2019 John Insall Award: Fructosamine is a better glycaemic marker compared with glycated haemoglobin (HbA1C) in predicting adverse outcomes following total knee arthroplasty

The July 2019 supplement of the The Bone and Joint Journal consists of papers from The Knee Society. In this episode, Mr Andrew Duckworth interviews Dr Noam Shohat, who is the lead author of one of the articles published in this supplement. The article "Fructosamine is a better glycaemic marker compared with glycated haemoglobin (HbA1C) in predicting adverse outcomes following total knee arthroplasty: a prospective multicentre study" was the recipient of the 2019 John Insall award.

Click here to read the article
Click here to view the table of contents for The Knee Society supplement

Transcript

[00:00:00] Hello everyone. I'm Andrew Duckworth and a warm welcome from your team here at The Bone & Joint Journal, to one of our podcasts in the month of July. As some of you may know for the months of June and July, we have released an overview podcast to accompany our supplements from the American Hip & Knee Society closed meetings. For this month we've already had a podcast with our Editor-in-Chief here at the journal, Professor Fares Haddad, and the guest editor for the Knee Supplement, Dr. Brian Springer, where we hope to provide you with a brief overview of the Knee Society, who the members are, as well as discussing how this collaboration came about and how we hope this will benefit you as our listeners and readers.

We also hope to have provided a behind the scenes insight into how the studies within the supplement have been chosen, as well as giving a brief discussion on a few select papers. With that in mind, over the next 15 to 20 minutes or so we'll be discussing in more detail one of the prize papers from The American Knee Society closed meeting in 2018.

I have the privilege today of being enjoined by Dr. Noam Shohat from the Rothman Orthopaedic Institute in Philadelphia, who is the lead author for the paper entitled Fructosamine is a better glycaemic marker compared with HbA1C [00:01:00] in predicting adverse outcomes following total knee arthroplasty, which was the winner of the John Insall Award from the Knee Society and will be published in the July supplement. Welcome Dr Shohat and a big thank you for taking the time to join us.

Thank you for having me today. 

So to kick us off Noam, as you state in your paper, diabetes is amongst one of the most prevalent and morbid conditions affecting millions of people worldwide and recent studies have reported anything from 30% to 50% of patients undergoing total knee arthroplasty have glucose levels in a diabetic or pre-diabetic range. So if you could just give us a brief background to your study in particular, looking at the potential current guidelines and consensus readings that have suggested with regards to monitoring glycaemic control in diabetic patients, particularly if there is an agreed gold standard. 

Yeah. So while total knee arthroplasty is a relatively safe and successful and operation complications do occur and among those PJI remains the most prevalent and dreadful with poor long-term outcomes. One of the ways to deal with this threat is to [00:02:00] prevent it from the first place and in recent years a tremendous amount of research and effort has been put into mitigating the risk by patient optimizations.

Diabetes has been a focus of much attention and a major concern to the medical community as its prevalence is on the rise with an expected 50% increase in the next 10 years. 

Now this is what we know about diabetes in the context of joint replacement. We know that diabetes is an independent risk factor for arthritis and consequently around 50% of our patients have diabetes or pre-diabetes with many of them unaware of their diabetic state. We also know from multiple studies and registry data, that diabetes is an independent risk factor for complications with PJI being the most studied. However, while there is an agreement that diabetes is an established risk factor for complications, the optimal marker for assessing glycaemic control and the threshold above which the risk for complication increases significantly remain [00:03:00] unknown.

So in the absence of a gold standard, glycated haemoglobin is a common use for assessing glycemia control. However, its association with complications is not clear. In a recent systematic review, looking at haemoglobin A1C and PJI, we found that more than 50% of the studies showed no association there. Furthermore, many physicians use a cutoff of seven based on the American Diabetes Association recommendation for defining inadequate control, but in a meta-analysis of these studies, clearly we show that this cutoff is not associated with an increased risk for PJI.

Another major limitation with haemoglobin A1C prior to surgery is the long time it takes to see a response to treatment. As the lifespan of haemoglobin is three months, in patients with inadequate control surgery will be delayed for three months and Nicolas Burey in a study a few years ago taught us that 70% of these patients will never reach surgery. So obviously [00:04:00] we must search for alternatives to haemoglobin A1C. And that was our thought when we first looked at fructosamine. 

So that's really, that's a really nice summary of the current legislation. Like you say, a lot of the emphasis has been on HBA, one C but obviously using it, hasn't really been the prediction we sort of wanted to be. So what is the current evidence or data about using fructosamine levels in these patients? 

So fructosamine was introduced in 1983 as an alternative to haemoglobin A1C. It's a measure of glycated serum protein, mostly being albumin with a life span of two to three weeks. That's reflecting glycemia control over a shorter period of time compared to haemoglobin A1C. It's been tightly correlated to continuous glucose monitoring, and it's been suggested to better reflect variability of glucose.

