Intra-peritoneal cisplatin and whole abdomen hyperthermia for relapsed ovarian carcinoma

Abstract

The study was designed to determine the maximum tolerated dose (MTD) of IP cisplatin [CDDP] combined with intravenous thiosulphate and concurrent whole abdomen hyperthermia for advanced, recurrent or progressive ovarian carcinoma. Between September 1991 and November 1998, 41 patients with advanced epithelial ovarian cancer received escalating doses of IP (IP) cisplatin (six cycles given every 3–4 weeks) and whole abdomen hyperthermia with intravenous thiosulphate as second line treatment. Whole abdomen hyperthermia was administrated using a BSD-2000 annular phased array system. Forty-one patients were enrolled in the phase I/II portions of the study. Forty-four per cent (18/41) had undergone sub-optimal cytoreductive surgery and 15% (6/41) had been optimally debulked of their disease. Ninety per cent (37/41) had platinum-resistant disease and 10% (4/41) had platinum-sensitive disease. No DLTs occurred in the phase I testing and the recommended dose for this combination schedule was 180 mg m⁻² of IP cisplatin with thiosulphate and whole abdomen hyperthermia. The overall response rate was 44% (10 CR, 8 PR) and the median survival for all patients from protocol entry was 30 months (range 2–107 months). Median duration and survival of those achieving a pathologic CR was 14 months (range 2–27 months) and 35 months (range 14–71 months, 95% CI 16–54 months), respectively. Salvage platinum based IP cisplatin with hyperthermia did achieve pathologic CR in selected patients and was well tolerated. These promising results suggest a role for the use of adjuvant whole abdomen hyperthermia as a means of augmenting chemosensitization.

• Ovarian cancer
• intra-peritoneal
• cisplatin
• hyperthermia

Introduction

Ovarian cancer is the fourth most frequent cause of cancer-related death in women in the US. Approximately 70% of women present with advanced disease. Although first-line therapy in advanced ovarian cancer is associated with high response rates, most women will relapse and ultimately succumb to their disease [1]. Despite advances in multi-disciplinary management that includes surgery, radiotherapy and chemotherapy, the 5 year survival of patients with advanced disease is 25–35% [2], [3]. Various clinical strategies have been developed for salvaging relapsed patients including high dose chemotherapy [4], whole abdomen radiation [5–7], intra-perionteal (IP) ³²P [8], IP chemotherapy [9–12], immunotherapy [13], monoclonal anti-HER2 antibodies [14], phase I chemotherapy agents and hyperthermia [15–18]. This study reports on a treatment strategy for relapsed ovarian cancer that includes IP cisplatin combined with intravenous thiosulphate and concurrent whole abdomen hyperthermia.

Intra-peritoneal drug delivery for ovarian carcinoma was initiated in the 1970s [11]. The rationale for this approach is that the peritoneal clearance rate of some drugs may be lower than that of plasma, allowing the use of higher drug concentrations within the peritoneal cavity than can be used in serum. This could maximize the local cytotoxic effects of the drug while minimizing systemic toxicity. During the last several decades, multiple phase I and II trials have reported the safety, pharmacokinetics and favourable clinical outcome associated with IP drug administration [19]. Three prospective randomized clinical trials have now been performed in the Gynecologic Oncology Group (GOG) and all have demonstrated an advantage favouring the intra-peritonal arm [9], [20]. Despite the confirmed improved response rates and survival, IP therapy has not become the standard of care most likely due to the increased haematologic and non-haematologic toxicities as well as the catheter-related toxicities present in the IP arm. However, IP therapy continues to be an intriguing therapeutic modality.

The concept of delivery of IP chemotherapy within a heated perfusate in order to potentiate drug activity was proposed nearly 20 years ago [21], [22]. The choice of chemotherapeutic agents to be used with hyperthermia would ideally include those to have known efficacy in treating ovarian carcinoma at normothermic conditions and have demonstrated potentiation of effect at elevated temperatures. The usefulness of cisplatin delivered either systemically or intra-peritoneally has been discussed above and further in vitro data have demonstrated that cisplatin cytotoxicity is increased by 2 logs at 42°C with clinically achievable concentrations of cisplatin in tissue [23], [24]. Cisplatin has a relatively greater degree of thermal potentiation at 42°C than carboplatin or other platinum complexes [25]. The effect of heat may be to increase the cisplatin-DNA reaction rate, although mechanisms are incompletely understood. Potentiation of cisplatin cytotoxicity at elevated temperatures is found only for simultaneous administration of heat and drug [26]. This potentiation persists in some cisplatin resistant lines and elevated temperatures reversed acquired cisplatin resistance in a murine tumour line [27], [28]. Thus, there is a strong rationale for selecting cisplatin as the first agent in investigating toxicity and efficacy of an IP drug combined with hyperthermia for the treatment of ovarian cancer.

