What would explain the sudden growth of lung cancer after minor trauma reported by El Saghir et al?
Michael Retsky, Children's Hospital and Harvard Medical School
10 August 2005
This is a case report in which a 43 year old male with a smoking history was diagnosed with inoperable stage IIIA or IIIB poorly differentiated non-small cell lung cancer [1]. The prognosis for this stage is very poor with 5-year survival of approximately 10% [2]. Treatment was chemotherapy and radiation. While relapse was inevitable, the patient was seemingly doing well 15 months after primary treatment. The patient then incurred a minor trauma to the right temporal skull bone; within a month, the patient reported a rapidly growing 7 cm tumor at the place of the trauma. Imaging revealed tumor had penetrated the skull with meningeal invasion, and while some compression was apparent, it did not yet invade the brain. (Lung cancer most often relapses to brain or adrenal glands but can recur virtually anywhere including the skeleton.) The patient died 15 days later of massive hemoptysis.
El Saghir et al suggest that it is not a coincidence that both the growth occurred at the location of the trauma and that the growth started shortly after the trauma. The mechanisms suggested include some type of trauma-induced angiogenesis of a dormant micrometastasis or dormant malignant cells that happened to be at that place, or perhaps, there were no cancer cells at the cite, but circulating cancer cells were entrapped there due to the trauma.
My colleagues and I have studied tumor growth in breast cancer [3,4]. We have reported that there is an early relapse mode 6-9 months after mastectomy for approximately 20% of premenopausal women with node-positive breast cancer. We proposed this is the result of surgery-induced angiogenesis of dormant distant avascular micrometastases. This may explain why adjuvant chemotherapy is particularly effective for that patient category and also why there is no early mortality benefit of mammography screening for women age 40-49. Since survival after relapse is 2 years, surgery-induced angiogenesis could quantitatively explain a possible excess mortality for the screened group 3 years after the start of screening trials for young women. It also suggests why breast cancer in young women is often termed “aggressive”. Can we explain the El Saghir et al case report of sudden growth in lung cancer after minor trauma with a similar mechanism as suggested by the authors?
We can analyze these data and draw some conclusions about what is possible and what is not. Since there were no biopsies taken of the growth, we do not know with certainty that it is metastatic lung cancer - although there is no reason to doubt that. Assuming a 1 mm dormant micrometastasis grew to 7 cm in one month exponentially, the doubling time would be a constant 2 days. As El Saghir et al report, this is indeed rapid growth for human cancer. That is approximately the growth rate of a multipassaged animal model tumor. Human cancer data are typically 2 to 5 month doubling time plus occasional dormancy but there are reports of growth with doubling time as short as 8 days after intervention [5]. It is difficult to accept that the tumor was the result of a 1 mm dormant micrometastasis suddenly undergoing angiogenesis and growing to 7 cm in 1 month. We can also rule out the possibility of a single dormant cell becoming a 7 cm diameter tumor in one month since doubling time would be less than one day - which is impossibly small.
What can we conclude from this information? The key aspect of this case is the fact that tumor grew at the site of trauma and started at that time; no other tumors grew at any other place or started at any other time. Also it is unlikely that the new growth is the result of sudden activation of cancer from some known dormant state previously at the trauma site. We need another explanation.
This is somewhat similar to the experiments of Martins-Green et al., whose studies may provide some insight [6]. This is a particular avian system in which a virus is the carcinogenic agent. Other than those two major differences, as we will see, there is a certain similarity to the case presented by El Saghir et al. When newly hatched chicks are given injections of Rous sarcoma virus, a tumor develops at the site of injection. In spite of the presence of the virus in the blood, no other tumors are found distant from the site of inoculation during the life span of the animal (4-6 weeks). However, if a wound is made away from the primary tumor, a tumor develops at the site of wounding.
Martins-Green et al report that these wound tumors do not develop as a result of metastasis. Rather, factors released upon wounding contribute to the development of the wound tumors. In particular, they showed that TGF¿, acidic FGF, and basic FGF, growth factors implicated in wound healing, can themselves produce tumor development in this system. Other growth factors (EGF and TGF¿), that also have roles in wound healing, do not induce tumors. They reported that wound tumor development correlates with inflammation and during the inflammatory response, blood vessel leakage occurs as tested by the release of fibrinogen into the tissues. The three factors (TGF¿, aFGF, bFGF) which promote tumors also induce inflammation, whereas the two factors (EGF and TGF¿) do not promote tumors or induce inflammation.
To test the possibility that inflammation is the key element in the development of these wound tumors, they used beta-methylprednisolone, a drug that inhibits inflammation (including blood vessel leakage), to determine if wound tumor development could be prevented. Martins-Green et al found that when inflammation was inhibited, tumors were also inhibited; when inflammation could not be stopped, tumors developed as before. These results indicate that the effect of wounding on the development of wound tumors in Rous sarcoma virus-infected chicks is accomplished through the cytokines released by the inflammatory cells at the site of wounding. Thus, at least in this avian virus system, inflammation plays a critical role in providing the conducive environment for oncogene integration and activation, and subsequent development of tumors.