Now glucose variability represents the fluctuation of glucose around the mean. So for two given people with the same haemoglobin A1C, the glucose [00:05:00] variability may be significantly different. And these variabilities have been associated with negative outcomes in critical and non-critical patients. And we recently showed this association with PJI in hospital complications and mortality, following joint arthroplasty.

This makes fructosamine a very attractive marker in the preoperative period and in a single centre study that was published two years ago, we showed that patients with high fructosamine levels were at significantly higher risk for complications, mainly PJI. 

Okay, that gives us a good basis for us to sort of move on to the, how your study was performed. So obviously the aim of the study was to determine if there was an association between fructosamine and early adverse outcomes following total knee replacement. And it was an enlarged cohort of both diabetic and non-diabetic patients. It was prospective multicentre across all academic institutions and included patients undergoing elective knee replacement, and they were screened for glycaemic control in the perioperative period and were followed up for a  minimum of [00:06:00] 90 days post-operatively. So now just for our listeners, could you give a bit more detail on how and when the glycaemic control was assessed to the patients included in this study? 

Yeah. So I think glycaemic control started at the pre-admission testing. And that occurred between two to three weeks prior to surgery and included blood samples that were sent off to glycated hemoglobin and fructosamine testing.

Okay. And how long were... so obviously that was the baseline sort of data, how long were the patients then followed up for and what was sort of the primary outcome measure that you used and any other secondary outcome measures you looked at? 

Yeah. So, as you mentioned, we followed up with patients prospectively for three months. The primary outcome was a periprosthetic joint infection as defined by the International Consensus Meeting. Secondary outcomes included hospital readmissions, wound complications and any return to the operating room or death. 

And that was all collected prospectively, obviously. [00:07:00] And the analysis for the paper is obviously very nicely laid out, but could you just for our listeners give a brief overview of your power calculation in terms of the numbers you needed, and the reasoning also, you divided the cohort into four groups for some of the analysis. So if you could just talk us, took us through that as well. 

Yeah, correct. So we performed a power analysis and sample size calculation based on our preliminary findings from a prior single centre study. Once we collected the data, we determined the cut-off for fructosamine above which the risk for complication increases significantly.

And we did that using *inaudible*. So these type of analyses point out the threshold that optimizes sensitivity and specificity of a certain test, in this case for detecting adverse outcomes, mainly PJI. We found that 293 micromoles per litre is the optimal threshold for dichotomizing patients to high and normal levels of fructosamine. And we compared between two groups. [00:08:00] We did the same for glycated haemoglobin. And in addition, we also looked at 7 and 7.5, as there is a controversy in the literature regarding the optimal threshold for haemoglobin A1C. 

Now to investigate the independent utility of fructosamine unrelated to haemoglobin A1C, we split the cohort into four groups - those who had both elevated fructosamine and haemoglobin A1C, those who had only one of them elevated, and in those who had both not elevated. And we performed a univariate followed by a regression analysis, accounting for various confounders. 

That explains it really clearly. So if we move on to the results of the study, obviously there were 1,119 patients recruited. It's a very large number. Of those, just under 14% who had a prior diagnosis of diabetes and 15 of those patients had *inaudible* such as kidney problems or peripheral vascular disease and the periprosthetic joint infection rate [00:09:00] in the study was just under 1% and eight patients had wound complications. So now with all that sort of the core sort of data,  can you sort of detail the key results for our listeners in relation to fructosamine and the baseline demographics, but I suppose more importantly as a predictor of adverse outcomes or complications. 

Yeah. So we found a total of 60 patients representing 6% of the cohort had high fructosamine levels, meaning a level above 293 micromoles per liter. There were no significant differences in baseline characteristics between those with normal and high fructosamine  except for their ASA score and as anticipated those with high fructosamine were more likely to have diabetes. But interestingly, 30% of these patients with high fructosamine did not have diabetes. And that's an interesting thing to have in mind. 

Now the results of the univariate analysis showed that high fructosamine levels were significantly associated [00:10:00] with all of their first outcomes that we looked into except for wound complications. PJI rates were 8.4 times higher in patients with high fructosamine. Readmission and reoperation rates were 3.6 and four times higher in the high fructosamine group and two patients from the elevated fructosamine group died compared to only one in those with normal fructosamine. Finally while we did not show a statistically significant difference, the rate of wound complications were three times higher in the group of patients who had high fructosamine.

I mean, they're all really quite stark values those. And I think that you say you highlight that, you know, 30% of patients who, you know, who did not have a background of diabetes, but had a normal abnormal fructosamine levels. So you obviously then performed the regression analysis. This was performed to determine the role of fructosamine as an independent predictor of PJI and what did this show?

So that first outcome seen in the elevated [00:11:00] fructosamine group remains significant for PJI readmissions and reoperations after controlling for potential confounders, including a primary versus revision surgery, operative time, length of stay, BMI, *inaudible*. And this was true in a multiple regression analysis. So mortality rates were too low to assess in a regression model. But this really emphasizes that fructosamine is not just a proxy for general illness, but rather is an independent predictor of complications.