The purpose of this study is to evaluate the overall toxicity and clinical activity of combination IP cisplatin and whole abdomen hyperthermia in patients with persistent or recurrent epithelial ovarian carcinoma.

Patients and methods

Between September 1991 and March 1999, 41 patients were enrolled in a Phase I/II trial of IP cisplatin with intravenous thiosulphate and whole abdomen hyperthermia for recurrent or persistent advanced epithelial ovarian cancer.

Eligibility requirements

Patient eligibility included those with either measurable or assessable disease Stage III or IV epithelial ovarian cancer with recurrent disease or disease which was initially refractory to standard cisplatin-based therapy. Patients were required to have macroscopic disease ≥1 cm and were stratified with respect to being 1–2 cm vs >2 cm. Patients were also required to have no serious underlying co-morbid diseases likely to limit survival or prevent protocol treatment such as pneumonia, sepsis or other serious infection. They were required to have normal renal, cardiac, hepatic and pulmonary functions and a GOG performance status of 0, 1 or 2. A signed informed consent was obtained for both IP cisplatin and regional hyperthermia as approved by the Institutional Review Board.

Cisplatin administration and dose escalation

Intra-peritoneal fluid and cisplatin was administered through a portacath or Tenckhoff catheter placed surgically prior to protocol enrollment. Serum electrolyte abnormalities were identified prior to each planned IP therapy and corrected before proceeding with treatment. Intravenous hydration was started for 2–4 h prior to IP fluid. All patients were pre-medicated with lorazepam, metoclopramide, diphenhydramine and dexamethasone, for anti-emetic coverage and to prevent cisplatin-associated hyper-sensitivity reactions. One thousand millilitres of normal saline were given intra-peritoneally using sterile technique. The fluid was preheated to 37°C prior to IP administration. Thirty minutes prior to complete installation of the first 1000 mL of normal saline, pre-medications were administered including sodium thiosulphate 4 gm m⁻² and 250 ml of sterile water over 10 min (loading dose) followed by an infusion of sodium thiosulphate 12 mg m⁻² IV at 167 ml h⁻¹ (maintenance dose). Sodium thiosulphate is used for renal protection given in a setting of IP cisplatin. As soon as the maintenance dose of sodium thiosulphate was initiated, the cisplatin was instilled intra-peritoneally in 1000 mL of normal saline as rapidly as possible. Treatment was to be delivered on an every-21 day schedule. Patients were permitted to receive a maximum of six courses of therapy. Patients experiencing excessive toxicity or disease progression were to be removed from study.

It was planned that the cisplatin dose would be escalated in this Phase I trial. Following the initial cohort where patients were administered cisplatin at a dose of 60 mg m⁻² (level 1), subsequent patients were to be treated with cisplatin doses of 100 mg m⁻² (level 2), 140 mg m⁻² (level 3) and 180 mg m⁻² (level 4). There were no intra-patient dose escalations in this trial. Assuming acceptable toxicity, a minimum of three patients were to be treated at each dose level for six cycles before dose escalation was allowed.

It was anticipated that renal toxicity would be the dose-limiting toxicity of the combination therapy. The following criterion were applied in the dose escalation protocol: (1) If no Grade II renal toxicities nor any Grade III toxicities occurred in three patients who complete six cycles of treatment at the starting dose, after at least 3 weeks have elapsed from the completion of the last treatment of this group, a new cohort of patients was treated on an escalated dose. If any patient had a serum creatinine of ≥2.0 mg ml⁻¹, any Grade II or Grade III toxicity before three patients completed six cycles of treatment, then additional patients were accrued until a total of six patients completed six cycles of therapy. If ≤2 patients experienced a rise in serum creatinine ≥2 or any Grade 2 renal or Grade 3 toxicity, then the dose of the next cohort was escalated by 40 mg m⁻². Prior published data suggested that doses >200 mg m⁻² will be unlikely to be tolerable [12], [29].