Exploring a relation of the Martins-Green et al experiment to the El Saghir et al case, oncogene activation is not relevant but it is certainly easy to imagine that the head trauma caused inflammation. The authors do not state what type of car the patient was in or what precisely was the object of impact so we cannot know the exact surface area of trauma. However, since the tumor was on a relatively flat part of the patient's head and cars do not have sharp areas (in order to prevent puncture injuries), we can assume the area of impact was at least several square cm. It would follow that the area of inflammation was likewise at least several square cm. There would be ample cancer cells in the blood to populate the entire area of inflammation via leakage since the patient obviously had significant existing tumor burden. That would explain the rapid growth since the tumor would grow from extensive seeding over a few cm2 surface together with the availability of growth factors.
Supporting this notion, Balkwill et al writes that if genetic damage is the “match that lights the fire” of cancer, then inflammation is the “fuel that feeds the flames” [7]. Also that inflammation may play a role in survival and proliferation of already initiated cancer cells [8].
We should also consider a few alternative hypotheses. For one, it may be that we are in error in estimating the neoplastic component of the tumor and that growth of a dormant micrometastasis might account for the sudden tumor. This can be rejected since it is very doubtful that even a factor of two error would make much difference. For a second alternative hypothesis, skull bones and associated marrow have sufficient vasculature to support many dormant micrometastases that perhaps could have been collectively activated into sudden angiogenesis with the trauma. This hypothesis would require a coincidence that the impact happened to occur where there were dormant micrometastases or the entire skull was laced with dormant micrometastases – both of which are unlikely.
There have been anecdotal reports for over a century of tumor growth after surgery or wounding. Angiogenic processes induced after surgery could reasonably explain those reports. The report by El Saghir et al is unusual in that there is no wound to repair and the growth is too rapid to be explained by activation of angiogenesis of a dormant micrometastasis. Their suggestion that circulating cancer cells were entrapped at the trauma site is probably close to the truth.
I agree that this case calls for research to understand the mechanism of this type of cancer inducement. The unusual isolated and exaggerated situation allowed El Saghir et al to observe what may be a new and possibly important hematologic metastatic pathway: inflammation as a facilitating precursor to tumor. Metastasis is a very inefficient process [9]. Millions of cancer cells might be found in a patient’s blood but only a few metastases occur. The inflammation sequence discussed by Martins-Green et al would certainly increase metastatic efficiency since it bypasses extravasation through an intact vessel wall and it provides growth factors in the microenvironment [10]. In the context of Paget’s 1889 “seed and soil” metaphor, we have a situation here where many viable seeds are planted in a large vacant field and good fertilizer is applied.
The possible connection between inflammation and tumor growth might also help explain why anti-inflammatory drugs such as aspirin, prednisone and Celebrex are often used in cancer prevention and therapy.
My thanks to Tatura Udagawa for reminding me of the Martins-Green et al experiment and Romano Demicheli for useful discussions. Thanks also to El Saghir et al and the patient's widow for presenting this valuable information. This effect is not amenable to analysis from clinical trial data so anecdotal reports of unusual events are essential and encouraged.
References:
1. Nagi S El Saghir, Ihab I Elhajj, Fady B Geara, Mukbil H Hourani. Trauma-associated growth of suspected dormant micrometastasis. BMC Cancer 2005, 5:94 doi:10.1186/1471-2407-5-94.
2. Beadsmoore CJ, Screaton NJ. Classification, staging and prognosis of lung cancer, Eur J Radiol. 2003 Jan;45(1):8-17.
3. M Retsky, G Bonadonna, R Demicheli, J Folkman, W Hrushesky, P Valagussa. Hypothesis: Induced angiogenesis after surgery in premenopausal node-positive breast cancer patients is a major underlying reason why adjuvant chemotherapy works particularly well for those patients. Breast-Cancer-Research, 2004, 6:R372-R374 (14 May 2004).
4. R Demicheli, G Bonadonna, WJM Hrushesky, MW Retsky. P Valagussa. Menopausal status dependence of the timing of breast cancer recurrence following primary tumour surgical removal. Breast Cancer Research2004, 6:R689-R696 . http://breast-cancer-research.com/content/6/6/R689
5. El Sharouni SY, Kal HB, Battermann JJ. Accelerated regrowth of non-small-cell lung tumours after induction chemotherapy. Br J Cancer. 2003 Dec 15;89(12):2184-9.
6. Martins-Green M, Boudreau N, Bissell MJ. Inflammation is responsible for the development of wound-induced tumors in chickens infected with Rous sarcoma virus. Cancer Res. 1994 Aug 15;54(16):4334-41.
7. Balkwill F, Mantovani A, INFLAMMATION AND CANCER: BACK TO VIRCHOW? Lancet, 0099-5355, February 17, 2001, Vol. 357, Issue 9255
9. Chambers AF, Groom AC, MacDonald IC. Dissemination and growth of cancer cells in metastatic sites. Nat Rev Cancer. 2002 Aug;2(8):563-72.
10. Fidler IJ. The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited. Nat Rev Cancer. 2003 Jun;3(6):453-8.
Competing interests
None
What would explain the sudden growth of lung cancer after minor trauma reported by El Saghir et al?
