Before closing up I washed the cavities of the heart carefully with saline solution. Children’s hearts start up very quickly and we didn’t want any infected debris in Sophie’s brain. We would soon have echo pictures of the repair from the probe in her oesophagus. There were the usual microbubbles of air whizzing around inside the left ventricle, like a snow storm on the screen. I made a small puncture hole in the highest point of the aorta and they all fizzed out into the atmosphere, where they belonged. As the heart flipped into regular rhythm we could see that the valve was competent, with just a trace of a leak – nothing more.
The time had come to think about separating from the bypass machine. This presented another challenge, but the heart was already looking better, with languid, coordinated contractions. All four valves opened and closed nicely on the echo, so I asked that it be slowly slid off from bypass. The little heart was soon on its own again, pumping away with pressure up to 100 mm Hg, and the repair was holding.
As if orchestrated to the minute, Nick Archer’s balding head suddenly appeared around the door, so I asked my bespectacled registrar to close up the chest and went over to speak with him. Archer wondered why the operation had taken longer than usual and was concerned for the parents. The more intelligent the waiting relatives, the greater their insight and the higher their anxiety levels. But he was particularly pleased by the word ‘repair’. Although we even had some babies with artificial mitral valves, striking a balance between inadequate and excessive anticoagulation was difficult, boiling down essentially to stroke versus bleeding. Anticoagulation was also a concern for young women who wanted babies, since warfarin can cause foetal abnormalities or bleeding into the placenta. It was difficult stuff to deal with, but now Sophie wouldn’t need it. Archer took me off to meet the divorced parents, with their respective new partners. They were all huddled together for moral support, each fearing the worst.
I have no affinity with triumphalism and I never wanted to play God. But parents do crave reassurance from the man that did the job. Indeed I had once been on the other side of the fence when a member of my family had heart surgery and I ached to be told that everything was fine. So that was precisely what I told them. Yet in the back of my mind I wondered whether an abscess remained in Sophie’s heart muscle, as I’d seen an unusual bulge beneath the repair in the wall of the left ventricle. I had cleaned out some pus and expected it to heal, but this staphylococcus was so bloody aggressive.
By the time I returned to the operating theatre my registrar had closed the chest and Sophie was being wheeled out. I still needed to draw the steps of the operation in her notes so that others could understand what had been done. Then, with the poor girl settled in intensive care, I went home the hero, while the complex little family took it in turns to sit by Sophie’s bedside through the night, loving her towards recovery. They had hope again. Fear for the time being was dispelled.
I was back with her by 6.30 the following morning. Because we had no other children’s heart surgeon in those days, no one else to take responsibility, there was no on call or off call, no day or night. It was just me if something happened. Sophie had been stable overnight and her mum was with her, grasping her hand, not about to let her go. But her temperature was high again. We had stirred up the hornets’ nest and staphylococcus was angry. Billions of them. She still needed powerful vasopressor drugs to maintain an adequate blood pressure, and her kidneys had stopped producing urine.
During the morning Archer did another echo. The left ventricle was working well and the mitral valve looked good. There was the usual accumulation of blood and blood clot around the heart – there always was at this stage – but Sophie remained stable for the next forty-eight hours and was taken off the ventilator. Next day she went to the adolescent ward known as Melanie’s, back to a single room. She still had that swinging temperature, and we blamed it on complement activation, my great discovery in Alabama.
4 March, 11.35 am. Crash call to Melanie’s Ward. One week after her emergency surgery, and fortunately with her mother Fiona in the room, Sophie suddenly collapsed without warning. She was lying prostrate on the floor, pulseless and not breathing, when the paediatric resuscitation team burst in and started cardiac massage. The anaesthetist hand-pumped the black gas bag, inflating her lungs with oxygen. An intravenous injection of adrenaline restored a modest blood pressure, just long enough for a quick echocardiogram. This showed her pericardium to be filled with blood. The little heart was squeezed and unable to fill. Worse still, there seemed to be a hole in the heart wall where there should be muscle.
