Salvage Therapy for Ovarian Cancer
Salvage Therapy for Ovarian Cancer
Of the 28,000 patients diagnosed with epithelial ovarian cancer in
1997, nearly 70% presented with advanced disease. Through the use of
primary taxane/platinum therapy, up to 50% of these patients achieved
a complete clinical remission. Unfortunately, the majority of
patients will have persistent disease after initial treatment or will
relapse within the first 3 years. As a result, salvage therapy
comprises the majority of patient encounters in the treatment of
epithelial ovarian cancer.
Many treatment options are now available for patients with persistent
or recurrent ovarian cancer; these include repeat treatment with a
taxane and platinum compound or a choice from among a menu of new
agents with activity in this disease. In addition to standard
systemic drug strategies, proposed alternatives for salvage treatment
include high-dose chemotherapy with stem-cell support, prolonged
maintenance chemotherapy, intraperitoneal chemotherapy, hormonal
therapy, secondary surgical cytoreduction, and radiotherapy.
Candidates for salvage therapy can be grouped into several different
categories. These differences are more than semantic, in that they
identify patients with markedly different prognoses and predict the
likelihood of response to treatment. Because of the heterogeneity of
these patients, they will be considered separately here.
The vast majority of patients with ovarian cancer respond to primary
therapy. A small percentage of patients (< 20%) may have
progressive tumor during primary treatment; this is defined as refractory
disease. In general, these patients are considered to have a poor
prognosis, and their poor performance status often precludes
A larger group of patients may show a partial clinical response to
six cycles of therapy. A partial response is characterized by
persistently elevated tumor markers or clinically evident disease at
the conclusion of treatment; this is termed persistent disease.
Despite a complete clinical remission, some patients may also have persistent
disease documented only at second-look surgical assessment. This
group can comprise up to 50% of patients with clinical complete remission.
A final group of patients may initially respond completely to primary
therapy but then relapse; these individuals are properly identified
as patients with recurrent disease. If the disease recurs in
< 6 months, it is defined as resistant to platinum (and
taxanes) and requires alternative therapies. If the disease recurs ³
6 months following primary therapy, it may be termed sensitive,
raising the potential for repeat treatment with platinum and taxane
compounds administered on the same or different schedules.
As mentioned above, the outlook for the < 20% of patients whose
disease progresses clinically during primary therapy with platinum-
and taxane-containing regimens remains poor. No curative strategies
are currently available for this group of patients, and any
intervention must be considered palliative.
Patients with primary refractory disease are generally offered
treatment with any of the various single-agent chemotherapies
discussed below. No randomized trial of second-line therapy in this
patient group has shown the superiority of one nonplatinum or
nontaxane agent over another, and participation in clinical trials is
particularly encouraged. Progress in testing new agents in this group
is confounded by the small number of patients, as well as their often
diminished performance status, which precludes additional
chemotherapy following primary treatment failure. Additional surgical
cytoreduction in patients with primary refractory disease has
significant morbidity and has not altered the median survival of 12 months.
Of all of the ovarian cancer salvage treatment groups, patients with
persistent disease at second-look surgical assessment are the only
group that appears to have the small possibility of achieving a cure
with currently available treatment strategies. This possibility of
definitive therapy provides a strong argument for accurate
post-treatment assessment in patients with apparent complete clinical
responses. Because of the limitations of computed tomography (CT) and
available biochemical markers, an invasive second-look assessment
remains the most accurate way to assess response to primary treatment.
Although the sensitivity and specificity of laparoscopic reassessment
vs formal laparotomy are under investigation, second-look procedures
have not been shown to prolong survival in previous comparative
trials. For example, a recent prospective trial randomized 102
patients in complete remission (as documented by clinical findings,
CT, laparoscopy, and serum markers) to laparotomy or observation.
Survival was not prolonged in patients who were surgically assessed.
However, it must be noted that a positive outcome from any
second-look procedure depends on the efficacy of salvage treatment,
since the intent of surgical reassessment is primarily diagnostic
rather than therapeutic.
No standard therapy currently exists for patients with persistent
disease. Three forms of salvage treatment can be considered for these
patients: prolonged chemotherapy with the same or different agents,
intraperitoneal therapy, or high-dose therapy with hematopoietic support.