Michael Retsky, Children's Hospital/Harvard Medical School
17 August 2005
This is a case report in which a 43 year old male with a smoking history was diagnosed with inoperable stage IIIA or IIIB poorly differentiated non-small cell lung cancer [1]. The prognosis for this stage is very poor with 5-year survival of approximately 10% [2]. Treatment was chemotherapy and radiation. While relapse was inevitable, the patient was seemingly doing well 15 months after primary treatment. The patient then incurred a minor trauma to the right temporal skull bone; within a month, the patient reported a rapidly growing 7 cm tumor at the place of the trauma. Imaging revealed tumor had penetrated the skull with meningeal invasion, and while some compression was apparent, it did not yet invade the brain. (Lung cancer most often relapses to brain or adrenal glands but can recur virtually anywhere including the skeleton.) The patient died 15 days later of massive hemoptysis.
El Saghir et al suggest that it is not a coincidence that both the growth occurred at the location of the trauma and that the growth started shortly after the trauma. The mechanisms suggested include some type of trauma-induced angiogenesis of a dormant micrometastasis or dormant malignant cells that happened to be at that place, or perhaps, there were no cancer cells at the cite, but circulating cancer cells were entrapped there due to the trauma.
My colleagues and I have studied tumor growth in breast cancer [3,4]. We have reported that there is an early relapse mode 6-9 months after mastectomy for approximately 20% of premenopausal women with node-positive breast cancer. We proposed this is the result of surgery-induced angiogenesis of dormant distant avascular micrometastases. This may explain why adjuvant chemotherapy is particularly effective for that patient category and also why there is no early mortality benefit of mammography screening for women age 40-49. Since survival after relapse is 2 years, surgery-induced angiogenesis could quantitatively explain a possible excess mortality for the screened group 3 years after the start of screening trials for young women. It also suggests why breast cancer in young women is often termed “aggressive”. Can we explain the El Saghir et al case report of sudden growth in lung cancer after minor trauma with a similar mechanism as suggested by the authors?
We can analyze these data and draw some conclusions about what is possible and what is not. Since there were no biopsies taken of the growth, we do not know with certainty that it is metastatic lung cancer - although there is no reason to doubt that. Assuming a 1 mm dormant micrometastasis grew to 7 cm in one month exponentially, the doubling time would be a constant 2 days. As El Saghir et al report, this is indeed rapid growth for human cancer. That is approximately the growth rate of a multipassaged animal model tumor. Human cancer data are typically 2 to 5 month doubling time plus occasional dormancy but there are reports of growth with doubling time as short as 8 days after intervention [5]. It is difficult to accept that the tumor was the result of a 1 mm dormant micrometastasis suddenly undergoing angiogenesis and growing to 7 cm in 1 month. We can also rule out the possibility of a single dormant cell becoming a 7 cm diameter tumor in one month since doubling time would be less than one day - which is impossibly small.
What can we conclude from this information? The key aspect of this case is the fact that tumor grew at the site of trauma and started at that time; no other tumors grew at any other place or started at any other time. Also it is unlikely that the new growth is the result of sudden activation of cancer from some known dormant state previously at the trauma site. We need another explanation.
This is somewhat similar to the experiments of Martins-Green et al., whose studies may provide some insight [6]. This is a particular avian system in which a virus is the carcinogenic agent. Other than those two major differences, as we will see, there is a certain similarity to the case presented by El Saghir et al. When newly hatched chicks are given injections of Rous sarcoma virus, a tumor develops at the site of injection. In spite of the presence of the virus in the blood, no other tumors are found distant from the site of inoculation during the life span of the animal (4-6 weeks). However, if a wound is made away from the primary tumor, a tumor develops at the site of wounding.
Martins-Green et al report that these wound tumors do not develop as a result of metastasis. Rather, factors released upon wounding contribute to the development of the wound tumors. In particular, they showed that TGF-beta, acidic FGF, and basic FGF, growth factors implicated in wound healing, can themselves produce tumor development in this system. Other growth factors (EGF and TGF-alpha), that also have roles in wound healing, do not induce tumors. They reported that wound tumor development correlates with inflammation and during the inflammatory response, blood vessel leakage occurs as tested by the release of fibrinogen into the tissues. The three factors (TGF-beta, acidic FGF, basic FGF) which promote tumors also induce inflammation, whereas the two factors (EGF and TGF-alpha) do not promote tumors or induce inflammation.
To test the possibility that inflammation is the key element in the development of these wound tumors, they used beta-methylprednisolone, a drug that inhibits inflammation (including blood vessel leakage), to determine if wound tumor development could be prevented. Martins-Green et al found that when inflammation was inhibited, tumors were also inhibited; when inflammation could not be stopped, tumors developed as before. These results indicate that the effect of wounding on the development of wound tumors in Rous sarcoma virus-infected chicks is accomplished through the cytokines released by the inflammatory cells at the site of wounding. Thus, at least in this avian virus system, inflammation plays a critical role in providing the conducive environment for oncogene integration and activation, and subsequent development of tumors.
