Background

In the United States, lung cancer has been the leading cause of cancer deaths in both men and women. It’s surpassed both breast cancer and colon cancer [1]. Despite the progress made in the last years in the diagnosis and treatment of this disease, men and women die each year from lung cancer. In recent years, lung cancer mortality rates in women have dramatically increased in developed countries, contrasting with a levelling off or decrease among men [2]. In women, this dramatic increase can be explained by multiple factors. The primary cause remains tobacco, which has rapidly gained social acceptance among women after World War II [3]. In contrast, the decrease of cancer mortality in men is probably attributable to the decrease in smoking rates that began earlier in men than in women [3].

According to the Moroccan registry of cancer of Rabat of 2005, lung cancer is the leading cancer in men and comes at the eleventh place in women [ 4 ]. In Morocco, a study conducted by Lalla Salma’s foundation: “Prevention and Treatment of Cancer”, showed that over 15 years the prevalence of active smoking among elderly Moroccans was estimated at 18 %, 31.5 % of which were men and 3.3 % were women. Moreover, 41 % of the population was exposed to passive smoking [5].

Purely descriptive epidemiological data on primary lung cancer in women are scarce in Morocco. The aim of this study, conducted in the National Institute of Oncology in Rabat, was to describe the epidemiological profiling, in particular the smoking status, to determine the most frequent type of lung cancer, and to analyse the survival of Moroccan women with lung cancer diagnosis.

Methods

A retrospective study was performed enrolling 1680 cases of lung cancer, 101 (6 %) were women diagnosed between the period of January 2004 and December 2008 at the National Institute of Oncology. All these incident cases were within the respective time period.

The study respected the ethical rules of medical research involving human subjects as stipulated by the World Medical Association in the Declaration of Helsinki. The local ethical committee of the National Institute of Oncology of Rabat also approved this study; with the consentment of patients or their families.

The hospital charts, including medical history, were reviewed and the following information was collected: age at diagnosis, smoking status, histology, stage, presenting symptoms, investigations, treatment (surgical resection, radiotherapy, chemotherapy, or no treatment) and survival.

All diagnosis were proved by anatomopathologic analysis of biopsy specimen. Non small cell lung cancer was divided into adenocarcinoma, squamous cell carcinoma and large cell carcinoma.

The descriptive analysis was made by Statistical Package for the Social Science version 13.0 (SPSS). The progression free survival (PFS) was defined as the time from the start of treatment to the first documented disease progression or death from any cause. For PFS, Kaplan–Meier curves were constructed and compared by using a long-rank test. Descriptive statistics with 95 % of confidence interval were calculated according to standard procedure.

Results

We found 101 women among 1680 cases of lung cancer (versus 1579 men). The mean age was 55.68 years (standard deviation was 13.06).

The percentage of never-smokers was 75 %, 14 % of patients had an environmental tobacco smoke exposure (the husband or another person in the family who smoke) and 11 % reported personal smoking (Table 1). The most frequent mode of presentation was chest pain (60 %), cough (54 %), dyspnea (38 %) and hemoptysis (27 %). Proportion of non small cell lung cancer at stage IV was higher (82 %), 16 % of cases were at stage III and only 2 % were classified as stage II (Table 1).

Table 1 Patient characteristics
Full size table

The proportion of adenocarcinoma among never and passive smokers was higher than that of squamous cell carcinoma (69.4 versus 30.6 %), while among women who were smokers, the most frequent histological type was squamous cell carcinoma (63.6 %).

The median follow-up was 2 months (1–24 months). The histological diagnosis was based on biopsy during bronchoscopy (80 %), as well as biopsy guided by CT scan (16 %) or thoracotomy (4 %).

Palliative chemotherapy was planned in 43 % of cases, 7 % were treated by chemoradiation, 8 % by radiotherapy, 2 % received a neoadjuvant chemotherapy before chemoradiation and 2 % were treated by surgery and radiotherapy. The best supportive care were proposed in 34 % of cases and 4 % of patients received a whole brain irradiation (Table 1).

For follow-up cases (88 %), the survival rates of cases with no small cell lung cancer at 6 months were at 27.7 %. Cases with squamous cell carcinoma had higher survival rates than the ones with adenocarcinoma (44.4 versus 25 %; p = 0.01), but we had no significant difference in progression free survival at 6 months between squamous cell carcinoma and adenocarcinoma (20 versus 25 %; p = 0.81) (Figs. 1, 2). In cases with small cell lung cancer, the survival rate at 6 months was 0 %. 13 cases were lost to follow-up.

Fig. 1
figure 1

Time to death, survival curve in terms of histological type

Full size image
Fig. 2
figure 2

Time to progression, survival curve in terms of histological type

Full size image

The Cox regression analysis showed that smoking and passive smoking were not significantly associated with survival [HR: 0.62 (95 % CI 0.31, 1.30); p = 0.19] [HR: 0.56 (95 % CI 0.29, 1.08); p = 0.08] respectively.