Prolonged Systemic Chemotherapy
Platinum Agents--In general, extended administration of
platinum agents has shown no benefit in patientswith persistent
disease. Randomized trials of 6 vs 12 cycles or 5 vs 10 cycles of
cyclophosphamide (Cytoxan, Neosar), doxorubicin, and cisplatin
(Platinol) in advanced ovarian cancer demonstrated no survival
advantage for the longer courses of treatment.[4,5] These
observations are consistent with those seen in the treatment of
Hodgkins disease or adjuvant breast cancer chemotherapy.
In contrast, a retrospective M. D. Anderson Cancer Center (MDACC)
review of 116 optimally debulked patients investigated the
relationship between duration of chemotherapy with platinum-based
regimens and survival. Median progression-free survival of patients
receiving 12 vs 6 planned cycles of therapy was 30 vs 15 months (P =
Thus, while additional cycles of platinum therapy may be considered
for patients who may still be responding by CA-125 measurements, the
preponderance of evidence does not support this "more of the
Paclitaxel--Recently, the addition of paclitaxel (Taxol) to
primary therapy regimens has reopened the sustained chemotherapy
question. No trials have compared longer courses of paclitaxel to the
standard six cycles of treatment.
Inferential data can be gleaned from the large body of literature
describing paclitaxel treatment of refractory patients. In 103
heavily pretreated (more than three prior regimens) patients
receiving paclitaxel at Memorial Sloan-Kettering Cancer Center
(MSKCC) who showed a minimal objective response rate of 4%, the 2-
and 3-year survival rates were 18% and 11%, respectively. Of these
patients, 21% received six or more courses of paclitaxel, and
treatment-related disease stabilization may have had a greater impact
on the natural history than is predicted by the response rate in
The concept of prolonged delivery of paclitaxel and other
cell-cycle-specific cytotoxic agents remains to be tested in a
randomized, prospective trial.
"Consolidation" Treatment--The obvious alternative
strategy is to select a different single agent or combination regimen
for use as "consolidation" treatment. There is very little
long-term information on such an approach, and no data exist
regarding the impact of consolidation treatment on time to treatment
failure or survival. However, the sequential administration of
non-cross-resistant chemotherapy has been advocated by Norton and
others. Such sequential strategies are currently being tested in
breast cancer and may be worthy of consideration in ovarian cancer as well.
There is preclinical evidence of a cytotoxic dose-response curve with
many chemotherapeutic agents, and the strategy of increasing drug
delivery via the intraperitoneal route has been extensively studied
and reviewed.[9-11] Several basic pharmacologic principles can be
generalized to intraperitoneal regimens.
In a recent review, Dedrick and Flessner developed a spatially
distributed pharmacokinetic model to examine the two most common
problems to be overcome when designing intraperitoneal regimens: (1)
poor tumor penetration by drugs, and (2) incomplete irrigation of
serosal surfaces by the drug-containing solution. The model
predicts, in a mechanistic way, the characteristic penetration
distance of a drug and the apparent permeability of a peritoneal
surface. The goal is to provide insights into the expected effects of
such procedures as pharmacologic manipulation or physical mixing.
To date, most of the information about intraperitoneal therapy has
come from clinical studies with drug sampling and pharmacokinetic
analysis. The pharmacokinetic advantages of intra-peritoneal drug
delivery include the slow systemic uptake of drug from the peritoneal
cavity and rapid subsequent drug clearance from the plasma. For
hydrophilic drugs, passive absorption into peritoneal capillaries is
directly proportional to the overall peritoneal surface area and
inversely proportional to the square root of the molecular weight of
the agent. As shown in Table 1, an
increased mean peritoneal cavity/plasma concentration ratio has been
demonstrated for many chemotherapeutic agents with known activity
against ovarian cancer.
No randomized clinical trials of salvage intraperitoneal therapy have
been conducted in patients with persistent ovarian cancer. The only
available randomized data evaluating intraperitoneal therapy come
from a study of patients with previously untreated stage III ovarian
epithelial cancer conducted by Alberts et al. Among the 546
eligible patients in this study, estimated median survival was longer
in the group receiving intraperitoneal cisplatin (49 months; 95%
confidence interval [CI], 42 to 56 months) than in the group
receiving intravenous cisplatin (41 months; 95% CI, 34 to 47 months).