Exploring a relation of the Martins-Green et al experiment to the El Saghir et al case, oncogene activation is not relevant but it is certainly easy to imagine that the head trauma caused inflammation. The authors do not state what type of car the patient was in or what precisely was the object of impact so we cannot know the exact surface area of trauma. However, since the tumor was on a relatively flat part of the patient's head and cars do not have sharp areas (in order to prevent puncture injuries), we can assume the area of impact was at least several square cm. It would follow that the area of inflammation was likewise at least several square cm. There would be ample cancer cells in the blood to populate the entire area of inflammation via leakage since the patient obviously had significant existing tumor burden. That would explain the rapid growth since the tumor would grow from extensive seeding over a few square cm surface together with the availability of growth factors.
Supporting this notion, Balkwill et al writes that if genetic damage is the “match that lights the fire” of cancer, then inflammation is the “fuel that feeds the flames” [7]. Also that inflammation may play a role in survival and proliferation of already initiated cancer cells [8].
We should also consider a few alternative hypotheses. For one, it may be that we are in error in estimating the neoplastic component of the tumor and that growth of a dormant micrometastasis might account for the sudden tumor. This can be rejected since it is very doubtful that even a factor of two error would make much difference. For a second alternative hypothesis, skull bones and associated marrow have sufficient vasculature to support many dormant micrometastases that perhaps could have been collectively activated into sudden angiogenesis with the trauma. This hypothesis would require a coincidence that the impact happened to occur where there were dormant micrometastases or the entire skull was laced with dormant micrometastases – both of which are unlikely.
There have been anecdotal reports for over a century of tumor growth after surgery or wounding. Angiogenic processes induced after surgery could reasonably explain those reports. The report by El Saghir et al is unusual in that there is no open wound to repair and the growth is too rapid to be explained by activation of angiogenesis of a dormant micrometastasis. Their suggestion that circulating cancer cells were entrapped at the trauma site is probably close to the truth.
I agree that this case calls for research to understand the mechanism of this type of cancer inducement. The unusual isolated and exaggerated situation allowed El Saghir et al to observe what may be a new and possibly important hematologic metastatic pathway: inflammation as a facilitating precursor to tumor. Metastasis is a very inefficient process [9]. Millions of cancer cells might be found in a patient’s blood but only a few metastases occur. The inflammation sequence discussed by Martins-Green et al would certainly increase metastatic efficiency since it bypasses extravasation through an intact vessel wall and it provides growth factors in the microenvironment [10]. In the context of Paget’s 1889 “seed and soil” metaphor, we have a situation here where many viable seeds are planted in a large vacant field and good fertilizer is applied.
The possible connection between inflammation and tumor growth might also help explain why anti-inflammatory drugs such as aspirin, prednisone and Celebrex are often used in cancer prevention and therapy.
My thanks to Tatura Udagawa for reminding me of the Martins-Green et al experiment and Romano Demicheli for useful discussions. Thanks also to El Saghir et al and the patient's widow for presenting this valuable information. This effect is not amenable to analysis from clinical trial data so anecdotal reports of unusual events are essential and encouraged.
References:
1. Nagi S El Saghir, Ihab I Elhajj, Fady B Geara, Mukbil H Hourani. Trauma-associated growth of suspected dormant micrometastasis. BMC Cancer 2005, 5:94 doi:10.1186/1471-2407-5-94.
2. Beadsmoore CJ, Screaton NJ. Classification, staging and prognosis of lung cancer, Eur J Radiol. 2003 Jan;45(1):8-17.
3. M Retsky, G Bonadonna, R Demicheli, J Folkman, W Hrushesky, P Valagussa. Hypothesis: Induced angiogenesis after surgery in premenopausal node-positive breast cancer patients is a major underlying reason why adjuvant chemotherapy works particularly well for those patients. Breast-Cancer-Research, 2004, 6:R372-R374 (14 May 2004).
4. R Demicheli, G Bonadonna, WJM Hrushesky, MW Retsky. P Valagussa. Menopausal status dependence of the timing of breast cancer recurrence following primary tumour surgical removal. Breast Cancer Research2004, 6:R689-R696 . http://breast-cancer-research.com/content/6/6/R689
5. El Sharouni SY, Kal HB, Battermann JJ. Accelerated regrowth of non-small-cell lung tumours after induction chemotherapy. Br J Cancer. 2003 Dec 15;89(12):2184-9.
6. Martins-Green M, Boudreau N, Bissell MJ. Inflammation is responsible for the development of wound-induced tumors in chickens infected with Rous sarcoma virus. Cancer Res. 1994 Aug 15;54(16):4334-41.
7. Balkwill F, Mantovani A, INFLAMMATION AND CANCER: BACK TO VIRCHOW? Lancet, 0099-5355, February 17, 2001, Vol. 357, Issue 9255
What would explain the sudden growth of lung cancer after minor trauma reported by El Saghir et al?