If two instances of dose-limiting toxicity were observed at any dose level, the immediately preceding dose level would be considered the MTD to be recommended for phase II testing.

Thermometry catheter placement

Patients enrolled on this protocol had two IP catheters placed prior to protocol enrolment. One catheter was an IP portacath where drugs and fluids are delivered to a subcutaneous reservoir which has a direct out-flow through a catheter into the peritoneal cavity. The other catheter was a Tenckhoff catheter which provided external access to fluid administration. In addition, the Luxtron thermometry probes were inserted into the Tenckhoff catheter under sterile conditions for monitoring of the IP temperature. Both of these catheters were placed within the abdominal cavity, however they were not localized to any exact location within the abdomen and likely moved with patient position.

Hyperthermia treatment and thermometry

Immediately after the cisplatin completed IP installation, regional hyperthermia was administered to the whole abdomen using the BSD Sigma 60 hyperthermia system. The initial power was limited to ≤1500 W with phase and amplitude adjusted for equal surface electrical fields in each quadrant of the applicator. Since the exact temperatures (oral and IP) must be determined empirically, power was adjusted for a T_oral ≤ 38.5°C as a first priority goal and the average IP fluid temperature (T_IP) ≥ 41.5°C as a second priority goal. Power was also adjusted to minimize patient perceptions of discomfort that appeared to be power related. It was also attempted to achieve a T_IP ≥ 41.5°C within 30 min after initiation of heating. Heating continued for 60 min after the average IP temperature of 41.5°C had been achieved or after 30 min of power application, whichever was shorter. The bolus temperature (which can be regulated) was 37°C at the initiation of power and was reduced as necessary for patient comfort and/or to help maintain the oral temperature ≤38.5°C. All temperatures were measured with calibrated BSD thermistor sensors or Luxtron fibre-optic sensors. The oral temperature was recorded as a measure of systemic temperature every 5 min starting 5 min before initiation of power and continuing after cessation of power until oral temperature exhibited a downward trend and was ≤38.0°C. IP temperatures were monitored and recorded at least every 5 min using sterile 4 sensor Luxtron thermometer (sensors 2 cm apart) inserted into a percutaneously placed catheter (Tenckhoff catheter). Thermometer and catheter were inserted to a depth ensuring that all sensors were within the IP catheter within the peritoneal cavity. Average IP temperature is defined as the arithmetical average of the temperatures of the four sensors at a given time. Additionally, temperatures from the rectum and vagina were monitored routinely.

Treatment modifications

Weekly blood work was obtained for creatinine and blood counts (white blood cells, haemoglobin/haematocit and platelets). For renal toxicity ≥ grade 1, serum increase of creatinine ≥0.5 mg%, serum creatinine ≥2.0 mg dl⁻¹ or any other toxicity ≥2, the treatment was withheld until the toxicity lowered to ≤grade 1. If the toxicity delay was renal or if the delay was ≥2 weeks the cisplatin dose was reduced one level for all further treatments of that patient. If toxicity had not sufficiently abated by 6 weeks, no further treatments given.

Evaluation of toxicity

The National Cancer Institute Common Toxicity Scale was employed in evaluating the toxicity experienced by patients treated on this trial. All patients initiating therapy were evaluable for toxicity of the treatment programme.

Response criteria

Response to treatment was assessed following completion of therapy. Patients without clinical evidence of disease were offered a third look laparoscopy. Not all patients who achieved a clinical complete response agreed to third look laparoscopy. Complete response was defined as the disappearance of all gross disease including radiographic and physical disease as well as normalization of CA 125 if previously elevated. A partial response is defined as a 50% or greater reduction in the maximal dimensions of each lesion or the disappearance of gross disease with persistent CA 125 elevations. Pathologic complete response is defined as the disappearance of all gross disease clinically, radiologically and at the time of third look laparoscopy with normalization of CA 125. Stable disease was defined as no significant changed in measurable or evaluable disease for 4 weeks or more and no increase in the size of any known or development of new disease. Progression was defined as an increase in size (25% or more) or the appearance of new malignant lesions.

