The objective of this paper was to determine the role of radical hysterecto
my in persistent or recurrent cervical cancer after primary radiation thera
py.
Between 1982 and 1995, 34 patients underwent radical hysterectomy for persi
stent (n = 15) or recurrent (n = 19) cervical cancer after primary radiothe
rapy. Univariate analysis using log-rank comparison of survival curves was
conducted to identify clinical and pathologic factors predictive of surviva
l. The median tumor size at the time of recurrence or persistence was 3.2 c
m (range 1-6 cm). 24 patients (70%) had recurrence limited to the uterine c
ervix; four (12%) had vaginal involvement and six (18%) had early parametri
al involvement. No treatment-related deaths were observed. Eighteen major c
omplications (grade III-IV) occurred in 15 cases (44%); 5 patients experien
ced a fistula. Mean follow-up time was 81 months (range 33-192 months). Rec
urrent disease was documented in 20 patients (59%), and median time to recu
rrence was 37 months (range 4-56 months). Fifteen patients (44%) are alive
without evidence of disease at a median survival of 81 months (range 33-192
), and 18 patients (53%) died of disease with a median survival of 22 month
s (range 7-106). One patient died of intercurrent disease. Actuarial 5-year
survival rate for the whole group is 49%. Patients with FIGO stage IB-IIA
at primary diagnosis, no clinical parametrial involvement, and small (less
than or equal to 4 cm) tumor diameter at the time of recurrence show a good
prognosis (11/17 alive NED) compared to patients who do not fit the above
mentioned criteria (4/17 NED, P 0.01). We conclude that radical hysterectom
y can be offered as an alternative procedure to exenteration only in highly
selected patients.