Injection Therapy for Erectile Dysfunction: Escalating Past Alprostadil Alone
A single patient with vasculogenic erectile dysfunction who has already failed maximum-dose oral therapy. The disagreement is whether to step through the one FDA-approved injectable first or go straight to the compounded combination his diabetes makes more likely to actually work.
Felix N., a 58-year-old HVAC technician, has lived with type 2 diabetes for fourteen years, long enough that peripheral neuropathy has already set in — and long enough, it turns out, that his erectile dysfunction has outgrown the two oral agents most men start with. He tried sildenafil, then tadalafil, both titrated to their maximum doses, with inadequate response either time. That failure pattern is itself informative: diabetic vasculogenic erectile dysfunction, layered on more than a decade of an A1c running around 8.2 percent, is exactly the population where oral PDE5 inhibitor response rates run lowest — the question the Sildenafil Diabetes Study Group trial (Rendell and colleagues, 1999) was built to answer in diabetic men specifically — since the vascular and neural damage driving his dysfunction sits downstream of where an oral agent works.
Intracavernosal injection is the next step, and here the choice is less about drug class than about how much drug and how it’s compounded. Alprostadil alone is the only agent in this space with FDA approval, dosed and titrated in-office to minimize the risk of a prolonged erection. Papaverine and phentolamine in combination — bimix — or all three drugs together — trimix — are compounded regimens no regulator has formally evaluated, valued in real-world practice precisely because men like Felix, with vasculogenic disease that blunts monotherapy response, often need the added potency a combination provides. That added potency is also what raises the priapism risk, since a regimen effective enough to work in a patient whose own vascular resistance has already reduced his response threshold is, by the same logic, more likely to produce an erection that doesn’t resolve on its own. His neuropathy adds a further wrinkle the group has to weigh honestly: reduced penile sensation can delay his own recognition of a problem developing, which argues for conservative in-office titration and explicit counseling regardless of which regimen the group ultimately chooses. It also means the usual patient self-report that guides in-office dose-finding — does this feel like enough, does it feel like too much — carries less information for Felix than it would for a man with intact sensation, another reason to titrate more conservatively than his vasculogenic profile alone might otherwise suggest.
In the andrology clinic, planning injection therapy
Start with alprostadil monotherapy, in-office titration, standard first step. It's the only FDA-approved option in this class, and stepping through it first gives us a clean, well-documented escalation pathway if we do end up needing a compounded combination.
I'd skip that step for him specifically. Diabetic vasculogenic erectile dysfunction with this degree of neuropathy predicts a lower monotherapy response rate — that’s the population Rendell’s diabetes trial was built around — we already know his oral agents underperformed for the same vascular reasons. Sending him home on an agent likely to underwhelm again just costs him another visit and another disappointment before we get to what probably works.
Calling it 'the clean first step' assumes it's actually likely to work for him — in a vasculogenic pattern this established, that's not a safe assumption to build a stepwise plan around.
I'd land between you. Trimix carries real priapism risk precisely because it's the most potent of the compounded options, and compounding quality varies by pharmacy in ways no FDA process oversees. Bimix captures most of the combination benefit — papaverine and phentolamine working through separate mechanisms than alprostadil alone — without committing to the most complex three-drug mixture as a starting point.
Start with careful in-office dose titration of bimix, with explicit priapism-recognition counseling before he self-injects at home — any erection past four hours is an emergency. If bimix underwhelms at a reasonable dose, trimix is the next step, not the first one.
Agreed: start with careful in-office dose titration of bimix, given his diabetic vasculogenic profile and prior PDE5 inhibitor failure, with explicit at-home priapism-recognition counseling before any self-injection; escalate to trimix only if bimix underperforms at a reasonable dose.
Not agreed: whether a single formal alprostadil trial should still have happened first purely to document a clean escalation pathway for insurance purposes. The urologist wanted that step regardless of predicted response; the endocrinologist saw it as a visit spent confirming what his oral-agent history had already shown.