Immune therapies and natural killer cell testing
The test does not predict the outcome. The treatments have real side effects.
The package usually arrives together: a blood test said to show that your immune system is rejecting embryos, followed by treatment to suppress it. The treatments include intravenous lipid emulsion, intravenous immunoglobulin, steroids, and sometimes TNF blockers such as adalimumab.
The test. The best systematic review pooled 60 studies. Levels of natural killer cells in the womb lining were modestly higher in women with recurrent miscarriage and recurrent implantation failure than in controls, with very high variation between studies. But the finding that matters:
There was no difference in pregnancy outcome in women with RM/RIF stratified by uNK level, and no significant correlation between pNK and uNK levels in women with RM/RIF.
Two failures in one sentence. Sorting women by their natural killer cell level did not predict whether they had a baby. And the blood test did not correlate with the uterine measurement it is supposed to stand in for.
The biology, from the UK regulator's own patient information. Despite their name, these uterine NK cells are never in contact with the foetus and do not attack the embryo. Indeed, it is now becoming clear that they are beneficial for pregnancy and work in cooperation with the placenta so it can successfully become established in the uterus.
On the blood test: because it looks at cells in the blood rather than in the uterus, they offer no useful information in relation to pregnancy outcomes.
The treatments.
- Intravenous immunoglobulin for recurrent miscarriage: pooled across 8 trials in 303 women, odds ratio for live birth 0.98 (95% CI 0.61 to 1.58). No effect.
- Steroids around implantation: 16 randomized trials in 2,232 couples, of which only two reported live birth. Odds ratio 1.37 (95% CI 0.69 to 2.71) at very low certainty. Translated, a baseline of 9% becomes somewhere between 6% and 21%, which is compatible with meaningful harm and with meaningful benefit.
- Intralipid: no Cochrane review exists. The regulator rates it grey for every outcome it assesses.
- TNF blockers: no randomized trial of these for implantation failure or recurrent miscarriage could be found. That absence is itself the finding, for drugs that carry infection and malignancy warnings in their licensed uses.
The fertility society's own guideline concluded that the immunotherapies it reviewed are either not associated with improved live-birth outcome in IVF or have been insufficiently studied to make definitive recommendations.
Immunologic testing appears on its do-not-order-routinely list.
Because there is no evidence that any immune cells, including uterine NK cells, ever do prevent a pregnancy, there is no reason for any patient without an immunological disease to take these therapies. None of these treatments are harmless and some of their side-effects are serious... All the good evidence to date shows that there is a risk of considerable harm without any increase in the chances of having a baby.
Steroids and intravenous immunoglobulin are both rated red.