Colorectal Cancer Metastases In The Reproductive Tract – A Systematic Review


 BACKGROUND: Colorectal cancer metastases are common and usually affect the liver, lungs, peritoneum or bone. However, unusual cases of metastatic disease within the reproductive tracts of both male and female patients have been described. This review aims to collate and summarise the available data to aid clinical decision making and patient counselling. METHODS: Pubmed, MEDLINE and the Cochrane databases were searched from inception till November 2020 for case reports and cases series describing original data on colorectal cancer metastases within the reproductive tract. RESULTS: 29,216 results were filtered down to 89 studies for inclusion. 43 and 46 related to the male and female reproductive tracts respectively, with individual patient outcomes reported for 48 males and 461 females. Genital metastases were associated more with rectal cancers whilst ovarian and testicular tumours more commonly derived from colonic primaries. Treatment was with curative intent for the minority of male patients (32%) with an associated poor prognosis (4% 5-year overall survival). Female reproductive tract metastases with ovarian lesions in particular were more likely to receive curative intent treatment (63%) and superior long-term survival (17% 5-year overall survival)CONCLUSIONS: Reproductive system colorectal cancer metastases are rare with limited cases reported. They are associated with advanced primary tumours and disseminated disease with an associated poor prognosis.


Introduction
Globally, colorectal cancer (CRC) is the third most common cancer for both men and women and the second most common cause of cancer-related death 1 . 20% of patient have metastases on diagnosis 2 with a further 20% of patients developing metachronous disease after initial surgery 3 . The most commonly affected sites are the liver (70%), thorax (32%), peritoneum (21%) and bone (12%) 4 . Despite progress in the treatment of metastatic CRC, prognosis remains poor. Relative 5-year survival rates with Stage IV disease are at most 12% 5 .
Treatment of metastatic disease is increasingly routine and effective. Surgical resection of solitary liver or lung metastases can lead to 5-year survival rates of 40% 6 and 50% 7 , respectively. However, metastatic disease at unusual sites is more poorly understood. There are published cases of cutaneous 8, 9 , centralnervous system 10,11 , gastro-intestinal 12,13 , urinary system 14,15 and skeletal muscle 16 metastases amongst others. Although a rare occurrence, the high prevalence of CRC and increasing survival amongst patients with advanced disease means it is not uncommon for the average colorectal cancer centre to see unusual patterns of disease. When this occurs, an understanding of the likely progression of the disease and the optimal treatment is vital for clinical decision making and patient counselling.
One potential site for metastases is the reproductive tract of both males and females. Ovarian metastases (OM) or 'Krukenberg' tumours are the most common of these with reported prevalence rates of 3-5% amongst patients with CRC 17 . These patients however, retain a poor prognosis in comparison to liver or lung metastases with a 5-year survival rate of 12% 17 . In addition to this, genital, cervical, uterine of scrotal metastases have all been described. The rarity of these cases means prognosis and optimal treatment for these patients is very poorly understood amongst clinicians.
Data on reproductive tract CRC metastases are largely con ned to individual case reports or within small case series. This review aims to collate and summarise these cases in order to aid clinician and patient understanding of these rare cases.

Methods
This systematic review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines 18 .

Search strategy
A systematic search was performed using Pubmed, MEDLINE and the Cochrane Library databases, from inception till September 2020. Results were ltered to include human studies only.
The following search algorithm was used: (Colorectal cancer metastasis OR rectal cancer metastasis OR colon cancer metastasis) AND (rare OR unusual OR atypical OR non-typical OR uncommon OR abnormal OR irregular OR isolated OR peculiar).

Eligibility criteria
Inclusion: All case series or case reports reporting original data on colorectal metastases in the male or female reproductive system Studies published on any date Exclusion:

Non-English language studies
Non-human subject studies

Study selection
One author (JA) manually screened each article identi ed by the initial search, using the study title and abstract in reference to the eligibility criteria. Screened articles were then included in a full text review to con rm nal eligibility. Full-text review and data extraction was completed independently by ve authors using a standardised tailored form.

Data extraction and synthesis
The following data was extracted from all included papers: site of primary tumour, site of metastasis, age, sex, ethnicity, signi cant past medical history, histology of primary tumour, initial staging, treatment of primary tumour, synchronous vs metachronous tumours, time interval from primary tumour and metastasis, mode of presentation, histology of metastasis, treatment of metastasis, length of follow-up, death due to disease progression.
Pooled rates have been calculated for outcomes with common denominators.

Selected studies
A total of 29,216 records were identi ed from the initial database searches. Initial screening excluded 29,092 irrelevant records. 7 non-English language records and 20 abstract only records were excluded. 8 studies did not meet the eligibility criteria on full text review leaving a total of 87 studies eligible for inclusion in this study. 42 relate to the female genital tract  , 44 relate to the male genital tract 63-106 and one study reports on both 51 .

Study characteristics
71 studies are case reports and 16 are case series. There is a broad publication date range from 1950-2020. The characteristics of studies involving the male and female reproductive systems are summarised Table 1 and Table 2 respectively. The total number of individual patient data reports are 47 and 461, respectively.

