Managing CKD Anemia: Ten to Twelve isn't being nice, it's just being stubborn

In response to my article on the NICE anemia guidelines, the following comment was made:

"You have mentioned that the evidence for Hb targets of 10 to 12 g/dl of NICE was weak. You have then proposed that "In fact, the RCT's would support a Hb range of 9 to 11 g/dL.". Does this statement have enough strong evidence or is it a simple projection after observing the dangers of higher (usually above 13 g/dl) targets in RCTs. (Anon Jan 25, 2012)”

Source: FDA CDRAC Briefing Document 2007

The evidence to support 9 to 11 g/dL is based on 2 studies: the Normal Hematocrit study that comprised of “symptomatic dialysis patients” – i.e., patients on dialysis with a history of cardiovascular disease – the typical dialysis patient. In this study, subjects were randomized to normalization of Hb (Hb target of 13-15 g/dL) or partial correction (9-11 g/dL). Guess what? The patients assigned to the Hb target range of 9 to 11 g/dL did better (see Table 1 and Fig.1).

Source: FDA CDRAC Briefing Document 2007

It is important to note that according to the US FDA (FDA Briefing document CDRAC 2007 and Singh testimony to the FDA) the NEJM paper reflected data from the third interim analysis and not from the final study report – so that the “non-significant” data in the NEJM paper when more rigorously examined was in fact statistically significant. Moreover, 25 of the patients who had a non-fatal MI subsequently died so that there were 221 deaths (35%) in the high Hb arm and 185 deaths (29%) in the low Hb arm (P<0.01 by the log rank test).

The other study is the “Trial to Reduce Cardiovascular Events with Aranesp Therapy” (TREAT).  In TREAT 4,038 subjects were treated with darbepoetin or placebo, with a goal Hb of 13g/dL in the active treatment arm. Rescue darbepoetin treatment was given to those with a Hb below 9 g/dL in the placebo arm.  The median achieved Hb in the intervention arm was 12.6 g/dL (IQR of 12.0-12.8 g/dL) and in the placebo arm 10.6 g/dL (IQR of 9.9-11.3).

TREAT was neutral for its primary composite endpoints: death or a cardiovascular event or death or end-stage renal disease. For the primary composite of death or a cardiovascular event (hazard ratio for darbepoetin versus placebo, 1.05; P = 0.41), but, there was a significantly higher rate of strokes in patients treated with darbepoetin. Fatal or nonfatal strokes occurred in 101 patients assigned to darbepoetin and 53 patients assigned to placebo (hazard ratio, 1.92; P<0.001). Death or end-stage renal disease occurred in 652 patients in the darbepoetin alfa group (32.4%) and in 618 patients in the placebo group (30.5%) (Hazard ratio for darbepoetin alfa vs. placebo, 1.06; 95% CI, 0.95 to 1.19; P = 0.29). A higher rate of both thromboembolism in the darbepoetin-treated patients was also observed.

The bottom-line? Normal Hematocrit demonstrates that a target of 9 to 11 g/dL is better than a higher target in dialysis patients, and TREAT demonstrates that patients randomized to placebo (keeping the Hb above 9 g/dL) did better with respect to stroke and no worse with respect to mortality and cardiovascular and renal complications.

The evidence to support a target Hb range of 10 to 12 g/dL is illusory and pharma-industry driven. Pushing a target of 10-to-12 g/dL is more dogma than science. The best evidence is for a Hb target of 9 to 11 g/dL.

Notwithstanding all of this, the goal should be tailoring anemia management by developing an individualized Hb trigger and using the lowest possible dose of ESA. The hemoglobin trigger is defined as the Hb concentration at which patients become symptomatic. The intervention could be ESA, iron, blood or a  combination thereof. This would mean that some patients who become symptomatic at 10 g/dL would require treatment, whereas others who are asymptomatic at lower Hb concentrations will need no intervention.