Vol. 55 No.7 627-33
Article type: Clinical Study
Authors: Lewis K.P. et al

3. Three other articles!

An editorial accompanied the first article (with two more articles in the same issue). Although the title of the editorial was Phosphodiesterase inhibitors: The inotropes of choice for the new millenium?, the editors provide surprisingly little information to justify their title. The accompanying studies are those of Laitinen et al., who looked somewhat vaguely at amrinone pharmacokinetics in neonates and infants, and of Lobato et al, who unsurprisingly showed that milrinone, known to have lusiotropic effects, improved LV compliance post-bypass. All in all, rather disappointing, including the web pointer to

PDE inhibitors - we need more hard evidence!

There are at least eleven types of phosphodiesterases (PDEs) in man, creatively labelled PDE1 to PDE11. This is far from being the whole story, as not only are there several different genes coding for some of the PDEs (for example, HSPDE4A, HSPDE4B, HSPDE4C, and HSPDE4D), but most of the PDEs recovered from different tissues look different owing to alternative splicing.

The PDEs and their locations are summarised in the following table.

Human Phosphodiesterases

PDE isozyme Where it's found Affects Inhibitors
* heart, brain, kidney, liver, skeletal muscle, smooth muscle (vascular + visceral, airway), vagus nerve cGMP > cAMP 'MIMX', vinpocetine, phenothiazines!
2 adrenal cortex, brain, corpus cavernosum, heart, liver, kidney, airway smooth muscle, platelet cGMP, cAMP 'EHNA' (erythro-9-(2-hydroxy-3-nonyl)adenine)
3 heart , corpus cavernosum, platelets, smooth muscle (vascular + visceral, bronchial ), vagus n., liver, kidney (inhibition increases renin secretion)
T-lymphocyte, B lymphocyte, basophil, mast cell, monocyte, macrophage, endothelial cell
cAMP enoximone, piroximone, olprinone; motapizone, cilostamide, cilostazol; milrinone and amrinone; pimobendan, imadazodan OPC-33540, and more..
4 kidney, lung (4A5, 4C1, 4D2, 4D3), involved in asthma, heart, skeletal muscle, smooth muscle (vascular, visceral, airway), vagus n., platelet most inflammatory cells
(T-lymphocyte, B lymphocyte, basophil, mast cell, monocyte, macrophage, endothelial cell, eosinophil, neutrophil)
cAMP rolipram (nanomolar), numerous others (piclamilast, tibenelast, CDP840, SB207499); benafentrine, zardaverine and tolafentrine inhibit both PDE3 and PDE4.
5 corpora cavernosa , platelets, skeletal muscle, smooth muscle (vascular, visceral, airway), vagus n., kidney, platelet cGMP sildenafil ; zaprinast; 'T1032'; dipyridamole
6 retina (rods and cones) cGMP E4021; PDE5 inhibitors.
7 skeletal muscle, heart, kidney, airways
T-lymphocyte, B lymphocyte, monocyte, eosinophil
cAMP PDE7B: dipyridamole !!
8 8A: testis, ovary, ileum, colon; also heart, brain, kidney, pancreas, airways, monocyte,..
8B: thyroid.
cAMP dipyridamole! , not 'IBMX'!
9 spleen, small intestine, and brain cGMP zaprinast (as for PDE5)
10 brain (putamen, caudate nucleus) cAMP; cGMP IBMX
11 skeletal muscle, prostate, kidney, liver, pituitary, salivary glands and testis cAMP; cGMP zaprinast, dipyridamole (as for other cGMP PDEs)
Nonspecific PDEIs include papaverine, theophylline, caffeine, and IBMX (isobutyl methyl xanthine). *Note that PDE 1 is dependent on calcium ions and calmodulin. Two good reviews of PDEIs are [Am J Resp Crit Care Med 1998 157 351-70], and an older overview in the BJA. [BJA 1992 68 293-302]

It would appear from their wide distribution in normal tissues that agents inhibiting PDE might be promising in a variety of therapies; conversely, because of the wide distribution of PDEs, such agents may have substantial side effects. Unfortunately, to date, 'selective' PDE inhibitors appear to have shown the latter tendency. Note how in the above studies, vasodilatation seems to have been clinically significant (and bothersome) - not for nothing have these agents been called ino dilators !

