Aims: To evaluate the relationship between late morbidity (i.e. > or =6 months) and a tumour-negative sentinel lymph node biopsy (SLNB) in primary breast cancer patients by using a systematic review approach, and to identify the predictors of late morbidity.

Methods: We performed a systematic review of the literature for studies concerning the late morbidity of patients who had undergone SLNB alone or SLNB followed by ALND when SLN metastases were found. A literature search over the last 16 years (1993-2008) was performed in the databases MEDLINE and EMBASE. The methodological quality of the selected studies was assessed according to a list of predefined criteria. The data of assessment and predictors of late morbidity were collected.

Results: We identified a total of 32 papers reporting 27 independent cohort studies, of which 17 were high quality studies and were further analysed in this review. There was a great variation in the prevalence of pain (7.5-36%), impairment of range of motion (0.0-31.0%), oedema (0.0-14.0%), decreased strength (11.0-19.0%) and sensory disorders (1.0-66.0%). Factors such as time after surgery and young age were strong predictors of late morbidity. Breast surgery, radiation to axilla, tumour location, body mass index (BMI) and two-step procedure, especially lymph mapping techniques, could also predict the late morbidity to different extents.

Conclusions: SLNB-associated late morbidity, even with a low prevalence, remains a clinical problem which cannot be neglected in primary breast cancer patients. Time after surgery and young age are the important predictors for late morbidity in primary breast cancer patients after SLNB; breast surgery, radiation to axilla, tumour location, BMI and two-step procedure also have limited prognostic value.