OBJECTIVETo investigate the efficiency of multi-round fluorescence in situ hybridization (FISH) and its influencing factors in preimplantation genetic diagnosis (PGD).

METHODSA total of 48 couples accepted PGD because of various reasons: 24 with Robertsonian translocations, 16 with reciprocal translocations, 2 with pericentric inversions, one with advanced maternal age who had a previous liveborn of Down syndrome, 3 suffered from sex chromosome abnormalities and 2 repeated spontaneous miscarriages. After 72 retrieval cycles, 432 cleavage stage embryos with more than six cells were biopsied on day three. Only intact nuclei (396) were hybridized in order to verify the chromosomal status of the individual embryos. If previous FISH has failed to give conclusive results while the nuclei remained undamaged, the nuclei were hybridized once again. A total of 870 times of hybridization were conducted to 396 nuclei. Signal identification rates of each round as well as the influence of different probes to the hybridization efficiency were compared. Factors leading to inconclusive FISH results were analyzed as well.

RESULTSFive hundred and thirty five out of 870 hybridizations gave identifiable signals (61.5%). The second and third round FISH showed the best signals with an identification rate of 71.8% and 77.4%, respectively, which were significantly higher than those of the first round (52.8%, P < 0.01), the fourth round (55.8%, P < 0.05, P < 0.01), the fifth round (54.5%, P < 0.05) and the sixth round (27.3%, P < 0.01). The identification rate of centromere specific probe signals (CEP group) was 80.3% and the former three rounds in this group got the best quality of signals with an identification rate of 85.7%, 85.1% and 88.0%, respectively, which was significantly higher than that of the latter three rounds. The identification rate of other probe was much lower than with the CEP probe (55.2% vs. 80.3%, P < 0.01) and the best quality of signal in this group was achieved in the fifth round (72.7%), followed by the second round (66.1%) and the third round (63.8%). The identification rate of the first round (50.3%) and the sixth round (22.2%) were significantly lower compared with the second round (P < 0.01). During the 6 rounds of FISH, 335 hybridizations did not give conclusion results (38.5%, 335/870). The main cause of unidentification was weak signals (20.9%, 182/870). Other common factors included background interference (7.6%, 66/870) and failed hybridization (6.1%, 53/870). Rare causes included nucleus damage (1.8%, 16/870), nucleus loss (1.1%, 10/870) and signal split/overlap (0.9%, 8/870).

CONCLUSIONMulti-round FISH can improve the utility of single nucleus in PGD and the former three rounds have the highest efficiency. The hybridization effect of CEP is better than other probe. Poor signal quality is the common cause of unidentification results.