Let's look in a little more detail at just some potential uses of PDE inhibition (PDEI):

PDEI in Asthma

We know that inflammation is central to the pathogenesis of asthma, and the mainstay of asthma therapy is still inhibition of inflammation. With practically every cell involved in asthma containing PDE3 or PDE4, it's attractive to hypothesise that selective PDEIs could be the new wonder drugs for asthma. So attractive, that since 1996 there have been over 1000 patent applications for new selective PDEIs. Unfortunately, these smart new drugs have not yet lived up to expectation - the PDE3 inhibitors have unacceptable cardiovascular side effects (and relax smooth muscle but aren't good for inflammation), and the PDE4 inhibitors are either poorly bioavailable, or when they are available, often cause a lot of nausea with little good anti-inflammatory effect!

Sadly, although we now appreciate that low dose aminophylline may suppress airway inflammation in asthma, this effect is almost certainly not related to PDE inhibition. In fact, one of the goals in designing newer aminophyllin-like drugs for asthma would seem to be decreasing their PDE inhibition! A good contemporary review of PDEIs in asthma can be found at: Drugs 2000 February 59(2) 193-212.

PDEI in Cystic Fibrosis!

There has been speculation (but only this) that PDE3 inhibitors such as milrinone may be useful in enhancing chloride channel activity in cystic fibrosis. Such speculation seems far removed from clinical practice.

PDEI in Heart Failure

The above articles are symptomatic of the widespread confusion about whether we should use PDE3 inhibitors to treat myocardial dysfunction. The FDA has approved the short-term (48 hour) use of milrinone for heart failure, as here it appears to work. However, withdrawal of milrinone may simply return the patient to a decompensated state. Others have even used long-term infusions to keep the patient going while a transplantable heart is found, but the reports are mainly anecdotal. An earlier, well-designed, long term study (the PROMISE study) certainly showed increased mortality [N Engl J Med 1991 Nov 21;325(21):1468-75]. Much controversy (but few facts) seem to have been generated by those who advocate intermittent short-term infusion of milrinone.

PDEI in renal disease

Despite the widespread presence of several PDEs in the kidney, little potential for use in renal disease seems to have been realised so far. PDE3 and 4 may play a role in some rat models of glomerulonephritis, but little human relevance is apparent. PDE5 activity seems increased in sodium-retaining states (thus antagonising the effect of atrial natriuretic peptide - 'atriopeptin'), but the clinical role of antagonists doesn't seem to have been explored.

PDEI in erectile dysfunction

Now we're talking! Everyone, simply everyone, is aware of the (ahem) potency of sildenafil in erectile dysfunction. The drug works - even the FDA has approved it. (We know that nitrergic nerves innervating the corpora cavernosa mediate erection through increased cGMP levels, and cGMP removal is inhibited by sildenafil). Sildenafil is remarkably selective in its PDE5 inhibition: IC50 of 3.5 as compared to values of 280 - 68 000 for PDEs 1 to 4, and 34-38 for PDE6, explaining the ~3% of users who have trouble with blue-green colour discrimination.

As an aside, don't forget that sildenafil was initially developed as an anti-anginal agent, and that PGE5 receptors are present on vascular endothelia. Unsurprisingly, hypotension can occur as a side effect, and this is exacerbated by concomitant use of antihypertensives, especially calcium channel blockers. Also remember its metabolism by CYP-3A4, inhibited by a vast number of other agents including macrolides, ketoconazole, cimetidine, ritonavir, resulting in profound hypotension as a drug interaction. And Viagra sales are drooping.

Even the few selective PDEIs that actually work have their problems!


Old scopes logo