Since May 2006, trastuzumab (HERCEPTIN) has been licensed for the adjuvant treatment of HER2-positive breast cancer after surgery, chemotherapy and possibly radiotherapy (1). As yet, only interim results are available from the four important randomised studies of the antibody, which is being investigated in different regimes. The rate of disease-free survival is increased by 5.5% to 7.5% with trastuzumab at one to three years of follow-up (2-4). The overall mortality is recorded in three of the four studies only as a secondary end point (5). It is 0.5% to 2.9% lower with trastuzumab. So far, a significant reduction has been found only in the pooled analysis of two studies (3), where the reliability is doubtful because of methodological defects (a-t 2005; 36: 96-8), and in a study available only as an abstract and therefore not assessable (4).

Two-year results of the HERA* study (2) have now been published (6): more than 5,000 women were given trastuzumab (6 mg/kg body weight i.v. every three weeks) for one or two years following surgery and completion of adjuvant chemotherapy or were only observed. So far, only the arm with one year of the treatment has been analysed. On randomisation, 84% of the patients were younger than 60 years, every second woman had positive hormone receptor status and in every third patient the axilla was free from involved lymph nodes. Patients with impaired left ventricular function or signs of manifest heart failure after the predominantly (94%) anthracycline-containing adjuvant chemotherapy were excluded (2, 6).

After a median study duration of 23.5 months, 218 (12.8%) of 1,703 patients in the trastuzumab group and 321 (18.9%) of 1,698 in the observation group suffered a disease recurrence, defined as recurrence of the primary tumour, new diagnosis of a breast cancer, another malignancy or death without evidence of tumour. The reduction is significant (hazard ratio [HR] 0.64; 95% confidence interval [CI] 0.54-0.76; number needed to treat [NNT] = 16) and is due particularly to a lower rate of distant metastases (9% vs. 14%; NNT = 20). The recurrences were diminished only in the first 12 to 18 months. Later, the annual rates become largely similar. In the trastuzumab group, 59 women died (3.5%), which was significantly fewer than in the observation group with 90 deaths (5.3%, NNT = 56). For three-year follow-up, a survival advantage of 2.7% is projected, though data over three years are available for fewer than 10% of the patients (6).

190 patients suffered major complications on trastuzumab, but only 88 in the observation group (11% vs. 6%; number needed to harm [NNH]) = 20). 115 women stopped the trastuzumab therapy early for safety reasons, 72 (4%) because of cardiac problems. With the antibody, there is a marked increase in symptomatic heart failure (2.1% vs. 0.1%; NNH = 50) and asymptomatic impairment of left ventricular function (3.0% vs. 0.5%; NNH = 40) (6).

The study has several methodological problems. After publication of the one-year results (2), the patients of the observation group were offered the option of switching to one of the trastuzumab groups. Up to the two-year analysis, 51% of the women availed of this offer (6). Because of this, positive effects in the trastuzumab group could be underestimated and distorting effects could be overestimated in the observation group. Concrete data on the patients were not reported.

58 patients (3.4%) on trastuzumab and 97 (5.7%) in the observation group were reported as "lost to follow-up". Taking into account these rates of loss, the results on overall survival would not stand up to "worst case" analysis**. However, according to a graph in the article, there appear not to be data on either disease recurrences or deaths for more than 150 women in each group. Because of incomplete follow-up observation of the patients, the reliability of the results is considerably impaired. In a "worst case" analysis a reduction in disease recurrences would also no longer be detectable. It is known from systematic investigations that an overestimation of effect can occur with incomplete follow-up (7).

The cardiotoxicity of the antibody is disturbing and not assessed for periods of more than three years. It is claimed to be usually reversible with immediate discontinuation of trastuzumab and symptomatic treatment of heart failure (8). However, it should be borne in mind that in HERA - as in other studies of adjuvant therapy - over 80% of the patients are younger than 60 years of age. Breast cancer occurs mainly after the age of 60 years in Germany (9), and age is a known risk factor for heart failure with trastuzumab (10). For over-60 year-old patients there is no proof of benefit and only inadequate safety data. The applicability of the study results is doubtful also because the percentage of women remains uncertain who had no evidence of impaired cardiac function after anthracycline-containing chemotherapy and who could therefore be selected for inclusion.

The British National Institute for Clinical Excellence (NICE) gave a positive assessment last year of trastuzumab for the adjuvant therapy of HER2-positive breast cancer, even taking economic aspects into account (11). In contrast, the Pharmaceutical Management Agency (PHARMAC), which is part of the New Zealand health ministry, considers that the benefit of trastuzumab is insufficiently proven even after publication of the recent HERA results (12, 13). It wants to support further investigation of a Finnish regime on the adjuvant use of trastuzumab (14), which has so far been tested only in a small study but uses only one fifth of the total trastuzumab dose with a similar reduction in the recurrence rate as in the larger studies (15).

	In the HERA study, one year of treatment with trastuzumab (HERCEPTIN) following adjuvant chemotherapy in women with HER2-positive breast cancer reduces the rate of disease recurrences from 19% to 13% after two years.
	The overall mortality is recorded only as a secondary endpoint. It is said to be 1.8% lower with trastuzumab.
	The high rate of loss in patient follow-up markedly diminishes the validity of the data.
	There are inadequate data on patients over the age of 60 years.
	With trastuzumab there is a relevant degree of impairment of cardiac function. The long-term consequences cannot be assessed with certainty at present.
	In our opinion, a benefit of trastuzumab in adjuvant therapy is inadequately proven.
	If use is considered in an individual case, the antibody should be used according to the HERA regime. Strict monitoring of cardiac function before and during the therapy is absolutely essential.

(R = randomised study, M = metaanalysis)

1

Roche: official prescribing information HERCEPTIN, Stand Aug. 2006

2

PICCART-GEBHART, M.J. et al.: N. Engl. J. Med. 2005; 353: 1659-72

3

ROMOND, E.H. et al.: N. Engl. J. Med. 2005; 353: 1673-84

4

CHUSTECKA, Z.: Anthracyclines may not be necessary in adjuvant therapy of breast cancer; http://www.medscape.com/viewarticle/549502

5

TAN, A.R., SWAIN, S.M.: Semin. Oncol. 2003; 30: 54-64

6

SMITH, I. et al.: Lancet 2007; 369: 29-36

7

TIERNEY, J.F., STEWART, L.A.: Int. J. Epidemiol. 2005; 34: 79-87

8

EWER, M.S. et a.: J. Clin. Oncol. 2005; 23: 7820-6

9

Gesellschaft der epidemiologischen Krebsregister in Deutschland e.V.: Krebs in Deutschland. Häufigkeiten und Trends [Epidemiological cancer registry society in Germany. Incidence and trends], 5th edition, Saarbrücken 2006, available at: http://www.gekid.de

10

GONZALEZ-ANGULO, A.M. et al.: Oncologist 2006; 11: 857-67

11

NICE: TA 107 Breast cancer (early)-trastuzumab: Guidance; http://www.nice.org.uk/guidance/TA107/Guidance/pdf/English

12

PHARMAC: Herceptin not funded, under continuing review; http://www.pharmac.govt.nz/pdf/280706.pdf

13

PHARMAC: Press release of 19th Jan. 2007; http://www.pharmac.govt.nz/pdf/190107.pdf

14

HIND, D. et al.: Lancet 2007; 369: 3-5

15

JOENSUU, H. et al.: N. Engl. J. Med. 2006; 354: 809-20

HERA study = Herceptin Adjuvant Study

Worst case analysis: "lost" patients in the trastuzumab group dead, "lost" patients in the observation group alive.