An abscess cavity had ruptured directly below the repaired valve, so my worst fear had indeed materialised. The cardiology registrar attempted to aspirate blood with a needle but it clotted off rapidly. Then she arrested again, needing more cardiac massage. The lady registrar called me in my office. I just told them to get Sophie round to theatre immediately, because if I opened her on the ward she’d die. I knew what to expect and needed to rush her back on the bypass machine. I could readily visualise the problem, and for once I was concerned that I couldn’t fix it. I had that gut-wrenching feeling that she wouldn’t even make it to the operating table. I hoped she would – but I didn’t expect it.
With repeated boluses of adrenaline and intermittent cardiac massage, Sophie did reach theatre, and her arrival was luckily well timed between my colleague’s planned operations, so God must have been sitting on her shoulder. She came directly through to the operating table, where I was scrubbed and waiting, and we hurriedly sliced through the skin stitches and snipped the wires holding the sternum together, then I unceremoniously ripped them out and ratchetted open the retractor. The heart was encased in blood clot, a purple gelatinous mass just like fresh liver that all had to be scooped out by hand so the heart could fill and eject blood again. It was beating like the clappers in response to the drugs, but then fresh blood suddenly gushed out from behind, filling the pericardium. I needed to see where it was coming from. I suspected that the abscess below the mitral valve had eroded through the wall of the ventricle, and that the muscle itself was now mush and wouldn’t hold stitches. The nightmare scenario.
The anaesthetist’s echo probe in her oesophagus confirmed my worst fears. There was a ragged abscess cavity that had ruptured through the ventricular wall beneath the important circumflex coronary artery. I had never seen or heard of this before in hundreds of cases of bacterial endocarditis, nor had I even read about it in the surgical journals. So this was going to have to be a ‘make it up as I go along’ job. One thing was certain, however. Had I implanted a rigid artificial valve first time around, that whole fragile junction between the left atrium and left ventricle would have fallen apart and been irretrievable. At least the valve repair remained intact. It was clear that I should attempt to close the abscess from outside the heart, because if I messed with my previous fix I would never get it together again. A complex conglomeration of human and cow was all that held this part of Sophie’s heart together.
I slipped in the cannulas and rushed her back onto the bypass machine, intending to empty the heart before I tried to lift it. I then stopped it in a chilled, flaccid state with cardioplegia so it became still and cold like a heart on the butcher’s slab. This time, when I lifted it, the bulge at the back was obvious. Bacterial enzymes had dissolved the muscle protein and the antibiotics had failed to protect it from liquefaction. As a result, the abscess had expanded. I asked for a stitch on a large needle and attempted to draw the edges of healthy muscle together.
I knew where the circumflex coronary artery should be but couldn’t see it within the inflammatory morass. In went those deep stitches through what I thought was uninfected tissue around the margins of the mush. As I tied the knots I was concerned lest the suture material cheese-wire through, with fearful consequences. At this point, there was no bleeding because the heart was empty and had no pressure inside it. And, sure enough, the bulge was gone.
When I let blood back into the heart it began to squirm, then contract again. But from the ECG I could see that we were in trouble. Instead of discrete spikes we had the rounded hills characteristic of heart muscle starved of blood – myocardial ischemia is the medical term. I knew that I had trapped that critical coronary artery in the stitches. A list of suitable expletives came to mind, but for everyone’s sake I kept my mouth shut. Sophie couldn’t survive the inevitable heart attack. I needed to take down the repair and start all over again. Battle of the Bulge, round two.
We gave more cold cardioplegia solution, then I picked up the heart and carefully cut away the stitches and replaced them with fresh ones at a different angle, this time further away from where I anticipated the coronary artery to be. Still not confident with the repair on which her life depended, I tried a belt-and-braces approach, covering the area with biological glue and a haemostatic gauze patch, just like repairing a hole in my trousers. Then we tried coming off bypass once again. This time the ECG returned to normal – pointed peaks in the Dolomites rather than the whalebacks of the Brecon Beacons. That part of the ventricle had its blood supply again. Now we needed the botch job to hold up. My operation note was prescriptive and cautious: ‘Keep the blood pressure below 90 mm Hg. Keep Sophie asleep on the ventilator for seven days. It will not be possible to retrieve another catastrophe.’