The effect of residual tumor size on predicting response to
intraperitoneal therapy has been repeatedly demonstrated in the
clinical arena, and the direct penetration of drug into tissue is
limited, ranging from 1 to 3 mm. Small-volume disease, generally
microscopic or < 0.5 to 1.0 cm, is necessary for the rational
application of this approach. The phase III randomized trial of
intraperitoneal therapy by Alberts et al in previously untreated
patients with stage III ovarian cancer showed marked differences in
survival based on tumor size at the initiation of intraperitoneal
therapy: Median survival was 76 months in patients with microscopic
disease; 42 months in those with £ 0.5-cm
disease; and 32 months in those with > 0.5- to 2.0-cm disease.
Cisplatin--Intraperitoneal cisplatin remains the most
extensively studied agent in the setting of small-volume persistent
disease after surgical reassessment. Overall, approximately 20% to
30% of patients with persistent ovarian cancer after initial
chemotherapy have been shown to respond to intra-peritoneal cisplatin
treatment. Surgically defined response (s-R) rates of 40% to 50% and
surgically defined complete response (s-CR) rates of 25% to 35% have
been reported in patients with small-volume residual disease
(microscopic or all tumor nodules £ 0.5 cm in diameter) treated
with a variety of regimens containing cisplatin.
Important predictors of response to intraperitoneal cisplatin include
not only size of the residual tumor but also prior response to
systemic cisplatin. In one series, even in the subset of patients
with < 0.5-cm residual disease, only 9% of platinum nonresponders
achieved a complete response (CR), as compared with 43% of such
patients with a previously documented response to intravenous platinum.
In the MSKCC experience, a small (< 15%) group of patients achieve
a pathologic CR after intraperitoneal cisplatin therapy and remain
disease-free for more than 7 years of follow-up. If this experience
can be duplicated by others, it may imply that salvage
intra-peritoneal cisplatin therapy may be effective in a small number
of patients with persistent disease.
Investigators at MSKCC have also recently reported preliminary data
on the use of three cycles of intraperitoneal cisplatin and etoposide
in patients with negative second-look surgical assessments. At
36-month follow-up, the median disease-free survival is 28.5 months
in the untreated group (concurrent historical matched controls) and
has not yet been reached in the treated group. The contribution of
etoposide in this regimen is unknown. A phase II trial of intravenous
cisplatin plus etoposide in 21 patients in complete clinical
remission (10 of whom had minimal residual disease at second-look
assessment) reported a median progression-free interval of 26 months,
which is similar to data with cisplatin alone.
Other Agents--Numerous other agents have been evaluated for
intraperitoneal administration, including paclitaxel, floxuridine
(FUDR), mitomycin (Mutamycin), carboplatin (Paraplatin), mitoxantrone
(Novantrone), interleukin-2 (Proleukin), interferon-alfa (Intron A,
Roferon-A), and interferon-gamma.[11,18,19] Intraperitoneal delivery
of mitoxantrone is limited by chemical peritonitis, making this drug
poorly tolerated and not recommended for this route of
administration. Responses have been seen in some patients, however,
including those who did not respond to intraperitoneal cisplatin.
Paclitaxel likewise causes abdominal pain at doses > 125
mg/m², and the recommended phase II dose with acceptable
toxicity determined by phase I studies is 60 to 65 mg/m² weekly.
A phase II Gynecologic Oncology Group (GOG) trial testing
intraperitoneal paclitaxel in patients with persistent disease
following primary therapy was closed to accrual in 1995 and has 76
evaluable patients; as yet, it is too early to assess outcome in
New intravenous agents with activity in epithelial ovarian cancer
that have recently been identified and tested largely in the advanced
salvage setting include liposomal doxorubicin (Doxil), topotecan
(Hycamtin), vinorelbine (Navelbine), docetaxel (Taxotere), and
gemcitabine (Gemzar).[21-25] If intraperitoneal administration of
cisplatin proves to be superior to intravenous administration for
small-volume disease, an important question will be to identify which
new agents may have a dose-response relationship and pharmacologic
properties appropriate for possible intraperitoneal administration in combination.
Summary--With the exception of one published trial, data
suggesting a benefit for intraperitoneal therapy come from
nonrandomized phase II trials. It is important to note that,
although the response rates recorded in the various trials may be
impressive, biases, such as treatment limited to patients with
responsive tumors or those with very small residual disease, cannot
be excluded or compared directly. Thus, intraperitoneal therapy
remains an investigational approach at present.