10 August 2005
This is a case report in which a 43 year old male with a smoking history was diagnosed with inoperable stage IIIA or IIIB poorly differentiated non-small cell lung cancer [1]. The prognosis for this stage is very poor with 5-year survival of approximately 10% [2]. Treatment was chemotherapy and radiation. While relapse was inevitable, the patient was seemingly doing well 15 months after primary treatment. The patient then incurred a minor trauma to the right temporal skull bone; within a month, the patient reported a rapidly growing 7 cm tumor at the place of the trauma. Imaging revealed tumor had penetrated the skull with meningeal invasion, and while some compression was apparent, it did not yet invade the brain. (Lung cancer most often relapses to brain or adrenal glands but can recur virtually anywhere including the skeleton.) The patient died 15 days later of massive hemoptysis.
El Saghir et al suggest that it is not a coincidence that both the growth occurred at the location of the trauma and that the growth started shortly after the trauma. The mechanisms suggested include some type of trauma-induced angiogenesis of a dormant micrometastasis or dormant malignant cells that happened to be at that place, or perhaps, there were no cancer cells at the cite, but circulating cancer cells were entrapped there due to the trauma.
My colleagues and I have studied tumor growth in breast cancer [3,4]. We have reported that there is an early relapse mode 6-9 months after mastectomy for approximately 20% of premenopausal women with node-positive breast cancer. We proposed this is the result of surgery-induced angiogenesis of dormant distant avascular micrometastases. This may explain why adjuvant chemotherapy is particularly effective for that patient category and also why there is no early mortality benefit of mammography screening for women age 40-49. Since survival after relapse is 2 years, surgery-induced angiogenesis could quantitatively explain a possible excess mortality for the screened group 3 years after the start of screening trials for young women. It also suggests why breast cancer in young women is often termed “aggressive”. Can we explain the El Saghir et al case report of sudden growth in lung cancer after minor trauma with a similar mechanism as suggested by the authors?
We can analyze these data and draw some conclusions about what is possible and what is not. Since there were no biopsies taken of the growth, we do not know with certainty that it is metastatic lung cancer - although there is no reason to doubt that. Assuming a 1 mm dormant micrometastasis grew to 7 cm in one month exponentially, the doubling time would be a constant 2 days. As El Saghir et al report, this is indeed rapid growth for human cancer. That is approximately the growth rate of a multipassaged animal model tumor. Human cancer data are typically 2 to 5 month doubling time plus occasional dormancy but there are reports of growth with doubling time as short as 8 days after intervention [5]. It is difficult to accept that the tumor was the result of a 1 mm dormant micrometastasis suddenly undergoing angiogenesis and growing to 7 cm in 1 month. We can also rule out the possibility of a single dormant cell becoming a 7 cm diameter tumor in one month since doubling time would be less than one day - which is impossibly small.
What can we conclude from this information? The key aspect of this case is the fact that tumor grew at the site of trauma and started at that time; no other tumors grew at any other place or started at any other time. Also it is unlikely that the new growth is the result of sudden activation of cancer from some known dormant state previously at the trauma site. We need another explanation.
This is somewhat similar to the experiments of Martins-Green et al., whose studies may provide some insight [6]. This is a particular avian system in which a virus is the carcinogenic agent. Other than those two major differences, as we will see, there is a certain similarity to the case presented by El Saghir et al. When newly hatched chicks are given injections of Rous sarcoma virus, a tumor develops at the site of injection. In spite of the presence of the virus in the blood, no other tumors are found distant from the site of inoculation during the life span of the animal (4-6 weeks). However, if a wound is made away from the primary tumor, a tumor develops at the site of wounding.
Martins-Green et al report that these wound tumors do not develop as a result of metastasis. Rather, factors released upon wounding contribute to the development of the wound tumors. In particular, they showed that TGF¿, acidic FGF, and basic FGF, growth factors implicated in wound healing, can themselves produce tumor development in this system. Other growth factors (EGF and TGF¿), that also have roles in wound healing, do not induce tumors. They reported that wound tumor development correlates with inflammation and during the inflammatory response, blood vessel leakage occurs as tested by the release of fibrinogen into the tissues. The three factors (TGF¿, aFGF, bFGF) which promote tumors also induce inflammation, whereas the two factors (EGF and TGF¿) do not promote tumors or induce inflammation.
To test the possibility that inflammation is the key element in the development of these wound tumors, they used beta-methylprednisolone, a drug that inhibits inflammation (including blood vessel leakage), to determine if wound tumor development could be prevented. Martins-Green et al found that when inflammation was inhibited, tumors were also inhibited; when inflammation could not be stopped, tumors developed as before. These results indicate that the effect of wounding on the development of wound tumors in Rous sarcoma virus-infected chicks is accomplished through the cytokines released by the inflammatory cells at the site of wounding. Thus, at least in this avian virus system, inflammation plays a critical role in providing the conducive environment for oncogene integration and activation, and subsequent development of tumors.