As the chest was closed again, the back of the heart remained dry with no bleeding at all. The whole team felt a huge sense of relief when Sophie made it back to paediatric intensive care. The poor family were still in shock, waiting huddled together in the relatives’ room and mentally disintegrating with the intolerable suspense. I explained that an abscess had eaten its way through the heart wall, that I had never seen this terrible problem before but we had done our best to fix it. Then the usual defensive crap. The next twenty-four hours would be critical. The outcome remained uncertain, but where there was life there was hope. This was all true, but it felt hollow on my tongue as I faced the three sad faces, each too stunned to ask any questions apart from when could they see her. I quietly withdrew and flipped my emotion switch to the off position.
Leaving the grief behind, I passed Archer on the corridor on his way to see Sophie. In an understated fashion he said what he always says: ‘Well done, Westaby.’
I appreciated that, and subsequently learned that he wasn’t expecting to see her again. Archer and the intensive care doctors took over for the night. I should have called in to apologise to my cancelled patient but I didn’t. I wasn’t in the mood to apologise to anyone for anything by that point. I used to lose track of time when I was operating, and by now it was 9 pm. I needed a beer and some downtime. As was often the case I couldn’t sleep, expecting the phone to ring at any time. Eventually I pre-empted the call. At 3.30 in the morning I rang the unit to see how Sophie was getting on. She was stable but still had that swinging fever. They were actively cooling her and she was still not passing urine. Then the important phrase: ‘There’s no bleeding.’ So I lapsed gratefully into unconsciousness.
Less than twelve hours after that, Sophie and I were back in theatre. Battle of the Bulge, phase three. At 1.30 pm the chest drains rapidly filled with blood and her blood pressure disappeared. Now she was bleeding to death. Again, I knew there was no point opening her in the intensive care unit as they wanted me to do. Either we let her go in peace – the logical option – or I needed her back on the bypass machine while I reconsidered the options, if there were any.
My intensive care colleagues kept squeezing in uncrossmatched donor blood through the drips, managing to hold the blood pressure at around 60 mm Hg. Then we rushed her bed and overflowing chest drains down the corridor, scattering wide-eyed hospital visitors and leaving a stream of blood in our wake. Had this happened during the night there would have been no chance. Once more my great team pulled together and we got her onto the operating table before she completely exsanguinated.
Her sternum was open again in minutes, the chest full of fresh blood and clot compressing the heart. Bags of O negative blood were still being squeezed in through her neck veins, but within minutes we were back on cardiopulmonary bypass for the third time, albeit with a sense of resignation. Thinking, ‘What the hell do we do now?’, we had pulled back from the brink again – to what end? We had to draw the line somewhere, but not yet. As Albert Camus wrote, ‘Where there is no hope, it is incumbent upon us to invent it.’ The bottom line was that Sophie was only fifteen.
I was gripped by a fierce determination to get her through, but doing it by the book was not going to work here. Grim Reaper was outmanoeuvring me with my conventional approach. If this infected muscle was ever going to heal, it simply couldn’t happen while the left ventricle was continuously generating pressure and supporting the circulation, as it was the relentless pressure swings within the chamber that was causing it to rupture. This time Sophie had started to bleed when her sedation wore off. Her conscious state lightened, her anxiety rose, so her blood pressure shot up. Then whoosh. Torn muscle, followed by cardiac tamponade.
