Effect Size of Denosumab versus Zoledronic Acid on Skeletal-Related Events in Myeloma

confidence: high · 2 sources

⚠
Working research — not a recommendation.

This page records what the evidence says a policy does, not what anyone should do about it. It has been synthesized from the cited sources but has not been independently reviewed.

What the evidence says

The main trial in newly diagnosed myeloma met its prespecified non-inferiority endpoint. Denosumab against zoledronic acid gave a hazard ratio (HR) of 0.98 (95% CI 0.85–1.14) for skeletal-related events (SREs). Rates were similar in the first 6 months, at 44% versus 45%. The trial was large, prospective and randomized, and radiologists reviewed skeletal outcomes blind. Evidence for non-inferiority on SREs is strong in both direction and magnitude.

A post-hoc analysis of the same trial found a lower SRE risk with denosumab: HR 0.66 (95% CI 0.44–0.98), roughly 34% lower. That interval comes close to 1.0, and the analysis was not prespecified, so the magnitude is less certain than the primary result. A pooled review of four randomized trials reports hazard ratios of 0.78–0.83 for the first SRE. That is a 17–22% reduction in the rate of first SRE, or a delay in its occurrence. The review's text describes this as non-inferiority in delaying the first SRE.

Secondary outcomes point the same way for denosumab on some measures. Median progression-free survival (PFS) was 46.1 months with denosumab and 35.4 months with zoledronic acid (HR 0.82, 95% CI 0.68–0.99). This was not the primary endpoint. Overall survival showed no difference, and the four-trial review reports similar survival in all studies. On pain, two of the four trials found that denosumab delayed pain worsening by 17–22%. Median time to worsening was 5.6–9.7 months with denosumab and 4.6–5.8 months with zoledronic acid. A secondary analysis of the Henry et al. trial also reports denosumab ahead on pain prevention, opioid use, quality of life and palliative radiotherapy. That text gives no effect sizes for those outcomes.

Safety differs by organ system. Denosumab produced fewer renal adverse events. A creatinine rise from baseline occurred in 9% of patients on denosumab and 16% on zoledronic acid. Creatinine doubling occurred in 3% versus 7%. The pooled review also reports fewer acute phase reactions with denosumab. In three of the four trials, hypocalcemia and osteonecrosis of the jaw were more common with denosumab. Most hypocalcemia was grade 1–2. The sources give no incidence rates for these harms.

The evidence on a mechanism is thin. One source notes that denosumab suppressed bone turnover markers more than zoledronic acid. Despite that, it found no difference in the primary outcome in the myeloma subgroup.

Where the evidence splits

The SRE comparison gives different answers depending on the analysis. The prespecified primary analysis shows non-inferiority (HR 0.98). The post-hoc analysis of the same trial shows a lower risk with denosumab (HR 0.66, 95% CI 0.44–0.98, p = 0.039). The sources say the trial was designed for non-inferiority and that the superiority result was post hoc. They do not say more than that about why the estimates differ. These are different questions, so the two results are not averaged here.

In the myeloma subgroup of Henry et al., time to first SRE did not differ (HR 1.03, p = 0.89). That subgroup had 180 patients, a minority of a trial that mostly enrolled other cancers. The four-trial review reports HRs of 0.78–0.83 for the first SRE. The sources on this page do not state whether those four trials were myeloma-specific or mixed-population. They also do not report how much of the pooled estimate comes from myeloma patients. The split is unexplained by these sources beyond the small subgroup size.

Within Henry et al., the primary outcome favoured denosumab on time to first SRE overall (20.6 versus 16.3 months, p = 0.06). The myeloma subgroup showed no difference. The sources do not say whether the subgroup's size accounts for the gap.

Thresholds and dose-response

No threshold established in the current sources.