Exploring a relation of the Martins-Green et al experiment to the El Saghir et al case, oncogene activation is not relevant but it is certainly easy to imagine that the head trauma caused inflammation. The authors do not state what type of car the patient was in or what precisely was the object of impact so we cannot know the exact surface area of trauma. However, since the tumor was on a relatively flat part of the patient's head and cars do not have sharp areas (in order to prevent puncture injuries), we can assume the area of impact was at least several square cm. It would follow that the area of inflammation was likewise at least several square cm. There would be ample cancer cells in the blood to populate the entire area of inflammation via leakage since the patient obviously had significant existing tumor burden. That would explain the rapid growth since the tumor would grow from extensive seeding over a few cm2 surface together with the availability of growth factors.
Supporting this notion, Balkwill et al writes that if genetic damage is the “match that lights the fire” of cancer, then inflammation is the “fuel that feeds the flames” [7]. Also that inflammation may play a role in survival and proliferation of already initiated cancer cells [8].
We should also consider a few alternative hypotheses. For one, it may be that we are in error in estimating the neoplastic component of the tumor and that growth of a dormant micrometastasis might account for the sudden tumor. This can be rejected since it is very doubtful that even a factor of two error would make much difference. For a second alternative hypothesis, skull bones and associated marrow have sufficient vasculature to support many dormant micrometastases that perhaps could have been collectively activated into sudden angiogenesis with the trauma. This hypothesis would require a coincidence that the impact happened to occur where there were dormant micrometastases or the entire skull was laced with dormant micrometastases – both of which are unlikely.
There have been anecdotal reports for over a century of tumor growth after surgery or wounding. Angiogenic processes induced after surgery could reasonably explain those reports. The report by El Saghir et al is unusual in that there is no wound to repair and the growth is too rapid to be explained by activation of angiogenesis of a dormant micrometastasis. Their suggestion that circulating cancer cells were entrapped at the trauma site is probably close to the truth.
I agree that this case calls for research to understand the mechanism of this type of cancer inducement. The unusual isolated and exaggerated situation allowed El Saghir et al to observe what may be a new and possibly important hematologic metastatic pathway: inflammation as a facilitating precursor to tumor. Metastasis is a very inefficient process [9]. Millions of cancer cells might be found in a patient’s blood but only a few metastases occur. The inflammation sequence discussed by Martins-Green et al would certainly increase metastatic efficiency since it bypasses extravasation through an intact vessel wall and it provides growth factors in the microenvironment [10]. In the context of Paget’s 1889 “seed and soil” metaphor, we have a situation here where many viable seeds are planted in a large vacant field and good fertilizer is applied.
The possible connection between inflammation and tumor growth might also help explain why anti-inflammatory drugs such as aspirin, prednisone and Celebrex are often used in cancer prevention and therapy.
My thanks to Tatura Udagawa for reminding me of the Martins-Green et al experiment and Romano Demicheli for useful discussions. Thanks also to El Saghir et al and the patient's widow for presenting this valuable information. This effect is not amenable to analysis from clinical trial data so anecdotal reports of unusual events are essential and encouraged.
References:
1. Nagi S El Saghir, Ihab I Elhajj, Fady B Geara, Mukbil H Hourani. Trauma-associated growth of suspected dormant micrometastasis. BMC Cancer 2005, 5:94 doi:10.1186/1471-2407-5-94.
2. Beadsmoore CJ, Screaton NJ. Classification, staging and prognosis of lung cancer, Eur J Radiol. 2003 Jan;45(1):8-17.
3. M Retsky, G Bonadonna, R Demicheli, J Folkman, W Hrushesky, P Valagussa. Hypothesis: Induced angiogenesis after surgery in premenopausal node-positive breast cancer patients is a major underlying reason why adjuvant chemotherapy works particularly well for those patients. Breast-Cancer-Research, 2004, 6:R372-R374 (14 May 2004).
http://breast-cancer-research.com/content/6/4/R372
4. R Demicheli, G Bonadonna, WJM Hrushesky, MW Retsky. P Valagussa. Menopausal status dependence of the timing of breast cancer recurrence following primary tumour surgical removal. Breast Cancer Research2004, 6:R689-R696 . http://breast-cancer-research.com/content/6/6/R689
5. El Sharouni SY, Kal HB, Battermann JJ. Accelerated regrowth of non-small-cell lung tumours after induction chemotherapy. Br J Cancer. 2003 Dec 15;89(12):2184-9.
6. Martins-Green M, Boudreau N, Bissell MJ. Inflammation is responsible for the development of wound-induced tumors in chickens infected with Rous sarcoma virus. Cancer Res. 1994 Aug 15;54(16):4334-41.
7. Balkwill F, Mantovani A, INFLAMMATION AND CANCER: BACK TO VIRCHOW? Lancet, 0099-5355, February 17, 2001, Vol. 357, Issue 9255
8. Balkwill F, Coussens LM. Cancer: an inflammatory link. Nature. 2004 Sep 23;431(7007):405-6.
9. Chambers AF, Groom AC, MacDonald IC. Dissemination and growth of cancer cells in metastatic sites. Nat Rev Cancer. 2002 Aug;2(8):563-72.
10. Fidler IJ. The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited. Nat Rev Cancer. 2003 Jun;3(6):453-8.
Competing interests
None
What would explain the sudden growth of lung cancer after minor trauma reported by El Saghir et al?