Presentation
The mode of presentation was not reported for 407/509 (80%) patients. In studies which did report this it was possible to describe the mode of presentation as either 'symptomatic', 'incidental' or from a formal 'surveillance' programme. A summary of presentation mode is found in Table 3.
For females. the majority (92%) of ovarian lesions do not have mode of presentation speci ed. However, in vulvar, cervical, salpingeal and uterine metastases, the was a trend towards a symptomatic presentation.
Vaginal metastases presented with vaginal bleeding or pelvic pain, vulvar metastases as palpable super cial lesions or with bleeding/discharge. Symptomatic uterine and cervical metastases all presented with vaginal bleeding. The one patient with salpingeal metastases presented with watery vaginal discharge.
Reporting was more complete for males and again demonstrated a trend towards symptomatic presentations. Penile metastases presented as painless or painful super cial lesions and scrotal secondaries displayed ulcerating skin lesions. Testicular and spermatic cord metastases tended to be painful with associated palpable masses. All studies report the presentation of metastases as either synchronous or metachronous. Synchronous is used to describe metastases which are diagnosed at the same time as (or in some papers within 6 months of) the diagnosis of the original tumour. Metachronous is used to describe recurrence of previously treated disease.
Mode of presentation was reported on for >99% (504/508) of included patients. Table 4 summarises this outcome for each site of metastasis and demonstrates a largely comparable distribution between synchronous or metachronous presentations. All studies that reported the histopathological ndings from the metastases were consistent with spread from the primary colorectal cancer.

Primary tumour location
The location of the primary colorectal tumour was speci ed in 344/461 (75%) and 44/47 (94%) of the male and female patients, respectively. See Table 4.
Patients with ovarian metastases demonstrated a variety of primary tumour locations with no demonstrable trend towards any one location. 34% of ovarian metastases were associated with a primary tumour which had arisen in the sigmoid or rectum. This is in contrast to lower genital tract (cervix, vagina, vulva) lesions in females which demonstrate a trend towards rectosigmoid primary lesions (50%, 100%, 86% respectively). Similarly, in men, the rectosigmoid is the most common location for a primary tumour in all metastatic sites (penis 100%, scrotum 78%, prostate 66%) with the possible exception of spermatic cord (50%) and testicular metastases (50%).   For individual sites of metastasis this relationship is largely maintained with the possible exception of penile metastasis where the rate of T1-4N0M0 primary cancers is 47% (8/17) Cervix 0 n/a n/a n/a 4 0 3 1 3 0 1 Fallopian tube 0 n/a n/a n/a 1 0 0 1 0 0 1 Uterus 0 n/a n/a n/a For these patients, the rates of palliative intent treatment were 74% (17/22), 100% (6/6), 57% (4/7), 33% (2/6) and 33% (1/3) for penile, scrotal, testicular, spermatic cord and prostatic metastases respectively and 68% for all male patients.
For male reproductive tract metastases prognosis was very poor. 80% (24/30) of patients were reported as having died, with 23 of these patients dying within the rst year after diagnosis of metastatic disease and the remaining patient dying at 18 months. Of the 6 living patients just one was con rmed to be alive at 5 years after diagnosis which was a patient with a penile metastasis reported in 1950. 5 further patients were reported as alive at time of publication but follow-up time was limited to 6, 12, 16, 18 and 30 months.
In female patients, 66% (162/245) were reported as having died within 5 years with 2 further deaths occurring at unspeci ed times. However, 41 (17%) patients were reported as alive at or beyond 5 years and a further 40 patients were alive with a median follow-up time of 24 (1-54) months. Of the patients alive at 5 years, all were patients with ovarian metastases and had received surgical treatment.

Discussion
In this review, the ovaries were the most frequently encountered site for metastases and demonstrated the largest patient population. Ovarian metastases (OM) are reportedly found in 3-5% 17 of patients with colorectal cancers and as such represent the most common site of spread explored in this review. Thus, they are likely the best understood. They are found largely in pre-menopausal women with advanced colonic cancers 107, 108 as either synchronous or metachronous presentations. The optimal treatment remains controversial. One study examines the role of prophylactic oophorectomy at the time initial cancer resection 109 . However, the negative effects of oophorectomy in pre-menopausal women including hormonal, metabolic and psychological consequences and inferior long-term overall survival are now well understood 110 and prophylactic surgery is not routinely practiced. Oopherectomy is indicated in con rmed metastases and has been reported to have reasonable outcomes with 5 year and 10 year overall survival rates of 14% and 5% respectively 17 . This review con rms OMs are associated with the best prognosis of reproductive tract metastases and the suitability for potential curative surgery likely accounts for this nding.
In contrast to this, all sites of metastatic disease in men are associated with a dismal prognosis with just one patient (2%) surviving for 5 years. Although poorly reported in the source material, these patients tended to have advanced disease on presentation and as such were largely treated with palliative intent. In those patients who did have potentially curative treatment, they commonly had metastatic disease elsewhere and succumbed rapidly to their disease.
With the exception of gonadal disease, reproductive tract metastases are more likely to arise from rectal than colonic primaries. Several authors suggest various methods of spread for genital metastases including implantation during surgery, retrograde spread to the pudendal venous system or from obstructed inguinal or hypogastric lymphatic channels 47,56,63 . The tendency for ovarian and possibly testicular metastases to arise from colonic primaries suggests an alternative dissemination pathway such as transcoelomic spread.
The rarity of these cases means the source material for this review is limited to case reports and cases series. As such, the review is descriptive only and no comparative analysis is possible. Furthermore, such reports are inherently subject to signi cant selection bias and cannot be considered to be representative of CRC reproductive tract metastases as a whole.
However, this review is the largest and most up-to-date summary of these unusual cases. Undoubtedly, patients with CRC metastases to the reproductive system have a very poor prognosis and often suffer from advanced and disseminated malignancy. However, there is evidence of long-term survival in selected patients after surgical treatment. Perhaps one way to learn more about these rare occurrences with worse prognosis is through establishing registries that include such patients. This would provide a hub for advancing clinical care and research into this area and further our understanding of these rarities. Proteomics, genomics and metabolomics of these unusual metastases would shed more light on pathogenesis and perhaps, help us to improve outcomes of such patients in future.