Yeah. Yeah, absolutely. It's quite impressive that model and how it is still controlling for those factors, is still a predictor.

So obviously that is fructosamine, but how did that then compare as you compared that with HBA1C as a predictor of complications and what were those results? 

Right. So unlike fructosamine, haemoglobin A1C failed to show a significant association with all of the adverse outcomes. We saw similar rates of adverse outcomes in patients with high and normal haemoglobin A1C. [00:12:00] And this was true for all haemoglobin A1C thresholds that were assessed. 

Okay. That's really interesting. And so obviously, if I could just finish it off with the results, you've got a very nice sort of figure where you demonstrate the results according to those four groups we just described earlier. Can you just summarize that for our listeners? 

So the figure shows the patients stratified into four groups based on their haemoglobin and fructosamine levels. Each column represents a group. Those with both fructosamine and haemoglobin A1C elevated. Those with only haemoglobin A1C elevated. Those with only fructosamine elevated. And those with both were not elevated. So out of the complications were significantly higher in the group who had only fructosamine elevated. You can see the rates of PJI readmissions and reoperations reaching 15 and 20% in that group. Really reflecting the added value of fructosamine over haemoglobin A1C and it's independent utility. 

Yeah, no, I think [00:13:00] I agree. I think it's a really great figure that, cause it just very clearly shows the role of the usefulness of fructosamine. So if we sort of move on to sort of your discussion about, you know, how the study fits into the current literature, I mean the study strengths are obviously without question. It's a prospective design, very large number of patients enrolled. And as we've obviously  discussed and heard the very robust analysis performed, and it's obviously highlighted very key findings and cutoffs with regards the role of fructosamine as a glycaemic control marker and its association with periprosthetic joint infection.

And I think one of the start findings is as you highlight in the paper is in patients with high fructosamine levels who have normal HBA1C levels, the rate of PJI readmission and reoperation reached 15%. So how do you feel that this data adds to the current literature in the area and potentially current guidelines in the future?

So our findings questioning the role of haemoglobin A1C as an appropriate [00:14:00] marker for glycaemic control in the preoperative period are consistent with two recent systematic reviews showing no or only a weak association between haemoglobin A1C and PJI. 

One of the major limitations of those prior studies was that they may have used a cutoff, which was too low. So we therefore examined both haemoglobin A1C of 7 and 7.5 as cutoff for determining inadequate control and still we were not able to show any association between haemoglobin A1C and any of the adverse outcomes assessed.

When comparing high haemoglobin A1C and high fructosamine the latter was much more predictive of adverse outcomes. Another interesting finding of the current study was that majority of patients with elevated fructosamine who developed a complication, had normal haemoglobin A1C levels. Further supporting the independent fructosamine then suggesting this should be the marker for assessing glycemic control in the future.

That's very nice. [00:15:00] So in terms of, you know, you talk about the potential financial implications of using fructosamine as a glycaemic marker. Could you just go into that in a bit more detail for our listeners? 

So fructosamine cost in the US is around $20 per test. Based on our findings, we would need to screen around 280 patients at a cost of just above $5,000 to detect and avoid a single PJI case, which is considerably less than the healthcare costs for treating a case of PJI. Even without accounting for morbity and the social cost. The general estimates for cost of treating a PJI in the United States is around a hundred thousand dollars. So these findings together with the fact that 50% with high fructosamine, who went on to develop PJI had normal haemoglobin A1C levels support the notion that all patients undergoing total knee arthroplasty should be screened for glycaemic control using fructosamine.

No, absolutely. And you know, [00:16:00] with that in mind, Noam, I think, you know, from your point of view, as the authors for our listeners, what do you feel your key take home messages are in light of all that and in light of any potential limitations of the study and where do you feel we really should go next with this research?

So fructosamine is an inexpensive task, which is readily available. It's more accurate and responds quicker to treatment compared to haemoglobin A1C. Our findings of high rates of diabetes and prediabetes as well as the increased risk goes with unknown diabetes caused for screening all patients undergoing total knee arthroplasty using fructosamine. And in those with the level above 293 micromoles per liter, the risk for surgery should really be carefully weighed against expenses. 

Yeah, that's very nicely put, I think that's right. And I think it's clearly, you know, with the data you have presented, I think it really adds to, like you say, backing up the systematic reviews that you've seen before and [00:17:00] emphasizing it's much, I would suggest, a much more important role than HBA1C. So I think, thank you so much, Noam. I think that's all we have time for. It's been a really excellent, informative discussion and a real congratulations on a great study and an invaluable addition to the literature. 

So to our listeners, we do have enjoyed joining us, and we encourage you to share your thoughts and comments through Facebook, Twitter, and a like. Feel free to post or tweet about everything we have discussed here today. And thanks again for listening.