Results

Patient characteristics

A total of 41 patients, with a median age of 55 years (range 24–70), were entered on this Phase I/II trial. Thirty-seven patients (90%) had FIGO stage III disease and four patients (10%) had FIGO stage IV disease. Forty-four per cent (18/41) had undergone sub-optimal cytoreductive surgery and 44% (18/41) had been optimally debulked of their disease. In six cases, the extent of cytoreductive surgery was unknown. Ninety per cent (37/41) had platinum-resistant disease and 10% (4/41) had platinum-sensitive disease. Fifty-nine per cent (24/41) had persistent disease at the time of the second look laparotomy. On pathology review, 38 had moderate-to-poorly differentiated lesions, 31 serous, two epithelial, five adenocarcinoma NOS. Two patients were found to have well differentiated tumours of low malignant potential tumours of the ovary (borderline). On review, one patient had peritoneal mesothelioma. All patients underwent laparotomy with placement of an IP port-a-cath and a Tenckhoff catheter at the time of protocol enrolment and the extent of residual disease was estimated. Sixty-three per cent (26/41) had disease at entry measuring between 1–2 cm and 37% (15/41) patients had disease >2 cm at the time of enrollment (Table I).

Table I.  Patient characteristics.

Characteristic	Number of patients (%)
Stage
III	37/41 (90%)
IV	4/41 (10%)
Initial cytoreductive surgery
Optimal	17/41 (42%)
Sub-optimal	18/41 (44%)
Unknown	6/41 (14%)
First-line chemotherapy
Cisplatin/cyclophosphamide	13/41 (32%)
Cisplatin/paclitaxel	11/41 (27%)
Carboplatin/paclitaxel	10/41 (24%)
Cisplatin/cyclophosphamide/adriamycin	2/41 (5%)
Paclitaxel/cisplatin/carboplatin	2/41 (5%)
Paclitaxel	1/41 (2%)
Melphalan	1/41 (2%)
No therapy	1/41 (2%)
Platinum-free interval
≤6 months	37/41 (90%)
>6 months	4/41 (10%)

Cisplatin toxicity

At the first three dose levels (60, 100 and 140 mg m⁻²) there was no renal toxicity noted (Table II). There were five episodes of allergic reaction (two grade 1, three grade 2) to cisplatin in these first three cohorts. An interim analysis at the 140 mg m⁻² dose level lead to the discovery of the infusion rate correlating with the anaphylactoid reaction and when the infusion was prolonged to <3 ug m⁻² min⁻¹ no further episodes of anaphylaxis were observed [30].

Table II.  Treatment toxicity by CDDP dose.

CDDP dose (mg m^−2)	Renal toxicity events/#patients	Allergic reaction events/#patients	Catheter complications events/#patients	Hyperthermia complications events/#patients
60	0/3	1/3	0/3	0/3
100	0/9	2/9	1/9 (peritonitis)	2/9
				2nd degree burns
140	0/9	2/9	3/9 (obstruction)	0/9
180	3/20	1/20	5/20	0/20
			3 obstruction
			2 infection

At the 180 mg m⁻² dose three patients were initially evaluated and no DLT occurred and a total of 20 patients were enrolled at this dose level. Subsequently, there were three cases of renal toxicity (one grade 4 and two grade 2) as indicated by an increase in the serum creatinine. The cisplatin dose was reduced and one of these two grade 2 patients completed all six cycles. The other grade 2 patient completed only five cycles because of the transient increase in creatinine. One patient had a markedly elevated BUN and creatinine following the first IP treatment (creatinine 7.3 mg dl⁻¹). She also had evidence of hyperglycemia with glucose in the 800 mg dl⁻¹ necessitating in-patient admission for hyperglycemia and renal failure. These toxicities resolved slowly with conservative management but she did require short-term dialysis. A thorough evaluation with regard to the aetiology of her acute renal failure did not reveal any other factors which may have been the cause of her transient renal failure. There was also one episode of cisplatin allergy (grade 1) at the 180 mg m⁻² dose. The other acute toxicities relating to cisplatin were generally mild and consisted principally of Grade I to II nausea and vomiting and blood count suppression which occurred transiently within a 2–3 week interval following each cycle of IP therapy.

Catheter complications

There were nine catheter related toxicities; one case of cellulitis, two episodes of peritonitis and six instances of catheter obstruction. Cases of obstruction involved the Tenckhoff catheter and occurred after one cycle in four patients and after two cycles in one patient and after five cycles in another patient. As IP thermometry could not be obtained in cases with obstruction, therapy was discontinued on this basis. The catheter obstructions were likely to be related to a malfunction in the catheter and/or obstruction related to IP adhesions which may have mechanically compressed the catheter. The majority of catheter-related complications (5/9: 56%) occurred at the highest dosing level, 180 mg m⁻² and included two cases of infection (one cellulitis and one peritonitis). The higher risk of infection may be secondary to increased myelosupression at the highest cisplatin dose.

