Start neratinib at a lower dose and build up over two weeks, as the default

The evidence. In the open-label phase II CONTROL trial, 563 people starting a year of extended adjuvant neratinib were enrolled into six cohorts in turn. Grade 3 or worse diarrhoea occurred in 31 per cent with scheduled loperamide alone, 28 per cent with budesonide added, 21 per cent with colestipol added, 33 per cent with colestipol and loperamide as needed, 27 per cent with a four-week dose escalation and 13 per cent with a two-week escalation (120 mg, then 160 mg, then 240 mg from week 3). Diarrhoea-related discontinuation was 20 per cent with loperamide alone and 3 per cent with the two-week escalation. Without any prophylaxis, in ExteNET, grade 3 diarrhoea occurred in 40 per cent.

Where it stands. The US label (July 2026) lists the two-week escalation as an alternative to 56 days of scheduled loperamide; its own table of the two CONTROL cohorts shows grade 3 diarrhoea of 32 against 13 per cent and discontinuation for diarrhoea of 18 against 3.3 per cent. Neither the label nor any guideline read prefers one strategy, and how often each is used has not been measured in any source read.

The doubt. The cohorts were sequential and compared with each other and with historical controls, not randomised; later cohorts may have benefited from teams' growing experience. Whether escalation affects how well neratinib prevents recurrence has not been tested, though the label's table shows similar median dose intensity in the two cohorts (234 and 230 mg a day).