Open empirical questions

Findings

Denosumab shows no measured difference from zoledronic acid on skeletal-related events in newly diagnosed myeloma (non-inferiority)

confidence — direction: highconfidence — magnitude: high (primary trial); low (Henry et al. myeloma subgroup)consensus: settled for non-inferiority; the lean is toward no difference.
mechanism
Not established in the claims. Denosumab suppressed bone turnover markers more than zoledronic acid in the Henry et al. myeloma subgroup.
effect
HR 0.98 (95% CI 0.85–1.14) for SREs. SRE rates in the first 6 months were 44% (denosumab) and 45% (zoledronic acid). In the Henry et al. myeloma subgroup (N = 180), time to first SRE showed no difference (HR 1.03, p = 0.89).
conditions / moderators
The primary result comes from a prospective randomized non-inferiority trial with blind radiologist review. The Henry et al. subgroup is small, and the broader trial mostly enrolled non-small-cell lung cancer and other cancers (breast and prostate excluded).

Denosumab decreases SRE risk relative to zoledronic acid in post-hoc and pooled analyses

confidence — direction: high (pooled); medium (post hoc)confidence — magnitude: mediumconsensus: contested, because prespecified results show non-inferiority and post-hoc or pooled results lean toward denosumab.
mechanism
Not established in the claims.
effect
Post-hoc HR 0.66 (95% CI 0.44–0.98), roughly 34% lower risk. Pooled review of four randomized trials: HRs of 0.78–0.83, a 17–22% reduction in the rate of first SRE or a delay in its occurrence.
conditions / moderators
The post-hoc result was not prespecified, and its interval nearly reaches 1.0. Effect sizes vary across the four pooled studies, and their confidence intervals overlap. The sources do not say whether the pooled trials were limited to myeloma.

Denosumab increases progression-free survival and shows no measured effect on overall survival

confidence — direction: highconfidence — magnitude: medium (PFS); high (no difference in overall survival)consensus: settled for no difference in overall survival; thin for the PFS difference, which comes from one trial.
mechanism
Not established in the claims.
effect
Median PFS was 46.1 months (denosumab) versus 35.4 months (zoledronic acid), HR 0.82 (95% CI 0.68–0.99). Overall survival showed no difference. The four-trial review reports similar survival rates in all studies.
conditions / moderators
PFS was not the primary endpoint, and the interval nearly reaches 1.0. The review finding on survival comes from all four pooled trials.

Denosumab decreases pain worsening and related secondary outcomes relative to zoledronic acid

confidence — direction: highconfidence — magnitude: mediumconsensus: thin, leaning toward denosumab on these secondary outcomes.
mechanism
Not established in the claims.
effect
Time to pain worsening was 17–22% longer with denosumab. Median time to worsening was 5.6–9.7 months (denosumab) versus 4.6–5.8 months (zoledronic acid). A secondary analysis of Henry et al. reports denosumab ahead on pain prevention, opioid use, quality of life and palliative radiotherapy need, without effect sizes.
conditions / moderators
Only 2 of the 4 pooled studies reported pain worsening as a secondary outcome. The Henry et al. analysis draws on a small myeloma subgroup.

Denosumab decreases renal adverse events relative to zoledronic acid

confidence — direction: highconfidence — magnitude: high (primary trial); medium (pooled review)consensus: settled, leaning toward fewer renal events with denosumab.
mechanism
Not established in the claims.
effect
A creatinine rise from baseline of at least 2 mg/L occurred in 9% (denosumab) versus 16% (zoledronic acid). Creatinine doubling occurred in 3% versus 7%. The pooled review reports fewer elevated creatinine events and fewer acute phase reactions.
conditions / moderators
The direction was consistent across multiple renal endpoints in the primary trial. The pooled review gives no incidence rates.

Denosumab increases hypocalcemia and osteonecrosis of the jaw relative to zoledronic acid

confidence — direction: highconfidence — magnitude: lowconsensus: thin, because it rests on one pooled review without quantified rates.
mechanism
Not established in the claims.
effect
Both events were more common in the denosumab group in three of the four pooled studies. No incidence rates are given. Most hypocalcemia cases were grade 1–2.
conditions / moderators
The claim is associational. The source gives no quantified rates, and the severity data suggest mostly mild cases.