She needed a new heart, but that was not possible. Her desolate mother would have gladly given hers, but even if we’d had an organ donor in the adjacent theatre no one would contemplate a transplant in an infected child. No one except me, that is, but we could never find a heart in time, unless I commandeered one from a philanthropic medical student … Then suddenly, amid the panic and delirious fantasy, the penny dropped. My only practical option was to actively empty the left side of the heart and keep it empty. That would remove all the pressure from within the chamber. I could use a left ventricular assist device (LVAD) to suck down the left ventricle, maintain the circulation and rest the damaged muscle while antibiotics dealt with the infection. This might enable it to heal. Had the technology ever been used for this purpose before? Absolutely not. But that was even more reason for me to give it a try. In the highly unlikely scenario of it proving successful, I could publish a paper about it.
I was then reminded that we no longer had a Levitronix pump as my charitable funds had run out and we had used the last one recently on the baby. To my knowledge we didn’t have any more lifesaving equipment, and in the mortality statistics only one individual would carry the can for this death – me. The blame-and-shame system we worked in would record this one as a death after mitral valve repair. Lifesaving equipment is expensive but death is cheap. Push mortality rates into the public arena, but don’t give the stroppy surgeons the tools to do the job. Pause for a moment to reflect on the morality of that.
Then Brian the perfusionist came to my rescue. He had a different type of centrifugal blood pump on trial in one of our heart–lung machines. It was said to be safe to use continuously for three weeks at a time, unlike the conventional roller pump, where three hours was too long. I felt that three weeks would be sufficient – inflammatory adhesions and fibrous scar ought to plug the hole in that time. We should go for it since there were no other options.
While the perfusion team assembled the equipment, I stopped Sophie’s heart for the very last time and searched for the rent in the ventricular wall. Luckily it was on the edge of the infamous bulge, well away from that coronary artery. The stitches had cheese-wired through the fragile muscle again, so with more stitches and much more expensive tissue glue I stuck the back of the heart to the fibrous pericardium, using every possible trick to lessen the risk of further catastrophe because the next exsanguination would undoubtedly be her last.
This Rotaflow pump was very simple – there was just one knob to boost the flow or turn it down – and I emptied out the left ventricle by sucking blood from its apex through a wide-bore cannula, with the tubing that connected to the external pump circuit emerging from below Sophie’s ribs. Through a second tube the pump returned this blood to the aorta. Bleeding was still going to be our biggest problem; after three cardiopulmonary bypass runs within a short period of time, Sophie’s blood wouldn’t clot
, so we needed masses of donor clotting factors and yet more blood transfusion. The perfusionist started the whirling rotary pump at its lowest speed, aiming to transfer gradually from the one extracorporeal circuit to the next. Once more there would be no pulse pressure on the monitors, just the average pressure of the continuous, non-pulsatile blood flow. Meanwhile the wounded left ventricle was still contracting but not pumping, while the right ventricle continued to push blood through the lungs. Magic. So far, so good. We had renewed grounds for hope.
Because of the coagulation problems and diffuse bleeding, I decided to leave Sophie’s sternum wide open for forty-eight hours. We packed swabs around the heart, placed an adhesive plastic drape over the chest, and drainage tubes were brought out alongside the assist device cannulas. With tubes emerging from everywhere it was a terrifying spectacle for the family, and was also a little too much for the paediatric intensive care unit this time. We took her back to adult intensive care, where our senior nurses had more experience with pulseless patients and where the parents of other sick children wouldn’t be spooked by the sight.
As the bleeding damped down, Sophie remained stable. Both her kidneys and her liver had taken a hit, but dialysis could cope with that. Mum Fiona stayed remarkably calm, although the gore took its toll – it was a terrible strain for Sophie’s poor sister, who stayed away from school. Two days later we removed the blood-soaked packs and closed the chest over the hardware to reduce the risk of further infection. By now the blood was clotting, the oozing had stopped and the front of the heart looked fine. I was definitely not going to explore the back. The pump was working well and I was determined to leave it in place for at least another ten days to give the abscess site the best chance to heal. In the meantime, we learned that the mighty staphylococcus was resistant to our second combination of antibiotics, so we changed the drugs again. At last the high temperature abated.
The Knife's Edge Page 18