Hyperthermia complications

In general, hyperthermia was well tolerated. There were two instances of second degree burn, which both occurred at the second dose level (100 mg m⁻²) which resolved with conservative management. At the 180 mg m⁻² dose level there was one patient who received six IP chemotherapy treatments but only five hyperthermia treatments due to computer failure during one of the whole abdomen hyperthermia sessions. Toxicities for the entire series are summarized in Table II.

Thermometry considerations

Analysis of thermometry data from normal tissue temperature measurements as well as IP measurements has been performed on all patients who completed the six cycles of therapy. Results for individual patients are presented in Table III, which shows for each cisplatin dose level the highest IP temperature and highest normal tissue temperature noted for each session. The exact IP location of the Tenckhoff catheter was not quantitatively localized before each session and it is possible that the position within the peritoneal cavity may have shifted from treatment to treatment. In general, the normal tissue temperatures derived from rectal and vaginal measurements tracked closely with the IP temperatures.

Table III.  Thermal data on individual patients.

CDDP=	HT#1 Max tumour temp	HT#1 Max normal temp	HT#2 Max tumour temp	HT#2 Max normal temp	HT#3 Max tumour temp	HT#3 Max normal temp	HT#4 Max tumour temp	HT#4 Max normal temp	HT#5 Max tumour temp	HT#5 Max normal temp	HT#6 Max tumour temp	HT#6 Max normal temp
60	42.1	41.9	40.9	41.3	41.8	40.9	42.4	43.2	41.1	40.5	41.6	41.7
60	41.7	40.9	42.3	41.7	42.1	43.0	44.6	42.2	42.3	41.6	42.2	41.8
60	42.2	40.4	41.8	41.0	41.2	41.0	40.9	41.6	41.5	41.3	42.2	41.2
100	42.9	41.3	41.9	40.4	42.0	41.4	42.7	41.6	41.1	41.5	39.9	41.4
100	41.6	41.7	41.9	41.0	41.9	41.1	41.5	41.9	41.4	41.0	41.6	40.5
100	41.8	41.7	41.9	40.2	41.8	41.8	41.8	41.0	41.9	41.7	41.9	41.8
140	42.4	40.8	41.9	39.9	41.2	41.1	41.3	41.2	41.7	40.9	41.6	41.0
140	42.3	41.5	41.3	39.7	41.1	40.7	41.7	39.9	41.8	41.8	43.0	42.4
140	40.3	41.5	40.4	41.8	42.1	40.9	41.3	41.8	41.4	41.4	42.0	41.5
140	42.4	42.4	40.6	42.7	42.3	42.1	41.8	42.0	42.0	41.0	42.7	42.7
180	40.1	40.8	41.1	41.9	40.9	41.4	41.3	*	41.1	41.6	40.9	41.5
180	42.5	41.7	38.3	42.8	42.3	41.4	41.8	40.8	42.2	40.9	41.7	41.0
180	42.1	40.7	43.0	40.1	41.9	39.6	41.7	40.9	42.2	40.5	42.0	39.5
180	41.9	41.5	42.3	43.1	42.6	42.0	43.2	41.6	42.8	41.1	43.0	41.8
180	41.1	41.7	42.0	41.6	41.5	39.7	40.8	42.1	41.7	*	41.8	41.4
180	41.7	40.5	42.4	40.8	41.5	41.7	42.2	41.8	42.1	41.8	41.4	41.1
180	40.1	41.7	41.0	41.3	41.3	41.4	41.7	41.8	40.7	41.7	41.2	*

* Data missing.