Adenocarcinoma was significantly associated with poorer survival [HR: 1.73 (95 % CI 1.05, 2.85); p = 0.03], brain metastasis was not significantly associated with survival [HR: 1.01 (95 % CI 0.53, 1.93); p = 0.97].

Subgroup analysis (85 patients who were never or passive smokers) showed that there was no significant difference in survival at 6 months between squamous cell carcinoma and adenocarcinoma (35 versus 20 %; p = 0.08) (Table 2).

Table 2 Subgroups analysis
Full size table

90.9 % (10 patients) of active smoking patients died, and one patient (accounting for 9.09 %) was lost to follow. 85 % (12 patients) of passive smoking patients died and 14.28 % (2 patients) were lost to follow.

Discussion

The mean age was 55.68 years (standard deviation was 13.06). Our patients developed the disease at an earlier age than other women in other studies [ 6, 7 ].

In our study, the percentage of never-smokers was 75 %. 14 % of patients had an environmental tobacco smoke exposure (the husband or another person in the family who smoke) and 11 % were reported personal smoking. In the majority of American and European reviews, most of the patients were current or ex-smokers [7–9]. Recently, approximately 23.1 % of men and 18.3 % of women in the United States have been cigarette smokers so far [10].

Tobacco does not explain all the women cases with lung cancer, because in this study, 75 % of patients were never smokers. Multiple factors such as environmental exposures, genetic mutations, hormonal factors, and some infections have also been implicated in the development of lung cancer in women who have never smoked [1]. A possible role of circulating steroid hormones particularly of oestrogens, in the etiology of lung cancer has been hypothesized [11]. It should be noted that oestrogen receptors have been detected on cancerous lung cells and estrogens induce the differentiation and maturation of the foetal lung [11–13]. Different studies have suggested that there is a potential protective effect of increased or prolonged estrogen exposure, others have found the opposite [14, 15].

Furthermore, a large meta-analysis reported a 24.5 % worldwide incidence of HPV in lung cancer, particularly in China [16] and recently, many Asian studies indicate that longer telomere length may increase lung cancer risk [17].

Studies evaluating gene environment interactions may provide important insights into carcinogenesis pathways of lung cancer in never smokers.

In European and American series, most cases of non-small cell lung cancer were diagnosed at a relatively advanced stage according to the TNM classification system [6]. In our population, proportion of non small cell lung cancer at stage IV was higher (82 %), 16 % of cases were at stage III and only 2 % were at stage II. This high percentage can be explained by the delay of diagnosis.

At the end of 1960s, changes in the frequency of different histological subtypes of lung cancer have been observed worldwide, with an increasing proportion of adenocarcinoma and a declining proportion of squamous cell carcinoma [18]. In our study, the proportion of adenocarcinoma among never and passive smokers was higher than that of squamous cell carcinoma (69.4 versus 30.6 %), while among women who were smokers, the most frequent histological type was squamous cell carcinoma (63.6 %). These results are almost similar to those reported in other studies. Adenocarcinoma represented 68.1 % in the Mayo Clinic’ study, 71 % in the Bennet’s series and 43.4 % in the Grivaux study [6, 9, 19].

Moreover, women’s 5-year relative survival rates have improved across nearly all stages with similar histologies, the benefit was more important in those with adenocarcinoma [20, 21]. In this study, adenocarcinoma was significantly associated with poorer survival [HR: 1.73 (95 % CI 1.05, 2.85); p = 0.03], while in a Radzikowska study published in Annals of Oncology 2002, the squamous cell carcinoma was a negative predictor of survival in univariate analysis (p = 0.008), but this result was no significant in multivariate analysis (p = 0.29) [7].

Lung cancer survival rates for non smokers have improved compared with those of smokers [22]. In this series, the Cox regression analysis showed that smoking and passive smoking were not significantly associated with survival [HR: 0.62 (95 % CI 0.31, 1.30); p = 0.19] [HR: 0.56 (95 % CI 0.29, 1.08); p = 0.08] respectively. Even though HR are neither statistically significant for smoking nor passive smoking, HR estimates seem to show a protective effect for tobacco smoking, which is highly unlikely. This may indicate some bias or confounding within the study population. However, the data does not allow additional analysis to understand these estimates.

This study had potential limitations, one of which was the retrospective nature of the analysis. Definitely, a case control study with healthy controls would have been methodologically more powerful.

Conclusions

Many Moroccan women affected by lung cancer have never smoked (75 %). The relationships between histological type and tobacco found in our series are in agreement with those reported in the literature: adenocarcinoma appears to be the most frequent cell type among never and passive smokers. Furthermore, adenocarcinoma is significantly associated with poorer survival.

Smoking does not explain all cases of lung cancer. Environmental exposures, genetic predisposition, hormonal factors, and viral infection may all play a role in lung cancer in women.