17 August 2005
This is a case report in which a 43 year old male with a smoking history was diagnosed with inoperable stage IIIA or IIIB poorly differentiated non-small cell lung cancer [1]. The prognosis for this stage is very poor with 5-year survival of approximately 10% [2]. Treatment was chemotherapy and radiation. While relapse was inevitable, the patient was seemingly doing well 15 months after primary treatment. The patient then incurred a minor trauma to the right temporal skull bone; within a month, the patient reported a rapidly growing 7 cm tumor at the place of the trauma. Imaging revealed tumor had penetrated the skull with meningeal invasion, and while some compression was apparent, it did not yet invade the brain. (Lung cancer most often relapses to brain or adrenal glands but can recur virtually anywhere including the skeleton.) The patient died 15 days later of massive hemoptysis.
El Saghir et al suggest that it is not a coincidence that both the growth occurred at the location of the trauma and that the growth started shortly after the trauma. The mechanisms suggested include some type of trauma-induced angiogenesis of a dormant micrometastasis or dormant malignant cells that happened to be at that place, or perhaps, there were no cancer cells at the cite, but circulating cancer cells were entrapped there due to the trauma.
My colleagues and I have studied tumor growth in breast cancer [3,4]. We have reported that there is an early relapse mode 6-9 months after mastectomy for approximately 20% of premenopausal women with node-positive breast cancer. We proposed this is the result of surgery-induced angiogenesis of dormant distant avascular micrometastases. This may explain why adjuvant chemotherapy is particularly effective for that patient category and also why there is no early mortality benefit of mammography screening for women age 40-49. Since survival after relapse is 2 years, surgery-induced angiogenesis could quantitatively explain a possible excess mortality for the screened group 3 years after the start of screening trials for young women. It also suggests why breast cancer in young women is often termed “aggressive”. Can we explain the El Saghir et al case report of sudden growth in lung cancer after minor trauma with a similar mechanism as suggested by the authors?
We can analyze these data and draw some conclusions about what is possible and what is not. Since there were no biopsies taken of the growth, we do not know with certainty that it is metastatic lung cancer - although there is no reason to doubt that. Assuming a 1 mm dormant micrometastasis grew to 7 cm in one month exponentially, the doubling time would be a constant 2 days. As El Saghir et al report, this is indeed rapid growth for human cancer. That is approximately the growth rate of a multipassaged animal model tumor. Human cancer data are typically 2 to 5 month doubling time plus occasional dormancy but there are reports of growth with doubling time as short as 8 days after intervention [5]. It is difficult to accept that the tumor was the result of a 1 mm dormant micrometastasis suddenly undergoing angiogenesis and growing to 7 cm in 1 month. We can also rule out the possibility of a single dormant cell becoming a 7 cm diameter tumor in one month since doubling time would be less than one day - which is impossibly small.
What can we conclude from this information? The key aspect of this case is the fact that tumor grew at the site of trauma and started at that time; no other tumors grew at any other place or started at any other time. Also it is unlikely that the new growth is the result of sudden activation of cancer from some known dormant state previously at the trauma site. We need another explanation.
This is somewhat similar to the experiments of Martins-Green et al., whose studies may provide some insight [6]. This is a particular avian system in which a virus is the carcinogenic agent. Other than those two major differences, as we will see, there is a certain similarity to the case presented by El Saghir et al. When newly hatched chicks are given injections of Rous sarcoma virus, a tumor develops at the site of injection. In spite of the presence of the virus in the blood, no other tumors are found distant from the site of inoculation during the life span of the animal (4-6 weeks). However, if a wound is made away from the primary tumor, a tumor develops at the site of wounding.
Martins-Green et al report that these wound tumors do not develop as a result of metastasis. Rather, factors released upon wounding contribute to the development of the wound tumors. In particular, they showed that TGF-beta, acidic FGF, and basic FGF, growth factors implicated in wound healing, can themselves produce tumor development in this system. Other growth factors (EGF and TGF-alpha), that also have roles in wound healing, do not induce tumors. They reported that wound tumor development correlates with inflammation and during the inflammatory response, blood vessel leakage occurs as tested by the release of fibrinogen into the tissues. The three factors (TGF-beta, acidic FGF, basic FGF) which promote tumors also induce inflammation, whereas the two factors (EGF and TGF-alpha) do not promote tumors or induce inflammation.
To test the possibility that inflammation is the key element in the development of these wound tumors, they used beta-methylprednisolone, a drug that inhibits inflammation (including blood vessel leakage), to determine if wound tumor development could be prevented. Martins-Green et al found that when inflammation was inhibited, tumors were also inhibited; when inflammation could not be stopped, tumors developed as before. These results indicate that the effect of wounding on the development of wound tumors in Rous sarcoma virus-infected chicks is accomplished through the cytokines released by the inflammatory cells at the site of wounding. Thus, at least in this avian virus system, inflammation plays a critical role in providing the conducive environment for oncogene integration and activation, and subsequent development of tumors.