Treatment response

Forty-one patients in the trial were evaluable for response in this intent-to-treat analysis. The overall response rate was 44% (10 CR, 8 PR) and consisted of five pathologic CR and five clinical CR. Response was documented for patients at the various dose levels, as shown in Table IV. Only one patient with a pathologic CR (at 100 mg m⁻²) had disease >2 cm at protocol entry. Only one patient who achieved a clinical CR (at 140 mg m⁻²) had initial disease >2 cm. All other patients achieving a clinical or pathologic CR (eight out of 10) had disease <2 cm at initiation of protocol therapy. The response rate for patients with platinum-resistant disease was 43% (23% CR and 20% PR) and for patients with platinum-sensitive disease the response rate was 75% CR and 25% NR. For the entire cohort of 41 patients enrolled on this protocol, the overall median survival was 30 months following protocol enrolment (range 2–107 months). There were a total of four patients who achieved a pathologic CR and the median relapse free survival for these patients was 16 months (range 5–27 months). The median survival for those patients achieving a pathologic CR was 74.5 months (range 23–96 months). For those achieving a clinical complete response the median relapse free survival was 13 months (range 5–34 months) and the overall survival for this cohort was 62 months (range 14–107 months). Some of these patients did receive further systemic therapy following completion of this protocol, such as gemcitabine or Doxil™ (liposomal doxorubicin).

Table IV.  Disease free survival (DFS) and overall survival (OS).

CDDP dose (mg m^−2)	Pathologic CR # pts/OS (months)	Clinical CR # pts/OS (months)	Clinical PR # pts/OS (months)	Clinical NR # pts/OS (months)
Disease free survival
60	0	1/12
100	1/27	1/27
140	1/14	2/14
		24
180	2/1	7/1
18	5
		10
		11
		13
		18
		34
Overall survival
60		1/33	1/34	1/14
100	1/71	1/71	2/26	4/2
			39	7
				26
				62
140	1/96	2/96+	2/13	4/3
		107+	28	8
				13
				42
180	2/23	7/14	2/30	9/3
78	23	37	3
		32		5
		57		10
		62+		10
		67+		34
		78		35
				58
				75+

On pathology review, two patients had low malignant potential tumours of the ovary. One of these patients was treated at the 140 mg m⁻² CDDP dose and completed the full six cycles of therapy. She was found to have a clinical complete response with a duration of 24 months and is still alive at 107 months after entry on protocol. The second patient received only one cycle of therapy (at 180 mg m⁻²) due to disease progression and catheter obstruction. She died at 26 months following entry on protocol. The patient with peritoneal mesothelioma was enrolled at 180 mg m⁻², achieved a complete response with a duration of 10 months and died at 32 months following protocol entry.

Discussion

Platinum chemotherapeutic agents are the most active agents in the treatment of epithelial ovarian cancer. However, despite excellent response rates for initial treatment, most patients develop recurrent disease and the 5 year progression free survival rarely exceeds 15–20% [31]. Thus, most patients with advanced stage epithelial cancer will require salvage therapy. Many new agents and approaches have been explored, including dose intensive therapy with or without stem cell support, novel cytotoxic and targeted biologic therapies, in addition to IP therapy.

Three promising consecutive clinical trials evaluating IP therapy have been conducted by the GOG. The first randomized trial to demonstrate an advantage to IP cisplatin was the GOG#104 protocol (SWOG 8501) which randomized patients between intravenous cisplatin and intravenous cyclophosphamide vs IP cisplatin and intravenous cyclophosphamide. Patients were optimally debulked and stratified according to the amount of residual tumour <0.5 cm vs >0.5 to 2 cm. There was overall survival advantage in the IP group with an 8 month increase in median survival from 41 months in the intravenous group to 49 months in the IP group [9]. The second trial, GOG#114, confirmed the clinical usefulness of regional drug delivery as front-line ovarian cancer treatment. Patients in the control group of this trial received intravenous cisplatin (75 mg m⁻²) plus intravenous paclitaxel (135 mg m⁻² for 24 h), while patients in the experimental group were treated with two courses single-agent intravenous carboplatin prior to receiving IP cisplatin (100 mg m⁻²) and intravenous paclitaxel (135 mg m⁻² for 24 h). Patients in the experimental group showed a statistically significant improvement in progression-free survival (PFS) (28 vs 22 months; p = 0.01) and overall survival (63 vs 52 months; p = 0.05) [20]. The third randomized trial, GOG#172, was initiated to examine the IP delivery of paclitaxel in addition to cisplatin in ovarian cancer. Patients in the control group received intravenous cisplatin (75 mg m⁻²) plus intravenous paclitaxel (135 mg m⁻²), while the experimental group was treated with intravenous paclitaxel (135 mg m⁻² for 24 h day 1) prior to IP cisplatin (100 mg m⁻² day 2) and IP paclitaxel (60 mg m⁻² day 8). Preliminary results showed that the risk of recurrence was 28% lower compared to the intravenous arm [32]. The higher toxicity and catheter-related complications associated with IP therapy have prevented it from being incorporated into the clinical arena and established as the new standard in care for the initial treatment of optimally debulked advanced stage ovarian cancer.

Given the promising results of IP therapy in the front-line setting, one sought to evaluate this therapy for the treatment of recurrent and persistent ovarian cancer. Whole abdominal hyperthermia was combined with IP cisplatin to enhance the cytotoxic chemotherapeutic effect of cisplatin. In vitro studies have demonstrated that the addition of hyperthermia to cisplatin increased cell kill by ∼2 logs with clinically achievable concentrations of cisplatin [23]. Cisplatin has a relatively greater degree of thermopotentiation at 42°C than carboplatin or other platinum complexes [25]. In addition, in vitro studies have demonstrated that hyperthermia can sensitize cisplatin resistant cell lines [28].

Similar to the findings in the three GOG trials evaluating IP therapy that have all demonstrated a survival advantage, this study has also demonstrated strong clinical activity of IP cisplatin combined with hyperthermia in the recurrent and persistent disease settings. The majority of the patients included in this trial had platinum-resistant disease yet the response rate was over 40%, an extremely high response rate for this patient population that typically has response rates in the 10–20% range. Therefore, whole abdomen hyperthermia may augment chemosensitization or conversely reverse chemoresistance in patients with platinum-resistant disease. Furthermore, patients with disease <2 cm were found to have the best response with eight of 10 (80%) achieving a pathologic CR which translated to an improved survival. Overall, the findings indicate that IP cisplatin therapy combined with hyperthermia is clinically active in patients with small volume recurrent disease and even those with persistent refractory disease.

Alternatively, combined IP cisplatin and hyperthermia may have therapeutic potential for consolidation. Nearly half of patients with stage III, optimally debulked ovarian cancer who achieve a pathologic CR will relapse with a median interval of 24 months from second look laparotomy. The majority of these recurrences are contained within the peritoneal cavity and, therefore, regional IP therapy may provide a logical strategy for consolidation. Vergote et al. [33] evaluated IP cisplatin (90 mg m⁻²) vs observation for consolidative therapy of ovarian cancer patients with stage IIb and IIIc disease. Even though only 16 patients were included in the study there was a suggestive trend toward improved survival in the experimental IP arm. The overall survival and PFS with 95% C.I. are 0.89 (0.59–1.3) and 0.82 (0.52–1.3), respectively, in favour of the IP group. These results did not achieve statistical significance most likely due to the small sample size and the low number of events. In the future, the authors hope to use combination IP cisplatin and whole abdomen hyperthermia in patients with a negative second look laparotomy following standard up-front therapy for consolidation. Whole abdomen hyperthermia may play a role in this context as a means of augmenting chemosensitization with minimal treatment-related toxicity. In this context, IP platinum with hyperthermia will be used as a technique for consolidative thermochemotherapy.

In general, this study has found that the combination of whole abdomen hyperthermia with IP cisplatin is a reasonably well tolerated regimen. The most frequent treatment limiting complication related to the Tenkhoff catheter, whether it consisted of catheter obstruction, precluding thermometry measurements or infection. While meticulous sterile technique was used for thermometry placement, the more frequent instrumentation of the Tenckhoff catheter during thermometry placement may have provided an avenue for infection. Certainly the catheter complication rate of 22% is reasonable and consistent with the literature. Recently, Walker et al. [34] presented date regarding intra-pertioneal catheter outcomes in patients enrolled on GOG#172 and 42% (86/206) had catheter-related complications. These complications consisted of 11 obstructed catheters, 14 leaking catheters, five access problems and 22 infected ports. Some clinicians have recommended using the intravenous Port-a-caths which are associated with fewer catheter related complications.

Further exploration of a role for combined IP cisplatin and whole abdominal hyperthermia in the management of ovarian cancer is indicated. In future studies, the authors plan to evaluate the use of intravenous port-a-caths and assess the frequency of catheter-related complications as well as assess 3-dimensional hyperthermia dose distribution using non-invasive MRI-based thermometry techniques.

Acknowledgements

The authors thank Dr Gary Rosner for statistical input during protocol design. Supported in part by P01 CA42745.