Exploring a relation of the Martins-Green et al experiment to the El Saghir et al case, oncogene activation is not relevant but it is certainly easy to imagine that the head trauma caused inflammation. The authors do not state what type of car the patient was in or what precisely was the object of impact so we cannot know the exact surface area of trauma. However, since the tumor was on a relatively flat part of the patient's head and cars do not have sharp areas (in order to prevent puncture injuries), we can assume the area of impact was at least several square cm. It would follow that the area of inflammation was likewise at least several square cm. There would be ample cancer cells in the blood to populate the entire area of inflammation via leakage since the patient obviously had significant existing tumor burden. That would explain the rapid growth since the tumor would grow from extensive seeding over a few square cm surface together with the availability of growth factors.
Supporting this notion, Balkwill et al writes that if genetic damage is the “match that lights the fire” of cancer, then inflammation is the “fuel that feeds the flames” [7]. Also that inflammation may play a role in survival and proliferation of already initiated cancer cells [8].
We should also consider a few alternative hypotheses. For one, it may be that we are in error in estimating the neoplastic component of the tumor and that growth of a dormant micrometastasis might account for the sudden tumor. This can be rejected since it is very doubtful that even a factor of two error would make much difference. For a second alternative hypothesis, skull bones and associated marrow have sufficient vasculature to support many dormant micrometastases that perhaps could have been collectively activated into sudden angiogenesis with the trauma. This hypothesis would require a coincidence that the impact happened to occur where there were dormant micrometastases or the entire skull was laced with dormant micrometastases – both of which are unlikely.
There have been anecdotal reports for over a century of tumor growth after surgery or wounding. Angiogenic processes induced after surgery could reasonably explain those reports. The report by El Saghir et al is unusual in that there is no open wound to repair and the growth is too rapid to be explained by activation of angiogenesis of a dormant micrometastasis. Their suggestion that circulating cancer cells were entrapped at the trauma site is probably close to the truth.
I agree that this case calls for research to understand the mechanism of this type of cancer inducement. The unusual isolated and exaggerated situation allowed El Saghir et al to observe what may be a new and possibly important hematologic metastatic pathway: inflammation as a facilitating precursor to tumor. Metastasis is a very inefficient process [9]. Millions of cancer cells might be found in a patient’s blood but only a few metastases occur. The inflammation sequence discussed by Martins-Green et al would certainly increase metastatic efficiency since it bypasses extravasation through an intact vessel wall and it provides growth factors in the microenvironment [10]. In the context of Paget’s 1889 “seed and soil” metaphor, we have a situation here where many viable seeds are planted in a large vacant field and good fertilizer is applied.
The possible connection between inflammation and tumor growth might also help explain why anti-inflammatory drugs such as aspirin, prednisone and Celebrex are often used in cancer prevention and therapy.
My thanks to Tatura Udagawa for reminding me of the Martins-Green et al experiment and Romano Demicheli for useful discussions. Thanks also to El Saghir et al and the patient's widow for presenting this valuable information. This effect is not amenable to analysis from clinical trial data so anecdotal reports of unusual events are essential and encouraged.
References:
1. Nagi S El Saghir, Ihab I Elhajj, Fady B Geara, Mukbil H Hourani. Trauma-associated growth of suspected dormant micrometastasis. BMC Cancer 2005, 5:94 doi:10.1186/1471-2407-5-94.
2. Beadsmoore CJ, Screaton NJ. Classification, staging and prognosis of lung cancer, Eur J Radiol. 2003 Jan;45(1):8-17.
3. M Retsky, G Bonadonna, R Demicheli, J Folkman, W Hrushesky, P Valagussa. Hypothesis: Induced angiogenesis after surgery in premenopausal node-positive breast cancer patients is a major underlying reason why adjuvant chemotherapy works particularly well for those patients. Breast-Cancer-Research, 2004, 6:R372-R374 (14 May 2004).
http://breast-cancer-research.com/content/6/4/R372
4. R Demicheli, G Bonadonna, WJM Hrushesky, MW Retsky. P Valagussa. Menopausal status dependence of the timing of breast cancer recurrence following primary tumour surgical removal. Breast Cancer Research2004, 6:R689-R696 . http://breast-cancer-research.com/content/6/6/R689
5. El Sharouni SY, Kal HB, Battermann JJ. Accelerated regrowth of non-small-cell lung tumours after induction chemotherapy. Br J Cancer. 2003 Dec 15;89(12):2184-9.
6. Martins-Green M, Boudreau N, Bissell MJ. Inflammation is responsible for the development of wound-induced tumors in chickens infected with Rous sarcoma virus. Cancer Res. 1994 Aug 15;54(16):4334-41.
7. Balkwill F, Mantovani A, INFLAMMATION AND CANCER: BACK TO VIRCHOW? Lancet, 0099-5355, February 17, 2001, Vol. 357, Issue 9255
8. Balkwill F, Coussens LM. Cancer: an inflammatory link. Nature. 2004 Sep 23;431(7007):405-6.
9. Chambers AF, Groom AC, MacDonald IC. Dissemination and growth of cancer cells in metastatic sites. Nat Rev Cancer. 2002 Aug;2(8):563-72.
10. Fidler IJ. The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited. Nat Rev Cancer. 2003 Jun;3(6):453-8.
Competing interests
None